**Preface to "Bacterial Meningitis: Epidemiology and Vaccination"**

The suddenness, severity and dire consequences of meningitis remain a challenge for all countries of the world. Although medical countermeasures to date, such as vaccines, diagnostics and therapies, are generalized to prevent, rapidly identify and treat acute bacterial meningitis, these still have major limitations. At the same time, support for people and their families living with long-term disability after meningitis is largely insufficient or non-existent. In addition, the occurrence of meningitis epidemics, which are difficult to predict, continues to pose a threat to communities in several countries. The unprecedented progress made in the fight against meningitis in recent decades has brought hope that the disease could be defeated. It is in this context that a call to action has gradually materialized and given rise to a collective and collaborative effort by many stakeholders to develop, under WHO leadership, the global roadmap to defeat meningitis by 2030. This roadmap, at the heart of the current strategy of the World Health Organization, is an essential element in achieving universal health coverage. It was approved by the 73rd World Health Assembly in November 2020, when Member States endorsed the first ever resolution on meningitis prevention and control. Collaboration among stakeholders from different fields and perspectives has enabled the development of an ambitious but achievable global roadmap, which is a powerful means, integrated with other initiatives, to advance primary health care, protection against health emergencies and enable more people to enjoy better health and well-being.

In the same spirit of a complete, global and multidisciplinary approach, the editor of this book, James Stuart, a long-standing expert in meningitis who has been closely involved in the development of the road map, has brought together a wide range of different experts in the field of bacterial meningitis. The articles included in this book cover various important and complementary aspects of the strategic goals of the global roadmap and, as such, they constitute valuable and timely documentation on the eve of the launch of this roadmap for 2030. I am very pleased to see publication of this book on bacterial meningitis and grateful to the editor and contributors to this broad range of substantive papers, that together recognize the importance of the roadmap and support our drive to defeat meningitis across the world and improve the care of those affected by this devastating infection.

#### **Dr Marie-Pierre Preziosi**

Meningitis lead, Division for Universal Health Coverage / Life Course, WHO, Geneva, Switzerland

### *Editorial* **Editorial for the Special Issue: Bacterial Meningitis—Epidemiology and Vaccination**

**James M. Stuart**

Population Health Sciences, Bristol Medical School, University of Bristol, Bristol BS8 1QU, UK; james.stuart@bristol.ac.uk

Bacterial meningitis has serious health, economic, and social consequences with a high risk of death and lifelong disability. WHO has published the first global road map on meningitis "Defeating meningitis by 2030" to tackle the main causes of acute bacterial meningitis: *Neisseria meningitidis, Streptococcus pneumoniae, Haemophilus influenzae*, and *Streptococcus agalactiae* (Group B *Streptococcus* (GBS)) [1,2]. The road map was endorsed by the World Health Assembly in November 2020 [3].

The three main goals of the meningitis roadmap are to eliminate epidemics of bacterial meningitis, reduce cases and deaths from vaccine-preventable bacterial meningitis, and reduce disability and improve quality of life after meningitis of any cause. Proposed measures to achieve these goals include development of new vaccines, increased effectiveness of prevention and control strategies, efficient global surveillance with accurate estimates of disease burden including sequelae, and global availability of and access to rapid diagnosis and high-quality treatment of meningitis and its after-effects.

This Special Issue includes a wide range of original research articles and review articles on epidemiology and vaccination of bacterial meningitis that have direct relevance to advancing the goals of the road map.

A fundamental step in establishing the importance of meningitis and in monitoring progress toward prevention and care is quantifying the burden from illness, death, and disability. Wright et al. [4] described wide variation in different modelled estimates of the global burden and advocated for alignment with improving surveillance data to improve the accuracy of model parameters. The consequences of meningitis are even harder to measure. Schiess et al. [5] underlined the social and economic costs of meningitis, the lack of recognition of more subtle sequelae, and the lack of knowledge on long-term effects, especially in low- and middle-income countries. Building care services for those affected by meningitis across the world will be a challenging objective for the meningitis strategy.

The principal means of achieving targets to reduce cases and deaths from meningitis will inevitably be through vaccination. Alderson et al. [6] gave a comprehensive overview of past and present developments in meningitis vaccines. They drew attention to the importance of low-cost vaccines for global introduction, the expanding range of conjugate vaccines and the more recently developed meningococcal protein vaccines, and the challenges in reaching prevention goals. As vaccines are developed and vaccination programmes expanded, Deghmane and Taha [7] made the case that preventing disease among those at higher risk will become increasingly important.

For meningococcal meningitis, the high-burden region of the meningitis belt in sub-Saharan Africa deserves particular attention. Karachaliou Prasinou et al. [8] showed how mathematical models can be used to optimise the effectiveness of vaccination programmes, with two key parameters being the duration of protection and age at vaccination. Such models are relevant both for the serogroup A vaccine currently being deployed in the meningitis belt and for the anticipated roll out of pentavalent (ACWXY) conjugate vaccines [6]. The need for broader-valency vaccines in the global control of invasive meningococcal disease was well demonstrated in the paper by Tzeng and Stephens [9], describing the changing epidemiology and emerging disease due to serogroups other than A, B, and C.

**Citation:** Stuart, J.M. Editorial for the Special Issue: Bacterial Meningitis—Epidemiology and Vaccination. *Microorganisms* **2021**, *9*, 917. https://doi.org/10.3390/ microorganisms9050917

Received: 22 April 2021 Accepted: 23 April 2021 Published: 24 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Slack [10] documented how meningitis due to *Haemophilus influenzae* fell rapidly with the introduction of conjugate Hib vaccines from the 1990s such that, by 2015, the burden of meningitis due to serotype b was limited to a few countries that had not introduced these vaccines into the national immunization programmes. However, as she pointed out, invasive disease due to non-typable organisms and other serotypes is of increasing concern. Serotype A has emerged as a significant cause of meningitis in indigenous populations of North America and has stimulated the development of a new conjugate vaccine.

A series of papers from the impressive PSERENADE project [11–13] demonstrated the substantial global impact of multivalent pneumococcal conjugate vaccines on invasive pneumococcal disease, including meningitis, after their introduction into infant immunisation programmes. Six or more years after introduction, they found a 95% reduction in serotype 1 disease in all age groups. Measuring the impact does depend on robust serotype surveillance systems, and they acknowledged the need for more data from the meningitis belt countries that are at high risk of pneumococcal meningitis and serotype 1 outbreaks.

Vaccines are in development but not yet available to protect against disease due to GBS [6]. Prevention of GBS in newborns currently relies on risk-based or microbiological screening for infection in pregnancy. However, a study of meningitis among infants under 90 days of age in a large paediatric hospital in the USA [14] showed that the majority of cases of bacterial meningitis were due to GBS, despite universal screening and intrapartum prophylaxis. This only emphasises the importance of a vaccine that could hopefully have more impact than prepartum screening with the additional protection of stillbirths due to GBS infection and late-onset GBS disease.

Tsang [15] focused on the molecular epidemiology of the four main bacterial causes of meningitis in the roadmap and the power of conjugate vaccines to both reduce the burden and drive the evolution of these bacteria, thus underlining the need for improved surveillance and expansion of whole-genome sequencing.

Zainel et al. [16] highlighted neurological complications from bacterial meningitis in children such as hearing loss, cognitive impairment, and epilepsy, as well as the importance of prompt effective treatment regimens in improving outcomes. A key component of prompt treatment is rapid accurate diagnosis of meningitis through bedside tests that can be applicable in low- and middle-income countries. Rondy et al. [17] reported on a field evaluation of a rapid test that should aid timely decisions on vaccine deployment in meningitis epidemics.

Meningitis can be caused by many infectious organisms: bacteria, viruses, fungi, and parasites. The focus in the "Defeating meningitis by 2030" strategy is on the main bacteria responsible for the overall global burden with potential for prevention by vaccination. Another major cause of bacterial meningitis, *Mycobacterium tuberculosis*, was given prominence in this issue by Basu-Roy et al. [18]. Their review highlighted how the "Defeating meningitis" roadmap can be applied to the prevention and control of tuberculosis in children, affirming the need for a collaborative endeavour and linking with activities of other initiatives such as the WHO TB roadmap [19]. The fact that many elements of the roadmap apply to TB and all other causes of meningitis must not be forgotten in the drive to defeat meningitis.

The global roadmap to defeat meningitis is an ambitious strategy, particularly in the context of the Covid pandemic. As shown by the contributions to this Special Issue, a concerted drive to reduce the burden of this illness is, without question, a worthy ambition. The theme of World Meningitis Day 2021 is "Take Action #DefeatMeningitis" [20,21]. Start now!

**Funding:** This work received no external funding.

**Acknowledgments:** I am very grateful for the contributions in time and effort from all the authors, and I am particularly thankful for the support and encouragement from Marie-Pierre Preziosi at WHO and Brian Greenwood at the London School of Hygiene and Tropical Medicine.

**Conflicts of Interest:** The author was employed as a consultant by WHO to help develop the meningitis roadmap.

#### **References**


### *Article* **The Global Burden of Meningitis in Children: Challenges with Interpreting Global Health Estimates**

**Claire Wright 1,\*, Natacha Blake <sup>1</sup> , Linda Glennie <sup>1</sup> , Vinny Smith <sup>1</sup> , Rose Bender <sup>2</sup> , Hmwe Kyu <sup>2</sup> , Han Yong Wunrow <sup>2</sup> , Li Liu <sup>3</sup> , Diana Yeung <sup>4</sup> , Maria Deloria Knoll <sup>5</sup> , Brian Wahl <sup>5</sup> , James M. Stuart 6,7 and Caroline Trotter <sup>8</sup>**


**Abstract:** The World Health Organization (WHO) has developed a global roadmap to defeat meningitis by 2030. To advocate for and track progress of the roadmap, the burden of meningitis as a syndrome and by pathogen must be accurately defined. Three major global health models estimating meningitis mortality as a syndrome and/or by causative pathogen were identified and compared for the baseline year 2015. Two models, (1) the WHO and the Johns Hopkins Bloomberg School of Public Health's Maternal and Child Epidemiology Estimation (MCEE) group's Child Mortality Estimation (WHO-MCEE) and (2) the Institute for Health Metrics and Evaluation (IHME) Global Burden of Disease Study (GBD 2017), identified meningitis, encephalitis and neonatal sepsis, collectively, to be the second and third largest infectious killers of children under five years, respectively. Global meningitis/encephalitis and neonatal sepsis mortality estimates differed more substantially between models than mortality estimates for selected infectious causes of death and all causes of death combined. Estimates at national level and by pathogen also differed markedly between models. Aligning modelled estimates with additional data sources, such as national or sentinel surveillance, could more accurately define the global burden of meningitis and help track progress against the WHO roadmap.

**Keywords:** meningitis; child mortality; neonatal sepsis; global health; global health estimates; modelling; *Streptococcus pneumoniae*; *Haemophilus influenzae*; *Neisseria meningitidis*

### **1. Introduction**

The world saw great progress in reducing child mortality over the lifetime of the United Nations (UN) Millennium Development Goals (MDGs) with an estimated 54% decline in children under five years of age from 93 deaths per 1000 live births in 1990 to 43 per 1000 live births in 2015 [1]. The successor UN Sustainable Development Goals (SDGs) are more ambitious again, and urge that by 2030 we should "end preventable deaths of newborns and children under five years of age, with all countries aiming to reduce neonatal

**Citation:** Wright, C.; Blake, N.; Glennie, L.; Smith, V.; Bender, R.; Kyu, H.; Wunrow, H.Y.; Liu, L.; Yeung, D.; Knoll, M.D.; et al. The Global Burden of Meningitis in Children: Challenges with Interpreting Global Health Estimates. *Microorganisms* **2021**, *9*, 377. https://doi.org/10.3390/ microorganisms9020377

Academic Editor: Glenn S. Tillotson Received: 11 January 2021 Accepted: 4 February 2021 Published: 13 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

mortality to at least as low as 12 per 1000 live births and under-five mortality to at least as low as 25 per 1000 live births." However, with the majority of an estimated 38 deaths per 1000 live births in 2019 being caused by preventable and treatable diseases [1], we are a long way from achieving this target.

Among these preventable diseases, meningitis has one of the highest fatality rates and the potential to cause devastating epidemics. Since the turn of the century, we have seen advances as a result of widespread global introduction of *Haemophilus influenzae* type b (Hib) and pneumococcal vaccines as well as the roll out of the meningococcal A vaccine, MenAfriVac, across some of the highest incidence areas of sub-Saharan Africa. Despite this, recent estimates have identified that the global burden of meningitis in all age groups remains high and progress lags substantially behind that of other vaccine preventable diseases [2]. Whilst deaths from measles and tetanus in children under five years are estimated to have decreased by 86% and 92% respectively, between 1990 and 2017, over the same time period deaths from meningitis are estimated to have decreased by just 51% [3]. Despite its burden, meningitis is seldom, if at all, mentioned in key global and regional health documents [4–9].

In response to calls from governments, global health organisations, civil society, public health bodies, academia and the private sector, a World Health Organization (WHO) led collaboration is developing a Defeating Meningitis by 2030 Global Roadmap [10]. The Roadmap focuses on the four leading global causes of bacterial meningitis; *Neisseria meningitidis* (meningococcus), *Streptococcus pneumoniae* (pneumococcus), *Haemophilus influenzae* (Hi), and *Streptococcus agalactiae* (group B streptococcus (GBS)).

To advocate for a global roadmap to defeat meningitis, the global burden of meningitis as a syndrome in relation to other infectious causes of death needs to be accurately described, and countries with the highest burden identified, so that efforts and resources can be targeted effectively. Estimates of pathogen-specific meningitis incidence and mortality at the global level can identify the need for new vaccines or support wider access to existing ones. Tracking trends in pathogen-specific meningitis and syndromic disease over time at the national and international level is vital to assess the impact of interventions such as vaccines implemented as part of the global roadmap to defeat meningitis.

Vital registration systems and disease surveillance platforms are limited across many geographies and regions, so there is a reliance on modelled estimates to get a complete global picture of disease across all settings but cause of death estimates have been found to differ across these different modelling efforts [11]. Modelled estimates also attempt to account for changes in causes of death over time, but to do so accurately they must be informed by reliable data to make accurate predictions where real data is lacking.

In this paper we aim to compare the available modelled estimates for cases and deaths from meningitis as a syndrome, by causative pathogen and the methods used, in order to assess whether these models can be used with confidence by decision makers to prioritise recommendations from a plan to defeat meningitis, and by those needing to track progress on the WHO Defeating Meningitis by 2030 Global Roadmap'.

#### **2. Materials and Methods**

#### *2.1. Identification of Data Sources*

Through attending key stakeholder meetings, we identified three modelling efforts that estimate the global burden of meningitis and neonatal sepsis: (1) WHO and the Johns Hopkins Bloomberg School of Public Health's Maternal and Child Epidemiology Estimation (MCEE) group's Child Mortality Estimation (WHO-MCEE), which estimates15 causes of death for children under five years of age [12]; (2) The Institute for Health Metrics and Evaluation (IHME) Global Burden of Disease Study (GBD 2017) which estimates age specific mortality for 282 causes of death in all ages [3]; and (3) The WHO's Global Health Estimates (WHO GHE) which estimates age specific mortality for 136 causes of death in all ages [13].

Two additional models were also identified that estimated disease burden caused by pathogens of particular relevance to the WHO Defeating Meningitis by 2030 Global Roadmap: (1) the WHO-MCEE group's estimates of the burden of pneumococcal and Hib disease in children [14], and (2) the London School of Hygiene and Tropical Medicine (LSHTM) Burden of Group B Streptococcus Worldwide for Pregnant Women, Stillbirths, and Children [15].

Not all of these efforts were directly comparable because they did not provide the same level of data or use the same indicators of burden (Table 1).


**Table 1.** Models estimating the global burden of meningitis and neonatal sepsis.

DALYs = Disability Adjusted Life Years; GBS = Group B streptococcus; Hib = *Haemophilus influenzae* type b; Nm = *Neisseria meningitidis* (meningococcus); Spn = *Streptococcus pneumoniae* (pneumococcus).\* WHO GHE use a ratio of meningitis to encephalitis deaths obtained from IHME data to separate out MCEE under-five meningitis/encephalitis estimates. \*\* Estimates derived from UN World Population Prospects 2017. Differences between WHO GHE and WHO-MCEE population estimates likely due to draft estimates circulating prior to final publication. \*\*\* Derived from UN World Population Prospects 2015

> As WHO GHE estimates were an amalgamation of historical models (WHO-MCEE's 2000–2016 and IHME's GBD 2016) we did not consider them further in our analysis. We did not include GBS estimates from LSHTM in our analysis because the age categorisation

(0–89 days) did not correspond with the disaggregated age categories of the other models and so did not allow for meaningful comparison.

#### *2.2. Analysis of Data Sources*

The scale of the global burden of meningitis deaths relative to all causes and leading infectious causes of death was assessed by comparing, death and mortality estimates from GBD 2017 and the WHO-MCEE's 2000–2017 model according to the following syndromic cause of death categories "All causes", "Infectious disease", "meningitis/encephalitis" and "neonatal sepsis".

We considered the burden of meningitis and neonatal sepsis together for the purposes of comparison with other leading infectious causes of death because distinguishing between these syndromes is almost impossible based on clinical signs alone in the neonate [16,17]. Lumbar puncture (LP) and analysis of the cerebrospinal fluid is the only reliable way of confirming a case of meningitis. However, in many countries there is a shortage of trained staff to perform LP [18], and in low-income settings as few as 2% of neonates with infection might have an LP or blood sample taken [19].

The WHO-MCEE have historically estimated sepsis and meningitis in the neonatal period within the same cause category because of difficulties in distinguishing between these clinical syndromes in this age group. These causes were estimated separately for the first time in their latest modelling round by using the ratio of neonatal meningitis and neonatal sepsis deaths derived from IHME estimates. Because WHO-MCEE estimate meningitis/encephalitis as one cause category, GBD 2017 meningitis and encephalitis deaths were amalgamated for the purpose of comparison.

Denominators used to report mortality rates were standardised across the models and, where necessary, recalculated to be expressed as deaths per 1000 live births in the neonatal period and deaths per 100,000 population in the post neonatal period. GBD 2017 mortality rates in the neonatal period were calculated from IHME live birth estimates for the year 2015. WHO-MCEE postneonatal mortality rates were calculated using UN population estimates for the year 2015 [20].

Priority geographical areas for targeting a plan to defeat meningitis were identified from country-specific GBD 2017 and WHO-MCEE meningitis/encephalitis mortality estimates for the year 2015 in children under five years.

Meningitis mortality and incidence estimates according to pathogen over time (2000– 2015) were analysed using estimates produced by GBD 2017 and the WHO-MCEE pathogen model. Meningococcal meningitis is commonly associated with epidemics. As WHO-MCEE meningococcal meningitis estimates did not account for deaths and cases resulting from epidemics, estimates for 'Hib meningitis' and 'pneumococcal meningitis' mortality and incidence in the post neonatal period (28 days–<5 years) were the categories and age group used for comparison.

An analysis of the estimation methodology for each model was also undertaken in an attempt to explain any inconsistencies between models.

#### **3. Results**

*3.1. Global Meningitis and Neonatal Sepsis Mortality Estimates in Children Aged Under Five Years*

Overall, the WHO-MCEE estimated there to be approximately 100,000 fewer deaths in the under-five age group than the GBD 2017, with proportionally more under-five deaths occurring in the neonatal period (46% compared to GBD 2017's 43%).

The GBD 2017 estimated 34% more deaths from meningitis/encephalitis than the WHO-MCEE in the year 2015 (190,515 and 142,841, respectively) (Table 2). Meningitis made up the majority of the GBD 2017 combined meningitis/encephalitis category; 87% in under five-year-olds, 86% in 1–59 months and 93% in 0–28 days.


**Table 2.** Estimated deaths by cause and model for the year 2015 in children under five years of age.

\* Percent difference (n) = (GBD 2017–WHO-MCEE)/((GBD 2017 + WHO-MCEE)/2) × 100. \*\* Sum of specific infectious diseases from WHO-MCEE cause list (HIV/AIDS; diarrhoeal diseases; tetanus; measles; meningitis/encephalitis; malaria; acute respiratory infections; sepsis and other infectious conditions of the newborn). † Rates per 100,000 population in 'Under 50 and '1–59 months', and per 1000 livebirths for '0–28 days'. a Uncertainty intervals not available–rate calculated using n and under-5 population statistic from UN WPP 2017 Revision–year 2015 (1–59 months calculated using 59/60 months population). b Figures only account for neonatal sepsis deaths in China.

> However, the WHO-MCEE estimated >100,000 more deaths than the GBD 2017 when neonatal sepsis deaths were combined with meningitis/encephalitis, due to the WHO-MCEE's much higher estimate of neonatal sepsis deaths. Uncertainty intervals do not overlap between modelled estimates of deaths from neonatal sepsis in any of the age categories. This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn.

> The WHO-MCEE model estimated meningitis/encephalitis and neonatal sepsis as the second largest infectious cause of death, co-ranked with diarrhoeal diseases, in children aged under five years in 2015, after acute respiratory infections (Figure 1). In contrast the GBD 2017 estimated this cause category to be the third largest infectious case of death after acute respiratory infections and diarrhoeal diseases.

**Figure 1.** Meningitis/encephalitis and neonatal sepsis mortality burden estimates by model in relation to other selected infectious causes of death in children under five for the year 2015. **Figure 1.** Meningitis/encephalitis and neonatal sepsis mortality burden estimates by model in relation to other selected infectious causes of death in children under five for the year 2015.

At the country level, there was considerable variability in estimates of burden per population. For meningitis/encephalitis mortality rates, the WHO-MCEE model ranked Somalia highest for the year 2015 (139.7 deaths per 100,000 population), whilst the GBD 2017 ranked Somalia 11th highest for the same year (68.6 deaths per 100,000). At the country level, there was considerable variability in estimates of burden per population. For meningitis/encephalitis mortality rates, the WHO-MCEE model ranked Somalia highest for the year 2015 (139.7 deaths per 100,000 population), whilst the GBD 2017 ranked Somalia 11th highest for the same year (68.6 deaths per 100,000).

For numbers of deaths, both models attribute approximately 70% of all meningitis/encephalitis deaths in children under five years to just 12 countries including India, Nigeria, Pakistan, Democratic Republic of Congo (DRC), Ethiopia, Niger, Afghanistan, Mali, Uganda and China. However, whilst Somalia and Chad feature in the top 12 (ranked 7th and 8th respectively) in the WHO-MCEE estimates, they did not feature in the GBD 2017 top 12, where Indonesia and Burkina Faso featured instead (ranked 8th and 9th highest respectively) (Figure 2). For numbers of deaths, both models attribute approximately 70% of all meningitis/encephalitis deaths in children under five years to just 12 countries including India, Nigeria, Pakistan, Democratic Republic of Congo (DRC), Ethiopia, Niger, Afghanistan, Mali, Uganda and China. However, whilst Somalia and Chad feature in the top 12 (ranked 7th and 8th respectively) in the WHO-MCEE estimates, they did not feature in the GBD 2017 top 12, where Indonesia and Burkina Faso featured instead (ranked 8th and 9th highest respectively) (Figure 2).

### *3.2. Meningitis Incidence and Mortality Estimates by Aetiology in Children Aged Under Five Years*

The GBD 2017 and WHO-MCEE's pathogen models both estimated pneumococcal and Hib meningitis mortality and incidence in children aged 1 to 59 months at the national and global levels.

A comparison of the global estimates for the year 2015 (Table 3) showed that both models agree there were more cases of pneumococcal meningitis than Hib meningitis in 2015. However, whilst the GBD 2017 estimated around twice as many deaths from Hib meningitis compared to pneumococcal meningitis, the WHO-MCEE estimated around five times more deaths from pneumococcal meningitis than from Hib meningitis in the same year.

**Figure 2.** Top twelve ranking countries by meningitis/encephalitis mortality (number and rate) according to model for the year 2015. **Figure 2.** Top twelve ranking countries by meningitis/encephalitis mortality (number and rate) according to model for the year 2015.



\* Percent difference = (GBD 2017–WHO-MCEE)/((GBD 2017 + WHO-MCEE)/2) × 100. † Rates per 100,000 population

Pneumococcal meningitis

> Despite major differences in the relative proportions of meningitis deaths attributed to Hib and pneumococcal bacteria between models, both models agreed that Hib and pneumococcal meningitis combined were the underlying cause of approximately 40% of all meningitis/encephalitis deaths globally. Despite major differences in the relative proportions of meningitis deaths attributed to Hib and pneumococcal bacteria between models, both models agreed that Hib and pneumococcal meningitis combined were the underlying cause of approximately 40% of all meningitis/encephalitis deaths globally.

**GBD 2017 WHO-MCEE Pathogen Model Difference \*** 

**n Rate** † **n Rate** † **%** 

Cases 267,686 40.10 83,809 13 105% (179,314–374,902) (26.86–56.16) (36,160–168,500) (5–26)

Deaths 20,156 3.02 37,964 5 −61% (16,114–25,199) (2.41–3.78) (15,397–79,718) (2–11)

Cases 208,658 31.26 31,243 5 148% (139,815–304,035) (20.95–45.55) (13,386–50,595) (2–8)

*Microorganisms* **2021**, *9*, x FOR PEER REVIEW 8 of 17

**Table 3.** Global aetiology-specific meningitis deaths and cases, 2015, in children aged 1–59 months.

\* Percent difference = (GBD 2017–WHO-MCEE)/((GBD 2017 + WHO-MCEE)/2) × 100. † Rates per 100,000 population

When comparisons were made between the modelled estimates for Hib and pneumococcal meningitis incidence and mortality over time (Figure 3), both models showed a steeper decline in Hib meningitis incidence and mortality compared to pneumococcal meningitis mortality, which is consistent with wider roll-out of Hib vaccination globally compared to pneumococcal vaccination. However, the GBD 2017 consistently reported much higher incidence of pneumococcal and Hib meningitis over time compared to the WHO-MCEE. The GBD 2017 estimated Hib and pneumococcal incidence to be 31 and 40 cases per 100,000, respectively, in 2015 compared to the WHO-MCEE estimates of around five and 13 cases per 100,000 for Hib and pneumococcal meningitis, respectively. When comparisons were made between the modelled estimates for Hib and pneumococcal meningitis incidence and mortality over time (Figure 3), both models showed a steeper decline in Hib meningitis incidence and mortality compared to pneumococcal meningitis mortality, which is consistent with wider roll-out of Hib vaccination globally compared to pneumococcal vaccination. However, the GBD 2017 consistently reported much higher incidence of pneumococcal and Hib meningitis over time compared to the WHO-MCEE. The GBD 2017 estimated Hib and pneumococcal incidence to be 31 and 40 cases per 100,000, respectively, in 2015 compared to the WHO-MCEE estimates of around five and 13 cases per 100,000 for Hib and pneumococcal meningitis, respectively.

**Figure 3.** Estimated Hib/pneumococcal meningitis mortality and incidence amongst children aged 1–59 months according to model in relation to the proportion of children unimmunised with Hib vaccine and pneumococcal conjugate vaccine (PCV) over time.

Of note is that case fatality rates (CFRs) differed dramatically between the two sets of estimates. CFRs derived from WHO-MCEE global cases, and deaths estimates for Hib and pneumococcal meningitis in 2015, were 23% and 45%, respectively. However, CFRs calculated from GBD 2017 estimates were 8% for pneumococcal meningitis, 19% for Hib meningitis and 8% for meningococcal meningitis. Evidence from the literature closely agrees with the WHO-MCEE CFRs, consistently reporting higher CFRs from pneumococcal meningitis compared to Hib meningitis and meningococcal meningitis [21–26].

(PCV) over time.

#### *3.3. Modelling Methodology Which Could Account for Differences in Mortality and Pathogen Specific Estimates*

Figure 4 depicts a simplified methodology for both modelling approaches. A more detailed explanation is provided in the appendix, and full methodological approaches are also outlined elsewhere [3,27]. When calculating the meningitis death envelope, both models used country-specific death data from vital registration and other sources and applied statistical modelling to fill gaps in the data using country-specific covariates and drawing on trends observed where data was more complete. Whilst the GBD included intervention covariates (such as vaccine coverage) within their cause of death ensemble modelling (CODEm) (Figure 4), the WHO-MCEE model used intervention covariates in both their modelling, and also in post hoc adjustments, to redistribute causes accounting for interventions. Details of the covariates used by the models are available in the Supplementary Materials. Both models ensured that the sum of deaths attributed to different causes fitted within a total all-cause mortality envelope calculated from surveys, censuses and vital registration data. *Microorganisms* **2021**, *9*, x FOR PEER REVIEW 9 of 17 **Figure 3.** Estimated Hib/pneumococcal meningitis mortality and incidence amongst children aged 1–59 months according to model in relation to the proportion of children unimmunised with Hib vaccine and pneumococcal conjugate vaccine Of note is that case fatality rates (CFRs) differed dramatically between the two sets of estimates. CFRs derived from WHO-MCEE global cases, and deaths estimates for Hib and pneumococcal meningitis in 2015, were 23% and 45%, respectively. However, CFRs calculated from GBD 2017 estimates were 8% for pneumococcal meningitis, 19% for Hib meningitis and 8% for meningococcal meningitis. Evidence from the literature closely

Whilst there was little difference between estimated mortality from all causes and infectious diseases in children under five years (2% and 4% difference in estimated deaths, respectively), between models there was a marked difference between meningitis/encephalitis and neonatal sepsis mortality estimates in this age group (29 and 53 percent difference, respectively) (Table 2). agrees with the WHO-MCEE CFRs, consistently reporting higher CFRs from pneumococcal meningitis compared to Hib meningitis and meningococcal meningitis [21–26]. *3.3. Modelling Methodology Which Could Account for Differences in Mortality and Pathogen Specific Estimates* 

Further investigation into the modelling methods and underlying data showed that countries with the highest meningitis burden have the lowest quality death registration data. Whilst this is also the case for all causes of death, a higher proportion of meningitis/encephalitis death estimates were based on extrapolating from low-quality underlying data compared to all-cause death estimates. For example, 77% of meningitis/encephalitis deaths came from countries with no or very low-quality death registration data (scaled 0 to 1) compared to 60% of deaths due to all causes in the GBD 2017 model. Likewise, in the WHO-MCEE model, 95% of meningitis/encephalitis deaths were estimated using modelling underpinned by verbal autopsy (VA) studies compared to 90% of all cause deaths due to these countries having poor quality death registration data (see Supplementary Materials). As would be expected, there were greater differences between estimates from countries with low-quality underlying data compared to those with higher quality data (Figure 5). Figure 4 depicts a simplified methodology for both modelling approaches. A more detailed explanation is provided in the appendix, and full methodological approaches are also outlined elsewhere [3,27]. When calculating the meningitis death envelope, both models used country-specific death data from vital registration and other sources and applied statistical modelling to fill gaps in the data using country-specific covariates and drawing on trends observed where data was more complete. Whilst the GBD included intervention covariates (such as vaccine coverage) within their cause of death ensemble modelling (CODEm) (Figure 4), the WHO-MCEE model used intervention covariates in both their modelling, and also in post hoc adjustments, to redistribute causes accounting for interventions. Details of the covariates used by the models are available in the Supplementary Materials. Both models ensured that the sum of deaths attributed to different causes fitted within a total all-cause mortality envelope calculated from surveys, censuses and vital registration data.

**Figure 4.** *Cont*.

(Figure 5).

**Figure 4.** Simplified schematic of the different mortality modelling approaches. VR Data—Data from 76 countries with high quality VR data covering >80% of the population was mapped directly to cause of death categories (see appendix for ICD10 codes mapped to meningitis and sepsis and other severe infections in the neonatal period). VRMCM—Data from the countries with high quality VR data was used to fit a multinomial logistic regression model which was used to predict cause of death proportions in 38 low mortality countries (<35 deaths/1000 live births 2000–2010) with low quality VR data. Covariates used in the model are provided in appendix. VAMCM—In 78 high mortality countries (>35 deaths/1000 live births 2000–2010) verbal autopsy data from 119 research studies in 39 high mortality countries was used to fit a multinomial model to predict causes of death. Cause of death proportions for India were estimated using a combination of VAMCM for the neonatal period and data from the million deaths study and INDEPTH sites in India for the post neonatal period. See appendix for model covariates Other – Cause of death proportions for China were estimated using data from the China Maternal and Child Health Surveil-**Figure 4.** Simplified schematic of the different mortality modelling approaches. VR Data—Data from 76 countries with high quality VR data covering >80% of the population was mapped directly to cause of death categories (see appendix for ICD10 codes mapped to meningitis and sepsis and other severe infections in the neonatal period). VRMCM—Data from the countries with high quality VR data was used to fit a multinomial logistic regression model which was used to predict cause of death proportions in 38 low mortality countries (<35 deaths/1000 live births 2000–2010) with low quality VR data. Covariates used in the model are provided in appendix. VAMCM—In 78 high mortality countries (>35 deaths/1000 live births 2000–2010) verbal autopsy data from 119 research studies in 39 high mortality countries was used to fit a multinomial model to predict causes of death. Cause of death proportions for India were estimated using a combination of VAMCM for the neonatal period and data from the million deaths study and INDEPTH sites in India for the post neonatal period. See appendix for model covariates Other – Cause of death proportions for China were estimated using data from the China Maternal and Child Health Surveillance system. A complete explanation of methods used to produce WHO/MCEE estimates is outlined elsewhere [27].

lance system. A complete explanation of methods used to produce WHO/MCEE estimates is outlined elsewhere [27]. Whilst there was little difference between estimated mortality from all causes and infectious diseases in children under five years (2% and 4% difference in estimated deaths, respectively), between models there was a marked difference between meningitis/encephalitis and neonatal sepsis mortality estimates in this age group (29 and 53 percent difference, respectively) (Table 2). Further investigation into the modelling methods and underlying data showed that countries with the highest meningitis burden have the lowest quality death registration data. Whilst this is also the case for all causes of death, a higher proportion of meningitis/encephalitis death estimates were based on extrapolating from low-quality underlying data compared to all-cause death estimates. For example, 77% of meningitis/encephalitis To estimate meningitis mortality by aetiology, both models applied a proportional split by pathogen to the country-specific meningitis death envelope and adjusted for vaccine coverage. Whilst GBD 2017 pathogen specific mortality proportions were informed by vital registration (VR) data from data rich locations, the WHO-MCEE model based mortality proportions on studies reporting the distribution of pathogen-specific meningitis cases adjusted by pathogen-specific CFRs to derive proportions of deaths. This approach was used due to a lack of literature reporting meningitis mortality fractions by pathogen. To adjust pathogen-specific estimates according to vaccine coverage, IHME ran a metaregression model (DisMod-MR 2.1) with pneumococcal and Hib vaccine coverage as covariates driving down the proportions of disease attributed to those pathogens. The WHO-MCEE model used a deterministic approach to account for vaccine use by calculating the percentage reduction in disease as a result of vaccine efficacy, coverage and, in the case of PCV, the vaccine product and proportion of disease caused by vaccine-specific serotypes.

deaths came from countries with no or very low-quality death registration data (scaled 0 to 1) compared to 60% of deaths due to all causes in the GBD 2017 model. Likewise, in the WHO-MCEE model, 95% of meningitis/encephalitis deaths were estimated using modelling underpinned by verbal autopsy (VA) studies compared to 90% of all cause deaths due to these countries having poor quality death registration data (see Supplementary Materials). As would be expected, there were greater differences between estimates from countries with low-quality underlying data compared to those with higher quality data The models used very different approaches for estimating incidence by aetiology. The GBD 2017 calculated meningitis incidence independently from meningitis mortality using incidence data gathered from hospital records, claims data and a systematic review of the literature capturing incidence studies. The WHO-MCEE incidence estimates were derived by dividing pathogen-specific death estimates by literature-derived CFRs. The WHO-MCEE also published an update to a previous incidence-based model for Hib and pneumococcal meningitis [28], which predicted even lower incidence rates for pneumococcal meningitis and similar rates for Hib.

**Figure 5.** Absolute difference between WHO-MCEE and GBD 2017 meningitis/encephalitis mortality estimates according to country. **Figure 5.** Absolute difference between WHO-MCEE and GBD 2017 meningitis/encephalitis mortality estimates according to country.

#### **4. Discussion**

To estimate meningitis mortality by aetiology, both models applied a proportional split by pathogen to the country-specific meningitis death envelope and adjusted for vaccine coverage. Whilst GBD 2017 pathogen specific mortality proportions were informed by vital registration (VR) data from data rich locations, the WHO-MCEE model based mortality proportions on studies reporting the distribution of pathogen-specific meningitis cases adjusted by pathogen-specific CFRs to derive proportions of deaths. This approach was used due to a lack of literature reporting meningitis mortality fractions by Despite major differences in the number of deaths attributed to meningitis, both models agree that there is a substantial burden of disease, with meningitis as either the 2nd or 3rd most important infectious syndrome. By far the biggest burden of meningitis is estimated to occur in countries with low quality or no death registration data where these models rely heavily on extrapolating from VA studies. Accurately attributing meningitis as a cause of death using VA is extremely challenging [29–31] and could lead to meningitis as a cause of death being underestimated. VA has a high specificity but low to moderate sensitivity for meningitis [32–34] and can easily attribute death from meningitis to a different cause, especially in malaria endemic regions where severe febrile illness is often assumed to be malaria [35–37].

If these syndromic models systematically underestimate deaths from meningitis, this would result in an underestimate of incidence by pathogen in the WHO-MCEE model because incidence is derived by dividing estimated deaths by CFR based on location and pathogen. The GBD 2017 estimated pathogen-specific incidence separately to pathogenspecific deaths and produced higher estimates than the WHO-MCEE model, but the incidence estimates were out of line with deaths when literature-derived CFRs were applied. Following a meeting where results from this analysis were presented to all modelling groups, the IHME amended their methodology for calculating pathogen-specific incidence. In the recently published GBD 2019 model [38], published studies and hospital data were used to estimate pathogen-specific CFRs as a function of healthcare access and quality. Pathogen specific mortality was then derived from estimates of pathogen-specific incidence and CFRs.

Using global health estimates to derive baseline numbers and targets against which progress can be measured is challenging. Estimates for the entire time series are updated with successive model iterations as new input data are considered and amendments are made to statistical modelling processes. This means that baseline estimates for a given year fluctuate with successive model iterations.

It is vital that the methods used to derive estimates are clearly communicated. Across models it was unclear from published methods exactly how neonatal meningitis as a cause was disaggregated from neonatal sepsis, and other infectious conditions of the newborn, when we know that the majority of the underlying input data does not distinguish between these two causes of death. Unless methods are made transparent, it is difficult for policy makers to understand, and therefore trust, model outputs [39].

Experts responsible for monitoring progress also need to know exactly how estimates were derived in order to assess whether they are capable of measuring progress against certain indicators. Whilst both models accounted for PCV and Hib vaccine impact, they did so using substantially different methods. The IHME's GBD 2017 study accounted for vaccine impact by finding existing relationships between vaccine coverage and the proportion of pathogen-specific meningitis targeted by the vaccine (from countries where data is available) and using these existing relationships to make predictions where data is unavailable. Whilst this approach has an advantage of using as much raw data as possible, it does not distinguish between differences in vaccine products and the varying efficacy associated with different dosing schedules between countries. Although incidence proportion models included data from some countries in sub-Saharan Africa and Asia, the use of VR data alone to determine proportional cause of death means that vaccine effects on pathogen-specific mortality in high mortality countries with no vital registration data are heavily reliant on effects demonstrated in data-rich low-mortality countries. The WHO-MCEE, on the other hand, make predictions where data is sparse/unavailable by simulating the effect of a given vaccine over time on a country specific basis. Assumptions about vaccine impact are transparent and take into account differences in vaccine formulations and dosing schedules, but they may be applied to a pathogen specific meningitis death estimate which is highly uncertain.

It is also important for decision makers to be aware that even in data-rich locations, global health estimates for the most recent year can be based on predictions rather than real underlying data. These estimates may, therefore, be unsuitable for tracking change as a result of a recent intervention, especially if the intervention has not been accounted for as a covariate in the model.

The IHME's GBD model is currently the only available complete source of information about the global and national burden of meningitis amongst all age groups and for most of the pathogens of interest to the global roadmap to defeat meningitis. The IHME have also improved some of their methods for the latest round of estimates by including more surveillance data from high mortality settings in the GBD 2019 than was included in the GBD 2017. Additionally, there are plans for future versions of the IHME's model to include estimates on the incidence and mortality from GBS meningitis, one of the major causes of meningitis in neonates worldwide. However, tracking outputs from multiple models in parallel has advantages in identifying areas of higher uncertainty, generating opportunities for modellers to improve methods and prioritising further primary data collection/strengthening surveillance. An interactive visualisation has been created to track progress using estimates from all of the major global health estimation models [40].

None of the models we assessed were able to accurately account for the fluctuating scales of periodic, large epidemics of meningitis, which are irregular and unpredictable in nature. Whilst GBD 2017 attempted to account for epidemic meningococcal meningitis deaths by adding these to the meningitis death envelope, they did not use equivalent methodology to account for epidemic meningococcal meningitis cases. The WHO-MCEE syndromic model attempted to account for epidemic disease by estimating the average increase in deaths in epidemic years relative to nonepidemic years and adding these to estimates in years with epidemics identified by WHO surveillance reports and published literature. This increases estimates during an epidemic year, but the underlying data from the country are not always reliable, and it does not accurately reflect the variation in the size of the epidemic for a given year. The WHO-MCEE pathogenic model only estimated pathogen-specific deaths for endemic disease, removing the simulated effects of epidemics from the syndromic model before applying proportional splits to the remaining meningitis envelope. Therefore, neither model estimating pathogen specific causes of meningitis was able to account for epidemic pneumococcal meningitis, yet this is an important consideration because it has been demonstrated as having a significant mortality burden [41].

Considering the current limitations of modelled meningitis estimates, it is desirable to track progress alongside additional data where possible. Countries across the African meningitis belt experience the highest burden of meningitis globally because they are susceptible to large and devastating outbreaks of meningococcal disease linked to climatic factors such as dry winds, low humidity and high levels of dust in the air [42]. Whilst many of these countries have poor death registration systems, they have relatively rich and complementary meningitis surveillance systems. Since 2003 an enhanced meningitis surveillance network has been established across the meningitis belt to strengthen outbreak detection and enable a rapid response to outbreaks of meningococcal disease across the region [43]. The network now covers 24 countries, reporting suspected cases and deaths from meningitis to the WHO intercountry support team (WHO/IST) each week during the meningitis season and every month for the rest of the year [44]. Case-based surveillance systems have been established in five countries within the region allowing for comprehensive information on CFRs by age [45].

Triangulating modelled estimates against surveillance data provides the opportunity to reality-check modelled outputs. Utilising surveillance data in combination with evidence of age and regionally specific CFRs has already successfully been used by experts wishing to monitor global progress towards the 2005 measles mortality reduction goal because measles mortality estimates calculated from vital registration data were considered an unreliable way to track progress [46]. Surveillance data for meningitis is not currently available for every country worldwide. However, comprehensive roll out of Hib and pneumococcal vaccines is driving down incidence and mortality from meningitis caused by these pathogens across the globe. Improved pathogen-specific surveillance informed by accurate and timely laboratory diagnosis is required to adequately assess the impact of these important life-saving interventions. This is particularly important for countries transitioning out of Gavi support which need to justify national investments in these vaccines. Additionally, all member states of the UN have committed to achieving universal health coverage by signing up to the SDGs, so there is reason to believe that the availability of good quality surveillance data will improve over time as health systems are strengthened.

More work is required to provide credible meningitis burden estimates for measuring progress. Currently meningitis mortality estimates are highly uncertain because the models rely heavily on death registration data, which is largely missing or incomplete in countries with the highest meningitis burden. Additionally, since postmortem examination is rarely performed in countries without vital registration systems, and the symptoms of meningitis can easily be mistaken for other diseases, there is a risk that the mortality burden of meningitis could be underestimated. Encouragingly, better data on cause of death are becoming available in regions where child mortality rates are the highest through the use of minimally invasive tissue sampling [47,48] and inclusion of these data in future models could considerably improve the reliability of their outputs.

#### **5. Conclusions**

Global meningitis estimates should be interpreted with caution. Tracking progress towards controlling this disease should also include analysis of real surveillance data where available. The WHO Defeating Meningitis by 2030 Global Roadmap will improve awareness, diagnosis and surveillance of meningitis. As the roadmap drives more comprehensive data on meningitis, a convergence in modelled estimates and a more reliable picture of reductions in the burden of meningitis are anticipated.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2076-260 7/9/2/377/s1 [49–51].

**Author Contributions:** Conceptualization, C.W., N.B., L.G., V.S., J.M.S. and C.T.; methodology, C.W., N.B., L.G., J.M.S., C.T.; validation, R.B., H.K., H.Y.W., L.L., D.Y., M.D.K., B.W.; formal analysis, C.W. and N.B.; writing—original draft preparation, C.W. and N.B.; writing—review and editing, C.W., L.G., V.S., R.B., H.K., H.Y.W., L.L., D.Y., M.D.K., B.W., J.M.S., C.T.; visualization, C.W.; supervision, L.G., J.M.S. and C.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no specific funding, but Meningitis Research Foundation has received unrestricted educational grants from GSK, Pfizer and Sanofi which allowed this study to take place.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Publicly available datasets were analyzed in this study. GBD 2017 estimates are available from http://ghdx.healthdata.org/gbd-2017; WHO-MCEE syndromic meningitis estimates are available from http://158.232.12.119/healthinfo/global\_burden\_disease/estimates/ en/index2.html and WHO-MCEE syndromic estimates are available from their publication [14].

**Acknowledgments:** This work arose from a meeting to evaluate the global burden of meningitis estimates convened by the Gates Foundation in November 2018.

**Conflicts of Interest:** C.T. received a consulting payment from GSK in 2018 outside the submitted work. C.W., N.B., L.G., V.S. took part in this research as employees of Meningitis Research Foundation, which has received unrestricted educational grants from GSK, Pfizer and Sanofi. All other authors declare no conflict of interest. GSK, Pfizer and Sanofi had no role in the design of this study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


**Nicoline Schiess 1,\*, Nora E. Groce <sup>2</sup> and Tarun Dua <sup>1</sup>**


**Abstract:** The burden, impact, and social and economic costs of neurological sequelae following meningitis can be devastating to patients, families and communities. An acute inflammation of the brain and spinal cord, meningitis results in high mortality rates, with over 2.5 million new cases of bacterial meningitis and over 236,000 deaths worldwide in 2019 alone. Up to 30% of survivors have some type of neurological or neuro-behavioural sequelae. These include seizures, hearing and vision loss, cognitive impairment, neuromotor disability and memory or behaviour changes. Few studies have documented the long-term (greater than five years) consequences or have parsed out whether the age at time of meningitis contributes to poor outcome. Knowledge of the socioeconomic impact and demand for medical follow-up services among these patients and their caregivers is also lacking, especially in low- and middle-income countries (LMICs). Within resource-limited settings, the costs incurred by patients and their families can be very high. This review summarises the available evidence to better understand the impact and burden of the neurological sequelae and disabling consequences of bacterial meningitis, with particular focus on identifying existing gaps in LMICs.

**Keywords:** meningitis; burden; social and economic costs; neurological sequelae; WHO meningitis roadmap; tuberculous meningitis; disability

#### **1. Introduction**

Many different bacteria can cause meningitis; however, *Streptococcus pneumoniae* (Sp or pneumococcus), *Haemophilus influenzae* type b (Hib) and *Neisseria meningitidis* (Nm or meningococcus) are the most common pathogens other than those in infants, who are most commonly affected by *Streptococcus agalactiae* (group B streptococcus or GBS) [1]. Prior to the advent of widespread vaccination campaigns, bacterial meningitis outbreaks imparted a significant toll, with some pathogens, such as group A *Neisseria meningitidis*, having meningitis rates as high as 1% of the population during major African epidemics in the last century [2]. Tuberculosis, which affects millions of people each year worldwide, predominantly in low- and middle-income countries (LMICs) [3] affects the central nervous system in approximately 1% of cases [4] yet can also result in profound mortality and morbidity [5]. Multiple factors contribute to the impact or severity of different pathogens causing meningitis. Meningococcus and pneumococcus can cause severe central nervous system damage and have the propensity to cause sepsis, a significant cause of mortality. However, other comorbid conditions can also impact the severity and sequelae of meningitis-causing pathogens. These include malnutrition, immunocompromising conditions, and delays in diagnosis and treatment.

Globally, the epidemiology of bacterial meningitis has changed dramatically with the introduction of conjugate vaccines [6,7]. The Hib conjugate vaccine has essentially eradicated Hib meningitis [6,8–10], and with widespread use of the meningococcal serogroup

**Citation:** Schiess, N.; Groce, N.E.; Dua, T. The Impact and Burden of Neurological Sequelae Following Bacterial Meningitis: A Narrative Review. *Microorganisms* **2021**, *9*, 900. https://doi.org/10.3390/ microorganisms9050900

Academic Editor: James Stuart

Received: 16 March 2021 Accepted: 19 April 2021 Published: 22 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

A conjugate vaccine (MACV), the overall burden of suspected meningococcal meningitis cases has been reduced by almost 60% in high-risk countries across northern Africa ("Meningitis belt") with near-complete elimination of confirmed serogroup A disease [11]. Pneumococcal conjugate vaccines (PCVs) have also resulted in a slight decrease in pneumococcal disease [12], and in many countries, this pathogen has overtaken *H. influenzae* as the most common cause of meningitis [6,13,14]. Despite these advances, there were still over 2.5 million new cases of bacterial meningitis and over 236,000 deaths worldwide in 2019 alone [15].

Meningitis survivors can be left with disabling neurological sequelae such as seizures, hearing and vision loss, neuromotor disability and hydrocephalus. Cognitive and behavioural sequelae following bacterial meningitis have also been reported [16,17]; however, it is likely that these more subtle sequelae may sometimes go undiagnosed and can have profoundly detrimental effects on school and work performance. The burden of disabling sequelae is highest in low- and middle-income countries (LMICs) as these countries have high rates of meningitis [16].

Over the past several years, the expansion of meningitis-related vaccination programs, increasing research and intervention efforts, and growing advocacy on behalf of meningitis survivors and their families have presented significant possibilities for both meningitis prevention and life improvement for survivors. However, coordination of these advances has been lacking. In response to this, a new international response to meningitis is now underway; WHO's Defeating Meningitis by 2030 Global Roadmap [18] intends to address the global issues around bacterial meningitis (meningococcus, pneumococcus, *Haemophilus influenzae* and group B streptococcus), with one of the main goals focusing on the long-term sequelae of meningitis and quality of life. A key activity proposed in the meningitis roadmap is to conduct research on the socioeconomic impact of sequelae on children, adults and their families/carers and on the availability and effectiveness of aftercare/support interventions.

In this review, we summarise the evidence to better understand the impact and burden of the neurological sequelae and disability of bacterial meningitis, with a focus on LMICs and with particular attention to the long-term impact of meningitis on those who survive, thus advising the third goal of the Defeating Meningitis Roadmap.

#### **2. Global Burden of Meningitis**

In 2019, worldwide mortality from all causes of meningitis (excluding tuberculous and cryptococcal meningitis) was over 236,000 deaths, with approximately 2.5 million new cases [15]. Additionally, in 2019, meningitis ranked sixth in the top causes of disability adjusted life years (DALYs) in children under 10 years of age [19]. While global deaths due to meningitis decreased between 1990 and 2016, the 21% decrease pales in comparison to the dramatic reductions in mortality from other diseases such as measles (93%) and tetanus (91%) [20].

#### *Meningitis Belt*

In 2019, there were over 22,000 suspected cases of meningitis, with 1261 deaths reported to the WHO in African countries sharing data [21]. A disproportionally high rate of bacterial meningitis occurs in Africa due to elevated endemic disease, a younger population and regularly occurring epidemics across the "meningitis belt"—a span of countries between Ethiopia and Senegal that includes Nigeria, Burkina Faso and Sudan (See Figure 1). Outbreaks in these countries are characterised by sporadic seasonal infections, with periodically superimposed larger epidemics. Although the burden of meningitis in this region has declined following the introduction of a MACV in 2010, other meningococcal serogroups and bacterial pathogens continue to cause endemic and epidemic disease [22,23]. As these epidemics have a profound effect on the population, meningitis is considered a priority disease in the WHO integrated disease surveillance and response platform [24].

**Figure 1.** African meningitis belt. Source: "Meningitis outbreak response in sub-Saharan Africa: WHO Guideline. Geneva: World Health Organization; 2014".

The costs of meningitis outbreaks for governments and ministries of health are significant as well. A Colombini et al. review of the Burkina Faso public health response [25] estimated the total cost for the 2006–2007 epidemic season to be 9.4 million USD—three quarters of which was covered by the government and the Ministry of Health's financial and technical partners. The remaining cost was absorbed by the families of meningitis victims. The review noted challenges that included medicine shortages, a paucity of healthcare workers and a lack of government funding for medication. [25]. The highest cost were the vaccine and injection supplies themselves. Vaccine transportation and personnel costs were the next highest cost although they were a fraction of that of the vaccines and injection supplies. The cost of potential long-term neurological sequelae and the associated expenses of rehabilitation were not evaluated.

#### **3. Neurological Sequelae**

#### *3.1. Frequency and Types of Neurological Sequelae Following Meningitis*

Acute bacterial meningitis can have severe complications with long-term neurological sequelae resulting in disability even in high-income countries (HICs) with appropriate antibiotic therapy and vaccine availability [26]. For example, a recent study in the United States on paediatric bacterial meningitis demonstrated a 45.9% complication rate at 30 days for community-acquired bacterial meningitis, with hydrocephalus (20.8%), intracranial abscess (8.8%) and cerebral oedema (8.1%) being the most common short-term neurological sequelae [27].

A large systematic review and meta-analysis by Edmond et al. estimated the risks of neurological sequelae globally by region and socioeconomic status from 1980 to 2008 and determined that the risk of suffering from some type of sequelae after bacterial meningitis was 20%. The risk was almost threefold higher in Africa and Asia compared to Europe [16]. Treatment delay [28,29], length of travel to receive care, lower immune defences as a result of chronic malnutrition and cost of hospital care [30] have all been reported to contribute to this elevated risk in low-income countries.

While hearing loss and seizures were the most common sequelae among the 132 studies in the Edmond et al. meta-analysis, cognitive impairment clearly affects a large proportion of survivors and, in LMICs, is no doubt underestimated considering that only two studies from Africa and Asia specifically evaluated cognitive domains. In many studies from LMICs, standardised assessment tools and thorough neurological examinations are not utilised and therefore do not capture possible subtle manifestations such as neurocognitive impairment or behavioural changes [16].

In addition, most studies do not compare the rates of sequelae among children with or without a history of meningitis. This method might provide a more accurate picture of the risk of sequelae after bacterial meningitis by controlling for baseline rates of neurological disorders within a population. This method was utilised in a prospective cohort study in Senegal that used standardised assessment tools on both the control and affected groups, making comparison and categorisation more reliable. The affected children in Senegal were found to have 3 times higher odds of major disability (such as cognitive/motor deficits, hearing loss or seizures) after suffering from bacterial meningitis when compared with a community control group. Multiple domains were often involved, the most frequent being cognitive and motor deficits with seizures [31]. Almost 40% of affected children did not attend pre-school or school compared with 16.7% of the control group. The importance of including a control group is underscored by the results in this study that showed that, while 51.8% of children with prior meningitis had hearing loss, a substantial number (30.3%) of children in the control group also had hearing loss, possibly due to untreated otitis media within the population.

#### *3.2. Persistence of Sequelae over Time*

The study follow-up time after acute infection is also an important component as subtle deficits, including poor school performance, behavioural issues and undiagnosed attention deficit disorder, may not be appreciated initially and can affect survivors for many years [32,33]. A survey of parents and teachers in the United Kingdom on 739 infantile meningitis cases and 606 matched controls was conducted years later when the subjects were teenagers. The results of the study showed that 46% of parents of affected children reported behavioural problems compared to 21% in the control group. The percentages of behavioural problems reported by their teachers were 37% and 23%, respectively [34].

A 2011 systematic literature review by Chandran et al. focused specifically on neurological sequelae five years or more after the acute attack. Searching all globally published articles of the consequences or sequelae of bacterial meningitis in children (one month to 18 years), they identified that almost one-half of survivors five years out or longer suffered from some type of sequelae, with over three fourths having intellectual or behavioural problems [32]. This study is particularly important as it defined "long-term" as five years or longer in contrast to other observational studies that either specified "long-term" as any time post-discharge or had no defined follow up [33,35–37].

Control-based studies examining the sequelae of meningitis ten years or longer after infection also have the potential to parse out the risks of sequelae according to age of infection. In other words, does the age of meningitis onset contribute to the severity of long-term sequelae or predict outcome? This question was examined by Anderson et al. in a longitudinal, prospective study that focused specifically on the age of illness and long-term sequelae in meningitis survivors 12 years later [38]. Reassuringly, those who had had meningitis did not show progressive deterioration when compared to healthy controls, indicating the ability to developmentally compensate in executive functioning. However, a clear difference showed that those who had had meningitis prior to one year of age had poorer performances in certain domains such as language and executive functioning compared to those who had meningitis after 12 months.

#### *3.3. Neurological Sequelae in LMICs*

Few studies in LMICs have examined the long-term neurological sequelae following bacterial meningitis. The large systematic review and meta-analysis of all sequelae post discharge by Edmond et al. [16] revealed that the number of studies published on disabling sequelae was much higher in regions such as Europe (40%) and the Americas (24%) versus Asia (6%) and Africa (10%). A different systematic review in 2009 by Ramakrishnan et al. included 6029 African children under age 15 years with confirmed meningitis in 21 African countries and revealed that nearly 20% of bacterial meningitis survivors experienced neurological sequelae while in the acute hospitalised setting [35]. Notably, only seven of these countries had post-discharge follow-up studies with the follow-up time ranging from 3 to 90 months. The total number of patients included in these studies was much lower (Table 1. Significantly, the analysis found that 10% of children died after discharge and that 25% (range 3–47%) had neurological sequelae 3–60 months after diagnosis based on clinical exam alone [35].

**Table 1.** The post-discharge sequelae in children with all causes of bacterial meningitis for studies with >25 subjects.


Neurological sequelae defined as behavioural problem, cognitive delay, speech or language disorder, seizures or vision loss. Hib = *Haemophilus influenzae* type b; Nm = *Neisseria meningitidis*; Spn = *Streptococcus pneumoniae*; adapted from Ramakrishnan et al., 2009 [35].

> Similarly, in Bangladesh, a study on children with pneumococcal meningitis showed that many survivors had hearing (33%), vision (8%), mental (41%) and psychomotor deficits (49%) within 40 days post-discharge. A second group of pneumococcal meningitis survivors in the study were followed up at 12–24 months and showed deficits in hearing (18%), vision (4%), and mental (41%) and psychomotor development (35%) [49].

#### **4. Social and Economic Burden of Neurological Sequelae**

Globally, but particularly in Africa, there are limited data on the long-term social and economic burden of neurological sequelae among meningitis survivors and their families. Social and economic factors can dramatically affect survivors' ability across the life course to perform in school or to obtain gainful employment, particularly as the risk of sequelae in children under five years has been found to be double that for children older than five [16]. While some children have very severe sequelae, there are many other children who are less severely affected. Neurodevelopmental delays can often be subtle and may not be adequately diagnosed in routine clinical exams. Cognitive and behavioural difficulties may only be noticed once a child has started school [17], or they may remain undiagnosed. Whether undiagnosed or simply unable to access adequate resources for help and support, these children may struggle to keep up, be labelled as delayed, and drop or flunk out, setting themselves up for a lifetime of limited opportunities.

For children in particular, downstream consequences of neurological sequelae can be dire for the whole family, with studies showing that caregivers are often forced to choose care for their disabled child versus working to generate an income or provide for other siblings [49,50]. A study of 107 South African children with TB meningitis who lived in low socioeconomic environments showed that 19% of all mothers reported experiencing financial difficulty after their child fell ill [50]. A reported case in Bangladesh painfully illustrates what a profound impact a disabled child can have on the whole family. A young boy, initially misdiagnosed and thus treated late for pneumococcal meningitis, lost key developmental milestones. The family's socioeconomic status underwent a dramatic change as a result of his disabilities. To pay for medical bills, the father was forced to sell his small piece of land and to work several jobs, barely earning enough to feed the family. The mother attends to all his needs, neglecting care for the rest of the family. An elder sibling's education was disrupted since they could not afford school supplies [49]. As noted by the authors:

"In countries like Bangladesh, (the impact of impairments) is quite different from that in developed parts of the world, because of very limited facilities for the education of these children and almost no priority for the facilitation of a normal life. As a result, most of these children cannot have an independent life, are unable to participate in any social activities, and remain confined at home. All these factors have psychological, social, and financial impacts on the entire family and on society." [49].

Even if patients have access to public health services, along with care in a timely manner and a reasonable distance, costs associated with medical care can be financially devastating to families and communities as personal household earnings or savings cover many expenses of medical care in countries such as Kenya [51] and Burkina Faso [52]. A study in Burkina Faso estimated the total average cost for each family to treat a child with a meningitis episode to be approximately 34% of the GDP per capita. For children with additional neurological sequelae, the total cost over the course of the two-year 2006–2007 epidemic was near the GDP per capita level. With little or no disposable income, most households were forced to sacrifice one or more basic necessities to pay for care [52].

#### **5. Neurological Disability, Quality of Life and Access to Care**

The challenges of those living with disability—as a permanent sequela from meningitis or from any other cause—are coming to the forefront of discussion within the scientific and public health community [53,54]. Stigma, restriction to education or employment opportunities, and a lack of specialised follow-up healthcare further result in a negative feedback loop that creates an economic gap between households with a disabled member and those without [55]. To illustrate this, a prospective cohort study of disabling meningitis sequelae in Senegal revealed that 40% of children who had had meningitis did not attend school compared to 17% of children with no history of meningitis [31]. Another study of 112 confirmed meningitis patients admitted to a children's hospital between 1992 and 2007 in the United Kingdom revealed that, 8 years after acute meningitis, both parents (32%) and teachers (19%) reported behavioural problems and lowered health-related quality of life (HRQoL) on Pediatric Quality of Life inventory (PedsQL) measurements. The authors of this review highly recommended that meningitis survivors be specifically screened for psychiatric and neurobehavioral difficulties at certain stages of development [56].

As limited as the data are on children in LMICs regarding the long-term impact of meningitis, even more striking is the lack of data related to how adolescents and adults fare in the aftermath of meningitis. HRQoL studies assessing the emotional, psychological, social and behavioural effects of meningitis are lacking in both HICs and LMICs. A 2018 systematic review of the quality-of-life impact on both patients and carers following invasive meningococcal disease in HICs found no studies describing HRQoL for patients who had meningitis-induced sequelae [57]. However, in survivors, particularly adolescents and young adults, self-esteem, friendships, well-being and school performance are important aspects of a good quality of life and problems in these areas also affect caregivers and the

community. The implications for someone disabled as a child are profoundly different than when disabled as an adult.

Recognition of those suffering from meningitis-induced disability and their access to (or lack of) resources is an important first step in order to provide equal opportunities for care, rehabilitation, specialised education and employment. For example, a study looking at 107 South African children with TB meningitis showed that, overall, less than half of children with documented neurological sequelae attended specialty clinics for follow-up care and that those in rural settings did not have access to these services [50]

The ramifications of meningitis in adults is no less significant. A range of shortand long-term sequelae including vision loss, neurological (cranial nerve palsies, aphasia, paresis and seizures) or neurobehavioral sequelae and cognitive impairment are found in adults [58,59], even among those considered to have made a "good" recovery from bacterial meningitis [60]. One of the few large studies looking at cognitive sequelae in adults was conducted by van de Beek et al. in Denmark in 2002. Fifty-one adult survivors "with good recovery" after bacterial meningitis were evaluated 6–24 months following meningitis. Cognitive disorders and lower scores in general health and quality of life were found in 27% of cases [60]. The social and economic impacts on individuals thus affected by the disease are profound even following a reported recovery. A study in the UK focusing on tuberculous meningitis in adults found that over one-third of survivors had residual neurological sequelae one year later [5].

A significant number of children and adults permanently affected by meningitis will live with one or more permanent disabilities. Increasingly, it is recognised that, in addition to medical and (neuro-focused) rehabilitative supports, where available, the lives of these individuals and their families can be dramatically improved by ensuring that they are also linked to a rapidly evolving global disability rights effort to improve the lives of persons with disabilities. Improving access to care by strengthening referral systems and health systems can subsequently also improve care for people who have disability from other types of meningitis or even other nervous system diseases.

In addition to services and support that may be available to children and adults disabled by meningitis, it is important to emphasise that additional resources for people with disability are often available and overlooked by individuals and clinical services that are wholly focused on meningitis. This includes Disabled People's Organizations (DPOs), organisations run by and for persons with disabilities, and disability-focused government services and charities that are available to all disabled members of the community. Such organisations can be found at both the local and national levels in both HICs and LMICs. Such support services often can help with education, employment and advice on social services and economic support programmes available through government agencies and local charities. Importantly, such organisations can advise people disabled by meningitis on their rights and entitlements designated under local and national disability law. For example, currently 164 countries are signatories to the United Nations Convention on the Rights of Persons with Disabilities, which means that their national laws should be in alignment with this international human rights declaration [61]. These advances are not limited to only improved access to health care and social services but have broader educational and socioeconomic implications. For example, the identification and inclusion of disabled people and their households into development efforts has been a significant part of this new global disability effort, with initiatives underway towards improving disabled children's right to education and efforts for adults to improve their socioeconomic status and involvement in the workforce, their right to self-determination, and their right to equal involvement in their communities and their societies. The resources available to DPOs and disability-focused services vary from one country to the next and, in LMICs, are often limited, but these organisations are an important and growing resource for people with disabilities around the world. Those disabled as the result of meningitis and those involved in providing care and support to those disabled by meningitis should be aware

of the potential benefits that links with the DPOs, government services, charities and the broader Disability Rights Movement can provide.

#### **6. Conclusions**

The burden, impact, and social and economic costs of neurological sequelae following meningitis can be devastating to patients, families and communities. Severe sequelae can present as seizures, hearing and vision loss, and neuromotor disability; however, it is likely that more subtle effects such as cognitive impairment, memory and behaviour changes are often overlooked and can have detrimental effects on school and work performance. Importantly, the majority of studies have not followed patients after five years. The longterm consequences, socioeconomic impact and demand for medical follow-up services for these patients and their caregivers is essentially unknown in many LMICs such as those located in the meningitis belt of Africa. More research on the care and support needs of patients and families would be valuable, and early recognition, improved management, support services, and access to care should be priority areas for research and funding programs. Building links to local, regional and global organisations that advocate on behalf of broader disability issues also provides additional support for improving the lives of children and adults with long-term sequelae of meningitis and their families.

**Author Contributions:** Conceptualization, N.S., N.E.G., T.D. writing—original draft preparation, N.S., N.E.G.; writing—review and editing, T.D.; supervision, T.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors gratefully acknowledge the edits and insightful comments to the manuscript provided by Lucy McNamara and Christine Dubray.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Disclaimer:** The authors are responsible for the views expressed in this article. Those views do not necessarily represent the views, decisions, or policies of the institutions with which they are affiliated.

#### **References**


### *Review* **Vaccines to Prevent Meningitis: Historical Perspectives and Future Directions**

**Mark R. Alderson \*, Jo Anne Welsch, Katie Regan, Lauren Newhouse, Niranjan Bhat and Anthony A. Marfin**

Center for Vaccine Innovation and Access, PATH, Seattle, WA 98121, USA; jwelsch@path.org (J.A.W.); kregan@path.org (K.R.); lnewhouse@path.org (L.N.); nbhat@path.org (N.B.); aamarfin@path.org (A.A.M.) **\*** Correspondence: malderson@path.org; Tel.: +1-206-302-4855

**Abstract:** Despite advances in the development and introduction of vaccines against the major bacterial causes of meningitis, the disease and its long-term after-effects remain a problem globally. The Global Roadmap to Defeat Meningitis by 2030 aims to accelerate progress through visionary and strategic goals that place a major emphasis on preventing meningitis via vaccination. Global vaccination against *Haemophilus influenzae* type B (Hib) is the most advanced, such that successful and low-cost combination vaccines incorporating Hib are broadly available. More affordable pneumococcal conjugate vaccines are becoming increasingly available, although countries ineligible for donor support still face access challenges and global serotype coverage is incomplete with existing licensed vaccines. Meningococcal disease control in Africa has progressed with the successful deployment of a low-cost serogroup A conjugate vaccine, but other serogroups still cause outbreaks in regions of the world where broadly protective and affordable vaccines have not been introduced into routine immunization programs. Progress has lagged for prevention of neonatal meningitis and although maternal vaccination against the leading cause, group B streptococcus (GBS), has progressed into clinical trials, no GBS vaccine has thus far reached Phase 3 evaluation. This article examines current and future efforts to control meningitis through vaccination.

**Keywords:** meningitis; meningococcus; pneumococcus; *Haemophilus influenzae*; Hib; group B streptococcus; conjugate vaccine

#### **1. Introduction**

Despite advances against individual pathogens, bacterial meningitis and sepsis remain public health challenges globally. Meningitis, characterized by inflammation of the meninges, is swift and severe and is associated with significant morbidity and mortality. Low- and middle-income countries (LMICs) suffer the greatest burden, with the African Meningitis Belt, a string of 26 countries from Senegal and The Gambia in the west to Ethiopia in the east, experiencing a disproportionate share of disease [1]. Bacterial meningitis epidemics are common in this region and many have been large-scale, threatening economic stability alongside human life. However, outbreaks and epidemics can occur globally [2,3]. There are an estimated 5 million cases of meningitis each year, with up to 300,000 deaths—nearly half of which are in children younger than five years of age (u5) [4]. Survivors are not always spared; a high proportion suffer long-term after affects including hearing loss, visual, physical, and cognitive impairment, and limb loss. Despite this sobering reality, progress against meningitis lags that of other vaccine preventable diseases [4].

The Global Roadmap to Defeat Meningitis by 2030, an initiative to raise awareness of bacterial meningitis as a public health problem and create a framework for addressing it, aims to reverse this trend. Critical goals include eliminating bacterial meningitis epidemics and reducing cases and deaths from the most significant causes of bacterial meningitis: *Haemophilus influenzae* type B (Hib), *Neisseria meningitidis* (meningococcus), *Streptococcus pneumoniae* (pneumococcus), and *Streptococcus agalactiae* (group B streptococcus (GBS)) [5].

**Citation:** Alderson, M.R.; Welsch, J.A.; Regan, K.; Newhouse, L.; Bhat, N.; Marfin, A.A. Vaccines to Prevent Meningitis: Historical Perspectives and Future Directions. *Microorganisms* **2021**, *9*, 771. https://doi.org/10.3390/ microorganisms9040771

Academic Editor: James Stuart

Received: 13 March 2021 Accepted: 2 April 2021 Published: 7 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Vaccines will play an essential role in preventing these diseases and fulfilling the roadmap vision. Effective vaccines exist and have been in use for years against some of these pathogens, and while there have been significant successes, there also remain significant challenges. As recognized in the World Health Organization (WHO) Immunization Agenda 2030, too many children have insufficient access to vaccines, driven in part by high prices for some of the most effective conjugate vaccines—resulting in limited availability in LMICs [6]. Moreover, existing vaccine formulations do not necessarily reflect the disease serogroups and serotypes most prevalent in the highest burden countries. And even in countries where vaccines are accessible, there is no standard approach to vaccination.

To defeat meningitis, it is critical that we advance new and better vaccines that will be affordable and accessible globally. This will not be easy, but past vaccine development efforts offer direction for future ones. From development of the first conjugate vaccine for humans to the groundbreaking Meningitis Vaccine Project (MVP) that effectively eliminated serogroup A meningococcal meningitis in Africa, the vaccine development landscape is rife with important lessons for developing and delivering vaccines to prevent meningitis [7].

Progress against the four key pathogens identified in the roadmap spans the vaccine development and delivery lifecycle. Hib conjugate vaccines (HibCVs), pneumococcal conjugate vaccines (PCVs), and meningococcal conjugate vaccines (NmCVs) have been in use for decades; vaccines against GBS are on the horizon. This article will explore the challenges, successes, and lessons learned through the development and introduction of meningitis vaccines—lessons critical for successful implementation of the roadmap and the strategy to defeat meningitis by 2030.

#### **2. History and Status of Meningitis Vaccines**

Hib, pneumococcus, meningococcus, and GBS are encapsulated bacteria that cause sepsis, meningitis, and other invasive and mucosal diseases [8]. Capsular polysaccharides are important virulence factors and have become the major vaccine target for all four pathogens. HibCVs, PCVs, and NmCVs are highly successful at preventing meningitis and other disease manifestations caused by these organisms. Conjugate vaccines against these bacteria not only protect against disease in multiple age groups, but also confer herd protection via reductions in pharyngeal carriage [9]. GBS is amenable to conjugate vaccine development but development thus far has targeted maternal immunization, given that the greatest disease burden occurs in the first three months of life [10].

HibCV was the prototype for targeting capsular polysaccharides. Polysaccharidealone vaccines (purified Hib polysaccharide) were certified for use in the US in 1985 but suffered from an inability to elicit immunological memory, poor persistence of immunity, and poor immunogenicity in children under 2 years of age. Covalently coupling the polysaccharide to a protein carrier transformed the vaccine into T-dependent antigens and as such elicited strong immune responses, immunological memory, and immune responses in infants. The first approved HibCV in the US was manufactured using polyribosylribitol phosphate (PRP) conjugated to diphtheria toxoid (DT), though it was eventually replaced by more effective vaccines using meningococcal outer membrane protein (OMP), crossreactive material 197 (CRM197), or tetanus toxoid (TT) carriers. HibCVs are usually used in combination with other pediatric vaccines including tetanus, diphtheria, hepatitis B, and pertussis (Table 1). Importantly, many developing country vaccine manufacturers (DCVMs) have licensed and secured WHO prequalification for low-cost Hib-containing vaccines, resulting in wide-spread introduction globally. The development, licensure, and introduction of HibCVs paved the way for other conjugate vaccines.

The prototypic pneumococcal vaccine was also polysaccharide-based, covering 23 serotypes, and was licensed in 1983 primarily for use in high-risk adults [11]. The first PCV (Prevnar®, PCV7) was licensed in the US in 2000 and was designed to protect against the seven most prevalent invasive disease serotypes in the US and Europe. PCV7 did not, however, protect against the serotypes responsible for considerable disease in LMICs, such as serotypes 1 and 5. Additionally, the introduction of PCV7 led to the

emergence of non-vaccine serotypes, a phenomenon referred to as serotype replacement. This experience prompted the development of 10- and 13-valent PCVs that offer broader coverage (Table 1). Next generation PCVs that extend coverage up to 24 serotypes are currently in mid- to late-stage development.

The first licensed meningococcal vaccines were also polysaccharide based. More recently, NmCVs containing various combinations of serotypes A, C, W, and Y have been licensed and introduced (Table 1). There are two licensed vaccines for serotype B, both protein-based, though neither is WHO prequalified. Vaccines are in development that combine either serotypes A, C, W, X, and Y or serotypes A, B, C, W, and Y [12,13].

There are currently no licensed GBS vaccines, but several candidates are in early- to mid-stage clinical assessment, including a hexavalent version formulated with serotypes Ia, Ib, II, III, IV, and V undergoing Phase 1/2 clinical study [14]. A protein based GBS vaccine has advanced into multiple clinical studies and has demonstrated encouraging safety and immunogenicity data [15,16].

Conjugate vaccines are not without limitations, though; limited serotype coverage and serotype replacement have resulted in the need to make higher valency vaccines for pneumococcus and meningococcus. This, in turn, contributes to manufacturing complexity and difficulty in ensuring affordability for LMICs. Protein-based vaccines are a possible alternative for meningococcus and GBS, but the vaccine against serogroup B is the only protein-based vaccine (OMP/outer membrane vesicle [OMV]) currently licensed [17–22]. This approach is not needed for Hib as the current conjugate vaccines are effective, and it has been difficult to develop a protein-based vaccine for pneumococcus.


**Table 1.** World Health Organization (WHO) prequalified Hib, pneumococcal, and meningococcal conjugate vaccines [23].

#### **3. Early-Stage Meningitis Vaccine Development**

The development of monovalent meningitis vaccines paved the way for newer licensed multivalent vaccines that have broader coverage, including the multivalent GBS vaccines currently under development. Multivalent conjugate vaccines (PCVs, NmCVs, and GBS conjugate vaccines) must be fit for purpose and are highly complex products from a manufacturing perspective—and, as such, are challenging from development and costeffectiveness perspectives.

#### *3.1. Considerations*

The Target Product Profile or Preferred Product Characteristics are critical to guide early strategic decisions for all stages of meningitis vaccine development, from drug substance formulation to presentation, preclinical through clinical studies, product licensure, and introduction (Table 2). Additionally, WHO publishes technical report series (TRS) documents that provide guidance for assuring the quality, safety, and efficacy of

vaccines—including for HibCVs, PCVs, and serogroup A and C NmCVs. The TRS includes recommendations on vaccine manufacturing, nonclinical evaluation, clinical evaluation, and for national regulatory authorities.

**Table 2.** Components of a meningitis vaccine Target Product Profile [24–27].


a serum bactericidal activity, <sup>b</sup> opsonophagocytic killing.

#### *3.2. Manufacturing*

Chemistry, Manufacturing and Controls (CMC) are well described for conjugate vaccines. Many vaccine attributes must be considered during development, including selection of the serotypes and carrier protein conjugation technology, formulation, presentation (Table 3), LMIC needs, and the cost of goods sold (COGS) [28]. The considerations to plan for include saccharide antigen, carrier protein, preservation of immunogenic epitopes, conjugation chemistry, stability of both drug substance and drug product, formulation, consistency of quality, analytics, preclinical models, and commercially viable manufacturing process. Due to the safety considerations for vaccines used in healthy humans, carrier

protein choice is currently limited to CRM197, TT, OMP, DT, and *H. influenzae* protein D, each of which has nuances with antibody avidity and quantity of antibodies elicited.


**Table 3.** CMC general considerations [28–31].

Other considerations include the optimal methodologies for the fermentation and purification of polysaccharides and whether they should be native or size-reduced for conjugation. Several technologies have been used to conjugate the polysaccharides to the proteins in currently licensed meningitis vaccines, the most common of which involve reductive amination or cyanylation chemistry [28]. Newer technologies under development are designed to increase conjugation efficiency, simplify the manufacturing processes, and better preserve immunological epitopes on both the saccharide and protein components [32–34]. The use of an adjuvant is an important consideration and is often driven by either licensed vaccines or clinical assessment, as preclinical models are not good indicators of adjuvant benefits on immunogenicity.

#### *3.3. Nonclinical Assessment*

Evaluating meningitis conjugate vaccines in animal models provides an initial assessment prior to clinical evaluation. For licensed vaccines such as Hib, NmCVs, and PCVs, however, demonstrating protection against disease in a preclinical model is not required and assessment focuses on immunogenicity. Preclinical animal models usually differentiate between the antibody responses of the formulations being tested [28]. Cost, availability, study duration, cross-reactivity, and applicability to humans contribute to animal model selection, though ultimately the choice relies on published work on similar vaccines and compares the responses of the candidate vaccine to a licensed vaccine for the serotypes they have in common. In vivo experiments may not predict the human response but are the best way to distinguish between vaccine formulations. In vitro assays to measure antibody responses in the animal models are, ideally, identical to the assays used in human antibody evaluation. Non-human primates are sometimes used to evaluate immunogenicity for advanced candidates; however, they may not predict immune responses in humans [35,36]. Preclinical animal immunogenicity assessments and toxicology study data (to indicate safety if the product does not elicit a toxic response) are required by regulatory authorities prior to the first in-human clinical study.

If the vaccine is intended for maternal immunization, as is the case for GBS vaccines currently in development, or may be used in a campaign setting that includes pregnant women, as is the case for meningococcal vaccines, a developmental and reproductive toxicology study is required to understand the impact of the vaccine or vaccine candidate on fertility and developmental toxicity. Pre- and post-natal development studies are also necessary to understand the full spectrum of potential reproductive impacts.

#### *3.4. Importance of Functional Assays*

Preclinical and clinical measurements of immune responses to Hib, pneumococcus, meningococcus, and GBS conjugate vaccines have focused on binding and/or functional assays. Binding assays (typically enzyme linked immunosorbent assays) are simple, can be multiplexed, and are highly quantitative in nature. Additionally, it is critical to measure functional antibody responses, whether serum bactericidal activity (SBA) titers for NmCVs, or opsonophagocytic killing assay (OPK/OPA) titers for PCVs and GBS conjugate vaccines. Both SBA and OPK assays demonstrate the ability of vaccine-elicited antibodies to kill live bacteria and are considered to correlate better with clinical efficacy than IgG binding assays.

The use of standardized assays and reagents for both pre-clinical and clinical trial assessment is essential for comparing data between trials, establishing a correlate of protection, and understanding results in the absence of a comparator vaccine. Standardized assays and reagents exist for HibCVs, PCVs, and NmCVs and are in development for GBS conjugate vaccines [37–39].

#### *3.5. Phase 1 Clinical Trials*

Phase 1 trials obtain initial safety, reactogenicity, and immunogenicity data in healthy adults. When licensed vaccines exist, such as for HibCVs, PCVs, and NmCVs, the candidate vaccine is measured against a licensed one. Phase 1 studies may provide initial assessment of different dose levels and formulations both with and without adjuvant, though for HibCVs, PCVs and NmCVs these parameters are becoming well defined with multiple licensed products (Table 1). Notably, for conjugate vaccines, aluminum adjuvants are sometimes incorporated for vaccine stabilization rather than to enhance immune responses. Phase 1 trials are usually small (<100 subjects) so the dose range and adjuvant must be definitively assessed in a Phase 2 trial.

#### *3.6. Phase 2 Clinical Trials*

Phase 2 trials assess the dose selection, adjuvant need, safety, and antibody response to a licensed vaccine (when available) in a larger number of subjects in the target age group. This ensures sufficient statistical power to determine whether the vaccine is promising enough to advance to the next phase of clinical study. For GBS vaccines in development, immunogenicity will be assessed in pregnant women, in cord blood, and in the newborns to determine whether there is adequate transplacental transfer of antibodies and how well they persist.

#### **4. Late-Stage Clinical Development**

HibCVs, PCVs and NmCVs have followed distinct scientific and regulatory pathways in the late stages of their clinical development. However, their licensure strategies have certain aspects in common, based on similarities shared across the three targets, including the type of pathogen, the vaccine platform, and the clinical outcomes targeted. For instance, experience with conjugate vaccine technology allows developers to make initial assumptions regarding dose range and schedules for early clinical development and likely methods for immunological assessment. Similarly, all of these pathogens exhibit a wide spectrum of clinical disease, ranging from asymptomatic carriage to invasive disease, including sepsis and meningitis. Protection against these more severe conditions formed the basis for initial licensure of the early vaccine candidates—but the rare occurrence of these conditions in the population has had similar implications for subsequent vaccine development.

This last consideration has been one of the more consequential factors in shaping latestage development of recent Hib, meningococcal, and pneumococcal vaccines. Licensure of the earliest conjugate vaccines was based on clinical efficacy trials against invasive bacterial disease outcomes, including meningitis, whose relatively low incidence required tens of thousands of participants. For instance, the efficacy of HibCVs was initially established through several randomized placebo-controlled clinical trials conducted in the late 1980s and early 1990s with invasive disease as the primary endpoint [40,41]. Conducted in both high-resource (California, UK, Finland) and lower-resource (Chile, The Gambia, US Alaskan Natives and Navajo) settings, these trials established the clinical efficacy of PRP conjugate vaccines based on four different protein carriers [41]. Having established the presence of safe and efficacious vaccines to protect against invasive Hib disease, it was considered unethical to conduct subsequent placebo-controlled efficacy trials that would leave a subset of participating infants unprotected. However, conducting a comparative efficacy trial between a new and an established vaccine would have been prohibitively large, given the low incidence of vaccine failures likely to occur in either arm. Therefore, later trials of HibCVs, either as new products, newer formulations (such as in combination vaccines), or in alternate schedules have relied on immunologic outcomes (anti-PRP serum IgG levels) for licensure.

In the case of meningococcal vaccines, the low incidence and sporadic epidemiology of disease in industrialized countries pushed this concept even further. The clinical efficacy of meningococcal vaccination was initially established with polysaccharide A and A/C vaccines more than 40 years ago. Effectiveness was demonstrated in closed populations of high-risk adults, demonstrating the vaccines' utility in controlling outbreaks [42–44]. Later, when the UK became the first country to introduce NmCV (against serogroup C) in 1999, licensure was not granted on the basis of clinical efficacy, but rather on the demonstration of adequate immunogenicity [45]. The licensure of all subsequent NmCVs has been granted based on immunogenicity relative to an accepted surrogate of protection, with later demonstration of protection against clinical disease achieved following broader use [45]. Notably, this approach was used for vaccines containing additional meningococcal serogroups, including W and Y, despite having no studies linking specific antibody levels to clinical protection. Licensure was nevertheless granted based on the assumption that these conjugate vaccines would behave similarly, given the infeasibility of conducting efficacy trials for these serogroups. In contrast, serogroup B meningococcal vaccines were relatively delayed, as similarities between group B capsular polysaccharides and host epitopes prevented use of the polysaccharide conjugate platform. Instead, vaccines based on protein subunits were developed. Nevertheless, licensure was still granted based on the induction of serum bactericidal antibody, an immunological outcome, with a postmarketing commitment to demonstrate clinical benefit [46].

MenAfriVac®, a monovalent group A meningococcal conjugate vaccine (NmCV-A) developed through MVP (a partnership between WHO, PATH, and SIIPL), has been deployed through two strategies, first a series of national mass vaccination campaigns throughout the African meningitis belt covering a broad age group (1 to 29 years of age), followed by incorporation of the vaccine into the routine infant immunization (EPI) schedules of the affected countries. To accomplish this, the vaccine's licensure strategy involved two stages. Initial licensure and WHO prequalification was based on a series of clinical trials in individuals 1 to 34 years of age demonstrating the safety and immunologic superiority of a full dose (10 µg PsA-TT) to a group A-containing polysaccharide vaccine [47–49], thus allowing the start of mass campaigns. Subsequently, an indication for a 5 µg single-dose regimen in children 3 to 24 months of age was achieved based on demonstration of immunologic non-inferiority to the 10 µg dose in two trials in infants [50].

More recently, the licensure strategy for a new pentavalent NmCV containing serogroups A, C, W, X and Y has followed a parallel path relying on demonstration of immunologic non-inferiority to established quadrivalent conjugate vaccines. Two ongoing Phase 3 trials, one in 2- to 29-year-old individuals in Mali and The Gambia [51–54], and another in adult and elderly individuals in India, both using Menactra as the comparator, are intended to gain licensure for use in mass campaigns and travelers. Another Phase 3 trial is planned for younger infants and toddlers in Mali to allow use in routine infant immunization. This trial will use Nimenrix as the comparator because, unlike Menactra, Nimenrix is licensed for use as a single dose down to 6 months of age.

Finally, for PCVs, the clinical efficacy of initial 7- and 9-valent vaccines against invasive pneumococcal disease (IPD) was established in four large-scale trials conducted in the late 1990s and early 2000s in both high- and low-income settings [52–55]. The observed efficacy in these studies ranged between 76.8 and 97.4 percent for IPD caused by serotypes contained in the vaccine, with higher efficacy seen in more industrialized settings. A later 10-valent vaccine was initially licensed using immune correlates of protection, with effectiveness subsequently established through two randomized double-blind controlled trials in the late 2000s in Finland (in a cluster-randomized design) and Latin America [56,57]. Vaccine development expanding the initial 7-valent vaccine to a 13-valent formulation and comparisons for different immunization schedules for the PCV13 and PCV10 vaccines subsequently relied on immunologic endpoints [58], as did the development and licensure of a newer 10-valent PCV in India [59].

In the evaluations of efficacy noted above, the clinical endpoints were chosen by balancing a need for the specificity and clinical relevance of laboratory-confirmed severe disease with the practicality of measuring relatively uncommon outcomes in a population. By necessity, meningococcal vaccine trials were limited to evaluation of protection against meningitis in the case of polysaccharide vaccines, and immunologic outcomes for conjugate vaccines. For Hib and pneumococcal vaccines, initial clinical trials assessed efficacy against all invasive disease, including bacteremia, bacteremic pneumonia, and meningitis, typically in such low numbers that these presentations were not differentiated in their reporting. The effectiveness of these vaccines in the prevention of meningitis specifically has been demonstrated in multiple later studies following implementation in various countries.

An important consideration for the overall clinical development plan as specified in WHO TRSs is the incorporation of antibody persistence studies to inform vaccine implementation strategies and schedules that may potentially require booster doses. For example, in the case of NmCV-A, antibody persistence analysis was used to estimate that protective immune responses would persist for at least 10 years following immunization [60]. As mentioned earlier, a critical feature of conjugate vaccines is their ability to invoke herd protection. The ability to prevent acquisition of carriage, an indicator for herd immunity, can be assessed in Phase 3 trials or in post-licensure studies.

#### *Immunological Correlates of Protection*

Despite the similarities among these vaccines, there are also aspects that were unique or assumed special prominence for each pathogen. Ideally, the reliance on immunologic endpoints for regulatory or policy decision-making should be based on a true immune correlate of protection. However, such a correlate is not always available. In the case of Hib vaccines, two immunologic correlates were established. Based on initial experimental data, an anti-PRP IgG level of 0.15 µg/mL indicated ongoing protection from invasive Hib disease, while field studies indicated that a peak post-vaccination response level of 1.0 µg/mL was needed for long-term protection (Table 4). As a result, both thresholds were ultimately considered for regulatory approval and post-licensure evaluation of new vaccines and schedules [41]. The presence of immune correlates proved to be particularly useful for assessing the adequacy of different infant schedules, especially those that were accelerated (2, 3, and 4 months) or early (6 weeks) [61]. Immune correlates were also instrumental in evaluating potential immunological interference between Hib and other childhood vaccines. For instance, a resurgence of Hib cases in the UK in the early 2000s was attributed to interference between Hib vaccine and the recently adopted acellular pertussis vaccines. Evaluation of antibody levels in cohorts receiving both vaccines revealed lower anti-PRP IgG levels in later toddler years compared to prior cohorts, prompting

the addition of a Hib booster dose at school entry [41]. Benchmarking antibody levels to short- and long-term thresholds became prominent again in subsequent years, as more complex combination infant vaccines were developed. Immunologic evaluation of these formulations revealed not only interactions between Hib, other antigens, and their carrier proteins, but also incompatibilities among adjuvants [62]; nevertheless, multiple Hibcontaining pentavalent and hexavalent vaccines have ultimately come to market.

**Table 4.** Immunological correlates of protection for Hib, meningococcal, pneumococcal, and group B streptococcus (GBS) vaccines [37–39,63,64].


<sup>a</sup> Human complement serum bactericidal activity, <sup>b</sup> rabbit complement serum bactericidal activity.

> For meningococcal vaccines, maintaining adequate levels of circulating serum antibody is considered most important, as the onset of severe clinical disease upon exposure is too rapid to allow time for generation of an immune memory recall response [45,65]. Therefore, assuring serum antibody persistence has been an important feature of meningococcal vaccine evaluation. The immunological evaluation of NmCVs has focused on functional immune responses, namely SBAs. In comparison with HibCVs or PCVs, NmCVs require only one or two doses for durable protection, which may be partly due to the older ages at which they are generally given [42].

> In the case of pneumococcal vaccines, a meta-analysis of humoral responses using pooled results from three of the original efficacy trials was conducted, allowing the scientific community to establish a non-inferiority threshold of 0.35 µg/mL capsular polysaccharide antibody against each serotype for the evaluation of newer PCVs. While this threshold is not serotype-specific, and true correlates of protection for specific serotypes may ultimately vary [66], this benchmark has allowed the development of later PCV formulations with higher valency based on immunologic outcomes [67].

> Among the major causes of bacterial meningitis, GBS has remained a challenge for vaccine developers. Notably, the early age at which this pathogen acts indicates the best approach to vaccination would be administration during pregnancy to transfer protection to the infant through maternal antibody. While regulatory guidance has been proposed for this novel indication, no "maternal" vaccine has yet been licensed for this purpose, and several uncertainties remain, particularly regarding late-stage development [68].

> Several GBS vaccine candidates are currently in Phase 2 development, and progression to licensure will follow one of two main pathways: efficacy trials demonstrating protection against specific clinical outcomes, or immunogenicity trials that target immunologic correlates of protection. Each developmental program has its own strengths and challenges.

> Demonstration of clinical efficacy through randomized controlled trials would be the most direct route to licensure. As with the other pathogens discussed in this review, GBS is associated with a wide spectrum of disease, with laboratory-confirmed invasive disease (early- and late-onset meningitis being particularly prominent) the most likely clinical endpoint, given its specificity and relevance to clinical care and public health [69]. Similarly, this outcome is relatively uncommon, particularly if focused on neonatal disease alone, and thus would require relatively large clinical trials to establish efficacy. For this

reason, composite endpoints that incorporate additional important laboratory-confirmed fetal and obstetric outcomes, such as stillbirth and maternal sepsis, have been proposed to reduce study size [69]. Neonatal invasive GBS disease occurs at a rate of 1 to 3 per 1000 live births in many geographies, and in those areas, best practices associated with prenatal and perinatal care and intrapartum antibiotic prophylaxis can reduce this rate to 0.5–1.0 per 1000 live births. Given these incidence rates, an efficacy trial could require between 30,000 and 1.8 million mother-infant pairs [39]. While some infant vaccine trials have included up to 70,000 participants, evaluating maternal immunization would also be more resource-intensive on a per-subject basis by comparison. Other clinical endpoints could be considered, including maternal urinary tract infection and colonization, but are unlikely to be included, as they do not directly correlate with invasive disease and otherwise do not pose a significant clinical or public health burden.

Given the impracticality of conducting clinical trials of this size, developers must consider pathways that utilize an immunologic endpoint. However, without prior vaccine efficacy trials, a correlate of protection must be established through sero-epidemiological studies that examine naturally occurring disease. Since the 1970s, serotype-specific maternal capsular antibodies were known to correlate with a reduced risk of invasive GBS disease. However, differences in methodology prevented the establishment of protective thresholds. More recently, larger-scale studies have been initiated in South Africa and the UK using a standardized approach to more definitively establish these associations. These efforts, along with data from animal models, will hopefully produce suitable criteria for pivotal Phase 3 vaccine trials based on immunologic endpoints [39].

Several aspects of the immune response to vaccination are particularly relevant to the maternal immunization model. Since fetal and infant protection is primarily generated through passive transfer of IgG antibody through the placenta during gestation, achieving a high peak maternal serum IgG antibody response to maximize infant levels by the time of birth is a key objective. Therefore, longevity of the immune response, generation of durable immune memory, and even protection of the mother, are secondary—although important goals. In addition, since this model involves adult vaccine recipients who likely have been previously exposed to GBS, a single vaccine dose to boost pre-existing memory responses is likely to be sufficient. Finally, either before or after licensure, vaccine manufacturers will need to demonstrate a lack of immune interference between their GBS vaccine and other vaccines currently given to pregnant women, including tetanus, pertussis, and influenza, or under development, such as respiratory syncytial virus. Moreover, compatibility studies among these vaccines could allow their incorporation into a combination maternal vaccine, which could greatly improve affordability and access.

#### **5. Accelerating Vaccine Introduction to Prevent Meningitis**

Introducing HibCVs, NmCVs, and PCVs and optimizing their coverage in affected populations has been critical for reducing meningitis morbidity and mortality in the last 20 years. However, the availability of effective and safe vaccines alone is insufficient to increase LMIC uptake. Despite the success of these vaccines in high-income countries, overcoming barriers to introduction and sustaining vaccine delivery in LMICs—where the greatest meningitis burden persists—remains a major challenge to global meningitis control [70].

#### *5.1. HibCV: Developing New Approaches to Increase Meningitis Vaccine Uptake*

In 2000, 13 years after HibCV was licensed, Hib still caused 8 million meningitis cases and about 400,000 deaths in u5 children in LMICs [71,72]. No Asian countries and only one sub-Saharan African country had introduced HibCV. By 2008, 70 percent of WHO members had introduced HibCV; Hib deaths in u5 children were cut in half [73]. Despite this remarkable impact, HibCV uptake remained low in LMICs. New LMIC introduction approaches were needed.

In the late 1990s, public-private partnerships started to develop new policies, strategies, and priorities for vaccine introduction and to financially support HibCV procurement dramatically increasing uptake in LMICs [74]. In 2006, combining HibCV into WHOprequalified quadri-, penta-, and hexavalent vaccines accelerated uptake and contributed to sustain HibCV use in Gavi-eligible countries [75]. These strategies and approaches would be replicated to increase uptake of NmCV-A and PCVs (Table 5). Incorporating similar approaches will lead to successful introduction of GBS vaccines and boost uptake of multivalent NmCVs and higher valency PCVs. In addition, as countries become ineligible for Gavi support, three approaches (vaccine procurement groups; lower-price, high-quality, WHO-prequalified vaccines from DCVMs; and combination vaccines) will allow middleincome countries (MICs) to continue to introduce new meningitis vaccines.

**Table 5.** Introducing polysaccharide conjugate vaccines to prevent meningitis due to Hib, meningococcus, and pneumococcus in LMICs [1,29,58,76].


<sup>1</sup> RI–routine administration within EPI schedule; <sup>2</sup> SIA–supplementary immunization activity (mass campaign).

#### *5.2. Defining Meningitis Burden to Justify Vaccine Introduction*

Poor understanding of the meningitis burden is a major hurdle that requires considerable time, effort, and resources to overcome. Laboratory-based meningitis surveillance to identify at-risk populations, detect outbreaks, and define the potential impact of meningitis control shows the public health value of these vaccines. Meningitis surveillance is

challenging and requires significant technical capacity to culture blood/cerebrospinal fluid and identify serogroups/serotypes of meningitis pathogens. However, without evidence that a specific pathogen is a public health problem, countries will be slow to commit to vaccine introduction.

The highest meningococcal disease burden is in the 26 countries of the African meningitis belt. From 1970 through 2010, recurring explosive serogroup A meningococcal (Nm-A) meningitis epidemics in sub-Saharan Africa increased in frequency and magnitude [77]. In 1992, WHO country offices, UNICEF, and non-governmental organizations (NGOs) began submitting outbreak data to WHO to justify release of stockpiled meningococcal vaccines. Because these periodic Nm-A epidemics largely defined the meningitis burden, these surveillance data-containing requests yielded data to support NmCV-A introduction. In 2014, MenAfriNet, a case-based meningitis surveillance system, began monitoring meningitis outbreaks, which will be important in future decisions to introduce a multivalent NmCV.

Hib meningitis results from endemic transmission. Because u5 children accounted for 90 percent of Hib meningitis cases and parents often seek hospital care for ill children, hospital-based surveillance of 0- to 59-month-old children was used to define disease burden [78]. Because only one serotype caused disease, the laboratory demands were much less than those for meningococcus and pneumococcus. This surveillance resulted in high quality burden data in Africa, where Hib was well recognized as a meningitis pathogen. However, most Asian countries did not show sufficient burden to justify HibCV introduction; that changed when a landmark vaccine probe study in Indonesia showed that Hib accounted for a large portion of meningitis and pneumonia not found in routine surveillance [79,80]. Subsequent vaccine probe studies showed significant reduction of meningitis was possible through vaccination and greatly accelerated HibCV uptake [81].

In LMICs that successfully introduce HibCV and NmCV-A, pneumococcus becomes the most common cause of meningitis in all age groups—yet, defining pneumococcal meningitis burden can be difficult [82]. Because of the disease's endemic transmission, broad age distribution, and multiple serotypes, defining the best surveillance is a challenge. As a result, pneumococcal meningitis burden data are often underestimated and insufficient alone to justify PCV introduction. It is better justified by the much higher burden of community-acquired pneumococcal pneumonia and PCVs' cost-effectiveness in preventing pneumonia. Compared to the pneumonia burden, except for periodic serotype 1 pneumococcus meningitis epidemics in Africa, pneumococcal meningitis surveillance has played a small role in accelerating PCV uptake.

#### *5.3. Highly Directive Policies from Global Public Health Authorities*

Global public health authorities have highly influential voices that can be used to advance vaccine introduction. Because of the challenges in diagnosing Hib meningitis, its high treatment costs, high mortality, and the severe neurologic impacts in survivors, HibCV was clearly cost-effective in most LMICs [83]. Yet, decisions to introduce HibCV lagged for many reasons, including inadequate in-country technical capacity to assess the value and potential impact of vaccines [84]. In 2006, WHO overcame this barrier when it universally recommended the implementation of Hib vaccination in all infant immunization programs worldwide without accumulating more surveillance data [76]. Such a statement was possible because the global risk of Hib meningitis was roughly the same for all children, the potential impact of vaccination was similar globally, and HibCV had an excellent safety and efficacy profile [81]. This statement was critical in the LMIC decisions to introduce HibCV [85].

#### *5.4. Structuring Vaccination Strategies for Success*

Successful vaccine introduction strategies can have a high impact in a short time and can motivate decision-makers in other countries to introduce new vaccines. Successful introduction strategies begin by clearly defining target populations. Several factors come

into play when defining this target, such as peak-incidence age, opportunities to vaccinate, persistence of immunity, and need for a booster vaccine. HibCV and NmCV-A introduction showed that well-targeted strategies can quickly achieve near elimination of disease and that successful introduction in early-adopting countries led to decisions to introduce in other countries.

The vaccine introduction strategy for HibCV was relatively straight-forward. Because WHO recommended vaccination for all children, there was no need to develop surveillance systems to identify at-risk countries, districts, or populations or to develop subnational introduction plans. Because peak-incidence age was in the first two years of life, vaccine had to be delivered to infants, and national immunization programs had well-developed opportunities to vaccinate 6-, 10-, and 14-week-old infants.

In contrast, NmCV-A does not universally benefit all children because Nm-A meningitis is not equally distributed globally [86]. Although epidemics were reported globally until the 1940s, Nm-A meningitis outbreaks had become restricted to African meningitis belt countries. Moreover, Nm-A meningitis was not equally distributed within countries. Consequently, highly granular disease surveillance data was needed to allow subnational NmCV-A introduction. Whereas the goal of HibCV was to prevent endemic disease, the goal of NmCV-A was to prevent periodic epidemics driven by meningococcal nasal carriage in children 10 to 14 years of age. The decision to conduct introduction campaigns in 1- to 29-year-olds was a strategy that stopped outbreaks and prevented meningitis in young adolescents. However, this strategy had to be balanced by the fact that routine vaccine delivery to school-aged children is not well-developed. Currently, NmCV-A vaccination occurs in children younger than 2 years of age. Whether bactericidal antibodies persist beyond 10–12 years at a level that will later suppress nasal carriage and prevent Nm-A epidemics is unknown.

#### *5.5. Public-Private Initiatives to Provide Vaccine and Improve Vaccination Practices*

Public-private partnerships have been critical to HibCV, NmCV-A, and PCV introduction by providing support for vaccine delivery, procurement, and technical assistance to low-income countries (LICs). These partnerships will remain critical for new vaccine introduction going forward.

In 1998, the William H. Gates Foundation donated \$100 million to establish the Children's Vaccine Initiative (CVI) to improve vaccine delivery to LICs [87]. Prior, many LICs used funds intended to support delivery to buy HibCV. To reverse this, CVI proposed funding to make vaccines more available and to improve the quality of vaccine delivery, rather than to procure vaccines. Through partnerships with WHO, UNICEF, PATH, and other international NGOs, CVI funded guideline development, vaccination worker training, model immunization programs, cost-effectiveness studies, and advocacy and communication programs to increase HibCV acceptance.

In 2001, the Bill & Melinda Gates Foundation funded MVP, which successfully developed, tested, licensed, WHO-prequalified, and introduced MenAfriVac®, an affordable NmCV-A. The keys to MVP's success included developing strong public private partnerships [7,88,89]; engaging SIIPL, a DCVM, to develop a low-cost, high-quality NmCV-A; providing technical assistance to SIIPL to acquire WHO prequalification; conducting clinical trials in Africa alongside African researchers; and supporting operational costs of introduction. Since 2010, more than 340 million Africans have been vaccinated with NmCV-A and Nm-A meningitis has been eliminated from this region.

In 2002, Gavi and key partners, including Johns Hopkins University, established the Pneumococcal Vaccines Accelerated Development and Introduction Plan (PneumoADIP) to increase uptake of PCVs in Gavi-eligible countries [90]. The keys to PneumoADIP's success included supporting PCV procurement and the operational cost of vaccine introduction, standardizing pneumococcal disease surveillance, developing advocacy and education activities to inform country decision-makers within national immunization programs regarding PCV and HibCV introduction, and providing technical assistance for

vaccine introduction campaigns and the transition to routine immunization. As a result of PneumoADIP, between 2000 and 2018, 59 of 73 Gavi-eligible countries introduced PCV.

In 2005, Gavi's Hib Initiative (GHI), a consortium of WHO, Johns Hopkins University, London School of Hygiene and Tropical Medicine, and the Centers for Disease Control and Prevention, was funded to help Gavi-eligible countries make evidence-based decisions regarding HibCV introduction [85]. Through these CVI and GHI activities, the number of LICs introducing HibCV increased from 13 in 2004 to 66 in 2008 [91]. Currently, all Gavi-eligible countries use HibCV-containing vaccines.

#### *5.6. Developing Innovative Vaccine Financing Options*

Defining the cost-effectiveness of HibCV, NmCV-A, and PCVs has been important for new vaccine decision-makers and has accelerated the uptake of these vaccines in LMICs. Studies have shown that HibCV is cost saving or highly cost-effective in essentially all settings. Cost-effectiveness has further increased due to the recent decline in HibCV prices, integration of HibCV into quadri-, penta- and hexavalent combination vaccines, and data showing the loss of productivity in meningitis survivors [92]. Similarly, compared with a reactive vaccination strategy, prevention strategies using NmCV-A were shown to be significantly cost saving in Burkina Faso [93]. Such analyses will be important for decisionmakers considering whether the higher price of the next generation of meningococcal or pneumococcal vaccines or the price of new vaccines to prevent GBS meningitis are justified by their benefits [94].

Prior to 2000, vaccine cost was often the greatest barrier to meningitis vaccine introduction. Since then, LICs have greatly benefited from Gavi's vaccine investment strategy and procurement of meningitis vaccines through The Vaccine Fund [95,96]. Unfortunately, many MICs that procure their own vaccines face financial challenges to introduction. In addition, LIC decision-makers are more widely considering the long-term costs of vaccination, not just the initial introduction costs.

To address this, in 2009, the Advance Market Commitment (AMC) for pneumococcal vaccines was launched. In the AMC, donors commit funds to guarantee the price of vaccines once they have been developed. In exchange, manufacturers make a legally binding commitment to provide the vaccines at a price affordable to LICs [97]. Although the AMC has been recognized as a valuable way to make effective and affordable pneumococcal vaccines available, it has also been criticized for not encouraging innovation, discouraging competition from new market entrants, and raising vaccine costs [98,99].

Another financing option is multi-country procurement groups, such as the Pan American Health Organization (PAHO) Revolving Fund. Since 1977, this fund has pooled the resources of 41 mostly middle-income Latin American countries to procure vaccines at a lower cost through consolidated ordering [100]. Currently, the fund is used to procure HibCV-containing vaccines and PCV, which has resulted in sustained use of these vaccines throughout Central and South America.

Finally, for countries that purchase their own vaccines, the availability of lower-cost, high-quality, WHO-prequalified vaccines produced by DCVMs has been an important alternative to vaccines produced by multi-national vaccine manufacturers.

#### *5.7. Implications for New Meningitis Vaccines*

The lessons learned from HibCV, NmCV-A, and PCV introduction will likely be applied to the introduction of new meningitis vaccines. For example, there has been development and successful use of several other meningococcal vaccines, including monovalent meningococcal vaccines against serogroups C and B and multivalent NmCVs against serogroups A, C, W, X and Y. WHO has stated the decision to use other meningococcal vaccines or to replace NmCV-A with a multivalent NmCV will depend on the locally prevalent meningococcal serogroup(s), identification of the best target group for vaccination, and opportunities to vaccinate within national immunization programs [42]. This underscores

the importance of meningitis surveillance. Discussion is ongoing regarding the use of new multivalent NmCVs being developed by DCVMs.

New vaccines are being developed against GBS to prevent meningitis in neonates and young infants [27]. Some of the approaches described above will likely be used to increase uptake (e.g., combination vaccines, support for vaccine procurement) [27]. However, because the goal of a GBS vaccine is to prevent invasive disease in neonates and infants, the target group for vaccination is pregnant women. Given the challenges of accessing obstetric care in LICs and the lower emphasis on vaccination in antenatal care clinics compared to EPI clinics, new approaches will be needed with special attention to advocacy and communication and antenatal healthcare worker training to introduce a GBS vaccine.

#### **6. Conclusions and Future Directions**

The development and global introduction of low-cost vaccines to prevent Hib and pneumococcus has had a significant impact on meningitis and other disease manifestations caused by these pathogens. DCVMs have become the major suppliers of affordable Hib combination vaccines and the recent licensure and WHO prequalification of a 10-valent PCV by SIIPL, in partnership with PATH, is poised to increase availability of low-cost PCVs for LMICs, notably in those countries that have not introduced PCVs into their routine immunization programs. Like with Hib vaccines, it is anticipated other DCVMs will license PCVs and increase the global supply of affordable vaccines. Despite the considerable success in reducing the burden of pneumococcal disease globally, serotype replacement and emergence has resulted in significant residual disease burden. Higher valency (15–24 serotypes) PCVs are in development, though there are considerable manufacturing and licensing challenges for such vaccines and LMIC affordability is uncertain.

Meningococcal vaccines present a dichotomy: Quadrivalent NmCV-ACWY and meningococcal serogroup B protein vaccines manufactured by multinational vaccine manufacturers are cost prohibitive for widespread use in LMICs, while a low-cost NmCV-A that has had incredible impact in the African meningitis belt has limited utility in other parts of the world. The development and licensure of low-cost NmCV-ACWY(X) and meningococcal B vaccines has the potential for broad appeal and to greatly reduce the burden of meningococcal meningitis globally.

In addition to reducing the per dose cost of meningitis vaccines, strategies to increase cost-effectiveness by minimizing the number of doses administered are in development. For example, WHO currently recommends a single dose of NmCV-A at 9 to 18 months of age for routine immunization and studies to assess whether a 2-dose schedule (1 + 1) instead of 3-dose schedule for PCVs may be sufficient to maintain adequate herd immunity are underway [101,102].

What about other meningitis pathogens that are potentially vaccine preventable? *Haemophilius influenzae* type A (Hia) causes meningitis in certain regions and populations globally, including indigenous populations in North America and Australia. Development of a Hia vaccine should be technically feasible but a limited market would likely require donor support to incentivize a manufacturer. *Klebsiella pneumoniae* is becoming increasingly recognized as an important cause of sepsis and meningitis in neonates in LMICs and as such could be targeted for maternal vaccine together with GBS. The relatively high number of *K. pneumoniae* capsular serotypes makes this a challenging approach, although targeting a more limited number of O antigens or protein antigens is also being considered [103].

Defeating meningitis is an ambitious undertaking that will require significant time, effort, and resources—particularly when it comes to developing new or improved meningitis vaccines. There are hurdles along the vaccine development and delivery spectrum but well-established vaccines like HibCV, PCV, and NmCV offer lessons for what does and does not work, how to successfully advance products toward market, and how to ensure they reach the populations in need—and where gaps remain that need to be filled. Despite their challenges, vaccines are a public health best-buy and have been critical to the progress we have made against meningitis thus far. Vaccines have saved millions of lives around

the world and new entrants are poised to take that success further to make the vision of defeating meningitis by 2030 a reality.

**Author Contributions:** Conceived and plotted the manuscript, M.R.A., J.A.W., K.R., L.N., N.B. and A.A.M.; drafted the original manuscript, M.R.A., J.A.W., K.R., N.B. and A.A.M.; reviewed, edited, and formatted the manuscript, K.R. and L.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported, in part, by the Bill & Melinda Gates Foundation, Seattle, WA, USA (OPP1116594, OPP1150281, OPP1054296) and the Foreign, Commonwealth, & Development Office, London, UK (PATH MVPP 300341-108). Under the grant conditions of the Foundation, a Creative Commons Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from this submission. The findings and conclusions contained within are those of the authors and do not necessarily reflect positions or policies of the Foundation or the Foreign, Commonwealth, & Development Office.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Acknowledgments:** The authors thank all their global partners in PATH's work on pneumococcal, meningococcal, and group B streptococcal vaccines.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


13. Zahlanie, Y.C.; Hammadi, M.M.; Ghanem, S.T.; Dbaibo, G.S. Review of meningococcal vaccines with updates on immunization in adults. *Hum. Vaccin. Immunother.* **2014**, *10*, 995–1007. [CrossRef]


## *Review* **Invasive Bacterial Infections in Subjects with Genetic and Acquired Susceptibility and Impacts on Recommendations for Vaccination: A Narrative Review**

**Ala-Eddine Deghmane \* and Muhamed-Kheir Taha**

Invasive Bacterial Infections Unit, Institut Pasteur, 28 Rue du Dr. Roux, CEDEX 15, 75724 Paris, France; muhamed-kheir.taha@pasteur.fr

**\*** Correspondence: ala-eddine.deghmane@pasteur.fr; Tel.: +33-1-44-38-95-90

**Abstract:** The WHO recently endorsed an ambitious plan, "Defeating Meningitis by 2030", that aims to control/eradicate invasive bacterial infection epidemics by 2030. Vaccination is one of the pillars of this road map, with the goal to reduce the number of cases and deaths due to *Neisseria meningitidis*, *Streptococcus pneumoniae*, *Haemophilus influenzae* and *Streptococcus agalactiae*. The risk of developing invasive bacterial infections (IBI) due to these bacterial species includes genetic and acquired factors that favor repeated and/or severe invasive infections. We searched the PubMed database to identify host risk factors that increase the susceptibility to these bacterial species. Here, we describe a number of inherited and acquired risk factors associated with increased susceptibility to invasive bacterial infections. The burden of these factors is expected to increase due to the anticipated decrease in cases in the general population upon the implementation of vaccination strategies. Therefore, detection and exploration of these patients are important as vaccination may differ among subjects with these risk factors and specific strategies for vaccination are required. The aim of this narrative review is to provide information about these factors as well as their impact on vaccination against the four bacterial species. Awareness of risk factors for IBI may facilitate early recognition and treatment of the disease. Preventive measures including vaccination, when available, in individuals with increased risk for IBI may prevent and reduce the number of cases.

**Keywords:** susceptibility; invasive bacterial infections; complement; genetic factors; *Neisseria meningitidis*; *Streptococcus pneumoniae*; *Haemophilus influenzae*; *Streptococcus agalactiae*; group B streptococci

#### **1. Introduction**

Invasive bacterial infections (IBI) usually refer to those infections provoked by *Neisseria meningitidis* ((Nm), meningococcus), *Streptococcus pneumoniae* ((Spn), pneumococcus), *Haemophilus influenzae* (Hi) and *Streptococcus agalactiae* (group B *Streptococcus* (GBS)). The major form of these invasive infections is acute bacterial meningitis. However, other clinical forms are also encountered. The term "bacterial meningitis" is frequently used to refer to all invasive infections due to these agents. In 2020, a road map, "Defeating Meningitis by 2030" was endorsed by WHO. This road map includes an ambitious and broad multidisciplinary plan that includes five pillars to control and eradicate invasive bacterial infection epidemics by 2030: (i) diagnosis and treatment; (ii) prevention and epidemic control; (iii) disease surveillance; (iv) support and aftercare for people affected; and (v) advocacy and information. Actions to achieve the specific goal of prevention and epidemic control include the introduction of vaccines against the four causative agents, achieving equal access to these vaccines and maintaining high coverage of targeted population [1].

Risk factors for developing IBI are linked to bacterial factors (virulence factors). Certain genotypes of these bacterial agents have been reported to be more significantly associated to IBI. The virulence traits are frequently associated with growth in the host, evasion of host immunity, persistence in the host and transmission between hosts [2–5]. Next, there

**Citation:** Deghmane, A.-E.; Taha, M.-K. Invasive Bacterial Infections in Subjects with Genetic and Acquired Susceptibility and Impacts on Recommendations for Vaccination: A Narrative Review. *Microorganisms* **2021**, *9*, 467. https://doi.org/ 10.3390/microorganisms9030467

Academic Editor: James Stuart

Received: 15 January 2021 Accepted: 20 February 2021 Published: 24 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

are factors linked to the host that increase its susceptibility to IBI by enhancing acquisition and/or reducing the clearance of bacterial agents. IBI are often due to underlying anatomical or immune disorders, either of which may be inherited or acquired. Improving surveillance and implementation of vaccines will continue to reduce the incidence of IBI in the general population. However, the burden of these infections among subjects with enhanced susceptibility to IBI will increase proportionally. Another factor that also requires analysis is the severity of invasive bacterial infections. Better knowledge of these two facets (susceptibility and severity) of IBI is therefore warranted. Several aspects of these infections require exploring, for instance, little is known about the genotypes of the involved bacterial isolates and whether they differ from bacterial isolates encountered in the general population. Moreover, response to vaccination and vaccine failure in these subjects are less explored than in the general population. The need for special vaccination schedules also requires analysis. In this narrative review, we aim to summarize the genetic and acquired risk factors that increase the susceptibility to and severity of invasive infections related to the four above-mentioned pathogens and to discuss preventive measures under these conditions.

#### **2. Method**

We performed a search of PubMed with the objective of summarizing the inherited and acquired host factors associated with susceptibility of patients to invasive meningococcal, pneumococcal, *Haemophilus influenzae* and group B streptococci disease. The following Mesh terms were used: ((*Neisseria meningitidis*) OR ((*Streptococcus pneumoniae*) OR (*Haemophilus influenzae*) OR (*Streptococcus agalactiae*) OR (group B streptococc\*)) AND (((invasive) AND ((disease\*) OR (infection\*))) OR (bacterial meningitis) OR (meningitis) AND ((genetic) OR (acquired) OR (immunocompromised)\* or (deficien\*) OR (immunodeficient\*) OR (susceptibility) OR (predispose\*) OR (recurrent infection\*)). A built-in PubMed filter was used to limit the search to papers published in English or French up until 31 October 2020. Both authors independently screened titles and abstracts. Studies lacking outcomes of interest were considered not relevant to the aim of our review and were excluded. Relevant publications matching the criteria applied to the search results were identified, and the full text of each was reviewed by both authors separately.

#### **3. Susceptibility to Invasive Meningococcal Infections**

Nm is a human-restricted, Gram-negative encapsulated bacterium that is usually encountered as a member of the nasopharyngeal microbiota, which acts as a carriage. However, a few genotypes (hyper-invasive clonal complexes) are associated with invasiveness of the bloodstream and are responsible for most of the cases of invasive meningococcal disease (IMD). Carriage and hyper-invasive isolates differ genetically and phenotypically. Unlike invasive isolates, carriage isolates are more frequently non-capsulated and do not belong to hyperinvasive genotypes [6]. The incidence of IMD varies according to age, with three peaks: in infants < 1 year of age, in adolescents and young adults and in the elderly. This incidence also varies geographically and the epidemiology of IMD is continuously changing [7,8].

The meningococcal capsule is a polysaccharide, and when present, it determines the serogroup. Twelve serogroups have been described with serogroups A, B, C, W, Y and X being responsible for virtually all cases of IMD [8]. Capsular polysaccharide-based vaccines are available against Nm of serogroups A, C, W and Y, while subcapsular protein-based vaccines are available against Nm of serogroup B. Recommendations exist to use these vaccines in subjects with increased susceptibility to IMD. However, rational support for these recommendations may require clarification.

#### *3.1. Genetic and Acquired Susceptibilitiesy to IMD*

The ability of Nm to invade, to survive and to spread in the bloodstream is linked to its pathogenesis, which is correlated to the complement-dependent clearance of meningococci. Factors that lead to the absence of bactericidal activity in complement-dependent serum increase the susceptibility to IMD [9,10]. These factors can be inherited and/or acquired.

#### 3.1.1. Inherited Factors of Susceptibility to IMD

The three pathways of the complement system (the classical, the lectin and the alternative pathways) are major actors in the innate immune response. Activation of complement is tightly controlled with several regulators. Complement is activated through the early complement components of these three pathways to first form C3 convertases, then, they converge to form the C5 convertase, and subsequently, the membrane attack complex (MAC) through the activation of the late complement components (LCC) (C5 to C9). The MAC ultimately leads to the lysis of the targeted cell. Moreover, complement activation leads to the opsonization of the bacterial surface [11]. These two events (lysis and opsonophagocytosis) are directly responsible for efficient bacterial clearance [12]. For Nm, bactericidal activity (in the absence of blood inflammatory cells) is able to lyse bacteria through the insertion of the MAC at the bacterial surface [9,13]. Deficiencies in these late components of the complement system lead, therefore, to enhanced susceptibility to IMD, which can result in repeated IMD [13–15]. This is particularly the case in subjects with late components of complement deficiencies (LCCD), deficits of properdin deficiency or deficits of factor D deficiency [15,16]. Polymorphism of Factor H (a negative regulator of the complement) is also associated with an increased risk of IMD while deficiencies in the early components (such as C1) were not reported to be specifically associated with increased susceptibility to IMD [17,18]. The incidence of IMD among LCCD patients, in regard to number and proportion, will increase due to the decreasing incidence of IMD in immune-competent subjects upon implementation of vaccination strategies. The incidence of IMD is 1000 to 10,000 times higher among LCCD patients than among the general population [15]. The frequency of hereditary complement deficiencies varies according to their type, age, sex and geographical/ethnic distribution [15]. Terminal complement pathway, properdin and factor D deficiencies seem to lead specifically to an increased susceptibility to IMD. LCCD are the most frequent but seem to be associated with a low fatality rate (1%), and are usually detected in adolescents and young adults [15,19]. About 45% of these patients developed more than one IMD episode with a median interval of 6 years between episodes of IMD [19]. Meningococcal isolates from IMD in patients with LCCD are often of serogroup Y, non-groupeable isolates or serogroups/genotypes that are rare in typical cases of IMD. Moreover, IMD disease among LCCD patients seems to be less severe with lower mortality than IMD in the general population [15,19,20]. The median age for the detection of LCCD is 17 years and it is frequently suspected due to repeated IMD episodes, while the detection of properdin deficiencies occurs earlier [15]. Moreover, fulminant and fatal IMD in patients with properdin deficiencies has been frequently reported [21–24]. However, properdin deficiencies are not all complete and there are three types: total deficiency (type I), partial deficiency (type II), and deficiency due to a dysfunctional molecule (type III).

#### 3.1.2. Acquired Factors of Susceptibility to IMD

The complement system has two facets and it plays the role of the two characters in the Dr Jekyll and Mister Hyde story. Indeed, complement is a major and beneficial actor in immune response and host defense, however, its over-activation may lead to systemic effects such as systemic lupus erythematosus (SLE, a systemic autoimmune disorder in which multiple autoantibodies against cell nuclear constituents form immune complexes that effectively activate the classical complement pathway and cause tissue damage) [25], paroxysmal nocturnal hemoglobinuria (PNH, an X-linked hematological disorder that results from somatic loss-of-function mutations impairing membrane expression of two complement inhibitors, CD55 and CD59, on red blood cells, resulting in erythrocytescomplement mediated lysis) [26], age-related macular degeneration (AMD, characterized by the progressive destruction of neurosensory retina in the macular area, and which

contributes to vision loss) [27] and atypical hemolytic uremic syndrome (aHUS, a disorder related to mutations in complement regulators (such as the factor H), and that result in a renal disease that encompasses the triad of microangiopathic hemolytic anemia, thrombocytopenia, and acute renal failure) [28]. Several of these systemic diseases may benefit from anti-complement drugs, and in particular, monoclonal antibodies (Mabs) that inhibit the late complement components. This inhibition of the complement can therefore increase susceptibility to IMD. Mabs that inhibit the C5 (Eculizumab and Ravulizumab) have reached the market and are used to treat aHUS and PNH. Other drugs are under development, targeting other components such as C3, factor B and factor D [20,29]. Treating COVID-19 with compstatin-based complement C3 inhibitor (AMY-101) has also been reported [30]. The use of anti-complement drugs in the management of various pathologies is growing [31], including the treatment of COVID-19 to control the inflammatory response [32]. IMD frequency in these patients should therefore be kept under tight surveillance.

Other acquired susceptibilities to IMD are encountered in cases of anatomic or functional asplenia. The spleen plays a central role in mounting innate and adaptive immune responses against encapsulated pathogens such as Nm. Asplenia/hyposplenia (including sickle-cell disease) were reported as a recognized risk factor of IMD in a large case-control study (odds ratio, 6.7; 95% confidence interval (CI), 3.0–14.7). Patients with hematopoietic stem cell transplantation (hSCT) are also at high risk for IMD as well as HIV patients [33,34]. hSCT is a procedure in which the immune system is transferred from the donor to the recipient. This transfer is at best incomplete and vaccine protection from the donor is usually lost. This loss is observed in particular, when the patient suffers from a graft-versus-host disease (GVHD) that requires the administration of immunosuppressive treatments [33].

hSCT transplant recipients are at risk of IMD due to total body irradiation, which induces a hyposplenism, and especially the progressive loss of specific antibodies, which has been documented in the literature for meningococci [35]. Solid organ transplant recipients may also be at risk for IMD due to immunosuppressive treatment [36].

#### *3.2. Host Factors of Severity of IMD*

The severity of IMD is frequently linked to hyperinvasive clonal complexes, and particularly, the clonal complex 11 [37]. However, several host factors are reported to be associated with severity and/or bad evolution of the disease. The deficiency of either protein C or its cofactor, protein S (anticoagulant proteins) has been reported as being associated with an increased risk of severe meningococcal sepsis [38]. Moreover, high levels of the plasminogen activator inhibitor-1 (PAI-1) have been associated with poor outcome of IMD with high sequelae and mortality rates [39]. The exacerbated inflammatory response may lead to complications such as pachymeningitis, which can be linked to promoter variants in genes involved in the inflammatory response (IL6, PAI-1 and macrophage migration inhibitory factor, MIF) [40].

#### *3.3. Impact on Anti-Meningococcal Vaccination Strategies*

Exploring the complement is highly recommended in patients who develop recurrent/chronic forms and/or mild infections provoked by unusual serogroups/genotypes of Nm. This exploration should include assays for C3, C4, CH50 and AP50 in order to detect deficiencies in early and late components and alternative pathways. When detected in a patient, the investigation should be extended to the siblings. LCCD are inherited in an autosomal recessive manner while properdin deficiencies are usually inherited as an X-linked disorder.

These patients (with acquired or hereditary complement deficiencies) are increasing due to increasing detection and new indications for anti-complement drugs such as Mabs. These drugs are being investigated in the treatment of COVID-19 [41]. Moreover, the number of patients with spleen disorders is substantial, for example, 6000 to 9000 patients are splenectomized each year in France [42].

These patients with increased susceptibility to IMD require particular management strategies including:


The immunogenicity of meningococcal vaccines in these patients requires more exploration in order to adapt vaccination schemes. For example, in a study on adult asplenic patients, they were able to achieve protective bactericidal titers after vaccination against serogroup C meningococci. However, they showed a significantly lower geometric mean titer (GMT) (157.8; 95% CI, 94.5 to 263.3) of bactericidal antibody in serum (SBA) than an age-matched control group (1448.2; 95% CI, 751.1 to 2792.0). The primary vaccination schemes may require several doses in these patients in addition to repeated boosters [43]. Immunogenicity after one dose of tetravalent conjugated ACWY vaccine was also poor in recipients of allogeneic hematopoietic stem cell transplantation [44]. The administration of two primary doses of polysaccharide conjugated anti-meningococcal vaccines is therefore recommended in several countries for patients with asplenia, HIV, or complement disorders [31,45]. No immunogenicity data on vaccines against meningococcal B are available among these subjects.

#### **4. Susceptibility to Invasive** *H. influenzae* **Infections**

Like Nm, *H. influenzae* is also a Gram-negative human-restricted encapsulated bacterium that is a member of the nasopharyngeal microbiota. Hi is highly polymorphic with six different capsular types (serotypes a to f) as well as non-capsulated isolates (nontypeable isolates, HiNT). The incidence of Hib infection has been drastically reduced since the introduction of a vaccination against this serotype. Invasive disease due to other serotypes as well as non-typeable isolates persists and no vaccine is available against these non-Hib isolates.

#### *4.1. Genetic and Acquired Susceptibilities to Invasive Haemophilus influenzae Disease*

As for Nm, disorders that affect the immune defense mechanisms and mainly the complement system are expected to increase susceptibility to invasive *H. influenzae*. The frequency of Hi infection in patients with early component deficiencies (C1, C2, C4) seems to be similar to that of meningococcal infections. However, this frequency is lower in infections in patients with C3 deficiencies and LCCD, suggesting that functions other than the lytic functions of the MAC are involved in the defense against invasive Hi infections. However, Hi invasive infections are still higher among patients with complement deficiencies (including factors P or D) than in the general population [15].

Disorders that influence the efficiency of IgG2 binding, the main isotype produced in response to encapsulated bacteria may also increase susceptibility to Hi infections. For example, the His131Arg allele encoding Fcgamma RIIa receptor (rs1801274) binds IgG2 poorly, and therefore, increases the risk of Hi infections [46]. Patients with a single nucleotide polymorphism (SNP) in the *TIRAP* gene (Toll-interleukin 1 receptor domain containing adaptor protein, an adapter molecule associated with Toll-like receptor) (rs1893352) was reported to be strongly associated with non-meningitis cases of Hib in vaccinated children. Another SNP (rs1554286, a promoter SNP in the interleukin-10 encoding gene)

was associated with epiglottitis [47]. Patients with asplenia, hSCT, HIV are also at high risk for invasive Hi disease [48].

#### *4.2. Impact on Anti-Hi Vaccination Strategies*

There is an unmet medical need in the field of vaccination against *H. influenzae* among patients at high risk due to the absence of vaccines against non-Hib isolates, and particularly, non-typeable Hi (NTHi) isolates. Unlike Nm, only vaccines against serotype B are available. New vaccines, immunogenicity knowledge and vaccination strategies are therefore needed. Non-Hib invasive infections can be more prevalent in patients at risk for Hi invasive infections, underlying the need for vaccines against other serotypes and non-typeable isolates of Hi. Moreover, studies on the immunogenicity of Hib vaccine in these patients are lacking; however, the implications of genetic traits on vaccine efficacy have been suggested [49].

#### **5. Susceptibility to Invasive Pneumococcal Infections**

The Gram-positive bacterium *Streptococcus pneumoniae* is an endemic global pathogen that causes a wide range of non-invasive and potentially life-threatening invasive diseases in children and adults. Invasive pneumococcal disease (IPD) implies invasion of pneumococcus into a normally sterile site, leading to several forms of IPD such as bacteremia, empyema, meningitis, endocarditis, and osteomyelitis [50,51]. The incidence of IPD, which ranges from 11 to 27 per 100,000 in Europe, is highest in younger children and the elderly [52–54]. Mortality rates for IPD vary from 12% to 22% in adults in developed countries and are substantially higher in low-income countries. Neurological sequelae, including hearing loss, focal neurological deficits, and cognitive impairment occur in 30–52% of surviving patients [55–58]. Susceptibility to IPD relates to both the virulence of the pathogen and to host factors. The most relevant host factors responsible for the increased risk of IPD are related to defects involving the immune system [59].

#### *5.1. Genetic and Acquired Susceptibilities to IPD*

Several inherited and acquired host factors have been shown to confer predisposition to IPD. In particular, primary immunodeficiency states, dysfunction or absence of the spleen and human immunodeficiency virus (HIV) infection, confer a high degree of susceptibility to IPD [60]. Recently, increasing evidence supports a central role of the NF-κB pathway in susceptibility to severe IPD [61].

#### 5.1.1. Inherited Factors of Susceptibility to IPD

#### Congenital Deficiencies in Immunoglobulins

In contrast to *N. meningitidis* and *H. influenzae* (Gram negative bacteria), the thick cell wall of *S. pneumoniae* (Gram positive) renders it resistant to lysis by insertion of the complement MAC. Furthermore, the presence of a polysaccharide capsule (that can have a thickness of 175 nm in some serotypes) makes them even harder targets for complementmediated lysis. Antibody-initiated complement-dependent opsonization (opsonophagocytosis), which activates the classic complement pathway, is thought to be the major immune mechanism of pneumococcal killing. Opsonization, refers to the coating of bacteria with antibodies and complement ligands, mainly C3b and iC3b, to facilitate their elimination through phagocytosis by cells bearing complement receptors. Therefore, the production of specific polysaccharide antibodies (IgA, IgM and IgG) and complement activation are the cornerstones to trigger complement-mediated opsonophagocytosis of pneumococci and proper T-B lymphocyte cooperation for an efficient antibody response. Specific antibody deficiencies to *S. pneumoniae* contribute to the increased rates of invasive infection [62]. Although specific rates are not available, patients with agammaglobulinemia (absence of B cell immunoglobulins due to a defect in maturation of B cells) or hypogammaglobulinemia (characterized by reduced serum levels of immunoglobulins and a diminished vaccinal response) are susceptible to invasive *S. pneumoniae* infection [63–65]. Specifically, as IgG an-

tibody responses to bacterial capsular polysaccharide antigens are mostly restricted to IgG2, patients with IgG2 deficiency are more susceptible to infections with *S. pneumoniae,* presumably because of the proposed unique ability of IgG2 to support neutrophil phagocytosis of pneumococci in the absence of complement [66,67]. Moreover, hyper-IgM syndromes (HIGM) are a group of hereditary immune system pathologies, characterized by ineffective immunoglobulin class switching, resulting from interrupted B cell co-stimulation. Patients with hyper-IgM have ineffective production of specific IgG and are susceptible to IPD and sepsis [68].

#### Congenital Deficiencies in Complement

Only a few clinically defined groups of patients experiencing pneumococcal disease have been systematically examined for the frequency of complement deficiencies [69]. In particular, it has been shown that certain complement deficiencies predispose patients to pneumococcal infections with, in decreasing order of frequency, the C3, the C2 and the C4 defects [63]. Sporadic pneumococcal infections have been diagnosed in patients with C1 and alternative pathway defects (properdin, factor D or factor I deficiencies) [70]. Findings on the role and the link between Mannose-binding lectin (MBL) deficiency and increased susceptibility to pneumococcal infections are conflicting [71–73]. Nevertheless, Eisen et al. analyzed the association between MBL deficiency and the outcome of IPD using data pooled from five studies with adults and one study with children and concluded that the risk of death was increased among MBL-deficient patients with *S. pneumoniae* infection (odds ratio, 5.62; 95% confidence interval, 1.27–24.92) after adjustment for bacteremia, comorbidities and age [74]. MBL deficiency may therefore be considered as a factor of severity instead of a risk factor for developing IPD.

#### Toll-Like Receptor Signaling Deficiencies

TLR signaling is critically important in the first unspecific meeting between host and microbe. Specific defects of molecules in the TLR signaling pathway including interleukin-1 receptor associated kinase-4 deficiency (IRAK-4), myeloid differentiation factor 88 (MYD88) and nuclear factor-κB essential modulator deficiency (NEMO) [63,75–78] have recently been defined. IRAK-4, a serine threonine kinase, is essential for signal transduction downstream in TLR canonical pathways. IRAK-4 deficiencies are inherited in an autosomal recessive manner [79,80]. Selective susceptibility to *S. pneumoniae* infections is high and many experience recurrent IPD in early childhood. High mortality (40%) is reported before the age of 8 years; however, among survivors, clinical phenotype of patients with IRAK-4 and MyD88 deficiencies tend to improve with age [79].

NF-κB essential modulator (NEMO), encoded by the X-linked *IKBKG* gene, is a regulatory protein essential for activation of the ubiquitous transcription factor NF-κB [81,82]. Children with NEMO-related defects present variable levels of impaired host defenses, with severe susceptibility to IPD [83–86]. Patients with these disorders mount a weak inflammatory response with delayed fever or minimal change in inflammatory markers (e.g., leukocytosis and C reactive protein levels in serum), which may explain the mild inflammatory response elicited in vivo in these patients [87]. It is worth noting that patients with NEMO defects have persistent absence of anti-pneumococcal polysaccharides antibodies after naturally occurring pneumococcal infections and after challenge with polyvalent pneumococcal polysaccharide vaccine, whereas some IRAK-4-deficient patients do [82,87,88].

#### 5.1.2. Acquired Factors of Susceptibility to IPD

*S. pneumoniae* is overwhelmingly the most common infecting organism in functional or anatomic asplenic patients, accounting for 50–90% of isolates from blood cultures in many cohorts of patients, particularly in younger patients with sickle cell anemia [89]. Mortality from IPD in asplenic patients is more than 50% [90]. As the major site for T-cell independent antibody responses to bacteria and splenic mononuclear phagocytes, the

spleen plays a critical role in controlling pneumococcal infection. Patients with asplenia have reduced levels of IgM memory B cells and IgM anti-pneumococcal antibodies, causing reduced ability to produce protective antibodies against polysaccharide antigens, and hence, possible vaccine failure [91,92].

Several studies have shown that HIV-infected individuals and adults have a significantly higher risk of acquiring *S. pneumoniae* and developing recurrent IPD [93,94]. Although active antiretroviral therapy significantly reduces the overall burden of IPD in HIV-positive populations, the risk of IPD remains 35 times higher in HIV-infected individuals than in non-HIV-infected adults [95]. Several studies have underlined the increased susceptibility to IPD in respiratory viruses infected patients, including influenza and respiratory syncytial viruses, especially in children [96–98]. Moreover, patients being treated for underlying solid or hematologic malignancies have high rates of invasive pneumococcal disease, although, interestingly, less than one-fifth of these infections occur during periods of neutropenia [99,100].

#### *5.2. Impact on Anti-Pneumococcal Vaccination Strategies*

Systematic immunological exploration in patients hospitalized for recurrent IPD is advocated. Levels of plasma Ig and IgG subclasses should be determined, especially in children who have a history of recurrent infections. In addition, screening of component complement deficiencies can be accomplished by an assessment of total complement function (CH50). Splenic function should be evaluated. In case of inherited immune deficiencies, siblings should also be examined. When detected, prophylactic measures are required to prevent infection. Based on the type of abnormality detected, these prophylactic measures fall into the following major axes:


duration of the patient's life, are administered by intravenous (400 to 600 mg/kg every 3 to 4 weeks) or subcutaneous (100 to 150 mg/kg per week) routes to regularly ensure IgG trough levels in the normal range [109].

• Patient education. It is of utmost importance that individuals with altered immune competence be informed and educated about their increased risk for serious, lifethreatening infections and understand the importance of seeking prompt medical attention should situations of risk arise (e.g., high fever). When traveling, especially to high-risk geographic areas, a prior consultation is necessary to receive recommendations and update vaccinations.

#### **6. Susceptibility to Invasive GBS Infections**

Group B *streptococcus* (GBS) is a leading cause of neonatal and infant sepsis and meningitis globally [110,111]. GBS can also cause stillbirths, prematurity and disease in pregnant women, immunocompromised adults and the elderly, but the highest incidence of disease is in neonates and young infants [112].

#### *6.1. Genetic and Acquired Susceptibilities to Invasive GBS Disease*

The susceptibility of neonates to GBS is correlated with a deficiency of maternal (transplacental) specific antibody and the intrinsically immature immune system of neonates [113]. Moreover, GBS infections in nonpregnant adults typically present when the host is in an immunocompromised or relatively compromised state, such as diabetes, cancer, HIV, with diabetes being the predominating underlying condition [114–116]. The search for monogenetic immunodeficiency disorders underlying susceptibility to invasive GBS infections has only been partially successful so far. One patient with very late-onset GBS sepsis suffering from IRAK-4 deficiency has been reported, supporting that cellular innate immunity and the TLR system are important for resistance against GBS [117].

The severity of disease can be attributed, at least in part, to the virulence of the strain and its ability to avoid immunological clearance and adapt to changing environments throughout disease progression. Indeed, the ST-17 lineage responsible for severe neonatal disease, has a number of ST-17-specific genes that may contribute to its ability to cause meningitis [118].

#### *6.2. Impact on Preventive Strategies*

Intrapartum antibiotic prophylaxis (IAP) is the only preventive strategy currently available for the prevention of perinatal GBS early-onset disease (occurring from day 0 to day 6 of life) [117,119,120].

However, IAP coverage has no impact on late onset disease (LOD, which occurs from day 7 to 90 of life), stillbirths and prematurity due to GBS, as well as a limited impact on disease in pregnant women and it might be an issue for antimicrobial resistance [121,122]. Implementing a suitable vaccine for pregnant women could provide effective protection to those forms of invasive disease that cannot be prevented with IAP or where IAP is not feasible. This preventive strategy has been identified as a priority by WHO. Based on specific capsular polysaccharide antigens, 10 serotypes of GBS have been described. A hexavalent GBS glycoconjugate vaccine that covers the major six serotypes responsible for 99% of GBS infections is the most advanced vaccine candidate. Preclinical and human phase I and II studies have been completed, revealing the safety and immunogenicity of these vaccines [123–125]. However, a large number of participants would be required to undertake Phase III clinical efficacy trials. Protein vaccines that might confer protection irrespective of serotype, are in earlier stages of development. Future use of these vaccines raises the question of the adherence of pregnant women to routine vaccination.

#### **7. Conclusions**

Several inherited or acquired risk factors are responsible for increased susceptibility to invasive bacterial diseases (Table 1). The investigation of patients with repeated invasive

bacterial diseases and patients who developed these infections with unusual isolates is recommended. The genetic dissection of inherited factors will shed light on the molecular and cellular mechanisms underlying protective immunity to bacterial pathogens, and will improve our knowledge on the interaction of the pathogen with the human immune system to pave the way for the development of new, more appropriate treatments. Furthermore, early diagnosis and proper management of immune deficiencies are essential to avoid permanent damage and serious infectious complications. In addition to vaccination, antibiotic chemoprophylaxis (including intrapartum antibiotic prophylaxis for GBS infections) should be strongly considered. However, prolonged chemoprophylaxis using broad-spectrum antibiotics may select resistant bacterial isolates, increasing the risk of selective colonization with resistant isolates. Avoiding, when possible, the use of large-spectrum antibiotics and using vaccines, when available, can contribute to reducing antimicrobial resistance by reducing the selective pressure and preventing transmission of resistant isolates. Safe vaccination, when available, should be encouraged among high-risk patients and their close contacts to prevent these infectious diseases.


**Table 1.** Congenital and acquired deficiencies and anatomic conditions that may predispose to meningococcal, pneumococcal, *H. influenzae* or GBS invasive infections

adaptor protein.

**Author Contributions:** A.-E.D. and M.-K.T. contributed to the conceptualization, research strategy, writing, reviewing and editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We would like to thank all members of our laboratory for helpful discussions and support.

**Conflicts of Interest:** M-K.T reports grants from GSK, grants from Pfizer, grants from Sanofi, outside the submitted work, M.-K.T. and A.-E.D. have a patent 630133 issued.

#### **References**


### *Article* **Understanding the Role of Duration of Vaccine Protection with MenAfriVac: Simulating Alternative Vaccination Strategies**

**Andromachi Karachaliou Prasinou \*, Andrew J. K. Conlan and Caroline L. Trotter**

Disease Dynamics Unit, Department of Veterinary Medicine, University of Cambridge, Cambridge CB3 0ES, UK; ajkc2@cam.ac.uk (A.J.K.C.); clt56@cam.ac.uk (C.L.T.)

**\*** Correspondence: ak889@cam.ac.uk

**Abstract:** We previously developed a transmission dynamic model of *Neisseria meningitidis* serogroup A (NmA) with the aim of forecasting the relative benefits of different immunisation strategies with MenAfriVac. Our findings suggested that the most effective strategy in maintaining disease control was the introduction of MenAfriVac into the Expanded Programme on Immunisation (EPI). This strategy is currently being followed by the countries of the meningitis belt. Since then, the persistence of vaccine-induced antibodies has been further studied and new data suggest that immune response is influenced by the age at vaccination. Here, we aim to investigate the influence of both the duration and age-specificity of vaccine-induced protection on our model predictions and explore how the optimal vaccination strategy may change in the long-term. We adapted our previous model and considered plausible alternative immunization strategies, including the addition of a booster dose to the current schedule, as well as the routine vaccination of school-aged children for a range of different assumptions regarding the duration of protection. To allow for a comparison between the different strategies, we use several metrics, including the median age of infection, the number of people needed to vaccinate (NNV) to prevent one case, the age distribution of cases for each strategy, as well as the time it takes for the number of cases to start increasing after the honeymoon period (resurgence). None of the strategies explored in this work is superior in all respects. This is especially true when vaccine-induced protection is the same regardless of the age at vaccination. Uncertainty in the duration of protection is important. For duration of protection lasting for an average of 18 years or longer, the model predicts elimination of NmA cases. Assuming that vaccine protection is more durable for individuals vaccinated after the age of 5 years, routine immunization of older children would be more efficient in reducing disease incidence and would also result in a fewer number of doses necessary to prevent one case. Assuming that elimination does not occur, adding a booster dose is likely to prevent most cases but the caveat will be a more costly intervention. These results can be used to understand important sources of uncertainty around MenAfriVac and support decisions by policymakers.

**Keywords:** meningitis; vaccine; Africa; mathematical modelling

### **1. Introduction**

Countries in the meningitis belt of sub-Saharan Africa have been repeatedly devastated by meningitis epidemics since the early 1900s. Primarily, these epidemics are caused by the bacterium *Neisseria meningitidis* and a number of circulating meningococcal serogroups are responsible for causing disease in the meningitis belt [1]. Until 2010, the predominant serogroup responsible for frequent epidemic cycles was *N. meningitidis* serogroup A (NmA) [2]. Since the introduction of a tailor made vaccine, MenAfriVac in 2010, over 300 million 1–29 year olds have been vaccinated against NmA, resulting in a more than 99% decline in the number of confirmed group A cases in fully vaccinated populations [3].

We previously developed a transmission dynamic model of NmA with the aim of forecasting the relative benefits of different immunisation strategies [4]. The model high-

**Citation:** Karachaliou Prasinou, A.; Conlan, A.J.K.; Trotter, C.L. Understanding the Role of Duration of Vaccine Protection with MenAfriVac: Simulating Alternative Vaccination Strategies. *Microorganisms* **2021**, *9*, 461. https://doi.org/ 10.3390/microorganisms9020461

Academic Editor: James Stuart

Received: 14 January 2021 Accepted: 18 February 2021 Published: 23 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

lighted the importance of a long-term vaccination strategy following the introductory mass campaigns of 1–29 year olds. Of the long-term strategies we investigated, a combination strategy of routine immunisation within the Expanded Programme on Immunisation (EPI) together with a mini catch-up, targeting children born after the introductory campaign, was the most effective. After reviewing the model findings and additional comprehensive information from clinical trials, the World Health Organisation's recommendation for the countries of the African meningitis belt is to introduce MenAfriVac into routine immunisation programmes within 5 years after completion of the mass campaigns. The vaccine regimen is a 1-dose schedule given at 9–18 months of age. At the time of introduction into EPI, it is recommended that countries should also include a one-time catch-up campaign to immunise those born since the introductory campaigns [5].

One of the key assumptions in our previous work was that the duration of vaccine induced protection is the same for all ages. Due to limited data at the time, we assumed that MenAfriVac offered protection for an average of 10 years. Since then, several studies have investigated the persistence of vaccine-induced antibodies and the influence of age at vaccination. These studies provide empirical evidence on the duration of the immune response to MenAfriVac, which may be used as a proxy to the duration of protection. Correlates of protection for meningococcal disease are based upon serum bactericidal activity (SBA) [6]. The studies by White et al. [7] and Yaro et al. [8] suggest that vaccine protection is age-dependent and lasts longer for individuals targeted after the age of 2 years or 5 years, respectively. These new studies were consistent in suggesting that the duration may be age-specific, but inconsistent in their estimates of the duration.

The aim of this paper is to investigate the influence of both the duration and agespecificity of vaccine-induced protection on our model predictions and explore how the optimal vaccination strategy may change in the long-term. More specifically, we consider four scenarios that could be plausible alternative strategies to the current.

#### **2. Materials and Methods**

#### *2.1. Model Structure*

Details of the model structure have been previously published [4]. In brief, it is a compartmental model that divides the population into: susceptible state, carrier of NmA, disease due to NmA, and recovered and immune, with each of these states replicated for vaccinated and unvaccinated.

In a modification for this paper, instead of having broad age groups, we now divided the population into annual age cohorts. A model modification was also necessary in order to simulate an age-dependent duration of protection. We added four new compartments to the previous model. These four states represent the susceptible, carriers, diseased, and recovered/immune who receive vaccination before the age of five years. A table with all of the compartments and their descriptions can be found in Appendix A, together with a flow diagram of the model.

#### *2.2. Model Parameters*

Demographic data for Chad were used to estimate parameters for the model. Epidemics of NmA in Chad in the pre-vaccine era occurred every 8–12 years, which is representative of the epidemiology of NmA in the African meningitis belt [9]. The introduction of MenAfriVac in that country was completed in two phases during 2011 and 2012 [10]. In the model, because there is no geographic sub-division, we assume that 50% of the target population were vaccinated in 2011 while the remaining 50% received the vaccine in 2012.

We assumed that the coverage in Chad for routine immunisation of infants starts at 75% in 2017 and continues with annual increments of 1% until it reaches 90%. It is then assumed to stay constant for the remaining years. There is no vaccine currently being administered at 5 or 10 years of age. Hence, we explored the impact of vaccinations, assuming a coverage of 80% at these ages. Other parameters were based on the available literature wherever possible, as previously described [4].

To account for the uncertainty around the duration of protection, we ran each scenario outlined below under the following four different assumptions: (1) an average of 5 years duration of protection for all ages; (2) 10 years duration of protection for all ages; (3) 20 years duration of protection for all ages; and (4) 5 years duration of protection for <5 year olds and 10 years duration of protection for children at 5 years of age or older.

#### *2.3. Vaccination Strategies*

We considered a range of vaccination strategies (Table 1) that were elucidated through informal discussions with colleagues at WHO, PATH, and CDC to be of interest.


#### **Table 1.** Vaccination strategies considered in the model.

#### *2.4. Model Implementation*

The model is run for the time period 2010–2060 using a daily time step. For each model run, the number of cases by age is calculated per year. The average number of cases and the percentage of cases prevented by each of the strategies is calculated over 200 simulation runs per strategy. To account for the uncertainty due to the stochastic nature of the model, 95% confidence intervals were calculated using a student-t distribution as implemented by the *t*-test function in R version 3.4.2 [11].

To allow for a comparison between the different strategies, we report the time to resurgence, median age of infection, and the age distribution of cases for each strategy. As time to resurgence, we define the year in which the number of cases exceeds the threshold of 1 case per 100,000 population following the preventive campaigns. The comparison of each metric is based on non-overlapping confidence intervals. Due to the large range from 10 years to 20 years duration of protection for all ages, we also investigate the effect of all the intermediate years on the time to resurgence in a sensitivity analysis. As an additional measure of efficiency, we calculated the number of people needed to vaccinate (NNV) to prevent one case [12]. We define NNV as the total number of doses administered divided by the total number of cases prevented under each vaccination strategy over the time period under consideration. The total number of doses given for each scenario was calculated by multiplying the total number of people targeted with the assumed age-specific vaccine uptake.

#### **3. Results**

#### *3.1. Baseline Scenario (10 Years Duration of Protection)*

The model results suggest that if the assumed duration of protection is 10 years for all ages, routine immunization aimed at schoolchildren is not better than routine immunization of 1-year-old children. However, switching the age at vaccination from 12 months to 5 years is the single dose strategy with the lowest average number of cases predicted, albeit there is a 5-year period with two doses. The numerical results for the different strategies are given in Appendix B. The strategy that leads to the largest number of cases averted is the Booster strategy with a 79.3% (CI: 78.7–79.8%) predicted overall reduction, relative to a 66.8% (66.2–67.4%) predicted reduction if the current strategy remains unchanged until 2060, but the NNV is much higher (Table 2).

#### *3.2. Time to Resurgence*

The model predicts that when the assumed duration of protection is 12 years or shorter for all ages, a resurgence always follows the initial mass campaigns (Figure 1). The size of the peak as well as the year of resurgence both depend on the schedule and the duration of protection. The longer the duration of protection, the longer the honeymoon period is. No resurgence was seen in model runs (i.e., 100% of the 200 simulations result in elimination) when vaccine-induced protection was assumed to last for an average of 18 years or longer. For an assumed duration of protection of 16 years for all ages, 69.5% of the simulation runs resulted in elimination after the mass campaigns. Summary statistics showing the year the disease incidence exceeds the threshold of one case per 100,000 population for duration of protection between 10 and 20 years can be seen in Table 3 in Appendix B. *Microorganisms* **2021**, *9*, x FOR PEER REVIEW 5 of 14

**Figure 1.** Average disease incidence across the different vaccination scenarios and across the different assumptions regarding the duration of vaccine-induced protection. Shaded areas represent the 95% confidence intervals. **Figure 1.** Average disease incidence across the different vaccination scenarios and across the different assumptions regarding the duration of vaccine-induced protection. Shaded areas represent the 95% confidence intervals.

If we assume that duration of protection is 5 years regardless of the age at vaccination, the model predicts that the number of cases will start increasing only 10 years after the introduction of MenAfriVac, compared to ~17 years of honeymoon period when duration of protection is 10 years. Earlier resurgence does not necessarily translate to a larger If we assume that duration of protection is 5 years regardless of the age at vaccination, the model predicts that the number of cases will start increasing only 10 years after the introduction of MenAfriVac, compared to ~17 years of honeymoon period when duration of

Of the strategies considered, the Booster strategy resulted in the fewest cases across all different assumptions regarding the duration of protection (Figure 3, Table 2). Taking into consideration only the single-dose schedules, the model results suggest that if the duration of protection is assumed to be the same for everyone regardless of at what age they are targeted, routine immunization at 12 months of age (EPI@12m) is similar to routine immunization at older ages (EPI@5y and EPI@10y). There is considerable overlap in the results but strategy Switch is the strategy with the lowest average number of total cases predicted (Figure 3). However, assuming that vaccination of 1-year-old leads to a shorter duration of antibody persistence compared to vaccination at older ages, strategies EPI@5y and EPI@10y result in a lower number of predicted cases compared to the

*3.3. Burden of Disease* 

EPI@12m strategy.

protection is 10 years. Earlier resurgence does not necessarily translate to a larger number of total cases (Figure 2).

#### *3.3. Burden of Disease*

Of the strategies considered, the Booster strategy resulted in the fewest cases across all different assumptions regarding the duration of protection (Figure 3, Table 2). Taking into consideration only the single-dose schedules, the model results suggest that if the duration of protection is assumed to be the same for everyone regardless of at what age they are targeted, routine immunization at 12 months of age (EPI@12m) is similar to routine immunization at older ages (EPI@5y and EPI@10y). There is considerable overlap in the results but strategy Switch is the strategy with the lowest average number of total cases predicted (Figure 3). However, assuming that vaccination of 1-year-old leads to a shorter duration of antibody persistence compared to vaccination at older ages, strategies EPI@5y and EPI@10y result in a lower number of predicted cases compared to the EPI@12m strategy. *Microorganisms* **2021**, *9*, x FOR PEER REVIEW 6 of 14

**Figure 2.** Total number of cases plotted against the year of resurgence across all scenarios and all assumptions regarding duration of protection and coverage. Each strategy is represented with a different colour and each assumption about the duration of protection is represented with a different symbol shape. Note that 20 years duration of protection is not shown. Error bars show the 95% confidence interval. **Figure 2.** Total number of cases plotted against the year of resurgence across all scenarios and all assumptions regarding duration of protection and coverage. Each strategy is represented with a different colour and each assumption about the duration of protection is represented with a different symbol shape. Note that 20 years duration of protection is not shown. Error bars show the 95% confidence interval.

**Figure 3.** Box plot showing the median, interquartile range, and full range of the predicted total number of cases for different immunisation strategies in the time period 2010–2060 from 200 simu-

lation runs.

confidence interval.

**Figure 2.** Total number of cases plotted against the year of resurgence across all scenarios and all assumptions regarding duration of protection and coverage. Each strategy is represented with a different colour and each assumption about the duration of protection is represented with a different symbol shape. Note that 20 years duration of protection is not shown. Error bars show the 95%

**Figure 3.** Box plot showing the median, interquartile range, and full range of the predicted total number of cases for different immunisation strategies in the time period 2010–2060 from 200 simulation runs. **Figure 3.** Box plot showing the median, interquartile range, and full range of the predicted total number of cases for different immunisation strategies in the time period 2010–2060 from 200 simulation runs.

**Table 2.** Number of doses given, number of cases predicted and averted, and number of doses needed to prevent one case for each immunization strategy for the time period 2011–2060. All of the numbers, apart from the number of people needed to vaccinate (NNV), are in the thousands. Averages across 200 simulation runs.


Vaccination programmes raise the average age of infection since vaccinated children are protected against disease. Routine immunization at 10 years (EPI@10y) is associated with the lowest median age of infection as it results in a large number of unprotected children at a very young age leading to a large number of cases in the under 10-year-olds (Figure 4). *Microorganisms* **2021**, *9*, x FOR PEER REVIEW 8 of 14

**Figure 4***.* Box plot showing the median, interquartile range, and full range of the total number of cases by age group from 200 simulation runs aggregated over the time period 2010–2060. **Figure 4.** Box plot showing the median, interquartile range, and full range of the total number of cases by age group from 200 simulation runs aggregated over the time period 2010–2060.

The strategy with the highest median age of infection is the routine immunization targeting children on their first birthday, with or without a booster dose when they turn 5 years of age. Anyone can develop invasive meningococcal disease, but rates of disease are higher in children under the age of 5 years [13]. However, carriage prevalence is higher in individuals aged 5–19 years [14]. Routine immunization of 1-year-old children leads to waning of vaccine protection by the teenage years, when there is still heightened risk of meningitis. A booster dose extends the protection until the individuals age into a lower risk age group, which in turn results in a decreased transmission. The strategy with the highest median age of infection is the routine immunization targeting children on their first birthday, with or without a booster dose when they turn 5 years of age. Anyone can develop invasive meningococcal disease, but rates of disease are higher in children under the age of 5 years [13]. However, carriage prevalence is higher in individuals aged 5–19 years [14]. Routine immunization of 1-year-old children leads to waning of vaccine protection by the teenage years, when there is still heightened risk of meningitis. A booster dose extends the protection until the individuals age into a lower risk age group, which in turn results in a decreased transmission.

#### **4. Discussion 4. Discussion**

At the time of developing our previous model, data on the duration of vaccine-induced protection was limited. We based our assumption of an average of 10 years duration of protection on findings from unpublished trials and expert opinion. The initial mass campaigns in the countries of the meningitis belt started taking place in 2010, but vaccination in children under the age of 12 months did not start before 2016. Here, we update our previous model to take into account findings from two recent studies, suggesting that protection lasts longer in individuals receiving MenAfriVac after the age of two years At the time of developing our previous model, data on the duration of vaccine-induced protection was limited. We based our assumption of an average of 10 years duration of protection on findings from unpublished trials and expert opinion. The initial mass campaigns in the countries of the meningitis belt started taking place in 2010, but vaccination in children under the age of 12 months did not start before 2016. Here, we update our previous model to take into account findings from two recent studies, suggesting that protection lasts longer in individuals receiving MenAfriVac after the age of two years [6,7].

[6,7]. We used this updated model to assess the impact of a set of new vaccination strate-

Assuming that the duration of protection is at most 10 years, model results suggest that meningococcal disease cannot be eliminated within the first 50 years after the initial vaccination by the current or new strategies explored. On the contrary, provided that high We used this updated model to assess the impact of a set of new vaccination strategies and compared them to the current strategy followed by African countries, since 2015.

Assuming that the duration of protection is at most 10 years, model results suggest that meningococcal disease cannot be eliminated within the first 50 years after the initial vaccination by the current or new strategies explored. On the contrary, provided that high antibody levels persist for an average of 20 years, all strategies, including the current, result in a possible elimination of NmA cases since there are no predicted cases until at least 2060.

As a long-term strategy, in the absence of any catch-up campaigns, routine vaccination of 10 year olds would lead to the smallest average number of cases. However, including the campaigns, in the case of determining which strategy leads to the least number of cases, assuming that the duration of protection is the same across all ages, no single-dose strategy is superior to the rest as there was considerable overlap in the results. This is due to the mini catch-up campaign, which is part of only the current strategy (EPI@12m) and not the other two (EPI@5y and EPI@10y). The main difference in the results comparing the strategies is in the age distribution of cases. Reductions in the number of cases in one age group results in a rise of cases in another age group. Routine vaccination at 12 months offers better protection in young children, whereas vaccination at older ages reduces disease burden in adolescents and young adults. The risk of developing at least one major sequelae after meningococcal meningitis is higher in children under the age of 5 years [15]. In this study, we do not calculate Disability Adjusted Life Years (DALYs), where an age-specific weight may be appropriate. Assuming that vaccine protection is short-lived in children under the age of 5 years, the model suggests that it would be wiser to change the target age of routine immunization from 12 months to 5 years provided that coverage is at least 50%.

Routine immunization at 10 years of age (EPI@10y) is consistently the most effective strategy across all different assumptions about the duration of vaccine protection in terms of the number of people needed to vaccinate (NNV). This is due to the small number of doses administered, calculated based on Chad's population demography. The high annual growth rate of the country results in a triangle-shaped age distribution with the number of individuals declining with age. The strategy associated with the highest NNV is the strategy with the additional booster dose since the number of doses is almost double that of the rest of the strategies. NNV is widely used in the scientific literature. The nature of the disease (endemic, epidemic, high/low *R*0) as well as the way NNV is calculated can produce biased results [12] and, thus, caution should be taken when interpreting results or comparing NNVs with other diseases in the scientific literature. However, the highest NNV value of 485 produced by the simulations for the Booster strategy is far superior to NNV 2800–3700 estimated by Trotter et al. [16] when evaluating the response thresholds for reactive vaccination campaigns.

This is the first model to explore the potential benefits of targeting schoolchildren for routine immunization with MenAfriVac. As in all mathematical models, there is uncertainty around the model structure and certain key model parameters. The results from this work were generated using demographic data from Chad, a country lying entirely in the meningitis belt and which suffered from epidemics every 8 to 12 years before the introduction of MenAfriVac in 2011 [9]. The same structure is used to model different countries across the belt; here, we chose Chad as a typical example, but given that countryspecific demography is not substantially different, we believe the results are more broadly generalisable to other meningitis belt countries. A number of key parameters, such as the transmission rate and the duration of natural immunity remain unknown; therefore, were kept the same as in the original study, allowing for a more direct comparison. Mixing parameters are also important in age-structured models. The carriage prevalence produced by our model is consistent with contact studies in Africa, in which the highest intensity of contacts is observed in 5–15 year olds [17].

We used several metrics to compare the different strategies qualitatively and quantitatively. None of the strategies explored in this work is superior in all respects. This is especially true when vaccine induced protection is the same regardless of the age at

vaccination. Immunising infants (EPI@12m) offers protection to young children and raise the median age of infection. However, the NNV to prevent one case is higher than the NNV to prevent one case when EPI targets 10-year-olds (EPI@10y). Leaving children up to the age of 10 years unprotected, however, results in more cases in younger ages and less in older age groups. The Booster strategy may result in the least number of cases but it is the most costly intervention since it needs two doses and therefore we assume approximately double the cost of the others. A possible change in the current immunization schedule would have to be based in the prioritization of all the above factors.

The uncertainty around the assumptions regarding the duration of protection was also explored in another mathematical model forecasting the impact of MenAfriVac vaccination by Jackson et al. [18]. In their study, they mainly focused on updating and validating their previous model in light of newly data [19]. In contrast to our work, Jackson et al. assumed that routine vaccination solely targets 9 months old children. They also explored the benefits of adding a booster dose at 10 years of age in a sensitivity analysis. Despite their structural differences, both models highlight the critical need for a long-term immunization strategy to sustain low levels of infection as well as the importance of continuous updating of models when new data become available.

Since the start of immunization with MenAfriVac, there has been an increased disease incidence caused by serogroups other than serogroup A. A new pentavalent vaccine is being currently developed with the expectation of licensure by end of 2022 [20]. In order to estimate the impact of introducing this new pentavalent vaccine in an already vaccinated population, a more robust study, including a multi-serogroup model, should be performed. This will involve a number of new unknown parameters and further increase the complexity of the model structure. Yaesoubi et al. [21] developed a transmission dynamic model to investigate the cost-effectiveness of alternative vaccination strategies using the novel multivalent vaccine. They concluded that the inclusion of a catch-up campaign with the novel vaccine would be a cost-effective way to further reduce the meningococcal disease burden.

Despite the limitations of this work, and the uncertainty surrounding the introduction of the pentavalent vaccine in the countries of the African meningitis belt, this analysis and the conclusions drawn can be used in the future by policymakers to understand the importance of the duration of vaccine protection and support decision making around vaccine scheduling, such as a shift to routine immunization at an older age or the addition of a booster dose. This change can either be the addition of a booster dose at a later age or simply the age of the primary dose. The aim of this study is to identify the optimal way to maintain the success of MenAfriVac in reducing the number of MenA cases in the long-term. Additional work on the feasibility and cost-effectiveness of policy changes is also essential. In the future, with the advent and rollout of affordable multivalent vaccines, protection against NmA and other serogroups will be enhanced.

#### **5. Conclusions**

Models can be useful in investigating a range of assumptions and a variety of vaccine strategies. Further empirical studies of the duration of protection (or the duration of the immune response) following MenAfriVac will help to decrease uncertainty about the optimal vaccination policy.

**Author Contributions:** Conceptualization, A.K.P. and C.L.T.; methodology, A.K.P., A.J.K.C., C.L.T.; model coding, A.K.P.; formal analysis, A.K.P.; writing—original draft preparation, A.K.P.; writing review and editing, A.K.P., A.J.K.C., C.L.T.; visualization, A.K.P.; supervision, A.J.K.C., C.L.T.; funding acquisition, C.L.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Vaccine Impact Modelling Consortium (www.vaccineimpact. org). VIMC is jointly funded by Gavi, the Vaccine Alliance, and by the Bill & Melinda Gates Foundation. The views expressed are those of the authors and not necessarily those of the Consortium or its funders.

**Acknowledgments:** We thank Antoine Durupt, James Stuart, Andre Bita, and Marie-Pierre Preziosi for comments that helped to shape our modelling work.

**Conflicts of Interest:** Caroline L. Trotter reports receiving a consulting payment from GlaxoSmithKline in 2018, outside the submitted work. Other authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **Appendix A**

**Table A1.** List of the model compartments and their definitions.


People are born in the susceptible compartment (S). Children vaccinated up to the age of 5 years are transferred to the SE, CE, IE, and RE compartments while individuals who are targeted at 5 years or older are moved to the SV, CV, IV, RV compartments accordingly. For example, during the initial mass campaigns, children in the age groups 1–2 years, 2–3 years, 3–4 years, and 4–5 years are transferred to the early vaccination compartments (SE, CE, IE, RE) while individuals between 5 and 29 years of age are moved to the vaccinated compartments SV, CV, IV, and RV. Note that there is no movement to the IE or IV compartments upon vaccination as we assume that individuals with meningitis do not receive a vaccine dose. Individuals in the vaccinated states (SE, SV, CE, CV, IE, IV, RE, RV) revert to the equivalent unvaccinated S, C, I, R states at the age-specific rates *w*<sup>1</sup> and *w*2, depending on the strategy implemented (Figure A1). When duration of protection is the same for all ages, then *w*<sup>1</sup> = *w*2.

With the addition of the extra compartments, the force of infection for age group *j* becomes

$$\lambda\_{\rangle} = \theta \sum\_{k=1}^{100} \beta \left( z\_{\rangle}, z\_{k} \right) \left( I\_{k} + \mathbb{C}\_{k} + I V\_{k} + \mathbb{C} V\_{k} + I E\_{k} + \mathbb{C} E\_{k} \right) \tag{A1}$$

where θ is the stochastic term, which changes annually, and was previously described [4] and *β zj* , *z<sup>k</sup>* is the transmission rate between age groups *j* and *k*.

**Figure A1.** Flow diagram of the model with vaccination. Susceptible individuals become carriers with age and time dependent force of infection (λ(z,t)), which is reduced by the vaccine efficacy against carriage (δ) for vaccinated people. Similarly, the age and time dependent rate at which carriers develop disease (a(z,t)) is reduced by the vaccine efficacy against disease (ξ). Carriers and diseased individuals recover at a rate α and ρ, respectively. Temporary immunity wanes at a rate φ, while vaccine induced protection wanes at a rate <sup>1</sup> for children vaccinated before the age of 5 years and <sup>2</sup> for people vaccinated after their 5th birthday. People die at an age-specific natural mortality rate not shown here. **Figure A1.** Flow diagram of the model with vaccination. Susceptible individuals become carriers with age and time dependent force of infection (λ(z,t)), which is reduced by the vaccine efficacy against carriage (δ) for vaccinated people. Similarly, the age and time dependent rate at which carriers develop disease (a(z,t)) is reduced by the vaccine efficacy against disease (ξ). Carriers and diseased individuals recover at a rate α and ρ, respectively. Temporary immunity wanes at a rate ϕ, while vaccine induced protection wanes at a rate *w*<sup>1</sup> for children vaccinated before the age of 5 years and *w*<sup>2</sup> for people vaccinated after their 5th birthday. People die at an age-specific natural mortality rate not shown here.


NNV - 241 196 173 218 309

**Table A3.** Summary statistics showing the year disease incidence exceeds the threshold of 1 case per 100,000 population from 200 simulation runs for a range of values for the duration of protection. The scenario simulated to generate these

Minimum 2033 2035 2037 2040 2044 2047 2056 - 1st Quartile 2038 2041 2044 2048 2053 2053 2057 -

#### **Appendix B Appendix B**

Total number of doses given (in millions)

results is the EPI@12m.

**Table A2.** Numerical results for the different vaccination scenarios for the time period 2010–2060. Duration of protection is 10 years and vaccine uptake for children routinely immunized over the age of 12 months is assumed to be 80%. Each value presented is the mean and 95% confidence interval is given inside the brackets. **Table 2.** Numerical results for the different vaccination scenarios for the time period 2010–2060. Duration of protection is 10 years and vaccine uptake for children routinely immunized over the age of 12 months is assumed to be 80%. Each value presented is the mean and 95% confidence interval is given inside the brackets.



**Table 3.** Summary statistics showing the year disease incidence exceeds the threshold of 1 case per 100,000 population from 200 simulation runs for a range of values for the duration of protection. The scenario simulated to generate these results is the EPI@12m.

\* All 200 simulations resulted in elimination.

#### **References**


## *Review* **A Narrative Review of the W, X, Y, E, and NG of Meningococcal Disease: Emerging Capsular Groups, Pathotypes, and Global Control**

**Yih-Ling Tzeng <sup>1</sup> and David S. Stephens 1,2,\***


**Abstract:** *Neisseria meningitidis*, carried in the human nasopharynx asymptomatically by ~10% of the population, remains a leading cause of meningitis and rapidly fatal sepsis, usually in otherwise healthy individuals. The epidemiology of invasive meningococcal disease (IMD) varies substantially by geography and over time and is now influenced by meningococcal vaccines and in 2020–2021 by COVID-19 pandemic containment measures. While 12 capsular groups, defined by capsular polysaccharide structures, can be expressed by *N. meningitidis*, groups A, B, and C historically caused most IMD. However, the use of mono-, bi-, and quadrivalent-polysaccharide-conjugate vaccines, the introduction of protein-based vaccines for group B, natural disease fluctuations, new drugs (e.g., eculizumab) that increase meningococcal susceptibility, changing transmission dynamics and meningococcal evolution are impacting the incidence of the capsular groups causing IMD. While the ability to spread and cause illness vary considerably, capsular groups W, X, and Y now cause significant IMD. In addition, group E and nongroupable meningococci have appeared as a cause of invasive disease, and a nongroupable *N. meningitidis* pathotype of the hypervirulent clonal complex 11 is causing sexually transmitted urethritis cases and outbreaks. Carriage and IMD of the previously "minor" *N. meningitidis* are reviewed and the need for polyvalent meningococcal vaccines emphasized.

**Keywords:** *Neisseria meningitidis*; capsule; meningococcal group; nongroupable; meningococcal carriage; invasive meningococcal disease; meningococcal urethritis

#### **1. Introduction**

*Neisseria meningitidis* (the meningococcus), a Gram-negative pathogen of humans, causes epidemic meningitis and rapidly fatal sepsis in many parts of the world. *N. meningitidis* is usually a commensal inhabitant of the human respiratory tract, is isolated from the nasopharynx of 3–20% of healthy individuals in the absence of outbreaks or crowding [1,2] and is transmitted from person to person by close contact of large aerosolized droplets or with oral or nasal secretions. Recent studies have suggested a global decline in overall meningococcal carriage.

For both children and adults, carriage of *N. meningitidis* can be an immunizing event, resulting in systemic protective immune response. While in most instances, the acquisition of meningococci in the upper respiratory tract does not result in invasive disease, invasive meningococcal disease (IMD), even with antibiotic therapy and supportive care, has a mortality rate that remains at 10–15%. Factors determining the establishment of carriage versus the development of invasive meningococcal disease following acquisition include expression of capsule and other bacterial virulence determinants (reflected in virulent genotypes, such as clonal complex (cc) 5 and cc7 (group A), cc41/44, cc32, cc18, cc269, cc8, and cc35 (group B), and cc11 (group C)) and host susceptibility.

**Citation:** Tzeng, Y.-L.; Stephens, D.S. A Narrative Review of the W, X, Y, E, and NG of Meningococcal Disease: Emerging Capsular Groups, Pathotypes, and Global Control. *Microorganisms* **2021**, *9*, 519. https://doi.org/10.3390/ microorganisms9030519

Academic Editor: James Stuart

Received: 5 February 2021 Accepted: 26 February 2021 Published: 3 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

The meningococcus often produces a capsular polysaccharide (CPS) in which structural differences form the basis of the historic serogroup typing system. Although there are 12 capsular groups expressed by *N. meningitidis* and defined genetically [3], three groups, A, B, and C, have been associated with significant invasive disease. Group A (MenA) *N. meningitidis* expressing a homopolymeric (α1 → 6) *N*-acetylmannosamine 1-phosphate capsule caused large pandemic outbreaks globally through much of the 20th century that persisted especially in the meningitis belt of sub-Saharan Africa until introduction of the MenA conjugate vaccine, MenAfriVac, in 2010. Meningococci of groups B and C (MenB and MenC), which are (α2 → 8)- and (α2 → 9)-linked homopolymers of sialic acid (*N*acetylneuraminic acid), respectively, cause clusters or local outbreaks (MenC) or localized longer outbreaks and hyperendemic disease (MenB) throughout the world and have been responsible for most sporadic meningococcal disease in developed countries. The "minor" groups, X, Y, and Z were first identified by Slaterus in 1961 and groups E (29E) and W (W135) were identified in 1968. The group Y (MenY) capsular polymer is composed of alternating D-glucose and partially O-acetylated sialic acid, while the group W (MenW) capsular polysaccharide is an alternating D-galactose and partially O-acetylated sialic acid. MenY and MenW have emerged as groups causing epidemic outbreaks and global disease since 1995. Group X meningococci (MenX) expressing homopolymer of (α1→4) *N*-acetylglucosamine 1-phosphate [4] also now cause outbreaks and endemic disease in parts of sub-Saharan Africa. Meningococcal group E (MenE) expresses a capsule consisting of alternating D-galactosamine and 2-keto-3-deoxyoctulosonate (KDO) residues [5] and nongroupable (MenNG) strains, either with capsule null locus (*cnl*) or unencapsulated due to inactivation of capsule synthesis, rarely but are now identified as a cause of invasive disease especially in immunocompromised individuals. Recently an unencapsulated cc11 meningococcal pathotype causing sexually transmitted urethritis cases, outbreaks, and disease clusters has also been recognized [6–8].

Meningococcal disease incidence has historically been cyclical in nature [9]. Epidemiological studies of *N. meningitidis* have clearly shown that IMD varies over time, influenced by the circulating meningococcal groups and clonal complex genotypes, by geographic locations and by the host populations affected [10]. In this report, we provide an overview of carriage and invasive disease caused by "minor" *N. meningitidis* groups W, X, Y, and E as well as nongroupable meningococci causing invasive disease and sexually transmitted infections. During the COVID-19 pandemic, many countries have experienced a significant, sustained reduction in invasive diseases due to *N meningitidis* [11] and other invasive respiratory pathogens (*Hemophilus influenzae* and *Streptococcus pneumoniae*, but not *S. agalactiae*), coinciding with the introduction of COVID-19 containment measures (medRxiv preprint, https://doi.org/10.1101/2020.11.18.20225029, access on 20 November 2020).

#### **2. Minor** *N. meningitidis* **Capsular Groups**

#### *2.1. Group W*

Group W meningococci (MenW) prior to 2000 was an infrequent cause of meningococcal disease. The first global epidemic caused by MenW was detected in 2000, after the Hajj pilgrimage to Mecca, Saudi Arabia [12] and was attributed to a specific MenW:cc11 lineage, referred to as the Hajj lineage [13]. The MenW attack rate was high among the pilgrims and household contacts of returning pilgrims. The emergence of MenW in a background of the hypervirulent cc11 lineage was likely related to capsule switching events [14]. Subsequently, MenW strains have continued to evolve and cause global disease. The global spread of diverse cc11 lineages expressing capsular groups C, B, and W, resolved by genomic typing, divided the invasive MenW:cc11 lineage into the Hajj and the South American sublineages [14]. The South American sublineage, with the spread in Brazil, Argentina, and Chile, emerged in the mid-2000s [15] and has been further divided into the original U.K. 2009 lineage [14] and the newly emerging novel U.K. 2013 lineage [16].

Surveillance data from 13 European countries revealed an increase in MenW IMD in the period of 2013–2017 [17]. While the annual incidence of IMD remained stable dur-

ing that time, the incidence and the proportion of MenW IMD among all IMD increased significantly. Average annual percentage increase in MenW incidence during this period was significant for the Netherlands (133%), Norway (86%), Spain (62%), Sweden (58%), Switzerland (44%), Germany (35%), and England (23%). The proportion of MenW among all IMD cases varied considerably between countries. The proportion of MenW was lowest in Portugal, Greece, and Poland (2–3%), while it was highest in Switzerland, the Netherlands, and England (22–24%) [17]. Of the MenW IMD isolates analyzed by multilocus sequence typing (MLST), 80% belonged to cc11 but cc22, cc174, and cc865 also caused disease. The proportion of MenW cc11 increased from 64% in the year 2013 to 86% in the year 2016. The increase in MenW IMD in England and Wales since 2009 has been mainly due to the novel U.K. 2013 lineage [16]. MenW IMD incidence, with an associated case fatality rate of 28.6%, increased from 0.02/100,000 in 2013 to 0.29/100,000 in 2017 in the Republic of Ireland and the Ireland MenW isolates clustered among both the original UK 2009 and the novel U.K. 2013 lineages [18]. Sweden had a low incidence of MenW IMD with an average incidence of 0.03 case/100,000 population from 1995 to 2014; however, the incidence of MenW increased 5-fold in 2015. This increase in MenW IMD was due to isolates belonging to the novel U.K. 2013 lineage that were introduced into Sweden in 2013 and have since been the dominant lineage of MenW [19]. The increases seen in Europe follows an increasing incidence of MenW in South America since 2004 [15], in Australia since 2013 [20], and in Canada since 2015 [21].

MenW became the predominant meningococcal capsule group in Australia in 2016 [20]. In 2017, an unprecedented outbreak of MenW infection occurred among the Indigenous pediatric population of Central Australia. Among these cases were atypical manifestations, including meningococcal pneumonia, septic arthritis, and conjunctivitis [22]. The Canadian MenW:cc11 isolates have been shown to be distinct from the traditional MenW:cc22. Both the Hajj-related and non-Hajj MenW:cc11 strains were associated with IMD in Canada [21].

A review of IMD in the Asia–Pacific region conducted by Global Meningococcal Initiative (GMI) recently reported that the predominant capsular groups were B, W, and Y in Australia, New Zealand, and China [23]. MenW circulation is significant across the Asia–Pacific region. The Philippines reported that 16.7% of sterile specimens collected in 2018 were MenW. As noted, a higher percentage (28%) of MenW was reported in Australia and a similar proportion (30%) of MenW cases was reported in New Zealand during the same period. An update on the global spread of cc11 provided during the GMI meeting highlighted the presence of the MenW:cc11 Hajj strain sublineage in Russia and Bangladesh; the MenW:cc11 South American strain sublineage in Russia, Japan, and New Zealand; the MenW:cc11 Chinese strain sublineage in China and Japan; and a further distinct MenW:cc11 strain in Bangladesh [23].

The introduction of the MenAfriVac vaccine in 2010 dramatically reduced MenA cases in 26 countries of the meningitis belt but magnified other groups as significant problems in the region, in particular groups C, W, and X. While MenW:cc11 cases have been reported in the African meningitis belt since the late 1990s and no epidemics have occurred since 2001, MenW:cc11 seems to have reemerged after 2010 [24]. In 2016, Togo experienced its second largest epidemic of bacterial meningitis since 1997, where 91.5% were due to MenW:cc11 [25]. The MenW:cc11 isolates collected in Burkina Faso during 2011–2012, Mali during 2012, and Niger during 2015 have been shown to descend from the strain identified during the Hajj-related outbreak of 2000 [26–28]. On the other hand, the MenW:cc11 isolates from Central African Republic in 2015–2016 grouped together in a genetic cluster separated from the Anglo-French Hajj sublineage and the South American/UK sublineage. These data appear to support a multifocal emergence of MenW:cc11 strains. The epidemiology of IMD in South Africa over 14 years [29] shows that MenW accounts for 49.5% IMD. Patients with MenW were 3 times more likely to present with severe disease than those with MenB, and HIV was associated with an increased risk of IMD, especially for MenW and MenY diseases.

#### *2.2. Group X*

Sporadic cases of IMD caused by *N. meningitidis* group X (MenX) have been reported in industrialized countries since 1980s [30–32] but since the late 1990s, MenX has emerged as a cause of IMD outbreaks in sub-Saharan African countries [33–39]. The PubMLST database (>75,000 isolate records, updated on 01/03/2021) contains a collection of 636 MenX isolates, 1961–2019. MenX is the latest group to cause large localized outbreaks in Kenya [36,37], Niger [34,35], Ghana [33], Mali, and Burkina Faso [38]. A study examining MenX burden and epidemiological patterns during 2006–2010 [38] showed that in Togo during 2006–2009, MenX accounted for 16% of the bacterial meningitis cases; while in Burkina Faso during 2007–2010, MenX accounted for 7% of meningitis cases, with a significant increase from 2009 to 2010 (4–35% of all confirmed cases, respectively) [38]. With the successful vaccination campaign of the MenA conjugate vaccine starting in 2010, the significance of MenX in the African Meningitis belt has become more evident. In Burkina Faso, a few months after the introduction of MenAfriVac in 2011, among the 258 confirmed meningococcal cases, only 1.6% were MenA, whereas 59% were MenX [39]. Thus, MenX, along with MenC and MenW disease in the meningitis belt, is a major driver for a new pentavalent conjugate (ACXYW) vaccine in clinical trials for sub-Saharan Africa [40]. Of note, IMD due to MenX (cc750) has also been seen in the United Kingdom (Scotland) and elsewhere in Europe. MenX has also been identified rarely in the hypervirulent cc5, cc11, and cc41/44 backgrounds.

MenX expressing genotypes (X:4) can be efficiently transmitted and colonize the nasopharynx as was seen in military recruits in the United Kingdom [41]. In a longitudinal carriage study investigating the dynamics of meningococcal carriage during an interepidemic period in Ghana, the disappearance of MenA was accompanied by a sharp increase in carriage of MenX, reaching 17% and coincided with an outbreak of MenX disease [33,42]. During the peak of the MenX wave, the ratio of MenX cases to carriers was found to be between 0.1 and 0.3 per 1000 cases; while the ratio of MenA during the outbreak was between 16.8 and 42.3 per 1000 cases in the respective dry seasons [42]. These studies suggest that MenX has a disease-to-carriage ratio significantly lower than MenA and that MenX have a lower invasive potential. Like other outbreak-causing meningococci, dominant virulent clones are responsible for the majority of MenX disease. Most MenX carrier and disease isolates recovered in the African meningitis belt belonged to cc181, which has been circulating in Africa since the 1970s [43].

#### *2.3. Group Y*

Group Y meningococci (MenY) are frequently recovered from the nasopharynx but have historically considered less invasive than groups A, B, and C [44]. However, in the mid-1990s, the rates of IMD due to MenY increased in the United States [45], and subsequently in several European countries [46–49] as well as Israel, South America, and South Africa. Clonal complexes cc23, cc167, and cc175 have been linked to the majority of MenY IMD, but MenY IMD has also been seen with cc22 (Europe), cc174 (the United Kingdom), cc92 (Europe and South America), and cc103 (Europe).

In the mid-1990s, MenY (cc23 and cc167) emerged as a major cause of significant sporadic and hyperendemic disease in the United States. The proportion of MenY IMD cases in the United States was 2% during 1989–1991 [50], increased to 10.6% in 1992, and increase to 32.6% of reported cases in 1996 [51]. Subsequently, the proportion of MenY cases decreased in the United States [45], although still causing 15% of IMD in 2018. The increase in MenY cases has not been as prominent in neighboring Canada [52]. In 1998 at the peak of MenY incidence in the United States, a carriage study of high school students from counties in the metropolitan area of Atlanta, GA, found the rate of meningococcal carriage to be 7.7% and of these isolates, 48% were MenY [53]. However, in 2006–2007, a similar carriage study in high school students found a much lower carriage rate of <3% and a lower proportion of MenY carriage [54]. Thus, like MenX, high rates of acquisition and carriage were associated with increased disease and lower MenY carriage correlated with the decrease in MenY IMD cases [45].

Meningococcal quadrivalent conjugate vaccines against groups A, C, Y, and W (Men-ACWY) were licensed in the United States beginning in 2005 and coverage has steadily increased among children aged 13–17 years, from 11.7% in 2006 to 86.6% in 2018. A study comparing group distribution of IMD isolates prior to (2000–2005) and post vaccine introduction (2006–2010) reported that among all age groups, the overall IMD incidence declined over time, but there was no evidence of vaccine-induced capsular group replacement. While the incidence of IMD significantly declined in the United States, the proportion of MenY varied from 33% in 2000–2005 and 37% in 2006–2010 to 27% in 2011–2015 [55,56]. Changes in group and clonal complex were observed in isolates of both vaccine targeted and non-targeted groups. These changing profiles are likely representative of natural variation and fluctuations within meningococcal population structure. As noted, clonal complexes cc23 and cc167 accounted for most of MenY disease in both the United States and Canada [55,57,58], again suggesting that closely related strains circulate at high frequencies in a community causing sporadic disease.

MenY disease has recently emerged in Latin American countries and is characterized by clear differences from country to country [59]. Molecular characterization of MenY IMD isolates during 2000–2006 showed variable trends among 5 countries. While no increase in the frequency of MenY isolates was observed in either Brazil or Chile, the proportion of MenY IMD isolates increased in Argentina from 2002, to a level similar to those of groups C and W by 2006. In Colombia, MenY IMD isolates increased from 4% in 2000 to 50% in 2006 [60]. Venezuela also reported an increase in the proportion of cases due to MenY in 2006, representing 50% of all cases identified [59]. Again, most of the IMD isolates belonged to cc23 and cc167 [61]. Recently, IMD cases caused by penicillin- and ciprofloxacin-resistant cc23 MenY were found in El Salvador. These isolates contained a β-lactamase gene (*blaROB-1*) and a mutated DNA gyrase gene (*gyrA*) [62].

Until the last decade, MenY cases were rare in Europe, accounting for <2% of cases [63]. An emergence of MenY IMD cases was noted in several European countries after 2010. For example, in France, MenY accounted for 3% of IMD cases in 2000 to 2005, but increased to 10% in 2013 [64,65]. In Scotland, MenY IMD cases increase from 2.3% in 2010 to 17% in 2013 [65]. Further, significant increases in the incidence and the relative proportion of MenY IMD cases were found in Scandinavian countries: in Norway, the 4-year trend between 2010 and 2013 for Norway is 31–55–25–26% and in Finland, it was 38–21–24–40% [48,65]. Sweden had the highest relative proportion of MenY IMD in Europe—39% in 2010 and ~50% in the following 3 years [65]. The significantly increased MenY IMD in Sweden is mainly due to the emergence of specific cc23 clusters [47,48]. Whole-genome sequencing data of invasive MenY isolates from 1995 to 2012 in Sweden found at least three related but distinct cc23 clusters causing disease in Sweden. Thus, the increase in MenY IMD cases was not caused by the expansion of a single virulent variant [47], but was linked to increased virulence, host adaptive immunity, and transmission dynamics. Comparison to a collection of MenY isolates from England, Wales, and Northern Ireland during 2010 to 2012, which had relatively low MenY incidence, and MenY from the United States showed that the MenY cc23 clusters have a distribution spanning North America and Europe, including Sweden, over a number of years [47] but different strain types were prevalent in each geographic region.

Several carriage studies conducted in the United Kingdom have detected changes in MenY carriage in young adults over the last three decades [66–69]. MenY constituted approximately 8% of recovered isolates when carriage was assessed during 1997–1998 in first-year university students at Nottingham University, the United Kingdom [66]. During 1999–2001 in >48,000 samplings of 15–17 years old throughout the United Kingdom, MenY strains accounted for ~10% of the carriage isolates [70]. A later 2008 carriage study of first-year students carried out again at Nottingham University, the United Kingdom, found that MenY carriage reached 26% [68]. Core genome analysis of carriage-associated MenY isolates recovered in the United Kingdom during 1997–2010 reveals extensive genetic similarities to disease-associated MenY recovered during 2010–2011 [5]. Again, the majority

of these MenY belong to cc23 (58% in carriage and 79% in disease) and a long-term temporal stability of MenY clones was suggested [5]. However, in South Africa, a different clonal complex was responsible for increases in MenY disease. MenY cc175 caused significant IMD and was dominant in South Africa in the early 2000s [58]. MenY is also expressed in cc11, cc32, and cc41/44 clonal complex backgrounds that are more frequently associated with other capsular groups. Interestingly, comparison of IMD cases during a 2-month lockdown period in 2020 and the same periods of 2018 and 2019 in France found significant decrease in all IMD cases from prior years, and seemed to mainly involve IMD cases due to groups B and C and W, but not IMD due to group Y and other groups or nongroupable isolates [11]. The MenY genotypes had not changed in 2020 and were cc23 [11]. The observed IMD decreased mainly in the highly transmissible and hyperinvasive isolates belonging to cc11.

#### *2.4. Group E*

IMD due to meningococcal group E (MenE), previously known as 29E and Z' and first identified in 1968 [71], is infrequent and has been most often associated with immunocompromised patients [71,72]. Query of the PubMLST database shows 1003 MenE isolates in the collection. The vast majority (>75%) were from pharyngeal carriers with cc60 (42%) and cc1157 (29%) dominating. The earliest invasive MenE recorded in PubMLST is in 2000, and are predominantly cc60 (33%), cc1157 (26%), cc254 (8%), and cc178. A recent study reported the molecular characterization of three MenE IMD cases in Queensland, Australia; the emergence of these cases was attributed to a circulating cc1157 clone [73]. Globally as noted, MenE carriage is not uncommon. Historic carriage studies of first-year college students in the United Kingdom in 1997 [66] and young adults in the Czech Republic during 1993 [74] found ~6% and ~5% MenE in the respective isolate collections. A 2008 carriage study of first-year students at Nottingham University, the United Kingdom, found MenE clones highly prevalent (21–32%) in residential halls, indicative of rapid clonal expansion [67]. However, a recent carriage study performed in Australia in 2017 identified a single individual with MenE carriage from 421 first-year university students (0.2%) [75]. In contrast, an ongoing study of meningococcal carriage in participants in an STI clinic, MenE was identified in ~13% (Tzeng et al. unpublished data) of carriage isolates. While no group-specific vaccine is currently available for MenE, the protein-based group B vaccine, MenB-4C (Bexsero), contains outer membrane vesicles with multiple surface antigens that can provide cross-reactive protection. Similar data are available for MenB-FHbp (Trumenba) where bactericidal responses to groups C, W, Y, and X expressing different fHbp peptides have been shown.

#### *2.5. Nongroupable*

The meningococcal nongroupable (MenNG) phenotype is a result of elimination or minimal capsule production. Responsible mechanisms include down-regulation of capsule gene expression, phase variation in the capsule synthesis genes, transient or permanent inactivation of genes by insertion element movement into the capsule gene cluster (*cps*), frame-shift point mutations within an otherwise intact biosynthesis genes, or transformation/recombination events resulting in major deletion of the *cps* locus [76–78]. Meningococci with a capsule null locus (*cnl*), similar to *N. gonorrhoeae*/*N. lactamica*-like genetic configuration at the *cps* locus, were first identified in healthy carriers in Germany in 2000 and constituted ~16% of all recovered isolates [78]. Subsequently, additional carriage studies showed that nongroupable *cnl* meningococci are prevalent in carriage [79,80].

MenNG rarely cause invasive disease; however, *cnl* isolates have been described as a cause of IMD in immunocompetent individuals [81–85]. Most invasive *cnl* meningococci belong to cc198 [81–83] and cc192 [84,85], with cc192 being most commonly identified in Africa, but rarely elsewhere in the world [79]. Two cc198 invasive *cnl* isolates from Canada were examined using murine intraperitoneal infection model. Although no mortality was seen upon infection with the non-encapsulated MC58 derivative, 18% succumbed to infection with one *cnl* strain and 50% died after infection with the other *cnl* strain. Thus, although virulence potentials of both *cnl* strains were below that of encapsulated strain MC58, both strains exhibit a virulence phenotype [83].

While case reports in the literature across several decades have indicated that *N. meningitidis* is capable of colonizing the urogenital tract and causing sporadic cases of urethritis, cervicitis, or proctitis, very low overall incidence has been reported [86–89]. In one study of 23 meningococci isolated from the urogenital tract and rectum, two are *cnl* isolates [90]. Another collection of 39 urethritis-associated *N. meningitidis* identified 4 *cnl* isolates and 17 MenNG isolates due to various mutations in the *cps* locus [91]. More recently, a meningococcal clade of cc11.2 lineage (US\_NmUC) with a nongroupable phenotype due to deletion of capsule biosynthesis genes has caused unprecedented clusters of meningococcal urethritis in heterosexual men [6,8,92]. As one example, in Columbus, Ohio from 2015 to 2016, ~25% of presumed gonococcal urethral infections were determined to be meningococcal urethritis with clinical presentation mirroring that of gonococcal urethritis [8,92]. Other mucosal infections, e.g., neonatal conjunctivitis, [93] and at least five cases of IMD (meningitis and meningococcemia) were also reported with this clade [6], although it is not known if these patients were immunocompetent.

While urogenital colonization and sporadic cases of urethritis caused by *N. meningitidis* are documented across many genotypes and groups [90,91,94], the US\_NmUC is unique in its capability of causing multicity epidemiologically unlinked urethritis clusters and US\_NmUC appears to be sexually transmitted, like gonococci [8,92]. These observations suggest that a phylogenetically distinct nongroupable cc11 US\_NmUC has emerged as a new urotropic pathotype to cause meningococcal urethritis. Specific signatures universal to the US\_NmUC include (1) an IS1301-mediated specific deletion of the group C capsule biosynthesis genes, (2) expression of a unique FHbp ID896 protein [95], and (3) the acquisition of gonococcal NorB-AniA denitrification apparatus [7]. These unique features differ from other urogenital meningococcal isolates, many of which express capsule, encode a frame-shifted *fHbp* allele, and have a meningococcal denitrification pathway [91,94]. Loss of capsule has been demonstrated to enhance colonization at the mucosal surface, confer increased invasion into epithelial cells [96], and facilitate biofilm formation [97,98]. The acquisition of gonococcal denitrification pathway likely contributes to the success of this clade in better adapting to the male urethra.

Enhanced national surveillance and whole genome sequencing analysis of invasive, urogenital and rectal isolates at CDC has identified ~300 isolates from over 13 states to be members of US\_NmUC [6]. This is certainly an underestimate due to possible misidentification by *N. gonorrhoeae* diagnostic assay [99]. In 2019, two US\_NmUC isolates were reported in MSM in the United Kingdom [100]. One of the UK isolates had acquired a frameshifted gonococcal maltose phosphorylase gene, resulting in a carbohydrate utilization profile more typically associated with gonococci [100]. Ecological separation within the human host is proposed as an explanation for the lower frequency of interspecies recombination noted between naturally competent *N. meningitidis* and *N. gonorrhoeae* [101]. However, among the members of the US\_NmUC, the genome content of total length of DNA sequences inferred to have originated from *N. gonorrhoeae* varied substantially from ~5 to ~30 kb [6]. Further, one 2015 isolate had a gonococcal-like *mtrR* allele that is associated with elevated azithromycin MICs [6] and 7 out of 10 isolates recovered during 2018–2019 in St. Louis, MO, are non-susceptible to azithromycin [99]. In addition, intermediate penicillin resistance was seen in the clade, and one UK isolate, having acquired part of a gonococcal DNA gyrase (*gyrA*) gene, was resistant to ciprofloxacin [100]. These multiple recombination events demonstrated widespread acquisition of gonococcal DNA by US\_NmUC and suggested that co-colonization of these two species had facilitated genetic exchanges and raised the prospect of further acquisition of gonococcal antibiotic resistance determinants [100].

#### **3. Population Structure of Invasive "Minor" Capsular Groups and Nongroupable** *N. meningitidis* **3. Population Structure of Invasive "Minor" Capsular Groups and Nongroupable** *N. meningitidis*

~5 to ~30 kb [6]. Further, one 2015 isolate had a gonococcal-like *mtrR* allele that is associated with elevated azithromycin MICs [6] and 7 out of 10 isolates recovered during 2018– 2019 in St. Louis, MO, are non-susceptible to azithromycin [99]. In addition, intermediate penicillin resistance was seen in the clade, and one UK isolate, having acquired part of a gonococcal DNA gyrase (*gyrA*) gene, was resistant to ciprofloxacin [100]. These multiple recombination events demonstrated widespread acquisition of gonococcal DNA by US\_NmUC and suggested that co-colonization of these two species had facilitated genetic exchanges and raised the prospect of further acquisition of gonococcal antibiotic re-

*Microorganisms* **2021**, *9*, 519 8 of 16

sistance determinants [100].

The genetic relationships of W, X, Y, E, and NG capsular groups associated with disease are shown in Figure 1. Genome allelic profile comparisons were made based on the core genome MLST (cgMLST v1.0) scheme with a set of 1605 loci present in ≥95% *N. meningitidis* isolates [102]. The analysis examines 1158 disease-causing isolates: 29 MenE, 575 MenW, 39 MenX, 453 MenY, and 62 MenNG (48 *cnl* and 14 urethritis clade) isolates, compiled by selecting a representative isolate that had a unique country/year/ST combination. Isolates not assigned to a clonal complex or without records of year and country origin were excluded. The resultant minimum spanning tree was visualized by GrapeTree [103], which is integrated into the BIGSdb functionality [104]. As shown, the geographically and temporally diverse collection of disease-causing X, E, *cnl,* and urethritis isolates displayed two distinct major groupings that are dominated by the W:cc11 and Y:cc23 isolates. However, the emergence of multiple distinct clonal lineages—cc11, cc22, cc174, and cc865 in MenW; cc181 and cc750 in MenX; cc60 and cc1157 for MenE; and the distinct cc11 urethritis clade—is evident. The genetic relationships of W, X, Y, E, and NG capsular groups associated with disease are shown in Figure 1. Genome allelic profile comparisons were made based on the core genome MLST (cgMLST v1.0) scheme with a set of 1605 loci present in ≥95% *N. meningitidis* isolates [102]. The analysis examines 1158 disease-causing isolates: 29 MenE, 575 MenW, 39 MenX, 453 MenY, and 62 MenNG (48 *cnl* and 14 urethritis clade) isolates, compiled by selecting a representative isolate that had a unique country/year/ST combination**.**  Isolates not assigned to a clonal complex or without records of year and country origin were excluded. The resultant minimum spanning tree was visualized by GrapeTree [103], which is integrated into the BIGSdb functionality [104]. As shown, the geographically and temporally diverse collection of disease-causing X, E, *cnl,* and urethritis isolates displayed two distinct major groupings that are dominated by the W:cc11 and Y:cc23 isolates. However, the emergence of multiple distinct clonal lineages—cc11, cc22, cc174, and cc865 in MenW; cc181 and cc750 in MenX; cc60 and cc1157 for MenE; and the distinct cc11 urethritis clade—is evident.

**Figure 1.** Meningococcal invasive isolates of capsule groups E, W, X, Y, nongroupable *cnl,* and urethritis clade isolates that have whole genome sequencing data and "disease" record entries of "invasive," "meningitis," "septicemia," or "meningitis and septicemia" were retrieved from PubMLST database (http://pubmlst.org/neisseria/, access on 2 March 2021). Isolates without records of "year" and "clonal complex" were excluded. A single isolate that has a unique combination of country/year/sequence type (ST) definition was selected and the clonal complexes with at least two isolates were included in the analysis. The minimum spanning trees were generated and visualized with GrapeTree [103], a plug-in analysis tool at PubMLST, with the scheme of *N. meningitidis* core genome MLST (cgMLST) v 1.0 and default parameters. The trees are **Figure 1.** Meningococcal invasive isolates of capsule groups E, W, X, Y, nongroupable *cnl,* and urethritis clade isolates that have whole genome sequencing data and "disease" record entries of "invasive," "meningitis," "septicemia," or "meningitis and septicemia" were retrieved from PubMLST database (http://pubmlst.org/neisseria/, access on 2 March 2021). Isolates without records of "year" and "clonal complex" were excluded. A single isolate that has a unique combination of country/year/sequence type (ST) definition was selected and the clonal complexes with at least two isolates were included in the analysis. The minimum spanning trees were generated and visualized with GrapeTree [103], a plug-in analysis tool at PubMLST, with the scheme of *N. meningitidis* core genome MLST (cgMLST) v 1.0 and default parameters. The trees are colored by capsular groups (**left**) or clonal complexes (**right**). The clonal complex breakdowns of each capsule groups are also listed.

#### **4. Meningococcal Vaccines**

Capsular polysaccharides have historically been the targets for group-specific meningococcal vaccine development. The first capsular polysaccharide vaccines were developed in the early 1970s. While a major step forward, they were generally not effective for children less than 2 years and failed to induce long-term memory responses. The subsequent development and introduction of polysaccharide–protein conjugate vaccines in the late 1990s markedly accelerated the prevention of meningococcal disease. Meningococ-

cal polysaccharide–protein conjugate vaccines against groups A, C, Y, and W, developed as monovalent, bivalent, or quadrivalent products, have considerably greater effectiveness than the polysaccharide vaccines and induce immune memory responses [105]. The polysaccharide conjugate vaccines reduce transmission by prevention of meningococcal acquisition, resulting in significant herd or community protection at quite modest levels of vaccine coverage. The introduction of a monovalent MenC conjugate vaccine in 2000 virtually eliminated the incidence of MenC disease in United Kingdom, an effect that has persisted for well over two decades, demonstrating 90% effectiveness at 3 years in 11–18-year-olds [106]. Vaccination against MenC induced herd protection and reduced the rates of MenC carriage and disease in non-vaccinated individuals by more than 50% [107]. A MenACWY conjugate vaccines was first licensed in the United States in early 2005. Additional quadrivalent MenACWY conjugate vaccines were subsequently introduced and are now in use globally (Table 1). The group B capsule has not been developed as a vaccine target given its structural similarity with human polysaccharide antigens. However, two outer membrane protein-based vaccines targeting MenB (Table 1), also with activity against non-MenB *N. meningitidis,* have been licensed and are now in use in meningococcal disease prevention strategies.

**Table 1.** Meningococcal protein andpolysaccharide conjugate vaccines \*.


Abbreviations: MenACWY-CRM = meningococcal groups A, C, W, and Y capsular polysaccharide-diphtheria CRM<sup>197</sup> conjugate; -D = diphtheria toxoid conjugate; -TT = tetanus toxoid conjugate vaccine; MenB-4C = fourcomponent meningococcal group B vaccine; MenB-FHbp = meningococcal group B bivalent factor H binding protein vaccine. # MenQuadfi is indicated for <sup>≥</sup>2 years in the U.S. Bexsero is licensed for 10–25 years in the U.S. \* Monovalent C conjugate vaccines, including Meningitec (MenC-CRM197), Menjugate (MenC-CRM197), NeisVac-C (MenC-TT), and Menitroix (MenC-TT+Hib), are still in use in some countries. Additional conjugate vaccines directed against MenAC and MenC are also available in China. Pentavalent meningococcal conjugate vaccines: MenABCWY (MenACWY-CRM-197 combined with MenB multicomponent recombinant proteins, GlaxoSmithKline), MenABCWY (bivalent FHbp-containing pentavalent vaccine, Pfizer), and MenACXWY (NmCV-5, Serum Institute of India) are in phase 3 clinical trials.

However, meningococcal vaccination strategies with limited capsular group coverage will eventually select for or uncover previously "minor" capsular groups, now causing significant endemic and epidemic meningococcal disease in the last two decades (e.g., groups W, X, and Y). Examples are the Hajj MenW outbreaks and the emergence of MenW disease in South America and Europe. In addition, the dramatic control of MenA disease in the African meningitis belt achieved by the introduction of MenAfriVac in 2010 uncovered outbreaks of MenX and MenW [39]. Due to horizontal gene transfer and recombination *N. meningitidis,* like *Streptococcus pneumoniae,* can undergo capsule structural change, e.g., "capsule switching" [55,108–111] lessening herd immunity. Transformation and homologous recombination of capsule genes with the appearance of otherwise identical MenC strains was first noted during a prolonged MenB outbreak in the 1990s [108]. The MenW outbreaks associated with the Hajj in 2000 may have been the result of a historic capsule switching event from cc11 MenC strains. In large meningococcal isolate collections, capsule switching is detected in ~3% of isolates [112].

Pentavalent meningococcal conjugate vaccines, i.e., MenABCWY (MenACWY-Oligosaccharide diphtheria CRM<sup>197</sup> conjugate, combined with MenB multicomponent recombinant, GlaxoSmithKline), MenABCWY (bivalent FHbp-containing pentavalent vaccine, Pfizer), and MenACXWY (NmCV-5, Serum Institute of India), are in phase 3 clinical trials and are

a next step, if widely implemented, for global control of meningococcal disease. The broad capsule focused coverage together with the MenB protein component(s) can potentially provide protection [113] against other minor disease-causing groups and nongroupable strains [95,114,115]. The broad capsule focused coverage together with the MenB protein component(s) can potentially provide protection [113] against other minor disease-causing groups and nongroupable strains [95,114,115]. **5. Conclusions** 

Pentavalent meningococcal conjugate vaccines, i.e., MenABCWY (MenACWY-Oligosaccharide diphtheria CRM197 conjugate, combined with MenB multicomponent recombinant, GlaxoSmithKline), MenABCWY (bivalent FHbp-containing pentavalent vaccine, Pfizer), and MenACXWY (NmCV-5, Serum Institute of India), are in phase 3 clinical trials and are a next step, if widely implemented, for global control of meningococcal disease.

disease in South America and Europe. In addition, the dramatic control of MenA disease in the African meningitis belt achieved by the introduction of MenAfriVac in 2010 uncovered outbreaks of MenX and MenW [39]. Due to horizontal gene transfer and recombination *N. meningitidis,* like *Streptococcus pneumoniae,* can undergo capsule structural change, e.g., "capsule switching" [55,108–111] lessening herd immunity. Transformation and homologous recombination of capsule genes with the appearance of otherwise identical MenC strains was first noted during a prolonged MenB outbreak in the 1990s [108]. The MenW outbreaks associated with the Hajj in 2000 may have been the result of a historic capsule switching event from cc11 MenC strains. In large meningococcal isolate collec-

*Microorganisms* **2021**, *9*, 519 10 of 16

tions, capsule switching is detected in ~3% of isolates [112].

#### **5. Conclusions** Capsular groups W, X, and Y now cause significant IMD as reflected in the higher

Capsular groups W, X, and Y now cause significant IMD as reflected in the higher numbers of invasive isolates deposited into PubMLST since 2000 (Figure 2) as well as the country and global surveillance data noted above. In addition, group E and nongroupable meningococci have appeared as a cause of invasive disease, and a nongroupable *N. meningitidis* pathotype of the hypervirulent cc11 is causing sexually transmitted urethritis cases and outbreaks. *N. meningitidis* is a human microbe circulating within populations. Due to factors including the introduction of highly effective meningococcal vaccines of limited coverage, the capsular groups causing IMD has changed over time and across geographic regions. Pentavalent meningococcal conjugate vaccines in phase 3 clinical trials appear to be an important next step for enhanced global control. However, the capacity of meningococci to continue to evolve is significant. Genetic transformation and recombination, including transfer of genes between meningococci, gonococci, and commensal *Neisseria* spp. [6,116] and immune selection can all result in the rise, diversification, and disappearance of virulent meningococcal clones. Continued surveillance including molecular characterization is key to recognizing the changing epidemiology of meningococcal disease. numbers of invasive isolates deposited into PubMLST since 2000 (Figure 2) as well as the country and global surveillance data noted above. In addition, group E and nongroupable meningococci have appeared as a cause of invasive disease, and a nongroupable *N. meningitidis* pathotype of the hypervirulent cc11 is causing sexually transmitted urethritis cases and outbreaks. *N. meningitidis* is a human microbe circulating within populations. Due to factors including the introduction of highly effective meningococcal vaccines of limited coverage, the capsular groups causing IMD has changed over time and across geographic regions. Pentavalent meningococcal conjugate vaccines in phase 3 clinical trials appear to be an important next step for enhanced global control. However, the capacity of meningococci to continue to evolve is significant. Genetic transformation and recombination, including transfer of genes between meningococci, gonococci, and commensal *Neisseria* spp. [6,116] and immune selection can all result in the rise, diversification, and disappearance of virulent meningococcal clones. Continued surveillance including molecular characterization is key to recognizing the changing epidemiology of meningococcal disease.

**Figure 2.** *N. meningitidis* capsular groups E, W, X, Y, and cnl from invasive meningococcal disease and urotropic meningococci submitted to PubMLST database, 1991–2020.

**Author Contributions:** Y.-L.T. and D.S.S. wrote and reviewed the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported in part by the NIH Grants R01AI127863 and R21AI128313.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** This publication made use of the *Neisseria* Multi Locus Sequence Typing and analysis tools hosted on Bacterial Isolate Genome Sequence Database (BIGSdb) (http://pubmlst.org/ neisseria/ accessed on 20 November 2020).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Review* **Long Term Impact of Conjugate Vaccines on** *Haemophilus influenzae* **Meningitis: Narrative Review**

**Mary Paulina Elizabeth Slack**

Gold Coast Campus, School of Medicine & Dentistry, Griffith University, Southport, QLD 4222, Australia; m.slack@griffith.edu.au

**Abstract:** *H. influenzae* serotype b (Hib) used to be the commonest cause of bacterial meningitis in young children. The widespread use of Hib conjugate vaccine has profoundly altered the epidemiology of *H. influenzae* meningitis. This short review reports on the spectrum of *H. influenzae* meningitis thirty years after Hib conjugate vaccine was first introduced into a National Immunization Program (NIP). Hib meningitis is now uncommon, but meningitis caused by other capsulated serotypes of *H. influenzae* and non-typeable strains (NTHi) should be considered. *H. influenzae* serotype a (Hia) has emerged as a significant cause of meningitis in Indigenous children in North America, which may necessitate a Hia conjugate vaccine. Cases of Hie, Hif, and NTHi meningitis are predominantly seen in young children and less common in older age groups. This short review reports on the spectrum of *H. influenzae* meningitis thirty years after Hib conjugate vaccine was first introduced into a NIP.

**Keywords:** *Haemophilus influenzae*; Hib; impact of Hib conjugate vaccine; Hia; NTHi

**Citation:** Slack, M.P.E. Long Term Impact of Conjugate Vaccines on *Haemophilus influenzae* Meningitis: Narrative Review. *Microorganisms* **2021**, *9*, 886. https://doi.org/ 10.3390/microorganisms9050886

Academic Editor: James Stuart

Received: 17 March 2021 Accepted: 19 April 2021 Published: 21 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

*Haemophilus influenzae* is a small, pleiomorphic Gram-negative coccobacillus, which is restricted to humans. It is fastidious in its growth requirement, only growing in culture media supplemented with both X factor (hemin) and V factor (nicotinamide adenine dinucleotide, NAD), for example chocolate agar. *H. influenzae* strains can be differentiated into two major groups: capsulated and non-capsulated strains (generally referred to as non-typeable strains, NTHi). The capsulated strains are further divided into six groups (a to f) based on the chemical structure of their polysaccharide capsules [1]. The most virulent type of *H. influenzae* is type b (Hib) and the major virulence determinant of Hib is its polysaccharide capsule, composed of polyribosyl ribitol phosphate (PRP).

*H. influenzae* colonizes the nasopharynx [2] and to a lesser extent the conjunctivae [3] and genital tract [4–6]. The respiratory tract is mainly colonized by *H. influenzae* and to a lesser extent *H. parainfluenzae* [2]. Approximately 80% of individuals carry NTHi strains in the nasopharynx, while 3–5% carry capsulated strains in the upper respiratory tract [7,8]. Spread from one person to another occurs via respiratory droplets or by direct contact with secretions [4].

Before the introduction of Hib conjugate vaccines, Hib was the commonest cause of bacterial meningitis in young children in the United States [9,10], Sweden [11], Iceland [12], the Netherlands [13], and England and Wales [14]. Seventy five percent of Hib meningitis cases occurred in children between the ages of three months and three years [15,16]. The case fatality ratio of Hib meningitis was ~5 to 10% in high-income countries [17].

In 1933, Fothergill and Wright [18] reported that blood from children aged less than two years lacked bactericidal activity against Hib, whereas blood from older children and adults demonstrated bactericidal activity. They speculated that naturally acquired antibodies to Hib were protective and as the mean level of Hib antibodies increased through exposure to the organism, so Hib meningitis incidence declined. The paucity of cases of Hib meningitis in infants aged <two months correlates with the presence of maternal Hib antibodies. This was confirmed by Peltola et al. [19] who demonstrated the incidence of Hib meningitis declined as the mean level of anti-Hib antibodies increased. Studies on un-immunized individuals established a putative short-term correlate of protection against Hib infection of ≥0.15 µg/mL anti-PRP antibodies [20]. Later studies established that an anti-PRP antibody titer of ≥1.0 µg/mL was required for long-term protection [21].

It is now more than three decades since Hib conjugate vaccines were first developed and a variety of vaccine formulations, with a Hib component, are now included in the NIP of almost all countries in the world. Wherever Hib conjugate vaccine has been used the epidemiology of *H. influenzae* meningitis has changed, with Hib meningitis now infrequently seen in young children [22]. However, *H. influenzae* serotype a (Hia) has emerged as a significant cause of meningitis in Indigenous children in North America [23], and nontypeable strains of *H. influenzae* (NTHi) are associated with invasive infections, including meningitis, in neonates, older adults, and other vulnerable patient groups [24]. In 2020, the World Health Organization (WHO) published the document" Defeating meningitis by 2030: a global road map" [25]. The aims of the road map include the reduction of cases and deaths from vaccine-preventable meningitis; introduction of new vaccines; increasing vaccine coverage; and improving surveillance and advocacy. This short review will review the current epidemiology of *H. influenzae* meningitis in the second decade of the twenty first century to assess the progress made to date in achieving the goals set out in this document.

#### **2. Method**

A PubMed search was performed to identify published papers on the epidemiology of *H. influenzae* meningitis, before and after the introduction of Hib conjugate vaccines, using the terms: (((invasive) AND haemophilus) AND influenzae) AND ("meningitis" OR "nontypable" OR "NTHi" OR "serotype a" OR "serotype b" OR "serotype c" OR "serotype d" OR "serotype e" OR "serotype f" OR "non-b" OR "Hib") AND (epidemiology OR "burden" OR "risk factor" OR "impact" OR "Hib vaccine" OR "Hib conjugate vaccine" OR "surveillance" OR "review" OR "clinical" OR "outcome" OR "neonate" OR "adult" OR "children") for papers published between 1985 and 2020. Relevant papers on *H. influenzae* meningitis were reviewed.

#### **3. Global Burden of Hib Meningitis before the Introduction of Routine Hib Immunization**

Acute meningitis was the most serious presentation of Hib infection, following invasion of the blood stream. Often the child initially developed upper respiratory tract symptoms or otitis media before signs of meningeal involvement [4]. Before the introduction of Hib conjugate vaccines, *H. influenzae* serotype b (Hib) was the commonest cause of bacterial meningitis in young children in the US [9], Sweden [11], Iceland [12], the Netherlands [13], and the UK [14]. The mean annual incidence of Hib meningitis in the US was 54/100,000 (range 19–69/100,000) in children aged < five years and ~120 to 130/100,000 in infants aged < 12 months [26]. Annual rates in Europe, Australia (non-Indigenous children) and South America ranged from <20 to 50/100,000 children aged <5 years [17]. A much higher rate was reported from The Gambia (60/100,000 < 5 years and 297/100,000 < one year of age) [27]. Incidence rates of 282/100,000, 254/100,000, 152/100,000, and 450/100,000 in children aged < five years were reported in Alaska Native [28], White Mountain Apache [29], Navajo Indian [30], and Indigenous Australian [31] children, respectively. A rate of 530/100,000 in children aged <five years was reported in the Keewatin District of Northern Canada, mostly afflicting Inuit children [32].

The majority of cases of Hib meningitis cases occurred in children aged between three months and three years [15,16]. The proportion varied in different parts of the world, with approximately 50%, 40%, and 80% of cases of Hib meningitis occurring in infants aged < 12 months in the US, Europe, and Africa, respectively [17]. In the US and Europe, the peak incidence occurred at eight to nine months of age with less than 10% of cases occurring before the age of six months, and approximately 40% of all cases of Hib meningitis occurred in the first year of life [33]. In Indigenous communities in North America and Australia, and in low and middle income countries (LMICs), the proportion

of cases of Hib meningitis occurring in the first six months of life was higher than in industrialized communities [33]. In Australia, the median age of onset of Hib meningitis (and the proportion of cases in the first 12 months of life) in Indigenous and non-Indigenous children was six months (60%) and 15 months (17%), respectively [34]. In The Gambia, 44% and 84% of cases occurred in the first six and twelve months of life respectively [33]. In Alaska Native children, 34% and 67% of cases of Hib meningitis occurred in the first six and twelve respectively [27] (Table 1).



Data derived from: USA [9,10]; North American (Indigenous) [28–30,35]; Europe [36–40]; Israel [41]; The Gambia (Reference [27]; Australia and New Zealand (non-Indigenous) [34,42]; Australia and New Zealand [31,43]; Latin America [44]; Asia [45]; and Mongolia [46].

> The mean case fatality ratio (CFR) of Hib meningitis ranged from approximately five to ten % in high-income countries to 28% in Africa [17]. Fifteen to 30% of survivors had long-term sequelae, including sensorineural hearing loss, intellectual impairment, epilepsy, cerebral palsy, or hydrocephalus [26,47–50]. Thirty eight percent of children who survived an episode of Hib meningitis in The Gambia had long-term sequelae [51].

#### **4. Hib Vaccines**

#### *4.1. Plain PRP Vaccine*

The first Hib vaccine was a plain polysaccharide vaccine consisting of PRP. It was used in a large field trial in Finland involving 100,000 children aged three months to five years [52]. Although efficacious in children >18 months, it did not induce protective levels of anti-PRP antibodies in children aged <18 months, i.e., those most at risk of Hib meningitis [20,53]. It also failed to have any impact on nasopharyngeal carriage of Hib and so had no impact on transmission [52]. Plain polysaccharide vaccines activate B cells via a T-cell independent pathway, which is poorly developed in children <18 months of age [54]. The antibody response is short-lived, mainly IgM with little isotype switching and no induction of immune memory [55].

#### *4.2. Hib Protein-Conjugate Vaccines*

In the late 1980s conjugate Hib vaccines were developed in which PRP was covalently linked to a protein carrier. The PRP-protein conjugate induces a T-cell dependent response, which develops at a much younger age in infants, who are able to respond to conjugate vaccines from the age of six to eight weeks [56]. The protein antigen encourages class switching from IgM to IgG via T-helper cells [55]. The IgG generated is predominantly IgG1, which in vitro induces complement-mediated opsonization and bacteriolysis. The antibodies produced are of a higher avidity than those produced by a plain polysaccharide vaccine [55]. Furthermore, PRP-conjugate vaccines have a marked impact on nasopharyngeal carriage of Hib [57]. By reducing nasopharyngeal carriage, transmission of Hib to other susceptible children and adults is interrupted, thereby reducing infection in other non-immunized groups. This is called "herd effect" or "herd protection".

Four different protein carriers were initially used for Hib conjugate vaccines: tetanus toxoid (PRP-TT), diphtheria toxoid (PRP-D), a non-toxic mutant *Corynebacterium diphtheriae* protein CRM 197 (PRP-CRM) and an outer membrane complex of *Neisseria meningitidis* (PRP-OMP) [58]. The different Hib vaccines were equally immunogenic in adults but elicited different responses in infants <18 months of age. PRP-D was the least immunogenic, generating antibody titers of ≥1.0 µg/mL in approximately 30% of infants after two or three doses [59]. This vaccine was subsequently withdrawn. PRP-OMP vaccine generated antibody titers ≥1.0 µg/mL in 70–80% of infants at two months of age [60] and was the preferred vaccine for use in Indigenous populations in North America and Australia, where there was a very high burden of disease in very young infants [60]. The PRP-TT and PRP-CRM vaccines were similar in their immunogenicity eliciting antibody titers ≥ 1.0 µg/mL after three priming doses [61]. Over time monovalent Hib conjugate vaccines have largely been replaced by combination vaccines, including a bivalent Hib + meningococcus serogroup C vaccine (Hib-MenC), and pentavalent and hexavalent vaccines, where Hib is combined with diphtheria toxoid (D), tetanus toxoid (T), pertussis whole cell (wP) or acellular (aP), and/or hepatitis B (HepB), and/ or inactivated polio vaccine (IPV).

#### **5. Introduction of Hib Conjugate Vaccine in National Immunization Programs (NIPs)**

Hib vaccine was introduced into the NIP of Finland in 1986 [52], followed by the US in 1987 [62]. In the early 1990s Hib vaccine was added to the NIP in many Western European countries. By 2004, Hib vaccine had been included in the NIP of all European countries and ≥90% high-income countries. The introduction of Hib vaccine into the NIP of LMICs has taken longer, because of several factors. These include a lack of local data on the burden of Hib disease as a result of the difficulties in culturing this fastidious organism, widespread use of antibiotics before collection of blood and cerebro-spinal fluid (CSF) samples for culture and the relatively high cost of the vaccine. In 2004 WHO and the Global Alliance for Vaccines and Immunization (GAVI) sought to address this. Vaccine probe studies [63], in which a randomized controlled trial assesses the difference in incidence of meningitis between children immunized with Hib vaccine and unimmunized children, and the Hib Rapid Assessment Tool (HibRAT) [64] provided data on the burden of Hib meningitis for many LMICs. In 2005, GAVI established the Hib Initiative to accelerate the introduction of Hib vaccine in GAVI-eligible countries [65]. In 2006, WHO recommended the use of Hib conjugate vaccines in all countries [66], thereby allowing GAVI-eligible countries to apply for Hib vaccine without the need to have local data on Hib disease burden. With these measures, the number of countries using Hib vaccine increased from 89/193 (46%) in 2004 to 158/193 (82%) in 2009 [67]. Hib vaccine has now been added to the NIP of all countries in the world, except China, where it is available in the private market and in the Russian Federation, where it is recommended for certain groups of children [68].

#### **6. Impact of Hib Conjugate Vaccine on Hib Meningitis**

Wherever Hib vaccine has been introduced there has been a significant and sustained decline in Hib meningitis [69–71]. In 2000, the global incidence of Hib meningitis was estimated to be 31 (uncertainty range (UR) 16–39) cases/100,000 children aged < five years [72]. The estimated incidence varied considerably by region (Table 2). At that time, the only regions that had widespread use of Hib vaccine were the Americas and Europe. A further analysis of the burden of Hib meningitis in 2000–2015 [73] estimated the global incidence of Hib meningitis had declined to five (UR 2–8) cases/100,000 children aged < five years. There were still regional variations, with the highest estimated incidences in the South East Asian and Western Pacific Regions, which may reflect the lack of introduction of Hib vaccine into some countries in these regions at that time.


**Table 2.** Estimated incidence and case fatality ratio of Hib meningitis (with uncertainty estimates) by WHO region in 2000 and 2015.

Data are estimates (uncertainty range) Incidence is /100,000 children aged <5 years. CFR: case fatality ratio. Data derived from: Watt et al. [61] and Wahl et al. [62].

> By 2015, the burden of Hib meningitis was limited to a small number of countries that had not yet or only recently introduced Hib vaccine in their NIP. In the six years since this study almost all countries have now introduced Hib vaccine and the global burden will have been further reduced. This excellent control depends on maintaining high coverage of Hib vaccine combined with on-going surveillance of all cases of Hib meningitis in all ages of patients.

> The estimated global CFR of Hib meningitis in 2000 was 43% (UR 23–55%), ranging from 22% (8–34%) in the Western Pacific Region to 67% (44–75%) in the African Region [72]. By 2015, the global CFR had declined to 19% (7–29%), ranging from 5% (2–8) in Europe and the Western Pacific Region to 61% (20–98%) in the African Region [73].

> In 2013, a systematic review of the impact of Hib conjugate vaccine on childhood meningitis mortality, estimated the dose-specific impact (one dose: relative risk, RR = 0.64, 95% CI 0.38–1.06; two doses; RR = 0.09, 95% CI 0.03–0.27; three doses: RR = 0.06, 95% CI 0.02–0.22) [74]. The relative risk (RR) or risk ratio is the ratio of the probability of meningitis in children vaccinated with Hib vaccine to the probability of meningitis in unvaccinated children. This review estimated that three doses of Hib vaccine would prevent 38–43% of childhood meningitis mortality [74].

> After the introduction of Hib immunization into several NIPs in the 1990s, the incidence of Hib meningitis declined rapidly [26]. Hib conjugate vaccines have proved to be highly effective in all countries, where there is sustained high coverage of the vaccine [75]. In the US active surveillance of invasive *H. influenzae* disease is undertaken in the Active Bacterial Core Surveillance (ABC) sites, coordinated by the Centers for Disease Control and Prevention (CDC). This surveillance system covers a population of over 42 million in five states and five metropolitan areas across the US [76]. In the 1990s, the rate of bacterial meningitis declined by 55% in the USA following Hib vaccine introduction [77]. Between 1998 and 2007, there were 187 cases of *H. influenzae* meningitis cases identified in the CDC ABC surveillance sites, 9.4% of cases were due to Hib. The overall incidence of *H. influenzae* meningitis declined between 1998–1999 and 2006–2007, from 0.12/100,000 population (95% CI, 0.09 to 0.17) to 0.08/100,000 (95% CI, 0.05 to 0.11) [77]. In 2018, only 38 cases of invasive Hib infection in children aged <five years (incidence 0.19/100,000) were notified throughout the US [78].The number of cases of Hib meningitis was not specified.

> In a population-based observational study in Finland, where Hib conjugate vaccine was introduced in 1986, there were 1361 reported cases of bacterial meningitis between 1995 and 2014. Four percent of cases were caused by *H. influenzae* (incidence 0.06/100,000 population) and 92% of the isolates were non-b [79].The median age of *H. influenzae* meningitis was 29 years. From 2004 to 2014 two of 26 *H. influenzae* isolates were Hib [79].

> Hib meningitis incidence declined by 72–83% at sentinel hospitals in Pakistan and Bangladesh, respectively, within two years of implementing nationwide Hib conjugate vaccination [80]. In a hospital-based multi-center prospective survey of bacterial meningitis in Turkey from 2015 to 2018, 994 cases of suspected bacterial meningitis in children, aged

one month to 18 years, were identified [81]. Three (2.4%) of the 125 culture-positive cases were caused by Hib. Hib conjugate vaccine was introduced in the Japanese NIP in 2013, although Hib vaccine had been available on a voluntary basis since 2008. A nationwide population-based surveillance of invasive *H. influenzae* diseases in children in Japan [82] identified 336 cases of *H. influenzae* meningitis between 2008 and 2017. Between 2008–2012 and 2013–2017 there were 336 and 6 cases of *H. influenzae* meningitis, respectively. No cases of invasive Hib meningitis have been identified since 2014.

Although Hib meningitis has been virtually eliminated in almost all countries with established immunization programs and high vaccine coverage, there have been a few examples of countries that have experienced a re-emergence of invasive Hib infections, including Hib meningitis.

#### **7. Resurgence of Hib Meningitis in Some Countries**

#### *7.1. Resurgence of Hib in the UK*

In the UK there was a resurgence in cases in the late 1990s. The UK introduced Hib vaccine in 1992 as a three-dose infant schedule of PRP-TT (at two, three, and four months) with no booster dose in the second year of life, together with a catch-up campaign for all children <five years of age. Hib infections declined rapidly in all age groups through direct and indirect (herd) protection. The incidence of invasive Hib disease in England and Wales declined from 22.9/100,000 children < five years in 1990 to 0.65/ 100,000 in 1998 [83]. From 1999 Hib infections began to increase, especially among toddlers, most of whom were fully immunized. After 1999, the incidence of Hib disease increased to 4.6/100,000 in children aged <five years [84], with many of the infections, including meningitis, occurring in toddlers [55]. Studies established that there was a greater than expected decline in Hib antibodies after primary immunization, which had been initially masked by the catchup campaign [85–87]. The catch-up campaign also contributed to indirect protection by reducing nasopharyngeal carriage. By 1998, all children aged <five years had received three priming doses of Hib vaccine in infancy. A single dose of Hib vaccine administered at the age of 12 months was more immunogenic than three doses given in infancy. Another factor was the use of a less immunogenic Hib combination vaccine with diphtheria, tetanus, and acellular pertussis (DTaP-Hib) in 2000–2001 [84,88]. The resurgence was controlled by the re-introduction of a whole-cell pertussis-containing Hib vaccine (DTwP-Hib) in 2002, an Hib booster campaign for toddlers in 2003, and the introduction of a routine 12-month Hib booster in 2006 [89,90].

Since that time Hib infections, including meningitis, have remained at a very low level in the UK, A review of invasive Hib infections in England and Wales, between 2009 and 2012, identified only 14 cases in 2012 [22]. Hib incidence was 0.06/100,000 (two cases) in children aged <five years [22]. Most of the cases that occurred over those four years were in adults (73%), many of whom had underlying comorbidities and presented with pneumonia (56%) [22]. The Hib-associated case fatality rate was 9.4% (10/106 cases) [22]. There were 20 cases (18.9%) of meningitis: ten in children aged < one year; five in children aged one to five years, two in adults aged 20 to 44 years, two in adults aged 45 to 64 years and one case in an older adult aged ≥65 years [22]. There was only one death in the vaccine-eligible age cohort: a child with Hib meningitis who was partially vaccinated and had a complement deficiency [22]. Hib meningitis is now uncommon in the UK.

The current Hib vaccination program in the UK is hexavalent vaccine (DTaP-Hib-HepB-IPV) administered at two, three, and four months, with a 12-month booster dose of Hib-MenC vaccine [91]. The number of cases of invasive Hib infection is at a very low level, with only five cases of invasive Hib disease (cases that were meningitis not specified) in the vaccine eligible population in 2017–2018 [92].

#### *7.2. Resurgence of Hib in South Africa*

South Africa introduced Hib conjugate vaccine (PRP-TT) in 1999 as an early accelerated schedule of three doses at six, ten, and fourteen weeks without a booster dose in the second year of life [93]. The number of cases of invasive Hib infection initially declined, but from 2005 increasing number of cases in fully vaccinated children were detected [94]. Despite high vaccination coverage the detection rate of invasive Hib infection in children aged < five years increased from 0.7/100,000 in 2003 to 1.3/100,000 in 2009 (*p* < 0.001), and 135/263 (51%) of cases in children with known vaccination status were Hib vaccine failures [93]. From 2003 to 2009 the surveillance program (GERMS) identified 349 cases of invasive Hib infection in children aged <five years, of which 211 (60%) presented as meningitis [94] with a CFR of 19%. Fifty-five% of the children, where HIV status was documented, were HIV negative. Following the addition of a booster dose of Hib vaccine in 2009, as a pentavalent vaccine (DTaP-Hib-IPV) the incidence of invasive Hib declined [92]. In 2018, GERMS identified 327 cases of invasive *H. influenzae* infection, of which 201 were available for typing. Seventeen percent (34/201) were Hib, of which eight cases presented with meningitis [95].

### *7.3. Resurgence of Hib in the Gambia*

The Gambia introduced Hib vaccine in 1997. Before the Gambia introduced routine Hib vaccination, Hib meningitis incidence was 297/100,000 in infants <one year of age and 60/100,000 in children aged <five years [27]. The Gambia used a three-dose primary series of PRP-TT Hib vaccine, administered at two, three, and four months without a booster dose. For 14 years invasive Hib disease was well controlled in this country with consistently high coverage, low carriage rates and high levels of protective antibodies [96]. On-going surveillance in eastern Gambia identified an increase in Hib infections between 2011 and 2013, with 17 cases of invasive Hib infection, including 14 cases of Hib meningitis [97]. Although the reason for this re-emergence is not entirely clear, it does emphasize the importance of on-going surveillance.

#### *7.4. Is a Booster Dose of Hib Vaccine Needed?*

Although these instances where invasive Hib infections have emerged were in countries using a three dose primary series of Hib vaccine without a booster dose, Kenya and most LMICs use this schedule with no evidence of a resurgence of invasive Hib cases [98]. A three dose primary series of Hib vaccine without a booster dose is recommended by WHO [66]. A meta-analysis of 20 RCTs, conducted in 15 countries comparing different Hib vaccination schedules (3 + 0, 3 + 1, and 2 + 1) and different intervals between the primary, and the primary and booster doses, concluded that there was no difference between the schedules in terms of preventing invasive Hib disease, clinical effectiveness or immunologic response. All of the schedules protected against Hib infection and local epidemiology should determine the schedule, with three doses in the first six months of life being more appropriate where the greatest burden of Hib infection is in the first year of life, as in sub-Saharan Africa. Where the burden of infection occurs at a later age, the third dose could be given in the second year of life. In countries like the UK, where Hib infections resurged with a 3 + 0 schedule, a booster in the second year of life may be required [99]. Children who are HIV infected may require a booster dose of vaccine [100].

#### **8. Current Burden of** *H. influenzae* **Meningitis**

When Hib vaccine was first introduced there were concerns that Hib meningitis might be replaced by infections caused by other serotypes of *H. influenzae*. This has generally not happened, except in the Indigenous communities of North America, where *H. influenzae* serotype a (Hia) has emerged as a significant pathogen [23,101]. There has also been a slight increase in infections, including meningitis, caused by Hie and Hif in Europe [102,103]. Invasive infections caused by non-typeable strains of *H. influenzae* (NTHi) have increased significantly in many regions of the world [103,104].

#### **9. Meningitis Due to Non-b Serotypes of** *H. influenzae*

#### *9.1. Meningitis Due to Serotype a (Hia)*

Before the introduction of Hib vaccine, invasive *H. influenzae* serotype a (Hia) disease was very uncommon, although Hia was responsible for 12% of cases of bacterial meningitis in young children in Papua New Guinea before the introduction of Hib immunization [105,106]. Over the last two decades Hia has emerged as a significant pathogen, particularly in Indigenous populations in North America [23]. High incidences of Hia infection have been reported in Alaska Native, American Indian, and Canadian Inuit children [29,107–110]. In 2011, a population-based study in 12 Canadian pediatric tertiary care centers reported an Hia incidence of 418.8/100,000 in Inuit children aged < five years in the Keewatin region [111]. Hia is the second most virulent capsular serotype of *H. influenzae* [112] and can cause meningitis, pneumonia, septic arthritis, and bacteremia [23]. Most Hia infections occur in children aged six months to two years [23]. Between 1998 and 2003 38/76 (50%) of cases of Hia infection identified in Navajo and White Mountain Apache children presented with meningitis [107]. Hia meningitis was the commonest presentation in Indigenous children in the North American Arctic and Northern Canada [108,110]. Hia has also emerged as a significant pathogen in Utah and North and South Dakota [113–117]. In a study in Utah from 1998 to 2008, 28% of all invasive disease in children aged < five years was due to Hia, and 18% due to Hib. Fifty percent of the Hia cases presented as meningitis [115]. Hia infections in these states were not exclusively in American Indian children. Hia infections have also been reported from Brazil [118–120]. The case fatality rate of Hia meningitis was 14% in Brazil [120], 16% in Northern Canada [110], and 6% in the North American Arctic [112]. Hia has also been reported in Italy [121] and England [122] but there were no cases of meningitis in these reports of infections, which predominantly occurred in adults. Hia meningitis in a 10 month old infant and a 3 year old child was reported from Saudi Arabia [123]. The emergence of Hia as a significant cause of invasive infections in certain populations has prompted the development of an Hia conjugate vaccine [124].

#### *9.2. Meningitis Due to Serotypes e and f (Hie and Hif)*

There has also been increasing recognition of cases of meningitis caused by Hie and Hif [102]. Between 2001 and 2010 the year on year incidence of Hie and Hif infections in England and Wales increased by 7.4% and 11.0% respectively [100]. In 2009–2010, the incidences of Hie and Hif infections were 0.03/100,000 persons and 0.09/100,000 persons, respectively, with the highest rates being seen in infants and older adults [102]. Nine of 10 cases occurring in infants aged <one year presented with meningitis (three Hie, six Hif). All of the infants with Hif meningitis survived, but one child with Hie meningitis died, one had severe bilateral sensorineural deafness and one developed seizures [102]. Meningitis was a less common presentation in older children and adults, with three cases of Hie meningitis (one child aged one to four years, one child aged five to 14 years, one adult aged 15–64 years) and four cases of Hif meningitis (one in a child aged one to four years, three in adults aged 15–64 years). The case fatality rates of Hie and Hif meningitis were 14.3% and 0%, respectively. In this study Hie meningitis was associated with more complications and a higher case fatality rate.

Whittaker et al. [125] analyzed reports of invasive *H. influenzae* infection reported by 12 European countries to the European Centre for Disease Prevention and Control (ECDC) between 2007 and 2014. Five hundred and ninety-six cases of meningitis were reported, representing 9% of all infections. Sixty percent and 40% of infants aged <one year with Hie or Hif infection were reported to have meningitis [125]. National surveillance in Germany between 2001 and 2016 identified 351 cases of capsulated *H. influenzae* invasive infection: 241 cases of Hif, 45 cases of Hie, seven cases of Hia, and 58 cases of Hib(126). Forty cases of Hif infection were in children aged < four years with 40% of these cases presenting as meningitis. There were 185 cases of Hif infection in adults aged ≥ 40 years

with meningitis accounting for 15% [126]. Hif meningitis has also been reported in the United States [127–129] and in Sweden [130].

#### *9.3. Meningitis Due to Non-Typeable H.influenzae (NTHi)*

Since the introduction of Hib vaccine, NTHi infections have emerged as the most common cause of invasive *H. infuenzae* infection in many parts of the world, where surveillance has been undertaken [104,113,125–128,130–139]. The highest burden of NTHi infections is seen in neonates, children aged <one year, pregnant/post-partum women, and in older adults (≥65 years) [104]. The clinical presentation varies by age, with meningitis more commonly seen in older infants and children and pneumonia more common in older adults [104].

Over a five year period (2009–2013), there were 115 cases of neonatal invasive NTHi infection in England and Wales (incidence 4.1/100,000; 95% CI 3.4–5.0) [24]. The incidence was significantly higher in premature babies (28.4/100,000; 95% CI 22.8–35.0) compared to those born at term (0.9/100,000; 95% CI 0.6–1.4) and increased exponentially with increasing prematurity. For infants born at <28 weeks' gestation the incidence was 342/100,000 (95% CI, 234–483). Most cases (110/115, 96%) presented within 48 h of birth. Although most of the infants developed a bacteremia, 11 (10%) presented with meningitis. One infant with meningitis died and five (50%) developed long-term sequelae [24].

Active surveillance for invasive *H. influenzae* disease in the US ABC surveillance sites from 2009 to 2015, reported that invasive NTHi infections had the highest incidence (1.22/100,000) [113]. Among 317 cases of invasive *H. influenzae* infection in children aged <one year, 25.1% presented with meningitis. One hundred and ninety six of 294 (66.7%) invasive infections (where the serotype was known) in this age group were due to NTHi. Although the serotyping of the meningitis cases was not reported it is probable that they included cases of NTHi meningitis.

Between 2001 and 2008, there were 396 cases of invasive NTHi infection documented by the Netherlands Reference Laboratory for Bacterial Meningitis [134]. Overall, the most common presenting clinical syndrome was invasive pneumonia (190/396, 48%) followed by bacteremia (75/396, 19%). Fifty-seven (14%) of the cases presented with meningitis. Among children aged seven weeks to <five years 28/60 (47%) of cases were meningitis. Nationwide active surveillance in Germany between 1998 and 2005 identified 70 cases of invasive NTHi infection. The median age of presentation was 26 months (0–73 months) and 34% presented with meningitis [135]. Thirty eight percent of children with NTHi meningitis had predisposing conditions, including prematurity, immunodeficiency, and Down's syndrome [135]. In a study from England [131] 26% of children who survived NTHi meningitis suffered long-term sequelae, including deafness, seizures, and hydrocephalus [131]. The case fatality rate of NTHi meningitis is similar to that of Hib meningitis [131].

#### **10. Conclusions**

Hib conjugate vaccine has been a remarkable success story, reducing the incidence of Hib meningitis to a very low level in countries with a well-established Hib immunization program and sustained high vaccine coverage [140]. There has been considerable progress in achieving the elimination of *H. influenzae* meningitis, but more still needs to be done. Cases of Hib meningitis do still occur, in unimmunized or partially vaccinated children, and as rare instances of true Hib vaccine failures. In 2015, Wahl et al. [73] estimated that there were still 12,900 cases (UR 6400 to 21,500) of Hib meningitis globally. Since then, Hib vaccine has been introduced into the NIP of almost all countries, including India and Thailand, except for China and the Russian Federation (where Hib vaccine is recommended for certain risk groups). Every child in the world should be offered Hib vaccine and vaccine coverage needs to be maintained at a high level in all countries. Hia has emerged as a significant cause of meningitis in Indigenous populations of North America, potentially requiring the use of Hia conjugate vaccine in these high-risk populations. Hie, Hif, and NTHi have also been associated with cases of meningitis. The changing epidemiology of *H.*

*influenzae* meningitis emphasizes the importance of on-going surveillance. Epidemiologic and microbiologic surveillance should be comprehensive, covering all ages and all types of *H. influenzae*. Accurate typing of strains, using molecular methods combined with clinical ascertainment of clinical presentation, underlying risk factors and outcome should be undertaken to fully document these changes. Considerable progress in achieving the elimination of *H. influenzae* meningitis has been made, but more still needs to be done.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data available in a publicly available repository (PubMed).

**Conflicts of Interest:** MPES has received personal fees from GSK, Pfizer, AstraZeneca, and Sanofi Pasteur as a speaker at international meetings and as a member of advisory boards (unrelated to the submitted work).

#### **References**


### *Article*
