Analysis of Beyfortus^® (Nirsevimab) Immunization Campaign: Effectiveness, Biases, and ADE Risks in RSV Prevention

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review the revised manuscript. I appreciate the author's efforts to improve the writing. Nonetheless, the temporal relationship between nirsevimab use and neonatal deaths is not convincing based on a mere visual inspection of Figure 1. If nirsevimab were indeed causing excess deaths, there should be a sustained spike rather than a fluctuating pattern. Additionally, statistical evidence, such as a time series analysis, should be provided to demonstrate a significant signal. This is particularly important since all quoted and referenced evidence overwhelmingly does not show increased neonatal deaths. Furthermore, the author should explain why the methodological limitations of existing evidence would suggest that actual harm was missed, given that actual harm has not been established as fact.

Other comments: Abbreviations such as "hRSV" should be expanded at their first mention.

Overall, despite the manuscript improvements, I do not feel that sufficient steps have been taken to address the major validity concern regarding establishing an association (let alone causation) between nirsevimab and neonatal deaths.

Author Response

Dear reviewer1

 

Thank you for agreeing to review my work once again.

I have tried to respond as best I can to your requests (see below).

In particular:

The mention of neonatal deaths has been removed from the abstract and the concluding sentence has been modified

“Similarly, the small but significant increase in the number of newborn deaths between 2 and 6 days of age during the campaign in France is a warning signal of a possible ADE, even if no causal link can be established.”

 

 

Comments and Suggestions for Authors

Thank you for the opportunity to review the revised manuscript. I appreciate the author's efforts to improve the writing. Nonetheless, the temporal relationship between nirsevimab use and neonatal deaths is not convincing based on a mere visual inspection of Figure 1. If nirsevimab were indeed causing excess deaths, there should be a sustained spike rather than a fluctuating pattern.

In rare individuals, the effective concentration of nirsevimab may only be reached in a few days: during the period of viral circulation, early ADE may result from sub-neutralizing concentrations of nirsevimab. This only occurs for a few days after injections, and so will not result in a sustained increase beyond a few days.

 

Additionally, statistical evidence, such as a time series analysis, should be provided to demonstrate a significant signal. This is particularly important since all quoted and referenced evidence overwhelmingly does not show increased neonatal deaths.

A figure has been added

Legend

Figure 2: Overlay of neonatal death rates between 2 and 6 days of life for each month between March 2018 and February 2024. Between March 2018 and February 2024 the monthly rate of deaths between 2 and 6 days of life per 1,000 births was calculated according to INSEE data [ , ]. The data have only been collected since 2018, the year for which 1 early neonatal mortality stabilized (after a drop between 2001 and 2011 and an increase between 2011 and 208) [COUR DES COMPTES 06.05.2024 La politique de périnatalité, p.37, Available on line: https://www.ccomptes.fr/fr/publications/la-politique-de-perinatalite, accessed August 6, 2024]

 

And the corresponding text in the main text

Superimposing the neonatal death rates for each month between 2018 and 2024 highlights the peaks in deaths that fall outside the average ranges and recur 2 months in a row (September/October and December/January) for the 2023-2024 season; the sharp drop in November is also found in 2019 (excluding the Covid pandemic and the nirsevimab immunization season).

It is impossible to perform a time series analysis, as monthly data on the number of doses of nirsevimab administered are not available: we only know the start and end of the immunization campaign, and can only assume that the number of doses administered in hospitals in November 2023 fell drastically. It is therefore impossible to establish a causal link between the administration of nirsevimab and the observed spike in neonatal deaths; however, the biological plausibility of this link is demonstrated by the review of the biological mechanisms of ADE and is suggested by the results of clinical and post-marketing studies.

 

The mention of neonatal deaths has been removed from the abstract and the concluding sentence has been modified

“Similarly, the small but significant increase in the number of newborn deaths between 2 and 6 days of age during the campaign in France is a warning signal of a possible ADE, even if no causal link can be established.”

 

Furthermore, the author should explain why the methodological limitations of existing evidence would suggest that actual harm was missed, given that actual harm has not been established as fact.

The main bias of these studies lies in the exclusion of many subjects, particularly those who developed an RSV infection in the days following nirsevimab injection. Careful examination of the few data available, however, suggests that RSV infections are facilitated during this time interval (sub-neutralizing antibody concentrations may explain this phenomenon).

I have highlighted another indication of early ADE in the Assad study, 2024: In the intention-to-treat analysis (including all patients regardless of the date of nirsevimab injection), there was no effectiveness during the first 7 days (19% of cases - 15/75- and 20% of controls - 23/120- received nirsevimab), this lack of efficiency compares with the advertised 80%. This does not rule out early ADE.

 

Other comments: Abbreviations such as "hRSV" should be expanded at their first mention.

This omission has been corrected

 

Overall, despite the manuscript improvements, I do not feel that sufficient steps have been taken to address the major validity concern regarding establishing an association (let alone causation) between nirsevimab and neonatal deaths

I have deleted any reference to a causal link between the excess neonatal mortality and the immunization campaign, but I feel it is important to publish this anomaly, which could be a signal.

 

 

At the request of the reviewer 2, the text has been reduced from 23 to 18 pages compared with version 1:

The discussion has been removed and the various paragraphs have been included either in the introduction, the main text or the conclusion, thus avoiding unnecessary repetition. Summaries in italics have been included at the end of long paragraphs to facilitate reading. Details have been corrected and a bullet summarizing the function of nirsevimab and its relationship with ADE has been added at the request of the reviewer2.

Reviewer 2 Report

Comments and Suggestions for Authors

The article by Hélène Banoun has been improved. However, some points still need to be addressed. Your work is highly valued and appreciated.

 

-              The author has to explain this statement: “Vaccination of infants does not appear to be a feasible solution in the current state of knowledge”. There are promising clinical trials for infants using the RSV vaccine.

-              It is crucial that after a relevant statement, the respective citation is provided to ensure accuracy and credibility.

-              The grammar needs to be deeply reviewed and corrected. There exists an incorrect use of :

-              This statement is untrue: “To date, no research favoring a vaccine with a Th1 response has succeeded”. The author must look at the literature and clinical trials.

-              Table 1: The missing data must be completed, indicating that it is not evaluated (N/E).

-              Add a legend in Figure 1.

-              Add another Figure that explains the function of niversimab and its relationship with ADE.

-              Sections 5.3 and 5.4 are too short to be a section. This information must be included in other ones. This is for all short sections.

-              A discussion and conclusion in a review is a lot for readers. Most of the discussion must be included in the section respective to the theme and only leave only the conclusion.

Comments on the Quality of English Language

Quality is fine, but the grammar is not.

Author Response

Dear reviewer2

Thank you for agreeing to review my work once again.

I have tried to respond as best I can to your requests (see below).

The author has to explain this statement: “Vaccination of infants does not appear to be a feasible solution in the current state of knowledge”. There are promising clinical trials for infants using the RSV vaccine.

This has been corrected

However, promising trials of RSV vaccines for children are underway (vector-based or live attenuated viruses)

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10847284/

Gong X, Luo E, Fan L, Zhang W, Yang Y, Du Y, Yang X, Xing S. Clinical research on RSV prevention in children and pregnant women: progress and perspectives. Front Immunol. 2024 Jan 24;14:1329426. doi: 10.3389/fimmu.2023.1329426. PMID: 38327765; PMCID: PMC10847284.

 

The grammar needs to be deeply reviewed and corrected. There exists an incorrect use of :

The English language has been reviewed by a bilingual person

The manuscript will be submitted for editing by the MDPI department if the publisher deems it necessary

This statement is untrue: “To date, no research favoring a vaccine with a Th1 response has succeeded”. The author must look at the literature and clinical trials.

This statement has been deleted and not discussed to avoid lengthening the manuscript.

Table 1: The missing data must be completed, indicating that it is not evaluated (N/E).

This has been corrected

Add a legend in Figure 1.

This has been done

 

Add another Figure that explains the function of niversimab and its relationship with ADE.

As a volunteer and independent researcher, I don't own any software that allows me to create diagrams, so I've added a bullet point.

 

Function of nirsevimab and its relationship with ADE

Function of nirsevimab

* Binds to RSV F (membrane fusion) protein by its variable part and binding to FcRn through its Fc fragment (to extend its lifespan)

* These two bonds are strongly enhanced compared to other mAbs directed against F

Relationship of this function with ADE

• Macrophages and lung epithelial cells strongly express FcRn : these cells are involved in ADE during RSV infections

• FcRn is able to facilitate entry of certain virus-antibody complexes into target cells at sub-neutralizing concentrations

• Strong binding to FcRn is able to modify other effector functions of nirsevimab that may be involved in its therapeutic effect (binding to FcγR and complement components), particularly at low concentrations

• Effective concentrations are reached within a few days in some subjects

• Non-specifically, injection of large quantities of mAbs capable of saturating FcRn may result in an immunosuppressive and pro-inflammatory effect

 

Sections 5.3 and 5.4 are too short to be a section. This information must be included in other ones. This is for all short sections.

This has been corrected

A discussion and conclusion in a review is a lot for readers. Most of the discussion must be included in the section respective to the theme and only leave only the conclusion.

The discussion has been removed and the individual paragraphs have been included either in the introduction, main text or conclusion. Summaries in italics have been included at the end of long paragraphs to facilitate reading.

Comments on the Quality of English Language

Quality is fine, but the grammar is not.

MDPI's editing service will be used if necessary after your proofreading to take account of any latest changes.

 

At the reviewer1's request, the mention of neonatal deaths has been removed from the abstract and the concluding sentence has been modified

“Similarly, the small but significant increase in the number of newborn deaths between 2 and 6 days of age during the campaign in France is a warning signal of a possible ADE, even if no causal link can be established.”

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thanks for the opportunity to review the revised manuscript, and I appreciate the author's efforts to address my prior concerns. Nonetheless, it is improbable that nirsevimab was given at only one point in time at the start of the vaccination campaign, which then does not explain any possible dip in the number of deaths. As pointed out by the author, bias affected data from other studies. Similarly, bias could also affect data used to infer harm from nirsevimab. A more robust approach would then be to use randomized trial data. Overall, while I appreciate the work done for this review, the available data do not suggest harm from nirsevimab.

Author Response

Dear reviewer 1

Thank you for your patience. I have tried once again to answer your major objection.

Comment :

Nonetheless, it is improbable that nirsevimab was given at only one point in time at the start of the vaccination campaign, which then does not explain any possible dip in the number of deaths. As pointed out by the author, bias affected data from other studies. Similarly, bias could also affect data used to infer harm from nirsevimab. A more robust approach would then be to use randomized trial data. Overall, while I appreciate the work done for this review, the available data do not suggest harm from nirsevimab.

Reply

Indeed, the small increase in neonatal deaths in September-October 2023 and December 2023-January 2024 cannot be attributed to nirsevimab with the data we have. However, the repetition of 2 peaks of excess mortality (each lasting 2 months) could be a signal to watch out for and it seems important to me to publish it.

This was made clear in paragraph 2.3 (The dip in the November mortality rate is observed not only in 2023, but also in 2019, and cannot therefore be explained only by a possible shortage of nirsevimab doses. This confirms that a causal link between the immunization campaign and the variation in death rates cannot be established with the data currently available. However, the repetition of 2 peaks of excess mortality (each lasting 2 months) could be a signal to watch out for.) and in the conclusion (Similarly, the small but significant increase in the number of newborn deaths between 2 and 6 days of age during the campaign in France is a warning signal of a possible ADE, even if no causal link can be established.).

And in the conclusion:

Similarly, the small but significant increase in the number of newborn deaths between 2 and 6 days of age during the campaign in France is a warning signal of a possible ADE, even if no causal link can be established.

Concerning the non-durability of the peak number of deaths observed, I have tried to clarify my answer (“6.1 Pharmacokinetics”).

With regard to the data on neonatal mortality in France during the immunization campaign presented above, it may be objected that mortality peaks are not sustained: if nirsevimab were causing excess deaths, there should be a sustained spike rather than a fluctuating pattern. This can be explained by the hypothesis of early ADE occurring only at sub-neutralizing concentrations in rare individuals, due to inefficient biodistribution in the lungs in the first few days post-injection. Newborns affected by this ADE but who do not die immediately are less severely affected and are saved (to prove this, we would need access to data on ICU stays and the immunization status of patients).

I hope you will accept the publication with these modifications.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear author,

In the statement, "However, promising trials of RSV vaccines for children are underway (vector-based or live attenuated viruses),"  it's crucial to bolster your argument with more references, particularly original articles. This will enhance the credibility of your research. 

The response to this: Add another Figure that explains the function of niversimab and its relationship with ADE is unacceptable. You do not need to have software to make a figure. You have free tools online and PowerPoint.

 "Infants have a reduced capacity to produce neutralizing antibodies against hRSV(human respiratory syncytial virus)." Unificate the nomenclatur in the text.

 This is not an appropriate figure legend: "Rates of death between 2 and 6 days of life per 1,000 births(France, INSEE data)." Please improve it. 

Also, this: "Overlay of neonatal death rates between 2 and 6 days of life for each month between March 2018 and February 2024 ".

Comments on the Quality of English Language

The quality is fine, but the grammar could be better.

Author Response

Dear reviewer 2

Thank you for your patience and the pertinence and precision of your criticisms to which I have endeavored to respond below.

Comment 1

In the statement, "However, promising trials of RSV vaccines for children are underway (vector-based or live attenuated viruses),"  it's crucial to bolster your argument with more references, particularly original articles. This will enhance the credibility of your research. 

Response 1


References for 2 clinical trials of vector-based vaccines and 2 trials of LAV vaccines have been added.

However, trials of RSV vaccines for children are underway but they do not concern newborn babies : vector-based in children over 12 months of age, with no proven efficacy against RSV infections [8] and with URTI more frequent in vaccinated children [9] or live attenuated viruses [in children over 12 months of age, with no proven efficacy against RSV infections [10] in children over 6 months of age, with an unexpected frequency of respiratory events in the treated groups [11]

Comment 2

The response to this: Add another Figure that explains the function of niversimab and its relationship with ADE is unacceptable. You do not need to have software to make a figure. You have free tools online and PowerPoint.

Response2

Figures 3 and 4 have been added

Comment 3

 "Infants have a reduced capacity to produce neutralizing antibodies against hRSV(human respiratory syncytial virus)." Unificate the nomenclatur in the text.

Response 3

The nomenclature has been unified: only the term RSV is used.

Comment 4

 This is not an appropriate figure legend: "Rates of death between 2 and 6 days of life per 1,000 births (France, INSEE data)." Please improve it. 

Response 4

The title of the caption has been corrected

Rates of death per month between 2 and 6 days of life per 1,000 births in France (from April 2018 to April 2024) displaying the time window of the nirsevimab immunization campaign (Beyfortus©)

The legend has been completed

Monthly numbers of births and deaths are taken from INSEE data (France- Institut National de la Statistique et des Études Économiques-France)[58,59].

Comment 5

Also, this: "Overlay of neonatal death rates between 2 and 6 days of life for each month between March 2018 and February 2024 ".

Response 5

The title of the legend has been corrected

Rate of deaths per month between 2 and 6 days of life per 1,000 births in France between March 2018 and February 2024: superposition of the curves obtained for each year.

Comments on the Quality of English Language

The quality is fine, but the grammar could be better.

The language has been revised once again, and MDPI's English grammar correction service will be used if the editor deems it necessary.

 

 

 

 

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

The author has addressed all my suggestions.

Comments on the Quality of English Language

MDPI has to review the grammar.