**The E**ff**ect of a 13-Valent Conjugate Pneumococcal Vaccine on Circulating Antibodies Against Oxidized LDL and Phosphorylcholine in Man, A Randomized Placebo-Controlled Clinical Trial**

#### **Hendrika W. Grievink 1,2 , Pim Gal 1,3, Maria Ozsvar Kozma <sup>4</sup> , Erica S. Klaassen <sup>1</sup> , Johan Kuiper <sup>2</sup> , Jacobus Burggraaf 1,5,6 , Christoph J. Binder <sup>4</sup> and Matthijs Moerland 1,3,\***


Received: 9 September 2020; Accepted: 17 October 2020; Published: 22 October 2020

**Simple Summary:** Atherosclerosis is the main underlying mechanism for cardiovascular disease. The main cause for atherosclerosis development is oxidized low density lipoprotein (oxLDL) accumulation in the vessel wall and a subsequent immune response. It has been established that immunoglobulin M antibodies against oxLDL help protect against atherosclerosis. It has been found in mice that vaccination with *Streptococcus pneumoniae* results in an increase of these protective antibodies and thereby decreases the development of atherosclerosis. In this study, we investigated if this increase of antibodies can be found in human as well. Twenty-four healthy male volunteers were vaccinated with Prevenar-13, a pneumococcal vaccine, using different dosing regimens. An increase in anti-Prevenar antibodies was found, showing that the vaccination worked. However, no increase in protective anti-phosphorylcholine or anti-oxLDL antibodies was observed. This work shows that vaccination against pneumococcal does not seem to be a suitable treatment option to help prevent atherosclerosis development, although further research would be required to test alternative pneumococcal-based vaccines, vaccination regimens or study populations.

**Abstract:** In mice vaccination with *Streptococcus pneumoniae* results in an increase in anti-oxLDL IgM antibodies due to mimicry of anti-phosphorylcholine (present in the cell wall of *S. pneumoniae*) and anti-oxLDL IgM. In this study we investigated the human translation of this molecular mimicry by vaccination against *S. pneumoniae* using the Prevenar-13 vaccine. Twenty-four healthy male volunteers were vaccinated with Prevenar-13, either three times, twice or once in a double-blind, placebo-controlled, randomized single center clinical study. Anti-pneumococcal wall, oxLDL and phosphorycholine antibody levels were measured at a fixed serum dilution, as well as circulating lipid levels over the course of 68 weeks. A significant increase in anti-oxLDL IgG and IgM was seen in the group receiving two doses six months apart compared to the placebo. However, these differences were not observed in the groups receiving a single dose, two doses one month apart, or three doses. This study shows that vaccination with Prevenar-13 does not result in robust anti-oxLDL IgM levels

in humans. Further research would be required to test alternative pneumococcal-based vaccines, vaccination regimens or study populations, such as cardiovascular disease patients.

**Keywords:** cardiovascular disease; vaccine; clinical trials; translational medicine; atherosclerosis; oxLDL; phosphorylcholine

#### **1. Introduction**

Oxidized low density lipoprotein (oxLDL) particles play a key role in the etiology of atherosclerosis [1]. In the vessel wall, oxLDL is recognized and phagocytosed by macrophages primarily via scavenger receptors leading to foam cell formation [2]. Macrophage foam cells are hallmark cells of atherosclerotic lesions and participate in the inflammatory responses that mediate smooth muscle cell migration and proliferation, and extracellular matrix production, and thereby stimulate atherosclerotic plaque progression.

Several mouse studies showed that IgM antibodies against oxLDL are atheroprotective [3,4]. IgM antibodies against oxidized particles facilitate the clearance of apoptotic cells, thereby promoting the resolution of inflammation [5,6]. Additionally, these antibodies neutralize the proinflammatory effects of oxidized phospholipids [7,8]. Inhibition of scavenger receptor-mediated oxLDL uptake by macrophages prevents the formation of foam cells and subsequent progression of atherosclerotic plaque formation [3,9]. In clinical studies, oxLDL-specific IgM has been reported to be a protective factor for atherosclerosis development, correlating with cardiovascular disease incidence and clinical outcome [10–13].

In contrast to IgM, the role of oxLDL–specific IgG in atherosclerosis is thought to be atherogenic. OxLDL–IgG complexes have been shown to induce survival of plaque-resident monocytes [14] and secretion of proatherogenic cytokines by mast cells [15]. Clinical studies showed a correlation between oxLDL–IgG antibodies and acute coronary syndrome, suggesting an untoward role of this antibody in plaque destabilization [16]. In human, oxLDL–specific IgG antibody titers correlated inversely to the oxLDL serum concentration [17] and, in mouse, serum cholesterol levels [18], which suggests that oxLDL–specific IgG facilitates phagocytosis of oxLDL by macrophages.

Mouse experiments showed that certain IgM clones binding oxLDL bind phosphorylcholine (PC) of oxidized—but not unoxidized—phospholipids [3,19]. Importantly, Binder et al. showed in Ldlr knockout mice that vaccination against *S. pneumoniae* using pneumoccocal extracts induced high titers of oxLDL–specific IgM, subsequently leading to a decrease in atherosclerotic lesions [19]. This effect was explained by the fact that PC is present as part of the capsular polysaccharide of *S. pneumoniae*. Moreover, immunization with PC conjugated to carrier proteins also induced oxLDL–IgM and decreased the extent of atherosclerosis in ApoE knockout mice [20,21].

Autoantibodies against PC are also found in humans, where low levels of PC–IgM autoantibodies correlate with a higher incidence of cardiovascular disease [22–25]. Moreover, pneumococcal-specific IgG and oxLDL-specific antibody titers correlated significantly in subjects who had received pneumococcal vaccination [26], although there are also reports of an absent effect of pneumococcal vaccination on oxLDL–specific IgM levels [27].

The present proof-of-concept study investigated the human translation of the observed effects of pneumococcal immunization in mice. Healthy human volunteers were vaccinated with a 13-valent conjugated pneumococcal vaccine (Prevenar-13®), and the induction of PC- and oxLDL-specific antibodies was measured.

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

This investigation was a double-blind, randomized, placebo-controlled, parallel, single-center study with twenty-four healthy males between 18 and 45 years of age. The study was performed at the Centre for Human Drug Research in Leiden, The Netherlands. Participants were recruited via advertisements and social media. Participants were assessed to be generally healthy based on a complete medical screening and had no previous exposure to the 13-valent pneumococcal vaccine. All participants gave written informed consent prior to any study-related activity. The study was approved by the Ethics Committee of the Leiden University Medical Centre (LUMC) and Declaration of Helsinki principles were followed. The study is registered in the Dutch Trial Registry (Nederlands Trial Register, NTR) under study number NTR5643 and took place for all participants simultaneously between March 2016 and October 2017. This study was funded by the European Union, call FP7-HEALTH-2013-INNOVATION, project ID 603131.

#### *2.1. Vaccination Schedule* endpoint, lipid levels in circulation were measured. Estimated differences were calculated between

The 13-valent conjugated pneumococcal vaccine (Prevenar-13®) used in this study was from a single batch (batch no. MU7958). The presence of residual PC in the vaccine preparation was confirmed by ELISA using the PC-specific mAb IgM E06. Placebo consisted of 0.9% NaCl solution. Since there are clear visual differences between these vaccinations, three physicians were unblinded for administration of the vaccine. These physicians were not otherwise involved in the study. the groups. A positive value indicates a higher estimated value for the active group, a negative value indicates a lower value for the active group. The analysis was performed in SAS v9.4 (SAS Institute, Cary, NC, USA). **3. Results** Twenty-four healthy volunteers were included in the study; their baseline characteristics can be found in Table 1. One subject withdrew consent after two weeks for non-study related reasons (Figure

*Biology* **2020**, *9*, x 4 of 10

Vaccinations took place at three time points: at baseline, at four weeks and at 28 weeks. Subjects were randomized in a consecutive order based on eligibility. The randomization code was generated using SAS v9.4 for Windows (SAS Institute Inc., Cary, NC, USA) by an independent statistician. The randomization code was only made available for data analysis after study completion. There were five different treatment arms, as displayed in Figure 1. In the mouse study three immunizations were enough for oxLDL–specific IgM induction [19]. In the mouse study proteinase-treated *S. pneumoniae* extracts were used. In this design, the power to detect differences between placebo and active treatment arms was optimized between baseline and at the 28 week time point (PP vs. AA vs. AP, *n* = 8 per group). 1). This subject was randomized to the active-active-placebo treatment arm and was not replaced. **Table 1.** Baseline characteristics. **Parameter** *n* **= 24** Age (years) 28.5 ± 8.5 Gender male (%) 100 Ethnicity Caucasian (%) 100 Height (cm) 180.5 ± 5.3 Weight (kg) 75.0 ± 11.0 BMI (kg/m<sup>2</sup>) 23.0 ± 3.2 Heart rate (min <sup>−</sup><sup>1</sup>) 58.5 ± 9.0 Systolic blood pressure (mmHg) 123 ± 9.3

Diastolic blood pressure (mmHg) 75.1 ± 6.7

**Figure 1.** Study flowchart. **Figure 1.** Study flowchart.

#### *3.1. Anti-Prevenar Antibodies 2.2. Antibody Measurements*

K2EDTA plasma antibody levels to Prevenar, PC–BSA, and CuSO4-oxidized LDL (oxLDL) were measured by chemiluminescent ELISA as reported previously [28]. In brief, Prevenar (Pfizer) was coated at 1:5000, PC–BSA (Biosearch Technologies, Novato, CA, USA) and oxLDL at 5 ug/ml in PBS/EDTA. IgM antibodies were measured at a dilution of 1:500 and IgG antbodies at 1:1000. Binding of IgG subclasses to Prevenar was measured at a dilution of 1:100 and to PC–BSA at 1:100 for IgG3 and IgG4, and 1:500 for IgG2.

Serum levels of total cholesterol, low density lipoprotein (LDL), high density lipoprotein (HDL) and triglycerides were measured by the chemistry lab of the Leiden University Medical Center on the Cobas P800 analyzer (Hoffmann–La Roche, Basel, Switzerland).

### *2.3. Power Calculation*

In humans, the median anti-oxLDL IgG levels in the healthy, unvaccinated population is around 50 U/l, with an interquartile range of around 25–75 U/l [26]. Anticipating an immune response minimally resulting in a five-fold rise in IgG and IgM antibody levels, and based on an inter-subject variability of 50% in basal IgG and IgM levels [19,24,26], a sample size of 4–8 subjects per group (dependent on the contrast) will be sufficient to meet the study objectives. This was a conservative approach considering in the magnitude of the oxLDL-specific IgM response observed in the murine model [19].

### *2.4. Statistical Analysis*

Data are presented as mean ± standard deviation (SD). In case of non-normal distribution, parameters were log-transformed. Repeatedly measured variables were analyzed with a mixed model analysis of variance with fixed factors treatment group, time and the interaction of treatment group and time as fixed factor and subject as random factor. Primary endpoints (Prevenar-specific Ig levels, oxLDL-specific Ig levels and PC-specific Ig levels) were compared between treatment groups for the 0–4 week window, the 4–28 week window, and the 28–68 week window. As a secondary endpoint, lipid levels in circulation were measured. Estimated differences were calculated between the groups. A positive value indicates a higher estimated value for the active group, a negative value indicates a lower value for the active group. The analysis was performed in SAS v9.4 (SAS Institute, Cary, NC, USA).

### **3. Results**

Twenty-four healthy volunteers were included in the study; their baseline characteristics can be found in Table 1. One subject withdrew consent after two weeks for non-study related reasons (Figure 1). This subject was randomized to the active-active-placebo treatment arm and was not replaced.


**Table 1.** Baseline characteristics.
