Antibody responses after sequential vaccination with PCV13 and PPSV23 in kidney transplant recipients

Immunocompromised patient cohorts such as solid organ transplant recipients are at major risk for infectious complications. These include lower respiratory tract infections, which can lead to severe disease with requirement of hospital treatment or even transfer to the intensive care unit [1]. Streptococcus pneumoniae (S. pneumoniae) is a capsulated gram-positive bacterium, that frequently colonizes the human nasopharynx, but can also lead to local and systemic diseases [2].

Apart from meningitis, otitis media and sinusitis, it is the most frequently identified bacterial pathogen in pneumonia [3]. Due to the administration of immunosuppressive agents, the risk of invasive pneumococcal disease (IPD) is dramatically increased in solid organ recipients. Therefore, vaccination in immunocompromised individuals is recommended to reduce the incidence of IPD [4‐6].

The capsule is the main virulence factor of S. pneumoniae and consists of different polysaccharides, which form the basis for the classification of pneumococci into over 90 serotypes. Twenty-three of these serotypes are responsible for 80–90% of infections nowadays [7].

Currently, two types of pneumococcal vaccines are licensed and used in routine clinical practice: the pneumococcal polysaccharide vaccines (PPSV) and pneumococcal conjugate vaccines (PCV). PPSVs act as T-cell independent type 2 antigens, inducing IgG responses and poor generation of memory B cells. PCV was developed to enhance immunogenicity by covalent conjugation to carrier proteins. These peptides induce a T helper cell response, which can promote B-cell differentiation into antibody producing plasma cells or memory B cells [8].

Apart from a reduced immunogenicity of vaccines in solid organ transplant recipients, another concern refers to the risk of triggering allograft rejection through stimulation of alloreactive T and B cells, which is of particular interest in PCVs as they are specifically engineered to increase immune activation [8, 9].

In Germany, a sequential administration of the 13-valent pneumococcal conjugate vaccine (PCV13) followed by the 23-valent pneumococcal polysaccharide vaccine (PPSV23) after 6–12 months is recommended for risk groups including solid organ recipients [6]. It is recommended to control humoral vaccination responses in this cohort. However, it is still unclear to what extent serological titers reflect protection against an infection with S. pneumoniae [6]. Studies in kidney transplant recipients (KTR), that examined the humoral response after administration of PCV13 revealed increased functional antibody responses after vaccination [8, 10] but could not match the responses in healthy controls [11]. The administration of PPSV23 in kidney KTR also led to a significant increase of antibodies, which was still detectable after a period of 15 months [12, 13].

However, there are currently no data on the serological response in KTR, who received a sequential vaccination with PCV13 and PPSV23. In addition to a measurement of a global pneumococcal antibody response (against 23 serotypes), we also determined specific immune responses to six pneumococcal serotypes. Therefore, this study aims to investigate the serological immunogenicity and safety of the aforementioned vaccination regiment in KTR.

The current study examined the serological immunogenicity and safety of a sequential vaccination with PCV13 and PPSV23 in kidney transplant recipients. We could demonstrate that this vaccination regiment induced a serological response indicated by a significant increase in both, global and serotype specific antibody levels. We recorded only one event each of biopsy proven rejection and de novo development of DSA in our cohort during a follow-up of 18 months after the first vaccination (with PCV13). Given the general risk of alloimmune processes in solid organ transplant recipients, we do not ascribe these events to the administered vaccines [20]. Therefore, we evaluate the sequential administration of both vaccines as immunologically safe in our cohort.

The age of patients in this study ranged from 22 to 76 years, with a median age of 57 years. For patients under and over 65 years of age, the Centers for Disease Control and Prevention (CDC) recommends different vaccination intervals. For younger patients, vaccination intervals vary depending on their specific risk factor, but should not be less than eight weeks between PCV13 and PPSV23. For healthy 65-year-old or older adults, the CDC recommends a minimum interval of one year. However, this can be shortened to a minimum interval of 8 weeks, as is the case in transplant patients with immunodeficiency presented here [21]. In our study, we have chosen a vaccination interval of 6 months, as indicated by the current vaccination recommendation against pneumococci for risk groups of the Robert Koch-Institute in Germany [22].

A common problem, that arises in studies with pneumococcal vaccines, is the lack of a robust threshold to define effectiveness in adults, especially in the case of immunocompromised individuals. The WHO ELISA has been used intensely in evaluation and licensure of pneumococcal vaccines, especially pneumococcal conjugate vaccines in children. Based on ELISA results from three clinical studies on the seven-valent pneumococcal conjugate vaccine (Prevenar), the WHO defined a correlate of protection of 0.35 μg/mL for serotype-specific antibody concentrations for the license of vaccines against invasive pneumococcal disease in children [23]. But this threshold does not aim to imply protective status in an individual, nor could it be used to assume protection against other pneumococcal infections like pneumonia or otitis media, which may require higher antibody levels [24]. Other authors suggested the aforementioned cut-off should be higher [8, 25, 26]. In our study, we observed that for some serotypes, e.g. serotype 14, the baseline antibody concentrations were already much higher than 0.35 μg/mL.

Therefore, we tend to focus more on the relative increase of antibody levels after vaccination in comparison to baseline values as it was also done in a phase III clinical trial to evaluate the safety and immunogenicity of PPSV23 [15] and in a previous study in KTR investigating the immunogenicity of a sequential vaccination with PCV7 and PPSV23 [26]. The vaccination regiment in the current study comprises two different vaccine types, one polysaccharide vaccine (PPSV23) and one conjugate vaccine (PCV13). Polysaccharide antigens are large molecules consisting of repetitive epitopes. These molecules are not processed by antigen-presenting cells and interact directly with B cells, inducing antibody responses in the absence of T cells. However, T cell-independent responses have several limitations, including poor induction of immunological memory. In contrast, antibody responses against protein antigens are T cell-dependent and result in long-lived immunity due to the generation of immunological memory. Pneumococcal polysaccharide vaccines elicit responses that mainly induce IgG2 in adults, whereas both IgG1 and IgG2 responses are induced by pneumococcal conjugate vaccines [27]. In our study, we could demonstrate that all measured serotype-specific geometric mean antibody concentrations showed more than a twofold increase at month 12 compared to baseline, indicating a sufficient immunogenicity. This includes serotypes, that are an ingredient of only PPSV13 (6A), only PPSV23 (2, 9N, 11A) or both (3, 14). The time courses of serotype 3 and serotype 14 exemplify that the vaccination with PPSV23 hardly boosts the antibody concentrations. Whereas PCV13 doubles them, PPSV23 only slightly increases them. We observed the strongest fold-increase for serotype 9N, which is in line with the mentioned phase III study for PPSV23. The highest absolute antibody concentrations were recorded for serotype 14 as also described in previous studies on different vaccination strategies [8, 26].

Our global anti-pneumococcal antibody data support the courses of serotype-specific antibody concentrations. With exception of IgM (1.7-fold increase), all other antibody subclasses showed more than a twofold increase of GMC at month 12 compared to baseline. We compared our data to the results of a previous study of our group, which comprised a comparable cohort of 47 KTR, who received a single dose of PCV13 [11]. The determination of global antibody concentrations was done with same commercial ELISA in the same lab. The median age of the current cohort was slightly higher (57 years vs. 53 years) whereas the median interval between first vaccination and (last) kidney transplantation was a bit lower (38 months vs. 49 months). Immunosuppressive medication did not deviate significantly as the vast majority in both cohorts received triple therapy with tacrolimus, MPA and corticosteroids. The comparison revealed higher relative increases and absolute values for concentrations of all antibody types at month 12 indicating an enhanced and longer lasting serological immune response after sequential vaccination compared to a single vaccination with PCV13. Our previous study showed that serotype specific IgG antibody concentrations as determined by ELISA correlated significantly with their functional activity measured by the opsonophagocytic assay, indicating that serotype specific IgG antibodies are functional in this immunocompromised cohort [11]. However, it should also be noted that functional antibody concentrations may be lower than the concentration of binding antibodies presented in the current study.

There are only limited data on global and serotype-specific anti-pneumococcal antibody concentrations in healthy adults before and after vaccination and validated protective cut-off values are lacking [28]. Therefore, we compared our data with two groups of healthy adults [18, 19, 29]. The first cohort consists of vaccine naïve healthy individuals [17, 30]. The comparison revealed reduced global antibody concentrations (all subclasses) in KTR before vaccination. One month after first vaccination, comparable antibody concentrations to healthy adults were reached except for IgM. This finding was in line with our previous study on a single administration of PCV 13 in KTR [16]. But 12 months after vaccination, the GMC of all immunoglobulin classes matched or even exceeded those of vaccine naïve healthy adults. This did not apply to the same extent to our previous study [11]. To evaluate serotype-specific antibody concentrations, we compared our patient cohort with healthy adults, that were vaccinated with a single dose of PCV13 [19]. With respect to the tested serotypes, we could only compare antibody concentrations of serotypes 3, 6A and 14. Antibody concentrations were higher for all serotypes in healthy individuals at one month after administration of PCV13. However, at 12 months after vaccination with PCV13 antibody concentrations of serotypes that are also part of PPSV23 were comparable (serotype 3) or even higher (serotype 14) in kidney transplant recipients. In contrast, antibody concentrations of serotype 6A (only part of PCV13) were still much higher in healthy individuals.

Correlation analysis for all antibody subclasses revealed that values before vaccination were predictive of antibody concentrations after vaccination with highest correlation coefficients for IgM antibodies. Overall, the strongest correlation was seen between IgG and IgG2 antibody concentrations, reaching highly significant results (p < 0.001) at each timepoint. Robbins et al. investigated the immunogenicity of PCV13 in an adult cohort with common variable immunodeficiency (CVID) or IgG subclass deficiency [31]. They observed that higher global IgG and IgG2 values at baseline were associated with protection at one year after vaccination. These findings are in line with a study on patients with systemic lupus erythematodes, for whom higher global IgG2 serum concentrations were associated with long-lasting protection three years after sequential PCV13/PPSV23 vaccination [32]. In general, IgG2 is known to play a key role in the defense against pneumococcal infections [33].

Decreased anti-pneumococcal IgA and IgM levels have been observed in healthy adult blood donors [18], but have also been associated with a pronounced rate of respiratory infections in patients with CVID [34] and primary antibody deficiency (PAD) [35]. It is also known that individuals characterized as having an intact humoral response based on measurement of serotype-specific IgG concentration can still display impaired anti-pneumococcal IgM and IgA levels [36]. Thus, an additional determination of anti-pneumococcal IgM and IgA concentrations could yield more precise information on the humoral response to pneumococcal vaccines in individuals, but the clinical benefit is questionable [36].

The global anti-pneumococcal IgG ELISA shows a good correlation to the GMC of serotype-specific antibodies, with exception of serotype 2, for most time-points in our study and was also reported previously [11]. This commercial assay therefore serves as a cost-effective and easy tool to monitor the humoral immune response to pneumococcal vaccination in clinical routine [28]. But to get more insight into serotype-specific serological responses, especially in the case of low-level global anti-pneumococcal IgG concentrations, serotype-specific WHO ELISA is required [28]. In particular, serotype 3 remains a dominant cause of invasive pneumococcal disease (IPD) in Europe [37]. In our study, only 56.5% of recipients showed a twofold increase of antibodies against serotype 3 after 12 months, compared to baseline. This is approximately 10% less than the mean value for the other serotypes. Moreover, concentrations were by far the lowest of the tested serotypes. Since serotype 3 is a main driver of IPD in Germany, these results corroborate the well-known lack of vaccine-protection [37‐39]. However, for a more accurate assessment of this problem, a longer follow-up of recipients and larger group of participants are necessary.

We compared the vaccine-induced immune responses with clinical patient data. We observed that patients with MPA treatment had significantly lower global IgG and IgG2 antibody concentrations compared to patients without MPA intake. We also tested, if we could confirm this finding for serotype-specific antibodies, but although we did see the same trend, it did not reach significance (data not shown). This may be explained by the small size of both groups (n = 33 vs. n = 13) in the sense that statistical significance was only reached by a summed effect of global antibody concentrations. Interval between (last) kidney transplantation and first vaccination were the only other variable with a significant association with global antibody GMC. In that case we recorded a positive correlation meaning that a longer interval, which normally implicates reduced immunosuppressive treatment, was associated with higher IgG and IgG2 GMCs. However, regression analysis revealed that only MPA treatment proved to be significant for IgG and IgG2 GMC at month 7 and month 12 after first vaccination, highlighting its relevance in case of long-term vaccine immunogenicity. This is in line with previous studies, which generally describe the dominant effect of MPA treatment on humoral immune responses [40] or specify its impact on serological responses after pneumococcal vaccination [11]. Mycophenolic acid inhibits the generation of guanine nucleotides. Unlike different other cell types (e.g., neurons, hepatocytes and renal cells), lymphocytes can only generate guanine nucleotides de novo. Therefore, they are a rather specific target of MPA, which leads to reversible inhibition of B and T cell proliferation. This explains the negative effect on humoral immune responses as described previously and in the current study. The calcineurin inhibitor tacrolimus primarily affects T cell functionality but can indirectly inhibit B cell functions that depend on CD4⁺ T cell interaction. Accordingly, tacrolimus is more likely to impair the efficacy of T cell-dependent conjugate vaccines like PCV13 than T cell independent polysaccharide vaccines. Glucocorticoids, which are also part of the conventional immunosuppressive triple therapy in kidney transplantation, have pro-apoptotic effects on B and plasma cells. The combination of MPA, tacrolimus and glucocorticoids therefore leads to a substantial suppression of humoral immunity [40, 41]. However, our data shows that MPA has the strongest influence on the production of anti-pneumococcal antibodies.

In this subgroup of organ transplant recipients, a third vaccination may be required to reach equivalent antibody concentrations. Pneumococcal booster vaccination has already been shown to improve overall immune protection against pneumococcal disease in immunocompromised patients such as HIV-positive adults and is highly recommended for adults with chronic obstructive pulmonary disease [42, 43]. Further studies are necessary to determine the most suitable vaccine for booster vaccination in kidney transplant recipients and to compare its impact on the incidence of pneumococcal disease.

There are several limitations to the study that should be acknowledged. First, the study lacked a control group that received the same vaccine sequence, which would have allowed a more robust comparison of the effectiveness and safety of the PCV13 and PPSV23 sequential vaccination. Direct comparison of PPSV23 vaccine efficacy between KTR and the healthy blood donor population is not ideal. Second, the follow-up period of 18 months may not be sufficient to assess the long-term effectiveness. Vaccine efficacy may decrease after 5 years, so CDC recommends booster vaccination after 5 years for adults 65 years and older, who received PCV13 at any age and PPSV23 before age 65 [21]. Third, the study was conducted in a single center in Germany, which may limit the generalizability of the findings to other regions or populations with different characteristics, prevalent serotypes or risk factors. However, we found that all measured serotype-specific geometric mean antibody concentrations showed more than a twofold increase at month 12 compared to baseline, indicating sufficient immunogenicity of sequential vaccination in KTR. In addition, diminished antibody rise was narrowed down to therapy with MPA.

To conclude, sequential vaccination with PCV13 and PPSV23 in kidney transplant recipients results in a superior antibody response compared with single vaccination. MPA treatment significantly reduced the antibody response in contrast to any other immunosuppressive therapy.