Routine pneumococcal vaccination is recommended at regular intervals posttransplant. However, there is limited data on durability of vaccine response and the impact of vaccine type on antibody persistence. We determined the durability of response for patients enrolled in a randomized trial of conjugate (PCV7) versus polysaccharide (PPV23) pneumococcal vaccination. Response was defined as a twofold increase from baseline and a titer ≥0.35 μg/mL using a pneumococcal ELISA for seven serotypes (measured at 8 weeks and 3 years). Forty-seven patients were evaluated and had received either PPV23 (n = 24) or PCV7 (n = 23). Response rates and geometric mean titers varied by serotype but declined significantly at 3-years for 6 of 7 serotypes (p < 0.001). No significant difference in durability was found in patients that had received PPV23 versus PCV7. Compared to the 8-week response, 20.6% fewer patients had a response to at least one serotype by 3 years. The largest relative declines were seen for serotype 4 (response dropped from 40.4% at 8 weeks to 17.0% at 3 years) and serotype 9V (44.7% dropping to 21.3%). The only factor predictive of response durability was a strong multiserotype initial response (p < 0.001). In conclusion, vaccine responses decline significantly by 3 years and conjugate vaccine does not improve the durability of response.
Renal transplant recipients are at greater risk of invasive pneumococcal disease than the general population (1). Pneumococcal polysaccharide vaccine (PPV23) is recommended in these patients posttransplant and includes the majority of serotypes circulating in the population (2–4). However, the immunogenicity of pneumococcal vaccination in transplant recipients has been shown to be lower than that commonly reported for healthy individuals. Alternative strategies have been proposed to improve the immunogenicity of pneumococcal vaccine in transplant recipients. For example, pneumococcal conjugate vaccine (PCV7), in which the antigens are conjugated to a protein carrier, has shown to be immunogenic in infants, persons infected with human immunodeficiency virus, and healthy adults (5–7). In addition, PCV7 appears to enhance the immunogenicity of the polysaccharide vaccine when given first as a priming vaccine to patients with Hodgkin's disease and sickle cell disease (8,9). In a randomized trial comparing PPV23 to PCV7 in adult kidney transplant recipients, we demonstrated that the conjugate vaccine is immunogenic in this patient population (10). However, there was no significant difference in immunogenicity demonstrated between the two vaccines at 8-weeks postvaccination.
Currently the American Society of Transplantation (AST) guidelines for vaccination of transplant recipients suggests pneumococcal vaccination be repeated at 3–5 year intervals (3). This is based on the assumption that vaccine titers may decline quickly in transplant recipients. Studies done several years ago have suggested that antibody titers in transplant recipients may decrease more rapidly than in patients receiving hemodialysis and that rapidly waning immunity may be seen in renal transplant patients given pneumococcal polysaccharide vaccine (11,12). However, there is a paucity of data on the durability of vaccine response in the current era of immunosuppression, especially in organ transplant recipients receiving a combination of calcineurin inhibitor plus mycophenolate mofetil (MMF) as part of their immunosuppressive regimen. In addition, the pneumococcal conjugate vaccine elicits a T-cell dependent response, results in memory B-cell formation and produces antibodies of higher avidity (13). It is therefore possible that conjugate vaccine may lead to a more durable antibody response compared to polysaccharide vaccine. There were two main objectives to the current study: (1) to assess the durability of response to pneumococcal vaccine in kidney transplant recipients and (2) to determine if differences exist in immunogenicity over time between conjugate versus polysaccharide vaccine. An assessment of response at the 3-year time point postvaccination was chosen since this reflects the current AST guidelines for revaccination at 3–5 year intervals.
In the original trial, kidney transplant patients were enrolled in a randomized double-blinded study comparing the immunogenicity of a single dose of 23-valent pneumococcal polysaccharide vaccine (PPV23) versus the 7-valent pneumococcal conjugate vaccine (PCV7). The primary outcome of the original trial was immunogenicity of the vaccines at 8-weeks postvaccination, and these results have been previously published (10). A total of 60 adult kidney transplant recipients that were between 3 months and 3 years posttransplant with stable renal function and immunosuppression were included. Patients who had received pneumococcal vaccine in the 5 years prior to enrolment were excluded. Patients were randomized to receive a single 0.5 mL dose of either pneumococcal conjugate or polysaccharide vaccine. At 3 years post initial vaccination, all available patients were approached for consent for the present study to evaluate durability of vaccine response. Of the 60 patients, 47 patients were available at 3 years postinitial vaccination. Thirteen patients were not available for follow-up for the following reasons: refused consent (n = 7), death (n = 1) and could not be reached (n = 5).
Serum was collected prevaccination at 8 weeks postvaccination and at 3 years. Clinical information such as immunosuppression and pneumococcal vaccination was recorded.
Pneumococcal antibody titers were determined for each of the seven serotypes contained in the conjugate vaccine (4, 6B, 9V, 14, 18C, 19F, 23F) for all patients at baseline, 8 weeks, and 3 years postvaccination. Titers were determined by a serotype-specific enzyme-linked immunosorbent assay (ELISA) performed at the pneumococcal reference laboratory at University of Alabama, Birmingham, Alabama (14). Briefly, microtiter plates were coated with serotype-specific pneumococcal polysaccharide antigen and incubated in a humidified environment at 37°C. Patient serum was preadsorbed with cell wall polysaccharide and pneumococcal type 22F capsular polysaccharide. Serum specimens were added to the microtiter plate and incubated at room temperature. Goat antihuman IgG-alkaline phosphatase conjugate was then added followed by phosphatase substrate. A reference plasma pool (89-SF) with preassigned values was used as control. Optical density was then measured and converted to antibody concentrations. The lower limit of detection of this assay is approximately 0.01 μg/mL. The detailed protocol can be found at http://www.vaccine.uab.edu.
A ‘serotype response’ was defined as a twofold increase in titer from baseline and an absolute titer ≥0.35 μg/mL. This value is used at the World Health Organization (WHO) pneumococcal reference laboratory as a protective titer, and is based on a pooled analysis of published clinical studies (15).
Titers at 3 years were compared to those obtained 8 weeks after vaccination for each serotype by Wilcoxon-signed rank sum test to investigate the influence of time on the durability of vaccine. Titers between vaccines were compared using a Wilcoxon-Mann-Whitney test. Serotype responses between the PPV23 and PCV7 groups were compared by chi-squared test. Logistic regression was conducted to explore available factors that might be associated with serotype response at year 3. All statistical analysis was done using SAS version 9.0 (SAS, Cary, NC) and the graphs were produced by R software.
The original double-blind randomized controlled trial enrolled 60 kidney transplant recipients who received either PPV23 (n = 30) or PCV7 (n = 30). Of these, 47 out of 60 (78.3%) of patients were available for follow-up at 3 years postinitial vaccination including 24 patients in the PPV23 arm and 23 patients in the PCV7 arm of the study. Demographic characteristics and immunsuppression regimens of the 47 patients are shown in Table 1. At 3 years, the majority of patients were receiving immunosuppression with calcineurin-inhibitors (95.7%) and mycophenolate mofetil (78.7%). There was no significant difference with regards to demographics (age, gender, time from transplant, immunosuppression) between the subjects that had received PPV23 or PCV7 (p = NS for all comparisons). No patient underwent a splenectomy or received immunoglobulin during the study period and no differences were seen in graft loss or serum creatinine at 3 years (Table 1). No episodes of biopsy prove acute rejected occurred between 8 weeks and 3 years postenrollment.
Table 1. Baseline characteristics of cohort
PPV23 (n = 24)
PCV7 (n = 23)
*p-value = nonsignificant for all comparisons.
Mean age (years ± S.D.)
48.3 ± 12.9
49.4 ± 13.1
Gender (M / F)
15 / 9
14 / 9
Time from Transplant (days ± S.D.)
534 ± 300
435 ± 303
Immunosuppression at baseline
Immunosuppression at 3 years
Ever received PPV23 prior to enrollment
Acute rejection between 8 weeks and 3 years postvaccine
Graft loss at 3 years
Serum creatinine at 3 years (μmol/L)
143 ± 44
129 ± 36
Serotype and vaccine responses
The number of subjects with a serotype specific response (absolute value ≥0.35 μg/mL and a twofold increase from baseline) to each of the seven serotypes measured is shown in Table 2. Overall response rates at 8-weeks postvaccination ranged from 12.8% (serotype 6B) to 44.7% (serotype 9V). Of subjects that had responded at 8 weeks, maintenance of that response at 3 years postvaccination was variable depending on the serotype (Table 2). For example, of the 19 patients with an initial response to serotype 4, 42.1% (8 of 19) maintained a response at 3 years postvaccination. The most durable response was observed for serotype 14. Of the 20 patients with an initial response to this serotype, 85% (17 of 20) maintained a response at 3 years. Comparing the durability of vaccine response in the PPV23 group versus the PCV7 group, there were no significant differences in the percentage of patients maintaining a serotype-specific response between these two arms (p = NS for all comparisons) (Table 2).
Table 2. Responses to each serotype at 8 weeks and continued response at 3 years (of those that responded at 8 weeks)
Continued response at 3 years
Geometric mean antibody titers for the cohort are shown in Table 3 and illustrated in Figure 1. Mean antibody titers significantly increased at 8-weeks postvaccination for all seven serotypes (p < 0.001; 8-week titer vs. baseline titer). Titers at 3 years declined significantly as compared to the eight-week titers (p < 0.001) for all serotypes except serotype 19F. However, at 3 years, titers were still significantly greater compared to prevaccination levels (p < 0.05) for all except serotype 14. The relative changes in titer at 3 years from a peak at 8 weeks are shown in Table 3 and vary by serotype. For example, for serotype 4, the geometric mean titer declined from 0.68 μg/mL at 8 weeks to 0.38 μg/mL at 3 years. There was no significant difference in decline of titer between subjects originally vaccinated with PCV7 or PPV23.
Table 3. Geometric mean titers at 8 weeks and 3 years postvaccination (μg/mL)
All (n = 47)
PPV23 (n = 24)
PCV7 (n = 23)
*p < 0.05 compared to titers at 8 weeks.
†No significant difference between vaccines for 3-year titers for all serotypes.
Predictors of durability of response
Thirteen of 47 patients (27.6%) had no vaccine response to any of the 7-serotypes at 8 weeks. The remaining 34 of 47 (72.3%) patients responded to one or more of the serotypes in the vaccine as shown in Figure 2. Although no difference was observed between the vaccine groups, response to 6 of 7 serotypes or 7 of 7 serotypes at 8 weeks and 3 years was seen only in those vaccinated with PCV7. Response to at least one serotype decreased to 27 of 47 (57.4%) at 3 years representing a loss-rate of 7 of 34 (20.6%). One patient who had no response at 8 weeks was found to have a 1-serotype response at 3 years. Factors were analyzed for their association with durability of response. The only significant predictor was the number of serotypes that patients responded to at 8 weeks. Logistic regression showed an odds ratio 2.63 (95% CI 1.49–4.76), p = 0.0007, which implies that the more serotypes the patient responds to at 8 weeks, the more likely he/she is to maintain a response at 3 years. Patient age, time from transplant, a history of prior pneumococcal vaccination (before enrolment), the use of MMF (based on dose, or analyzed as yes/no) or calcineurin inhibitors (based on cyclosporine or tacrolimus levels or analyzed as yes/no) did not have an impact on antibody titers, serotype response or durability of response.
Pneumococcal vaccine is widely recommended and used in the solid organ transplant population to prevent invasive disease (bacteremia, pneumonia, meningitis) due to Streptococcus pneumoniae. The current AST guidelines recommend revaccination at 3–5-year intervals. However, there is only limited data to support this time period as to the durability of immunologic response by vaccination. In transplant patients, it is hypothesized that vaccine-induced antibody titers may decline more rapidly than immunocompetent patients due to ongoing exogenous immunosuppressive use that inhibits B- and T-cell proliferation and responses. In this study, we demonstrate that antibody titers undergo a significant decline over a 3-year period for almost all serotypes. We also show that approximately 21% of subjects that initially respond to at least one serotype lose that response in 3 years. Several serotypes demonstrate very large declines in response rate over time. For example, with serotype 4, the response rate dropped from 40.4% at 8 weeks to 17.0% at 3 years and with serotype 9V the response rate declined from 44.7% to 21.3% by 3 years postvaccination. Patients more likely to lose the response were those that had a weaker initial response to vaccination.
The conjugate pneumococcal vaccine was developed to enhance the immunogenicity of capsular polysaccharides by covalent conjugation with carrier proteins. PCV7 elicits a T-cell-dependent response, which, in turn, induces the T helper cell to stimulate polysaccharide-specific B cells to mature into either antibody-producing plasma cells or memory cells (13). Conjugate vaccine results in increased antibody production, and the antibodies produced are of high avidity. The efficacy of conjugate vaccine has been demonstrated in infants (5). Enhanced immunogenicity has also been found in patients with sickle cell disease and Hodgkin's lymphoma (8,9). However, in our previous study of 60 renal transplant recipients, no differences in immunogenicity of PCV7 versus PPV23 were shown at 8 weeks postvaccination (10). We also demonstrate that conjugate vaccine does not lead to a more durable antibody response for any of the 7-serotypes covered in this vaccine when compared to polysaccharide vaccine. In fact, the only predictor of a durable response was the strength of the initial response regardless of the type of vaccine given.
Initial responses and durability of response were very variable depending on the pneumococcal serotype. For, example, serotype 6B gave a particularly poor response at 8 weeks (6 of 47 (12.8%)); response was maintained at 3 years in four of these patients all in the PCV7 group. Serotype 6B is known to be a poor immunogen and therefore, 4 μg of 6B is included in PCV7, in contrast to 2 μg of the other serotypes (16). In contrast, for serotype 14, both initial response (20 of 47 (42.6%)) and durability of response 17 of 20 (85%) were quite reasonable. It should be noted that antibody titers are a surrogate marker of pneumococcal disease and absolute titers may not predict protection from invasive disease. However, a protective titer of 0.35 μg/mL has been estimated through pooled analysis of vaccine efficacy studies with invasive disease endpoints and is discussed in the WHO technical report (15). Additionally, studies of pneumococcal vaccination in the elderly have shown a twofold response to be significant (17). Therefore, we chose a conservative and strict measure of response as a twofold rise in titer and an absolute titer greater than 0.35 μg/mL.
There is limited data on longitudinal follow-up of pneumococcal antibody levels in organ transplant recipients. McCashland et al. vaccinated patients with chronic liver disease who went on to receive a liver transplant. Pneumococcal antibody levels declined to baseline levels rapidly in the first few months posttransplant (18). In an early study of 33 renal transplant recipients on azathiaprine- and prednisone-based immunosuppression (precyclosporin era), patients were found to have significantly waning antibody levels at 2 years post-polysaccharide vaccine (11,12). In a more recent study, seven heart transplant recipients were found to have a decline in antibody titer of 50–80% over 2 years (19). Similar to our study, no correlation with any specific immunosuppressive agent and waning of antibody response has been demonstrated. The durability of vaccine response has also been assessed in nontransplant populations. For example, Sankilampi et al. evaluated 62 patients of age >65 who were followed for three years post-polysaccharide pneumococcal vaccination. Titers declined close to pre-vaccination values for five of the seven serotypes tested (20).
Our study had several limitations. First, this was a substudy of a randomized trial whose primary endpoint was an 8-week response. However, the three-year assessment of vaccine response was specified a priori in the initial study protocol. We are also limited by the relatively small sample size and the dropout of 13 of 60 (21.7%) of patients from the original cohort. Due to this small sample size, the logistic regression analysis should be regarded as exploratory only since the power to detect variables of significance is limited. Another limitation was that unlike our original study, we did not perform the opsonophagocytic assay at the 3 year time point. This assay provides information on the functionality of the antibodies and is shown to have a moderate correlation with the ELISA (8,21). For several reasons, we elected not to perform the OPA in this study. These include (1) only a modest correlation with antibody titer was observed in the original study; (2) the 7-serotype OPA test available to us is not a standardized or validated assay and (3) precise cutoffs and protective levels for OPA titers have not been as well-established as for quantitative antibody titers.
In summary, this is the first longitudinal study that provides data to 3 years postvaccination with both polysaccharide and conjugated vaccines. The current recommendations by the AST suggest revaccination of organ transplant recipients with polysaccharide vaccine at 3–5 years (3). Our findings support this revaccination practice. In adults, the use of conjugate vaccine does not appear to enhance durability of response. Overall, it is likely that more novel strategies are required to optimize responses. For example, the use of pneumococcal conjugate vaccine as a priming vaccine followed by a booster of polysaccharide vaccine (prime-boost strategy) may be one alternative that could be studied further.
We would like to thank Dr. Moon Nahm, PhD, and Mr. Robert Burton for assistance with laboratory procedures.