Cardiovascular, Neurological, and Immunological Adverse Events and the 23-Valent Pneumococcal Polysaccharide Vaccine

This cohort study evaluates the association between cardiovascular, neurological, and immunological adverse events and the 23-valent pneumococcal polysaccharide vaccine in older adults.


Introduction
Globally, pneumonia remains one of the leading causes of mortality in older adults, resulting in approximately 2.5 million deaths annually. 1,2Consequently, intensive management strategies are becoming increasingly important for public health, which has lead to increased immunization policies for prevention of pneumonia.According to the current recommendations of the Advisory Committee on Immunization Practices, all older adults should receive 1 dose of the 23-valent pneumococcal polysaccharide vaccine (PPSV23). 3Consistent with these recommendations, the Korea Disease Control and Prevention Agency implemented a single-dose schedule of the PPSV23 vaccine in its national immunization program beginning in May 2013 for adults aged 65 years or older. 4To date, over 5 million older adults have received PPSV23, and the vaccination rate is reported to be 56.2%. 5 PPSV23 is generally deemed safe based on more than 30 years of evolving clinical experience. 6st adverse events following PPSV23 vaccination are mild and transient in nature and resolve within a few days. 7Previous studies have reported no evidence of an increased risk of cardiovascular outcomes in older adults following PPSV23 vaccination. 8,9Notably, a meta-analysis suggested potential cardioprotective effects of the pneumococcal vaccine, including reduced risk of myocardial infarction and all-cause mortality. 10However, these studies were generally limited by statistical power or residual confounding, and the current knowledge is predominantly based on White populations.Moreover, recent pharmacovigilance studies have identified potential signals of PPSV23 for various outcomes, [11][12][13] necessitating further signal evaluation studies.
Given the importance of obtaining updated evidence and massive immunization with PPSV among the elderly population in South Korea, it is crucial to conduct a study that minimizes the proposed limitations by using a self-controlled design to provide public confidence regarding the safety of PPSV23.Therefore, our objective was to assess the potential association between cardiovascular, neurological, and immunological adverse events and PPSV23 vaccination in older adults using a nationwide large linked database.

Methods
This study was approved by the institutional review board of Sungkyunkwan University, which waived the requirement for informed consent because only deidentified data were used.This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Data Sources
We assembled a large linked database by linking the Korea Immunization Registry Information System (KIRIS) to the National Health Information Database (NHID) from July 2018 to June 2021.KIRIS encompasses all inoculation data for 18 types of vaccines covered under the National Immunization Program (NIP).The NHID includes demographic and clinical information of the entire population of 50 million South Koreans. 14Using anonymized identifiers, we extracted variables including age at vaccination, date of vaccination, type of vaccine, dosing schedule information, route of vaccination, and site of vaccination from the KIRIS.Additionally, we obtained data on medical diagnosis, date of diagnosis, prescriptions, procedures, health care utilization records, medical insurance type, region of residence, and income level from the NHID.A detailed description of this large linked database is provided elsewhere. 15 linking these 2 data sources, we could leverage information on exposure and outcomes to assess the association between PPSV23 and adverse events.All procedures and prescriptions were coded using the national codes.Diagnostic information was coded according to the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD -10).A previously conducted validation study reported a positive predictive value of 82% for diagnosis codes in claims data compared with diagnoses from electronic medical records, which are regarded as the reference standard. 16

JAMA Network Open | Pharmacy and Clinical Pharmacology
Adverse Events and the 23-Valent Pneumococcal Polysaccharide Vaccine

Study Population
Older adults aged 65 years or older who received PPSV23 between July 2018 and June 2021 were eligible for the study.Among these eligible individuals, we excluded individuals with the following conditions: those with a record that was inconsistent with current PPSV23 vaccination guidelines or NIP criteria for older adults (such as having received 2 or more doses of PPSV23, being younger than 65 years of age, or having pneumococcal conjugate vaccination records), missing information on age or sex, and those who died between the start date of the predefined risk interval and the end date of control intervals to ensure an adequate follow-up period for outcome assessment.
In South Korea, health care clinicians are required to report detailed information on immunizations to KIRIS to request reimbursement of immunization fees from the governmental body for all vaccines covered under the NIP.Therefore, records of PPSV23 immunization were obtained from KIRIS.Cohort entry was defined as the date of PPSV23 immunization.Baseline characteristics, such as age at vaccination, sex, region of residence, type of health insurance, income level, comorbidities, history of medication use, Charlson Comorbidity Index, health care utilization, and history of seasonal influenza vaccination were assessed at cohort entry.

Outcomes
We defined 6 cardiovascular outcomes (myocardial infarction, atrial fibrillation, cardiomyopathy, heart failure, hypotension, myocarditis or pericarditis, and stroke), 2 neurological outcomes (Bell palsy and Guillain-Barré syndrome), and 3 immunological outcomes (sepsis, thrombocytopenia, and anaphylaxis) as the primary outcomes of interest.13]17 Each outcome was identified using ICD-10 diagnostic codes in either the primary or secondary position in the inpatient or emergency department settings.The detailed diagnostic codes are provided in eTable 1 in Supplement 1.We also applied additional criteria to ensure the validity of the outcome definitions.For thrombocytopenia, we excluded cases with a concurrent diagnosis of cancer recorded on the same day to distinguish chemotherapy-induced thrombocytopenia.To reduce false positives in anaphylaxis cases, we only included cases with a diagnosis code for anaphylaxis and simultaneous prescription of epinephrine or systemic steroids. 18Only incident cases were included in this study, and individuals with a history of each outcome diagnosis within 1 year before cohort entry were excluded.

Self-Controlled Risk Interval Analysis
We used a self-controlled risk interval (SCRI) design that included older adults with records of both PPSV23 vaccination and at least 1 of the predefined outcomes (Figure 1).This design allowed us to minimize the impact of time-invariant confounders and exposure misclassification by comparing the risk between specific time periods within individuals who experienced an outcome, rather than For sepsis and anaphylaxis, the risk intervals were defined as days 0 to 7 and 0 to 2, respectively, including the day of vaccination.PPSV23 indicates the 23-valent pneumococcal polysaccharide vaccine; SAE, serious adverse event.
between different individuals.Moreover, it is less susceptible to the effects of unmeasured confounders than other study designs. 19Given the challenges in selecting appropriate comparator groups due to the heterogeneous characteristics of vaccinated and unvaccinated groups in older adults, the SCRI design can complement the limitations of cohort designs and serve as an alternative approach to assess the association of adverse events following PPSV23 vaccination.
The risk interval was defined as days 1 to 28 after PPSV23 vaccination, while the control interval was defined as days 57 to 112 after PPSV23 vaccination, with day 0 defined as the day of PPSV23 vaccination.1][22][23][24] We excluded outcomes that occurred on the day of vaccination to account for the possibility of reverse causation, unless stated otherwise.Additionally, we implemented a 28-day washout interval between the risk and control intervals to address the possibility that adverse events occurring beyond the predefined risk interval might be influenced by prior PPSV23 vaccination.This measure is implemented to mitigate the underestimation of risk due to outcome misclassification. 3,19For sepsis and anaphylaxis, the risk intervals were defined as days 0 to 7 and 0 to 2, respectively, including the day of vaccination, based on the pathophysiology associated with these outcomes.The same washout and control intervals as those for other events were applied to enhance statistical power for these 2 events.

Statistical Analyses
We summarized the sociodemographic, clinical, and vaccination-related characteristics of the study population.Categorical variables are presented as frequencies and proportions, while continuous variables are presented as means and SDs.To estimate the incidence rate ratio (IRR) with a 95% CI for each outcome, we used a conditional Poisson regression model.The calculation of IRR allowed us to adjust for differences in the durations between the risk and control intervals.The model was adjusted for age, which is a time-varying variable, throughout the risk and control intervals.Additionally, we presented the temporal distributions of case onset according to the number of days after PPSV23 vaccination to identify the overall clusters of adverse events following vaccination.All statistical analyses were performed using SAS Enterprise Guide 7.1 for Windows (SAS Institute).A 2-tailed P value of .05 was considered statistically significant.Data were analyzed from November 2022 to April 2023.
We conducted sensitivity analyses with different risk intervals (21, 42, 84, and 112 days) to ensure the robustness of our main analysis and to explore the possibility of capturing insidious outcomes that may not have been identified with the 28-day risk interval.We also varied the washout interval from 28 days to 14 or 42 days.To account for the potential impact of influenza vaccination or seasonality, we further adjusted for influenza vaccination status and seasonality during the risk and control intervals.Additionally, we conducted an analysis that included day 0 in the risk interval.Subgroup analyses were performed to assess the heterogeneity of the outcomes of PPSV23 vaccination based on patient characteristics such as sex, history of cardiovascular diseases (coronary artery disease, myocardial infarction, arrhythmia, hypertension, stroke, atherosclerosis, heart failure, and peripheral arterial disease), and immunocompromising status.These subgroup analyses allowed us to examine whether the association between PPSV23 vaccination and adverse events differed across specific subsets of populations.

Results
After linking the NHID of 12 222 807 individuals with the PPSV23 immunization records of 1 830 928 individuals from the KIRIS, the data of 1 830 245 adults were available (Figure 2).Of these, 27 506 were excluded for various reasons.Among the In each SCRI analysis, we did not find an increased risk of cardiovascular, neurological, or immunological events following PPSV23 vaccination (Figure 3).Specifically, for cardiovascular None of the subgroups showed a significant positive association between PPSV23 and cardiovascular (Figure 4), neurological (eFigure 1 in Supplement 1), and immunological (eFigure 2 in Supplement 1) events.Although the risk of cardiovascular events tended to increase in patients with a history of cardiovascular diseases, the association was not statistically significant.The risk of immunological events was higher in immunocompromised patients, but the small number of events led to wide CIs (sepsis IRR, to an increased risk of cardiovascular events.PPSV23 vaccination was likely associated with a higher risk of immunological adverse events such as sepsis, thrombocytopenia, and anaphylaxis in immunocompromised patients; however, the overall incidence of these events was very low.The results of the sensitivity analyses support the robustness of our main findings. Many studies and accumulated clinical experience have supported the fact that vaccination with PPSV23 is safe, and our findings reinforce this evidence.6][27] Postlicensure studies using data from the Vaccine Adverse Event Reporting System in the US 13 and General Practice data in Australia 28 also did not find any unexpected severe systemic adverse events following PPSV23 vaccination.Our results, applying a self-controlled design, provide additional evidence that there was no increased risk of adverse events of interest after vaccination, even after adjusting for all timeinvariant, unmeasured confounding factors by performing a within-person comparison. 29previous Korean study using the Adverse Event Reporting System database identified unexpected safety signals for PPSV23, including cardiovascular and immunological events.11 In line with pharmacovigilance efforts, 30 we conducted an epidemiological study to evaluate these signals but find no significant positive association between PPSV23 and adverse events.The discrepancy with the previous study may be attributed to different study designs; a signal does not indicate a causal relationship, and the information in a signal is not conclusive because it may change over time Incidence rate ratio (95% CI) a A risk interval of 28 days was applied unless otherwise specified in the main analysis.A washout interval of 28 days was applied between the risk and control intervals.For sepsis and anaphylaxis, risk intervals of 7 days and 2 days were defined, respectively, including the day of vaccination.
b Incidence rate ratio was adjusted for age.with additional data. 31The variation in age distribution among study populations could also contribute to the observed inconsistency.The former study included younger adults (19-65 years), whereas our focus was on older adults.Guidelines recommend PPSV23 for younger adults with risk factors for pneumonia, 32 potentially leading to a higher incidence of adverse events due to confounding by indication.Differences in the incidence of myocardial infarction and stroke between age groups after PPSV23 receipt were also identified in a US study. 8Future studies involving younger adults are required to assess the risk of serious adverse events associated with PPSV23 in different age groups.

JAMA Network Open | Pharmacy and Clinical Pharmacology
In a US cohort study using the Vaccine Safety Datalink project data, safety outcomes including acute myocardial infarction, atrial fibrillation, cardiomyopathy, heart failure, Guillain-Barre syndrome, thrombocytopenia, and allergic reactions in the elderly were less favorable with PPSV23 compared with pneumococcal conjugate vaccine 13 (PCV13), 33 but were comparable with the general US population aged 65 years or older. 34We also assessed the association between PPSV23 and outcomes using a nationwide linked database in Korea, demonstrating favorable overall safety.In our study, among the total PPSV23 vaccinated population, 180 (0.01%) experienced myocardial infarction within the risk interval (28 days after vaccination).Considering the incidence rates of myocardial infarction in the Korean population aged older than 60 years (1.18-3.21per 1000 person-year), 35 the occurrence in this study also seemed similar to that in the overall population.

Limitations and Strengths
This study had several limitations.First, the risk of adverse events may have been underestimated because patients who received more than 1 dose of PPSV23 (ie, prevalent users) were included in this study.Our data only had NIP-supported vaccination records; thus, we could not exclude patients who paid out of pocket for PPSV23 (before age 65 years) or PCV13.These prevalent users can contribute to mitigating early harm following drug exposure as survivors of the initial treatment. 36cond, the self-controlled design, involving only vaccinees, might also underestimate risk due to potential selection bias from including individuals eligible for vaccination.We also included individuals alive during the risk and control intervals, implying that vaccinees who died after adverse events in the risk period were excluded.Consequently, the risk of adverse events in the risk interval might be lower than that in the control interval among this population.However, we tried to reduce healthy vaccinee bias by excluding time period immediately before vaccination from our analysis. 19ird, the adverse events caused by the administration of PPSV23 may have occurred after the risk interval.We attempted to set the plausible lengths of the risk, washout, and control intervals by considering physiological mechanisms, but these operational definitions may have led to outcome misclassification.Nonetheless, this bias was unlikely to have substantially impacted our results, given the consistent results from the sensitivity analysis with varying interval lengths.
Despite limitations, our study has generalizability, using a nationwide large linked database with over 1.8 million PPSV23 recipients.To our knowledge, this is the first self-controlled risk interval study to investigate the safety of PPSV23.The risk interval design, focusing solely on vaccinated individuals, reduces bias from unmeasured differences that may exist between vaccinees and the unvaccinated, a limitation of the cohort design.Our findings align with those of previous cohort studies, supporting the overall safety of PPSV23 across diverse outcomes and study designs.

Conclusions
In this self-controlled study, our results corroborate previous findings that PPSV23 was not associated with a higher risk of serious systemic adverse events related to the cardiovascular, neurological, and immunological systems.Even among adults with a history of cardiovascular diseases or immunocompromised status, PPSV23 did not appear to elevate the risk of related events to worrying levels.These updated safety findings on PPSV23 can help alleviate concerns regarding the vaccine and encourage vaccination in elderly individuals.

Figure 1 .
Figure 1.Study Design of Self-Controlled Risk Interval Analysis Differential characteristics were observed between the SCRI population and PPSV23 vaccinated population.The mean (SD) age of the SCRI population was 72.4 (8.2) years, which was higher than that of the PPSV23 vaccinated population (68.3[5.6]years)(Table).A total of 2272 patients were male (52.1%).The proportion of patients with comorbidities was higher in patients with events 1 802 739 individuals who received PPSV23, those who experienced outcome events of interest within 6 months before vaccination were excluded, and JAMA Network Open | Pharmacy and Clinical Pharmacology 1.38; 95% CI, 0.65-2.95;thrombocytopenia IRR, 2.00; 95% CI, 0.28-14.20;and anaphylaxis IRR, 2.33; 95% CI, 0.26-20.88).Results from the sensitivity analysis, risk or control or washout intervals, adjustment for influenza vaccination status or seasonality, and day 0 inclusion in the risk interval, were consistent with the main findings (eTable 2, eTable 3, and eTable 4 in Supplement 1).The case distribution of events according to the days after vaccination with PPSV23 did not show any temporal trends of increasing risk, as shown in eFigure 3, eFigure 4, and eFigure 5 in Supplement 1.