# A Reassessment of Serious Adverse Events After Human Papillomavirus Vaccination in the Nagoya Survey in Japan by Using Incidentality Analysis

## TL;DR

This study uses incidentality analysis to reassess serious adverse events after HPV vaccination in Japan, finding possible associations with symptoms like chronic pain and memory issues.

## Contribution

The study introduces incidentality analysis as a novel method for detecting potential vaccine-symptom associations when traditional methods fail due to low incidence rates.

## Key findings

- HPV vaccination was associated with a higher number of cases where symptoms occurred after vaccination (MDif >0) for 22 serious symptoms.
- Subgroup analyses showed significant associations in both 15-18 and 19-21-year-old age groups, with more symptoms linked in older participants.
- The study highlights the potential of incidentality analysis for signal detection in vaccine safety assessments.

## Abstract

Introduction

Many studies have assessed the safety of human papillomavirus (HPV) vaccination. However, analyses based on statistical methods examining incidentality remain scarce. Such analyses may provide evidence for the compensation covering the healthcare costs or loss of income due to adverse vaccination events in affected patients, particularly when the incidence rate of adverse events is exceptionally low, to the extent that a significant difference is not detected in cohort and case-control studies. This study aimed to reassess whether HPV vaccination influences the occurrence of serious adverse events using incidentality analysis of data from the Nagoya City Cervical Cancer Immunization Program Survey.

Methods

The survey was conducted between September and November 2015, using a self-completed mail survey design. The data were obtained from 30,793 women aged 15-21 years living in Nagoya City (response rate of 43.3%; 30,793/71,177). Vaccines were administered using either the bivalent (Cervarix, GlaxoSmithKline, London, UK) or quadrivalent (Gardasil, Merck & Co., Rahway, NJ) HPV vaccine. From the survey data, the following items were selected for analysis: date of birth, the presence of 24 symptoms experienced from the sixth grade of elementary school to the survey date, the onset date of each symptom, and the date of HPV vaccination. The variable MDif was defined as the number of months from the date of the most recent vaccination before the symptom onset date to the symptom onset date. MDif was calculated for each symptom. The number of cases with MDif >0 and MDif <0 was compared across 24 symptoms using Fisher’s exact test. Initial analyses included all participants aged 15-21 years, followed by subgroup analyses based on two age categories: 15-18 and 19-21 years.

Results

The 30,793 survey responses included 20,912 vaccinated participants (mean age: 19.2 years, standard deviation (SD) 2.1 years), with 13,388 cases of the bivalent HPV vaccine, 4,244 cases of the quadrivalent HPV vaccine, and 3,280 unclear cases. In the initial analyses, the number of cases with MDif >0 was significantly greater than that with MDif <0 for 22 symptoms, including serious adverse events like chronic pain, motor impairment, memory impairment, poor concentration, visual disturbances, and autonomic neuropathy. In the subgroup analyses, the number of cases with MDif >0 was significantly greater than that with MDif <0 for 16 symptoms in the 15-18-year age group and for 21 symptoms in the 19-21-year age group.

Conclusions

This incidentality analysis suggests that HPV vaccination may influence the occurrence of some symptoms and might be associated with them, including serious adverse events. This approach may not provide conclusive evidence, but it can offer valuable insights into the assessment of HPV vaccine safety and may serve as a useful tool for signal detection. Furthermore, the results may contribute to determining appropriate compensation for affected patients when the incidence rates of adverse events are exceptionally low. As this study is based on questionnaire data from 2015, the robustness of the evidence is limited. Thus, to strengthen the evidence derived from this study, future research should be conducted by employing the same incidentality analysis based on up-to-date physician-recorded data with minimized reporting bias, rather than questionnaire-based data, focusing on symptoms rather than disease diagnoses.