Immune thrombocytopenic purpura after influenza vaccine administration; a systematic review and meta-analysis

Abstract

Background

The American Society of Haematology defines immune thrombocytopenic purpura (ITP) as a common hematologic disorder characterized by a transient or long-term decrease in platelet counts (< 100 × 109/L.), purpura, and haemorrhagic episodes caused by antiplatelet autoantibodies, with the exclusion of other clinical conditions. We aimed to systematically determine the incidence of ITP in adults and children following influenza vaccination, the duration between vaccination and the occurrence of ITP, and to identify predictors of ITP after the vaccine.

Methods

We searched PubMed, Cochrane Library, Google Scholar, Web of Science, Scopus, and Science Direct. We included primary studies that assessed the occurrence of immune thrombocytopenia in individuals who had received any influenza vaccine (primary or booster dose), regardless of the dosage, preparation, time of administration, or age of the participants. We excluded studies that were (a) Narrative, scoping, and umbrella reviews ;(b) studies with no accessible full text, abstract-only studies, or (c) Overlapping or unreliable data. The risk of bias in the included studies was assessed using the Joanna Briggs Institute (JBI) tool. We categorized studies for qualitative analysis based on study design. Descriptive statistics were used to summarize quantitative data, including the incidence of ITP after influenza vaccination.

Results

Out of 729 articles retrieved from the database search, we included 24 studies. All patients identified and included in this systematic review presented with immune thrombocytopenia, determined by their platelet count. The period between vaccination and the occurrence of ITP ranged from (2:35 days). The mean duration was 13.5 days. The analysis revealed a statistically significant incidence rate ratio (IRR) = 1.85,95% CI [1.03–3.32] of ITP occurrence after 42 days.

Conclusions

Influenza-associated ITP is uncommon, self-limiting, non-life-threatening, and curable. None of the patients reported having severe adverse events or death. Further studies are required to confirm the exact incidence of the ITP to better understand the pathophysiology of ITP development post-influenza vaccination.

Introduction

Immune thrombocytopenic purpura (ITP), also known as idiopathic thrombocytopenic purpura, is an acquired bleeding disorder in adults and children that is characterized by thrombocytopenia leading to bleeding episodes that range in severity from a purpuric rash to epistaxis and bleeding. In addition, patients may experience bleeding, such as intracranial and intestinal bleeding, in severe cases with a platelet count of < 20,000/uL [1]. ITP occurs with an incidence rate of 1.6 to 3.9 per 100,000 patient-years, which increases with age and has a slight female preponderance. The age-adjusted prevalence of ITP is estimated at 9.5 per 100,000 persons in the USA. In contrast, its annual incidence is estimated to be 2.68 per 100,000 in Northern Europe [2]. A bone marrow biopsy is used to confirm ITP and to rule out other common causes of thrombocytopenia [1].

Genetic variations may contribute to ITP susceptibility, although the cause is complex, poorly understood, and due to multifactorial causes. Autoantibody-induced platelet destruction reduces platelet synthesis, and T-cell abnormalities, such as Th1 polarization, Th17 overabundance, and regulatory T-cell deficiency, are possible reasons for immune dysregulation leading to immune thrombocytopenia [3].

Other factors associated with the pathogenesis of ITP are drugs, infections, cancers, autoimmune diseases, and vaccinations such as the influenza vaccine [4] [5]. Despite protective immunity, vaccines might induce autoimmune responses. Although the pathogenesis of influenza vaccine-induced thrombocytopenia is still unknown, the literature suggests that ITP risk increases following vaccination through the exact mechanism of anti-platelet autoantibodies formation by microbial infections. Vaccines stimulate protective immunity by mimicking body mechanisms; both live and inactivated vaccines cause the development of ITP [6] [7]. A systematic review (SR) described an association between the measles-mumps-rubella (MMR) vaccine and ITP in children. They concluded that only children with persistent or chronic ITP who must receive MMR require care [8]. In addition, surveillance systems revealed that vaccines, such as hepatitis A, diphtheria, tetanus, and pertussis (DTP), diphtheria, tetanus, acellular pertussis (DTaP), and varicella vaccines, were associated with ITP in Canada between 1992 and 2007 [9].

The severity of ITP, patients’ criteria, platelet count, bone marrow changes, and presence of antiplatelet antibodies vary between studies.

There is a scarcity of reliable data concerning the incidence of ITP following influenza vaccination. Surveillance system reports are prone to reporting bias because they depend on each country’s vaccination schedule. There is insufficient data on clinical outcomes such as bleeding manifestations and the development of chronic thrombocytopenia after the influenza vaccination. Therefore, we conducted a SR and meta-analysis of the available medical literature to calculate the incidence of ITP after influenza vaccination.

Methods

Study design

This SR and meta-analysis was conducted according to the PRISMA guidelines, as shown in (Supplementary Tables 1 and 2). The study protocol was previously registered with the International Prospective Register of Systematic Reviews (PROSPERO) database as CRD42022313947.

Search strategy

We conducted a systematic literature search in six databases: PubMed, the Cochrane Library, Google Scholar, Web of Science, Scopus, and Science Direct in March 2022. The search terms were modified to match the databases. The data were independently extracted using a standardized Excel sheet containing the primary baseline characteristics and the outcomes of interest.

Selection criteria

As per the study protocol, we included primary studies that reported the occurrence of ITP after influenza vaccination in adults or children. However, we excluded secondary studies, studies with no accessible full text, and abstracts. To ensure data quality, overlapping data were removed. We included all studies that met our criteria, regardless of age, country, language, or publication date.

Study selection

Our search results were exported into EndNote (Thompson Reuter, USA) and then to CADIMA to detect and remove duplications. All included studies were imported into CADIMA for the title and abstract screening by two authors independently, and a senior reviewer was consulted to make the final decision in case of any inconsistency. Full-text screening of included studies was performed by two independent reviewers using our criteria, and any disagreements were resolved through a discussion.

Data extraction

We developed a standardized extraction sheet based on the desired outcome to extract relevant data from the included studies; that assessed and answered our review questions. The extraction sheet was divided into two sections: one for case reports and case series and the other for other study designs. Two independent reviewers extracted all relevant data, and a consensus was reached on any differences after a discussion. If there was no agreement, a senior reviewer was consulted to make the final decision.

Assessment of risk of bias

Three independent reviewers evaluated the quality of the included studies for bias using the Joanna Briggs Institute critical appraisal tool. The overall risk of bias was categorized into good, fair, or poor. However, there was no tool for assessing the risk of bias for self-controlled case series studies. Thus, they were not applicable for assessment.

Analysis

Extracted data was cleaned and prepared for qualitative and quantitative analysis. The odds ratio was extracted and analysed to measure the burden of ITP after influenza vaccination compared to no exposure. The meta-analysis was performed using R software (version 4.1.3) and the (meta) package. A subgroup analysis was performed to assess different factors influencing the results. Heterogeneity was calculated and was considered significant when I² was greater than 50% with a p-value < 0.05. A fixed effect model was used unless heterogeneity was significant. On the other hand, data that was unsuitable for statistical analysis was assessed qualitatively through table presentation and interpretation.

Results

Search results

The initial electronic database search yielded 729 studies; 71 met our inclusion criteria, and 21 were eligible after the full-text screening. A total of 45 studies were excluded by full-text screening, according to the PRISMA flow chart (Fig. 1). A manual search yielded three additional studies. As a result, only 24 studies were included in our review. Additionally, a detailed illustration of the results of each database is provided in (Supplementary Table 3).

Fig. 1

PRISMA flow diagram of included studies

Risk of bias assessment

Out of 24 included articles, 10 received good scores, 11 got fair and 3 self-controlled case series (SCCS) received no quality assessment. The score of each article is shown in Supplementary Tables (4, 5, 6, 7 and 8). We have included 1 RCT study with a good score. Out of 11 case report studies, 4 received a good score, and 7 received a fair score. Two Cohorts and 2 case controls received a good score. Of 5 cross sectional studies, 1 received a good score and other 4 got a fair score.

Qualitative analysis of reported cases

Study characteristics

Eleven case series that comprised 13 patients were identified and described [10,11,12,13,14,15,16,17,18,19,20]. The mean age was 52.9 years (3–88), and 53.8% of patients were male. The most prevalent geographic location where thrombocytopenia was reported following influenza vaccination was Japan (6, 46%). Several patients presented with chronic diseases prior to receiving the influenza vaccine, including hypertension (2, 15%), primary biliary cirrhosis (1, 8%), and hepatitis C (1, 8%).

All patients had received an influenza vaccination; however, only 4 case series reported the name of the vaccine. Of these 4, 2 received Vaxigrip, and 1 received Fluarix, and 1 patient received Kitaken. Only 2 patients were co-administered with the influenza vaccination and other vaccines, including DTP/Polio/Hib, Hepatitis B, and pneumococcal conjugate vaccine. The ITP occurred following 1 dose of the influenza vaccine in 3 patients (23%), following three doses in 2 patients (15%), and following 2 doses in 1 patient (8%). The remaining 7 case reports did not present data on how many vaccine doses the patients received when ITP occurred (Table 1).

Table 1 Study characteristics of included case reports

ITP clinical presentation

All patients identified and included in this SR presented with ITP, diagnosed after examining their platelet count. Table 2 describes the platelet count before and after vaccination for each patient, with a substantial decrease being observed in all patients. The period between vaccination and the occurrence of ITP differed greatly between patients, ranging between 2 and 35 days. The mean duration was 13.5 days. Bone marrow biopsy findings were described in 8 cases, with hypercellularity observed in 5 cases (38%) (Table 2). Purpura or a bleeding event was described in 12 cases (92.3%), including purpuric rashes, epistaxis and petechiae, haemoptysis, bruising, buccal haematoma, bleeding gums, petechial lesions to the sclera, nasal bleeding, gross haematuria, cutaneous and mucosal bleeding, genital bleeding, and bleeding blisters.

Table 2 ITP clinical presentation in included cases

Quantitative analysis

A total of 13 studies [4, 13, 19, 21,22,23,24,25,26,27,28,29,30] reported the occurrence of ITP after influenza vaccination. Five [5, 27,28,29, 31]studies were surveillance representing 38.4% of included studies, while self-controlled case series were presented in 3 [13, 19, 25]studies. Cohort [24, 30]and case-control [4, 22] studies were conducted in 2 studies, while only 1 [23] study was a randomized controlled trial. More details regarding the country and type of influenza vaccine are provided in (Table 3).

Table 3 Study characteristics of included observational studies

Only 4 [24, 25, 27, 30] studies were included in the analysis of the incidence rate ratio (IRR) of ITP occurrence 42 days after receiving influenza vaccination. Generally, the analysis showed a positive statistically significant IRR = 1.85, 95% CI [1.03–3.32]. However, heterogeneity was statistically significant with I² = 0.66% and a p-value < 0.01. Thus, a subgroup analysis according to the age of patients was performed, and it was only applicable in 3 studies [24, 25, 30]for the age group above 65 years. This yielded a statistically insignificant positive IRR = 2.13, 95% CI [0.71–6.43] with I² = 83% and p-value < 0.01 (Fig. 2).

Fig. 2

Meta-analysis of incidence rate ratio (IRR) of ITP after 42 days of influenza vaccine administration

Discussion

This study reviewed 24 articles on thrombocytopenia following the influenza vaccination. All 13 cases reported a drop in platelet count post-vaccination, indicative of ITP development. Furthermore, the majority of authors stated that the influenza vaccination was responsible for the onset of ITP, occurring at a mean duration of 13.5 days post-vaccination. A large proportion of cases were reported from Japan, and this geographical location was discussed among observational studies that investigated ITP development following other live vaccinations.

Few patients had chronic disease prior to vaccination, implying that thrombocytopenia developed irrespective of pre-existing conditions. ITP is characterised by its heterogeneous manifestations and unpredictable outcomes; however, the current literature suggests several risk factors for ITP development and chronic disease. ElAlfy et al. sought to identify the predictors for ITP in 409 patients with a confirmed ITP diagnosis. Several relevant risk factors were observed for the development of chronic ITP, specifically an initial platelet count of less than 20 × 109/L and a presentation age of more than ten years [32]. The risk of the development of ITP is increased depending on a patient’s platelet count, if they are female, and their exposure to NSAIDs [33].

The pathophysiology of vaccine-induced thrombocytopenia is poorly understood, with few large-scale research studies observing significant findings or concerning patterns. Haber et al. conducted an observational study investigating adverse outcomes following quadrivalent live attenuated influenza vaccination in the United States. The most common adverse events were neurologic conditions, including seizures and Guillain-Barré syndrome. Asthma and wheezing were also reported in a few children. However, no occurrence of ITP was reported among all 779 participants. Besides, no concerning patterns of adverse events were identified [34].

The cases identified in this SR reported on the occurrence of thrombocytopenia in adult patients. However, few studies reported on the incidence of ITP in this population. Moreover, this review focused exclusively on the development of ITP following the influenza vaccination, yet, current research extends to other live, inactivated, and simultaneous vaccinations. Yokomichi et al. assessed the risk of ITP following live, inactivated, and simultaneous vaccination in children under two years old. Despite the limitation, the findings suggest no significant ITP risk following single vaccinations or simultaneous vaccination spanning all age groups, including children [22].

Although detailed data was presented in case reports, a few observational studies are worth mentioning. Nakayama et al. investigated several cases of vaccine adverse events reported in the Kitasato Institute’s post-marketing surveillance, categorising patient responses into allergic reactions and severe systemic illnesses. The evidence concerning the relationship between thrombocytopenia and the influenza vaccination describes 12 cases of ITP. However, the occurrence of ITP was not significant, given the large sample size included in this report. This is reflected in the estimated incidence of severe neurological illness post-vaccination at 0.1–0.2 per million immunisation practices [5]. Additionally, Villa et al. discussed the safety of MF59-adjuvanted influenza vaccination in the elderly population of Northern Italy. Despite the fact that 170,998 vaccine doses were administered to 107,661 individuals during the study period, adverse events that requiring hospitalisation was rare. This emphasizes the safety of seasonal influenza vaccinations in older people [24].

Lafaurie et al. assessed the risk of ITP in a nationwide study in France, comprising 4394 patients with incident primary immune thrombocytopenia. The number of patients that had received at least one dose of the influenza vaccination was included. However, there was no increased risk of ITP following the influenza vaccine [30]. O’Leary et al., on the other hand, investigated the risk of ITP following childhood vaccines besides MMR amongst a cohort of 1.8 million children aged six weeks to 17 years. In total, 197 confirmed ITP cases were identified, indicating no elevated risk of ITP after any vaccine in early childhood. However, a significantly increased risk of ITP was observed following the hepatitis A vaccination in children aged seven to 17 years and for varicella and tetanus-diphtheria-acellular pertussis vaccination in adolescents aged 11 to 17 years [27].

Our SR has a few limitations. The review is based on published cases and observational studies. However, most cases of ITP are known to be asymptomatic [35] and, therefore, might not have made it to scientific reportage. It is possible that the cases we reviewed may not have been representative of the whole population. The influenza vaccine is usually administered on a priority basis which tends to skew towards the elderly and those with other comorbidities [36]. Therefore, the reviewed studies might have inadvertently left out a critical sample. Furthermore, because case reports were included, the level of evidence may be low. We may have missed some of the investigation results due to reporting bias. However, the bias was mitigated by strict inclusion and exclusion criteria and a comprehensive analysis of all included articles by multiple researchers. Due to the small number of reported cases, it was impossible to compare the characteristics and severity of ITP across regions. Meta-analysis was not conclusive because of heterogeneity.

Conclusion and future directions

This SR corroborated that ITP post-influenza vaccinations is rare. Several observational studies also substantiate the safety profile of influenza vaccinations. None of the patients reported having severe adverse events or death. More research is required to ascertain the true incidence and pathophysiology of ITP post-influenza vaccination. Further large-scale prospective studies are required to establish the characteristics and risk factors of influenza vaccine‐associated ITP. Further research is also needed to explore the influenza vaccine’s causal role in the development of ITP by investigating epitope similarities between platelets and vaccine‐driven antigens. The potential disparity between the different influenza vaccine brands to induce ITP warrants further exploration and may aid in optimizing vaccine production to produce vaccines with a lower risk of triggering autoimmunity.