Population-Based Incidence of Guillain-Barré Syndrome During Mass Immunization With Viral Vaccines: A Pooled Analysis

Misunderstanding temporal coincidence of adverse events during mass vaccination and invalid assessment of possible safety concerns have negative effects on immunization programs, leading to low immunization coverage. We conducted this systematic review and meta-analysis to identify the incidence rates of GBS that are temporally associated with viral vaccine administration but might not be attributable to the vaccines. By literature search in Embase and PubMed, we included 48 publications and 2,110,441,600 participants. The pooled incidence rate of GBS was 3.09 per million persons (95% confidence interval [CI]: 2.67 to 3.51) within six weeks of vaccination, equally 2.47 per 100,000 person-year (95%CI: 2.14 to 2.81). Subgroup analyses illustrated that the pooled rates were 2.77 per million persons (95%CI: 2.47 to 3.07) for individuals who received the influenza vaccine and 2.44 per million persons (95%CI: 0.97 to 3.91) for human papillomavirus (HPV) vaccines, respectively. Our findings evidence the GBS-associated safety of virus vaccines. We present a reference for the evaluation of post-vaccination GBS rates in mass immunization campaigns, including the SARS-CoV-2 vaccine.


INTRODUCTION
The coronavirus disease-2019 (COVID- 19), induced by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has been challenging all over the world since December, 2019 (1). As of December 29, 2021, the total number of confirmed cases is over 281 million worldwide, including more than five million deaths (2). SARS-CoV-2 infection is commonly characterized by fever,

MATERIALS AND METHODS
This study was conducted by reference to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline (20), which is provided in Table S1.

Literature Screening and Selection
Firstly, the titles and abstracts of the publications were reviewed by two authors (FW and DW) independently. Secondly, the full text and online supplementary data were read to determine the eligibility of the publications. Any uncertainties and discrepancies were resolved by discussion with the third author (YW). The inclusion criteria were: 1) studies that reported temporal coincidence of GBS in mass immunizations; 2) participants received viral vaccines, including but not limited to influenza vaccine, HPV vaccine, polio vaccine, hepatitis vaccine, measles-rubella vaccine, rubella vaccine, or measlesmumps-rubella (MMR); 3) the following data were available or can be calculated: number of GBS patients, number of vaccinated populations, or background rate of GBS after vaccination. The quality of the literature was assessed by two authors (CL and YW) independently (Table S2).
Studies matching the following items were excluded: 1) reviews, case report studies, letters and conference abstracts; 2) animal studies; 3) clinical studies evaluating the safety of vaccines; 4) studies did not provide the information of vaccines in detail; 5) studies with the duplicate publication or overlapping data. ethnicity, region, gender, age-range); 4) information of vaccines (e.g., the target viruses of vaccines, type of vaccine, follow-up duration after vaccination, the valence of vaccines and adjuvants of vaccines); 5) the number of study participants; 6) background rate and/or number of coincident cases of GBS during vaccination; 7) sources of vaccination. If there were duplicate data, the studies with larger sample size or newly published ones were involved.

Statistical Analysis
The meta-analysis, a statistical procedure for the combination of the results from multiple independent studies, was performed using STATA version 16.0 (STATA Corp, College Station, TX, USA) and/ or R version 4.0.4 (Foundation for Statistical Computing, Vienna, Austria), by which the pooled background rate and its 95% confidence interval (CI) were calculated. Cochran's Q-test and I 2 statistics were applied to measure the significance of heterogeneity across eligible studies. Heterogeneity was assumed insignificant if P > 0.05 and I 2 < 50%, then a fixed-effect model meta-analysis was carried out, otherwise, heterogeneity was considered statistically significant, then the Der-Simonian Laird's random-effect model was implemented (21). Moreover, subgroup analyses were conducted on the basis of gender, age range, ethnicity, target virus, type of vaccines, follow-up duration after vaccination, the valence of vaccines and adjuvants of vaccines. To estimate the stability of the pooled results and distinguish the potential influence of individual studies, a sensitivity analysis was conducted by sequential removal of every single study one at a time. In addition, the publication bias was modeled by the funnel plots and analyzed by the Egger's test. Furthermore, the trim-fill method was used to adjust for publication bias when it is significant. A P < 0.05 was considered significant if not mentioned specifically.

Literature Search
A total of 2,201 publications (1,081 from PubMed and 1,120 from Embase) were retrieved, among which 943 duplicate records were excluded. After reviewed by titles and abstracts of the remained studies 1,124 publications were excluded for the following reasons: 188 studies were reviewed, case report studies, letters and conference abstracts; 102 were no-human-based researches; 179 were clinical trials; 226 were not studies on the incidence of GBS in vaccinees; 305 were not studies on virus vaccines; and 124 were researches on the mechanism of GBS. Among the 134 articles evaluated by full-text, 64 were excluded due to no details of vaccines or doses; 17 were excluded due to no number of coincident cases of GBS during vaccination; 5 were excluded due to duplicate data.
We checked the database of original studies on GBS after vaccination, and excluded one study of seasonal influenza vaccine in the U.S (22), one H1N1 vaccine study in the U.S (23), one HPV vaccine study in the U.S (24), and two studies of H1N1 vaccine in China due to duplicate data (25,26). Finally, 48 publications with 58 independent studies were included in our meta-analysis (12,. The flow chart of literature screening process is shown in Figure 1.

Characteristics of Included Studies
Of the 58 studies, 45 reported the incidence of GBS for influenza vaccines, seven reported HPV vaccines, one reported polio vaccine, one reported hepatitis vaccine, two reported Measles-Rubella vaccines, one reported Rubella vaccine, one reported measles-mumps-rubella vaccine (MMR) vaccine. There are 20 studies on inactivated virus vaccines, seven on live-attenuated vaccines, one on the recombinant vaccine and one on the splitvirion vaccine. With regard to the duration of follow-up, 32 studies reported the background rate of GBS within six weeks after vaccination. The details are listed in Table 1.

Pooled Results of Post-Vaccination GBS Rate
As shown in Figure 2, the pooled GBS rate, synthesized by a random-effects model, was 5.29 per million (95% CI: 3.66 to 6.93 per million) after immunization of viral vaccines. The heterogeneity test showed significant heterogeneity between studies (I 2 = 98%, P < 0.01). Since the time period between 0 and 6 weeks is considered as the risk window after vaccination, we evaluated the GBS rate in this period. As result, the pooled rate was 3.09 per million persons (95% CI: 2.67 to 3.51 per million) for the 42-day window, equally 2.47 per 100,000 person-year (95%CI: 2.14 to 2.81 per 100,000 person-year). In contrast, as shown in Table S3, the previous studies that estimated incidence rates of GBS within general populations showed a range from 0.42 to 2.42 per 100,000 per person-year, meaning that there was no significant increase in GBS among population received viral vaccines.

Subgroup Analyses
Subgroup analyses were performed on the basis of gender, age, ethnicity, target virus, type of vaccines, the valence of vaccines and adjuvants of vaccines. As shown in Table 2, the pooled incidence rates of GBS were 7.26 per million (95%CI: 3.11 to 11.41 per million) among people aged <18 years, 0.99 per million (95%CI: 0.24 to 1.73 per million) among people aged 18 to 59 years, and 6.06 per million (95%CI: 2.51 to 9.61 per million) among people aged ≥ 60 years of age. The pooled rates were 6.31 per million (95%CI: 0.81 to 11.82 per million) among men, and 6. 41

Publication Bias and Sensitivity Analyses
Funnel plot analysis and Egger's test were used to examine the significance of publication bias underlying our study, by which a statistical significance was identified (Table 2 and Figure 3). In order to control publication bias the trim-fill method was further performed, by which the pooled GBS rate was 1.71 per million (95% CI: 0 to 3.96 per million) after immunization of vaccine against virus, and 1.89 per million (95% CI: 1.48 to 2.30 per million) in 6 weeks follow-up after immunization.
To examine the strength of the pooled results, we performed a sensitivity analysis by omitting one study at a time.
Consequently, the pooled result was not dominantly affected by any of the individual studies ( Figure S1), indicating high stability of our results.

DISCUSSION
To the best of our knowledge, this study is the first to comprehensively summarize the incidence rates of GBS following mass immunizations of viral vaccines. Our metaanalyses, involving 58 original studies and 2,110,441,600 participants, identified a pooled rate 5.29 per million (95% CI:3.66 to 6.93 per million) among people received viral vaccines, and a pooled rate 3.09 per million (95% CI:2.67 to 3.51 per million) in 6 weeks of vaccination, equally 2.47 per 100,000 person-year (95%CI: 2.14 to 2.81 per 100,000 person-year). There was no significant increase in GBS incidence among population received viral vaccines compared to general population without prior vaccination. Subgroup analyses released the pooled rates of 2.77 per million (95%CI: 2.47 to 3.07 per million) for individuals received influenza vaccine and 2.44 per million (95%CI: 0.97 to 3.91 per million) for HPV vaccinees, respectively.
GBS is a demyelinating transient neurological disorder characterized by lack of paralysis and sensory impairment. GBS is an immune-related disorder, in which the immune response generates antibodies that cross-react with gangliosides (i.e., GM1, GD1a, GT1b and GQ1b) at nerve membranes (75).  This autoimmune response results in nerve damage or functional blockade of nerve conduction (76). Aberrant active immunization induced by artificial vaccines, hypothetically, is able to stimulate the immune system to produce specific antibodies, which contribute to cross reaction with epitopes on myelin or axons, leading to nerve damage (77). Vaccines might, as understood, damage the peripheral nerves directly (78). However, the causal associations between vaccines and GBS have not been substantially proved, i.e., the association might not be causally established. The mass immunization against COVID-19 has started unprecedentedly on a global scale. Recently, coincident GBS case was observed after administrated with COVID-19 vaccine (19). New considerations about vaccine safety will undoubtedly arise. Toward the public, it is critical to distinguish events that are temporally associated with vaccination from those directly caused by vaccines. Misinterpretation of GBS incidence that is only temporally coincident with but not caused by vaccination will not only obstruct the success of mass vaccination, but also hinder the development of newer vaccines (9).
During the 1976-77 A/H1N1 influenza immunization campaign, an increase of GBS was reported after vaccine administration (15), which suspended the immunization program temporarily, and initiated vaccine safety concerns. In the 1993-1994 influenza seasons, public concern of vaccine-related GBS arose again due to the increment of GBS (79). The 2009 H1N1 influenza pandemic motivated H1N1 vaccine campaigns in North America and Europe, where post-vaccination GBS concern was raised consequently (29). However, in 2009-2010, a surveillance of H1N1 influenza vaccine in 45 million persons showed a lower excess risk for GBS during the immunization campaign compared to earlier vaccination (73). In France, a study did not support the causation between GBS and H1N1 vaccination (80). Our pooled results show that the temporal coincidence of GBS in influenza vaccinees is not higher than that among general populations unvaccinated.
With regard to HPV, the debate on vaccine safety still exists, which remains one of the barriers to achievement of intensive global vaccination coverage. The Vaccine Adverse Event Reporting System (VAERS) in the United States reported a GBS rate of 0.2 per 100,000 dosages coincided with HPV vaccination from 2006 to 2008 (24). In a school-based HPV study in Canada, the overall background rate was 0.73/100,000 person-year for adolescents aged 7-17 years (81). Similarly, our study did not observe an increase in background rate of GBS after HPV vaccine administration.
The adjuvants of vaccines could affect the magnitude and quality of immune response. The AS03 adjuvant contains atocopherol, which might promote immune system activation in the nonregional lymph nodes (82), whereas MF59 might modulate cellular immune response at the injection site or regional lymph nodes (83). Moreover, an in vitro study demonstrated that atocopherol can raise the expression of hypocretin, leading to antigen presentation via human leukocyte antigens (84), which results in an autoimmune response, and damages hypocretinproducing neurons. In this current study, both vaccines adjuvanted with AS03 and those with MF59 have lower background rates of GBS than that of general populations, even though the background rate among individuals received vaccines with AS03 adjuvant is higher than that with MF59 adjuvant.
Our study has potential limitations that usually exist in observational studies and systematic reviews. Firstly, we aimed to summary data from studies that reported background rate of GBS in mass immunizations, which reflected the temporally coincidence of GBS in "real world". In accordance with the predefined protocol, we did not include clinical trials that   In conclusion, our findings evidenced a mild increase in coincidental GBS during virus vaccination. We presented a reference for evaluation of the coincidental occurrence of GBS in mass vaccination campaigns, including SARS-CoV-2 vaccine.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
HH, YXW, and WW designed this study. FW, DW, YJW, and CL contributed to literature search, review, data extraction. YLZ, ZG, PL, and YCZ conducted statistical analyses. FW, DW, YJW, and CL wrote the manuscript. HH, YXW, and WW contributed to manuscript revision. All authors have reviewed and approved the final version of this manuscript.