The Impact of the Bacillus Calmette-Guerin Vaccination on COVID-19: A Systematic Review and Meta-Analysis

Background: The Bacillus Calmette-Guerin (BCG) vaccine has been known to confer nonspecific benefits beyond its target pathogen, including immunoprotective effects against other respiratory infections and cancer. Recent evidence has shown that BCG vaccination could explain lower-than-expected coronavirus disease 2019 (COVID-19) incidence and mortality rates. The objective of this study is to compare the mortality and incidence for COVID-19 in both BCG-vaccinated and nonvaccinated populations.

Methods: A systematic review and meta-analysis were performed to identify relevant studies since the start of the COVID-19 pandemic according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The strictly standardized mean difference was calculated to measure the effect size of BCG vaccination on COVID-19 incidence and mortality rates.

Results: A total of 49 articles were reviewed, and eight eligible comparative studies were used in the meta-analysis. The results showed an overall protective effect of BCG vaccination against COVID-19 mortality and incidence rates. The strictly standardized mean difference for mortality rate was 348.76. An I² value of 91.57 was found for mortality rates with BCG vaccination and of 65.56 for mortality rates without BCG vaccination. Five studies were analyzed regarding COVID-19 incidence rates. The strictly standardized mean difference for incidence rates was 1,273,584.117. An I² value of 2.84 was found for incidence rates with BCG vaccination and of 78.23 for incidence rates without BCG vaccination.

Conclusions: Our study shows that populations with BCG vaccination programs have lower COVID-19 incidence and mortality rates and highlights the need for future studies to investigate if other variables may be behind the observed protective effect of BCG vaccination on COVID-19.

Keywords: Bacille Calmette-Guerin; COVID-19; SARS-CoV-2

Introduction

The coronavirus disease 2019 (COVID-19) pandemic has spread rapidly across the globe since the first reported cases in Wuhan, China in December 2019 [1]. COVID-19 is caused by a new viral pathogen named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that renders most infected individuals asymptomatic or developing a mild-to-moderate disease characterized by upper respiratory tract symptoms [2]. The spectrum of symptomatic infection ranges from mild to critical, with severe disease observed in 14% of cases and critical illness observed in 5% of cases with an overall case fatality rate (CFR) of about 2.3% [3]. Significant variation in prevalence and mortality from COVID-19 among countries exists [4]. This may be the result of several different factors, including the extent of containment measures (e.g., social distancing, lockdowns), availability of critical resources in healthcare systems, and the existence of a Bacillus Calmette-Guerin (BCG) vaccination program [5, 6].

Transmitted through respiratory droplets, tuberculosis (TB) and COVID-19 are deadly pulmonary infections [5]. Recent evidence shows that vaccination programs against TB could explain a lower-than-expected COVID-19 incidence and mortality [7, 8]. Apart from North American and Western European countries, national BCG vaccination programs are available in most countries. Furthermore, recorded deaths from COVID-19 in TB endemic countries with BCG vaccination programs remains low at fewer than six deaths per 1 million cases [9]. These lower incidences and mortality from COVID-19 in TB-endemic areas have been attributed to the protective role of BCG vaccinations [5, 7, 8, 10].

The BCG vaccine is a live, attenuated strain of Mycobacterium bovis developed in the early 20th century for protection against TB, although its protection is only effective for non-pulmonary forms of TB in children [11, 12]. It has been reported that the BCG vaccine could also protect against other respiratory pathogens such as influenza A, Streptococcus pneumoniae, and respiratory syncytial virus (RSV) [13-15]. Immunotherapeutic effects of the BCG vaccine have granted its use against some forms of bladder cancer as well [13]. While the exact pathway of its heterologous effects is still unclear, BCG can induce innate immunity via monocytes, natural killer (NK) cells, and other immune cells, suggesting activation and functional specialization of immune cells against non-tuberculous pathogens [13, 16].

The present study aims to investigate the COVID-19 incidence and CFR in countries with and without BCG vaccination programs by systematically reviewing the current literature and performing a meta-analysis of the available data.

Materials and Methods

Literature search

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used for this study. An electronic search was conducted in the following databases: MEDLINE, EMBASE, PubMed, and EBSCO libraries. The majority of all articles were published within the last 2 years. Language restriction was applied to only studies published in English.

The inclusion criteria were articles that provided information on countries and/or individuals with an existing protocol requiring BCG vaccination. Other studies included in this search provided information on the number of COVID-19 cases and deaths related to COVID-19 in countries and/or individuals with and without positive BCG vaccination status. These inclusion criteria allowed further examination of the mortality and incidence rates due to COVID-19 in the BCG-vaccinated and nonvaccinated population. Exclusion criteria included an absence of case numbers and/or death rates due to COVID-19, BCG vaccination rates, and studies published beyond our 5-year search criteria.

Based on these criteria, our first search aimed to obtain mortality rates due to COVID-19 on BCG-vaccinated and nonvaccinated individuals. The terms used were “Bacille Calmette-Guerin” or “BCG”, “COVID-19” or “SARS-CoV-2” or “coronavirus”, “morbidity”, and “mortality”. The second search included studies that examined the incidence rates of COVID-19, and the term “incidence” was added. Informative, observational, and descriptive studies were excluded from our meta-analysis due to an inability to extract data.

Data retrieval

Data were extracted from the articles after an independent review by three of the co-authors, allowing for correct identification of data that may have been omitted by a single reviewer and cross-checked information presented across all studies.

The following variables were obtained: total population (sample size), number of cases, number of deaths, total tested positive COVID-19 cases, mortality, and incidence rates. To calculate the COVID-19 incidence and CFR for two studies [5, 17], the number of COVID-19 cases were extrapolated along with the population data for the counties [18].

Statistical analysis

We calculated mortality and incidence rates for each study by dividing the number of events (either the number of deaths for mortality or the number of new cases for incidence) by the sample size presented in the article. Measures of dispersion, such as standard error and variance were also calculated for each study. The random effects model was chosen, as we assumed the variability is not due to sampling error but rather by variations in the population of effects. Due to this assumption, we adjusted the weight of each study with a constant (v). We used this constant to calculate the new weight (w_v), weighted effect size (w_v*es) and w_v*(es²) and w_v². We computed individual study weights (w) and each weighted effect size (w*es), as well as w*(es²) and w², needed to calculate the Q statistic. The Q test was used to measure heterogeneity amongst the studies, and the I² was calculated as a method to quantify heterogeneity. The latter is expressed as a percentage of the total variability in a set of effect sizes due to true heterogeneity (between-studies variability) [19].

Forest plots were created for each set of data analyzed, i.e., COVID-19 mortality and incidence with and without BCG. The strictly standardized mean difference (SSMD) was also calculated and included in our study to compare the independent groups.

Institutional Review Board approval is not applicable. The study was conducted in compliance with the ethical standards of the responsible institution on human subjects as well as with the Helsinki Declaration.

Results

A total of 49 studies were initially retrieved after the search. Two studies were removed as those were duplicates. Of the 47 remaining studies, 21 were excluded as those were non-relevant to the inclusion criteria and were review or editorial comment articles. From the remaining 26 studies, 18 studies were removed due to unclear data on death rates as well as confirmed COVID-19 cases. Eight articles fit the criteria for the first objective of our meta-analysis to examine the mortality rates of COVID-19 in BCG-vaccinated and nonvaccinated individuals (Supplementary Material 1, www.ciijournal.org). Five of the eight articles provided data for the second objective to examine the incidence rates of COVID-19 in BCG-vaccinated and nonvaccinated individuals (Suppl. Material 2, www.ciijournal.org).

Data from two studies [7, 8] directly presented mortality rate values, whereas the other articles required a calculation for mortality rate based on the tables presented in each study. One study [20] presented data from 24 countries, 12 countries with BCG vaccination policies and 12 without. Jirjees et al [5] compared data from 35 countries with BCG vaccination vs. 15 countries without vaccination, while Islam et al [21] presented data from 15 countries with BCG vaccination and 15 without BCG vaccination. The largest group comparison [17] presented data from 167 BCG-vaccinated countries and 64 nonvaccinated countries. Madan et al [7] reviewed COVID-19 incidence from 174 countries divided into low and high BCG coverage. One case-control study [22] compared 85 BCG-vaccinated vs. 101 nonvaccinated hospitalized patients with confirmed COVID-19.

Mortality with and without BCG vaccination

All eight of the studies in the systematic review were included for analysis. Tables 1 and 2 [5, 7, 8, 17, 20-23] show data regarding mortality rates with BCG and without BCG vaccination, respectively. Statistical analysis for heterogeneity showed an I² value was 91.57 for mortality rates with BCG vaccination and 65.56 for mortality rates without BCG vaccination. As seen in Tables 1 and 2 [5, 7, 8, 17, 20-23], there was a wide range in the sample size data across the studies, which could explain the high I² value across the populations analyzed in this study. The variations were assumed to be in the population characteristics, i.e., within each study itself rather than the combination of studies analyzed. A random effects model, which allows for making inferences on the population based on the assumption of normal distribution, was used for the meta-analysis. The weighted effect size, as presented in for mortality rates due to COVID-19 in the BCG-vaccinated group was 2.10 (1.69 - 2.51, 95% confidence interval (CI)) (Fig. 1), whereas the weighted effect size for mortality rates in the BCG nonvaccinated group was 5.80 (95% CI: 3.53 - 8.07) (Fig. 2). Overall, the mortality rates for mortality with BCG vaccination were lower compared to nonvaccinated populations (Figs. 1, 2). The SMD difference between the two groups is 3.70. We calculated the SSMD for mortality rates between the two groups, with and without the BCG vaccination, to be 3.22.

Table 1. Mortality With BCG Vaccination

Table 2. Mortality Without BCG Vaccination

Figure 1. Mortality with BCG vaccination. BCG: Bacillus Calmette-Guerin; CI: confidence interval.

Figure 2. Mortality without BCG vaccination. BCG: Bacillus Calmette-Guerin; CI: confidence interval.

Incidence of COVID-19 with and without BCG vaccination

Five of the eight studies in this systematic review provided data for analysis regarding the incidence of COVID-19. Tables 3 and 4 [5, 7, 8, 17, 20] show data extracted for incidence rates of COVID-19 in the BCG-vaccinated and BCG nonvaccinated countries, respectively. Statistical analysis for heterogeneity, gave an I² value was 2.84 for incidence rates with BCG vaccination and 78.23 for incidence rates without BCG vaccination. The incidence rates were also lower in the BCG-vaccinated group (Fig. 3) compared to the nonvaccinated (Fig. 4). The scale, as seen in Figure 3, is in the thousandths on the forest plot whereas the scale is in the hundredths for Figure 4. The weighted effect size for incidence rates due to COVID-19 in the BCG-vaccinated group was 0.023 (95% CI: 0.00673 - 0.0395) (Fig. 3), whereas the weighted effect size for incidence rates in the BCG non-vaccinated group was 0.233 (95% CI: 0.0196 - 0.446) (Fig. 4). The SSMD for the incidence rates of COVID-19 in the BCG-vaccinated and nonvaccinated groups was 49.56.

Table 3. Incidence of COVID-19 With BCG Vaccination

Table 4. Incidence of COVID-19 Without BCG

Figure 3. Incidence of COVID-19 with BCG vaccination. COVID-19: coronavirus disease 2019; BCG: Bacillus Calmette-Guerin; CI: confidence interval.

Figure 4. Incidence of COVID-19 without BCG vaccination. COVID-19: coronavirus disease 2019; BCG: Bacillus Calmette-Guerin; CI: confidence interval.

Discussion

The present meta-analysis supports the hypothesis that the BCG vaccine provides a protective effect against the incidence and mortality from COVID-19. Our results illustrate an increase in the incidence rate and in the mortality rate due to COVID-19 in the BCG nonvaccinated populations.

Although recent epidemiological studies suggest that countries with extensive BCG vaccination programs and coverage have lower-than-expected morbidity and mortality from COVID-19, most of them are retrospective in nature and have multiple limitations, including selection bias (Supplementary Material 3, www.ciijournal.org). Mortality data showed a high heterogeneity amongst studies. Given the endemicity of TB in most of the countries with a BCG immunization program, it is probable that the studies in our meta-analysis included patients in the unvaccinated group with a history of previously active or latent TB. Another limitation of our study was using the number of reported deaths divided by the number of confirmed cases to calculate CFR, which may not represent the most accurate measure of CFR during an ongoing global pandemic. Furthermore, the CFR depends on multiple factors such as the ability of governmental authorities to report the number of deaths and confirmed COVID-19 cases accurately, as well as the ability of national healthcare systems to respond to ongoing infections. Most of the countries with BCG vaccination programs have lower economic income compared to those without BCG vaccination programs. It is possible that this inequality may account for less COVID-19 case reports, due to lower number of tests done, lower exposure due to possible low travel rate during the pandemic period and other economic factors. Since developing countries have a lower life expectancy, an older population with probably higher rate of age-related comorbidities may have contributed to higher mortality rates or symptomatic disease in developed countries without BCG vaccination programs. These additional considerations may also challenge the validity of our findings.

The hypothesis that BCG-induced trained immunity could protect COVID-19 needs to be tested in rigorous randomized clinical trials. Currently, at least 20 clinical trials in various countries are ongoing to investigate the extent of protection from COVID-19 conferred by the BCG vaccine [24]. Nevertheless, the epidemiologic evidence of immunologic protection offers empirical support for the potential protective effect of BCG vaccination to reduce the number of severe COVID-19 cases.

A protective role of BCG against respiratory viral infections such as influenza and RSV, as well as its utility as first-line therapy for non-muscle invasive bladder cancer patients, suggest immunotherapeutic effects beyond its targeted pathogens [12]. The exact immunologic pathway of this broad immunity remains elusive, but it has been suggested that a “trained immunity” process, elicited in innate immune cells, along with tolerogenicity [25], accompanied by transcriptional, epigenetic, and metabolic reprogramming of the myeloid cells in healthy vaccinated individuals [14, 26] may be involved.

An immune dysregulation appears to underlie the mechanisms involved in severe COVID-19 leading to cytokines storm and lymphopenia [27]. Though the exact mechanism whereby BCG may induce protection against COVID-19 is not well established, significant sequence homology has been identified between BCG-derived and SARS-CoV-2 NSP3- or NSP13-derived peptides [28]. In recipients of the BCG vaccine, CD4⁺ and CD8⁺ T-cells primed with BCG-derived peptides develop an enhanced immune response to corresponding homologous SARS-CoV-2 peptides shown in vitro and in silico [28], suggesting an essential role in viral clearance [12, 28]. A nonspecific protective immunogenic response from the BCG vaccine against SARS-CoV-2 has also been suggested. Upregulation in interleukin (IL)-17, tumor necrosis factor (TNF), NOD-like receptors, and nuclear factor-κB signaling pathways in SARS-CoV-2 infections correlated inversely with the downregulation of pathways after BCG vaccination [12].

Further studies are needed first to confirm the results of this study, if the relationship between BCG vaccination and COVID-19 is not due to the heterogeneity of the populations and other epidemiological factors.

Conclusions

The BCG vaccine has long been known to confer nonspecific benefits beyond its target pathogen, including immunoprotective effects against other respiratory infections and cancers. Our study shows that populations with BCG vaccination programs have lower COVID-19 incidence and mortality rates, and highlights the need for future studies to investigate if other variables may be behind the observed protective effect of BCG vaccination on COVID-19. In addition, randomized clinical trials on the effectiveness of BCG vaccine to protect from severe COVID-19 are needed.

Supplementary Material

Suppl 1. Eight studies of incidence and mortality in BCG-vaccinated and nonvaccinated individuals.

Suppl 2. Flow chart of the article selection process to analyze the association between mortality and incidence rates due to COVID-19 in both BCG-vaccinated and nonvaccinated populations. PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Suppl 3. Risk of bias assessment.

Acknowledgments

None to declare.

Financial Disclosure

None to declare.

Conflict of Interest

None to declare.

Informed Consent

Not applicable.

Author Contributions

IK and SB performed literature search. IK and JC performed analysis. IK, AK, KA, and JC wrote the article.

Data Availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

References