Introduction
The coronavirus pandemic has spread like wildfire across the globe. Some countries
are more affected than others while children seem to have a milder form of the disease
as compared to adults.[1] We are still grappling to understand the reason behind these disparities. There
appears to be a correlation between Bacillus Calmette–Guérin (BCG) vaccination and
protection against COVID-19.[2],[3] Some experts hypothesize that countries offering BCG vaccine in the national immunization
program at birth seem to have a lower death rate from COVID-19.[3] Whether this can be attributed to the BCG vaccine or a less virulent virus strain
or social distancing measures or because countries like India are just entering phase
III of the epidemic will only be known retrospectively. If BCG is indeed the reason
for this protective effect, can it be attributed to persistence of “heterologous”
or “off-target” effects of the BCG vaccine, will unfold over the next few months.
Banking on the “heterologous effect” of BCG, two clinical trials NCT04327206 and NCT04328441
propose to administer prophylactic BCG vaccine to health-care providers in Australia
and the Netherlands respectively. At this point in time, there are 3 trials in the
Clinical Trials Registry of India using different BCG strains to harness this heterologous
effect against COVID-19. The research hypothesis is activation of the innate immune
system to protect against progression of the Coronavirus infection.[4] In this article, we try to understand the underlying mechanisms of immune protection,
“trained immunity” conferred by the BCG vaccine and review some epidemiological evidence
that support or refute the hypothesis.
Background
The BCG vaccine has been in use since 1921 and about 160 million newborn children
receive BCG vaccine annually (2017 data).[5],[6] The efficacy of BCG vaccine against pulmonary tuberculosis (TB) is highly variable
ranging from 0% to 80%, but it is highly efficacious against tuberculous meningitis
and miliary TB in children under 5 years of age.[6] BCG offers protection against leprosy and other mycobacteria due to cross-reactive
antigens. This effect is called “Cross-Protective effect.”[7] In addition, BCG vaccination has shown to reduce overall pediatric mortality, far
more than its target effect on pulmonary TB. This is also true for measles, smallpox,
and oral polio vaccine. In addition, concurrent or previous administration of BCG
is associated with significantly higher levels of antibodies against hepatitis B,
polio, and influenza vaccines.[8] Thus BCG vaccine acts as an adjuvant to concurrent or subsequent immunization through
its immunostimulatory effect on the innate immune system. Clinically, BCG reduced
all-cause mortality in children in several observational studies and randomized trial
settings and decreased hospitalization rates due to respiratory infections and sepsis.[5],[9],[10],[11] These effects of BCG are known as “off-target effect” or “heterologous effect.”[7] The heterologous effects of BCG are mediated through the cells of the innate immune
system including monocytes/macrophages and natural killer (NK) cells and are independent
of memory T or B cells. This phenomenon is associated with a memory-like response
of the innate immune cells and is known as “trained immunity.”[12]
Immunology of training the immune system
Understanding the underlying mechanism of BCG induced immune effects is still a work
in progress. Both the innate and the adaptive immune systems mount a response against
the bacillus in the BCG vaccine. The effects being both, specific on-target effects,
mediated mostly by the T and B cells (adaptive immunity), and off-target effects mediated
by actors of the innate immune system mainly monocyte/macrophage system and NK cells.
Until recently, it was thought that the innate immune system is the first line of
defense and does not have memory. This paradigm has changed recently. Here, we try
to understand at least some of the mechanisms of how the bacillus in the vaccine activates
the immune system and sustains the effect for a significant period of time called
“trained immunity.”[12]
Kleinnijenhuis et al. injected healthy volunteers, between 20 and 36 years of age with BCG vaccine.In
vitro cytokine assays were performed on peripheral blood mononuclear cells after nonspecific
stimulation on day 14 and 3 months later. There was an enhanced Th1 cytokine response
(Interleukin [IL] 1-β, Interferon-gamma [IFN-γ] and Tumor Necrosis Factor-α [TNF-α])
to nonspecific stimuli such as Staphylococcus aureus and Candida albicans that were significantly enhanced at day 14 and lasted at least 3 months. This was
attributed to an increased number and altered functional state of monocytes. The increased
cytokine response correlated with increased mRNA expression of these cytokines. This
was brought about by long-term transcriptional regulation though histone modification.
It was shown that the lysine 4 residue on histone H3 (H3K4) undergoes trimethylation.
H3K4 trimethylation at specific loci leads to increased transcription of proinflammatory
cytokine genes.[13]
Pattern recognition forms the mainstay of innate immune system. Innate immunity to
microbial pathogens relies on the specific detection of pathogen-associated molecular
patterns by specific host receptors. Muramyl dipeptide (MDP) is a peptidoglycan, common
to cell wall of Gram-positive and Gram-negative bacteria. NOD-2 is an intracellular
pattern recognition receptor expressed, specifically in monocytes and macrophages.
The heterologous immune response to BCG seems to be mediated by the interaction of
NOD2 with MDP in the BCG cell wall through activation of the Nuclear factor kappa
B (NFκB) pathway.[13]
Furthermore, Arts et al. demonstrated that the upregulation of Th1 cytokines after BCG vaccination is dependent
on changes in cellular metabolism. They noted an increase in glycolysis, upregulation
of glutamine metabolism, and oxidative phosphorylation after stimulation with BCG
at day 7 in healthy volunteers. These shifts in metabolism are dependent on the mammalian
target of rapamycin (mTJR) pathway and histone methylation and persisted even at 3
months post vaccination. These metabolic and epigenetic modifications are intertwined
and necessary for development of trained immunity.[14]
In the same cohort of healthy volunteers studied by Kleinnijenhuis et al.,[13] administration of BCG vaccine led to increased production of proinflammatory cytokines
upon stimulation with heat inactivated C. albicans at 3 months after vaccination.
It was also seen that in severe combined immunodeficiency mice with deficient T and
B lymphocytes but preserved NK cell lineage, BCG vaccination-induced protection against
disseminated Candida Albicans infection, establishing the protective effect of NK
cells.[15] This NK cell-mediated cytotoxic effect is exploited effectively in the treatment
of bladder cancers. Intravesicular instillation of BCG is an effective treatment option
in superficial bladder cancers.[16]
The non-specific effects of the BCG vaccine are also effective against viruses. In
another study (NCT02114255) by the same group at the Nijmegen Medical Center, healthy
volunteers were randomized to receive BCG vaccine (n = 20) or placebo (n = 20) 14 days prior to trivalent influenza vaccine. In the BCG-vaccinated group,
hemagglutination-inhibiting antibody responses against 2009 pandemic H1N1 vaccine
strain were significantly enhanced with a trend toward rapid seroconversion defined
as >4-fold increase in titers as compared to baseline.[17] In an experiment by Floc'h et al., mice were inoculated with Pasteur Institute BCG and challenged with different viruses
introduced by various routes 15–31 days later. BCG-inoculated mice exhibited a significantly
higher resistance to these lethal infections than control mice (overall survival 18%
vs. 41% in BCG inoculated mice).[18] Thus, BCG also acts as an adjuvant to potentiate responses to a viral challenge,
possibly viral infections like coronavirus.
The off-target or heterologous effects of the BCG vaccine are thus mediated by the
monocyte/macrophages and NK cells of the innate immune system. Now, we understand
that these effects can last longer through epigenetic changes that go hand-in-hand
with metabolic shifts in these cells.[13],[14] However, how long can these changes persist, what is their clinical relevance and
how can they be exploited in our fight against a novel virus? Is there any evidence?
Epidemiological evidence for trained immunity
In 1927, Carl Näslund was first to note that BCG vaccination at birth had a mortality
almost threefold lower than the very high rate of 10% among unvaccinated children.[19] A systematic analysis of several observational studies and clinical trials by the
WHO (Higgins et al.) has shown reduction in overall mortality by more than expected through their on-target
effects. The strongest effect was seen in the first 28 days and in premature infant.[9] BCG contributed to an overall 30% reduction in all-cause neonatal mortality (95%
confidence interval [CI]: 1%–51%). In the two randomized control studies with minimum
bias, the reduction in neonatal mortality was about 48% (95% CI: 18%–67%).[20],[21] In contrast. a study done by Jayaraman et al. that reported results of two trials in India did not show any protective effect
on neonatal mortality. This dichotomy could also be related to nutritional status
of the mother and child, immunization status of the mother, host genetics, and access
to health care. This could also be related to the genetic variation in BCG strain,
although there is no such direct evidence to prove so.[22] Where BCG is protective, the duration of protection is variable with an average
of about 10 years. Approximately 50% protection has been observed 15–20 years after
vaccination in Brazil and 40–50 years after vaccination in American Indians and Alaskan
natives. It is hypothesized that BCG vaccination has higher efficacy in more temperate
areas, probably related to lower exposure to nontuberculous mycobacteria (NTM).[23] Furthermore, the effects of environmental NTM can precede BCG vaccination and block
multiplication of the vaccine strain.[24] Therefore, to suggest that BCG at birth might offer protective immunity against
SARS-CoV-2 is preposterous to assume.
Discussion
Considering that BCG has heterologous effects that can last for a longer duration
of time sets a persuasive hypothesis that the vaccine might offer some protection
from severe infection from SARS-CoV-2. There are epidemiological data to suggest that
BCG offered protection from acute respiratory illnesses in addition to bacterial infections.[11] However, given that the efficacy is highly variable from 0% to 80% and wanes over
time, it might be premature to say that the low death rate due to coronavirus pandemic
in certain countries is related to the protective effect of BCG at birth.
The NCT04327206 clinical trial is a good idea as it is trying to activate the innate
immune system in preparation to face the virus. Does re-vaccination work? Again, the
jury is still out there. Retrospective observation in Hungary and Poland, in 1950s–60s,
showed reduced incidence of TB after re-vaccination. In Chile, where re-vaccination
was given at 6 years and 14 years; there was no difference in the proportion of young
adults with 1, 2, or 3 BCG scars.[25] In a randomized control trial in South Africa, where incidence of TB was high, re-vaccination
seemed to offer sustained reconversion of IFN-γ release after BCG re-vaccination.
This effect was measured at day 70 and up to 2 years after re-vaccination.[26] The off-target effects and clinical impact on non-TB infections were not evaluated
in these trials.[27] In a similar prospective randomized trial in South India, IFN-γ and IL-2 and TB
antigen-specific CD4 and CD8 T cell responses were boosted at 4 and 34 weeks after
re-vaccination. In addition, innate IFN-γ, NK cell, γδ T cell, and NKT cell responses
were higher in re-vaccinated persons as compared to placebo controls. The TB antigen-specific
innate immunity peaked transiently at week 4, while BCG specific peak was seen later.[28] One of the contentions of using BCG re-vaccination is the duration of onset of nonspecific
immunity is longer. If we look at antigen-specific immunity as in the Indian trial,
the peak was seen at 4 weeks. In the epigenetic modulation studies by the Dutch group,
non-specific immune responses were seen at day 14 post-vaccination.[13] The trained immune responses lasted for about 3 months to 1 year in different trial
settings.[26],[28]
In the Guinea Bissau study for low birth weight infants, a protection from infections
was seen as early as first 3 days of life after BCG vaccination at birth. Therefore,
a 14-day window might be sufficient to rev up the innate immune system as the front
line of defense against the virus.
Given the heterogeneity of the data on BCG in India, well-designed clinical trials
in the Indian setting will target two birds in the same stone, the virus and the mycobacteria.
