Vive la Résistance: T-cell Immunity in the Protection against SARS-CoV-2 Infection

As the SARS-CoV-2 pandemic reaches 2 years since its emergence, the role of host immune responses in both facilitating and protecting against virus-associated morbidity and mortality remains both prominent and complex. Soon after widespread infections were first reported, an excessive inflammatory response was found to be a key driver of toxicity and death in infected patients (1), and suppression of inflammation by dexamethasone remains the most important therapeutic strategy for actively infected patients (2). Yet the development of mRNA-based vaccines against SARS-CoV-2 and their continued efficacy despite the emergence of viral variants that potently evade antibody neutralization underscore the importance of a polyfunctional immune response in preventing COVID-19–associated morbidity and mortality. Finally, recent studies have found that persistent inflammatory responses may be associated with long-term sequelae of COVID-19 infection, otherwise known as “long COVID” (3).

There has, therefore, been widespread interest in identifying host immune factors associated with disease susceptibility and/or severity. Several such factors have been identified (Fig. 1). These include genetic variants of proteins required for SARS-CoV-2 entry, such as the ACE2 receptor and the protease TMPRSS2, as well as germline variants in genes encoding the production of innate immune cytokines such as type I interferons (4, 5). A role for the adaptive immune system in protecting against COVID-19 infection and severity is supported primarily by the efficacy of therapeutic vaccination, but also by recent studies of patients with hematologic cancer, who exhibit defective adaptive immune responses and consequently suffer from increased COVID-19–associated mortality (6, 7). However, few if any of these studies were able to distinguish factors associated with susceptibility to infection from those associated with disease severity in infected patients. Given the continued emergence of novel variants that exhibit a heightened capacity for immune evasion, understanding factors that reduce the likelihood of infection in the setting of viral exposure could help refine both preventive and therapeutic strategies. In this issue, Fahrner and colleagues tackle this question by analyzing immune factors associated with protection in individuals exposed to SARS-CoV-2, as well as host and viral factors associated with response to therapeutic vaccination (8).

The authors began by broadly characterizing the secretory profile of peripheral blood leukocytes (PBL) in acutely infected and convalescent individuals. To this end, the authors developed an assay that allows the measurement of 12 different cytokines secreted by the PBLs upon antigen presentation by autologous dendritic cells. Compared with the non–SARS-CoV-2–exposed controls, acutely infected and convalescent COVID-19 patients produced significantly higher levels of IL2, IFNγ, and IL5 in a SARS-CoV2–specific manner. These results are consistent with a combination of Th1/Tc1- and Th2/Tc2-type cellular immunity to SARS-CoV-2 infection. Notably, the authors demonstrated that there was no difference in peptide-specific cytokine secretion between cancer-free individuals and patients with cancer who successfully recovered from acute infection, suggesting that cellular immunity to COVID-19 develops in patients both with and without active malignancy.

Given the development of both Th1 and Th2 peptide-specific responses following SARS-CoV-2 infection, the authors next asked whether a predisposition toward either T-cell state might influence susceptibility to infection. To investigate this question, the authors collected research samples on a cohort of patients with cancer who had documented evidence of SARS-CoV-2 negativity and were subsequently followed longitudinally to determine whether they ultimately developed COVID-19 infection. Of the 214 patients with cancer enrolled in this prospective clinical trial, 19 patients went on to contract COVID-19 during subsequent waves of the pandemic and were therefore considered susceptible. Of the remainder, 42 of the patients were determined to have been exposed to the virus but remained disease-free and were thus considered resistant. The susceptibility status of the remainder of the enrolled patients could not be established (i.e., “unknown”). Notably, though the polyfunctionality of T-cell responses did not significantly differ between the two groups, susceptible individuals produced peptide-specific IL2 with lower frequency and to a lesser extent. Indeed, when comparing the ratio of IL2 and IL5 levels in the 61 patients with cancer who were exposed to SARS-CoV-2 during the second wave of the pandemic, most susceptible individuals exhibited a T-cell response with an IL2 to IL5 ratio of <1, with the two individuals who developed severe COVID-19 by World Health Organization criteria displaying a ratio of <0.01. Next, the authors asked whether cellular responses to specific regions of the SARS-CoV-2 proteome were associated with susceptibility or resistance to COVID-19 infection. By analyzing peptide-specific IFNγ production during a 7-day in vitro culture assay, the authors found that peptide-specific responses to the Spike 1 protein receptor binding domain (S1-RBD) region best correlated with resistance to infection in SARS-CoV-2–exposed individuals. This suggests that preexisting, Th1-polarized cellular immunity to the S1-RBD region is associated with resistance to COVID-19 infection.

Finally, Fahrner and colleagues sought to determine whether emergent mutations in the SARS-CoV-2 viral genome might evade cellular immune responses. Using a whole-blood IFNγ release assay following 22 hours of incubation with a panel of overlapping peptides from the wild-type RBD region of the SARS-CoV-2 proteome or a mutant pool comprised of documented variants in the RBD region, the authors found that both two doses of mRNA vaccination or a single dose of vaccination in previously infected individuals induced robust peptide-specific IFNγ responses to the wild-type, but not mutant, receptor-binding domain. They also found that mutations in the RBD region occurred at a frequency that was out of proportion to other regions of the viral genome, suggesting that RBD mutants that avoid cellular immune responses might be evolutionarily selected. Intriguingly, the authors observed no correlation between peptide-specific IFNγ responses and the same individual's serologic IgG titer against S1-RBD. This suggests that, on an individual level, humoral and cellular responses are not necessarily correlated and underscores the importance of tracking both types of immune responses independently in vaccinated individuals.

The work from Fahrner and colleagues reveals several exciting areas for future exploration. First, their observation that the Th1/Th2 balance of cellular response to SARS-CoV-2 raises several questions. Identifying either host or infectious features that promote Th1 versus Th2 differentiation may be of particular value for future vaccine design. It is indeed possible that altering either peptide or adjuvant selection to facilitate a Th1 response may be advantageous; however, characterizing the respective roles that Th1- and Th2-type responses play in mediating acute viral control, viral eradication, and prevention of off-target toxicity will be important to understand before attempting to therapeutically shift cellular immune responses toward Th1 or Th2 dominance.

Second, the authors’ finding that cellular and humoral immune responses to vaccination were poorly correlated is notable. Future vaccination studies may benefit from longitudinal assessment of both cellular and humoral immune responses, both to determine which arm of the adaptive immune system is most critical for eliciting a protective response and to understand how CD4⁺ T-cell responses to vaccination might facilitate antibody diversification, as has been previously reported (9).

Finally, why patients with cancer mounted a poor T-cell response to vaccination remains unclear. In the present study, patients with hematologic malignancies appeared to have not only a blunted peptide-specific response but also a poor response to polyclonal stimulation. Given that independent reports have suggested a preserved cellular response to vaccination in patients lacking B-cell function (10), understanding the predictors of poor vaccination response in this high-risk cohort will be of immense value, as patients who are unlikely to develop either humoral or cellular responses to vaccination may be appropriate candidates in whom to investigate either passive immunization or prophylactic antiviral therapy as a means of protection.