While the importance of inducing neutralizing antibodies by natural infection or vaccination to protect against (re)infection is becoming well established, the role of T cell immunity is much less clear. Nevertheless, there are indications both from experimental animals and observations during natural infection in humans that T cells could be protective as well.
Antibody-mediated protection by neutralizing antibodies (and potentially other mechanisms) is a straightforward process that can relatively easily be measured in vitro, although, as repeatedly mentioned, standardization is still lacking, which makes comparisons between different studies difficult.
On the contrary, T cell immunity is a very complex process, involving distinct T cell subsets, such as CD4+ and CD8+, follicular helper cells, regulatory T cells as well as various functional states, according to differentiation (stem cell like, naïve, effector, memory, exhausted…) and cytokine production profile (Th1, Th2, Th17 etc). Moreover, the recognition of particular peptides by T cells is highly dependent on the genetic HLA constitution. While a very easy and more or less standardized assay (based on IFN-g production by whole blood or mononuclear cells after peptide stimulation) is often used in clinical setting, it only provides a very rough and incomplete picture of T cell function and it can certainly not be directly interpreted as evidence of protective immunity.
Immuno-virology is an empirical science. Just let us consider what has been published recently and try to carefully make sense out of it. Bear with me and fasten your seatbelt: T cell immunology is a “hard nut to crack” for many….
See the added Word document with some nice illustrations.
Ep 177-1: A nice Swedish cohort study in J Int Med of Aug
Mainly focusing on 370 HCW, who got infected in April-June 2020, mostly asymptomatic or with mild-moderate symptoms, were followed up for 8 months.
A much smaller group of 52 hospitalized (hence severe) COVID patients and of 48 seronegative HCW were also followed.
- Anti-Spike antibodies remained positive in over 90 % of HCW and all COVID patients, but in the HCW (with milder disease) levels were lower and tended to slightly decline (Fig 1 A p. 18).
- “T cell memory” was evaluated in a whole blood stimulation with a peptide pool of 47 peptides, covering several SARDS-CoV-2 antigens and 97 % of HLA types, but avoiding overlap with endemic CoV: IFN-gamma responses were found in 63 % of HCW and 88 % of COVID patients. There was some tendency to decline in HCW already after 5 months (Fig 1B p. 18).
- The cumulative incidence of PCR-confirmed SARS-CoV-2 infection was 1% (3/252) among previously infected anti-spike IgG positive HCW, (0.13 cases per 100 weeks at risk) compared to 23% (11/48) among anti-spike IgG negative HCW (2.78 cases per 100 weeks at risk), resulting in a protective effect of 95.2% .
- The presence of anti-spike IgG antibodies is associated with a substantially reduced risk of infection in HCW with asymptomatic to moderate infection.
- No clear-cut conclusions about T cell immunity and about COVID patients
Ep 177-2: A nice Korean longitudinal study in JID of Sept, focusing on 97 COVID patients, subdivided in ASY, non-pneumonic and pneumonic, with matching for age, body mass index and comorbidities between these categories (see Table 1 p. 756).
- Qualitatively sustained, but quantitatively declining spike-antibodies and neutralizing antibodies: levels remained highest in pneumonic patients (Fig 1 p. 758). Also older age and prolonged shedding associated with stronger humoral responses.
- T cell responses were measured once (week 9-20) in IFN-gamma ELISPOT against a commercially available peptide pool of Spike, Membrane and Nucleoptotein: clearly more responses in pneumonic > non-pneumonic > asymptomatic patients and correlation with humoral, immunity.
Conclusion: Both humoral and cellular responses correlate positively with disease severity in COVID patients. (which is not a direct argument for role in protection). However, no indication on protective capacity against reinfection in this paper.
After mRNA vaccination
Ep 177-3: Temporal course of T- and B-cell-responses to vaccination with BNT162b2 (Pfizer) and mRNA-1273 (ModeRNA) in German HCW (CMI Sept)
- T cell responses, measured with “interferon release assay” in whole blood (similar to 177-1) became already positive within 1st week after 1st vaccine (p. 3 and 4)
- Antibodies (binding and neutralizing during 2nd week after 1st dose (p. 3 and 4)
- T and B cell responses strongly boosted after 2nd dose (p. 4)
- Influencing factors:
- Younger individuals respond stronger to 1st dose
- Previously infected subjects much stronger responses
- ModeRNA stronger responses than Pfizer
- No antinuclear antibodies (as sign of auto-immunity) induced
Ep 177-4: Corbett et al in Science (Sept) clearly show that antibody levels, induced by mRNA-1273 (ModeRNA) in macaques are a correlate of protection in an experiment with increasing doses of the vaccine:
- 10-fold increase in S-binding titers was associated with ~10-fold reductions in viral replication in BAL (broncho-alveolar lavage) and NS (nasal swab) after challenge.
- Threshold for protection against replicating virus in BAL = Spike specific IgG >336 IU/ml and in nasal swab S-specific IgG >645 IU/ml. So less IgG needed to protect against lower respiratory disease.
- Passive transfer of vaccine-induced IgG from NHPs to naïve hamsters was sufficient to mediate protection.
- No data on T cell immunity in this paper
Conclusions: Antibodies suffice for protection and lower titers can protect against lower respiratory disease. (while blocking upper respiratory infection/disease requires higher Ab levels)
After Adenoviral vaccination
Ep 177-5: Phase 2/3 study of ChAdOx-1 NCOV19 (Astra-Zeneca) vaccine (Lancet Nov 2020):
- T cell responses (measured with IFN-gamma ELISPOT) peaked at day 14, slightly declined thereafter and was slightly lower in the participants > 70 years (Fig 6 p. 1990)
- Antibody responses (S-specific and neutralizing) peaked after 42 days and remained at plateau at day 56: slightly lower in the older groups after 1 dose, but increased to same level as younger groups after two doses (Fig 4 p. 1989 and Fig 5 p. 1990)
Ep 177-6: A follow-up on 177-5, published in July as medRxiv preprint. An in-depth characterization of the T cell responses in subjects 18-85 years old.
- AZ vaccination induced a robust, polyfunctional Th1-dominated T-cell response, with broad CD4+ and CD8+ T-cell coverage across the SARS-CoV-2 spike protein.
- Remarkably, the responses were well maintained in the older subjects with a similar proportion of T cells, producing 2 or 3 cytokines (IFN-gamma, IL-2 and TNF). See fig 2 p. 44.
Ep 177-7: Kim from Korea in “Immune Network” (Aug 2021) has compared the humoral (S-binding and neutralization) and T cell responses (IFN-g ELISPOT) after 1st dose of Astra-Zeneca and Pfizer in HCW:
- Humoral responses were higher for Pfizer (Fig 1 p. 4)
- T cell responses against overlapping Spike epitopes similar after Pfizer and AZ (Fig 2 p. 5).
Ep 177-8: A mouse study from China, suggesting that prime with Adeno, followed by boost with mRNA induces the highest titers of neutralizing antibodies (Fig 2 B p. 633). It is difficult to extrapolate this study, since the vaccines used are different from the Adeno and mRNA, used for vaccination in the West. Moreover not all combinations were used in all assays (no info on T cell responses after Adeno prime + mRNA boost).
After vaccination with inactivated virus
Ep 177-9: Bueno evalua ted humoral and T cell responses to CoronaVac (inactivated Sinovac vaccine) in Chile
- Seroconversion for anti-S1 was observed at 4 wks after 2nd dose in 86.6 % of 18-59 yrs and 70.3 % of > 60 yrs old. Similar values for various neutralization tests
- Maga Pools of 15-mer peptides derived from the S protein of SARS-CoV-2 (MP-S) and the remaining proteins of this virus (MP-R) were evaluated by ELISPOT in a total of 47 volunteers. As can be seen in Fig 4 p. 41 significant, but weak responses were seen in a proportion of these patients.
T cell responses to variants of concern
Ep 177-10: Zhang (Sci Rep Sept) performed a very refined analysis of CD8 T cell responses from convalescent patients against peptides from various SARS-CoV-2 proteins
- On one hand CD8+ T cell epitopes highly conserved among human coronaviruses are identified.
- On the other the SARS-CoV-2 variants harbor multiple mutations that reduce cellular immune responses
→ Especially the K417N and L452R mutations in Delta spike reduce binding to HLA and hence decrease T cell responses
Ep 177-11: A somewhat less refined approach in CMI (Aug), using whole blood for stimulation with peptide pools and co-expression of IL-2/TNF in CD4 T cells versus IFN-g/TNF in CD8 T cells
- 88 % of previously (months ago) infected subjects showed T cell responses in either CD4 and/or CD8 T cells.
- T cell responses in recently fully Pfizer vaccinated subjects significantly higher
- Much more CD4 than CD8 T cell responses in both infected and vaccinated subjects.
- Positive correlation between CD4 T cell and humoral responses
- Only slightly lower responses to peptide pools from Alpha and Delta variants, as compared to wild type. (This is compatible with 177-10 as in that paper a much more refined peptide-per-peptide approach is used)
Ep 177-12: A very interesting preprint by Goel on mRNA vaccines induction of durable immune T and B memory to SARS-CoV-2 that continues to evolve over time:
- SARS-CoV-2-specific memory B cell responses were robustly induced following mRNA vaccination and continue to increase in frequency for at least 6 months, even as circulating antibody levels declined in the same individuals.
- mRNA vaccination generated highly mutated memory B cells that were capable of cross-binding VOCs, including Alpha, Beta and Delta.
- mRNA vaccine-induced memory B cells also appeared to have a qualitative advantage at binding variants compared to memory B cells generated by mild COVID-19
- Specific memory CD4+ T cells were relatively stable from 3-6 months post mRNA vaccination.
- Early CD4+ T cell responses correlated with 3- and 6-month humoral responses
Ep 177-13: Another recent preprint in bioRxiv is less reassuring
The next generation ModeRNA, designed on the Beta Spike (mRNA 1273.351) was tested in transgenic mice (susceptible to SARS-CoV-2):
- It induced neutralizing antibodies in serum against ancestral SARS-CoV-2 and several variants, although levels were lower particularly against the Delta virus.
- High dose formulation induced protection against lung pathology by all variants
- Low dose (as a model for waning immunity) showed breakthrough infection and pneumonia by Delta.
Question: Can low dose vaccination be considered as a good model for waning immunity, as Ep 177-12 shows that after a full-dose vaccination, even at a time when peripheral blood parameters decline (= apparently waning), the central T and B memory responses may evolve favorably….
mRNA vaccination in people with HIV (PLWH)
Ep 177-14: Good news from Israel in CMI : Pfizer mRNA vaccine appears equally immunogenic and safe in PLWH who are on antiretroviral therapy with unsuppressed CD4 count and suppressed viral load as compared to health care workers.
Limitation: no long-term follow-up on protection.
mRNA vaccination in cancer
Ep 177-15: 88 consecutive cancer patients (67 lung, 8 melanoma, 7 kidney and the remaining 6 had head and neck, bladder , breast and squamous cell skin cancer) all on checkpoint inhibitor (PD-1/ PD-L1 antibody) therapy were vaccinated with Pfizer: good T and B cell responses for their age (median 68) and no COVID cases.
Combined vaccination after solid organ transplantation (SOT)
Ep 177-16: Transplant (mainly kidney) responded clearly weaker both T cells and neutralizing antibodies to either mRNA (Pfizer or Moderna) or Adeno vector (AZ) vaccines, but:
- IgG and neutralizing activity were more pronounced after mRNA priming (p = .0001 each), whereas CD4 and CD8 T cell levels were higher after vector priming (p = .009; p = .0001).
- SARS-CoV-2–antibodies and/or T cells after second vaccination were induced in 100% of controls and 70.6% of transplant recipients.
- While antibodies were only detected in 35.3% of patients, cellular immunity was more frequently found (64.7%)
- From Fig 3 p. 9, we can see that heterologous (vector/mRNA) induces better B and T responses in SOT patients, as compared to homologous regimens.
- No data on subsequent protection from disease.
Vaccines in multiple sclerosis (MS) treated with B cell-depleting anti-CD20
Ep 177-17: Brill (Israel) in Jama Neurology: after Pfizer vaccination
- Preserved T cell response, irrespective of anti-CD20 treatment (in this case only IFN-g ELISPOT)
- Defective S1 and RBD specific humoral responses, only in those patients who received anti-CD20 (Fig 1 A/B p. 3)
Ep 177-18: Gadani (USA) very similar observations after either Pfizer, Moderna of Janssen vaccine
Ep 177-19: Apostolides in Nat Med (§Aug 2021) more detailed analysis various parameters after mRNA vaccination
- Most MS patients do not generate optimal antibody responses (Fig 1 c and d p. 5) as well as low B memory responses (Fig 1 f and g)
- T cell priming, especially of T helper 1 (IFN-g producing) CD4 T and CD8 T cells, is largely intact, while circulating T follicular helper cells (Tfh, the B cell helpers) were compromised (Fig 6 f p. 9).
- Unclear what the effect is on protection against wild type SARS-CoV-2 and VOC.
Comments and reviews
Ep 177-20: Comment by Yi Yun Noh (first author of Ap 177-2):
There are several arguments for an important role of T cell immunity in SARS-CoV-2 infection:
- In acute phase of the disease, there is a strong inverse correlation between the frequency of SARS- CoV-2- specific IFNγ- producing CD8+ T cells and disease severity.
- Patients with hematological malignancy, treated with B cell depleting anti-CD20 had no increased severity of disease or fatality, because of intact T cell immunity.
- Depletion of CD8 T cells in SARS-CoV-2 infected macaques: increased severity of disease.
This author also argues that, after vaccination, antibody responses wane, while T cell responses are more persistent.
Ep 177-12 however provides arguments for ongoing qualitative maturation of protective memory B cell responses even when levels of antibodies in the periphery decline.
Various attempts to promote T cell immunity as a vaccine strategy are briefly mentioned.
Ep 177-21: provides a nicely illustrated overview of the role of various players of the adaptive immunity: antibodies (neutralizing and non-neutralizing); memory B cells, CD4 T cell helpers and cytolytic CD8 T cells, with complementary role in mitigating infection and disease after natural infection or vaccination.
The authors foresee a similar scenario for SARS-CoV-2 as in case of “endemic” coronaviruses (HCoV) or pandemic versus seasonal influenza:
- An ‘endemic equilibrium’ between hCoV infection and human immunity, whereby the rate of decay of immune memory is balanced by regular reinfection and boosting of immune responses, leading to frequent mild infections in the context of short- lived memory
- The high pathogenicity of pandemic influenza virus strains during primary infection is reduced to a level equivalent to the pathogenicity of circulating (seasonal and/or endemic) influenza virus strains in the presence of an immune population
Some preliminary conclusions
- CD4 and CD8 T cell responses during natural infection may modulate disease expression:
- Early responses may prevent severe disease and preliminary CD8 T cell depletion in experimental animals aggravates disease expression.
- After severe disease, often in elderly, T (and B) cell responses are higher than after mild disease. The “protective quality” of these responses is questionable.
- After vaccination, T and B responses are clearly higher and more persistent in healthy younger adults. T cell responses peak earlier, suggesting a “helper” function for antibody production. Anyhow, until now, the protective ability of (neutralizing) antibodies after vaccination is much more evident than the protective capacity of any T cell function.
- In various conditions with reduced immunity, the T and B responses after vaccination are weaker. While mRNA vaccines, have the best record overall, adenoviral vaccines may induce relatively more T cell responses and an heterologous Adeno-RNA approach may be beneficial in some instances.
- Only some papers (e.g. 10, 12, 13…) dig a bit deeper in T and B cell function (than just IFN and antibody production). From these papers a “nuanced” picture arises:
- B cell memory may mature over months after vaccination to broaden recognition of VOC, even when peripheral antibodies decline. This phenomenon was previously also observed after infection and suggests persistence of the antigen within the immune system (e.g. associated with so called “follicular dendritic cells?)
- Persistent T cells (e.g. CD4 + T follicular cells) may have a role in this maturation process
- While CD8+ T cell responses to peptide pools and to conserved parts of the proteins may seems unaffected by mutations in VOC, some mutations may affect T cell recognition in an HLA-dependent way.
- At an individual level, especially older and immune-compromised individuals may be confronted with declining T and B immunity after vaccination and increasing risk of re-infection and potentially severe disease by new variants, requiring carefully adapted booster vaccines. Nevertheless, in the part of the world, where vaccination has been rolled out at a population scale, “herd immunity” is increasing, which will promote “endemization” of SARS-CoV-2, hopefully rendering the virus less pathogenic at a population scale (while still dangerous for frail individuals).
I hope this long episode has clarified your insights in the complex and fascinating world of the T cell-B cell- virus interactions….?
9 August Episode 279: BA.2.75, novel monoclonal Ab, polymerase and anti-inflammatory treatment options
> More info
2 August 2022 Episode 278: Follow up on novel vaccine concepts: mucosal application and broadening towards “pansarbeco”
> More info
19 July 2022 Episode 275 SARS-CoV-2 infection or vaccination, risk of reverse transcription
> More info