Since the onset of the SARS-CoV-2 pandemic, the heterogeneity of the clinical course has been an important issue for investigations. While in some patients the infection goes unnoticed, in others the same infective agent leads to a severe course of disease accompanied with high morbidity and mortality20,21. Furthermore, there could be a combination of several other factors that might lead to these outcomes like gender, distance regulations, length of the infection wave or season. We found elevated or stable period IgG antibody levels of up to 240 days after the end of symptoms. In all patients in whom elevated antiviral IgG antibodies had been observed the respective anti-S IgG antibodies remained detectable for at least 240 days. This is in accordance with our previous publication in the European Journal of Medical Research regarding the first infection wave with SARS-CoV-210 and our report to the New England Journal of Medicine22. In these publications, we demonstrated in detail, the time sequence of IgG, IgA and IgM plasma levels against the spike and nucleocapsid protein during a SARS-CoV-2 infection. Unfortunately, other institutions reported a very early and rapid decline of the antiviral IgG antibodies which is in contrast to our results23,24. This context may constitute the most important argument why a booster vaccination against SARS-CoV-2 is essential for all people23,24.
Individual immunologic pathways based on metabolism, age, sex and genetics are known to play a crucial role in disease outcome 25. Our study demonstrates a prolonged clinical course in carriers of some HLA haplotypes while the impact is less in those with concomitant favourable HLA variants. Our results fit perfectly with what we already know about HLA superhaplotypes26
The occurrence of one of three HLA haplotypes was a prognostic marker for a mild course of disease and three distinctive haplotypes were associated with a more severe disease course
We know that SARS-CoV-2 infection and its severity might be related to gender, and males are more likely to develop COVID-1927. However, in our study females predominated. One of the explanations for this is that the first group we analyzed in our previous reports were skiers who came from a SARS-CoV-2 hotspot in Austria. There were more men in this group than in the second wave of infection. The second group were blood donors following public call (newpaper and TV) for convalescent plasma donation. When we advertise for blood donations, usually more woman than men respond., We therefore remain to assume that men may be at higher risk for more severe SARS-Cov-2 infection, but this is not reflected in our investigation and may bias our results.
Another limitation of our study could be the characterization of our patients. We used the WHO classification to describe the symptoms. But still parameters like long Covid were not analysed. All patients were non hospitalized and healthy indiviudals with no co morbidities and and an ECOG performance status 0 before the infection. The fact that patients of a wide age range were grouped together could further limit the significance of our results.
Previously, we conducted and subsequently reported the results of an observational clinical study examining the impact of different HLA genotypes, ABO blood group antigens, and the CCR5Δ32 mutant on the clinical course of COVID-1910. We identified protective HLA alleles (HLA class II DRB1*03:01 and HLA class I B*35:01) associated with a shorter disease duration10. In addition, we demonstrated that patients with heterozygous CCR5Δ32 mutation, in which proper function is almost completely abolished, had a longer disease course10. Remarkably, these patients also had significantly reduced anti-S IgA levels after infection, suggesting that the nasopharyngeal humoral adaptive immune response may be impaired. With regard to blood groups, we also found that ABO heterogeneity correlated with a higher risk of a worse outcome.
Our current study analyses HLA haplotypes and the role of CCR5. Compared to our previous analyses, we expanded the sample size for the second wave of the SARS-CoV-2 pandemic. By regulating lymphocytes and stimulating leukocyte migration, the G protein-coupled receptor CCR5 can trigger inflammatory responses. CCR5 gained special attention because of its unique role in HIV infection. As a co-receptor on the cell surface of T-lymphocytes, CCR5 enables the HI-virus to migrate into cells. While patients with a homozygous CCR5Δ32 mutation appear to be resistant to HIV infection, heterozygous carriers are still partially protected from infection by reduced receptor activity. Furthermore, for early summer meningoencephalitis (TBE) and West Nile fever, Kindberg et al. showed that the allele frequency of the CCR5Δ32 mutation correlates with the severity of the clinical course28. In contrast to HIV infection, homozygous CCR5Δ32 carriers show an increased risk for a severe course of West Nile fever. In the case of SARS-CoV-2 virus, we have shown that CCR5Δ32 is associated with a significantly prolonged disease course during the first wave. This finding was not supported for the second wave. The data from the first and second waves are only comparable to a limited extent, as the courses in the first phase were generally longer than in the second. One explanation for this could be that we recruited mainly men in the first wave, whereas the majority of donors in the second wave were women. It is known that men tend to have longer disease courses. Another explanation could be that higher viremia due to lack of protective strategies such as clearance regimens or protective masks led to longer disease courses in the first wave. Furthermore, the first wave occurred in the cold season, whereas the second wave occurred in the summer.
At first glance, our results seem to contradict previous publications. Patterson et al. recently demonstrated that the CCR5 antagonist Leronlimab reduced SARS-CoV-2 plasma viremia and hyperinflammation in ten critically ill patients with severe COVID-1916. The difference between the two studies is that our cohort did not include critically ill patients and our end point is duration of symptoms. Brevity of disease progression is associated with rapid and targeted activation of the immune system, whereas interrupting the hyperinflammatory response, may improve prognosis in critically ill patients with overshooting infection. Thus, the two studies address different arms of the immune response. The overall view of our results and those of the previous study suggest that modulation of CCR5 requires adaptation to the phase of infection to prevent counteracting effects.
In a cohort of 416 patients recovered from SARS-CoV-2, Hubacek et al. demonstrated that the incidence of CCR5Δ32 was significantly higher in asymptomatic patients than in symptomatic patients18. However, the Tübingen group reported opposite results in 110,544 stem cell donors17. In this study, the CCR5 polymorphism did not appear to correlate with either symptomatic vs asymptomatic course or with the likelihood of infection with SARS-CoV-2. Whereas symptom duration time is the end point in our study, the previous work focussed on symptom onset. Given the precise CCR5-mediated initiation of infection, it is conceivable that patients may become symptomatic but have shorter disease courses. Thus, the initial, seemingly contradictory results of the first study can be explained by the different endpoints.
Having postulated that CCR5 leads to triggering of the immune response by clustering, we wondered whether previously described positive HLA variants could attenuate this effect in CCR5Δ32 mutant patients. We found that in patients with favourable HLA profiles, the CCR5Δ32 prolonged immune response was attenuated. Therefore, we hypothesized that CCR5Δ32 mutation presentation across HLA alleles may modulate this effect in the presence of impaired clustering.
We are well aware that the patient cohort may be too small and too heterogeneous to conclusively assess the role of CCR5 on the duration of clinical course. In addition, our study does not include severe courses of COVID-19. Their inclusion would be desirable to make a more sophisticated prediction regarding the different phases of the immune response. Prospective studies with larger patient collectives would be desirable. The candidate gene approach performed has the natural limitation that it is based on pre-existing knowledge and therefore may underestimate certain molecular processes. It does not analyze multiple genetic effects together, which risks overlooking existing molecular networks.
In conclusion we were able to identify the HLA haplotypes in 536 patients with SARS-CoV-2 and showcase that the presence of certain haplotypes is correlated with the severity of COVID-19. Patients with HLA haplotypes A*03:01_B*35:01_C*04:01, A*29:02_B*44:03_C*16:01 and DRB1*13:01_DQA1*01:03_DQB1*06:03 recovered in a shorter period. We also focused on CCR5 and its role in the duration of the disease course among the patients. In particular, the presence of mutated CCR5 resulted in a longer recovery period.