This nationwide study investigated the clinical and demographic profiles of HSCT recipients who also had an underlying HM and who were hospitalized during the COVID-19 pandemic. We also identified predictors for in-hospital mortality in this population. Our results provide insight into the patient characteristics of this population, and their risk factors for mortality, highlighting their vulnerabilities and elucidating their clinical course and outcomes.
Among the 2,324 HSCT recipients, the overall in-hospital mortality rate was 13%, notably lower than the 20% observed in non-HSCT patients. This difference highlights the complexity of HSCT patients, as this finding can be counterintuitive. When analyzed the entire cohort of 35,648 patients with HMs, our univariate and multivariate analyses did not show an independent protective effect of HSCT on COVID-19-related mortality. This finding contrasts with the lower mortality observed in transplanted patients compared to those who did not undergo HSCT (13% vs. 20%, p< 0.001). The apparent survival advantage in transplanted patients may be explained by selection factors, such as the inclusion of younger patients and those with fewer comorbidities in the HSCT group, rather than a direct effect of transplantation on immunity against COVID-19. Additionally, the lack of detailed data on the type of transplant (allogeneic vs. autologous) and pre-infection immunological status limits our ability to assess the specific impact of HSCT on clinical outcomes.
According to a recent narrative review, reports of COVID-19-related mortality among HSCT recipients are mostly based on single-center studies; those from multicenter studies involved a low number of individuals17. The review also indicated significant heterogeneity in patients, clinical settings, inclusion criteria, follow-up periods, and outcomes. Furthermore, most of these studies were conducted during the early pandemic, before vaccines and other specific treatments were yet available. Hence, the mortality rate among HM patients may be overstated. Nonetheless, it is clear that mortality in hematologic patients, regardless of HSCT status, is increased relative to the general population. However, there is evidence that morbidity and mortality outcomes may vary significantly between hematologic patients who do and who do not undergo HSCT. For example, Shah et al.18, found that, while HSCT recipients experienced high mortality (up to 22%), they had favorable outcomes compared to non-HSCT recipients. They interpreted their findings as being due to the adaptive immune response after HSCT against COVID-19. Thus, HSCT patients are able to modulate immune reconstitution following transplantation19,20. Although SARS-CoV-2 infection may cause lymphopenia in severe cases, HSCT recipients are able to recover T cells, in spite of marked lymphopenia and circulating B cells. Indeed, such patients can produce SARS-CoV-2 antibodies18,21. A later meta-analysis of 19 studies including 2,031 HSCT recipients reported a median age of 57 years, similar to our findings, and a mortality rate of 19%22.
In our study, we found that mortality was higher in 2022, albeit specifically in the first months of the year. This may reflect the evolution of the pandemic in terms of SARS-CoV-2 variants. The omicron subvariants BA.4, BA.5, and BA.2.75 were predominant in Spain beginning in the first months of 2022. They were reported to be more immune-evasive, and although they only caused mild disease in healthy individuals, they could cause severe disease in vulnerable populations, such as the elderly or immunocompromised, during these months6,23.
In line with previous studies, older age at diagnosis of COVID-19 was associated with poor outcomes24. We likewise found that dementia and leukemia were significant predictors of mortality, in both logistic regression and survival analyses. Dementia is associated with age, probably reflecting a certain degree of multicollinearity between these two variables. These comorbidities may be key determinants of poorer outcomes, not only in HSCT recipients, but also more generally in patients with HMs18,25.
We found that acute leukemias were also significantly associated with mortality, likely due to their aggressive clinical course and the profound immunosuppression inherent in their management26. Our results align with those of Sharma et al.14, who included 318 patients diagnosed with COVID-19 in a survival analysis and studied mortality during the first months of the pandemic. They found that solid tumors, plasma cell disorders (i.e., multiple myeloma and malignant plasma cell neoplasms), and lymphoma were relevant predictors of mortality. Our results also support Shah et al.18. , who also found that active malignancy, particularly leukemia, was associated with mortality in HSCT recipients.
In our study, ICU admission was also predictive of mortality in both logistic regression (OR = 0.13, 95%CI 0.09–0.19, p < 0.001) and in survival analysis (HR = 0.42, 95%CI 0.31–0.56, p < 0.001). However, it may act as a confounding factor rather than a direct protective factor, and our finding requires careful interpretation. We believe that this is likely to have been influenced by survivorship bias. That is, patients who survive long enough to be admitted to the ICU are younger or have a less severe case of the disease and may inherently differ from those who did not match ICU admission criteria. The latter group of patients might experience rapid clinical deterioration or death before ICU admission. Furthermore, ICU admission could act as a surrogate marker for the availability and intensity of advanced medical care, not as a factor that directly mitigates mortality risk. Thus, this finding should be interpreted with caution, and clinicians should consider their clinical decision-making and resource availability when interpreting the role of ICU admission in morbidity and mortality risk among HSCT recipients. Indeed, future studies would benefit from incorporating additional variables, such as the timing of ICU admission, the onset of clinical deterioration, the management of complications, and the immunological status of patients prior to assessing the role of ICU admission as a potential confounder versus a direct protective factor.
Limitations and strengths
The potential limitations of our study include the inherent characteristics of the registry data itself. Detailed immunological and treatment data were not available. We did not collect information on graft type, i.e., allogeneic and autologous HSCT. In addition, we lacked data on the year of transplantation, as these data were not available in the administrative dataset. Patients with graft-versus-host disease were not identified. More importantly, no information on the immunologic profile before the SARS-CoV-2 infection (CD4+/CD8 + count) or on the status of the HM (relapse or progression of disease after transplantation) were available. We could not assess the impact of specific treatments such as corticosteroids, antivirals, or immunosuppressive therapies on mortality.
The lack of data on immunologic status, the diversity of graft sources, and the distinct immunosuppression therapies employed may have impacted the interpretation of our findings, although we focused on trends over time and the description of HSCT recipients with HMs using available data from the national registry, reporting the results for a large cohort to fully characterize risk factors for mortality. Therefore, the strengths of our study include its nationwide scope and its large sample size. We also used a robust methodological approach. By analyzing this centralized administrative database, we captured comprehensive information to depict an overall picture and enhance the generalizability of our findings.
Nonetheless, further research is needed to assess the role of immunological recovery, graft-versus-host disease, other transplant-specific factors, and the biomarkers of immune function that may be associated with mortality due to COVID-19 infection to provide further insight into risk stratification for HSCT recipients.
