Convergence in age adjusted mortality rates and proportion of all deaths from COVID-19 among patients with hematologic malignancies, 2020-2024 Free

Introduction: Patients with hematologic malignancies (HM) are often immunocompromised, either as a result of their illness or its treatment (e.g., chemotherapy), and prior reviews in this population have been inconclusive. One retrospective cohort study found no effect of immunosuppression with prednisone, tacrolimus, or mycophenolate mofetil on ventilation, in-hospital mortality or length of stay among hospitalized COVID-19 patients in one hospital system (Andersen et. al., 2021). One registry study in Spain found a threefold increase in mortality from COVID-19 infections among patients with HM compared to the general population, and a twofold increase compared to patients with solid tumors (García-Suárez et al., 2020). One registry study in the U.S. showed a 14% proportion of COVID-19 deaths among chronic lymphocytic leukemia patients from 2020-2021 (Fedeli et. al., 2024). We compared the age-adjusted mortality rates (AAMR) of COVID-19, and the proportion of deaths attributed to COVID-19, in patients with HM compared to the general population.

Methods: National mortality trends from 2018 to 2024 in the United States were exported from the CDC WONDER database, specifically the “multiple causes of death” function, using the ICD codes C81-C96 (Malignant neoplasms of lymphoid, hematopoietic, and related tissue) AND the ICD code U07.1 (COVID-19). Data was analyzed as a share of all deaths (COVID-19 among patients with HM as a share of all HM-associated deaths, and COVID-19 as a share of all deaths) and rates were reported as deaths per 1,000,000 population. Data was limited to patients 25 years or older at the time of death to enable AAMR calculations per 10-year strata. Standard error was propagated in shares of total deaths by the formula Δ = c/b * sqrt((Δc/c)^2+(Δb/b)^2) in which c is defined as the death from COVID-19 rate per 1,000,000 population, Δc is the standard error of c as reported by CDC WONDER, b is the total HM-associated death rate or total death rate per 1,000,000 population, Δb is the standard error of b as reported by CDC WONDER, and Δ is the standard error of the resulting ratio. Provisional data from 2024 was included in the analysis.

Results: AAMRs from all HMs increased in most regions from 2019 to 2020-2021, in line with the peak of COVID-19 associated deaths. All regions demonstrated a decrease in AAMR from all HMs from 2021 to 2022-2023, in line with the moderation of COVID-19 associated deaths. In all regions, deaths from COVID-19 in patients with HM fell in terms of AAMR and as a share of all deaths among HM patients. The Western region enjoyed a faster moderation in the share of COVID-19 associated deaths towards 2% in 2023 among HM patients. Peak COVID-19-associated deaths were higher in the general population than among HM patients. Among the general population, peak COVID-19 deaths as a share of all deaths were 15% in the Northeast in 2020, 12% in the Midwest in 2021, 15% in the South in 2021 and 14% in the West in 2021. Among patients with HM, peak COVID-19 deaths as a share of all HM-associated deaths was 7% in all regions in 2021. These proportions of all deaths moderated and converged – as of 2024, COVID-19 accounted for 1.75-2.00% of deaths in all regions among both patients with HM and the general population. COVID-19 deaths among patients with HM as a share of all HM-associated deaths demonstrated a similar pattern in both sexes with an increase in death rates among males in all five years.

Conclusion: There was a higher peak AAMR from COVID-19 as a proportion of all deaths in the general population during the heights of the COVID-19 pandemic (2020-2021) than among patients with HM as a share of all HM-associated deaths in the same period across all regions. The all-cause AAMR is higher at baseline among patients with HM, so the effect of COVID-19 may have been smaller in this population as a share of all deaths. Patients known to have HM may access care, take airborne precautions and vaccinate more regularly. Treatment with dexamethasone, nirmatrelvir-ritonavir and the management of severe infection with low-tidal volume ventilation and pronation are now standards of care applied to all hospitalized patients regardless of immunocompromised status. Future directions for research should include calculating a case-fatality ratio among these populations over time, and deploying a principal component analysis to assess which underlying risk factors are most salient.