Prolonged SARS-CoV-2 Infection in Patients with Lymphoid Malignancies

Abstract

Coronavirus disease 2019 (COVID-19) infection results in both acute mortality and persistent and/or recurrent disease in patients with hematologic malignancies, but the drivers of persistent infection in this population are unknown. We found that B-cell lymphomas were at particularly high risk for persistent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) positivity. Further analysis of these patients identified discrete risk factors for initial disease severity compared with disease chronicity. Active therapy and diminished T-cell counts were drivers of acute mortality in COVID-19-infected patients with lymphoma. Conversely, B cell-depleting therapy was the primary driver of rehospitalization for COVID-19. In patients with persistent SARS-CoV-2 positivity, we observed high levels of viral entropy consistent with intrahost viral evolution, particularly in patients with impaired CD8⁺ T-cell immunity. These results suggest that persistent COVID-19 infection is likely to remain a risk in patients with impaired adaptive immunity and that additional therapeutic strategies are needed to enable viral clearance in this high-risk population. SIGNIFICANCE: We describe the largest cohort of persistent symptomatic COVID-19 infection in patients with lymphoid malignancies and identify B-cell depletion as the key immunologic driver of persistent infection. Furthermore, we demonstrate ongoing intrahost viral evolution in patients with persistent COVID-19 infection, particularly in patients with impaired CD8⁺ T-cell immunity.This article is highlighted in the In This Issue feature, p. 1.

Figure 1:. High rates of persistent SARS-CoV-2 PCR positivity in hematologic malignancy patients who received systemic chemotherapy or B-cell depleting agents.

Swimmer’s plot depicting duration of known SARS-CoV-2 PCR positivity relative to first known positive PCR test. Colored lines represent times of persistent positivity (at least two positive tests). Treatments, hematologic cancer subtypes, deaths during follow-up, and individual PCR test results as indicated.

Figure 2:. Predictors of mortality during acute SARS-CoV-2 infection in patients with lymphoma.

a) Multivariate analysis of clinical risk factors predicting mortality attributable to COVID-19 infection in patients with lymphoma. Statistically significant risk factors highlighted in red. b) Left, two-dimensional Uniform Manifold Approximation and Projection (UMAP) of laboratory data for 36 patients with lymphoma in whom at least 10% of all laboratory analyses were conducted. Missing data is imputed. Right, description of three patient clusters generated based on Earth Mover’s Distance (EMD) of each patient relative to each other in UMAP space. c) Laboratory markers in patients who were alive or dead following acute COVID-19 infection. d) Kaplan-Meier curve displaying probability of survival within 60 days of acute COVID-19 infection in patients with decreased CD8⁺ or CD4⁺ T-cell counts. **p<0.01. ***p<0.001. ****p<0.0001.

Figure 3:. Adaptive immune dysfunction in lymphoma patients with recurrent COVID-19 infection.

a) Multivariate analysis of clinical risk factors predicting re-hospitalization attributable to COVID-19 infection in patients with lymphoma and prior hospitalization attributable to COVID-19. Statistically significant risk factors highlighted in red. b) Kaplan-Meier curve displaying probability of re-hospitalization for COVID-19 in patients with lymphoma based on receipt of anti-CD20 therapy in the year prior to first known documented SARS-CoV-2 PCR positivity. c) Laboratory markers in patients with lymphoma who were either re-admitted or not re-admitted to the hospital for symptomatic COVID-19 infection following discharge from the hospital for acute COVID-19 infection. *p<0.05. **p<0.01.

Figure 4:. Increased SARS-CoV-2 viral genome entropy in patients with adaptive immune dysfunction.

a) Minor variants in SARS-CoV-2 sequences from 17 patients with longitudinal data. Samples from the same patient are grouped together and highlighted with the same background color. Location of each variant is indicated by a vertical line. Red lines represent the recurrent variants, which are seen in at least two samples from the same patient. Blue lines represent the variants unique to a particular sample. White gaps indicate regions with low sequencing coverage. b) Viral entropy of the SARS-CoV-2 genome in patients with normal or decreased CD8⁺ T-cell counts (above), or patients who had or had not received anti-CD20 therapy in the past year (below). *p<0.05. For patients with viral sequencing at multiple time points, the last time point was utilized.