To the Editor:

We read with great interest the paper by He et al. [1] on the outcome of COVID-19 infection in hematological malignancy patients. This study was one of the first to report the outcome in hematological malignancy patients and since then many others have reported theirs [2,3,4,5]. Most of these studies [1,2,3,4, 6] included none or very few hematopoietic stem cell transplant (HCT) recipients (N < 4) and therefore we like to present the largest data to our knowledge of the impact of COVID 19 infection in HCT recipients. We identified 34 HCT recipients [autologous, (autoHCT) N = 14 and allogeneic, (alloHCT) N = 20] diagnosed with COVID-19 across 4 cancer centers (Rush University, The University of Chicago, Mount Sinai and Northwestern University) in Chicago and New York City.

The median age for all patients was 57 years; 54 in alloHCT and 59 years in the autoHCT group (Table 1). The majority of patients were Caucasian males. Eleven patients (32%) were obese (BMI ≥ 30) and 20 (59%) had a history of smoking. Hypertension (N = 15, 44%) and diabetes (N = 13, 38%) were the most common comorbidities among all patients. Acute leukemia (N = 16, 47%), followed by multiple myeloma (N = 9, 26%) and lymphoma (N = 6, 18%) were the most common diagnoses. Peripheral blood (N = 30, 88%) was the stem cell source for the majority of them and most of the patients received myeloablative conditioning (N = 21, 62%).

Table 1 Patient and transplant characteristics and clinical features at presentation.
Full size table

The median time after HCT at which a COVID-19 infection occurred was 17.4 months for all patients; 18.9 months in the alloHCT and 13.2 months in the autoHCT recipients. Nine patients (26%) had active acute or chronic graft-versus-host disease (GVHD) at the time of COVID-19; 15 (44%) were on immunosuppressive drugs (ISP). Eleven (32%) patients were on steroids; ten (50%) in alloHCT (nine for GVHD, one for adrenal insufficiency), and one (7%) in autoHCT for relapse. The most common COVID-19 presenting symptom was fever (N = 25, 74%), followed by cough (N = 24, 71%) and dyspnea (N = 20, 59%) (Table 1). Lymphopenia (N = 12/30, 40%), anemia (N = 18/30, 60%) and thrombocytopenia (N = 11/30, 37%) were common. D-dimer was elevated in 11/19 patients (58%); 10/13 alloHCT patients (77%) and 1/6 (17%) autoHCT patients. Chest radiography obtained in 28 patients showed that bilateral infiltrates (N = 21, 75%) was the most common abnormality.

Twenty patients (59%) had mild/moderate and 14 (41%) had severe COVID-19 at presentation. Twenty-five of 34 patients (74%) [alloHCT N = 14, 41%; autoHCT N = 11, 32%] were hospitalized for COVID-19 (Table 2). The median duration of hospitalization was 8 days (range, 2–50 days). Eleven patients (32%) [alloHCT N = 8, 24%; autoHCT N = 3, 9%] were transferred to the medical intensive care unit (MICU). Eight of these patients (73%) were intubated [alloHCT group N = 6, autoHCT group N = 2], and two needed hemodialysis. Hydroxychloroquine (N = 15, 44%) was the most frequently administered therapy for COVID-19, followed by Tocilizumab (N = 6, 18%), Remdesivir (N = 5, 15%), convalescent plasma (N = 2), Lopinavir-Ritonavir (N = 1), Anakinra (N = 1), Sarilumab (N = 1), and Ribavirin (N = 1). None of the patients had their ISP discontinued, one hospitalized patient had tacrolimus dose reduced due to renal dysfunction. Seven patients (21%) died [alloHCT N = 5, autoHCT N = 2] at a median of 15 days from their COVID-19 diagnosis (range, 7–42); six of the 7 had severe COVID-19 at presentation and one had mild/moderate disease. The mortality rates among the mild/moderate and severe groups were 5% (N = 1/20) and 43% (N = 6), respectively (Supplementary Figs. 1–3).

Table 2 Hospital course of the admitted pts (N = 25).
Full size table

Analyzing the factors (Supplementary Table 1) associated with MICU admission and mortality, we found that age had a definitive impact on outcomes (Supplementary Fig. 4). For the entire cohort, patients <40 years avoided MICU transfer or death. AutoHCT recipients <60 years also avoided these outcomes, however, all allo-HCT recipients >60 years (n = 3) requiring MICU transfer died. Interestingly, being underweight (defined as BMI ≤ 20) rather than obesity was associated with a higher rate of MICU transfer. ISP use post alloHCT was not associated with poorer outcomes. Being on steroids at diagnosis of COVID-19 was significantly associated with an increased rate of MICU admission and mortality in the entire cohort (Supplementary Fig. 5). There was a trend toward increased mortality for the patients who had active GVHD at COVID-19 diagnosis (p = 0.09) (Supplementary Fig. 6). Patients with COVID-19 within the first year of alloHCT had poorer survival than that after the first year (0% vs 87%, respectively. p = 0.007). Patients with hemoglobin <10 g/dL or platelet count <150 (×109/L) had significantly increased mortality (Supplementary Figs. 7 and 8). Other factors associated with poor outcomes included a higher LDH, ferritin, and lymphopenia at COVID-19 diagnosis.

In conclusion, our study demonstrated that HCT recipients are at an increased risk of mortality compared to the general population [7]. Older age, being on steroids at diagnosis of COVID-19, and COVID-19 infection within 1 year of HCT, but not ISP use was associated with poor outcomes. Whether the steroid use by itself or comorbid conditions requiring steroid treatment is associated with inferior clinical outcomes deserves further investigation. Based upon our data, tapering or discontinuing ISP drugs in the setting of COVID-19 is not warranted.