Dear Editor,

Cancer has been shown to be associated with higher risk of COVID-19 complications [1]. However, data on patients with COVID-19 and an underlying hematological disease as well as on specific risks factors in this particularly immunocompromised population are scarce [2,3,4,5]. We conducted a retrospective study in a tertiary center of 1000 beds with a hematology reference center. Our ethics committee approved the study (N° CEHF 2020/06AVR/201). Between March 13 and May 15, 2020 a total of 375 consecutive patients were hospitalized with COVID-19 and among them 13 (3.4%) met the inclusion criteria of having an underlying hematological disease. Demographics, clinical characteristics and laboratory findings are summarized in Table 1. The median age was 70 years (IQR 59–79) and 77% of patients were male. COVID-19 pneumonia was the admitting diagnosis for the majority of patients (n = 10, 77%) and a delayed secondary diagnosis in 3 patients (23%) with one of them being highly suspect for nosocomial infection. Diagnosis was based on the association of positive RT-PCR and CT-scan in 11 patients (85%) and on compatible CT-scan only in the two remaining. Median duration of symptoms (after exclusion of patients presenting with symptom-overlapping acute conditions) was 8 days (IQR 3–10). The most common reported symptoms were fever (n = 12, 92%), shortness of breath (n = 8, 62%) and cough (n = 5, 39%). Lymphopenia was present in 5 patients (39%) and neutropenia (grade 3 or more) in 2 patients. Therapy directed against COVID-19 included hydroxychloroquine for 10 patients (77%) with addition of methylprednisolone in 2 patients, azithromycin in 1 patient, and lopinavir/ritonavir in 1 patient.

Table 1 Characteristics of 13 patients with hematological disease and COVID-19
Full size table

The underlying hematological diseases (Table 1) were distributed as following: 4 chronic lymphocytic leukemia’s (31%), 4 plasma cell dyscrasia’s (31%), 2 acute myeloid leukemia’s (15%) with one of them being secondary to primary myelofibrosis and the other one being a phenotype shift from early T cell precursor acute lymphoblastic leukemia, 2 non-Hodgkin lymphoma’s (15%) and 1 non-malignant condition which was a chronic hypogammaglobinemia of unknown origin. Two patients were stem cell transplant receptors (1 autologous and 1 allogeneic). Four malignant hematological diseases (33%) were newly diagnosed or in first line treatment, 3 were in remission or in a watch and wait strategy without ever having had any treatment, 2 were stable without remission and 3 were relapsed or refractory. Patients received a median of 1 (range 0–5) treatment lines.

Two patients were admitted to the intensive care unit (ICU) at presentation: one received high flow oxygen therapy and the other one invasive mechanical ventilation. Five more patients (39%) presented worsening respiratory state later during hospitalization and were medically eligible for an admission to the ICU but, owing to age and comorbidities, palliative care was provided instead. Four patients (31%) had a documented bacterial co-infection (Three urinary tract infections caused by Escherichia coli [2],and Enterococcus faecalis [1] and one septicemia caused by Escherichia coli probably secondary to mucositis in a patient with febrile neutropenia). Overall, 6 patients (46%) died during their stay in our COVID-19 units (n = 5) or ICU (n = 1). Comparing survivors and non-survivors, we observed that non-survivors were significantly older than survivors with a median age of 80 [interquartile range (IQR) 70–83] versus 60 (IQR 45–79; P = 0.043) respectively. Of interest, the hemoglobin level at day 1 was lower in non-survivors [mean ± standard derivation (SD) 10.0 ± 1.4] than in survivors (mean ± SD 12.5 ± 2.2; P = 0.037) besides not being correlated to age (r = − 0.09; P = 0.765). No specific type, status or treatment of hematological disease was shown to be associated with a higher mortality in our series.

We identified two covariates that were significantly associated with worse outcomes in our patients: older age and lower hemoglobin level at day 1.

A higher mortality in older patients was already shown elsewhere [3, 5]. The association between lower hemoglobin level at presentation and a higher mortality rate was also found by Mehta et al. [4]. Of interest, they demonstrate that myeloid malignancies show a trend for higher mortality compared to lymphoid malignancies. Like highlighted by Martin-Moro et al. [3], it can be discussed whether this effect is intrinsic to the myeloid character of the disease or rather due to other covariates [e.g. older age and presence of a symptom-reduction-intention treatment often present in patients with myeloproliferative neoplasm (MPN) or myelodysplastic syndromes (MDS)] [3]. In our small cohort of patients, we found no association between the myeloid/lymphoid character of the underlying disease and COVID-19 fatality (data not shown) but, to mention, our series did not include any MPN or MDS except under the form of a progression to secondary acute myeloid leukemia.

In conclusion, patients with hematologic malignancies are very vulnerable to COVID-19. Age and low hemoglobin level (on day 1) seems to be factors associated with poor outcome. Larger prospective and cohort study are needed to identify other factors associated with mortality in this population.