Background
Particularly older patients and patients with comorbidities have been reported to suffer from complications and fatalities due to coronavirus disease 2019 (COVID-19) [Citation1–5]. COVID-19 has given rise to a case-fatality rate (CFR) above 20% in cancer patients compared to 5–6% in non-cancer patients [Citation6–11]. Patients with haematological cancers have demonstrated the highest CFRs, often exceeding 30% [Citation9,Citation12,Citation13]. Emphasising the vulnerability, seroconversion 21 days after COVID-19 vaccination was 94% in healthy controls compared with 18% in patients with haematological cancer [Citation14].

Almost all previous studies have investigated prognosis using mortality rate at 30 days or CFR without specified follow-up period [Citation7,Citation15–17] and prognosis data beyond one month is scarce [Citation12,Citation18,Citation19]. Since patients with haematological cancer may have a delayed or even absent clearance of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [Citation20], studies with longer-term follow-up are relevant to monitor mortality and post infection symptoms. Recently it was reported that even SARS-CoV-2 infected patients that were not hospitalised, more frequently initiated bronchodilating agents or received a subsequent hospital diagnosis of dyspnoea than matched SARS-CoV-2 negative comparisons [Citation21]. Similarly, we found that among 66 patients with haematological disease and SARS-CoV-2 infection, 57% reported fatigue, 46% reduced functional abilities, and 33% dyspnoea one month after verified infection [Citation22]. In the present study, we extended inclusion period and follow-up to include additional patients and study mortality and functional capacity 90-days after SARS-CoV-2 infection. Herein we present data regarding clinical presentation and outcome for 108 adult Danish patients with haematological disease and SARS-CoV-2 infection verified before September 1, 2020.

Material and methods
Data sources and patients
Our study included patients with a haematological diagnosis ≥18 years with verified SARS-CoV-2 infection. Patients were in clinical follow-up at a Danish department of haematology for any subtype of haematological cancer or non-malignant blood or bone marrow disorder at the time of infection. Danish haematology departments are managed by public hospitals that provide universal free access to health care. SARS-CoV-2 testing was also facilitated by public authorities and was performed using nasal or oropharyngeal swabs and reverse transcriptase polymerase chain reaction techniques. Antigen tests was not routinely implemented during the inclusion period. Patients with respiratory tract symptoms or fever were routinely SARS-CoV-2 tested throughout the study period and by end of April 2020, extended to asymptomatic patients who were scheduled for any type of in-hospital admission. For this study, patients were identified during routine clinical practice through contacts with their haematological departments as described previously [Citation22]. The available clinical and para clinical data were abstracted from medical files by local investigators. The current study included patients for whom SARS-CoV-2 infection was verified before September 1, 2020, allowing at least three months of follow-up. Patients were eligible irrespective of severity of SARS-CoV-2 infection and therefore both hospitalised patients and patients managed at home were included. Mortality and self-reported symptoms (fatigue, dyspnoea, and functional abilities) were re-evaluated approximately three months after verified infection. Data regarding self-reported symptoms were only provided for patients evaluated through a routine virtual or an on-site hospital consultation between 80 and 120 days after confirmed infection. Follow-up terminated January 1, 2021.

Baseline data
We registered age, sex, body mass index (BMI), PS, living conditions, haematological diagnosis, previous (> 6 months), recent (0–6 months) and ongoing haematological treatment, remission status, comorbidity, and biochemical results at the time point of the verified SARS-CoV-2 infection. Charlson Comorbidity Index (CCI) and Cumulative Illness Rating Scale (CIRS) was used to summarise comorbidity [Citation23,Citation24]. Additionally, we registered comorbid conditions (e.g., diabetes and pulmonary disease) of particular interest [Citation5,Citation24]. Haematological diagnoses were aggregated in four groups including multiple myeloma (MM), chronic lymphocytic leukaemia (CLL)/lymphoma, acute leukaemia (AL)/myelodysplastic syndrome (MDS), and other haematological disorders (Chronic myeloproliferative neoplasm subtypes, chronic myeloid leukaemia, hairy cell leukaemia, T-cell large granular leukaemia, paroxysmal nocturnal haemoglobinuria, aplastic anaemia, and immune thrombocytopenia)

Definitions
SARS-CoV-2 infection was graded as either asymptomatic/mild, severe, or critical based on the available information about symptoms and observations in medical files[Citation5]. Patients with no or only mild non-pneumonia/pneumonia symptoms were classified as asymptomatic or mild. Severe infection corresponded to patients with fever, respiratory symptoms, dyspnoea, respiratory frequency ≥30/min, blood oxygen saturation ≤93%, partial pressure of arterial oxygen to fraction of inspired oxygen ratio <300, and/or radiology confirmed infiltrates in >50% of the lungs within 24 to 48 h from presentation [Citation5]. Patients with critical infection met one of three criteria: respiratory failure, septic shock, or multiple organ failure [Citation5].

Statistical analysis
We summarised baseline patient characteristics, SARS-CoV-2 findings, and outcomes using means/proportions by the pre-defined diagnosis groups and classification of SARS-CoV-2 severity. Patients were followed from the date of verified SARS-CoV-2 infection until death or censoring (January 1 2021), whichever came first. Overall survival (OS) was estimated using the Kaplan-Meier method. The risk of admission to intensive care unit (ICU) was estimated by the proportion of patients admitted to ICU within 90 days of verified infection. Ninety-five percent confidence intervals (95% CIs) for risk of ICU admission were computed according to the exact method described by Clopper and Pearson [Citation25]. Associations between SARS-CoV-2 severity or ICU admission and clinical risk factors were tested using Fishers’ exact test for grouped exposures or univariable logistic regression for continuous exposures. Analyses were conducted using the statistical programming language R (version 4.0.3).

Ethics
This study was approved by the Danish Region of Southern Denmark (record: 20/13067). The Danish council for patient safety (record: 31-1521-230) and ethics committee (record: 20202000, 53) waived informed consent due to non-interventional study design and registration of data from routine clinical care only.

Results
Baseline characteristics
The 108 included patients with haematological disorders and SARS-CoV-2 infection, had a mean age of 68.5 years (SD, 13.8) and 66 (61.1%) were male. The four aggregated diagnoses groups comprised; CLL/lymphoma (n = 49, 45.4%), MM (n = 20, 18.5%), AL/MDS (n = 14, 13.0%), and other haematological disorders (n = 25, 23.1%)) (Table 1). Therapy for the haematological disorder was ongoing in 60 (55.6%) patients, recent in 12 (11.1%), prior in 17 (15.7%) patients, and 19 (17.6%) patients were untreated (Supplementary Table 1). Prior stem cell transplantation was performed in 13 patients (3 allogeneic, 10 autologous; 12%). Comorbidity - CCI score was ≥1 in 57.4% and mean CIRS score was 6.4 (SD 4.5) (Supplementary table 1). The most common among the collected comorbid conditions were pulmonary disease (15.7%), heart disease (12.0%), and diabetes (11.1%) (Supplementary table 1). Body mass index ≥ 30 kg/m2 was present in 14.4%. Neutrophil granulocyte count was below 1.0 × 109/l in 13.8%, lymphocyte count below 1.0 × 109/l in 50.6%, and IgG blood levels below 5.0 g/l in 26.2% (Supplementary table 1).