COVID-19 in patients with chronic lymphocytic leukemia: a multicenter analysis by the Czech CLL study group

Abstract

Patients with chronic lymphocytic leukemia (CLL) have a high risk of poor outcomes related to coronavirus disease 2019 (COVID-19). This multicenter cohort study evaluated the impact of COVID-19 infection on the population of CLL patients in the Czech Republic. Between March 2020 and May 2021, 341 patients (237 males) with CLL and COVID-19 disease were identified. The median age was 69 years (range 38–91). Out of the 214 (63%) patients with the history of therapy for CLL, 97 (45%) were receiving CLL-directed treatment at diagnosis of COVID-19: 29% Bruton tyrosine kinase inhibitor (BTKi), 16% chemoimmunotherapy (CIT), 11% Bcl-2 inhibitor, and 4% phosphoinositide 3-kinase inhibitor. Regarding the severity of COVID-19, 60% pts required admission to the hospital, 21% pts were admitted to the intensive care unit (ICU), and 12% received invasive mechanical ventilation. The overall case fatality rate was 28%. Major comorbidities, age over 72, male gender, CLL treatment in history, CLL-directed treatment at COVID-19 diagnosis were associated with increased risk of death. Of note, concurrent therapy with BTKi compared to CIT was not associated with better outcome of COVID-19.

Introduction

Since its first report in 2019, the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) caused a global healthcare crisis [1, 2]. Given the predominantly advanced age [3] and comorbidities [4], patients with chronic lymphocytic leukemia (CLL) are at high risk for poor outcomes related to coronavirus disease 2019 (COVID-19).

Two large cohort studies of patients with CLL and COVID-19 have reported an identical overall case fatality rate around 27% [5–8]. The high mortality in CLL is not surprising. Patients with CLL have a significant and complex impairment of both humoral and cellular immunity[9, 10]. In addition, immunosuppression seen at presentation can be exacerbated by CLL-directed therapies [11, 12]. However, the impact of CLL therapy on the outcome of COVID-19 was not identical in previously published studies, probably due to heterogeneous baseline characteristics. Notably, patients receiving chemoimmunotherapy, a highly immunosuppressive treatment, were underrepresented in the previous reports. This study aimed to reassess the risk factors for COVID-19-related death and further elucidate the impact of chemoimmunotherapy and other CLL-directed treatments.

Patients and methods

Inclusion and exclusion criteria and data collection

This multicenter retrospective cohort study included 341 patients with CLL and confirmed SARS-CoV-2 RNA by a reverse transcriptase-polymerase chain reaction (PCR) between March 2020 and May 2021. We selected the enrollment cutoff date considering the major expansion of the reimbursement policy in the Czech Republic. In the second half of 2021, the increased availability of targeted oral inhibitors for CLL treatment in the Czech Republic led to a sharp reduction in the use of chemoimmunotherapy in CLL patients. These were consecutive patients from nine hematological centers cooperating within the Czech CLL Study Group. We also included data from extended follow-up of 103 patients from three Czech centers who participated in ERIC/CAMPUS analysis [7]. CLL diagnosis was based on the International Workshop on CLL criteria [13]. We included either inpatient or outpatient cases of COVID-19. Full patient data were obtained from the CLLEAR registry – the national CLL database of the Czech Republic. Major comorbidities were defined as Cumulative Illness Rating Scale (CIRS) score of ≥ 7[14]. All included patients provided their informed consent before their personal data were captured in the database.

Statistics

We performed statistical analyzes with the R software (version 4.1.1, www.r-project.org). Differences in proportion were determined using Fisher's exact test. Survival curves were constructed using the Kaplan–Meier method and differences in survival compared via the log rank test. Independent predictors of time to event were determined by Cox regression analysis. The receiver operating characteristic (ROC) curve analysis was used to assess the discriminatory power of age cut-off for predicting mortality due to COVID-19. P values < 0.05 were considered statistically significant; all P values were two-sided. The median follow-up was calculated using the reverse Kaplan–Meier method [15].

Results

Full details of baseline demographic and clinical characteristics are presented in Table 1. The median age at the time of COVID-19 diagnosis was 69 years (range 39–92 years), and 237 (70%) were men. Most of the patients (325/341, 95%) had symptomatic course of COVID-19 infection. We recorded the main comorbidities in 70% of the patients, among which arterial hypertension (50%) and diabetes mellitus (22%) were the most frequent. Regarding prognostic characteristics, we observed an unmutated immunoglobulin heavy chain variable region gene (IGHV) in 150/268 (56%) patients. Only 40/316 (13%) were affected by the del (17p) and/or TP53 mutation, which corresponds to the fact that a significant proportion of patients (37%) did not receive any previous therapy for CLL.

Table 1.

Patient demographics and baseline characteristics

Total number of patients	341
Age at CLL diagnosis, median (range)	62 (28–86)
Age at COVID-19 diagnosis, median (range)	69 (39–92)
Male, n (%)	237 (69.5)
Median follow-up since COVID-19 diagnosis, weeks	15.6
Unmutated IGHV, n (%)*	150 (56)
Mutated TP53 and/or deleted 17p†	40 (13)
Obesity, n (%)‡	84 (25)
Hypogammaglobulinemia, n (%)§	97 (41)
Major comorbidities, n (%)	238 (70)
CLL treatment history
Never treated	172 (37)
Previously treated	214 (63)
Receiving treatment for CLL at COVID-19 diagnosis	131 (38)
Chemoimmunotherapy, n (%)	34 (26)
BCL-2 inhibitor, n (%)	24 (18)
PI3K inhibitor, n (%)	8 (6)
BTK inhibitor, n (%)	63 (48)
Other treatment, n (%)	3 (2)

^*IGHV available in 268 pts; ^†TP53 status available in 316 pts; ^‡obesity defined as body mass index (BMI) ≥ 30; ^§data available in 235 pts; abbreviations: n, number of patients; CLL, chronic lymphocytic leukemia; IGHV immunoglobulin heavy chain variable region; TP53, tumour protein p53; BCL-2, B-cell lymphoma 2; BTK, Bruton tyrosine kinase; PI3K, phosphatidyl inositol 3-kinase

Of the 214 (63%) patients ever treated for CLL, 97 (28%) were receiving CLL-directed treatment while diagnosed with COVID-19. The majority of patients (96/130, 74%) underwent treatment based on targeted oral inhibitors; 63 (48%) patients were treated with Bruton tyrosine kinase (BTK) inhibitor; ibrutinib n = 45, zanubrutinib n = 10, and acalabrutinib n = 8. Twenty-four (18%) patients received venetoclax, and 8 (6%) idelalisib with rituximab. Thirty-four patients (26%) were treated with chemoimmunotherapy; bendamustine plus rituximab (BR) n = 21, obinutuzumab plus chlorambucil (G-CLB) n = 4, rituximab plus chlorambucil (R-CLB) n = 1, rituximab plus cyclophosphamide and dexamethasone (RCD) n = 3, fludarabine and cyclophosphamide combined with rituximab (FCR) n = 5 (detailed information is summarized in Table 2). Twenty-six patients did not interrupt CLL-directed therapy despite the diagnosis of COVID-19; 13 patients continued treatment with ibrutinib, 10 with venetoclax ± acalabrutinib, 3 with acalabrutinib, and one remained on zanubrutinib. All 34 patients receiving chemoimmunotherapy at the time of COVID-19 diagnosis discontinued the CLL-directed treatment, and 82% required admission to the hospital. One hundred and thirty-five patients (40%) of the 341 patients with COVID-19 and CLL were managed on the outpatient basis. Among those treated at a hospital, 72 (35%) patients were admitted to the intensive care unit (ICU), 44 (21%) required oxygen therapy with high flow nasal cannula (HFNC) or noninvasive ventilation (NIV), 40 (19%) received invasive mechanical ventilation, and only 30 (15%) patients did not require any oxygen therapy.

Table 2.

Characteristics and outcome of COVID-19 in patients treated with chemoimmunotherapy

Total number of patients	34
Age at CLL diagnosis, median (range)	67 (49–81)
Age at COVID-19 diagnosis, median (range)	72 (52–82)
Male, n (%)	25 (74)
Unmutated IGHV, n (%)*	18 (67)
Mutated TP53 and/or deleted 17p†	0 (0)
Obesity, n (%)‡	11 (32)
Hypogammaglobulinemia, n (%)§	13 (38)
Major comorbidities, n (%)	30 (88)
Chemoimmunotherapy regimen at COVID-19 diagnosis
BR, n (%)	21 (62)
FCR, n (%)	5 (15)
G-Clb, n (%)	4 (12)
RCD, n (%)	3 (9)
R-Clb, n (%)	1 (3)
Vaccination prior to SARS-CoV-2 infection, n (%)	0 (0)
Previous lines of therapy for CLL, median (range)	1 (1–4)
COVID-19 course and outcome
Hospitalization, n (%)	28 (82)
ICU Admission, n (%)	11 (32)
Deaths, n (%)	17 (50)

^*IGHV available in 27 pts; ^†TP53 status available in 30 pts; ^‡obesity defined as body mass index (BMI) ≥ 30; ^§data available in 32 pts; abbreviations: n, number of patients; CLL, chronic lymphocytic leukemia; IGHV immunoglobulin heavy chain variable region; TP53, tumour protein p53; BR, bendamustine and rituximab; FCR, rituximab, fludarabine and cyclophosphamide; G-Clb, obinutuzumab and chlorambucil; RCD, rituximab, cyclophosphamide, and dexamethasone; R-Clb, rituximab, and chlorambucil

The case fatality rate (CFR) was 28% (95/341) for all patients, with 45% in hospitalized patients and 73% in patients admitted to the ICU (Fig. 1). Conversely, the CFR in non-hospitalized patients was only 1.6%. Figure 2 depicts an epidemic curve of confirmed COVID-19 infections in CLL patients in the Czech Republic. In particular, none of the patients who continued Bcl-2 or BTK inhibitor died from COVID-19 infection or its complication. Six patients (2%) developed venous thromboembolism (4/6 pulmonary embolism).

Fig. 1.

Overall survival of patients with confirmed COVID-19

Fig. 2.

Epidemic curve of confirmed COVID-19 cases with CLL in the Czech Republic

Regarding treatments used for COVID-19, 142 patients were treated with steroids, 94 (66%) with hydroxychloroquine in combination with azithromycin or clarithromycin, 43 (30%) with remdesivir, 10 (7%) with bamlanivimab, and 5 (4%) with casirivimab/imdevimab. Twenty-six patients (18%) received convalescent plasma.

The presence of major comorbidities, age over 72, male gender, CLL-directed treatment in history (Fig. 3), CLL-directed treatment at COVID-19 diagnosis (Fig. 4), chemoimunnotherapy, use of steroids for treatment of COVID-19 were associated with increased risk of death (Table 3). There was a trend towards elevated risk of death in patients with unmutated IGHV and TP53 dysfunction. Conversely, we did not identify obesity, defined as body mass index (BMI) ≥ 30, as a risk factor for death. Of note, concurrent therapy with BTKi compared to chemoimmunotherapy was not associated with reduced risk of death (Fig. 5). In the multivariate analysis, we identified the use of steroids for treatment of COVID-19, history of CLL-directed treatment to be risk factors for death.

Fig. 3.

Kaplan–Meier survival curve stratified by CLL treatment status

Fig. 4.

Kaplan–Meier survival curve stratified by concurrent CLL treatment at COVID-19 diagnosis

Table 3.

Predictors of overall survival, univariate and multivariate analysis (Cox regression model)

	Univariate analysis	Multivariate analysis
Variables	HR (95% confidence interval, p value)	HR (95% confidence interval, p value)
Age over 72y at COVID-19 diagnosis	2.69 (1.76–4.09, p < 0.001)	1.62 (0.87–3.00, p = 0.125)
Gender male vs. female	1.80 (1.10–2.95, p = 0.019)	0.91 (0.47–1.74, p = 0.769)
Obesity	0.71 (0.43–1.17, p = 0.178)	0.59 (0.31–1.12, p = 0.105)
Major comorbidities	2.01 (1.20–3.35, p = 0.008)	1.30 (0.62–2.72, p = 0.482)
Hypogammaglobulinemia	1.38 (0.84–2.25, p = 0.208)	–
TP53 deletion/mutation present vs. absent	1.72 (1.00–2.96, p = 0.050)	1.29 (0.65–2.58, p = 0.469)
IGHV unmutated vs. mutated	1.85 (1.13–3.03, p = 0.014)	1.47 (0.80–2.71, p = 0.211)
Ever treated vs. "watch and wait"	3.52 (2.03–6.12, p < 0.001)	3.28 (1.26–8.58, p = 0.015)
Currently treated for CLL at COVID-19 diagnosis	2.60 (1.73–3.92, p < 0.001)	0.67 (0.14–3.07, p = 0.603)
Treatment with BTK inhibitor*	2.02 (1.29–3.16, p = 0.002)	2.59 (0.57–11.71, p = 0.215)
Chemoimmunotherapy*	2.14 (1.27–3.63, p = 0.004)	0.92 (0.23–3.62, p = 0.901)
Treatment with anti-CD20 monoclonal antibody*	2.44 (1.54–3.86, p < 0.001)	1.47 (0.22–9.77, p = 0.693)
Steroids for COVID-19	6.44 (3.62–11.45, p < 0.001)	12.13 (4.71–31.24, p < 0.001)

^*at COVID-19 diagnosis, multivariable Cox model containing all predictors with P < .20 from the univariable models was constructed, with each predictor adjusted for all others in the model. Numbers marked in bold indicate statistically significant results

Fig. 5.

Kaplan–Meier estimates for overall survival according to concurrent treatment modality (BTKi vs. CIT)

In order to understand the changes in the prognosis over time, we compared the outcome of patients diagnosed with COVID-19 during 2020 to a subset of patients diagnosed later in the course of the pandemic. In 2020, the case fatality rate was 32% (52/160), while CFR was 24% in those diagnosed during 2021. Therefore, the risk of death during the later phase of the COVID-19 pandemic was significantly decreased (hazard ratio 0.57; 95% confidence interval, 0.38–0.86; p = 0.005). Although 177 patients received a vaccine against SARS-CoV-2 infection during the follow-up, only four patients were vaccinated before the diagnosis of COVID-19. Consequently, we could not evaluate the efficacy of vaccines against SARS-CoV-2 infection in our cohort.

Discussion

This study presented the outcomes of 341 patients with CLL and COVID-19 reported from nine hematology institutions in the Czech Republic. Our study's case fatality rate (CFR) of 28% is in agreement with findings from two extensive reports on COVID-19 in patients with CLL. Roeker and colleagues reported an international analysis of 374 cases [5]. Of these, 85% were hospitalized, and 32% required ICU admission. The CFR of the entire cohort was 28%, and 36% for those admitted to the hospital. In the largest cohort to date, the European Research Initiative on CLL (ERIC) and Campus CLL analyzed the outcome of 941 cases [7]. In total, 75% of patients were admitted to the hospital, 26% required ICU stay, and the overall CFR was 27% for all patients. While considering these findings, the general population's mortality rate due to COVID-19 infection is markedly lower than that reported in hospitalized CLL patients [16]. The high mortality rate for COVID-19 is not surprising as CLL primarily affects older adults with a median age of 70 years; significant comorbidities are common [3]. Higher age and comorbidities are well-recognized risk factors for dismal outcomes due to COVID-19 infection [17–20]. Further, CLL is associated with hypogammaglobulinemia, T cell abnormalities, impaired complement, and phagocytic system[9, 10]. Specifically, a lack of plasmacytoid dendritic cells, key producers of type I interferon, was proposed as an underlying mechanism for the aggressive clinical course of COVID-19 in CLL [21]. These substantial immune defects result in severe disruptions in defense mechanisms against infections. Moreover, immunosuppression worsens throughout the disease course due to medications used in CLL treatment.

In the present study, prior or concurrent CLL-directed therapy was associated with an increased risk of death. These results are in line with those obtained by ERIC and CAMPUS analysis [7]. Roeker et al., on the other hand, did not report any differences in survival rates in treated or untreated patients [5]. The various baseline characteristics in each study can explain this finding. In the study reported by Chatzikonstantiou et al., a higher proportion of patients (n = 85) underwent CIT during the last 12 months before COVID-19 diagnosis. In addition, 9% of patients were receiving chemoimmunotherapy at the time of diagnosis of COVID-19. On the other hand, in the Roeker study, less than 3% of patients were on CIT. The cause of the differences in treatment representation can be found due to distinct reimbursement policies in many countries [22].

Given the limited access to targeted oral therapy, particularly for first-line CLL treatment in the Czech Republic, we could analyze the impact of CIT on the outcome of SARS-Cov-2 infection in CLL patients [23, 24]. Unsurprisingly, patients receiving CIT at the time of COVID-19 diagnosis were at an increased risk of death. On the other hand, we did not observe any beneficial effect of ibrutinib in this setting.

Corticosteroids seem to reduce all‐cause mortality in severely ill patients with COVID-19 [25, 26]. In contrast, in our analysis and series reported by Roeker, administration of corticosteroids was associated with an increased risk of death [5]. A possible explanation for this finding could be that steroids show inhibitory effects on a broad range of immune responses resulting in a significant risk for infections [27]. However, we cannot rule out the potential for bias, given that only patients in the severe course of COVID-19 received rescue treatment using corticosteroids. Moreover, we did not have detailed information (e.g., dose, length of treatment, etc.) on corticosteroid use. Therefore, a larger-scale and comprehensive prospective study is required to address the role of corticosteroids in CLL patients with COVID-19. Without data from such a study, steroids' impact has to be interpreted very cautiously.

Our study is not without potential limitations. First, asymptomatic or mildly symptomatic patients may be underrepresented, given the retrospective study design. Second, predominantly tertiary centers with a selected patient population collaborated on the study, which may have led to selecting cases with advanced disease and after repeated treatments. These patients are prone to severe immunosuppression with an unfavourable course of COVID-19. Third, though the patient's heterogeneity seems to be lower in our study compared to the aforementioned large-scale retrospective sets, the approach to COVID-19 was not unified among the participating centers. Finally, given the chosen recruitment period, we could not assess the outcome of CLL patients during the recent omicron spread.

Conclusions

In summary, in this large retrospective cohort study on COVID-19 in patients with chronic lymphocytic leukemia, we report 28% case fatality rates and identify age, and treatment status as the most significant risk factors for dismal outcomes. Of interest, patients receiving chemotherapy during the diagnosis of COVID-19 did not show deteriorating results compared to ibrutinib therapy.

Authors' contributions

MŠi participated in the study design, data analysis, and wrote the manuscript. PT assisted in the study design, data analysis and contributed to writing and/or commenting on the manuscript. All other authors contributed to writing and/or commenting on the manuscript.

Funding

This work was supported by DRO (UHHK, 00179906) and the Cooperatio Program, research area ONCO.

Data Availability

The data supporting our study will be available by contacting the corresponding author for reasonable reasons.

Declarations

Consent to participate / ethical approval statement

Full patient data was obtained from the CLLEAR registry, the Czech Republic's epidemiological and clinical database of CLL. All patients gave their informed consent before their personal data was captured in the CLLEAR database.