Severe Infections in Patients with Chronic Lymphocytic Leukemia Treated with (Immuno-)Chemotherapy: A Pooled Analysis of Gcllsg Trials

INTRODUCTION

Chronic lymphocytic leukemia (CLL) is frequently associated with an impaired humoral and cellular immunity. On a global scale chemoimmunotherapy (CIT) has remained one of the most frequently used treatment option. Thus, patients (pts) may experience further cytopenia, particularly treatment-related neutropenia, increasing the risk of infections. In order to better characterize incidence, characteristics and outcomes of infections during and after therapy, a pooled analysis of phase II and III German CLL Study Group trials was performed.

METHODS

Data of first line pts from 5 clinical trials (CLL7, pts treated with fludarabine, cyclophosphamide, rituximab [FCR]; CLL8, FC vs FCR; CLL10, FCR vs bendamustine-rituximab [BR]; CLL11, chlorambucil [CLB] vs CLB-R vs CLB-Obinutuzumab [CLB-Ob] and CLL2M, BR) were analyzed. Clinical, laboratory, genetic and event-related data were pooled. Infections defined as severe (CTC grade 3-5) from initiation of therapy until 4 weeks after completion of study treatment were considered related. Due to varying reporting periods for infections of the respective trials later events of infections were not included. Kaplan-Meier curves for landmark overall survival (OS) from completion of study treatment plus 4 weeks were plotted and compared by non-stratified log-rank test. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated using Cox proportional-hazard regression modelling.

RESULTS

Data from 2,291 pts receiving at least one dose of CIT were pooled. Median observation time was 71.7 months, ranging between 43.7 months (CLL2M) and 81.0 months (CLL10). Seven-hundred and twenty-seven pts received FCR, 396 pts FC, 395 pts BR, 116 pts CLB, 326 pts CLB-R and 331 pts CLB-Ob.

Overall, 274 severe grad 3/4/5 infections were reported in 229 pts (10.0% of 2,291 pts). Of those 189 pts (82.5%) had max. grade 3 infections, 22 (9.6%) pts had grade 4 infections and 18 (7.9%) pts died due to infectious complications. Median time to severe infection from start of treatment was 1.8 months (IQR 0.9-3.6), with a median number of infectious episodes per patient of 1 (range 1-4).

Thirty-one (13.5%) of 229 pts had bacterial infections, 35 (15.3%) viral infections, 5 (2.2%) fungal infections and 172 (75.1%) unspecified infections. Higher grade (i.e. ≥ CTC grade 3) leukopenia and/or neutropenia was reported in 121 (52.8%) pts with severe infections. Eighty-eight (12.1%) of FCR treated pts had severe infections, followed by BR 45 (11.4%), CLB 12 (10.3%), FC 35 (8.8%), CLB-Ob 25 (7.6%) and CLB-R 24 (7.4%).

Median age was 64 years in the entire cohort; no differences between pts with and without infections were observed with regards to age, sex, ECOG or creatinine clearance. Molecular and cytogenetic characteristics (deletion 17p, deletion 11q, trisomy 12) and IGHV status were similarly distributed between both groups. Median neutrophil count at enrolment was 4.4x10^-9/l in both groups, respectively. Prior to therapy, levels of immunoglobulin were comparable between pts with and without infections (median IgG 7.0 vs 7.5 g/L, IgM 0.3 g/L vs 0.3 g/L). Also, pts with at least one episode of ≥ CTC grade 3 leukopenia/neutropenia had comparable rates of severe infections to pts without higher grade leukopenia/neutropenia (147 [53.6%] vs 127 [46.4%] pts).

No differences were observed between pts with or without infections regarding the response to first line treatment (183 pts [79.9%] with complete response or partial response to treatment vs 1715 pts [83.2%]) as well as the rate of undetectable minimal residual disease levels (50 [21.8%] vs 477 [23.1%]).

Overall survival from 4 weeks after completion of study treatment was significantly shorter in pts with severe infections compared to pts without severe infections (median 73.7 months vs 97.3 months, HR 1.503, 95% CI 1.217-1.856, p < 0.001).

CONCLUSION

This analysis confirms that prognosis of CLL pts who received first line treatment with (immuno)chemotherapy is influenced by severe infections. This risk does not correlate with the explored cyto- or molecular genetic risk factors, nor with response to treatment, pre-therapeutic levels of immunoglobulins or occurrence of higher grade neutropenia. Pts who experience severe infections have a significantly shorter overall survival compared to pts without severe infections. Due to their vulnerability, careful management of infectious complications in CLL pts is warranted.