Comparison of Treatment-Emergent Adverse Events of Covalent BTK Inhibitors in Clinical Trials in B-Cell Malignancies: A Systematic Review and Meta-Analysis

Background: Bruton's tyrosine kinase inhibitors (BTKi's) have revolutionized the therapeutic landscape of B-cell malignancies such as chronic lymphocytic leukemia/small lymphocytic leukemia (CLL/SLL), mantle cell lymphoma (MCL), and Waldenstrom macroglobulinemia (WM). Ibrutinib was the first-in-class and first FDA-approved BTKi. Its off-target effects have resulted in several adverse events (AEs) of concern, such as atrial fibrillation, hypertension, hemorrhage, and diarrhea. Acalabrutinib and zanubrutinib are second-generation, more selective BTKi's that were developed with the goal of minimizing off-target effects. While multiple trials have been conducted for these BTKi's in B-cell malignancies, direct comparison of toxicity profiles between BTKi's has been limited. We sought to systematically analyze and compare treatment-emergent AEs reported with the three BTKi's in clinical trials. Methods: PubMed and major hematology/oncology conference abstracts were searched (last updated query on July 11, 2022) for trials of ibrutinib, acalabrutinib, and zanubrutinib in B-cell malignancies. AEs of clinical interest or AEs reported by ≥10% of the included trials were analyzed in this study. Bayesian multilevel regression models were applied to estimate incidence probabilities of AEs and odds ratios (ORs) between treatments. Partial information contained in the censored data was incorporated into the likelihood function using cumulative probabilities of binomial distribution. Results: A total of 61 trials were included involving 6458 patients and 68 treatment arms: ibrutinib (n=30; 44%), ibrutinib plus CD20-mAb (n=14; 21%), acalabrutinib (n=11; 16%), acalabrutinib plus CD20-mAb (n=2; 3%), zanubrutinib (n=10; 15%), and zanubrutinib plus CD20-mAb (n=1; 1%). Most trials were in CLL/SLL (n=32), MCL (n=10), or WM (n=9); 38 trials were relapsed/refractory setting and 14 in frontline. Three trials involved randomized comparison between different BTKi's (ASPEN, ELEVATE-RR and ALPINE). A total of 65 AEs were analyzed. The most common all grade AEs included infection (62.2%), hemorrhage (41.5%), diarrhea (34.3%), fatigue (22.8%), neutropenia (21.6%), and upper respiratory infection (URI, 21.0%). The most common grade ≥3 AEs included neutropenia (14.9%) and infection (12.3%). The incidences of all grade/grade ≥3 atrial fibrillation were 6.0%/2.5%, and for hypertension were 12.5%/7.8%. Mean incidences of AEs by drug type, follow-up time, and frontline versus relapsed setting are shown in Figure 1. Comparison of AEs between ibrutinib-based and acalabrutinib/zanubrutinib-based regimens are shown in Figure 2. Select all grade AEs that occurred more frequently with ibrutinib included atrial fibrillation (acalabrutinib/zanubrutinib vs ibrutinib OR 0.30), hypertension (OR 0.47), anemia (OR 0.67), thrombocytopenia (OR 0.57), nausea (OR 0.67), vomiting (OR 0.77), and diarrhea (OR 0.49). Grade ≥3 AEs that occurred more frequently with ibrutinib included atrial fibrillation (OR 0.29), hypertension (OR 0.37), diarrhea (OR 0.52), and rash (OR 0.30). All grade AEs that occurred more frequently with acalabrutinib/zanubrutinib-based regimens included contusions (OR 1.50), hematuria (OR 1.96), leukopenia (OR 2.93), second primary malignancies (OR 1.53), URI (OR 1.29), and headaches (OR 2.03). Grade ≥3 AEs that occurred more frequently with acalabrutinib/zanubrutinib-based regimens included leukopenia (OR 2.08), URI (OR 1.74), and pneumonia (OR 1.35). Conclusions: This is the first comprehensive and the largest meta-analysis of AEs reported with covalent BTKi's. Our results add substantial evidence regarding the AE profiles of BTKi's in B-cell malignancies and the comparative toxicity profiles between first- and second-generation BTKi's. Atrial fibrillation, hypertension, and diarrhea were more common with ibrutinib, providing additional evidence supporting observations made in the small number of head-to-head randomized trials. Additionally, several bleeding and infection events were more common with acalabrutinib/zanubrutinib, an observation that requires validation and clinical attention. This meta-analysis is limited by, although attempted to adjust for, reporting bias (eg, different threshold of AE reporting across trials), which necessitates caution in interpreting these exploratory data. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal