Findings from a real-world study demonstrated that zanubrutinib (Brukinsa) monotherapy was associated with significantly longer time to next treatment (TTNT) and overall survival (OS) than acalabrutinib (Calquence) monotherapy in patients with relapsed/refractory mantle cell lymphoma (R/R MCL) who received one of the covalent BTK inhibitors in the second line or later.1
Findings from this retrospective analysis of US administrative claims data, presented at the
Landmark TTNT rates favored zanubrutinib vs acalabrutinib across all time points, with a higher probability of patients not advancing to the next treatment at 12 months (66.3% vs 60.6%), 24 months (52.8% vs 46.7%), and 36 months (42.8% vs 37.5%).
Additionally, the median OS was not reached (NR; 95% CI, 56.0-NR) with zanubrutinib vs 60.6 months (95% CI, 52.5-81.9) with acalabrutinib, translating to a 26% reduction in the risk of death in the unadjusted analysis (HR, 0.74; 95% CI, 0.62-0.88; P < .001) and a 19% reduction in the IPTW-adjusted analysis (HR, 0.81; 95% CI, 0.68-0.98; P = .03).
Landmark OS rates showed a higher probability of survival at 12 months (86.4% vs 81.6%), 24 months (76.3% vs 68.6%), and 36 months (70.4% vs 62.6%) among patients receiving zanubrutinib vs those given acalabrutinib.
“These findings provide robust real-world comparative evidence among a large cohort of commercially insured patients in the US on the clinical effectiveness of zanubrutinib and complement emerging real-world data demonstrating clinically meaningful differences among covalent BTK inhibitors,” lead study author Yucai Wang, MD, PhD, of Mayo Clinic in Rochester, Minnesota, and colleagues wrote in a poster presentation of the data. “Real-world evidence provides a novel approach to compare the clinical outcomes when comparisons do not exist in clinical trials. However, unmeasured confounding factors and lack of randomization should caution interpretation and generalizability to other populations.”
In October 2017, the
In November 2019,
The retrospective, observational analysis drew on claims data from the Komodo Health database, a large US administrative claims database capturing longitudinal real-world data.1 Eligible patients were at least 18 years of age with a diagnosis of MCL who initiated zanubrutinib or acalabrutinib monotherapy in the second-line setting or later between January 2015 and August 2025. Those who participated in clinical trials, had end-stage renal disease, or had undergone a prior stem cell transplant were excluded.
From an initial pool of 53,474 patients with a diagnosis of R/R MCL, the final cohorts comprised 931 patients who received zanubrutinib and 1288 who received acalabrutinib.
TTNT was defined as the time from initiation of the BTK inhibitor to the initiation of next treatment or death. OS was defined as the time from initiation of BTK inhibitor therapy to death. Kaplan-Meier estimates and the log-rank test were used to assess and compare outcomes, and Cox proportional hazards models were used to estimate unadjusted and adjusted hazard ratios. The IPTW-adjusted model accounted for age, sex, US region, treatment initiation year, and Charlson Comorbidity Index (CCI).
The median age at the index date was 73 years in the zanubrutinib arm and 72 years in the acalabrutinib arm. Patients were predominantly male (71% vs 75%, respectively) and White (73% vs 74%). The South US region accounted for the largest proportion of patients in both groups (37% vs 32%). The mean CCI was 3.5 (SD, 3.2) and 3.6 (SD, 3.1), respectively. The median follow-up was 14.8 months for the zanubrutinib arm vs 18.2 months for the acalabrutinib arm.
Wang and colleagues noted that the follow-up period was imbalanced between the 2 groups, reflecting differing regulatory milestones, staggered clinical adoption, and enrollment timing. The analysis attempted to mitigate this by adjusting for treatment initiation year in the adjusted models.
They also explained that, as with any claims-based analysis, the potential for miscoding or delayed coding may have led to underidentification or misclassification of clinical events. Additionally, mortality data derived from multiple sources may be affected by delays in capture or misreporting.
The study authors also cited residual confounding and selection bias stemming from a lack of information on key clinical variables of disease aggressiveness, such as TP53 mutation status, Ki-67 index, and physician treatment-selection patterns, cautioning that unmeasured confounding and the absence of randomization should temper interpretation and generalizability.
