BTK Inhibitors Versus Venetoclax as First- or Second-Line Therapy in Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma: A Real-World Evidence Study

Abstract

Targeted therapies (e.g., Bruton tyrosine kinase inhibitors [BTKi; ibrutinib, acalabrutinib, zanubrutinib] or B-cell lymphoma-2 inhibitors [BCL-2i; venetoclax]) with or without anti-CD20 monoclonal antibodies (CD20 mAb), have demonstrated consistent survival benefits versus chemoimmunotherapy (CIT) in patients with chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) in the front-line (1L) and relapsed/refractory (R/R) settings (summarized in Eichhorst et al. [1]). However, there are currently no prospective data comparing BTKi-based and venetoclax-based therapies directly, and information to guide optimal sequencing strategies for targeted agents is limited [2-4]. An improved understanding of the comparative effectiveness of BTKi and BCL-2i in both 1L and R/R settings, in novel agent–naive and exposed patients, is needed to guide treatment sequencing decisions.

To better understand the performance of targeted agents, we conducted a retrospective, observational cohort analysis to evaluate clinical outcomes of comparable patients with CLL/SLL who received either BTKi monotherapy or BCL-2i+CD20 mAb in 1L or second-line (2L) settings.

This study used Flatiron Health electronic health record-derived de-identified data from August 1, 2014 to February 28, 2022. Adult patients (≥ 18 years) with CLL/SLL were included if they received BTKi monotherapy (acalabrutinib or ibrutinib) or BCL-2i (venetoclax) in combination with a CD20 mAb in the 1L or 2L setting and had 2 or more clinical encounters during the study period. The study compared time-to-next-treatment-or-death (TTNTD; defined as time from initiation of the current treatment to the starting time of a new line of therapy or death) for patients treated with BTKi monotherapy versus BCL-2i+obinutuzumab (1L), or BTKi monotherapy versus BCL-2i+CD20 mAb (ofatumumab, rituximab, or obinutuzumab) (2L).

Kaplan–Meier analysis was used to obtain unadjusted TTNTD curves for each of the cohorts. TTNTD was descriptively compared between cohorts using a Cox proportional-hazards model to estimate adjusted hazard ratios (HR). Propensity score matching (PSM) was used to improve the comparability between cohorts and reduce the effects of confounding. Each patient treated with BCL-2i+obinutuzumab (1L) or BCL-2i+CD20 mAb (2L) was matched with two patients treated with BTKi monotherapy in each setting. PSM balanced key demographic and disease characteristics: age, gender, race, Rai stage, time from diagnosis to index date, del(17p), del(11q), and immunoglobulin heavy chain variable region (IGHV) mutational status, Eastern Cooperative Oncology Group performance status, total number of therapy lines received during the study period, and prior novel agent exposure. Inverse probability of treatment weighting (IPTW) was used as an alternative to PSM to control for potential confounding and reduce noncomparability between cohorts (Supporting Information Methods).

A total of 14 602 CLL/SLL patients were included, among whom 6065 unique patients had 6743 treatment episodes that included BTKi monotherapy or BCL2i+CD20 mAb; some patients received both treatments. A total of 3643 patients in the 1L setting and 1772 patients in the 2L setting met the inclusion criteria (Figure S1). In the 1L setting, 93.2% (n = 3397) received BTKi monotherapy and 6.8% (n = 246) received BCL-2i+obinutuzumab. In the 2L setting, 84.4% (n = 1496) received BTKi monotherapy and 15.6% (n = 276) received BCL-2i+CD20 mAb.

After applying PSM, the two cohorts were well balanced on all selected matched baseline and disease characteristics in 1L (n = 738) (Table S1) and 2L (n = 828) settings (Table S2). In the 1L setting, the median age across cohorts was 67 years (interquartile range [IQR] 59.0–74.0), and 26.8% were female. In the 2L setting, the median age across cohorts was 70 years (IQR 63.0–77.0), 32.9% were female, and 32.6% had prior treatment with novel agents. In the 2L BTKi-treated cohort, 172 (31.2%) had received a prior BTKi, and 6 (1.1%) had received prior BCL-2i-based therapy. Among the 163 patients in the 2L BTKi-treated cohort who received 1L ibrutinib monotherapy, 158 (96.9%) received acalabrutinib in the 2L setting. In the BCL-2i-treated cohort, 75 (27.2%) had received prior BTKi, and 24 (8.7%) had received prior BCL-2i-based therapy.

Median TTNTD was not reached for the BTKi or BCL-2i cohorts in 1L or 2L settings following PSM analysis. The mean proportion of patients remaining on treatment or in treatment-free observation at 24 months in the 1L setting was 95.6% (95% CI 92.8–97.3) for the BTKi cohort and 92.7% (95% CI 84.8–96.6) for the BCL-2i+obinutuzumab cohort (Figure 1A); the corresponding proportions in the 2L setting were 84.9% (95% CI 81.0–88.1) for the BTKi cohort and 80.4% (95% CI 73.6–85.6) for the BCL-2i+CD20 mAb cohort (Figure 1B). Overall, the BTKi cohort had a significantly longer TTNTD versus the BCL-2i+CD20 mAb cohort in the 2L setting (log-rank p = 0.0071). A statistically significant difference was not observed in the 1L setting (log-rank p = 0.2608). Similar trends were observed when using an IPTW approach (Supporting Information Results, Tables S3 and S4, Figure S2).

This is the first time, and in the largest dataset to date, that a longer TTNTD was demonstrated with BTKi monotherapy versus BCL-2i+CD20 mAb in the 2L setting. The results did not change when utilizing different adjustment algorithms and key matching variables (PSM and IPTW).

BTKi and BCL-2i (with or without anti-CD20 mAb) have consistently demonstrated clinical benefit versus CIT in 1L (treatment-naive) and R/R settings [1]. In this analysis, both treatments were highly effective in the 1L setting, with similar TTNTD observed with continuous BTKi monotherapy versus BCL-2i+obinutuzumab. Other studies using indirect-comparison methods have not demonstrated significant differences between these classes of agents among patients with CLL treated in the 1L setting [5, 6]; therefore, the current approach to therapy selection relies largely on disease biology, patient-related factors, and patient preference. These findings emphasize the need for prospective head-to-head trials to guide treatment selection and support these current practices.

In the 2L setting, TTNTD was significantly improved with BTKi monotherapy versus BCL-2i+CD20 mAb regimens. Notably, this population included patients with (32.6%) and without (67.4%) prior novel agent exposure. The significant finding in only the 2L setting may be attributed to differences in patient characteristics in the 2L setting versus 1L, specifically cytogenetic profile and treatment history. While PSM was used to adjust for these baseline characteristics, resulting in statistically insignificant differences between cohorts, those in the 2L setting had a higher rate of del(17p) (13.3% vs. 9.6%) and del(11q) (17.5% vs. 14.1%), but a slightly lower rate of unmutated IGHV (34.5% vs. 36.3%) versus 1L patients. Additionally, a slightly higher proportion of patients in the BCL-2i+CD20 mAb cohort had received prior novel agents versus the BTKi monotherapy cohort (35.9% vs. 31.0%, respectively). Among those who received a novel agent as 1L treatment, BTKi-based regimens predominated for both cohorts; BTKi was included in 92.5% and 75.8% of 1L novel agent regimens in the BTKi monotherapy and BCL-2i+CD20 mAb cohorts, respectively. These real-world data suggest that transitioning from 1L BTKi to 2L BTKi therapy can be an appropriate choice when needed given intolerance, patient preference, or insurance decision. A notable difference in prior use of venetoclax-based regimens was also observed between the BCL-2i+CD20 mAb cohort and the BTKi monotherapy cohort (8.7% vs. 1.1% of all cohort patients), which may have impacted outcomes in the 2L setting (i.e., potential for a less durable response with repeated BCL-2i-based therapy vs. 1L BCL-2i-based therapy followed by 2L BTKi).

Although clinical trials have established the efficacy of BTKi and BCL-2i+CD20 mAb therapies versus CIT, data from prospective head-to-head studies comparing BTKi therapy and BCL-2i-based regimens for CLL/SLL are currently unavailable. Thus, retrospective analyses can offer insight on expected outcomes with these two drug classes. However, findings from this retrospective analysis does not obviate the need for prospective clinical trials. As with all studies that utilize electronic medical records, data are limited by the possibility of inaccuracy or incompleteness. Missing information was frequently noted, including IGHV status for approximately 40% of the PSM 1L cohort, which limited the ability to describe patients at the time of treatment initiation and may have impacted treatment selection. Additionally, reasons for initial treatment discontinuation (e.g., disease progression or intolerability to 1L therapy) were not available though they would not be expected to bias the efficacy endpoint given that the analysis focused on the drug class. While PSM weighting was used to balance cohorts, unmeasured confounding may remain. Generalizability of the results to patients outside of the Flatiron Health database may be limited. At the time of analysis, there was relatively limited follow-up for the cohorts; therefore, the study lacked data maturity to assess overall survival. Finally, the sample size of the BTKi cohort was much larger than the BCL-2i-based cohorts; thus, oversampling could potentially introduce bias.

In conclusion, this retrospective analysis demonstrated a longer TTNTD among patients with CLL/SLL treated with BTKi monotherapy versus BCL-2i+CD20 mAb in the 2L setting. These results may have implications for treatment selection in current clinical practice and inform the results of ongoing randomized studies. While benefit in the 2L setting was observed with BTKi monotherapy, studies evaluating BTKi combinations may demonstrate additional benefits in the 1L setting with the potential for time-limited therapy versus continuous therapy.

Study design: Anna Teschemaker, Anthony R. Mato, and Lindsey E. Roeker. Study investigator: All authors. Enrolled patients: n/a. Data analysis: Yi Han. Data interpretation: All authors. Manuscript review and revisions: All authors. Final approval of manuscript: All authors.

The authors have nothing to report; US nationwide electronic health record-derived de-identified database.

The authors have nothing to report; US nationwide electronic health record-derived de-identified database.

L.E.R. has served as a consultant for AbbVie, Ascentage, AstraZeneca, BeiGene, Janssen, Loxo Oncology, Pharmacyclics, Pfizer, and TG Therapeutics; is a member of a data safety monitoring committee (DSMC) for Ascentage; served as a CME speaker for DAVA, Curio, Medscape, and PeerView; holds a minority ownership interest in Abbott Laboratories; received travel support from Loxo Oncology; and received research funding (paid to the institution) from Adaptive Biotechnologies, AstraZeneca, Genentech, AbbVie, Pfizer, Loxo Oncology, Aptose Biosciences, Dren Bio, and Qilu Puget Sound Biotherapeutics. Y.H. is employed by Vinzent Strategies, which received funding from AstraZeneca for statistical analysis services. A.T. and A.R.M. are employees of and own stock in AstraZeneca. M.C.T. has served on consultancy/advisory boards for AbbVie, AstraZeneca, BeiGene, Janssen, and Loxo Oncology; received research funding (paid to the institution) from AbbVie, AstraZeneca, BeiGene, GenMab, Nurix Therapeutics, and Genentech; received travel support from GenMab, Nurix Therapeutics, and Dava Oncology; and received honoraria from PeerView Medical Institute, Dava Oncology, Philips Group Oncology Communications, MJH Life Sciences, Intellisphere LLC, Clinical Care Options, and Mashup Media.