Venetoclax or pirtobrutinib in relapsed/refractory Waldenström macroglobulinemia: Clinical and molecular predictors and sequencing implications

Abstract

Abstract Introduction: Venetoclax and pirtobrutinib are two novel targeted agents used for the treatment of relapsed or refractory Waldenström macroglobulinemia (WM). However, guidance on the comparative efficacy and predictors of response for each agent remains limited, particularly outside of clinical trials. Methods We retrospectively analyzed all WM patients treated with single-agent venetoclax or pirtobrutinib at our institution, excluding those treated for Bing-Neel syndrome. Responses were assessed per the IWWM-11 criteria. Each cohort was analyzed independently for factors associated with major response rate (MRR; partial response or better) and progression-free survival (PFS). Comparative efficacy was assessed in all patients (unmatched) and in a 1:1 matched cohort, with matching based on prior lines of therapy, progression on a covalent BTK inhibitor (cBTKi), and TP53 mutational status. Results 91 patients were included: 64 received venetoclax (V), 27 pirtobrutinib (P), including 11 who received both agents sequentially. Median age was 69 years (range 39-88) in V and 71 years (55-85) in P (p=0.39). Female sex (30% [19/64] vs. 26% [7/27]; p=0.72) and CXCR4 mutational status (46% [26/57] vs. 48% [11/23]; p=0.86) were evenly distributed. TP53 mutations were less frequent in V (25% [11/44] vs. 50% [9/18]; p=0.06). Prior progression on cBTKi was less common in V (38% [24/64] vs. 81% [22/27]; p<0.001), and V patients had fewer prior treatments (median 3 [range 1-7] vs. 4 [1-13]; p=0.22). A total of 18/64 (28%) in V and 5/27 (19%) in P received therapy within a clinical trial (p=0.34). Median follow-up was 36 months for V and 18 months for P. Among 31 patients who transitioned from a BTKi to V or P within one week: 12 received venetoclax without overlap, of whom 8 (67%) experienced IgM rebound (not classified as progression unless continued IgM increase >3 months into therapy); 4 received venetoclax with BTKi overlap, without IgM rebound; and 15 transitioned from cBTKi to pirtobrutinib without overlap, without IgM rebound. In the venetoclax cohort (n=64), MRR was lower in patients with TP53 mutations (82% vs. 45%; p=0.02). In multivariate analysis adjusting for CXCR4 status, ≥3 prior lines of therapy, prior cBTKi progression, maximum (≥400mg), and trial inclusion, TP53 mutations (OR 0.16; p=0.05) and receiving venetoclax outside of a clinical trial (OR 0.05; p=0.05) were associated with lower odds of MRR. Survival analysis showed shorter median PFS among patients with prior progression on a cBTKi (12 vs. 36 mo; p=0.003), TP53 mutations (10 vs. 36 mo; p<0.001), ≥3 prior lines of therapy (17 vs. 42 mo; p=0.002), and those not enrolled in a clinical trial (29 vs. 40 mo; p=0.03). On multivariate analysis adjusting for age (≥65), ≥3 prior lines, cBTKi progression, and trial inclusion, TP53 was the only factor associated with PFS (HR 4.2; p=0.02). In the pirtobrutinib cohort (n=27), MRR was lower among patients with CXCR4 mutations (40% vs. 91%; p=0.013). In multivariate analysis adjusting for trial inclusion and ≥3 prior lines, CXCR4 mutations remained associated with lower response (OR 0.08; p=0.05). CXCR4 mutations predicted shorter PFS (8.3 months vs. not reached; p=0.022), significant after adjustment for age (≥65) and trial inclusion (HR 5.9; p=0.04). TP53 status, prior venetoclax exposure, trial participation, prior cBTKi progression, and number of prior lines were not significantly associated with MRR or PFS. Venetoclax and pirtobrutinib showed comparable MRR (69% vs. 64%; p=0.63) and no statistically significant difference in median PFS, despite a numerical difference favoring venetoclax (32 vs. 16 months; p=0.16). In the 1:1 matched cohort, response rates remained equivalent (OR 0.96, p=0.93), and no difference in PFS was observed (HR 1.1, p=0.85). Conclusion Venetoclax and pirtobrutinib demonstrate comparable activity in relapsed or refractory WM, each with distinct limitations. TP53 and CXCR4 mutations predicted inferior outcomes to venetoclax and pirtobrutinib, respectively, suggesting non-overlapping molecular vulnerabilities. Venetoclax post-cBTKi was limited by reduced efficacy and IgM rebound. A prospective phase II study combining P and V as a fixed-duration regimen is underway (NCT05734495).