Introduction: Zanubrutinib (zanu) received accelerated U.S. Food and Drug Administration (FDA) approval for patients (pts) with relapsed or refractory (R/R) marginal zone lymphoma (MZL) based on the results of the pivotal phase 2 study (n=68), which showed an overall response rate (ORR) of 68.2%, complete response rate (CRR) of 25.8%, and 2-year progression-free survival (PFS) of 70.9% (Opat S, et al. Clin Can Res 2021). However, there is a paucity of real-world data with zanu in MZL. Furthermore, factors associated with ORR and primary progression (PP) on zanu are unknown, although the presence of MYD88 and TNFAIP3 mutations was associated with longer PFS in a subset of 18 pts from the MAGNOLIA trial (Tatarczuch et al., Blood Adv 2023). Hence, we sought to evaluate real-world outcomes with zanu in MZL including clinicopathologic characteristics predictive of treatment failure.

Methods: This multicenter retrospective cohort study included adults with MZL treated with zanu at 10 U.S. academic centers following the FDA approval of zanu in the U.S. The primary objective was to evaluate the real-world efficacy outcomes of zanu including ORR and CRR, PFS, and overall survival (OS). Secondary objectives included the evaluation of factors predictive of CRR, PP (defined as disease progression as best response to zanu), and factors prognostic for PFS. Response rates were assessed using Lugano criteria.

Results: The study included 78 pts. Median age was 66 years with 54% male. MZL subtypes included 33% splenic, 32% extranodal, 28% nodal, and 6%, unspecified MZL. Characteristics at diagnosis included: stage 3-4 MZL in 79%, elevated serum lactate dehydrogenase (LDH) in 32% (n=73 tested), MYD88 mutation in 12% (n=38 tested), and Ki-67>20% in 32% (n=46 tested). Zanu was administered as 1L treatment (tx) in14%, 2L tx in 36%, 3L tx in 30%, and 4L+ in 20% of pts.

Among the 68 pts evaluable for response, the best ORR and CRR were 75% and 44%, respectively, without significant differences across MZL subtypes (P=0.90 and 0.51, respectively). When evaluated by line of therapy (LOT), ORR/CRR in 1L vs R/R MZL were 75%/50% and 75%/43%, respectively. All pts with PP received zanu in R/R setting. Two clinicopathologic characteristics were significantly associated with PP on zanu: Ki-67 >20% (75% of PP pts vs. 14% of responders; p=0.02) and shorter duration of response to 1L tx (median 4.6 vs. 17.8 months; p=0.029). Among 38 pts tested for MYD88, presence of MYD88 mutation (38% vs. 6% of pts tested; p=0.035) was the only feature significantly associated with CRR to zanu.

The median follow up was 18 months. The median PFS was 45 months (95%CI, 29 to not reached [NR]), while 2-year PFS was 71.7% (95%CI, 57.4-81.9). There was no difference in PFS based on the zanu LOT (log-rank p=0.27). Among the clinicopathologic factors evaluated, only Ki-67 >20% was prognostic of inferior PFS (HR=8.86, 95%CI=2.35-33.39, p=0.001). Presence of MYD88 mutation was not associated with PFS (log-rank p=0.52). The median OS was not reached. The 2-year OS estimate was 96.1% (95%CI, 84.2–99.1).

The 1-year OS post zanu relapse/progression was 87% (95%CI, 55.2–96.6), and it did not significantly differ based on PP versus secondary progression (pts who initially responded but subsequently progressed; log-rank p=0.099).

Nine pts (12%) required zanu dose reduction with the most common causes being fatigue (n=2), and arthralgias (n=2). Ten pts (13%) discontinued zanu for toxicity with 3 related to cardiac adverse events (atrial fibrillation; palpitations; hypertension) and 2 related to bleeding (upper gastrointestinal; retinal).

Conclusions: In this first assessment of outcomes of pts with MZL treated with zanu in the real-world setting, we observed slightly higher response rates, but very similar PFS estimates compared to the clinical trial. Zanu demonstrated consistent efficacy regardless of line of use. Given the association of high Ki-67 (>20%) with significant risk for PP, consideration should be given to check Ki-67 on the biopsy samples when feasible.

This content is only available as a PDF.
2025
Sign in via your Institution