Dear Editor,

Since receiving regulatory approval in 2015, oral Bruton tyrosine kinase (BTK) inhibitors have revolutionized the treatment of Waldenström macroglobulinemia (WM) in both the frontline and relapsed/refractory settings [1, 2]. Covalent BTK inhibitors (BTKis), including ibrutinib, acalabrutinib, and zanubrutinib, have demonstrated robust response rates and prolonged progression-free survival, especially for patients with MYD88MUT WM [3]. However, BTKi treatment duration can be limited by adverse events (AEs), such as bleeding and cardiac toxicities, as well as drug resistance [4, 5]. Although long-term follow-up has been presented from some trials, there is a dearth of real-world data examining the outcomes of patients post-BTKi therapy [6, 7]. The objectives of this study were to evaluate the outcomes of patients with WM who previously received BTKi therapy (BTKi-discontinued), looking at their response to, and duration of, therapy, reasons for BTKi discontinuation, next-line treatment patterns, and associated outcomes. Furthermore, we sought to compare their outcomes with those of patients who remained on BTKi therapy (BTKi-active) at the time of data cutoff.

This retrospective study evaluated all patients with WM who were treated with a covalent BTKi at MD Anderson Cancer Center (MDACC) between January 2014 and March 2024. Both treatment naïve (TN) and relapsed/refractory (R/R) patients were included, and outcomes were evaluated relative to the first BTKi therapy utilized. All included patients had received at least 1 cycle (28 days) of BTKi. The VIth International Workshop on Waldenström Macroglobulinemia (IWWM) consensus response criteria were used to assess clinical response, as the majority of patients had been treated prior to the establishment of the IWWM-11 criteria [8]. This study received approval from MDACC’s Institutional Review Board (IRB).

A total of 150 patients (median age: 63 years; 38% frontline, 62% R/R) were treated with a covalent BTKi (n = 113 - 75% ibrutinib; n = 26 - 17% zanubrutinib; n = 11 - 7% acalabrutinib) during the study timeframe. With a median follow-up of 67 months (95% CI: 55-78 months) from the start of BTKi therapy, 74% of patients (111/150 patients) had discontinued taking their first BTKi by data cutoff. More patients had received ibrutinib (87% vs. 44%, p < 0.001) in the BTKi-discontinued compared with the BTKi-active group; however, two-year discontinuation rates were similar among all three BTKi agents (p = 0.86). Comparing agents, patients treated with ibrutinib (78 months) had a longer median follow-up than those treated with zanubrutinib (28 months) or acalabrutinib (33 months) (p < 0.001). Baseline demographics, IPSS-WM score and disease-related factors are included in Table 1 [9]. The majority of characteristics were comparable (p > 0.05) between patients who discontinued vs. remained on active treatment. Notably, BTKi-active patients had a higher median IgM level at the time of BTKi initiation (3133 vs 2027 mg/dL, p = 0.006). Mutational analyses prior to BTKi demonstrated comparable frequencies of MYD88 (86% vs 88%), CXCR4 (34% vs 33%), and TP53 (23% vs 5%) between discontinued and active groups (all p > 0.05).

Table 1 Baseline characteristics and response to Bruton tyrosine kinase inhibitor (BTKi) therapy.
Full size table

The median time on BTKi for all patients was 22.5 months (range, 0.3–136.3), with a longer median time on therapy for those who remained on active treatment than for those who discontinued it [36.6 (range, 13.2–136.3) vs. 12.0 months (range, 0.3–130)]. Major response (MR; i.e., partial response or better) was achieved by 69% of all patients but was higher among those who remained on active treatment (87% vs. 63%) (p = 0.009) (Table 1). The median progression-free survival (PFS) for all patients, from the date of BTKi initiation, was 71.9 months (95% CI, 51.5–97.8), with a 5-year PFS of 57% (95% CI, 48-67%) (Fig. 1A). Among those who discontinued therapy, median PFS (mPFS) was 36.0 months (95% CI: 25.6-69.1), and 5-year PFS was 41% (Supplementary Fig. 1A). Time to next treatment (TTNT) from first BTKi start date to next therapy start date was 38.3 months (95% CI, 32.3–71.9). Among patients discontinuing therapy, the median time from BTKi end date to next therapy start date was 1.6 months (95% CI, 1.0–2.1). The primary reasons for BTKi discontinuation were AEs in 45% (50/111 pts), progressive disease(PD)/large cell lymphoma transformation in 30% (33/111), non-WM-related comorbidities in 7% (8/111), patient death in 6% (7/111), and other factors in 12% (13/111) (complete details in Fig. 1B).

Fig. 1: Outcomes associated with Bruton tyrosine kinase inhibitor (BTKi) treatment and next line therapy.
figure 1

Progression-free survival (PFS) from BTKi initiation (A). Reasons for BTKi discontinuation (B). Overall survival (OS) from BTKi initiation (C). OS stratified by treatment status (D). OS from BTKi discontinuation date (E). Next line therapy stratified by first BTKi agent (F). Response rate associated with next line therapy (G). Overall survival (OS) and progression-free survival (PFS) analyses based on next line of treatment after BTKi discontinuation (H, I). AE adverse event, A-Fib atrial fibrillation, CR complete response, LBCL large B cell lymphoma, MR minor response, ORR overall response rate, PD progressive disease, PR partial response, Pt patient, RR relapsed/refractory, SD stable disease, T-Naïve treatment naïve, VGPR very good partial response, WM Waldenström macroglobulinemia.

Full size image

For the whole cohort, median overall survival (OS) from the start of first BTKi was 132 months (95% CI: 115–not reached (NR)), with a 5-year overall survival (OS) rate of 81% (95% CI, 75–89%) (Fig. 1C). Median OS (p = 0.23) and PFS (p = 0.61) were comparable among patients who were TN and R/R (Fig. 1D, Supplementary Fig. 1B). Furthermore, median time on therapy was comparable between these groups (TN: 10.6 vs R/R: 13.3 months, p = 0.22). Among those who discontinued BTKi, median OS was 106 months (95% CI: 94.3–NR; 5-year OS: 66%) from the date of BTKi discontinuation (Fig. 1E). No significant median or 5-year OS differences were observed between patients who discontinued BTKi due to PD (median: 94 months; 5-year: 63%) versus non-PD-related factors (median: 106 months; 5-year: 68%) (p = 0.59) (Supplementary Fig. 1C).

Following BTKi discontinuation, 80% (n = 89) of patients received subsequent treatment (Fig. 1F); the majority [51%;45/89 patients] received an anti-CD20 antibody ± chemotherapy, followed by 28% (25/89) who received an alternative BTKi-based regimen, 11% (10/89) who received venetoclax (B-cell lymphoma 2 inhibitor [BCL2i]), and 10% (9/89) who received other therapies (outlined in Supplementary Table 1). With this next line therapy, 59% (51/87, 2 patients not evaluable (NE)) of patients achieved a minor response or better. Major response rates were statistically comparable (all p > 0.05) between all next-line regimens (≥PR): BTKi (42%;10/24 patients, 1 NE), anti-CD20/chemo (52%;23/44, 1 NE), venetoclax (50%;5/10), and other (44%;4/9) (Fig. 1G). From the time of initiation of next line therapy, OS (p = 0.02) and PFS (p = 0.04) were longer among patients who received an alternate BTKi-based regimen (3-year: OS 95%, PFS 64%) than among those who received a CD20-based regimen (3-year: OS 65%, PFS 44%), venetoclax (3-year: OS 88%, PFS 29%), or other regimens (3-year: OS 62%, PFS 15%) (Fig. 1H, I). However, when limiting the analysis to just those who discontinued 1st BTKi due to disease progression, no significant differences in OS (p = 0.09) or PFS (p = 0.26) were observed across regimens (Supplementary Fig. 1D, E). Among the small cohort of patients who received subsequent BTKi (16/25 zanubrutinib; 4/25 acalabrutinib; 4/25 ibrutinib; and 1/25 pirtobrutinib), objective response rates (ORRs) (p = 0.47), OS (p = 0.89), and PFS (p = 0.86) were comparable between agents. In patients not receiving subsequent therapy after first BTKi discontinuation, 12/22 patients had died at last follow-up, and 3/22 had progression events; the median OS of this group from date of discontinuation was 10 months (95% CI, 3.0–NR), and median PFS was 10 months (95% CI, 1.8–NR).

Comparing our real-world long-term outcomes to the major studies of covalent BTKis, similar response rates were observed. In the pivotal phase 3 ASPEN trial, ORRs of 95% (zanubrutinib) and 94% (ibrutinib) were reported in MYD88MUT WM (Cohort 1), and 81% in MYD88Wild-Type(WT) disease (Cohort 2; all zanubrutinib) [6, 7, 10]. Other trials evaluating covalent BTKis, including a phase 2 trial of acalabrutinib for frontline or R/R WM, have demonstrated comparable ORRs of 81–100% [1,2,3,4, 11, 12]. In keeping with these data, ORR in the present study was similar, at 80% in the whole cohort. As expected, response rates were significantly higher in BTKi-active (ORR: 98%) compared with BTKi-discontinued (ORR: 74%) patients.

Extended follow-up results from the ASPEN trial demonstrated that 34% and 48% of MYD88MUT patients (median follow-up: 44.4 months) on zanubrutinib and ibrutinib, respectively, had discontinued therapy, compared with 64% of patients with MYD88WT WM (median follow-up: 42.9 months) [6]. The most common reasons for discontinuation were AEs (MYD88MUT: zanubrutinib—9% vs ibrutinib— 20%; MYD88WT—21%) and disease progression (MYD88MUT: zanubrutinib—14% vs ibrutinib—13%; MYD88WT—29%). In the phase 2 study of acalabrutinib, with a median follow-up of 27.4 months, 28% (n = 30/106) of patients treated had discontinued therapy, most commonly due to PD (30%) and AEs (23%) [12]. In the present study, 74% of 150 patients treated with BTKi had discontinued treatment at the time of data cut-off. Contributing factors to this higher percentage likely include the longer follow-up of patients (67 months) and the high percentage of patients treated with ibrutinib (75%), which is associated with more adverse events and a trend toward shorter PFS [10]. The fact that more patients in the BTKi-discontinued group received ibrutinib may be due to lead time bias, given its earlier availability compared with zanubrutinib and acalabrutinib. Finally, the differences in first BTKi discontinuation rates could be influenced by the inherent differences in patient selection and treatment adherence between clinical trials and real-world practice.

Following BTKi discontinuation, the majority of patients were treated with a second BTKi or anti-CD20-antibody-based regimen as next-line therapy. Over the 10-year timeframe of the study, the availability of therapies shifted, with drug classes such as non-covalent BTKis and BCL2is becoming available [13, 14]. In patients who did not have progression on BTKi, our results indicate improved OS and PFS outcomes with a second BTKi, highlighting the robust efficacy and tolerability of these agents. However, when discontinuation was caused by a progression event, no significant advantage was seen with any particular agent, and ORRs were between 50-64% with next-line strategies. Of note, venetoclax, which is often prescribed after BTKi progression/intolerance, was associated with an ORR of 50% in the present study. This is in comparison with an ORR of 75% among patients treated with prior BTKi on the phase 2 trial of venetoclax, and 63% in those having ≥3 lines of therapy [14]. The lower response rate with venetoclax seen in the present study is likely due to the small number of patients treated (n = 10), the majority of whom had received ≥3 lines of therapy.

The present study is limited by its single-center retrospective design and smaller sample size; however, it provides long-term follow-up data and treatment outcomes for patients following BTKi. It demonstrates that following the first BTKi therapy, patients can achieve long overall survival (median >8 years) regardless of the reason for discontinuation. For the next line of treatment, our analysis suggests that the use of an alternate BTKi-based regimen has an OS and PFS advantage if there has not been progression on the first BTKi. However, when a BTKi is discontinued due to disease progression, comparable outcomes can be achieved with a variety of strategies, including alternate BTKi, CD20-based regimens, and BCL2i. Prospective randomized studies and large multi-center real-world analyses are needed to further inform the selection of the best post-BTKi therapy.