Venetoclax retreatment in chronic lymphocytic leukemia is biologically rational and clinically effective

Abstract

Venetoclax therapy has evolved from continuous monotherapy to fixed-duration (FD) and minimal residual disease (MRD)–guided combination regimens in frontline and relapsed/refractory chronic lymphocytic leukemia. Compared with continuous therapy, FD regimens are associated with a lower incidence of emergence of BCL2 mutations and distinct mechanisms of relapse, supporting the rationale for investigating venetoclax-based combination retreatment strategies. In this review, we summarize the key studies for venetoclax rechallenge and discuss the available clinical evidence. To date, 10 abstracts or published articles have reported on venetoclax retreatment, including continuous monotherapy or with FD/MRD-guided combinations with anti-CD20 monoclonal antibodies (rituximab and obinutuzumab) or covalent Bruton tyrosine kinase inhibitors (ibrutinib and acalabrutinib). Patient and disease characteristics were heterogeneous across studies, with a substantial proportion of patients with high-risk disease (IGHV-unmutated or del17p/TP53 mutations). The median time off venetoclax following initial therapy ranged from 9 to 90 months. The reported overall response rates to retreatment ranged from 72% to 100%, and undetectable MRD rates in peripheral blood from 32% to 92%. Median progression-free survival, when available, ranged from 23 to 58 months. Optimal patient selection remains to be defined, but the depth of the initial venetoclax response and time to progression from last venetoclax exposure may predict rechallenge efficacy and are incorporated in new clinical trial criteria allowing previous FD venetoclax treatment.

Introduction

Venetoclax-based therapies have evolved from indefinite monotherapy used in early-phase clinical trials to fixed-duration (FD) and minimal residual disease (MRD)–guided duration combination regimens. Today, FD venetoclax combinations are considered standard of care for both frontline and relapsed/refractory (R/R) chronic lymphocytic leukemia (CLL) and small lymphocytic lymphoma (SLL) and demonstrate noninferior outcome to continuous covalent Bruton tyrosine kinase inhibitors (cBTKi).¹ Relapse after venetoclax therapy is associated with distinct biological features and resistance patterns, which vary according to the regimen of prior venetoclax. Emerging data support the feasibility of B-cell lymphoma 2 (BCL2) inhibitor rechallenge in selected CLL populations. In this review, we discuss the rationale for venetoclax retreatment and summarize the available supporting evidence. We also review the emerging and still limited evidence regarding next-generation BCL2 inhibitors, including lisaftoclax and sonrotoclax, for patients previously treated with venetoclax.

Initial development of continuous single-agent venetoclax

Venetoclax is a small-molecule inhibitor that selectively binds to the BH3 domain of the antiapoptotic BCL2 protein, which is overexpressed in many hematologic malignancies. In CLL/SLL, malignant B lymphocytes show constitutive BCL2 overexpression, mechanistically driven by the deletion of miR-15a and miR-16-1 on chromosome 13q14.² The inhibition of BCL2 disrupts the mitochondrial apoptotic pathway, thereby restoring programmed cell death in leukemic cells.^3,4

Initial clinical trials utilized indefinite venetoclax monotherapy in heavily pretreated high-risk populations and were limited by low complete response (CR) rates and inevitable eventual disease progression often characterized by the emergence of BCL2 mutations.⁵

The first-in-human phase 1 trial of continuous venetoclax therapy for patients with CLL demonstrated the safety of a step-wise dose-escalation strategy to mitigate the risk of tumor lysis syndrome and showed substantial efficacy, with an overall response rate (ORR) of 79%. CR or CR with incomplete marrow recovery (CRi) was achieved in 20% of patients, with a median duration of response (DOR) of 25 months in a heavily pretreated population characterized by prevalent adverse prognostic features, including unmutated IGHV and del17p.⁶ Subsequently, a phase 2 multicenter trial in patients with R/R CLL and del(17p) reported an ORR of 77% and a 1-year progression-free survival (PFS) of 72% with continuous daily dosing of 400 mg, establishing this dose as the recommended target for ramp-up in CLL.^7,8 Undetectable MRD (uMRD) at 10⁻⁴ cells (uMRD-4), assessed by flow cytometry in peripheral blood (PB), was achieved in 30% of patients after a median duration of treatment of 8.8 months. There was a high correlation between attainment of uMRD-4 in the PB and bone marrow (BM), with 20 of 28 available patients who achieved PB uMRD-4 also being uMRD-4 in the BM.⁸ Corroborating previous results, a pooled analysis of 4 early-phase studies assessing continuous venetoclax therapy included 436 patients and demonstrated an ORR of 75%, with CR/CRi of 22% and PB-BM uMRD-4 in 27% to 16% of patients. Median PFS was 30.2 months.⁹ Bulky adenopathy, ≥3 prior lines of treatment, and prior BTKi exposure were associated with a lower likelihood of achieving a CR and a shorter DOR, whereas adverse CLL-related factors (unmutated IGHV, del(17)p, TP53, and NOTCH1 mutations) and prior fludarabine exposure did not appear to affect the ORR but were associated with a shorter DOR.

In the phase 3 VENICE-1 trial assessing venetoclax continuous therapy in R/R CLL, 27% and 35% of patients with and without previous exposure to BTKi had CR/CRi, with a median PFS of 28.3 months.¹⁰ Continuous venetoclax therapy was also effective in patients with del17p/TP53 mutations, with ORR/CR of 77%/21% and a median PFS of 28.2 months.¹¹

Progressive disease (PD) on continuous venetoclax may result from transformation into a large-cell lymphoma (Richter transformation), which typically manifests within 12 months of commencement of treatment, However, it more commonly presents as progression of CLL/SLL without change in immunophenotype.¹² There are multiple mechanisms of resistance to venetoclax described, but the most common involves on-target acquired mutation of BCL2. Blombery et al analyzed the entire coding region of BCL2 by next-generation sequencing in CLL samples from 21 patients before and after progression on continuous venetoclax therapy in early clinical trials. Among recurrent BCL2 mutations, G101V was observed in 4 of 21 patients and conferred resistance to venetoclax while maintaining the physiologic antiapoptotic properties of BCL2 protein in vitro. This mutation was also subsequently reported in other series of venetoclax-resistant disease.^5,13, 14, 15, 16 Mutations can also occur in proximity to G101V (mainly D103 and F104 mutations), and A113G, V156D, and D103E were also reported in a smaller number of cases.^14,15,17 The prevalence of G101V and/or D103Y BCL2 mutations at progression in patients treated with continuous venetoclax varies between series but is generally ∼10% to 15%, and they were detected after a median of 36 months of therapy (range, 13-70 months).^5,14, 15, 16^,18 BCL2 mutations were often observed at low allele frequencies (1%-2%), suggesting other coexisting mechanisms of resistance.^5,13 However, it was subsequently demonstrated that many patients harbor multiple distinct BCL2-mutant subclones (eg, G101V alongside D103Y, Arg107_110dup, or other mutations) and that when the CLL cell fractions of these subclones are considered cumulatively, the total BCL2-mutant burden may represent a substantially higher proportion of the malignant clone than would be appreciated by focusing on a single mutation, suggesting that venetoclax resistance in CLL is often oligoclonal rather than driven by a single dominant BCL2-mutant clone.¹⁸ Similar observations and resistance patterns have previously been reported in the context of continuous BTKi therapy with oligoclonal BTK/PLCG2 mutations and have been described extensively.¹⁹ Non-BCL2 mutations have also been identified, including cell cycle regulators BTG1 and CDKN2A (particularly associated with Richter transformation), as well as new TP53, NOTCH1, KRAS, BIRC3, and SF3B1 mutations, which were also commonly observed at progression.^5,16,20,21 In addition, the upregulation of alternative prosurvival proteins MCL1 and BCL-XL has been described following both continuous and FD venetoclax therapy and is closely associated with the activation of the NF-κB pathway. MCL1 overexpression was initially attributed to chromosome 1q amplification, the MCL1 gene locus.²² However, subsequent observations demonstrated that MCL1 protein overexpression was frequent and predominantly driven by epigenetic upregulation because only a few patients (∼23% in a case series) with MCL1 overexpression harbored increased copy number at 1q. In contrast to MCL1 gene amplification, epigenetic upregulation appeared to be reversible and tended to dissipate after venetoclax discontinuation, in parallel with the resolution of NF-κB activation.^16,23

Development of venetoclax and anti-CD20 monoclonal antibody combinations

FD venetoclax combinations with anti-CD20 monoclonal antibodies (mAb) have been explored with the aim of improving disease control and limiting emergence of BCL2 mutations. They are now considered among standard-of-care treatment options for frontline and R/R CLL/SLL.

The first phase 1b study combining continuous venetoclax with 6 cycles of rituximab in patients with R/R CLL demonstrated a marked improvement in CR/CRi rates compared with those in earlier venetoclax monotherapy studies (ORR, 86%; CR/Cri, 51%), acknowledging that patients were less heavily pretreated (median number of prior therapies, 2).²⁴ In the long-term follow-up, 33 patients with deep response (CR/CRi or uMRD-4) were allowed to stop venetoclax therapy per protocol amendments. Among them, 19 stopped therapy after a median duration of venetoclax of 1.4 years. Interestingly, their 5-year PFS was not inferior (71% [39-88]; vs 79% [49-93]) compared with the 14 patients with a similar depth of response who elected to remain on therapy, thereby supporting further studies of FD venetoclax therapy.²⁵ The phase 3 MURANO trial demonstrated superiority of 2-year FD venetoclax in combination with rituximab (Ven-R) compared with bendamustine-rituximab (BR) in patients with R/R CLL (2-year PFS rates, 84.9% vs 36.3%; P < .001). Ven-R superiority was observed among all disease subgroups (including unmutated IGHV, del17p, and TP53 mutations).²⁶

Venetoclax and obinutuzumab (Ven-O) FD was first evaluated for patients with R/R or untreated CLL and demonstrated high and durable responses rates for this combination (ORR, 95% and 100%; CR, 37% and 78%; and PB uMRD, 64% and 91%, respectively).²⁷ Given that patients with R/R CLL are generally enriched for higher-risk features and the greater likelihood of detectable MRD after prior therapies, further treatment regimens strategies incorporated longer venetoclax exposure (2 years) compared with the shorter duration (1 year) used in the frontline setting.^3,26,28,29 In the frontline setting, the phase 3 trial CLL14 compared 1-year FD Ven-O with chlorambucil and obinutuzumab (O-Chlorambucil) in patients with comorbidities and demonstrated superiority of Ven-O in all subgroups as well (median PFS, not reached vs 35.6 months at 39.6 months median follow-up; P < .001).²⁸ For frontline therapy in fit patients with TP53-intact CLL, the phase 3 CLL13 study explored 1-year FD Ven-R, Ven-O, or Ven-O plus ibrutinib to FCR (fludarabine, cyclophosphamide, and rituximab) in patients aged ≤65 years and BR in patients aged >65 years. Ven-O (with or without ibrutinib) demonstrated superior PB-uMRD-4 rates compared with chemoimmunotherapy (92.2% and 86.5% vs 52%; P < .001). Ven-O (with or without ibrutinib) also demonstrated a superior 3-year PFS compared with FCR/BR (90.5% and 87.7% vs 75.5%; P < .001), whereas Ven-R did not (80.8%; P = .18), establishing obinutuzumab as the preferred anti-CD20 mAb partner to venetoclax in the frontline setting.²⁹ Whether extending venetoclax treatment to 2 years in Ven-O would provide more durable remissions in high-risk CLL remains unknown and is the subject of ongoing debate. However, the HOVON139/GiVe phase 2 trial showed that 12 cycles of venetoclax consolidation following Ven-O did not prevent loss of MRD response.³⁰

FD venetoclax combination regimens achieve high rates of uMRD, which translates to durable PFS comparable to those observed with continuous venetoclax combination regimens. Limiting time on therapy reduces the cumulative toxicity (mostly neutropenia and infection) associated with venetoclax. FD also minimizes the financial and logistical burden and may improve patients’ quality of life.^25,31,32 Moreover, CLL that relapses after FD venetoclax regimens does not show the same resistance pattern as described in the indefinite venetoclax regimens. G101V and D103Y BCL2 mutations were not observed in disease samples at progression after time-limited venetoclax therapy and FD treatment does not seem to drive the acquisition of TP53 or NOTCH1 mutations, because the mutational landscape at relapse generally reflects the baseline genetic profile rather than therapy induced clonal evolution.^14,16,33,34 As previously described, the upregulation of alternative prosurvival proteins (eg, MCL1 and BCL-XL) and activation of survival pathways (eg, NF-κB) have also been observed with FD venetoclax and tend to dissipate after drug discontinuation, supporting the hypothesis of preserved sensitivity of CLL to venetoclax reexposure while acknowledging the overall small number of patients studied (n = 13).^16,23

Broadening spectrum of frontline venetoclax-based FD treatments

Covalent BTKi continuous monotherapy usually leads to a high ORR with prolonged response but is associated with low CR rates and low uMRD rates.35, 36, 37 Moreover, continuous therapy with cBTKi is associated with the emergence of mutations in the BTK gene that may confer resistance to cBTKi (eg, C481S and L528W) and more rarely to noncovalent BTKi (ncBTKi; eg, T474I and L528W), as well as gain in function mutations of the downstream PLCG2.^16,38 Furthermore, continuous exposure to BTKi confers a cumulative risk of toxicity including hypertension and cardiac arrythmias.³⁹ In recent years, efforts have been made to deepen the response to BTKi, achieve uMRD, and use FD treatment to limit BTKi-associated side effects and the emergence of BTK mutations.

Combining venetoclax with cBTKi in a FD regimen is effective. FD venetoclax and ibrutinib were combined in the phase 2 CAPTIVATE trial and the phase 3 GLOW trial. They demonstrated high rates of CR and PB uMRD (52%-64% and 77%-90%, respectively) and superiority over O-Chlorambucil in patients with and without high-risk biology CLL.40, 41, 42 Similarly, FD acalabrutinib and venetoclax, with or without obinutuzumab, was evaluated in the phase 3 AMPLIFY trial and both demonstrated superiority over BR or FCR.³⁷

MRD-guided duration therapy has also gained increasing interest, because it may allow for short treatment duration in quick responders while delivering longer treatment exposure to achieve better disease control in slower responding disease, as well as potentially preventing toxicity and the emergence of BCL2 or BTK mutations associated with continuous therapy. However, treatment duration remains unpredictable and may be substantial, paradoxically especially so in the prognostically favorable IGHV-mutated subset. The phase 2 HOVON141/VISION trial compared patients with or without del17p treated with frontline ibrutinib-venetoclax, assigning them either to discontinue treatment upon achieving uMRD-4 after 15 cycles or to continue with ibrutinib maintenance. The study reported a 12-month PFS rate of 98% in the treatment discontinuation group.⁴³ Recently, the extended follow-up of the phase 3 FLAIR trial demonstrated an overall survival (OS) benefit of MRD-guided therapy with venetoclax-ibrutinib in patients aged up to 75 years (and without del17p in >20% of cells) compared with that with ibrutinib (with or without rituximab) or FCR (hazard ratio [HR], 0.41 [95% confidence interval (CI), 0.20-0.83]; HR, 0.26 [95% CI, 0.13-0.50], respectively). Patients receiving venetoclax-ibrutinib could discontinue treatment upon achieving a sustained MRD response, defined as uMRD4 in PB at +3 months and uMRD4 in both the PB and the BM at +6 months following the first documentation of uMRD. However, treatment duration with venetoclax-ibrutinib remained substantial, with a median of 27 cycles (range, 2-79) of 28 days each, and was paradoxically longer in patients with favorable-risk IGHV-mutated CLL than in those with unmutated IGHV, with >30% of patients with IGHV-mutated CLL still receiving treatment at 60 months.³⁵ Moreover, few deaths were due to CLL, and the OS benefit may have been due to lowering complication rates relative to continuous ibrutinib. Venetoclax-ibrutinib was also resumed upon MRD-5 reemergence, but this treatment resumption was not considered as a PFS event. Interestingly, no BTK mutations were observed among patients treated with MRD-guided venetoclax-ibrutinib. However, 1 patient developed a BCL2 G101V mutation. An MRD-guided strategy was also studied in the phase 2 BOVen trial, combining zanubrutinib, venetoclax, and obinutuzumab for 8 to 24 cycles in treatment-naïve patients. Treatment was stopped 2 months after obtaining uMRD-4 in PB and BM and could be restarted at detectable MRD-4 or PD. After a median follow-up of 25.8 months, 94% achieved uMRD-4 in PB and BM, with a median of 10 cycles to reach uMRD-4 and a median duration of MRD response of 34 months.44, 45, 46

FD combinations of venetoclax with cBTKi appear to result in a lower rate of BTK resistance mutations at relapse, likely due to shorter BTKi exposure and deeper remissions with combination therapy.⁴¹ After a median follow-up of 38.9 months in the CAPTIVATE trial, 15% of patients developed PD, with no BTK or PLCG2 mutations. An uncommon BCL2 mutation (A113G) was observed, but its potential role in venetoclax resistance remains unknown (response of retreatment with venetoclax-ibrutinib in this patient is pending).^47,48

Overview of venetoclax retreatment approaches and available evidence

Available evidence on venetoclax retreatment, either as monotherapy or in combination, is summarized in Table 1. To date, 10 abstracts or published articles have reported on venetoclax retreatment. Most patients exhibited high-risk CLL characteristics, most commonly IGHV-unmutated disease, with a substantial proportion also carrying del17p and/or TP53 mutations.

Table 1.

Summary of available evidence on efficacy of Ven-based retreatment strategies

Cohort	No. of patients	Disease features	First Ven regimen	Median time off Ven (range), mo	Second Ven regimen	ORR/CR rates (%) and DOR	uMRD rates, %	Comments
Shadman et al49	48	NR	Ven monotx: 58.3% Ven-O or Ven-R: 33.3% Ven-BTKi: 8.4% ≥2L: 91.7%	21.2 (9.4-32.1)	Ven monotx: 58.3% Ven-O or Ven-R: 27.1% Ven-BTKi: 12.6%	NR TTNTD: 25.9 mo	NR	68.8% of patients had received prior BTKi before first Ven. 85.4% of patient had no therapy between first and second Ven. Median duration of first Ven was 12.1 mo (range, 4.1-24). For Ven-O or Ven-R, median TTNTD was not reached. For continuous Ven monotx, median TTNTD was 25.9 mo.
Niemann et al50 Retreatment after GAIA/CLL13	49	Del17p or TP53 mut: 0 (protocol criteria) Among overall cohort of patients receiving first Ven regimen: IGHV unmutated: 93/112 (83%) Genomic complexity: 27/112 (24%)	Ven-R, Ven-O, or Ven-O–ibrutinib All frontline	Time from progression on first Ven regimen to initiation of 2L therapy: Ven-R: 12.8 Ven-O: 11.4 Ven-O–ibrutinib: 14.5	Ven-based: 23 Ven-Acala: 24 Ven-O–Acala: 2	NR 2-y TFS for second Ven: Ven-based, 81.4%; Ven + BTKi, 100%	NR	No patient receiving Ven-BTKi as second Ven regimen had received Ven-O-ibrutinib as first Ven regimen.
Brander et al51 M13-365	9	Del17p or TP53 mut: 2 (11%) IGHV unmutated: 5 (55%) Genomic complexity: NR	Ven-R Ven continuous therapy: patient could stop therapy if PB uMRD was achieved ≥2L: 100%	38.4 (17.3-68.7)	Ven-R	100/33 Median PFS on second Ven regimen, 4.9 y; median PFS from the start of first Ven regimen, 9.5 y	PB: 22 BM: 11	No patient had BCL2 mutation before second Ven.
Kater et al52 MURANO	25	Del17p: 4 (16%) TP53 mut: 7 (28%) IGHV unmutated: 22 (88%) Genomic complexity: 10 (40%)	Ven-R ≥2L: 100%	27.6 (14.4-37.2)	Ven-R	72/24 Median PFS was 23 mo and was longer in patients with a median time off treatment >2 y (24.3 vs 14.3 mo; HR, 0.28; 95% CI, 0.10-0.79)	PB: 32
Davids et al53 REVENG	25 (Co-1 22 and Co-2 3)	Del17p or TP53 mut: 8/21 (38%) IGHV unmutated: 18/21 (86%) Genomic complexity: -	Ven-O ≥2L: 42.8%	53.3 (19-90)	Ven-O (Co-1, 12 mo Ven; Co-2, 24 mo Ven)	100/20 among 15 available	PB: 85 (11/13 available)	Co-1 had PD after frontline Ven >2 y, and Co-2 after 1-2 y. No BCL2 mutation were observed before Ven retreatment. Median follow-up was 10.3 mo, with no PD.
Mazot et al54 CLL2-BAG	7	Del17p or TP53 mut: 4 (57%) IGHV unmutated: 7 (100%) Genomic complexity: 2 (29%)	Ven-O MRD-guided duration All frontline	68 (NR)	Ven-O MRD-guided duration	100/NR Median PFS, 33 mo	PB: 43	Median duration of first Ven-O was 6 mo (range, 12-27) Median duration of second Ven-O was 21 mo (range, 11-30).
Krestin et al55 HOVON/REVEAL	15/44 available	Del17p: 0 TP53 mut: 0 IGHV unmutated: 10 (67%) Genomic complexity: NR	Ven-O: 47% Ven-R: 53% All frontline	36 (31-43)	Ven-Acala (24 mo)	100/36	BM: 21 PB: 43	After a median follow-up of 30 mo, 1 patient experienced PD.
Ghia et al48 CAPTIVATE	6	NR	Ven-ibrutinib	NR	Ven-ibrutinib	83/33	NR
Thompson et al56	46	Del17p: 20.5% TP53 mut: 15.6% IGHV unmutated: 82.1% Genomic complexity: 20.5%	Ven monotx: 37% Ven-R: 47.8% Ven-O: 4.3% Ven-ibrutinib: 2.2% Other: 8.7% ≥2L: 91.3%	16 (3-52)	Ven monotx: 45.7% Ven-R: 28.2% Ven-O: 10.9% Ven-ibrutinib: 4.4% Other: 10.9%	79.5/33.3 Median PFS, 25 mo (95% CI, 17-42)	PB or BM: 41.7
Cramer et al57	13	Del17p and/or TP53 mut: 8 (62%) IGHV unmutated: 13 (100%) Genomic complexity: 6 (46%)	Ven-O–ibrutinib: 23% Ven-O: 76.9% ≥2L: 30%	29 (15-55)	Ven monotx: 15.4% Ven-O: 15.4% Ven-O–Acala: 69%	100/NR	PB: 92	No progression was observed after a median follow-up of 19 mo (range, 8-33).

2L, second-line; Acala, acalabrutinib; Co-1/2, cohort 1 and cohort 2; mut, mutation; NR, not reported; TFS, treatment-free survival; TTNTD, time to next treatment of death; Ven monotx, venetoclax monotherapy.

The MURANO retreatment cohort included 25 patients, most of whom had IGHV-unmutated CLL (88%).⁵² High-risk genomic features were present in a subset, with del17p and/or TP53 mutations detected in 16% and 28% of patients, respectively. Initial Ven-R had been administered as second-line therapy in 80% of cases. No new TP53 mutations were detected after frontline Ven-R; however, the prevalence of genomic complexity increased. Among the 14 patients re-treated with Ven-R, 50.0% and 7.1% had 3 to 4 and ≥5 copy number alterations before frontline Ven-R, respectively, compared with 21.4% and 42.9%, respectively, before Ven-R retreatment. The median treatment-free interval before retreatment was 2.3 years (range, 1.2-3.1). With Ven-R retreatment, the ORR was 72.0%, including 6 CR and 12 partial response (PR). PB uMRD-4 was achieved in 32% of patients at the end of combination therapy; however, no patient maintained uMRD at the end of the planned 2 years of venetoclax. Median PFS and DOR were 23.3 months (95% CI, 15.6-24.3) and 15.5 months (95% CI, 11.5 to not reported), respectively. Estimated OS rates at 1, 2, and 3 years were 96.0% (95% CI, 88.3-100.0), 79.6% (95% CI, 63.6-95.6), and 53.1% (95% CI, 25.1-81.0), respectively, noting that the median follow-up of 32.3 months limits the full capture of long-term outcomes. Moreover, compared with the 9 patients reported by Brander et al, the re-treated population was enriched for high-risk CLL features. Among the 8 deaths observed, 2 were unrelated to CLL progression (1 due to COVID-19 and 1 due to a myelodysplastic neoplasm). Patients re-treated after a treatment-free interval of ≥2 years experienced significantly longer PFS than those re-treated within 2 years (median PFS, 24.3 vs 14.3 months; HR, 0.28; 95% CI, 0.10-0.79; P = .0107). Among 4 patients with detectable BCL2 mutations following frontline Ven-R, 2 were re-treated with Ven-R. One patient developed PD during venetoclax treatment and subsequently underwent allogeneic stem cell transplantation, whereas in the other patient, the disease progressed 2.5 years after venetoclax reinitiation. Neither patient achieved uMRD.

Similarly, the REVENG trial evaluated FD Ven-O retreatment in patients with disease relapsing after initial Ven-O.⁵³ Venetoclax was administered for 12 cycles in patients whose PD occurred >2 years after the completion of the first-line Ven-O (cohort 1, n = 22) and for 24 cycles in those relapsing between 1 and 2 years (cohort 2, n = 3). Most patients had IGHV-unmutated CLL (86%), and 38% harbored del17p and/or TP53 mutations. No patient had acquired a BCL2 mutation after first-line Ven-O. The median interval between the last venetoclax dose and retreatment initiation was 53.3 months (range, 19-90). Among the 15 patients who completed the combination phase plus the 3-month postcombination assessment, all achieved a response, including 3 CR/CRi. PB uMRD by next-generation sequencing was obtained in 11 of 13 evaluable patients at the end of combination therapy (n = 2 at 10⁻⁴, n = 3 at 10⁻⁵, and n = 6 at 10⁻⁶).

Recently, Brander et al reported on 9 patients who received Ven-R retreatment after stopping therapy in the M13-365 trial.⁵¹ Patients enrolled in this phase 1b study had R/R CLL and were treated with Ven-R. They could discontinue venetoclax if they achieved CR or uMRD (regardless of International Workshop on Chronic Lymphocytic Leukemia response category). Of the 49 patients enrolled, 33 achieved a deep response, and among these, 16 elected to stop treatment after a median of 1.4 years of venetoclax therapy. Of these 16 patients, 9 were re-treated with Ven-R after disease progression and the development of indications for therapy per iwCLL criteria. Five patients had IGHV-unmutated CLL, and 1 had del17p mutation and 1 had TP53 mutation. The median number of prior lines of therapy before first Ven-R was 2 (range, 1-4). The median time off treatment before PD was 38.4 months (range, 17.3-68.7), and the median time from PD to meeting retreatment criteria was 4.9 months (range, 0.5-35.5). Disease in all patients responded to retreatment (ORR, 100%; CR, 33%), with 22% and 11% achieving uMRD-4 in PB and the BM, respectively. Median PFS was 4.9 years, and median PFS from the first Ven-R treatment (PFS2) was 9.5 years. Among the 6 patients tested, none had acquired a BCL2 mutation after their first Ven-based treatment.

These encouraging data are supported by preliminary reports presented at national meetings, involving more heterogeneous cohorts, and are described in more detail in Table 1. Frontline venetoclax-based regimens were predominantly Ven-R or Ven-O. With the exception of 3 cohorts, retreatment was uniformly administered in combination with an anti-CD20 mAb (rituximab or obinutuzumab) or a covalent BTKi (ibrutinib or acalabrutinib). The median treatment-free interval between initial venetoclax exposure and retreatment ranged from 9 to 90 months. Most retreatment regimens employed a FD approach, although 1 cohort (CLL2-BAG) adopted an MRD-guided strategy. Reported ORR were consistently high (72%-100%). However, CR rates were somewhat lower than those typically observed after frontline venetoclax-based combinations (range, 20%-36%), and PB uMRD rates showed substantial heterogeneity across cohorts (32%-92%). Median PFS, which was not consistently reported across studies, ranged from 23 to 58 months, indicating that clinically meaningful and durable responses can be achieved

Discussion

Venetoclax-based retreatment strategies have the potential to induce deep and, in a proportion of patients, durable responses in R/R CLL. However, optimal patient selection and a preferred venetoclax-based regimen remain to be defined, because available evidence is still emerging.

CLL characteristics such as IGHV mutation status or TP53 aberrations are highly likely to influence retreatment durability, as observed at first venetoclax exposure. However, they do not preclude response to venetoclax retreatment because high ORR and uMRD rates have also been reported in patients with high-risk CLL. Although TP53 mutations are consistently associated with shorter response durations to both chemoimmunotherapy and targeted agents (BTKi and anti-BCL2) in CLL, they are not associated with lower response rates to venetoclax-based therapies.^6,33,58 Among potential predictive factors, the treatment-free interval from frontline venetoclax-based therapy appears particularly relevant and is incorporated into eligibility criteria for ongoing clinical trials, which generally require >1 to 2 years off therapy before retreatment with venetoclax.^29,37,40,53 In addition, the depth of response (particularly the achievement of uMRD following frontline venetoclax-based therapy) may also affect retreatment outcomes, but this parameter is inconsistently reported in currently available venetoclax retreatment studies. The role of BCL2 mutations in influencing venetoclax retreatment response remains uncertain, with mutations being extremely rarely observed or reported in FD contemporary venetoclax regimens.¹⁴ However, we acknowledge the lack of large-scale systematic data to inform this issue, and it is possible that the impact of BCL2-mutated oligoclonal progression following FD venetoclax-based frontline therapy is underappreciated with currently available follow-up. This is illustrated in the long-term follow-up of MURANO, in which BCL2 mutations were detected by digital polymerase chain reaction (limit of detection, 0.1%) at various time points in 4 patients. Venetoclax reexposure likely exerts selective clonal pressure with implications for both the depth and durability of response to retreatment and highlights the need for longitudinal molecular monitoring in clinical trials.

The optimal retreatment duration, whether FD or MRD-guided, also remains unknown. In available venetoclax retreatment studies, most regimens used FD approaches, with the exception of CLL2-BAG, which is evaluating an MRD-guided Ven-O retreatment strategy. Moreover, there are no available data on venetoclax-based combinations directly comparing FD to MRD-guided approaches, making it difficult to extrapolate their efficacy to venetoclax retreatment. MRD-guided regimens offer the potential to further individualize treatment, potentially reducing overtreatment and cumulative toxicity for deep responders, while allowing for extended therapy in those with persistent MRD. However, they may be more difficult to implement in a nonclinical trial setting and stopping rules after achieving uMRD vary among studies, making uniformity challenging. Finally, the ideal venetoclax partner also remains uncertain and will likely depend on the frontline regimen, and patient comorbidities with further insights expected from ongoing clinical trials.

Future directions

With some primary results available and presented in this review, ongoing clinical trials dedicated to venetoclax retreatment include the ReVenG trial (ClinicalTrials.gov identifier: NCT04895436), the CLL2-BAG trial (NCT02040153) with Ven-O, and the REVEAL trial with FD acalabrutinib-venetoclax (NCT01252953). The MAVRiC trial (NCT07024706) evaluates acalabrutinib and venetoclax in patients previously treated with a venetoclax-BTKi regimen as frontline therapy. The trial also takes into account high-risk features, with patients with IGHV-unmutated and/or TP53-mutated CLL receiving 22 cycles of acalabrutinib and venetoclax, whereas patients lacking these features receive 12 cycles of treatment. In addition, the long-term follow-up of the CAPTIVATE trial (NCT02910583) with ibrutinib and venetoclax is ongoing.

The place of venetoclax retreatment will also be challenged by new and possibly more potent BCL2 inhibitors, namely lisaftoclax and sonrotoclax. Recently, Davids et al published the results of a phase 1b/2 trial evaluating lisaftoclax monotherapy or in combination with rituximab or acalabrutinib in untreated or previously treated CLL.⁵⁹ Fourteen patients had prior exposure to a BCL2 inhibitor (all in the lisaftoclax-acalabrutinib cohort), including 12 (13.2%) who had previously received venetoclax (10 of whom experienced disease progression within 12 months of frontline venetoclax). Among patients previously exposed to venetoclax, the proportion who had also been exposed to a BTKi was not specified. The ORR was 92.9%, and among patients with venetoclax-refractory disease, the 18-month PFS was 88.9%. It remains unclear to what extent this clinical activity can be attributed to lisaftoclax, because clinical evidence supporting lisaftoclax efficacy in patients with CLL previously exposed to venetoclax is currently lacking. Sonrotoclax (BGB-11417) is a next-generation BCL2 inhibitor with greater selectivity and pharmacologic potency than venetoclax, characterized by a shorter half-life and an absence of drug accumulation. Encouraging in vitro and in vivo data have shown that sonrotoclax retains activity against the BCL2 G101V mutation, and many ongoing clinical trials of sonrotoclax are allowing prior venetoclax exposure.⁶⁰

The role of venetoclax retreatment within the R/R CLL treatment landscape is yet to be fully defined. Several therapeutic options are available or under investigation, including ncBTKi and BTK degraders, bispecific T-cell–engaging antibodies, and chimeric antigen receptor T-cell therapies.61, 62, 63, 64, 65, 66 Current ongoing trials for R/R CLL assessing venetoclax-based treatment strategies with novel therapies may allow prior venetoclax exposure, but they generally require a minimum DOR to prior venetoclax-based therapy (typically >2 years). The novel ncBTKi pirtobrutinib is being tested in combination with venetoclax in phase 2 trials (ClinicalTrials.gov identifiers: NCT06466122 and NCT05536349), which include patients with prior venetoclax exposure. Bispecific T-cell–engaging antibodies, mosunetuzumab and epcoritamab, are also being combined with venetoclax for R/R CLL (NCT05091424 and AETHER trial NCT05791409, respectively) and allow previous venetoclax therapy. Phosphatidylinositol 3-kinase inhibitors are also being assessed in combination with venetoclax and allow prior venetoclax exposure. The phase 1/2 study evaluating venetoclax, umbralisib, and ublituximab included patients with R/R CLL or Richter transformation. Treatment could be discontinued after 12 cycles in patients achieving uMRD-4 in the BM. Patients with detectable MRD-4 in the BM at cycle 12 continued umbralisib until disease progression or intolerance. Preliminary reports showed ORR/CR rates of 98%/42% in CLL, and 77% of patients achieved uMRD-4 in the BM.⁶⁷ Finally, the phosphatidylinositol 3-kinase δ inhibitor roginolisib combined with venetoclax and rituximab is currently being studied for R/R CLL (NCT06644183), including patients with prior venetoclax treatment.

Conclusions

In conclusion, venetoclax-based combinations, whether as FD or MRD guided, may provide high and durable responses in patients with and without high-risk features in CLL. The DOR after the first venetoclax-based therapy may serve as a predictor of rechallenge efficacy. The ideal venetoclax partner remains unclear, although efficacy has been reported for both anti-CD20 mAb and cBTKi. Monitoring for the emergence of BCL2 mutations will be crucial during venetoclax rechallenge to understand the risk of this resistance mechanism emerging under the selective pressure of retreatment.

Conflict-of-interest disclosure: M.C. declares honoraria from Bristol Myers Squibb (BMS). M.A.A. declares honoraria from AbbVie, Jansen, AstraZeneca, BeiGene, Sobi, Specialised Therapeutics, Novartis, Gilead, Roche, Takeda, and CSL; and is an employee of the Walter and Eliza Hall Institute. P.T. declares honoraria from AbbVie, Ascentage, AstraZeneca, BeiGene, Genmab, Janssen, Lilly, and Merck; speaker fees from AbbVie, AstraZeneca, BeiGene, Janssen, Merck, and Roche; and travel support from Merck. J.F.S. declares honoraria for advisory board role from AbbVie, AstraZeneca, BeiGene, BMS, Gilead, Janssen, Loxo, and Roche; research funding from BMS, Janssen, Loxo, and Roche; speaker’s bureau fees from AbbVie, AstraZeneca, and Roche; fees for expert testimony from BMS, Roche, and TG therapeutics; and a consultant role for TG therapeutics.