Triplet regimens for frontline treatment of CLL—Great company or just a crowd?

Abstract

Standard frontline treatment of chronic lymphocytic leukemia (CLL) is with fixed‐duration venetoclax‐based doublets or indefinite covalent Bruton tyrosine kinase inhibitor (BTKI). Although these approaches achieve excellent results, venetoclax doublets have diminished efficacy in high‐risk biological subgroups, and indefinite covalent Bruton tyrosine kinase inhibitor (cBTKI) is associated with cumulative cardiovascular and infectious toxicity. Triplet regimens for treatment of CLL involve simultaneous use of cBTKI, venetoclax, and anti‐CD20 monoclonal antibody. Three major frontline Phase 3 trials (CLL‐13/GAIA, AMPLIFY, and A041702) have demonstrated higher rates of undetectable minimal residual disease (uMRD) and longer remissions with triplets than doublets, particularly in patients with IGHV‐unmutated (IGHV‐U) disease. However, this comes at the cost of increased infectious toxicity, particularly with COVID‐19, and thus has translated into a variable impact on progression‐free survival (PFS) and, to‐date, no overall survival (OS) benefit. Although there are promising Phase 2 data for triplets in patients with TP53 aberrant or relapsed disease, the heterogeneity of treatment duration/MRD definition, lack of control arm, and potential increased toxicity make it premature to use triplets in these groups. We recommend considering triplets in treatment naïve CLL patients with IGHV‐U, TP53 wild type, anticipated low incidence/good tolerance of Gr ≥ 3 infection (<70 years old, no major comorbidity and fully immunized) who are well informed and prioritize maximal time off therapy at the expense of increased short‐term logistical complexity. Future triplet research should focus on randomized trials in specific genomic subgroups, incorporating novel agents (e.g., non‐covalent BTKI, BTK degrader, and next‐generation BCL2 inhibitors) and new ways of adapting treatment duration to maximize efficacy and minimize toxicity.

INTRODUCTION: AN ABUNDANCE OF RICHES

The treatment of chronic lymphocytic leukemia (CLL) has been revolutionized by the development of three major classes of non‐cytotoxic therapies: anti‐CD20 monoclonal antibodies (mABs), BCL2 inhibitors (BCL2is), and covalent Bruton tyrosine kinase inhibitors (cBTKIs). ¹ , ² Currently, standard frontline treatment of CLL is with one or two of these drugs in three distinct paradigms: indefinite cBTKI (1) (e.g., acalabrutinib, zanubrutinib), fixed‐duration (FD) anti‐CD20 mAb/BCL2i (2) (e.g., venetoclax/obinutuzumab), or FD cBTKI/BCL2i (3) (e.g., ibrutinib/venetoclax or acalabrutinib/venetoclax).3, 4, 5, 6, 7, 8, 9 The benefit of adding anti‐CD20 mAB to BCL2i is undisputed, with higher rates of undetectable measurable residual disease (uMRD) and longer progression‐free survival (PFS). ¹⁰ Conversely, the benefit of adding anti‐CD20 to cBTKI is dependent on the specific mAB, cBTKI, and patient population. There was no benefit of adding rituximab to ibrutinib, ⁶ , ¹¹ but a significant improvement in PFS when obinutuzumab was added to acalabrutinib in the ELEVATE‐TN trial (6‐year PFS 62 vs. 78%). ¹² There are no randomized data available regarding the addition of obinutuzumab to ibrutinib or with any antibody and zanubrutinib.

Each of these approaches has its shortcomings. Indefinite cBTKI is hampered by the continuous nature of therapy, cumulative cardiovascular and infectious toxicities, and low rates of uMRD. ¹³ , ¹⁴ FD venetoclax doublets achieve high rates of uMRD but do not fully abrogate the negative prognostic impact of IGHV‐unmutated (IGHV‐U) status, karyotypic complexity, or TP53 aberrancy ¹⁵ and are associated with major logistic complexities with delivery. ¹⁶ Combining all three drugs in a “triplet” approach has been heralded as a potential solution to many of these shortcomings. ¹⁷ But is a third drug “great company or just a crowd”? Does the increased depth of response negate the increased toxicity and complexity? With the recent publication of two major Phase 3 trials assessing such triplet approaches and a plethora of Phase 2 studies, we are beginning to answer these questions. Establishing the efficacy and toxicity of triplets from these data is complex and requires consideration of the duration of therapy (fixed vs. MRD‐guided), definition of uMRD, genomic risk, and study timing relative to the COVID‐19 pandemic. Herein, we synthesize the data for triplets in upfront treatment of CLL, offer an expert opinion on when they could optimally be used, and frame the key questions for future research.

PHASE 3 DATA FOR TRIPLETS IN FRONTLINE CLL

There are three published Phase 3 trials including a triplet arm (Table 1).

Table 1.

Data for frontline treatment of chronic lymphocytic leukemia (CLL) using triplets.

Regimen	No pts/median age	Population	Years enrolled	Treatment	PFS	uMRDa	Gr3+ cytopenias	Gr 3+ infections	Gr5 infection/COVID‐19	Reference
Ibrutinib/venetoclax/obinutuzumab
CLL2‐GiVE	N = 41 62 years	All TP53 aberrant	2016–2018	C1–6 IVO → C7–12 IV → C13–15 I → continue until PB uMRD CR or C36	3‐Year: 80%	PB (10−4) 78% at C15	Neutropenia: 42% Thrombocytopenia: 22%	19.5%	0%/0%	18
Rogers et al. Phase 2	N = 25 59 years	TN or RR (50 total)	2017–2019	C1 O → C2 IO →C3–8 IVO → C9–14 IV. → C15 onwards I mono at investigator discretion	Median: 81.8 m	BM (10−4) 67% at 2 months after C14	Neutropenia: 66% Thrombocytopenia: 34%b	10.7%b	4%c/0%	17
CLL13‐GAIA	N = 1080 total N = 231 IVO 61 years	CIRS ≤ 6 TP53 wild type	2016–2019	CIT versus VO versus RV versus IVO IVO: C1–6 IVO → C7–12 IV → I monotherapy until PB uMRD ×2 or C36 (median 13C)	5‐year overall: IVO 81.3% versus VO 69.8% HR 0.61 (95% CI 0.41–0.91) 5‐year IGHV‐U: IVO 75.9% versus IV 59%CI	PB (10−4) 92% IVO versus 87% VO at month 15	Neutropenia: VO 56% versus IVO 48%	21.2% IVO versus 13.2% VO	IVO: 1%/0.5%% VO: 1%/0.5%	19 20 21
A041702	N = 465 total N = 233 IVO 74 years	>70 or > 65 with del(17p)	2019–2022	(1) IO: O C1–6 with I indefinite (2) IVO: Ven from C3 → stop at C14 if BM uMRD CR	3‐year: IVO 87.5% versus IV 85 HR 1.2 (95% CI 0.73–1.97)	PB (10−4) 86.8% IVO versus 33.3% IO at C14	Cytopenias: 61% IVO versus 38% IO	21.2%	IO: NR/4.7% IVO: NR/8.2%	22
Acalabrutinib/venetoclax/obinutuzumab
Davids et al. Phase 2	N = 72 63 years	63% TP53 aberrant	2018–2022	Acala C1 → AO from C2 → AVO C4–7 → AO C8–15 → AV stopped at C6 if uMRD CR or C25 if uMRD PR	4‐year TP53 aberrant: 70% TP53 wild type: 88%	BM (10−4) 78% (all), 71% (TP53 aberrant) at C16	Neutropenia: 36% Thrombocytopenia: 28%	10%	1.4%/1.4%	23
AMPLIFY	N = 1141 total N = 286 AVO 61 years	CIRS ≤ 6 TP53 wild type	2019–2021	CIT versus AV versus AVO AVO: Acala C1–14, Ven C3–15, Obin C2–7	3‐year overall AV: 76.5% AVO: 83.1% CIT: 66.5% 3‐year IGHV‐U‐AVO 82.8% versus AV 68.9%	PB (10−4) 45% AV versus 95% AVO versus 73% CIT at EOT	Neutropenia: 46.1% AVO versus 32.4% AV Thrombocytopenia: 6% AVO versus 1.4% AV	23.6% AVO versus 12.4% AV	AVO: 10.1%/8.7% AV: 3.4%/3.4%	9
Zanubrutinib/venetoclax/obinutuzumab
BOVEN	N = 52 62 years	17% TP53 aberrant	2019–2021	ZO C1–2 → ZVO C3–8→ ZV ceased at C8–24 cycles when uMRD PB and BM	NR	PB & BM 10−4) 92% as best response	Neutropenia: 18% Thrombocytopenia: 8%	10%	0%/0%	24 25

Abbreviations: A, acalbrutinib; BM, bone marrow; CIRS, cumulative illness risk score; CIT, chemotherapy; CR, complete response; EOT, end of therapy; HR, hazard ratio; I, ibrutinib; IGHV‐U, IGHV‐unmutated; NR, not reported; O, obinutuzumab; PB, peripheral blood; PFS, progression‐free survival; PR, partial response; RR, relapsed/refractory; TN, treatment naïve; uMRD, undetectable minimal residual disease; V, venetoclax; Z, zanubrutinib.

^a Rates of uMRD generally reported from evaluable subset.

^b Includes RR patients also.

^c Includes events at any time post study enrollment, including after salvage treatment. None of these infectious deaths occurred during study treatment.

CLL13

The CLL13 study randomized 1080 fit (cumulative illness risk score [CIRS] ≤ 6), young (median age 61 years), TP53 wild type patients to one of four treatment arms: CIT (6 cycles of bendamustine/rituximab [BR] or fludarabine/cyclophosphamide/rituximab [FCR]), venetoclax/rituximab (VR: 12 cycles), venetoclax/obinutuzumab (VO: 12 cycles) or ibrutinib/venetoclax/obinutuzumab (IVO). ¹⁹ The duration of IVO was MRD‐guided with cessation when uMRD (10⁻⁴, bone marrow [BM] or peripheral blood [PB]) was achieved in two consecutive 3‐monthly measurements or at Cycle 36 (3 years). The majority (94%) of patients in the IVO arm stopped treatment at C16 after achieving uMRD at C12 and C15. Treatment was largely completed before the onset of the COVID‐19 pandemic, with enrollment between December 2016 and October 2019. The initial publication focused on the composite primary endpoints, showing higher rates of uMRD for VO and IVO compared with CIT (but not VR vs. CIT) and improved PFS with IVO versus CIT (3‐year PFS 90.5% vs. 75.5%, hazard ratio [HR] 0.32, 0.19–0.54, P < 0.001). ¹⁹ , ²⁰

Recently, with >5 years median follow‐up, the IVO triplet has demonstrated superior 5‐year PFS to VO (81.3% vs. 69.8%, HR 0.61, 95% CI 0.41–0.91, P = 0.0046) and higher uMRD at month 15 (92.2% vs. 86.5%). ²¹ However, there was no difference in overall survival (OS). Even with the triplet, PFS appears inferior in patients with IGHV‐U versus IGHV‐M disease (estimated 5‐year PFS 75.9% vs. 89.1%). The magnitude of PFS benefit is larger in IGHV‐U (5‐year PFS 75.9% vs. 59%) but still apparent in IGHV‐M (5‐year PFS 89.1% vs. 82.1%). ²¹

Although the frequency of Grade 3–5 infections is higher with IVO than VO (21.2% vs. 13.2%), there is no significant difference when adjusting for duration of exposure (P = 0.093). Importantly, there was only one infectious death (fungal encephalitis) and no COVID‐related deaths on the IVO arm. Addition of ibrutinib did not appear to increase the risk of Gr3–4 neutropenia (48%) compared with VO (56%), but did increase the incidence of cardiac events (15 vs. 8 per 1000 patient months). ²⁰ Although quality of life improved at the end of IVO treatment, it improved much earlier with VR or VO, suggesting a higher symptom burden during treatment with cBTKI‐containing triplet. ²⁶ This is corroborated by higher rates of treatment discontinuation in the IVO arm. ¹⁹

AMPLIFY

The Phase 3 AMPLIFY trial explored doublet and triplet combinations with acalabrutinib, a second‐generation cBTKI with non‐inferior efficacy and more favorable toxicity than ibrutinib. ²⁷ The trial was similar in size (867 patients randomized) and included population (median age 61 years, TP53 aberrant excluded, CIRS ≤ 6) to the CLL13 trial. Patients were randomized 1:1:1 to CIT (six cycles of FCR or BR), AV (acalabrutinib C1–14, venetoclax C3–14), or FD AVO (AV with obinutuzumab Cycles 2–7). ⁹ Importantly, the trial was not powered to directly compare AV to AVO. The study ran during the height of the COVID pandemic, recruiting from 2019 till end of 2021, and included a significant proportion of patients from Eastern Europe. AV was superior to CIT in terms of PFS (HR 0.65, 95% CI 0.49–0.87), the primary endpoint of the trial, and OS (HR 0.33, 95% CI 0.18–0.56). Although the AVO triplet achieved higher rates of uMRD (95% AVO vs. 45% AV) and a numerically higher 3‐year PFS (83.1% AVO vs. 76.5% AV), OS with AVO was not superior to CIT (HR 0.76, 95% CI 0.49–1.18) and was, in fact, numerically lower compared with AV (3‐year OS: 87.7% AVO vs. 94.1% AV). In contrast to the IVO triplet in CLL13, AVO appeared to perform equally well with IGHV‐U and IGHV‐M disease (3‐year PFS 82.8% vs. 83.6%). This translates to a numerically larger PFS benefit over AV in patients with IGHV‐U (3‐year PFS 82.8% vs. 68.9%), compared with no significant difference in patients with IGHV‐M (3‐year PFS 83.6% vs. 86%). ⁹

The discordance between PFS and OS in the AVO arm is largely explained by higher rates of COVID‐19‐related death (25/8.7% with AVO, 10/3.3% with AV, and 21/7.2% with CIT). Excluding COVID, rates of Gr3–5 infections (4.1% AV, 5.2% AVO, and 4.6% CIT) were similar, whilst rates of Gr5 infection were numerically more common in AVO‐treated patients (0% AV, 1.4% AVO, and 0.3% CIT). When COVID‐19 deaths were censored in an Food and Drug Administration (FDA)–mandated pre‐specified analysis, AVO significantly improved OS compared with CIT (HR 0.46, 95% CI 0.22–0.95). Addition of obinutuzumab to AV is also associated with infusion reactions (19.7%) and a higher incidence of Gr ≥ 3 neutropenia (46.1% vs. 32.4%). ⁹

Interpretation of the COVID‐related toxicity in the AVO arm of AMPLIFY has been a topic of contention. Some experts perceive it as a generalizable signal of the increased immunosuppression with the triplet regimen. Others believe the signal is unique to the time (19/25 deaths in the AVO arm occurred before the availability of any vaccine, and only 7.6% of COVID‐19 infections in the AVO arm received any anti‐COVID therapy) and locations (69.6% of deaths were in Eastern Europe, where the impact of COVID‐19 was most pronounced) in which the AMPLIFY trial recruited. ²⁸ , ²⁹

A041702

Finally, the A041702 trial compared continuous ibrutinib and six cycles of obinutuzumab with or without the addition of venetoclax from Cycle 3 (IO vs. IVO). Venetoclax was ceased after 12 cycles if BM uMRD complete response (CR) was achieved. Patients were exclusively 65 years of age and older (median age 74 years) and enrolled during the onset of the COVID pandemic (2019−2022). ²² Despite high rates of PB uMRD at Cycle 14 (96.9% vs. 33.5%), the predefined futility boundary was crossed at 14 months of follow‐up with the PFS HR favoring the IO arm (1.2, 95% CI 0.73–1.97). Similar to the AMPLIFY trial, this was primarily due to an excess of COVID‐19‐related deaths in the IVO arm (19/233 vs. 11/232). As expected, Gr ≥ 3 cytopenias were also more common in the IVO arm (61% vs. 38%). ³⁰

Taken together, when comparing triplets to doublets in Phase 3 trials (Table 1):

• 1.Triplets are associated with a higher rate of uMRD: This effect is most pronounced when obinutuzumab is added to AV. It is notable that the rates of uMRD appear lower with AV (45% In AMPLIFY) than IV (77% in CAPTIVATE), and there is currently no Phase 3 data assessing the efficacy of IV versus IVO. ³¹
• 2.Triplets appear to have a longer PFS than the various doublets: This is with the caveat that the duration of therapy was unequal in CLL13, and PFS was not directly statistically compared in AMPLIFY. Although the PFS benefit appears most pronounced in patients with IGHV‐U disease, there is conflicting data about whether triplets overcome the prognostic significance of IGHV‐U and whether there is any PFS benefit in patients with IGHV‐M. Longer follow‐up may be required for the latter, given the predominance of late progressions in patients with IGHV‐M disease. Additionally, the potential impact of exposure to multiple drug classes on time to second disease progression or death (PFS2) remains unclear. In the limited data available to date, response rates to retreatment are high, and resistance mutations have not been identified.
• 3.Infectious toxicity is higher in triplets than doublets: Across the three studies, Gr ≥ 3 infections occur in 21%–23.6% of patients treated with triplets, compared to 13.2%–17.5% of patients treated with VenO, ⁷ , ¹⁹ and 8%–15.1% of patients treated with BCl2i/BTKI. ⁸ , ⁹ , ³¹ In Phase 3 trials, the difference is most apparent when the third drug is anti‐CD20 mAB (AMPLIFY) or BCL2i (A041702). Infectious toxicity is more pronounced in older patients (A041702) and studies conducted early in the COVID pandemic (AMPLIFY, A041702). Although unproven, it is unlikely that there is an enduring signal of increased fatal infections for well‐selected patients in the modern era of COVID vaccination and treatment.
• 4.Non‐infectious toxicities of triplets versus doublets depend on which drug is added: Adding ibrutinib to VO increases cardiac, but not hematological, toxicity. Adding obinutuzumab to AV increases Gr ≥ 3 neutropenia and infusion reactions.
• 5.At present, there is no demonstrated OS benefit for triplets versus doublets in any population: Notably, it is very difficult to show an OS advantage in CLL, particularly when the comparator arm is not CIT.

PHASE 2 DATA IN SPECIFIC SITUATIONS

Presently, there is no Phase 3 triplet data with zanubrutinib or pirtobrutinib combinations, in TP53 aberrant disease or in relapsed disease. However, there are Phase 2 data in these situations.

Other cBTKI and non‐covalent BTKI

The Phase 2 BOVEN study evaluated the combination of obinutuzumab (C1–8), zanubrutinib (from C1), and venetoclax (from C3). Duration was MRD‐guided (8–24 cycles), with discontinuation of zanubrutinib/venetoclax (ZV) when sustained (two assessments, 2 months apart) uMRD was achieved in both PB and BM (flow cytometry, 10⁻⁴). ²⁴ Fifty‐two patients (17% TP53 aberrant) were treated, with 92% of patients achieving the primary endpoint of BM/PB uMRD and stopping therapy after a median of 8 months. ²⁵ The rate of BM uMRD appeared higher than that seen with VO in CLL14 (57%), and 85% of patients with flow uMRD were MRD negative using next‐generation sequencing (Clonoseq, 10⁻⁶). ²⁴ Rates of infusion reaction were substantial (44%), reflecting the initiation of obinutuzumab before BTKI lead‐in, and the incidence of Gr ≥ 3 neutropenia was relatively low (18%), ²⁴ a surprising finding given the comparatively high rates of neutropenia with zanubrutinib in the ALPINE study. ³² This study introduced the concept of “delta MRD 400,” where a 400‐fold reduction in PB MRD level, as assessed by clonoseq at Cycle 5, predicted PFS (15 vs. 23 months, P < 0.001). ²⁵ Delta MRD400 is being explored as a very early means of adapting treatment duration in ongoing clinical trials. ³³ A caution is that the MRD response kinetics may differ by IGHV mutational status, and larger cohorts are needed to validate the predictive performance of MRD400 across the range of biological subsets. Interestingly, 16 patients received zanubrutinib‐venetoclax retreatment for clinical progression (n = 12) or MRD recrudescence > 1% (n = 4) with overall response rate (ORR) 92% and 46% again achieving PB uMRD (10⁻⁴, flow cytometry). ³³

Pirtobrutinib, a non‐covalent BTKI (ncBTKI), has been combined with venetoclax and obinutuzumab (six cycles) (PVO) for frontline treatment of CLL (all genomic risk) in 74 patients with short follow‐up (median 10.3 months). ³⁴ Treatment duration was 12 cycles if BM MRD negativity (clonoseq, 10⁻⁶) was achieved at C12 or else extended to 24 cycles. Amongst 27 patients with follow‐up beyond 12 months, 22/27 (81%) achieved BM uMRD and were able to discontinue therapy, suggesting extremely high efficacy for this triplet. ³⁴ Gr ≥ 3 neutropenia and thrombocytopenia occurred in 58 and 18% of patients with no fatal infections reported.

Triplets in TP53 aberrant disease

Although dependent on patient factors, variant allele frequency, and clinician preference, continuous BTKI is generally favored as first‐line treatment for TP53 aberrant disease.35, 36, 37 This is based on superior PFS with continuous BTKI compared with VO. Specifically, ibrutinib achieves a 5‐year PFS of 70%, zanubrutinib 5‐year PFS of 72.2% in Arm C of SEQUOIA, and acalabrutinib 6‐year PFS of 56% in ELEVATE‐TN (with no apparent benefit of adding obinutuzumab in TP53 aberrant subgroup).38, 39, 40, 41 This is compared to a 4‐year PFS of 53% with VO in CLL14 and a 5.5‐year PFS of 36% with FD I + V in CAPTIVATE. ⁴² , ⁴³ For patients with TP53 aberrant CLL, triplets have been touted to maximize depth and duration of response using MRD‐guided duration of therapy. Whilst Phase 3 data, such as that from the CLL16 study specifically addressing this biological subset, will be required to test this hypothesis, some lessons have emerged from Phase 2 data.

In a study by Davids et al. 72 patients with treatment naïve CLL were treated with AVO (acalabrutinib from Cycle 1, obinutuzumab Cycles 2–6, and venetoclax from Cycle 4). AV was ceased at C15 if BM uMRD CR achieved or at C24 if BM uMRD PR achieved. ²³ Forty‐five (63%) patients were TP53 aberrant. Duration of therapy was relatively long (because of the stringent requirements for CR), with only 36% of patients able to discontinue therapy at C15 and 11% of patients remaining on AV indefinitely. The rates of BM uMRD were high regardless of TP53 status (78% overall, 71% for TP53 aberrant), but 4‐year PFS remained inferior in TP53 aberrant patients (70% vs. 88%). This supports the idea that relapse is driven by early and rapid regrowth of TP53 aberrant clones after treatment cessation. ⁴⁴ It is notable that 4 of 10 progressions while receiving the AV doublet were with Richter transformation (RT). Finally, five patients (three with TP53 aberrancy) in this study have been retreated with AV (three for clinical PD and two for MRD recrudescence), with three of four evaluable patients responding. ²³

The German CLL2 GIVE study treated 41 patients (median age 59), all with TP53 aberrancy, using IVO (six cycles of IVO triplet, six cycles of IV doublet, and three cycles of ibrutinib alone). Ibrutinib monotherapy was then continued until PB uMRD CR was achieved or to a maximum duration of 36 cycles. ¹⁸ The primary endpoint of CR at C15 was achieved in 58.5% of patients, PB uMRD was achieved in 78%, and the 3‐year PFS was 79.9%. Two valuable observations have emerged from this trial. First, the rate of PB MRD positivity increased during ibrutinib monotherapy from C15–36, in contrast to the increasing uMRD rate seen in the Davids et al. study with AV doublet. ²³ Second, 3‐year PFS was lower for patients whose disease had both del(17p) and TP53 mutation compared to those with an isolated TP53 mutation (67.6% vs. 100%). This may have been due to the relative enrichment of IGHV‐M in the TP53‐mutated subgroup. ¹⁸

Taken together, these studies suggest that triplets can achieve high rates of uMRD in TP53 aberrant CLL but do not fully abrogate its negative prognostic impact when delivered as time‐limited therapy. Although the PFS seems superior to VO doublets, this is an unfair comparison given the longer duration of therapy with triplets, comparatively shorter follow‐up, and lack of data on the durability of response at retreatment. ¹⁵ For now, continuous BTKI has the highest evidence for TP53 aberrant disease and, in the authors' opinion, remains the standard of care. Results of the randomized Phase 3 CLL16 trial comparing AVO to AV (both FD) in patients with CLL harboring del(17p), TP53 mutation or complex karyotype are eagerly awaited. Future studies should compare PB MRD‐guided versus FD strategies and, given substantial evidence that uMRD is a more important prognostic marker than attainment of CR, ⁴⁵ , ⁴⁶ , ⁴⁷ remove the requirement for CR as a criterion for treatment discontinuation.

Triplets in relapsed disease

In patients with relapsed disease, time‐limited treatment with triplets to achieve a second treatment‐free interval is highly desirable. In this heterogenous population, MRD‐guided duration of therapy is particularly attractive. At present, triplet data for RR CLL are limited to several small Phase 2 trials.

First, the original IVO triplet data from Rogers et al. included both treatment naïve (TN‐25/50) and relapsed/refractory (RR) patients (25/50). ¹⁷ The RR cohort excluded patients with BTKI refractory disease, had a median of 1 prior line of therapy, and a median age of 58. The proportion of patients with prior BTKI or BCL2i exposure is not reported. Treatment involved one cycle of obinutuzumab monotherapy, one cycle of ibrutinib/obinutuzumab doublet, six cycles of IVO triplet, and six cycles of IV doublet. Continuation of ibrutinib monotherapy was at the investigator's discretion. Whilst the ORR was similarly high (92% vs. 96%), BM uMRD was achieved in only 50% of RR patients compared with 67% in TN. Rates of Gr3 neutropenia (66%) and thrombocytopenia (34%) were higher than in TN studies, particularly in the RR cohort (neutropenia 76%), suggesting tolerance may be reduced in pretreated patients, ¹⁷ particularly those with prior CIT exposure.

The German CLL2‐BZAG enrolled 42 patients with RR CLL, excluding patients with disease which had previously progressed on BTKI or BCL2i. ⁴⁸ After an optional bendamustine debulking (recommended if absolute lymphocyte count > 25 × 10⁹/L or lymph node size > 5 cm), obinutuzumab was given during C1–8, zanubrutinib from Cycle 2, and venetoclax from Cycle 3 (ZVO). Maintenance ZV was continued until PB uMRD CR was achieved in two consecutive 3‐monthly assessments (shortest duration 14 months) or until a maximum duration of 24 months. This was a high‐risk cohort, with 47.5% TP53 aberrant, 45% BTKI exposed, and 17.5% BCL2i exposed. Rates of PB uMRD were modest at 6 months (52.8%) but increased to a best uMRD rate of 85%. This translated to a median PFS of 29 months. Infectious toxicity included two fatal COVID‐19 infections and one fatal fungal pneumonia. Bendamustine debulking did not appear to increase rates of uMRD, but presumably increased net immunosuppression. ⁴⁸

The German CLL2‐BAAG enrolled 46 patients with RR CLL using the same inclusion criteria and treatment paradigm as CLL2‐BZAG, with acalabrutinib instead of zanubrutinib. ⁴⁹ PB uMRD was numerically higher than in BZAG (75.6% at 6 months and 93.3% at any time), and 3‐year PFS was excellent at 85%. ⁴⁹ Infectious toxicity included three fatal COVID‐19 infections, all occurring >18 months after treatment cessation and thus possibly unrelated. This study also assessed MRD by circulating tumor DNA (ctDNA), demonstrating additional utility when combined with flow cytometry and an increased sensitivity for relapsing disease in the nodal compartment. ⁴⁹

In the Phase 1b BRUIN study, 25 patients with RR CLL who were naïve to BCL2i were treated with 25 cycles of pirtobrutinib and venetoclax, 10/25 with rituximab (PVR, C1–6 only) and 15/25 without (PV). At a median follow‐up of 23 months, the triplet yielded an ORR 100%, CRR 30%, and 90% PB uMRD after C12. Gr ≥ 3 neutropenia and infections were higher in patients treated with PVR compared to PV (60% vs. 47%, 40% vs. 27%, respectively). ⁵⁰ The Phase 3 BRUIN CLL‐322 study comparing FD PVR versus VR in RR CLL is ongoing. ⁵¹

Taken together, these studies suggest a future role for triplets in the treatment of RR CLL. Further studies are needed, particularly in patients treated with novel agents (including BTKI/BCL2i) in first‐line therapy, and randomized data are required to determine the optimal duration of therapy. Bendamustine should not be a component of triplets in future studies.

PUTTING IT TOGETHER: CONTEXT MATTERS

Table 2 summarizes the major pros and cons of triplet therapy in CLL. At present, we believe that the clinical application of triplets should be limited to patients with treatment‐naive, TP53 wild‐type CLL (as in the Phase 3 trials). Specifically, we would prioritize consideration of a triplet in patients who meet the following three criteria:

• 1. Disease: Patients with IGHV‐U: There is a consistent 15%–17% 3‐year PFS benefit compared with doublet therapy (VO or AV) in this group. For IGHV‐M patients, there was only a 7% benefit in 5‐year PFS in CLL13, no benefit demonstrated in AMPLIFY, ⁹ , ²¹ and outcomes are already excellent with the Ven‐O time‐limited doublet.
• 2. Patient: Age < 70 without major comorbidities and fully immunized: In this age group, patients reap the most benefit from additional time off therapy and are expected to tolerate the modest increase in infectious events. Specifically, patients should be free of comorbidities that may compromise tolerance of Gr ≥ 3 infection (e.g., serious cardiac disease, airways disease, and unstable diabetes) or predispose to undue risk from BTKI (e.g., pre‐existing arrhythmia or dual anti‐platelet agent use). Ideally, patients should be fully vaccinated (as per recommendations for immunocompromised patients) well in advance of commencing triplet therapy. ³⁶
• 3. Preference: Patients commencing a triplet should be accepting of the increased short‐term logistical burden and reaction risk of monthly infusions (compared with AV), cardiovascular risks of BTKI use (compared with VO), aware that QOL may be temporarily lower during triplet therapy and prioritize the benefit of longer time off therapy.

Table 2.

Pros and cons of triplet therapy in frontline treatment of chronic lymphocytic leukemia (CLL).

Pros	Cons
− Higher rates of uMRD compared with the FD doublet − Longer PFS compared with FD doublets in the IGHV‐U subgroup − FD therapy with potential for retreatment compared with continuous cBTKI − Low rates of resistance mutations	− Higher incidence of fatal COVID‐19 compared with doublets − Potential increase in Gr ≥ 3 infection and hematological toxicity compared with doublets − Infusion reaction risk (compared with BTKI/BCL2i) − Cardiac toxicity risk (compared with BCL2i/anti‐CD20) − Logistical burden of infusions

Abbreviations: BCL2i, BCL2 inhibitor; BTKI, Bruton tyrosine kinase inhibitor; cBTKI, covalent Bruton tyrosine kinase inhibitor; FD, fixed duration; IGHV‐U, IGHV‐unmutated; PFS, progression‐free survival; uMRD, undetectable minimal residual disease.

FUTURE DIRECTIONS

Before we can confidently define the role of triplets in CLL therapy, the field must address the question of the optimal doublet. Thus far, doublets have been compared to CIT and not to one another.7, 8, 9, 19 There are several anti‐CD20/BCL2i versus BTKI/BCL2i clinical trials underway (CLL17, MAJIC, and CELESTIAL) that will help answer this question.52, 53, 54 The optimal duration of triplet therapy also remains unclear. Whilst MRD is clearly prognostic, its degree of impact is modulated by genomics, and future treatment paradigms may incorporate both of these factors in combination. ⁸ Additionally, alternate means of assessing MRD (e.g., Clonoseq, ctDNA) or prognosticating its impact (e.g., ΔMRD400) are likely to play a role in the future. ³³ In patients with TP53 aberrant disease, randomized triplet studies are required. The CLL16 trial is the only ongoing Phase 3 trial comparing a triplet to a doublet. In this study, high‐risk patients with del(17p), TP53 mutation, or complex karyotype (three or more abnormalities) are randomized to FD VO or IVO (ibrutinib continued from C15–24 for those not achieving uMRD after 12 cycles of triplet therapy). ⁵⁵ Whilst this will provide useful information, the omission of continuous BTKI as a control arm may limit interpretation. In this subgroup, future trials should compare the FD triplet versus FD induction followed by BTKI maintenance versus continuous BTKI. Finally, the incorporation of newer agents into triplet combinations is being explored. Studies including pirtobrutinib triplets and sonrotoclax triplets are underway, ³³ , ³⁴ with future trials likely to incorporate BTK degraders (in place of BTKI) or bispecific antibody (in place of anti‐CD20 mAB).

In conclusion, there is a rapidly expanding pool of data regarding the use of triplets in all lines and genomic subgroups of CLL. The question of crowd or company is nuanced and dependent on both disease and patient factors. For now, we recommend limiting use of triplets to younger, fit patients with IGHV‐U and TP53 wild type. It is unlikely that “one size will fit all,” and future triplet studies should focus on genomic subgroups with a modern control arm.

AUTHOR CONTRIBUTIONS

Sean McKeague: Conceptualization; writing—original draft; writing—review and editing. John F. Seymour: Conceptualization; writing—original draft; writing—review and editing.

CONFLICT OF INTEREST STATEMENT

S.M.—honoraria: Janssen. J.F.S.—honoraria: AbbVie, AstraZeneca, BeiGene, BMS, Gilead, Janssen, and Roche; consulting or advisory role: AbbVie, AstraZeneca, BeiGene, BMS, Genor Bio, Gilead, Janssen, Roche, and TG Therapeutics; speaker's bureau and travel, accommodation or expenses: AbbVie, AstraZeneca, and Roche; research funding: AbbVie, BMS, Janssen, and Roche; patents, royalties, other intellectual property: AbbVie; expert testimony: BMS and TG Therapeutics.

ETHICS STATEMENT

Not applicable as this manuscript is based exclusively on previously published data.

FUNDING

This research received no funding.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.