Resistance to Acalabrutinib in CLL Is Mediated Primarily By BTK Mutations

Introduction

Acalabrutinib (A) is a selective, second-generation irreversible Bruton's Tyrosine Kinase inhibitor (BTKi) that has shown outstanding clinical activity in chronic lymphocytic leukemia (CLL) and other hematologic malignancies. Acquired resistance to the first-generation BTKi, ibrutinib (I), is mediated most commonly by C481S mutations in BTK, which decrease binding affinity and change binding from irreversible to reversible. The mechanism of resistance to A has not been investigated.

Patients and Methods

All patients (pts) treated at The Ohio State University and enrolled on an IRB approved phase 1b/2 study in CLL were included in this analysis. Beginning 12 months (mos) after the initiation of A, pts underwent deep sequencing every 3-6 cycles using a digital droplet PCR assay for BTK C481S or Ion Torrent Sequencing for any BTK or PLCG2 mutations. At relapse, samples from each pt underwent full BTK and PLCG2 sequencing. Correlation between baseline characteristics and relapse was assessed using proportional sub-distribution hazards model, pts who were still on treatment were censored at last follow-up.

Results

105 pts were included in this analysis. 38 (36%) were treatment-naïve (TN), 50 (48%) were relapsed/refractory (RR), and 17 (16%) were previously intolerant of I (intol). The median age was 62 (range 33-84) and median number of prior therapies was 1 (range 0-11). The pts were generally high-risk, with 66% having unmutated IGHV, 24% with del(11)(q22.3), 15% with del(17)(p13.1), and 28% with complex karyotype (≥ 3 abnormalities). With a median follow-up of 47.5 mos (range 37.7-58.5) 30% of pts (n=31) have discontinued therapy, 17 for progression of disease (CLL in 16 pts, Richter's transformation in 1), and 14 for other reasons.

Cumulative incidence rate (CIR) of progression varied significantly by cohort. At 12 mos, CIR of progression ranged from 0% in both the TN and RR cohorts to 11.8% (95% CI 1.8-32%) in those who were intol. By 36 mos, CIR of progression was 5.3% (95% CI 0.9-15.7%) in TN cohort, 6% (95% CI 1.5-15%) in the RR cohort, and 35.3% (95% CI 13.7-58%) in the intol cohort (Figure 1).

In univariable analysis, higher risk of progression was associated with del(17)(p13.1) on baseline FISH (p<0.0001), number of prior therapies (p=0.009), female sex (P=0.01), increased baseline LDH (p=0.02), and presence of Myc abnormality on baseline FISH (p=0.05). Treatment status was associated with risk of progression as well, where TN pts had a similar risk to RR, but intol pts had a higher risk of progression (p=0.02). In multivariable analysis, del(17)(p13.1) (p=0.01) and sex (p=0.03) remained significantly associated with risk of progression after accounting for other variables. When taking into account total time of BTKi exposure, there was no statistically significant increased risk of progression for intol pts (p=0.13), however, at 48 mos of BTKi exposure, those who were only treated with A had a CIR of progression of 17.3% (95% CI: 7.9-29.6%), while those who switched BTKi had a CIR of 36% (95% CI 13.8-59.1%).

All 16 pts with CLL relapse had a sample evaluated for BTK C481S at relapse, and 14 had samples evaluated for full BTK and PLCG2 mutations using Ion Torrent deep sequencing. BTK C481 mutations were found in 11/16 (69%; C481S in 10, C481R and C481Y in 1). One pt with a BTK C481S mutation also had a BTK T474I gatekeeper mutation, and one had co-existing C481R mutation. Two pts with BTK C481S mutations had co-existing PLCG2 mutations previously associated with I resistance at <3% VAF.

103/105 pts were screened every 3-6 cycles for BTK mutations. 22 pts have had mutations detected at a median of 31.6 mos from A initiation. TN and RR pts had median time to mutation of 38.8 and 32.9 mo, respectively. For intol pts median time to mutation detection was 24.8 mos from start of A, 33.8 mos from start of any BTKi. Median time from BTK mutation detection to clinical relapse is 12 mos (95% CI: 4.8-not reached), with no significant difference among cohorts in time from mutation to disease progression.

Conclusions

Here we show for the first time that CLL relapse on A is mediated predominantly by mutations in BTK similar to I. While not unexpected, this is significant as resistant patterns could be different given the more selective nature of A as well as potentially higher BTK occupancy over time due to twice daily dosing. Monitoring for BTK resistance offers the opportunity to intervene clinically before symptomatic disease.

View largeDownload slide

View largeDownload slide

Close modal