Outcomes By Frailty in IC-Ineligible Newly Diagnosed AML – a Real-World Evidence Comparison By Treatment with Hypomethylating Agents + Ivosidenib or Venetoclax

BACKGROUND:

Nearly 50% of newly diagnosed (nd) patients with AML are ineligible for intensive chemotherapy (ICi) and are commonly referred to as “unfit”. However, the ICi population is heterogeneous, and evidence supporting treatment success by level of frailty is limited in the setting of lower intensity therapies.

Our prior study of patients with nd ICi AML and mutant IDH1 (mIDH1) treated with hypomethylating agents (HMA) + venetoclax (VEN) vs HMA + ivosidenib (IVO) found superior efficacy for HMA+IVO. Here we report outcomes for cohorts based on baseline characteristics indicative of relative frailty.

METHODS:

This retrospective chart review included ICi patients with nd mIDH1 AML, initiated in US community and academic settings with either IVO+HMA or VEN+HMA between Nov 2021 through Nov 2022.

Data included baseline patient and practice characteristics, treatment patterns, and safety and efficacy outcomes. “Less frail” was defined as age <65, or age 65-74 with ECOG 0-1. Endpoints included response rates, event-free survival (EFS), tolerability, bridge to transplant, and unscheduled acute care use. Multivariate regression was performed to adjust for differences in baseline characteristics.

RESULTS:

Of 280 patients, 181 received IVO+HMA and 99 VEN+HMA. Baseline characteristics were similar, median age was 61 (IQR 53-71) for IVO+HMA and 65 (55-75) for VEN+HMA with median follow-up from treatment of 7.1 mo and 8.8 mo, respectively. The “less frail” population comprised 147 patients age <65 and 53 patients age 65-74 with ECOG 0-1, a total of 200 patients (71%); the remaining 80 patients were “more frail”.

CR rates were 41.0% vs 26.3% (p=0.022) for those less frail vs more frail, respectively, while composite CR/CRi rates were 65.5% vs 38.8% (p<0.001). Median times to best response were 3.6 and 3.3 mo. A competing risks regression for 6-mo EFS (composite CR/CRi within 24 weeks, and no relapse or death) favored the less frail cohort (57.5% vs 30.0%, HR 0.55; p<0.0001). Among less frail patients, 22 (11.0%) had bridge to transplant, vs 1 (1.3%) more frail patient (p=0.0271).

VEN+HMA was used for 63 patients in the less frail cohort (31.5%), with 32 (50.8%) receiving ≤7 days (d) of VEN per 28-day cycle, 13 (20.6%) receiving 8-21 d per cycle, and 18 (28.6%) receiving VEN 22-28 d per cycle. VEN+HMA was received by 36 (45.0%) of the more frail patients, with 10 (27.8%) receiving ≤7 d, 18 (50%) with 8-21 d, and 8 (22.2%) receiving 22-28 d. In both cohorts, >90% of VEN schedule reductions occurred at initiation, prior to achievement of remission. IVO+HMA comprised the remaining 137 less frail and 44 more frail patients.

Efficacy outcomes were stratified by treatment among the less frail population. IVO+HMA CR rates were 47.4% vs 27.0% (p=0.007) for VEN+HMA. Composite CR/CRi rates were 69.3% vs 57.1% (p=0.093), respectively. A competing risks regression for median time to CR favored IVO+HMA vs VEN+HMA (HR 1.78 CI (1.00-3.15), p= 0.0485), resulting in improved 6-mo EFS with IVO+HMA (64.2% vs 42.9%, p=0.005). Bridge to transplant was more common with IVO+HMA (20 [14.6%] vs 2 [3.2%]); p=0.029. Results were consistent among the less frail after adjusting for baseline differences. In the smaller more frail cohort, findings were directionally similar in favor of IVO+VEN. Regardless of frailty, patients receiving VEN+HMA had greater relative risk of unscheduled acute care days 1-84 (less frail 73.0% vs 40.9%, p<0.001; more frail 66.7% vs 31.8%, p=0.002).

CONCLUSIONS:

Efforts to assess the impact of level of frailty among patients with nd AML ineligible for IC confirmed expectations that the subgroup of patients considered to be less frail had higher CR and composite CR/CRi, achieved best response faster, and had longer EFS than those more frail. With respect to treatment, IVO+HMA showed improved response and EFS compared with VEN+HMA in the less frail cohort. Patients receiving VEN were often treated with <21 days and this may have impacted the regimen’s efficacy. Even with most VEN doses being reduced, patients had more unscheduled acute care compared to IVO+HMA. The efficacy and tolerability of IVO+HMA suggest it may compare favorably to VEN+HMA as frontline therapy for patients with ICi nd mIDH1 AML regardless of the degree of frailty.