Venetoclax and idasanutlin in relapsed/refractory AML: a nonrandomized, open-label phase 1b trial

Abstract

This phase 1b trial (NCT02670044) evaluated venetoclax-idasanutlin in patients with relapsed/refractory (R/R) acute myeloid leukemia (AML) ineligible for cytotoxic chemotherapy. Two-dimensional dose escalation (DE, n = 50) was performed for venetoclax daily with idasanutlin on days 1 to 5 in 28-day cycles, followed by dosing schedule optimization (n = 6) to evaluate reduced venetoclax schedules (21-/14-day dosing). Common adverse events (occurring in ≥40% of patients) included diarrhea (87.3% of patients), nausea (74.5%), vomiting (52.7%), hypokalemia (50.9%), and febrile neutropenia (45.5%). During DE, across all doses, composite complete remission (CRc; CR + CR with incomplete blood count recovery + CR with incomplete platelet count recovery) rate was 26.0% and morphologic leukemia-free state (MLFS) rate was 12%. For anticipated recommended phase 2 doses (venetoclax 600 mg + idasanutlin 150 mg; venetoclax 600 mg + idasanutlin 200 mg), the combined CRc rate was 34.3% and the MLFS rate was 14.3%. Pretreatment IDH1/2 and RUNX1 mutations were associated with higher CRc rates (50.0% and 45.0%, respectively). CRc rate in patients with TP53 mutations was 20.0%, with responses noted among those with co-occurring IDH and RUNX1 mutations. In 12 out of 36 evaluable patients, 25 emergent TP53 mutations were observed; 22 were present at baseline with low TP53 variant allele frequency (median 0.0095% [range, 0.0006-0.4]). Venetoclax-idasanutlin showed manageable safety and encouraging efficacy in unfit patients with R/R AML. IDH1/2 and RUNX1 mutations were associated with venetoclax-idasanutlin sensitivity, even in some patients with co-occurring TP53 mutations; most emergent TP53 clones were preexisting. Our findings will aid ongoing/future trials of BCL-2/MDM2 inhibitor combinations. This trial was registered at www.clinicaltrials.gov as #NCT02670044.

Graphical abstract

Figure 1.

Best percent BM blast reduction and individual patient best responses. ∗Best percent change from baseline in BM blasts is >100. PR, partial response; RD, refractory disease; SD, stable disease.

Figure 2.

Kaplan-Meier curves for OS in the ven-idasa arm. Median OS was 5.1 months among all patients, 13.8 months patients with a CRc response, and 5.7 months in patients with a PR/MLFS response. Data for patients without an event were censored at the date of the last study visit or the last known date to be alive, whichever was later. ∗Other = stable disease, refractory disease, relapse, missing or not done. CI, confidence interval; PR, partial response.

Figure 3.

MCL-1 and BCL-xL PD data. MCL-1 and BCL-xL were assessed by flow cytometry in blood at C1D1, C1D5 pretreatment, and C1D15 pretreatment. Ven-idasa treatment resulted in modest decreases in MCL-1, and to a lesser degree BCL-xL, that trend with dose and response. Groups by treatment include patients who received indicated idasa dose, regardless of ven dose received. D, day; MFI, mean fluorescence intensity; ND, no data; PD, progressive disease; PR, partial response; RD, refractory disease; SD, stable disease.

Figure 4.

Mutational analysis describing (A) heat map of mutations and (B) outgrowth of TP53 mutations on study. In the heat map, detected mutations are indicated by shaded boxes. ∗Mutations that emerged on treatment and were detected at last BM aspirate sampling before therapy discontinuation. Green, favorable mutations; Red, unfavorable mutations (dark red, TP53; light red, RAS signaling mutations). The outgrowth graph indicates the VAF of TP53 mutations that emerged on therapy plotted over time. A total of 25 emergent TP53 mutations in 12 patients were noted on study. Most (22 of 25) emergent TP53 mutations were detectable at low levels in baseline samples either using duplex sequencing or by evaluating binary alignment map files provided by Foundation Medicine. Three were not detected at baseline (red). PR, partial response; ND, no data; RD, refractory disease; SD, stable disease.