Zella 201: A Biomarker-Guided Phase II Study of Alvocidib Followed By Cytarabine and Mitoxantrone in MCL-1 Dependent Acute Myeloid Leukemia (AML): Results of Newly Diagnosed High-Risk Exploratory Arm

Background
Alvocidib is an investigational cyclin-dependent kinase-9 (CDK9) inhibitor that can suppress RNA polymerase II-mediated transcription of genes implicated in leukemia cell survival, including myeloid leukemia cell-1 (MCL-1). MCL-1 is an anti-apoptotic BCL-2 family member that is a key mediator of apoptosis in AML. Alvocidib combined in a timed-sequential regimen with cytarabine and mitoxantrone (ACM) has shown clinical activity in newly diagnosed and relapsed/refractory (R/R) AML through Phase I and II clinical trials. Analysis of bone marrow samples from newly diagnosed AML patients (pts) treated with ACM showed an association of complete remission (CR) with MCL-1 dependence by a BH3 profiling biomarker assay. Zella 201 was initiated based on the hypothesis that ACM may have preferential clinical activity in pts with MCL-1 dependence. We report the findings from an exploratory cohort of newly diagnosed high-risk (NDHR) AML pts with MCL-1 dependence treated with ACM.

Methods
Zella 201 is a biomarker-driven Phase II study of ACM in R/R AML patients with MCL-1 dependence. Stage 1 included a cohort of R/R AML pts with various levels of MCL-1 dependence and an exploratory cohort of NDHR AML with MCL-1 dependence >40%, as determined by a BH3 profiling assay. Eligibility criteria for the NDHR cohort included pts 18-65 years with high-risk AML defined as one of the following: A) treatment-related AML, B) AML from preexisting MDS/MPN, C) adverse-risk by ELN 2017 criteria. Induction therapy consisted of alvocidib 30 mg/m² as a 30 minute IV bolus followed by 60 mg/m² over 4 hours on Days (D) 1-3, cytarabine 667 mg/m²/D by continuous IV infusion D6-8, and mitoxantrone 40 mg/m² IV on D9. Up to 3 additional cycles of the same regimen (with or without mitoxantrone) were permitted in responders. The primary endpoint was CR/CRi. Key secondary endpoints were overall survival (OS), relapse-free survival (RFS), overall response rate and safety.

Results

Thirteen NDHR pts were treated and evaluable in this cohort (Table 1). One pt received alvocidib on days 1-3 and withdrew from the study on day 6 due to grade 4 diarrhea, cytokine release syndrome, and acute kidney injury. This pt was excluded from the efficacy analysis. Median MCL-1 score was 56% (Range: 42-70%). This cohort was influenced by the following poor risk categories: secondary AML (n= 9; 69%), adverse-risk by ELN (n=8; 62%) and TP53 mutations (n=6; 46%). The most common ≥Grade 3 treatment-emergent non-hematologic AEs (n=14) were diarrhea (29%); TLS, hypocalcemia, sepsis, hypotension (21%), pneumonia, colitis, hyperglycemia, anorectal infection, dyspnea, and left ventricular dysfunction (all 14%).

Overall, CR/CRi was 62% with 7 (54%) pts responding following 1 cycle of therapy and another pt achieving CR after a second cycle. Two of six pts with TP53 mutation achieved CR. Although all pts included in this cohort were determined to be MCL-1 dependent, there was no association of CR with increasing MCL-1 dependence. Six (46%) pts went on to an allogeneic stem cell transplant (SCT). Sixty-day mortality was 0%. Median follow-up, OS, and RFS were 8.0, 8.5, and 6.1 months, respectively. Five of 8 (68%) CR/CRi pts have relapsed, and 10 pts (77%) have expired to date. The three pts still alive all received a post-study SCT.

Conclusion

ACM has clinical activity in a limited cohort of NDHR AML pts with MCL-1 dependence scores >40% in a biomarker assay. Despite observed CR rates, duration of CR was modest and overall outcomes were poor. These results are comparable to historical controls with conventional chemotherapy regimens given the high-risk subset (62% of pts had adverse-risk and 46% had TP53 mutations). Further study is warranted to better define subgroups of ND AML pts who may benefit from alvocidib-containing induction regimens.