Results of a Phase 2, Open-Label Study of Idarubicin (I), Cytarabine (A) and Nivolumab (Nivo) in Patients with Newly Diagnosed Acute Myeloid Leukemia (AML) and High-Risk Myelodysplastic Syndrome (MDS)

Background: The combination of an anthracycline and cytarabine has remained a standard of care induction regimen for newly diagnosed AML pts for more than 40 years. PD-1 positive CD8+ T-cells are increased in bone marrow (BM) of AML pts and blocking PD-1/PD-L1 pathways enhances anti-leukemia effect of cytotoxic chemotherapy by increasing CD8+ T-cells in murine models. Addition of Nivo to IA induction may prolong event-free, relapse-free and overall survival (EFS, RFS, OS; respectively). We report on updated feasibility and efficacy data.

Methods: Pts aged 18-65 yrs with newly diagnosed AML (≥20% blasts by WHO criteria) and high risk MDS (≥10% blasts) were eligible if they had adequate performance status (ECOG ≤2) and organ function. Treatment included 1-2 induction cycles of (A) 1.5 g/m² over 24 hours (days 1-4) and (I) 12 mg/m² (days 1-3). Nivo 3 mg/kg was started on day 24±2 and continued every 2 weeks for up to a yr. For pts achieving complete response (CR) or CR with incomplete count recovery (CRi), up to 5 consolidation cycles of attenuated dose I+A was given. Eligible pts received allogeneic stem cell transplant (alloSCT) at any time during or after consolidation.

Results: 3 AML pts were treated at a run-in phase with Nivo 1 mg/kg without drug-related toxicity. Subsequently, 41 pts were treated as above. Median (med) age was 54 yrs (range, 26-66) with 42 AML (32 de novo, 7 secondary, 3 therapy-related) and 2 high risk MDS. 19 pts had adverse ELN genetic risk with 8 TP53 mutations (Fig 1A).

All pts were evaluable for response: 34 (77%) achieved CR/CRi (28 CR, 6 CRi) and 18/34 (53%) had undetectable MRD by flow cytometry (FC) following induction. Med number of cycles to response was 1 (range 1-2) and med number of total cycles received was 2 (range, 1-6). Counting maintenance schedule, med number of Nivo doses was 2 (range, 0-25) and 15 pts (34%) received ≥4 doses; 4 pts did not receive Nivo: insurance issues (n=2), early death (n=1) and rapid disease progression (n=1). The 4- and 8-week mortality was 5%. At med follow-up of 17.25 mos (range 0.5-30.4), med OS was 18.54 mos (range, 0.5-30.4) and med RFS for the 34 CR/CRi pts was 18.5 mos (range 1.7-25.6); med EFS was not reached (range 0.5-13.7). When compared to a contemporary cohort of pts treated with I+A alone, there is a trend of improvement of OS at a short follow-up duration (med 18.54 vs 13.2 mos; p=0.2).

Six pts had grade 3/4 immune-related toxicities with rash (n=2), colitis (n=2), transaminitis and pancreatitis (n=1; each). Grade 3/4 cholecystitis was possibly attributed to Nivo in 1 pt. 18 (41%) pts proceeded to alloSCT. Donor source was matched related in 3, matched unrelated in 12 and haplo-identical in 3 pts. Conditioning regimen was Fludarabine+busulfan-based in 12, and fludarabine+melphalan in 6 pts. 13 (72%) pts developed graft versus host disease (GVHD)(grades I/II in 8, III/IV in 5), which responded to treatment in 8 (Fig 1B).

Multicolor FC studies were conducted on baseline (prior to 1st Nivo dose) and on-treatment BM aspirate and peripheral blood to assess T-cell repertoire and expression of co-stimulatory receptors and ligands on T-cells and leukemic blasts, respectively. Baseline BM was evaluated on 19 responders and 5 non-responders. Pts who achieved CR/CRi had a trend to higher frequency of live CD3⁺ total T cell infiltrate compared to non-responders in baseline BM aspirates (Fig 1C). We evaluated expression of immune markers on T cell subsets: CD4 T effectors [Teff]: CD3⁺CD4⁺CD127l^o/+Foxp3^-, CD4 T regulatory: CD3⁺CD4⁺CD127^-Foxp3⁺, and CD8 T cells. At baseline, BM of non-responders had significantly higher percentage of CD4 Teff co-expressing PD1/TIM3 (p<0.05) (Fig 1D) and trend towards higher percentage of CD4 Teff cells co-expressing PD1/LAG3 compared to responders (data not shown). TIM3/LAG3 co-expression on PD1⁺ T cells has been linked to an exhausted immune phenotype in AML. Using Cytof mass cytometry, we quantified leukemic progenitor cells and T-cells and demonstrated clearance of progenitors and reconstitution of T-cells in pts achieving CR.

Conclusion: Addition of Nivo to (I+A) induction is feasible and safe in younger AML pts. Post-transplant severe GVHD is not significantly increased and is manageable. CD4+ T-effector cells display exhausted phenotype in non-responders that can be potentially reversed. Studies with earlier initiation of checkpoint inhibitor therapy are planned

Figure.

View largeDownload slide

Figure.

View largeDownload slide

Close modal