Indoximod Combined with Standard Induction Chemotherapy Is Well Tolerated and Induces a High Rate of Complete Remission with MRD-Negativity in Patients with Newly Diagnosed AML: Results from a Phase 1 Trial

Background

Successful allogeneic stem cell transplantation (HSCT) in treatment of patients (pts) with acute myeloid leukemia (AML) is dependent upon graft-versus-leukemia, suggesting that intact immune surveillance is essential for eradicating minimal residual disease. Myeloblast-induced T-cell tolerance through overexpression of indoleamine 2,3-dioxygenase (IDO) is thought to play a significant role in immune evasion through upregulation of tryptophan (Trp) catabolism and kynurenine production, resulting in a Trp-poor environment that leads to immune system suppression. Indoximod is a small-molecule inhibitor of the IDO pathway that acts directly on immune cells to reverse IDO pathway-mediated suppression. We are assessing the safety and preliminary efficacy of indoximod in combination with standard induction chemotherapy in patients (pts) with newly diagnosed AML.

Methods

In this open-label, multicenter, phase 1 study (NCT02835729), eligible pts with newly diagnosed AML were treated with indoximod in combination with induction chemotherapy (idarubicin 12 mg/m²/d x3 days with cytarabine 100 mg/m²/d x7 days). Using a "3+3" design, indoximod (600 mg [dose level 0], 1000 mg [dose level 1], 1200 mg [dose level 2]) was given orally every 8 hours (Q8h) starting on day 9 of induction. Regimen limiting toxicity (RLT) was defined as any ≥ grade 3 non-hematologic adverse event (AE) that was not incontrovertibly related to the underlying AML or cytarabine or idarubicin. After induction, pts received up to 4 cycles of high dose cytarabine (HiDAC) consolidation while continuing indoximod. Patients continued on maintenance indoximod for up to 6 months from completion of consolidation therapy. Indoximod was discontinued 4 weeks prior to HSCT in eligible patients and not restarted as maintenance post-HSCT.

Results

As of July 15, 2018, 31 pts were enrolled (median age 55 years, range 18-78; 77% male). Six patients did not proceed with study therapy due to a diagnosis of acute promyelocytic leukemia, issues with medical insurance coverage, consent withdrawal, critical illness, and intestinal myeloid sarcoma preventing oral intake. Pts who received ≥1 dose of indoximod were included in the intention-to-treat (ITT) analysis (n=25), and pts who received ≥80% of their scheduled indoximod doses were included in the per-protocol (PP) analysis (n=19). Reasons for not completing ≥80% of indoximod doses were: consent withdrawal (n=3), inability to swallow (n=2) and physician decision (n=1). Of the 19 PP patients, 16 (84%) had either unfavorable karyotype or adverse mutation profile and 3 (16%) had secondary AML (s-AML). Indoximod combined with induction chemotherapy was well tolerated; no RLT was observed. The most frequent grade ≥3 non-hematologic treatment-emergent AEs in the ITT population, regardless of attribution, were febrile neutropenia (60%), hypoxia (16%), atrial fibrillation (12%), pneumonia (12%), hypocalcemia (12%), and hypotension (12%). Among 25 ITT pts, 21 (84%) achieved a remission (CR/CRh/CRi/CRp), and 15 of 19 (79%) in the PP analysis achieved remission. Among 12 pts with measurable residual disease (MRD) available in remission, 10 (83%) had MRD <0.02% (MRD-neg). Eleven of 19 pts (58%) received ≥1 cycle of HiDAC and 5 (26%) received maintenance indoximod. All 11 patients who received HiDAC #1 became MRD-neg. Median relapse-free and overall survival have not been reached. IDO Composite Scores in bone marrow were calculated by multiplying percentage of stained mononuclear cells by grade of staining intensity determined by 3 independent pathologists. Median composite IDO1 score in tested pt samples (n=11) was 0.76 (range, 0.1-2.2). Expression of IDO1 mRNA at baseline varied significantly among patient samples analyzed (fold changes (FC) range: 0.1-84, normalized to β-Actin expression). IDO1 mRNA was significantly upregulated in post-induction samples compared to baseline (FC range: 1.7-248) in 10 out of 12 paired samples.

Conclusions

Indoximod is well tolerated in combination with standard AML induction therapy. Rates of morphologic response and of MRD-neg status are very promising. The recommended phase 2 dose (RP2D) was 1200 mg oral Q8h and a placebo-controlled randomized phase 2 study is under development.