The Lysine-Specific Demethylase 1 (LSD1) Inhibitor Tranylcypromine (TCP) in Combination with ATRA Is Tolerable and Has Anti-Leukemic Activity in Adult Patients with Relapsed/Refractory AML and MDS

Introduction:

Tranylcypromine (TCP) is an irreversible monoamine oxidase inhibitor, a potent antidepressant that has been in use since the 1960s. Additionally, TCP has been demonstrated to inhibit lysine-specific histone demethylase 1A (LSD1), which is highly expressed in AML (Lee 2006; Berglund 2008). Preclinical studies combining TCP and ATRA induced differentiation and impaired clonogenic survival in non-APL AML cell lines and primary patient samples. These findings were supported by mouse xenograft models (Schenk 2011). Based on this preclinical work, we pursued an investigator-initiated Phase 1 study of this combination at the University of Miami Sylvester Comprehensive Cancer Center (NCT02273102).

Methods:

A Phase 1 study was initiated to evaluate the safety, PK/PD, and preliminary clinical activity of TCP in combination with ATRA in patients (pts) with relapsed/refractory AML and high-grade MDS. The study followed a traditional 3+3 dose escalation design. Safety for all pts and efficacy for all evaluable pts to date are reported. All adverse events were recorded per NCI CTCAE v4.03. All pts received continuous daily dosing of both ATRA (45 mg/m2 in divided doses) and TCP (3 escalating dose levels: 10mg BID, 20mg BID and 30mg BID), with a 3-day lead-in of TCP only during cycle 1. Cycles were 21 days and pts were allowed to remain on study until progression or unacceptable toxicity.

Results:

At the time of data cutoff, 15 pts had received therapy with combination TCP/ATRA (8 AML and 7 MDS). Median age was 74 years, 40% were female and 67% white/27% Hispanic/7% black. Overall, the combination was well tolerated, with the majority of treatment emergent adverse effects (TEAEs) Grade 1 and 2. The most common TEAEs (all grades, ≥20%) included dry mouth (33%), dry skin (27%), febrile neutropenia (27%), dizziness (27%), fatigue (27%), headache (27%), rash (27%), increase in creatinine (27%); and vomiting, nausea, diarrhea, infection, urinary frequency and thrombocytopenia all with a frequency of 20%. The most common Grade 3/4 TEAEs included febrile neutropenia (27%), thrombocytopenia (20%), sepsis (13%), lung infection (13%) and anemia (13%). There was 1 DLT of dizziness at the TCP 20mg BID dose level (out of 8 pts) and 2 DLTs of generalized weakness and nausea/vomiting, respectively, at 30mg BID (out of 3 pts). All DLTs were grade 2, but persistent and poorly tolerated. Therefore, TCP 20mg BID was determined to be the MTD and selected as the RP2D. Best evaluable responses per modified IWG/ELN criteria included 5 pts with prolonged stable disease (all 3 months or more) (2 AML, 1 CMML, 2 MDS), 1 marrow CR (MDS) and 1 MLFS (AML). Three of the 4 MDS/CMML responders had hematologic improvement (HI) (2 HI-P and 1 HI-P and HI-E). One AML pt also recovered neutrophils (0.62 to 14.75) with a decrease in blasts but did not meet response criteria. The 2 pts with best response of marrow CR and MLFS continued on study for 7 and 10 months, respectively. Importantly, these 2 pts and a third pt who had prolonged SD (5 months) plus HI-P/HI-E were all taken off study for cumulative skin toxicity (not progression), and the marrow CR and MLFS pts are both still alive.

Conclusions:

TCP/ATRA combination therapy has demonstrated an acceptable safety profile in pts with R/R AML and MDS, and additionally has demonstrated clinical activity. TCP 20mg BID is the RP2D, and a phase 1 dose expansion at this dose level is ongoing. In responders, skin toxicity may be treatment duration-limiting due to continuous exposure to ATRA, and an intermittent ATRA schedule after cycle 4 may be pursued for the phase 2 study. Additional data will be presented at the meeting, including myeloid mutational analysis, RNA-seq and ATAC-seq, in order to delineate pre- and post-treatment molecular profiles and chromatin accessibility in these pts. Preliminary data (not shown) suggest that a baseline gene expression pattern may predict sensitivity or resistance to TCP/ATRA.