Phase I/IB Study of Azacitidine and Hedgehog Pathway Inhibition with Sonidegib (LDE225) in Myeloid Malignancies

Background: Lower intensity therapies for patients (pts) with myeloid malignancies (AML, MDS, CMML, MPNs) are beneficial, but suboptimal despite the emergence of potentially active new agents. Even after an initial response to hypomethylating agents (HMAs) [5-Azacitidine (AZA) or decitabine], the majority of patients fail treatment. In large-scale RNAi screens we identified Hedgehog pathway (HP) genes as a potential rational combination to overcome a priori and acquired HMA resistance. Herein we report the first analysis of the final study results of the oral smoothened (SMO) inhibitor, Sonidegib (LDE225) in combination with AZA in 63 pts treated on a Phase 1/1B (Ph1/1B) trial.

Methods: Ph1 (3+3) dose escalation and Ph1B (expansion cohort) study to determine the MTD (Ph1) and best overall response (Ph1B) of Sonidegib (SON) combined with AZA. 2^nd endpoints included AEs, DOR, TTP, OS, time to AML for MDS pts and QOL. Pts with untreated and relapsed/refractory (rel/ref) AML, CMML, MDS, MDS/MPN overlap (MDS/MPN) or myelofibrosis (MF) were eligible. SON was started at dose level (DL) 0 at 400mg oral daily continuously (28 day cycle) and AZA at 75mg/m2 x 7 days or days 5-2-2 with dose adjustment per label. Dose-limiting toxicity (DLT) in Ph1 was assessed after 42 days at steady-state concentrations of SON. Pharmacokinetics were not performed.

Results:

Study population:From 05/2014 to 05/2017 (data cut-off) 63 pts were enrolled and followed for outcome: 10 pts were treated on the Ph1 and 52 in Ph1B part of the trial (1 pt withdrew). Median age was 72.5 years (range 39-86) and 66% pts were male. ECOG 0, 1 and 2 are 34%, 61% and 5% respectively. 45% (n=28) were newly diagnosed and 55% (n=34) rel/ref pts. 38 pts (61%) had AML and 13 of the 15 newly diagnosed/untreated AML pts had 2^nd AML due to AHD; 23/38 pts had rel/ref AML. For all AML pts 21/38 were intermediate risk and 16/38 had unfavorable cytogenetics (1/38 unknown). 18 pts (29%) had MDS, all in the Intermediate, High or Very High risk status (IPSS-R), 4 (6.5%) had CMML and 2 pts MF.

Dose levels, treatment administration and follow up: 6 pts were treated at DL0 SON 400mg/day continuously with AZA and 1 DLT of CPK elevation observed, while another pt had grade 4 neutropenia (at that time included in DLT definition, later removed). Due to overall tolerance of the combination at DL0 and investigator assessment, despite only 1/6 formal DLTs, DL -1 at SON 200mg x28 days was explored without further DLT (n=4 pts) and 52 pts accrued on Ph1B. For the 56 pts treated at the MTD dose of SON 200mg (Ph1B=52 and Ph1=4 pts), after a median of 14.5 months follow up, 218 cycles have been administered (median 3 cycles-range 1-25 ms). Median doses for SON and AZA were 200mg and 75mg/m2 respectively. 12.5% and 18% of pts had dose adjustments and 71.4% (i.e. missed only 1 dose) and 18% dose omissions at least once for SON and AZA respectively. 19/56 (34%) of pts are alive.

Safety and Tolerability : Of 56 pts treated at MTD dose100% experienced at least one grade (gr) 3 AE and 52 (93%) had a gr 4 AE regardless of attribution. 96% and 91% had a gr 3 and gr 4 Heme AE and 73% and 20% had a gr 3 and gr 4 Non-Hem AE respectively. Gr 3/4 AEs were as expected for either drug and the diseases under study and were: thrombocytopenia (n=51), neutropenia (47), anemia (43), leukopenia (41). Most common gr 1/2 AEs were: fatigue (n=48), constipation (29), nausea/cough (each 24), insomnia (19), diarrhea (18), abdominal pain (16). 4/56 pts died on study deemed not drug related (cardiac arrest, MOF, neoplasm, sudden death).

Efficacy: With a median follow up of 14.5 ms the best response outcome for untreated AML/MDS patients was 23.1% and for rel/ref 7.1%. However, the rate of SD (NR) was remarkably high particularly in the rel/ref AML population at 76% (16/21 pts, 2 not yet assessed). This high disease control rate was reflected in an OS of 7.6 ms (4.6-NE) in rel/ref AML, the majority of which failed previous HMAs .

Conclusion: In this, to date largest SMO inhibitor/HMA combination trial, the oral SMO inhibitor SON was safe with full dose AZA in untreated and previously treated AML, MDS/CMML and MF pts. Although remission rates are not superior to single agent AZA in the heavily pre-treated and advanced myeloid malignancy population, the OS and lack of progression are encouraging, particular in the rel/ ref AML population indicating that SMO/Hh pathway inhibition may add to HMA efficacy. Predictive biomarker studies are ongoing.