A Phase 1b Study of Eltrombopag and Azacitidine in Patients with High Risk Myelodysplastic Syndromes and Related Disorders

Background: Azacitidine (AZA) based therapy is a standard of care in IPSS INT-2/High MDS/low blast count AML and CMML-2. Thrombocytopenia is an adverse prognostic factor in MDS; increased severity correlates with haemorrhagic complications and shorter AML progression time, and early platelet recovery after AZA is linked to better outcomes. Accumulation of AZA doses is important to patient response, but initial treatment with AZA can aggravate thrombocytopenia and impact outcome. Eltrombopag (ELT) is a thrombopoietin receptor (TpoR) agonist approved for adult chronic immune thrombocytopenic purpura. ELT has shown efficacy in low risk MDS and reduces thrombocytopenic events in advanced MDS and AML. The ELASTIC trial (ISRCTN05858391) is a single-arm phase 1b dose-finding study examining the safety and tolerability of ELT/AZA in patients with MDS/AML with expansion cohort (n=10) at MTD. We investigated if ELT/AZA would lead to earlier platelet recovery and improve AZA efficacy. Although a Phase 3 study testing ELT/AZA (SUPPORT) was terminated due to futility during the ELASTIC trial, we reasoned that drug sequencing may affect outcome. As ELASTIC patients received an ELT pre-phase, ELASTIC was run to completion.

Methods: Patients with IPSS INT-2 or High MDS/CMML-2/AML <30% blasts were treated in a 3+3 trial design in cohorts of ≥3 patients with 25, 50, 100, 200 or 300mg/day ELT. AZA (75mg/m²) was administered for 7 days of a 28 day cycle. All patients received ELT and AZA and were treated for maximum 6 cycles. ELT was administered from day -7 of cycle 1. ELT continued through cycles 1-2 and 4-6 with no ELT during cycle 3. Patients were evaluated after cycle 1 for dose-limiting toxicities (DLTs).

The primary outcome was safety and tolerability (including establishing the MTD) of ELT in combination with AZA. Key secondary outcomes were the effect of ELT on platelet counts, platelet transfusions and marrow blast percentage. Full blood count was measured weekly over the first 13 weeks and monthly thereafter; bone marrow was taken on day 8 of ELT and at the end of cycles 3 and 6.

Results: Thirty-one patients were recruited from November 2014 to August 2018. Thirty (median age 74, range: 62-86) were evaluable for analysis in the safety population (AML=5, IPSS INT-2 MDS=12, IPSS High MDS=12, CMML-2=1); 1 patient was ineligible post registration. 11/30 (37%) were IPSS good risk, 3/30 (10%) intermediate risk and 14/30 (47%) high risk (2 unknown).

ELT was well tolerated up to 300mg (25mg=5, 50mg=3, 100mg=4, 200mg=4, 300mg=14) and was safely combined with AZA. There were no DLTs in the DLT evaluable patients (n=15). The MTD was established as 300mg (n=14). There were 3 SUSARs, 19 SAEs, 28 SARs and 87 episodes of grade 3/4 AEs. There were 16 episodes of grade 3/4 neutropenia and 11 episodes of grade 3/4 anaemia. Five patients discontinued treatment due to toxicity (25mg=2, 300mg=3), 6 due to death (pneumonia=2, disease-related=4, ischaemic heart disease=1) and 1 withdrew. Platelet responses were seen at ELT 50mg and higher. At the MTD, median platelet count at baseline was 37 (IQR: 25-52), rising to 77 (IQR 50-108.5) and 74 (IQR: 32-430) at cycle 2 and 3 respectively. The median platelet nadir was 20.5 (IQR: 11-25) and 42 (IQR 18.5-51.5) after cycle 1 and 2 dropping to 20 (IQR: 9-160) after cycle 3 when ELT was stopped. Mean platelet transfusions did not change with ELT dose. There was no increase in delays to AZA dosing across the ELT dose cohorts. There was no change in bleeding incidents when patients received ELT. Overall, there was no increase in bone marrow blast percentage from baseline. The cycle 3 marrow aspirate was reached by 27/30 (90%) and 11/30 (37%) reached the cycle 6 marrow aspirate. At the end of 3 cycles there were 21 non-responders and 10 responders (CR=3, marrow CR=4, PR=3, SD=9). At the end of 6 cycles there were 22 non-responders and 9 responders (CR=3, marrow CR=4, PR=2, SD=5). 6/14 (43%) treated at the MTD responded while 3/16 (19%) treated at below the MTD responded.

Conclusions: The combination of ELT with AZA is safe and tolerable. Clinically meaningful responses in platelets were seen in patients receiving 50mg or more ELT. At the MTD, there was a noticeable drop in platelet nadir in cycle 3 when patients were not receiving ELT. No change in platelet transfusion was identified across cohorts. The results suggest that an ELT pre-phase may improve platelet responsiveness in patients treated with AZA and warrants further exploration.