Clinical Activity and Safety of the Oral ATR Inhibitor Ceralasertib (AZD6738) in Patients with MDS or CMML after Prior HMA Therapy

Background: Myelodysplastic syndromes (MDS) and chronic myelomonocytic leukemia (CMML) progressing after hypomethylating agent (HMA) treatment have very poor outcomes. MDS and CMML frequently harbor mutations in genes that regulate pre-mRNA splicing, including SF3B1, SRSF2, ZRSR2, and U2AF1. These mutations result in increased R-loops, which are resolved via ATR, suggesting ATR inhibition may be a therapeutic target in MDS, particularly in splicing factor mutant (SF mut) disease.

We evaluated the oral ATR inhibitor ceralasertib in MDS and CMML and previously reported that monotherapy (RP2D 160mg PO BID d1-14 q28d) had an overall response rate (ORR) of 27% in patients with relapsed or refractory (R/R) MDS and CMML, while thrombocytopenia was the primary toxicity. We now describe response rates, toxicities, and patient outcomes with an alternate schedule of ceralasertib (NCT03770429).

Methods: We enrolled adult patients with SF mut MDS or CMML, R/R after prior HMA exposure. Patients received ceralasertib 160mg BID on an alternative schedule of days 1-7 and 15-21 of a 28 day cycle (7on/7off). We evaluated this alternative schedule for dose limiting toxicities (DLTs) in the first 6 patients, and enrolled an expansion cohort of 14 patients in total. Responses were assessed according to IWG 2006 response criteria. This dose schedule was compared to the prior RP2D schedule (160mg BID 14on/14off). We assessed overall and progression-free survival for patients enrolled on both dose schedules.

Results: A total of 44 patients were treated, including 30 patients on the 14on/14off schedule, and 14 patients on the 7on/7off alternate schedule of ceralasertib. The 7on/7off patients included 8 LR-MDS and 6 HR-MDS patients. The median age was 76 years (range 71-87) and 12/14 were male (86%).

Reported grade 3 or higher adverse events (AEs) with 7on/7off dosing included febrile neutropenia (n=4), thrombocytopenia (n=2), and pneumonia (n=2), while 1 patient each reported anemia, GI bleeding, folliculitis, infection, sepsis, leukopenia, muscle weakness, dyspnea, and hypoxia. Fewer patients reported grade 4 thrombocytopenia in the 7on/7off schedule (1/14) than the 14on/14off dosing (11/30). No deaths occurred during the 7on/7off schedule, and there were no DLTs. Patients on the 7on/7off schedule were less likely to require a platelet transfusion compared to the 14on/14off schedule (p=0.038). Fewer patients with a baseline platelet count of 50k or higher required a platelet transfusion during the first cycle on the 7on/7off schedule (1 of 10) than the 14on/14off schedule (8 of 16).

The ORR on the 7on/7off schedule was 29% (4 of 14 patients) and included one patient with marrow CR (improvement in BM blasts from 10% to 4%) without hematologic improvement (HI), one patient with RBC transfusion independence (TI) (LTB at baseline requiring 2U / 8 weeks, and achieving TI for 12 weeks), one patient with HI-E, and one patient with HI-N. 4/14 patients had progression to AML, all of whom had 10% or higher marrow blasts at study entry. The efficacy was similar to that reported on the 14on/14off schedule (ORR 27%, p=1.00).

Overall survival (OS) was assessed for the entire study; the median OS was 12 months (95%CI 11, 27). There was no significant difference in OS between patients on the 14on/14off schedule (12mo, median f/u 40mo) vs 7on/7off schedule (13mo, median f/u 10mo; p=0.44). The median PFS for the entire cohort was 5.9mo (95%CI 4, 13) with no difference between dose schedules (p=0.98).

We performed serial gene panel sequencing to assess clonal dynamics during treatment and relapse. Overall clonal structure was relatively stable during the treatment period, including splicing factor mutations. However, at the time of progression, RUNX1 mutation VAFs were enriched compared to enrollment (expanded n=6, new n=1), and generally were part of the dominant clone. Serial testing of serum inflammatory markers and cytokines is ongoing.

Conclusion: Ceralasertib is an oral ATR inhibitor with clinical activity in patients with post-HMA MDS and CMML. The schedule of 160mg BID 7on/7off had similar response rates, OS, and PFS compared to a 14on/14off schedule, with a lower rate of severe thrombocytopenia, and this schedule merits further evaluation. At the time of progression, RUNX1 mutations were enriched; future efforts are ongoing to evaluate the mechanisms of response to ATRi in MDS and CMML.