Frontline Guadecitabine Falls Short in Phase III AML Study

Mohammad Azab, MD

Guadecitabine (SGI-110) did not meet the coprimary endpoints of the phase III ASTRAL-1 study comparing the hypomethylating agent with physician’s choice of standard therapy in treatment-naïve adult patients with AML who are not candidates for intensive induction chemotherapy.

Astex Pharmaceuticals and Otsuka Pharmaceutical announced in a press release that guadecitabine failed to improve complete response rate (P >.04) and overall survival (OS; P >.01)—the 2 primary endpoints of the study—compared with physician’s choice of azacitidine, decitabine, or low-dose cytarabine.

Investigators continue to evaluate secondary outcome measures from the trial, as well as safety. Data from the study will be presented at an upcoming medical conference, according to Astex and Otsuka. Additionally, 2 ongoing trials are examining guadecitabine in the relapsed/refractory setting—ASTRAL-2 (NCT02920008) in AML and ASTRAL-3 (NCT02907359) in myelodysplastic syndromes and chronic myelomonocytic leukemia.

“We are disappointed in the outcome of the ASTRAL-1 study,” Mohammad Azab, MD, Astex’s president and chief medical officer, said in a statement. “The study used very strict criteria of ineligibility to receive intensive chemotherapy based on age (over 75 years) or poor performance status (ECOG PS of 2 or 3) or comorbidities, which made it a difficult population to show superior benefit of guadecitabine,” added Azab.

Early-phase results had been positive for guadecitabine in AML. More than half of elderly, treatment-naïve patients with AML assigned to treatment with guadecitabine had a composite complete response irrespective of dosage or schedule, according to results from a phase 1/2 study published in The Lancet Oncology.

Hagop M. Kantarjian, MD, with MD Anderson Cancer Center and colleagues recruited 107 treatment-naïve patients at 14 medical centers in the United States for this multicenter, randomized, open-label study. From August 2012 to September 2014, patients were randomly assigned to 60 mg/m² (n = 26) or 90 mg/m² (n = 28) of guadecitabine on days 1 to 5 (5-day schedule) of a 28-day cycle, or 60 mg/m² (n = 53) on a 10-day schedule, also in a 28-day cycle. Median age was 77 years and median follow-up was 953 days.

Composite complete response was defined as complete response, complete response with incomplete platelet recovery, or complete response with incomplete neutrophil recovery regardless of platelets.

The number of patients achieving a composite complete response was similar for all dosing groups or schedules: 13 of 24 treated patients [54%; 95% CI, 32.8-74.4] receiving 60 mg/m² on the 5-day schedule; 16 of 27 treated patients [59%; 95% CI, 38.8-77.6] receiving 90 mg/m² on the 5-day schedule; and 26 of 52 treated patients [50%; 95% CI, 35.8-64.2) receiving 60 mg/m² on the 10-day schedule.

Eligible patients were not candidates for chemotherapy either due to age ≥75 years or a combination of age ≥65 years or at least one of the following criteria: poor cytogenetics, secondary AML after a previous diagnosis of a myelodysplastic syndrome, ECOG performance score of 2, or poor cardiopulmonary function that was unrelated to leukemia. Patients with acute promyelocytic leukemia were excluded.

All patients were recommended to receive at least 6 cycles of therapy and to continue treatment until disease progression or unacceptable toxicity.

At the time of database lock on August 24, 2016, 15 patients were still being followed for survival, including 6 who were continuing treatment. Overall median follow-up was 31.8 months, 33.9 months for patients on the 5-day schedule and 23.5 months for patients on the 10-day schedule. Investigators said the clinical data cutoff date of June 13, 2016, was the date of the last patient observation included in the database for patients in both schedule groups, so the length of follow-up was longer for patients on the 5-day schedule based on the enrollment dates.

Median OS was 316 days (95% CI, 160-420) for the 5-day cohort and 44 patients died. Median OS was 284 days (95% CI, 140-478) in the 10-day cohort with 42 deaths.

Median OS was 574 days (95% CI, 420-907), with 24 deaths, for 37 patients who achieved a complete response and 476 days (95% CI, 108-510), with 17 deaths, for 18 patients who achieved a complete response with incomplete platelet recovery or with incomplete neutrophil recovery regardless of platelets. Median OS was 93 days (95% CI, 68-150) for 48 patients with no response. There were 45 deaths in that group. OS for patients with a composite complete response was longer than for patients with no response (P <.0001).

Among responders, median time to best response was shorter for the 10-day cohort than for the 5-day cohort (69 days vs 89 days). Median duration of response was similar for the 5-day and 10-day cohorts (186.0 days vs 269.5 days).

Patients in the 5-day cohort reached complete response in a median of 4 cycles of treatment compared with 3 in the 10-day group.

Posthoc exploratory multiple logistic regression analysis did not detect any significant predictors for composite complete response, and multiple Cox regression did not reveal any significant predictors for OS.

Almost 80% of patients received the planned dose in all treatment cycles. Of those who received treatment in cycles 2 and 3, no patient on the 5-day schedule had a dose reduction. In the 10-day schedule arm, 5% of patients in cycle 2 and 9% of patients in cycle 3 had dose reductions. The proportion of dose reductions increased with subsequent cycles. In cycle 6, 23% of patients on the 5-day schedule and 22% of those assigned to the 10-day schedule had a dose reduction.

According to the researchers, the most common grade ≥3 adverse events were febrile neutropenia (61% [31/51] of patients receiving the 5-day schedule vs 69% [36/52] of patients receiving the 10-day schedule), thrombocytopenia (49% vs 42%), neutropenia (39% vs 35%), pneumonia (29% vs 37%), anemia (29% vs 23%), and sepsis (16% vs 27%).

Kantarjian HM, Roboz GJ, Kropf PL, et al. Guadecitabine (SGI-110) in treatment-naive patients with acute myeloid leukaemia: phase 2 results from a multicentre, randomised, phase 1/2 trial. Lancet Oncol. 2017;18(10):1317-1326. doi: 10.1016/S1470-2045(17)30576-4.