AMSTERDAM – For treatment-naive patients with acute myeloid leukemia (AML) who are ineligible for chemotherapy, guadecitabine offers similar efficacy to other standard treatment options until four cycles are administered, after which guadecitabine offers a slight survival advantage, based on results from the phase 3 ASTRAL-1 trial.
Complete responders also derived greater benefit from guadecitabine, a new hypomethylating agent, reported lead author, of the Hôpital Saint Louis, Paris.
With 815 patients, ASTRAL-1 was the largest global, randomized trial to compare low-intensity therapy options in this elderly, unfit population – specifically, patients who were at least 75 years old or had an Eastern Cooperative Oncology Group (ECOG) performance status of 3 or more, Dr. Fenaux said at the annual congress of the European Hematology Association.
They were randomized in a 1:1 ratio to receive guadecitabine or one of three other treatment options: azacitidine, decitabine, or low-dose cytarabine. The coprimary endpoints of the trial were complete response rate and median overall survival. Safety measures were also investigated.
A demographic analysis showed that almost two-thirds of patients were at least 75 years old (62%), and about half had an ECOG status of 2 or 3, or bone marrow blasts. Approximately one-third of patients had poor-risk cytogenetics and a slightly higher proportion had secondary AML.
After a median follow-up of 25.5 months, patients had received, on average, five cycles of therapy. However, many patients (42%) received three or fewer cycles because of early death or disease progression. This therapy cessation rate was similar between the guadecitabine group (42.4%) and the other treatment group (40.8%).
The study failed to meet either coprimary endpoint across the entire patient population. Median overall survival was 7.10 months for guadecitabine versus 8.47 months for the other treatments, but this difference was not statistically significant (P = .73). Similarly, the complete response rate was slightly higher for guadecitabine (19.4% vs. 17.4%), but again, this finding carried a nonsignificant P value (P = .48).
The benefit offered by guadecitabine was realized only with extended treatment and in complete responders.
Patients who received a minimum of four cycles of guadecitabine had a median overall survival of 15.6 months, compared with 13.0 months for other treatments (P = .02). This benefit became more pronounced in those who received at least six cycles, which was associated with median overall survival of 19.5 months versus 14.9 months (P = .002). Complete responders also had extended survival when treated with guadecitabine, although this benefit was of a lesser magnitude (22.6 vs. 20.6 months; P = .07).
Most subgroup analyses, accounting for various clinical and genetic factors, showed no significant differences in primary outcomes between treatment arms, with one exception: TP53 mutations were associated with poor responses to guadecitabine, and a lack of the TP53 mutation predicted better responses to guadecitabine.
Adverse events were common, although most measures were not significantly different between treatment arms. For example, serious adverse events occurred in 81% and 75.5% of patients treated with guadecitabine and other options, respectively, while grade 3 or higher adverse events occurred in 91.5% of guadecitabine patients and 87.5% of patients treated with other options, but neither difference was statistically significant.
Adverse events leading to death occurred in 28.7% of patients treated with guadecitabine versus 29.8% of other patients, a nonsignificant difference. In contrast, Dr. Fenaux noted that patients treated with guadecitabine were significantly more likely to develop febrile neutropenia (33.9% vs. 26.5%), neutropenia (27.4% vs. 20.7%), and pneumonia (29.4% vs. 19.6%).
“In those patients [that received at least four cycles], there seemed to be some advantage of guadecitabine, which needs to be further explored,” Dr. Fenaux said. “But at least [this finding] suggests once more that for a hypomethylating agent to be efficacious, it requires a certain number of cycles, and whenever possible, at least 6 cycles to have full efficacy.”
The study was funded by Astex and Otsuka. The investigators reported additional relationships with Celgene, Janssen, and other companies.
SOURCE: Fenaux P et al. EHA Congress, .