Turning to the efficacy outcomes, Dr. Levis reported that the trial did not meet its primary endpoint, with no significant difference in RFS between the gilteritinib and placebo arms, at a hazard ratio of 0.679 (P = .0518). There was also no significant difference in the key secondary objective of overall survival, at a hazard ratio of 0.846 (P = .4394).
However, Dr. Levis noted that there was a “clear difference in the benefit of gilteritinib by region,” and, “at every level,” MRD predicted a benefit from gilteritinib, which he said was a “big surprise” and “really leapt out in the subgroup analysis.”
He explained that the researchers used a modified version of a two-step assay that has been used in previous studies, and was able to detect MRD at a sensitivity of approximately 1x10-6. “In our study, 98% of participants had samples pre- and post-[transplant].”
Regardless of treatment arm, MRD positivity measured at that sensitivity was associated with a significant reduction in overall survival, at a hazard ratio versus MRD– status of 0.514 (P = .0025).
When stratifying the patients by MRD status, the researchers found that, among MRD+ participants, gilteritinib was associated with a significant improvement in RFS, at a hazard ratio versus placebo of 0.515 (P = .0065), while there was no significant difference in MRD– patients.
Stratifying the patients by their conditioning regimen prior to HCT also revealed differences, with those undergoing myeloablative conditioning having significantly greater overall survival than those who underwent reduced-intensity conditioning, at a hazard ratio for death of 0.529 (P = .0027).
Dr. Levis said there is “no surprise there,” and the result could reflect the selection of fitter, younger patients to undergo the more intensive regimen.
He then showed that MRD+ patients who had undergone myeloablative conditioning had better overall survival with gilteritinib than placebo, at a hazard ratio for death of 0.418 (P = .0087). Again, the difference disappeared when looking at MRD– patients.
“So conditioning doesn’t help you in the setting of MRD,” Dr. Levis said.
Finally, he took a deeper dive into the regional differences in outcomes, noting that patients in the Asia/Pacific region, where gilteritinib showed no benefit over placebo, “were 10 years younger” than those in other regions, “tended to get myeloablative conditioning, and hardly ever used FLT3 inhibitors.”
In contrast, North American patients, who experienced a significant gilteritinib benefit in terms of RFS, underwent HCT an average of 26 days earlier than those elsewhere, and received fewer courses of chemotherapy pre-HCT. Moreover, 93.5% received an FLT3 inhibitor pretransplant.
The study was funded by Astellas Pharma Global Development. Dr. Levis declares relationships with Abbvie, Amgen, Astellas, Bristol-Myers-Squibb, Daiichi-Sankyo, GlaxoSmithKline, Jazz, Menarini, Pfizer, Sumitomo-Dainippon, Syndax, Takeda. Dr. Brunstein declares no relevant relationships.