Outcomes with Subsequent FLT3-Inhibitor (FLT3i) Based Therapies in FLT3-Mutated (mu) Patients (pts) Refractory/Relapsed (R/R) to One or More Prior FLT3 Inhibitor Based Therapies: A Single Center Experience

Background: The addition of midostaurin to induction therapy has become a standard approach in newly diagnosed FLT3-mu AML. Second gen FLT3i's quizartinib and gilteritinib demonstrated single-agent CR/CRi rates of 45-55% in pts with R/R FLT3-mu AML in phase II/III trials. The majority of pts in these trials were prior FLT3 inhibitor naïve, a population likely to become obsolete soon. The response rates to second and third FLT3i exposure remain poorly defined. This will be an important clinical question as multiple FLT3i are approved.

Methods: Adult pts with FLT3-ITD mu AML who received at least one FLT3i based therapy at MDACC from 2007 to 2018 were included. Single agent FLT3i and FLT3i combinations with cytotoxic chemotherapy (CCT) and low intensity therapy (LIT) (hypomethylating agents and low-dose cytarabine) were included. FLT3 testing was performed as previously reported (Luthra R et al., Haematologica 2014). All FLT3-ITD positive pts were included irrespective of the allelic frequency.

Results: 217 pts with FLT3-ITD mu AML received one FLT3i therapy, 109 pts received a second FLT3i, 29 received a third FLT3i, and 4 received a fourth FLT3i (Table.1). Overall response rates (ORRs) (ORR = CR+CRp+CRi+PR) were 49%, 27%, 17% and 25% with the first, second, third, and fourth FLT3i (composite CR (CRc) rates of 49%, 27%, 13% and 25%), respectively (Table.1A). The median OS was 8.7, 4.2, 3.5 and 3.9 months with the first, second, third, and fourth FLT3i, respectively.

In the first FLT3i exposure, the ORRs with CCT, LIT, and single-agent FLT3i were 57%, 62%, and 32%. The median (med) prior AML treatment in this group was 1 (range, 0 - 7), however none had received a prior FLT3i. In the second FLT3i exposure, the ORRs with CCT, LIT, and single-agent FLT3i were 27%, 46%, and 15%, respectively. The med prior AML treatments in this group were 3 (range, 1 - 8), including one prior FLT3i. In the third FLT3i exposure, the ORRs with CCT, LIT, and single-agent FLT3i were 0%, 29%, and 18%. The med prior AML treatments in this group were 4 (range, 2 - 10), including two prior FLT3i's. Only 4 pts received a fourth FLT3i and 1 had a CRi.

Analyzing by individual FLT3i, 359 FLT3i's were used at different time points, excluding duplication of FLT3i's in any pt. The most frequently used was quizartinib (131 pts, 37% of all pts). Quizartinib was used in combination with LIT (63 pts) or as a single-agent (68 pts) only, with ORRs in all FLT3i exposure groups of 73% and 38% (CRc 70% and 38%), respectively. In the first (n=102), second (n=23), and third (n=6) FLT3i exposures, ORRs with quizartinib-based therapies were 59%, 39%, and 50% (CRc 59%, 35%, and 33%), respectively (Fig.1B). Quizarinib is a potent FLT3i. Switching to another FLT3i after failing first line quizartinib (n=33) still produced ORRs of 24% (CRc 21%).

The second commonest used FLT3i was sorafenib (108 patients, 30% of all pts). Sorafenib was used in combination with CCT (n=38), LIT (n=57), and as single-agent (n=13) with ORRs in all FLT3i exposure groups of 54%, 42%, and 7% (CRc 54%, 39%, and 7%). Single agent sorafenib was not as effective. In the first (n=69), second (n=33), and third (n=6) FLT3i exposures, ORRs with sorafenib-based therapies were 49%, 30%, and 17% (CRc 46%, 30%, and 17%), respectively. Quizartinib-based therapies had an ORR of 37% (32% CR/CRi) as second line post-sorafenib (Fig.1C).

Gilteritinib was used in 12 pts as a single-agent. 3, 8 and 1 pts received gilteritinib in first, second and third FLT3i exposures with ORRs of 67%, 38% and 100% (all CRc responses), respectively.

Midostaurin was always used in combination, with CCT (n=2) or LIT (n=10) with ORRs of 50% and 40% (all CRc responses) in all FLT3i groups, respectively. In this small set, post-midostaurin ORRs to second FLT3i's were 57% (all CRc).

Conclusions: ORRs dropped from 49% to 27% to 17% with first, second, and third FLT3i-based therapies. Combining FLT3is with low or high intensity chemotherapy appeared superior to single agent in all exposure groups. LIT (HMA and LDAC)-based combinations had encouraging ORRs of 46% and 29% in the second and third FLT3i exposure, likely because many pts received CCT-based combinations as initial therapy. Quizartinib and gilteritinib were effective with ORR of 35-40% even when used as a second FLT3i. This is a first attempt at identifying benchmark response rates for second and third FLT3i exposures, for developing novel FLT3i combinations and expectations with sequential FLT3i usage.

View largeDownload slide

View largeDownload slide

Close modal