A phase I /II Trial of Erlotinib in Higher Risk MDS After Azacitidine (AZA) Failure

Prognostic of IPSS high and int-2 risk MDS after azacitidine failure or relapse is very poor, with a median survival < 6 months (Prebet, JCO 2011). We previously showed that Erlotinib (ERLO), an oral inhibitor of EGF-R tyrosine kinase approved for the treatment of several solid tumors, also has “off target” in vitro effects in myeloid leukemic cell lines and primary blast cells of higher risk MDS and AML (Boehrer, Blood 2008). ERLO induced, cell cycle arrest, myeloid differentiation, apoptosis and also reduced the growth of xenografted human AML cells. We therefore assessed safety and efficacy of ERLO in higher risk MDS with AZA failure in a phase I/II trial.

This trial (NCT 01085838), conducted by the GFM, had the following inclusion criteria: (1) MDS according to WHO classification (also including RAEBt /AML with 20–30% blasts), (2) IPSS Int-2 or High, (3) marrow blasts >10% and <30% (4) no response to treatment with AZA for at least 6 cycles, or relapse after a response. Patients with CMML, prior history of MPN, HIV infection, concomitant treatment with CYP3A4 inducers or inhibitors were excluded. The phase I part was a 2 cohort escalating dose study of ERLO at 100 mg/d in the first cohort, and 150 mg/d in the second. Response was evaluated according to IWG2006 criteria after 12 weeks of treatment. In the absence of limiting toxicity or progression at 12 weeks, treatment was to be continued until progression.

Between July 2010 and July 2012, 29 patients (19M, 10F) were included in 9 centers. Median age was 77 years (range 65–85). Median number of cycles of AZA before inclusion was 18 (5–41). Before AZA, 6 pts had received intensive chemotherapy (with 3 CR, 1 PR and 2 failures) and 1 Lenalidomide, without response. At inclusion, 17 pts had RAEB-2 and 12 had RAEB-t/AML with 20–30% blasts. IPSS cytogenetic risk was poor in 6 pts (21%) including 5 complex, intermediate in 9 pts (31%) and good in 8 pts (38%, all normal karyotype), while 6 pts had cytogenetic failure.

During the phase I part, five patients received ERLO at 100 mg/day, and 5 pts ERLO at 150 mg/d. After review of toxicities by the independent DSMB, the ERLO dose of 150 mg/d was retained for the phase II part, that included 19 patients, of who 7 had not yet completed 12 weeks of treatment and were too early for evaluation.

Overall, 22 patients were evaluable (including 6 AZA failure and 16 relapses after response to AZA) 5 of them had early death (before week 12) due to disease progression (n=2), sepsis (n=3) and 9 additional pts stopped treatment before week 12 due to grade 2 skin rash (n=2), bleeding (n=1), fatigue (n=1), investigator's decision(n=1), consent withdrawal (n=1), disease progression (n=3).

Three patients, ie 13.6% of the 22 evaluable, and 37.5% of the 8 pts who received at least 12 weeks of treatment achieved IWG2006 response (2 at 100mg/d and 1 at 150 mg/d) including 1 marrow CR, 2 hematological improvement (1 HI E and 1 HI P). Six pts had stable disease without HI, and 13 pts progressed. All 3 responders had RAEB-2, with normal, int and failed karyotype, respectively. Two had received intensive chemotherapy before AZA (with 1CR and 1PR), 2 had relapsed after AZA response and 1 had not responded to AZA. Response duration was 3.4 (pt with marrow CR) 5.3 and 11.7+ months, respectively. Survival of the 3 responders from inclusion was 6.1,15 and 14.5 + months, respectively. Median OS from inclusion was 4 months. In the 100 mg/day cohort, 7 SAE occurred (3 sepsis, 1 grade I liver toxicity, 1 FUO and 1 grade III anorexia). No hematological toxicity was described and no DLT was seen. In the 150 mg cohort, 5 SAE occurred (3 sepsis, 1 bleeding and 1 grade III diarrhea). The most common toxicities reported were grade I rash (80%) and grade I diarrhea (30%). Bone marrow cells were collected in all the patients at inclusion and at week 12, and study of in vivo ex vivo correlations is in progress.

Erlotinib at 100 to 150 mg/ d is well tolerated in higher risk MDS resistant to AZA. A few (13.6% in the whole cohort ) significant responses occurred including 1 marrow CR, 1 HI E and 1 HI P. Based on our previous in vitro work showing that ERLO can potentiate the effect of AZA on MDS and AML cells, particularly by inhibiting drug efflux via ABC transporters. (Lainey et al, ASH2010,abstract 974,) combinations of AZA and ERLO could be considered in the treatment of higher risk MDS.

Off Label Use: Erlotinib is approved to treat non small cell lung cancer and pancreatic cancer.