A phase II NCCTG/CALGB trial of imatinib (STI571) in patients (pts) with c-kit-expressing relapsed small cell lung cancer (SCLC)

Abstract

Listen

Translate article

7048 Background c-Kit over-expression is found in multiple cancers including up to 70% of SCLC. STI571 can inhibit c-kit tyrosine kinase activity and has high efficacy in chemoresistant c-kit positive (c-kit+) gastrointestinal stromal tumors. A phase II study of STI571 in unselected SCLC pts by Johnson et al., was negative, but inconclusive for proof- of - principle purposes, since 80% of tumors did not express c-kit. We therefore sought to evaluate clinical activity of STI571 in the treatment of patients with recurrent and refractory c-kit+ SCLC. Methods Only c-kit+ SCLC cases were enrolled on this study (≥ 1+ by immunohistochemistry). Arm A included pts progressing < 3 months and Arm B included pts progressing ≥ 3 months after previous treatment. STI571 was administered at a dose of 400 mg bid continuously, with a cycle length of 28 days. A single stage Simon design with a planned interim analysis was used to evaluate the 16 week progression free rate in each arm. Results A total of 29 evaluable pts were entered into the study (7 arm A, median age 68; 22 arm B, median age 65). Median number of treatment cycles was 1 in both arms. All pts are now off active treatment. Grade3+ non-hematologic adverse events were seen in 15(52%) pts, with nausea, vomiting, dyspnea, fatigue, anorexia and dehydration occurring in at least 10% of pts. Median survival was 3.9 and 5.3 months and median time to progression was 1 and 1.1 month for arms A and B, respectively. Enrollment to arm A was temporarily suspended prior to reaching interim analysis due to progression (29%), early deaths (29%), and refusal (42%). No objective responses and no confirmed stable disease ≥ 6 weeks were seen in either arm. Accrual was permanently terminated to both arms as no patient was progression-free at 16 weeks. Conclusion STI571 was inactive in spite of pt selection for c-kit+ SCLC. Our results strengthen the collective evidence that prediction of efficacy of novel therapeutic agents based on target expression, rather than pathway activation (for example through activating mutations), may not be a valid paradigm for drug development. No significant financial relationships to disclose.