Subgroup analysis of efficacy and safety analysis of a randomized, double-blinded controlled phase II study of STA-4783 in combination with paclitaxel in patients with metastatic melanoma

Abstract

8528 Background: STA-4783 (S), an inducer of heat shock protein 70 (hsp70) is a bis-thiobenzoylhydrazide compound. S leads to up-regulation of hsp70 in tumor cell lines. Xenograft models of solid tumors showed synergistic anti-tumor activity in combination with paclitaxel (P). The combination P + S, in phase I and II studies, showed dose-related hsp70 induction (evidence of biological activity) and tolerability. Methods: Eligibility was based on a diagnosis of metastatic cutaneous melanoma, ECOG <=2, and prior treatment with 1 or no chemotherapy regimens. A total of 81 patients (pts) were randomized 2:1 (P 80 mg/m2 + S 213 mg/m2:P 80 mg/m2) 3 weeks out of 4 at 21 US clinical sites. The primary endpoint was progression free survival (PFS); secondary endpoints were response rate (RR), and adverse events (AEs). Results: Based on intent-to-treat analysis, the median PFS was 3.68 months (m) for P + S vs. 1.84 m in the P only arm (p=.035). RR was 15.1% in the P + S arm and 3.6% in the P arm. Subgroup analysis showed chemo- naive pts (n=23) with P + S showed a median PFS of 8.28 m vs. 2.40 in the P arm (n=9). For pts with 1 prior chemotherapy, (n=29), PFS on P + S was 3.12 m vs. 1.77 m on P (n=19). Of 19 pts who crossed over at progression, data are available for 14. PFS ranged from 0.72 to 5.5 m. Three of the 14 evaluable pts treated with P alone had rapid progression (0.95, 1.6, and 1.7 m) then significant inversion of the time to progression with the addition of S to P (2.3, 5.5, and 4.2 m) suggesting study drug effect. Scans were done at identical intervals (8 weeks). The proportion of pts with AEs of grade 3 or higher was 54% (n=52) in the P + S group and 57% in the P group (n=28); pts on P received a median of 2 cycles, while pts in the P + S group received a median of 4. Adverse events leading to discontinuation were low in both groups: 10% for the P + S, and 14% for P. Conclusions: The addition of S to P showed an increase in PFS vs. P alone particularly in chemo-naïve pts. A few pts failing single agent P appeared to benefit from P + S. Despite the additional treatment duration in the P + S group the drugs were well- tolerated, and showed mainly P related adverse events. A phase III study is planned to confirm a role for P + S in metastatic melanoma. [Table: see text]