B212

Background: 17DMAG is a water-soluble and potent geldanamycin analog, which inhibits HSP90. The objectives of this first-in-human study were to: establish the dose-limiting toxicity (DLT); recommend a phase 2 dose; & characterize the PK & PD of 17DMAG.
 Methods: Patients (pts) were accrued to a modified accelerated schema. 17DMAG was given IV over 1 h daily x 5 (schedule A) or daily x 3 (schedule B) every 3 weeks. Plasma 17DMAG concentrations during cycle 1 were quantitated by LC/MS assay. HSP27, HSP70, and 4 client proteins CDK4, RAF-1, AKT, and ILK were assessed by western blot at baseline & 24 h on d 1. Pre- & post-treatment (at 24h) biopsies were done in selected patients at the phase 2 dose (n=7).
 Results: 56 pts were entered. Sequential cohorts of patients on schedule A received 1.5, 3, 6, 9, 12, 16 or 22 mg/m2/d (n = 26). On schedule B, the starting dose was 2.5 mg/m2/d. Based on safety information from schedule A, subsequent schedule B dose levels were 14, 19, 25, 34 & 46 mg/m2/d (n = 30). Dose-limiting toxicities on both schedules were pneumonitis, transaminitis, thrombocytopenia and fatigue. Selected grade ¾ toxicities are liver function test elevation (14%), pneumonitis (9%), diarrhea (4%), nausea (4%), fatigue (4%) & platelets (4%). No objective responses were noted, stable disease was seen in 4 pts. Day 1, 17DMAG PK were linear over 1.5-46 mg/m2 and Cmax increased linearly with dose (0.071-1.7 mg/ml). The AUC increased linearly with dose (0.7-14.7 mg/ml•h). Both clearance and T1/2 did not vary systematically with dose (79 + 40 ml/min/m2 and 24 + 15 h). The 24-h urinary excretion accounted for 20 + 9% of dose. The mean HSP70 and HSP27 were 114% (range 58-155) and 154% (range 71-229) of baseline respectively in PBMCs at 24 h at 12, 16, & 25 mg/m2. HSP27, HSP70, AKT and CDK4 levels were decreased in the majority of post treatment biopsy samples, but did not correlate to changes in corresponding PBMC samples.
 Conclusions: The recommended phase 2 of 17DMAG is 16 mg/ m2 x 5 days or 25 mg/ m2 x 3 days every 3 weeks. Therapy was well tolerated at the phase 2 doses. Reversible pneumonitis was a new DLT and not predicted by animal toxicology. Pneumonitis and transaminitis appear to be schedule-dependent. Cardiac toxicity was not seen. 17DMAG plasma PK are linear over the doses delivered to date, and there is a suggestion of a target effect as manifested by HSP and client protein changes in PMBCs and paired tumor biopsies.
 Support: U01CA099168-01, U01CA62502, U01CA69912, R01CA90390 & NIH/NCCR/GCRC grants #5M01 RR 00056 & M01 RR00080

AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics-- Oct 22-26, 2007; San Francisco, CA