Mendelson, David S.

10062 Background: 17-AAG is a benzoquinone ansamycin that binds to and inhibits the HSP90 family of molecular chaperones leading to the proteasomal degradation of client proteins critical in malignant cell proliferation and survival. We have undertaken a Phase 1 trial of CNF1010, an oil-in-water nanoemulsion of 17-AAG. Methods: Patients (pts) with advanced solid tumors, ECOG PS 0–2, and adequate hematologic, hepatic, renal and cardiac functions received CNF1010 by 1 h intravenous infusion, twice-a-week, three weeks out of four, starting at 6 mg/m2 per dose. Doses were escalated sequentially in single pts (6 and 12 mg/m2) and 3–6 pts (≥ 25 mg/m2) cohorts according to a modified Fibonacci’s schema. Plasma pharmacokinetic (PK) profiles were obtained on days 1 and 18. Biomarkers were measured in PBMC’s (HSP70) and plasma (HER-2 ectodomain (HER-2 ECD)). Results: 30 pts (M/F: 14/16; median age 63, range 48–78) with colorectal cancer (11), pancreatic cancer (5), melanoma (5), ovarian (2), others (7) were treated with a median of 2 courses (range: 1–10). There was no dose-limiting toxicity up to 175 mg/m2. One pt at 175 mg/m2 died on study, but drug relation was unclear. Grade 1–2 gastrointestinal toxicities (nausea, vomiting, diarrhea) and serum creatinine elevation were observed. Severe toxicities (grade 3 but no grade 4) consisted of reversible hepatic enzyme elevation, hyperbilirubinemia, fatigue, anemia and hyperglycemia. There were no hematological toxicities. Plasma 17-AAG PK appeared dose-proportional (AUC, Cmax); CL (17 L/h/m2) and t1/2 (5.2 h) were dose independent and unchanged after repeated dosing (day 18). Post-treatment increases in HSP70 were observed in PBMCs and decreases in plasma HER-2 ECD were observed at doses ≥ 83 mg/m2. Minor tumor regressions were seen in 1 pt with duodenal cancer (83 mg/m2), 1 pt with gastric carcinoid (175 mg/m2) and 1 pt with melanoma (175 mg/m2). Conclusion: The threshold of biologic activity for CNF1010 administered by a twice a week schedule appears to be 83 mg/m2. This dosing regimen appears to be optimal and supported by the pharmacokinetic and pharmacodynamic data. The MTD has not yet been determined. Dose escalation of CNF1010 continues at 225 mg/m2. [Table: see text]

265 Background: Gemcitabine (G) is the standard treatment for first- line PanC. PanC is one of the most hypoxic solid tumors. TH-302 is an inert prodrug of brominated isophosphoramide mustard and undergoes selective activation in deep hypoxia. As a single agent, tumor responses were reported in patients (pts) with metastatic melanoma, SCLC, and head/neck cancer at TH-302 weekly doses of 480-575 mg/m2. Methods: Eligible pts for the PanC expansion of this phase I/II study ( NCT00743379 ) had ECOG <1, locally advanced or metastatic PanC previously untreated with systemic chemotherapy other than adjuvant G, 5FU, and/or radiation. IV TH-302 was dosed at 240-575 mg/m2 (240 or 340 in expansion) with standard dose G (1000 mg/m2) on days 1, 8 and 15 of a 28-day cycle. Serum protein and microRNA hypoxia biomarkers were analyzed at baseline, start of cycle 3 and end of study. Results: 46 PanC subjects (12 locally advanced, 34 distant mets); median age: 63 (range 41-83); 24 male; ECOG 0/1 in 29/17 pts; RECIST response rate (RR) of 21%, median PFS of 6.1 mo (95%CI 4.8, 7.7) and median survival of 11.4 mo (95%CI 6.0, not reached) were observed. RR was 23% with median survival of 7.4 mo in pts with distant mets. 52% of pts had a >50% decrease in CA19-9. Common adverse events were skin or mucosal toxicity, nausea, fatigue and vomiting; most grade 1/2. Grade 3/4 neutropenia, thrombocytopenia and anemia in 68%, 64%, and 20% of pts respectively. The dose intensities at 240 mg/m2 and 340 mg/m2 were similar and related to hematologic toxicities. Skin toxicities were less common at 240 mg/m2. A TH-302 dose response was present with higher RR and PFS at 340 mg/m2. Initial serum hypoxia biomarkers did not identify a preferential pt population. Conclusions: The activity and clinical benefits of the combination of TH-302 with G in first line PanC are promising as compared to previous studies of G alone. TH-302 adds to the hematologic toxicity of G, but the regimen is well tolerated. The safety and activity provided rationale for comparing TH-302 plus G versus G alone in a randomized phase II trial ( NCT01144455 ) and indicate TH-302 may complement G by penetrating into the hypoxic regions of the PanC tumors where activation induces cytotoxicity. No significant financial relationships to disclose.