A phase II trial of oblimersen sodium (G3139) in combination with doxorubicin (DOX) in advanced hepatocellular carcinoma (HCC). NCI protocol # 5798

14072 Background: Advanced HCC is refractory to most standard forms of chemotherapy, however responses to DOX are seen. The bcl-2 protein confers resistance to apoptosis in cancer cells and is important in tumor progression and resistance to chemotherapy. The bcl-2 antisense oligonucleotide, G3139 (G), has been shown to enhance the activity of DOX in tumor models by blocking bcl-2 synthesis. This argues for evaluating G + DOX in combination in HCC. By decreasing tumor bcl-2 protein levels, HCC may be sensitized to the apoptotic effects of DOX. Methods: We completed a phase II trial evaluating treatment with G at 7 mg/ kg for 7 days cont. i.v. infusion (d1–8) plus DOX at 45 mg/m 2 i.v. bolus d5, every 28 d (as determined from our phase I HCC study). Eligible patients (pts) had path-confirmed, measurable, advanced HCC. Minimal eligibility included Childs-Pugh A cirrhosis, adequate hematological (hem) parameters and ECOG PS <2. Tumor biopsies for correlative studies were obtained at baseline and cycle 1 d 4 in consenting pts. Results: 19 patients were accrued, 1 was ineligible, 18 evaluable for toxicity, 17 evaluable for response; receiving a median (med) of 2 cycles (range 1,10). Risk for HCC was 39% HBV, 22% HCV, 17% alcohol, 22% other. Most common toxicities were hem and could be attributed to both G+DOX and to G alone. Overall grade 3–4 toxicities seen were: ANC- 67% (med nadir d 24–25), lymphopenia - 44%, thrombocytopenia - 6%, transaminitis - 33% and grade 1–2 G-fever - 67%. No responses were seen and the trial was stopped at stage 1. Six patients (35%) had stable disease, with one pt completing 10 cycles as per protocol (pt # 22). Med TTP is 1.8 months (1.7-NA) and 6-month PFS is 17.2% (5.3–56.4). 18 of 19 pts have died with med OS of only 5.4 months (2.7–11.6). Correlative studies on 3 available pts’ paired tumor biopsies showed absent baseline bcl-2 expression but moderate expression of both bcl-xl and BAX protein and with no change after exposure to G (includes pt #22). Conclusions: G + DOX is inactive in HCC at this dose and schedule. The overlap of hem toxicity may have resulted in suboptimal DOX dosing in HCC. Low baseline bcl-2 tumor expression relative to bcl-xl seen may suggest a relative insensitivity to the effects of bcl-2 inhibition in these HCC tumors. [Table: see text]