Treatment of patients with upper gastrointestinal carcinomas.

Carcinomas of the stomach and esophagus are a major health problem worldwide. Cancer remains incurable when it is metastatic or unresectable, and new active agents are needed to improve the outcome for these patients. Three phase II studies of paclitaxel (Taxol; Bristol-Myers Squibb Company, Princeton, NJ) are described here. The first, at M.D. Anderson Cancer Center, examined the role of single-agent paclitaxel in untreated patients with advanced gastric carcinoma. Thirty-three patients were accrued. The starting dose of paclitaxel was 200 mg/m2 and it was repeated every 21 days. The first 15 patients received paclitaxel administered over 3 hours, and the subsequent 18 patients received paclitaxel administered over 24 hours. Three patients were not evaluable. Among the first 15 patients (two not evaluable) receiving paclitaxel over 3 hours, there was one partial response (PR) (8%) and three minor responses (MRs) (23%). Among the 18 patients (one not evaluable) receiving paclitaxel over 24 hours, however, there were four PRs (24%) and three MRs (18%). The overall PR rate was 17% (five of 30 patients). The median duration of PR was 6.5 months (range, 2.3 to 11.3+ months). Myelosuppression was more severe with the 24-hour schedule than with the 3-hour schedule. Our data suggest that paclitaxel is active against gastric carcinoma, albeit to a lesser degree than in esophageal adenocarcinoma. Paclitaxel should be investigated further in combination therapy with other active agents. In the second phase II study, paclitaxel was administered via a 24-hour intravenous infusion in chemotherapy-naive patients with unresectable locoregional or metastatic carcinoma of the esophagus. The starting dose of paclitaxel was 250 mg/m2, repeated every 21 days. In addition, patients received granulocyte colony-stimulating factor (5 microg/kg/d subcutaneously) 24 hours after the completion of treatment. Thirty-three patients had adenocarcinoma and 18 had squamous cell carcinoma. Sixteen patients (31%) achieved a response (one complete response [CR] and 15 PRs), and 11 (22%) achieved an MR. Of the 33 patients with adenocarcinoma, 12 (36%) achieved either a CR (one patient) or a PR (11 patients); six patients had an MR. Four of 18 patients (22%) with squamous cell carcinoma had a PR and four (22%) had an MR. Paclitaxel was well tolerated. Granulocytopenia was frequent, resulting in 11 hospitalizations in nine patients. Thus, paclitaxel is active against both adenocarcinoma and squamous cell carcinoma of the esophagus. In a subsequent phase II trial, paclitaxel has been studied in combination with cisplatin and 5-fluorouracil in 30 patients with advanced adenocarcinoma and 22 patients with squamous cell carcinoma of the esophagus. The doses are paclitaxel 175 mg/m2 administered over 3 hours on day 1, cisplatin 20 mg/m2 on days 1 to 5, and S-fluorouracil 1,000 mg/m2/d (in the first 10 patients, but then reduced to 750 mg/m2/d as continuous infusion on days 1 to 5) given every 28 days. Among the 47 patients evaluated for response so far, four have had a CR and 17 have had a PR (overall response rate, 45%). Toxicity with this combination has been moderate, and there have been no treatment-related deaths. With the reduced starting dose of 5-fluorouracil, the tolerance of this combination has improved substantially both in the inpatient and outpatient settings, resulting in a low frequency of grade 3 or 4 nonhematologic toxicity. Response has been higher in patients with squamous cell carcinoma.