No abstract

Patients with locoregional gastric carcinoma often die because of the low rates of curative resection and frequent appearance of distant metastases (mainly peritoneal and hepatic). To evaluate the feasibility of preoperative and postoperative chemotherapy, 25 consecutive previously untreated patients with potentially resectable locoregional gastric carcinoma received two preoperative and three postoperative courses of etoposide, 5‐fluorouracil, and cisplatin (EFP). Ninety‐eight courses (median, five courses; range, two to five courses) were administered. Six patients had major responses to EFP. Eighteen patients (72%) had curative resections, and three specimens (12%) contained only microscopic carcinoma. At a median follow‐up of 25 months, the median survival of 25 patients was 15 months (range, 4 to 32+ months). Peritoneal carcinomatosis was the most common indication of failure. One patient died of postoperative complications, but there were no deaths due to chemotherapy. EFP‐induced toxic reactions were moderate. Preoperative and postoperative chemotherapy for locoregional gastric carcinoma is feasible, and additional studies to develop regimens that could result in 5% to 10% complete pathologic responses may be warranted.

The leukocyte migration in agarose assay recently developed by Clausen (Clausen, 1971), was used to test 22 lung cancer patients against soluble extracts of allogeneic lung cancer and allogeneic normal lung. Seventeen were inhibited to a significantly greater degree by at least one tumor extract (average migration index (MI equals 0.58) than by the corresponding normal lung extract (average MI equals 0.83). DNCB-positive and DNCB-negative patients reacted with equal frequency to tumor extracts. Three patients tested against their autologous tumor and normal lung extracts were specifically inhibited by the tumor extract (average MI equals 0.53) but not by the normal lung extract (average MI equals 0.83). None of seven lung cancer patients tested against non-pulmonary tumor extracts was significantly inhibited (average MI equals 1.1). Only 6/53 controls including patients with other tumors, patients with emphysema and age-matched non-smokers showed significant inhibition against any of the lung cancer extracts. These findings strongly suggest the presence of tumor-associated antigens in KCl-solubilized extracts of human lung cancer.

In 135 patients with regional cutaneous recurrence of extremity melanoma, the prognostic significance of 12 clinical and pathologic variables was analyzed in four alternative Cox stepwise regression models and by single variable analysis. A highly significant fit of the regression (P less than 0.01) identified four factors that particularly influenced survival: the presence of intradermal or mixed (as opposed to purely subcutaneous) metastases (P less than 0.001), sex (P = 0.032), excision of regional cutaneous metastases with or without perfusion (P = 0.033), and the presence of subcutaneous metastases (P = 0.201). Not predictive of survival were age at diagnosis; site of primary; anatomic location, number, size, or distance from the primary of the regional cutaneous metastases; time since primary treatment; number of positive regional lymph nodes; and single- or triple-drug perfusion.