Evaluation of response to preoperative chemotherapy in esophageal and gastric cardia cancer using biphasic esophagrams and surgical-pathologic correlation.

Twenty-two patients with newly diagnosed carcinoma of the esophagus and gastric cardia were treated with combination chemotherapy prior to transhiatal esophagectomy. The clinical response to chemotherapy was determined by obtaining measurements of the vertical length of tumor, tumor depth, lumen size, and extent of ulceration on serial biphasic contrast esophagrams. Response was defined as a greater than 50% decrease in the sum of the products of the maximum vertical length and depth of the esophageal tumor mass. To obtain an assessment of the accuracy of the radiographically determined tumor measurements the esophagram results were correlated with surgical pathologic tumor measurements in 19 patients. The technique of esophagram and the determination of tumor measurements were standardized by having one radiologist perform and interpret all studies. All patients were clinically staged using the TNM classification system of the American Joint Commission on Cancer Staging. Three patients had T1 lesions, fifteen were classified as T2, and four as T3 lesions. Radiographic measurements of the tumor from the preoperative esophagram were correlated with the surgical pathologic specimen in five patients prospectively and 14 patients retrospectively. Response was seen in nine patients. The radiographically measured vertical length of the tumor was within 1 cm of the surgical specimen measurement in four of five (80%) prospectively evaluated patients and in nine of 14 (64%) retrospectively evaluated patients. These results suggest that chemotherapy-induced tumor regression can be objectively evaluated with a reasonable degree of accuracy using serial biphasic esophagrams.