Bronchogenic carcinoma located in the aortic window. The importance of the primary lesion as a determinant of survival.

Thirty-four patients with an aortic window lesion were carefully staged with gallium scans and mediastinoscopy according to the TNM classification system for carcinoma of the lung. All were in Stage III. Twenty-five patients had non-oat cell carcinomas (15 squamous cell, eight adeno-, two large cell) and nine had oat cell carcinomas. Quantitative ventilation-perfusion lung scans were particularly helpful in verifying the subaortic location of the tumor by showing a less than 20% interference with pulmonary blood flow or ventilation secondary to left mainstem bronchus or pulmonary artery invasion. Decision for resectability in 13 Stage III M0 patients was based on the length of the uninvolved proximal left main pulmonary artery seen on pulmonary arteriogram. Eight patients (seven non-oat cell and one oat cell) had resection after radiation and prior to chemotherapy (after two cycles of chemotherapy and prior to radiation therapy for the oat cell) with a resultant survival rate better than those of M0 and M1 non-oat cell or oat cell patients without resection. The survival rates of nine non-oat cell M0 patients, nine non-oat cell M1 patients, and eight oat cell patients, all without resection, were not statistically different. This similarity in survival rates is explained by the observation that 38% of the non-oat cell M1, 71% of the non-oat cell M0, and 63% of the oat cell patients died from complications of their primary tumor. Patients with aortic window lesions, irrespective of their histology, have an extremely poor prognosis due to the high incidence of lethal complications of their primary tumor. Complete resection when feasible, as judged by pulmonary arteriography, provides the best control of the primary tumor and, as a consequence, gives longer survival.