Objective: The objective was to determine whether tumor volume, calculated by multiplying 3-dimensional diameters recorded from surgical specimens, could represent tumor size more precisely than the currently used greatest tumor diameter (T status) as a prognostic factor in early-stage non-small cell lung cancer (NSCLC).
Methods: From 1991 to 2001, a total of 236 stage Ia NSCLC patients underwent curative resection in our institution. Their clinicopathological factors were retrospectively reviewed. The cases were grouped according to stereographic tumor size. Group 1 included tumor volumes less than 0.5236 (1(3) x pi/6) cm3, Group 2 included tumor volumes between 0.5236 - 4.1888 (2(3) x pi/6) cm3, while Group 3 included tumor volumes between 4.188 - 14.1372 (3(3) x pi/6) cm3. Overall survival and disease-free interval analyses were performed with the Kaplan-Meier method and multivariable Cox's proportional hazard model. Tumor volume and other clinico-pathological factors were included for analysis.
Results: Median follow-up was 55.5 months. The overall 5- and 10-year survival rates were 78 % and 71 %, and the 5- and 10-year disease-free intervals were 68 % and 64 %, respectively. The overall survival ( P = 0.0075) and disease-free interval ( P = 0.0025) showed significant differences between tumor volume groups. Smoking history, presence of symptoms and the number of resected lymph nodes were also significantly related to overall survival and disease-free interval.
Conclusion: The products of 3-D diameters (tumor volume) more precisely evaluated tumor size as well as survival of patients with stage Ia NSCLC.