Estimation of tumor volume in radical prostatectomy specimens in routine clinical practice.

Tumor volume may be an independent prognostic factor in prostatic adenocarcinoma but is too difficult, expensive, and time consuming to measure in routine clinical practice. We sought to evaluate several simple estimates of tumor volume in radical prostatectomy specimens. Specimens from 86 radical prostatectomies were completely sectioned, and the true tumor was volume calculated using a computer-assisted image analysis technique. True tumor volume was then compared with the results of several estimation techniques. True tumor volume ranged from 0.004 to 9.74 cc (mean, 1.59 cc). The Pearson correlation coefficient of the length of the maximum dimension of tumor as measured from the slide with true tumor volume had an r2 of 0.688. However, 15 (75%) of 20 tumors less than 0.5 cc had a single maximum diameter of less than 10 mm, and only 3 (4%) of 66 tumors with volumes greater than 0.5 cc had a maximum diameter of less than 10 mm. Fifteen (68%) of 22 tumors with true tumor volumes greater than 2 cc had a maximum dimension greater than 20 mm, and only 2 (3%) of 64 tumors with volumes less than 2 cc had a maximum dimension greater than 20 mm. Increasing correlation with true tumor volume could be obtained from the largest single tumor area (r2 = 0.749), the sum of the largest dimensions of two separate tumor foci (r2 = 0.759), and the sum of the two largest areas (r2 = 0.859). For tumors with true volumes less than 0.5 cc, only 1 (5%) of 20 cases had a sum of the two largest areas greater than 65 mm2, and no tumor larger than 0.5 cc had a sum of less than 65 mm2. For tumors larger than 2 cc, 19 (86%) of 22 tumors had a sum of the two largest areas greater than 250 mm2, and only 1 (2%) of 64 cases with a true tumor volume less than 2 cc had an area greater than 250 mm2. We conclude that tumors in radical prostatectomy specimens can be stratified by size based on simple measurements obtainable during routine pathology practice.