The area of cancer in transurethral resection specimens as a prognostic indicator in carcinoma of the prostate: a computer-assisted morphometric study.

We used a computerized interactive morphometric method to determine total area of tissue and total area of cancer in routinely stained sections from each of 79 cases of prostatic adenocarcinoma, diagnosed by transurethral resection (TURP). The ability of two morphometrically determined parameters (area of cancer and percent area of cancer) to predict survival was compared with results obtained from two estimates of tumor area commonly used in clinical practice (number of chips and percent of chips involved by cancer). Total Gleason score was also determined. All patients were diagnosed prior to 1981, allowing follow-up of at least 5 years, or until death. Using the Cox proportional hazards regression analysis, our two morphometrically determined parameters, as well as the percentage of chips involved by cancer and total Gleason score, were significant predictors of survival. In contrast, the total number of chips involved by cancer did not reach statistical significance as a predictive factor for survival. By using our morphometrically determined area measurements as a bench mark for clinical utility of area estimates of cancer in TURP specimens, we concluded that calculating the percentage of involved chips yields prognostic information that closely approaches the "ideal" of these more time-consuming computer-assisted techniques.