Anatomic Pathology / PREDICTING TUMOR VOLUME IN RADICAL PROSTATECTOMY SPECIMENS

Tumor volume has prognostic value in numerous malignant neoplasms; however, the determination of tumor volume in prostatic adenocarcinoma remains problematic. We tested the hypothesis that the diameter of the largest focus of carcinoma in whole-mount prostate sections predicts the volume of adenocarcinoma in the entire prostate. We evaluated 184 radical prostatectomy specimens by whole-mount processing of the entire prostate. The maximum diameter of the largest focus of carcinoma was measured directly on glass slides. Tumor volume in the entire prostate was calculated by the grid method. The maximum tumor diameter ranged from 0.1 to 4.1 cm (median, 1.6 cm). The total tumor volume ranged from 0.1 to 12.5 cm3 (median, 1.6 cm3). There were significant correlations between maximum tumor diameter and tumor volume (Spearman correlation coefficient = 0.84; P < .0001), surgical margin status (P < .001), perineural invasion (P < .001), serum prostate-specific antigen level at diagnosis (P = .004), Gleason score (P = .004), and pathologic stage (P < .0001). Maximum tumor diameter is a predictor of tumor volume and might be useful for the assessment of tumor volume in routinely processed prostatectomy specimens. During recent decades, our understanding of the natural history of prostate carcinoma has expanded rapidly. Tumor volume has emerged as a significant and valuable prognostic factor.1 Quantitative assessment of the extent of prostatic adenocarcinoma has been problematic because gross identification of prostatic carcinoma is subtle and may be difficult,2 and the cancers often are multifocal.3,4 Inclusion of tumor size in radical prostatectomy specimens in surgical pathology reports has been suggested,5,6 but the exact method for reporting has not been standardized. The Association of Directors of Anatomic and Surgical Pathology recommends that the percentage of the prostate involved by carcinoma in relation to the weight of the specimen be reported for radical prostatectomy specimens.7 Multiple techniques have been used to assess tumor size, including visual inspection with the pathologist’s percentage estimate,8,9 the number of blocks involved by tumor,10 the diameter of the largest tumor focus,8,10,11 the maximal area of the tumor,11 grid morphometric analysis,12 3-dimensional estimate,13 and computer-assisted image analysis.14-17 The latter methods, while accurate, are time consuming and expensive. Renshaw et al11 noted that the simple measurement of the maximum diameter of the largest tumor focus can stratify prostatic adenocarcinoma in radical prostatectomy specimens based on size. Maximum dimension of the largest tumor focus was particularly useful for identifying very small tumors (<0.5 cm3) or large tumors (>2 cm3).11 In the present study, radical prostatectomy specimens were analyzed by whole-mount processing to determine whether the maximum diameter of the largest focus of tumor correlated with total tumor volume and other known prognostic factors. Anatomic Pathology / ORIGINAL ARTICLE Am J Clin Pathol 2003;120:386-391 387 387 DOI: 10.1309/82U1089XLQGKMMN1 387 © American Society for Clinical Pathology Materials and Methods The study group consisted of 184 consecutive patients treated by radical retropubic prostatectomy with or without bilateral pelvic lymphadenectomy between April 1999 and July 2001 at Indiana University Hospital, Indianapolis. Patients who received preoperative androgen-deprivation therapy were excluded. Serum prostate-specific antigen (PSA) levels were measured using the Immulite PSA assay (Diagnostics Products, Los Angeles, CA). The research was approved by the Indiana University Institutional Review Board. The radical prostatectomy specimens were examined as previously described.18-21 Prostates were weighed, measured, inked, and fixed in 10% neutral formalin. Following fixation, the apex and base were amputated and serially sectioned at 3to 5-mm intervals in the vertical, parasagittal plane. The seminal vesicles were sectioned parallel to their junction with the prostate and entirely submitted for examination. The remaining prostate was sectioned serially, perpendicular to the long axis from the apex of the prostate to the base, and whole-mount sections were prepared. The greatest diameter of the largest single focus of tumor was obtained by marking both ends of the tumor on the glass slide and measuring this distance with a ruler marked in millimeters. If the tumor size was less than 0.5 cm, an ocular micrometer was used. The volume of carcinoma in the entire prostate was determined by using the grid method12,18,19,22-24 and was the sum of the volumes of individual foci of tumor. In this method, the sum of each area was multiplied by the thickness of the average slice, and the sum of these volumes was multiplied by a factor of 1.25 to account for tissue shrinkage during processing.22,23 All cancers were graded according to the Gleason system.25,26 The 1997 TNM (tumor, nodes, metastasis) system was used for pathologic staging.27 Surgical margins were considered positive when carcinoma cells were in contact with the inked margin.18,19 SAS software, version 8 (SAS Institute, Cary, NC) was used for analysis. Maximum tumor diameter and tumor volume were analyzed as continuous variables. The maximum tumor diameter was dichotomized for illustrative purposes using the median value of 1.6 cm as the cutoff. Polynomial regression predicting maximum tumor diameter (mm) with tumor volume (cm3) was performed. The Spearman coefficient of rank correlation was calculated for maximum tumor diameter with all continuous variables (tumor volume, Gleason score, and preoperative PSA level). The Wilcoxon rank sum test was used to compare maximum tumor diameter among binary variables, and the KruskalWallis test was used to compare maximum tumor diameter between pathologic stages. Logistic regression was used to predict tumor at surgical margins and perineural invasion with either tumor volume (cm3) or maximal tumor dimension (cm) while controlling for TNM pathologic stage. Because of the small frequencies, stages T2a and T2b were combined, and stages T3a and T3b were combined for analysis purposes. Therefore, a categorical variable for TNM stage with 2 categories was used as an independent variable in the logistic model. P values less than .05 were considered statistically significant.