Weak Association of Prostatic Atrophy with High-Grade PIN and Low-Grade Cancer: Topographic Digital Study of 48 Whole-Mounted Prostatectomy Specimens
Kenneth A Iczkowski, Kathleen C Torkko, Robert S Wilson, Marshall S Lucia, David G Bostwick. University of Colorado Denver, Aurora, CO; Bostwick Laboratories, Inc., Glen Allen, VA
Background: Controversy persists regarding the relationships, if any, between atrophy and high-grade prostatic intraepithelial neoplasia (HGPIN), and between atrophy and cancer.
Design: Foci of atrophy without inflammation (A), atrophy with chronic inflammation (AI), HGPIN, and 9 different cancer patterns as defined by Gleason, were digitally encircled using separate colors on 238 virtual whole-mount slides from 48 radical prostatectomies with cancer. The frequency of abutment of A or AI to cancer, and of HGPIN to cancer, was tallied. Nearness (defined as <2 mm) of these areas was also recorded. The significance of these differences was tested by chi-square test. The percentage of these foci abutting each of the cancer patterns was tested by chi-square Goodness-of-Fit. Area sums of patterns within a specimen were defined as the sum of areas of all foci annotated a given color, and their correlations were studied by Pearson correlation coefficient.
Results: 1163 foci of A, 402 of AI, and 363 of HGPIN were annotated. AI did not abut cancer more frequently than did A (21% vs. 23%, p=0.31) but nearness of AI to cancer was more common than nearness of A to cancer (29% vs. 12%, p=0.0001). A total of 273 (75%) HGPIN foci abutted cancer, with another 15% near. There was abutment to A in 2.4% of HGPIN foci and to AI in 2.0%; corresponding percentages for nearness were 1.4% and 1.5%. Small separate cancer glands (S) abutted A in 40% of foci, AI in 16%, and HGPIN in 36% (all p<0.0001 compared to % of foci of all other cancer patterns). For larger separate glands with undulating lumens (U) abutment was seen in 21%, 6%, and 33%, and for fused small glands (F), 6%, 3%, and 3% respectively. Abutment to high-grade cancer patterns other than F was rare (<1%). Nearness showed similar trends. The area sum of A or AI did not correlate with that of total cancer. The area sum of all atrophy showed negative correlations with area sums of S (r=-0.46, p=0.001), of U (r=-0.37, p=0.022), and of F (r=-0.47, p=0.007), Gleason score (r=-0.30, p=0.037), and computed tumor volume (r=-0.45, p=0.001); and it had no correlation with the presence of any cancer pattern (t-tests).
Conclusions: Atrophy showed a much weaker relationship to cancer than HGPIN did. Frequency of AI merging with HGPIN was far less than reported by others. Atrophy area did not correlate with cancer area or volume. While a role for A or AI as promoters of neoplasia is possible, a strong spatial association with HGPIN or cancer was not evident.
Category: Genitourinary (including renal tumors)
Tuesday, March 5, 2013 1:00 PM
Poster Session IV # 172, Tuesday Afternoon