Tumor Grade at Margins of Resection in Radical Prostatectomy Specimens Is an Independent Predictor of Prognosis
F Brimo, JI Epstein. The Johns Hopkins Hospital, Baltimore
Background: We sought to assess whether reporting the grade of cancer at the site of positive margins in a radical prostatectomy specimen was independently prognostic.
Design: We restricted our study to 108 cases (1995-2008) of Gleason score 7 with non-focal extra-prostatic extension (pT3a) and positive surgical margins. Cases with tertiary pattern 5, neoadjuvant therapy, or positive margins as a result of intraprostatic incision were excluded. When more than one section showed a positive margin, the overall length of positive margin was obtained by adding the lengths of all areas of cancer in contact with ink regardless of their location. When two consecutive slides going from apex to base showed positive margins, the length of positive margin in these two slides was calculated as 3 mm (1.5 mm being half of the thickness of each tissue section on a slide) provided that none of these slides showed a positive margin on the one side larger than 3mm. Cancer present at a margin was assigned a Gleason score independent from the cancer score of the entire case. In slides where cancer at the margin showed obscuring cautery artifact, the Gleason score at the margin was assigned based on the grade of uncauterized cancer in direct continuity with the cauterized cancer.
Results: The overall Gleason score was 3+4 in 73 (67%) and 4+3 in 35 (33%) cases. The median length of positive margin was 3.0 mm (0.5-10 mm). Gleason score at the margin was 3+3, 3+4, 4+3, and 4+4 in 40 (37%), 41 (38%), 16 (14.8%) and 11 (10.2%) cases respectively. 45 (42%) patients remained free of disease post-radical prostatectomy (median follow-up=6 years, range 3-13 years). Univariate and multivariate analysis showed no correlation between biochemical recurrence and both pre-operative serum PSA levels (p=0.7) and overall Gleason score (p=0.5). A strong association was noted between biochemical recurrence and Gleason score at the margin (p=0.007) with length of cancer at the margin also predictive (p=0.015) in multivariate analysis. Using the median length of positive margin (3mm) as a cutoff, the association with biochemical recurrence was significantly different between the two groups (p=0.004) on Kaplan-Meier analysis.
Conclusions: The current study is the first to prove that the grade of cancer at the site of a positive margin influences outcome. We were able to stratify grade into 3 categories: 3+3; 3+4; and 4+3 or higher (4+3 and 4+4 at the positive margin carried equal prognostic information).
Category: Genitourinary (including renal tumors)
Monday, March 22, 2010 1:00 PM
Poster Session II # 118, Monday Afternoon