Assessing the Extent of Biopsy Core Involvement by Prostate Cancer and Its Impact on the Selection of Patients for Active Surveillance: Is “Eyeballing” Accurate Enough?
Osman Mahamud, Theodorus H van der Kwast, Michelle R Downes, Andrew J Evans. University Health Network, Toronto, ON, Canada
Background: Active surveillance (AS) is an established management option for patients with low-risk prostatic adenocarcinoma (CaP). Biopsy criteria for AS include Gleason score < 6/10, involvement of < one third of cores in a biopsy set and < 50% core involvement in any positive core. There is no standardized method for assessing % core involvement and a variety of methods can be used including estimation, or “eyeballing” (EB), and direct measurement. In this study, we assessed the concordance between EB and measurement in a typical cohort of positive needle biopsies.
Design: As part of a prospective digital pathology validation project in our department, digital slides from 95 sets of prostate biopsies were available for review, 43 (45%) of which had least one core with CaP. CaP was annotated on the digital slides by two experienced urological pathologists. Open source java script software (Image J, http://rsbweb.nih.gov/ij) was used to measure the linear extent of CaP in relation to total core length. Image J permits precise length measurements of non-linear objects such as prostate cores. The results, expressed as % core involvement, were compared to values obtained by EB of the digital slides. A sub-set of cores determined to have 40-60% involvement by EB was identified to assess biopsies particularly relevant to the issue of AS patient selection.
Results: A total of 192 cores contained CaP in the 43 positive cases. There was no significant difference between EB and Image J in terms of mean % involvement when considering all positive cores (EB - 32.7 % + 29.0 (SD) versus Image J - 31.4 % + 26.4 (SD), p=0.08). There were 38/192 cores (20%) determined to have 40-60% core involvement by EB. In this sub-set there was a significant difference between the two methods (EB - 54.8 % + 13.4 (SD) versus Image J - 47.0 % + 11.6 (SD), p= 0.0004), with EB tending to overestimate % involvement by an average of 12%. Issues complicating the assessment of % core involvement by either method included: core contour (straight, folded or bent), core fragmentation, discounting rectal mucosa or periprostatic soft tissue from total core length and discontinuous core involvement by CaP.
Conclusions: There was excellent overall concordance between EB and Image J, however direct measurement from digital images may be more robust when assessing % core involvement for AS patient selection. Our study also raises the question of a need for standardization for determining extent core involvement given its importance in the setting of AS.
Category: Genitourinary (including renal tumors)
Monday, March 4, 2013 1:00 PM
Poster Session II # 159, Monday Afternoon