Interobserver Variability in Histologic Evaluation of Radical Prostatectomy (RP) between Central Pathologists (CP) and Local Pathologists (LP): Findings of a Multinational Clinical Trial
GJ Netto, JI Epstein. Johns Hopkins University, Baltimore, MD
Background: TAX 3501 is a randomized phase III multinational trial comparing outcome of post RP adjuvant androgen deprivation vs. docetaxel in high-risk patients. Eligibility is determined by a predicted biochemical progression free survival of 60% based on Kattan's nomogram following CP review.
Design: All RP sections were blindly rereviewed by one of two urologic CP on the trial. Data on Gleason Score (GSc), organ confined vs focal vs extensive extraprostatic extension (OC vs FEPE vs EEPE), margin (MG), seminal vesicle (SV) and lymph node (LN) status was compared between CP and LP in a total of 257 consecutive RP.
Results: GSc: Agreement was found in 181/257 (70%) RP. Among 76 cases with GSc discrepant diagnoses: CP upgrade occurred in 57 (75%) and downgrade in 19 (25%) cases. Most frequent upgrade was from 7 to 8/9 and most frequent downgrade was from 8 to 7. 37% and 2% of upgrades were of 2 and 3 GSc increments, respectively. 21% of downgrades were of 2 GSc with all remaining GSc changes being of 1 increment. OC/FEPE/EEPE: Agreement was achieved in 179/256 (70%) of RP. Among the 77 (30%) cases with stage discrepant diagnoses: CP upstaging occurred in 70 (91%) and downstaging in 7(9%) cases. Most frequent upstage was from FEPE to EEPE (44%) followed by OC to EEPE (27%). 25% of RP that were staged as capsular invasion by LP, were restaged as EEPE. Reasons that led to understaging of EEPE included: ambiguity at apex, tumor with desmoplasia lacking surrounding fat, ambiguity of capsular invasion definition. SV Status: Agreement was encountered in 238/257 (93%) cases. Among the 7% RP with disagreement, almost an equal number of under/overcalling was observed in LP compared to CP evaluation. Overcalling was due to presence of tumor in peri-SV tissue (not in muscle) or in ejaculatory duct and prostate tissue adjacent to SV. MG Status: Agreement was present in 229/256 (89%) RP. Among the 27 (11%) with discrepant diagnoses: CP review lead to reclassification from (-) to (+) in 17 (62%) and (+) to (-) in 10 (38%) cases. Overcalling due to artifactual tissue tear and undercalling due to missed cauterized/crushed tumor cells was observed. LN status: Positive LN missed by the LP was identified on CP review in 2/208 (1%) RP with agreement observed in all remaining cases.
Conclusions: Significant interobserver variation exists between CP and LP interpretations that could affect patient accrual to clinical trials and impact prognostication and management.
Category: Genitourinary (including renal tumors)
Wednesday, March 11, 2009 9:30 AM
Poster Session V # 117, Wednesday Morning