Sensitivity of frozen section examination of pelvic lymph nodes for metastatic prostate carcinoma.

BACKGROUND A Total Quality Improvement review of frozen section diagnoses yielded four sequential false-negative frozen section diagnoses in lymph nodes in the evaluation of metastases from prostate carcinoma. Because these results appeared to differ from the department's overall frozen section experience, frozen section diagnoses of pelvic lymph nodes in patients with prostate carcinoma were reviewed to benchmark the department's performance. METHODS All 220 pelvic lymph node frozen section samples from 110 patients with prostate carcinoma from January, 1986 to July, 1993 at Bridgeport Hospital (BH) were reviewed, and the department's frozen section diagnostic efficiency was compared with: (1) all BH frozen section procedures performed during the same period; (2) pelvic lymph node frozen section analysis and (3) imaging techniques for prostate carcinoma from other institutions found in MEDLARS literature searches from 1973 to 1993. RESULTS Frozen section diagnostic efficiency for patients with prostate carcinoma from BH was 93.6% (sensitivity 63.2%, specificity 100%) compared with (1) 98% for all BH frozen section analyses (sensitivity 99%, specificity 98%); (2) 90.1% for MEDLARS search pelvic lymph node frozen section prostate carcinoma analyses (sensitivity 66.5%, specificity 100%); and (3) 88% for MEDLARS search magnetic resonance imaging (MRI) pelvic lymph node prostate carcinoma (sensitivity 41%, specificity 97%). CONCLUSIONS Frozen section diagnosis of pelvic lymph nodes approaches the overall diagnostic efficiency of frozen section analysis and is more sensitive for the diagnosis of prostate carcinoma in pelvic lymph nodes than is MRI because more than half of the metastases are smaller than the 1.0-cm resolution limit of the MRI. False-negative frozen section diagnoses (67% sensitivity) occur because of errors in sampling microscopic metastases.