Concordance and prediction ability of original and reviewed vascular invasion and other prognostic parameters of clinical stage I nonseminomatous germ cell testicular tumors after retroperitoneal lymph node dissection

Abstract

Purpose: We reviewed the slides of patients with clinical stage I nonseminomatous germ cell testicular tumors who underwent retroperitoneal lymph node dissection to evaluate the concordance between original and reviewed vascular invasion status, and other histological correlates.

Materials and methods: Between 2002 and 2007 at our institution 202 consecutive patients underwent retroperitoneal lymph node dissection. We requested the slides of 183 patients who underwent orchiectomy elsewhere. The risk of nodal metastasis was considered high in those with vascular invasion and/or greater than 90% embryonal carcinoma, and low in those with no vascular invasion and embryonal carcinoma less than 90%. Using Cohen's κ we assessed the concordance index between original and reviewed parameters (vascular invasion and risk category). Using the chi-square test we also evaluated the association between nodal status at retroperitoneal lymph node dissection and original vs reviewed parameters.

Results: The original report did not contain vascular invasion information on 98 of 183 cases (53.4%). A total of 164 patients were evaluable since we had no slides for 19. Vascular invasion absence and presence were confirmed in 27 (73.0%) and 30 (78.9%) of 37 patients, respectively (Cohen's κ = 0.16). Low and high risk status was confirmed in 20 of 28 patients (71.4%) and in 47 of 64 (50.6%), respectively (Cohen's κ = 0.22). Reviewed vascular invasion and risk category were significantly associated with nodal status at retroperitoneal lymph node dissection (chi-square test p = 0.03 and 0.01, respectively), although the original parameters were not.

Conclusions: In half of the patients no information was available on vascular invasion in the original reports. Concordance between original and reviewed reports was generally poor. Reviewed parameters better predicted nodal status at retroperitoneal lymph node dissection. These findings may have important implications in clinical practice.