Longitudinal evaluation of the concordance and prognostic value of lymphovascular invasion in transurethral resection and radical cystectomy specimens

OBJECTIVE: To evaluate the concordance transurethral resection of bladder tumour (TURBT) and radical cystectomy (RC) specimens with regard to the presence of lymphovascular invasion (LVI). Additionally, to evaluate the prognostic value of LVI in the prediction of lymph node metastases, overall survival, disease-specific survival and recurrence-free survival following RC. PATIENTS AND METHODS The records of 487 patients who underwent RC at our institution between 1987 and 2008 were retrospectively reviewed and evaluated for the presence or absence of LVI as determined by pathological evaluation. The presence or absence of LVI was then evaluated on previous transrectal resection specimens of this cohort of patients undergoing RC. Cox regression and Kaplan-Meier analysis were undertaken to evaluate the contribution of LVI to various outcomes. RESULTS Of 474 patients with complete LVI data, 60 (12.3%) were found to have LVI at TURBT compared to 161 (33.1%) at RC. Although the presence of LVI at TURBT was more significantly associated with the presence of LVI at RC, only 42.9% of patients in whom LVI was documented at TURBT were found to harbour LVI at RC. The risk of nodal disease was higher in those patients with LVI at TURBT than in those with no evidence of LVI at TURBT (48.3% vs 25.0%, P < 0.001). Additionally, LVI at TURBT was associated with an increasing risk of pathological upstaging and the receipt of adjuvant chemotherapy. Survival analysis showed a significant decrement in overall and recurrence-free survival among those with LVI at TURBT compared to those with no evidence of LVI. CONCLUSIONS Lymphovascular invasion at TURBT provides useful prognostic information that should be incorporated into clinical decision-making, particularly with regard to cystectomy for nonmuscle-invasive carcinoma and the administration of neoadjuvant chemotherapy.