Patterns of lymphatic metastases in upper tract urothelial carcinoma: Experience from three comprehensive cancer centers.

Abstract

349 Background: Data on patterns of lymph node metastasis (LNM) for upper tract urothelial carcinoma (UTUC) is sparse despite its critical relevance for staging and therapeutic potential, with only 1 prior study detailing patterns of spread in 23 pathologically confirmed cases (Kondo T et al, Urol. 2007). Methods: We retrospectively identified and reviewed patients with pTany N+ M0 UTUC at 3 comprehensive cancer centers which routinely perform template-based dissections for UTUC. Anatomic locations of the primary tumor and positive lymph nodes based on the pathological specimens were analyzed, and maps of the location and incidence of LNM constructed. Results: A total of 92 patients with pN+ UTUC were identified, 63 of whom had renal pelvis (n=43) or proximal ureteral (n=20) tumors. LNM from left renal pelvis tumors occurred to the hilar (29.4%), pre- and para-aortic (50.0%) regions, with secondary spread to suprarenal (1.8%), interaortocaval (5.9%), and common iliac (2.9%) regions. LNM from right renal pelvic tumors occurred to the hilar (24.1%), pre- and para-caval (38.0%), retrocaval (13.8%), and inter-aortocaval (24.1%) areas. Left proximal ureteral tumors primarily had LNM to the hilar (46.1%), para-aortic (46.1%), and retro-aortic (7.7%) regions. Right proximal ureteral tumors had LNM to the hilar (36.4%), paracaval (54.5%), and retrocaval (9.1%) regions. A significant difference in survival was noted with patients who had organ confined tumor (p = 0.002) and a lymph node density of less than 25% (p = 0.005). The 1, 2, and 5-year cancer-specific survival rates of the overall cohort were 69%, 49%, and 27%, respectively. Conclusions: Our multi-institutional study provides data regarding the primary landing zones for LNM in renal pelvis and proximal ureter UTUC as well as guidance for developing anatomic dissection templates for prospective studies in UTUC incorporating surgical methodologies as a variable.