Presentation Authors: Ziho Lee*, Philadelphia, PA, Opeyemi Jegede, Boston, MA, Naomi B Haas, Philadelphia, PA, Michael R Pins, Park Ridge, IL, Edward M Messing, Rochester, NY, Judith Manola, Boston, MA, Christopher G Wood, Houston, TX, Christopher J Kane, San Diego, CA, Michael AS Jewett, Toronto, Canada, Keith T Flaherty, Boston, MA, Janice P Dutcher, New York, NY, Robert S DiPaola, Lexington, KY, Robert G Uzzo, Philadephia, PA
Introduction: Local recurrence (LR) after extirpative surgery for renal cell carcinoma (RCC) refers to cancer that has recurred at or near the site of the primary tumor. Currently, no existing definition of LR accounts for differences in location or extent of recurrence, and the potential prognostic implications of these differences are unknown. We describe a novel anatomic classification system for LR, and evaluate clinicopathologic risk factors associated with potential survival differences based on LR type.
Methods: Recurrence data from the ASSURE (ECOG-ACRIN 2805) trial were queried for all patients with fully resected intermediate or high risk non-metastatic RCC with LR. All patients with concurrent metastasis (any extra-abdominal recurrence) at time of LR were excluded. The cohort was divided into four LR groups: Type I: remnant organ or adjacent soft tissue recurrence; Type II: ipsilateral vein (vein remnant or IVC), gland (adrenal) or node recurrence; Type III: distant intra-abdominal soft tissue or visceral recurrence; and Type IV: any combination of Types 1-3 LR. Multivariable logistic regression was used to identify clinicopathologic predictors. The covariates assessed included age, gender, postoperative performance status, AJCC stage, Fuhrman grade (FG), sarcomatoid differentiation, positive surgical margin, papillary pathology, and use of minimally-invasive surgery. Logrank test was used to compare RCC-specific survival (RCCS) and overall survival (OS).
Results: Of 1,943 patients in ASSURE, 300 (15.4%) patients had LR. There were 65 (21.7%) Type 1, 99 (33.0%) Type 2, 93 (31.0%) Type 3, and 43 (14.3%) Type 4 LR patients. On multivariable analysis, use of minimally-invasive surgery did not predict any type of LR. Type I LR was predicted by higher AJCC stage (p=0.001) and FG (p=0.034), and papillary pathology (p=0.039). Type II LR was predicted by higher AJCC stage (p < 0.001) and FG (p=0.001). Worse postoperative performance status (p=0.032) was the only variable associated with Type III LR. Type IV LR was predicted by male gender (p=0.014), and higher AJCC stage (p < 0.001) and FG (p=0.003). Five-year RCCS (p < 0.001) and OS (p < 0.001) were worse for patients with Type 4 LR (median 37.1% and 31.2%, respectively) compared to those with Types 1-3 LR (median 61.5-71.7% and 57.9-65.7%, respectively). There was no difference in 5-year RCCS and OS among Types 1-3 LR.
Conclusions: Our anatomic classification system may be used to categorize LR based on location and extent of tumor burden. Although there was no difference in survival when LR was limited to a single anatomic subdivision, LR that involved multiple subdivisions (Type 4) was associated with worse survival.
Source of Funding: This study was supported by the National Cancer Institute of the National Institutes of Health under the following award numbers: CA180820, CA180794, CA180867, CA180858, CA180888, CA180821, CA180863, and Canadian Cancer Society #704970.