Kozo Kataoka1, Hannes Ysebaert2, Manabu Shiozawa3, Dries Reynders2, Masataka Ikeda4, Naohiro Tomita4, Els Goetghebeur2, Wim Ceelen5. 1. Division of Lower GI, Department of Surgery, Hyogo College of Medicine, Japan; Department of Gastrointestinal Surgery, Kanagawa Cancer Center, Japan. Electronic address: kozokataoka@hotmail.co.jp. 2. Department of Applied Mathematics, Computer Science and Statistics, Ghent University, Belgium. 3. Department of Gastrointestinal Surgery, Kanagawa Cancer Center, Japan. 4. Division of Lower GI, Department of Surgery, Hyogo College of Medicine, Japan. 5. Department of GI Surgery, Ghent University Hospital, and Cancer Research Institute Ghent (CRIG), Ghent University, Belgium.
Abstract
INTRODUCTION: Debate persists on the ideal extent of lymphadenectomy for colon cancer (CC). Specifically, it is unknown whether the anatomical location of positive lymph nodes (LN) has any independent prognostic significance. We assessed the prognostic value of positive LN location in stage III CC patients who underwent extensive (D3) lymphadenectomy. METHODS: Patients from Kanagawa Cancer Center, Japan, who underwent D3 dissection for CC from 2000 to 16 were analyzed. Mesenteric LN were classified according to location as paracolic (L1), intermediate (L2), or central (L3). Recurrence-free survival (RFS) and the corresponding hazard function were evaluated with their trends over the L groups. Multivariate Cox models were used to evaluate the association of LN location with RFS. RESULTS: Four hundred forty-six stage III patients were analyzed. The mean number of examined/positive nodes per patient was 42.5/2.6 in L1 (n = 310), 40.9/4.8 in L2 (n = 111), and 44.0/9.8 in L3 (n = 25). RFS was worse for L3 vs. L2 (HR: 2.00, 95%CI [1.05-3.75], p = 0.034) and for L3 vs. L1 (2.62 [1.45-4.71], p = 0.001), but not significantly different between L2 and L1 (1.32 [0.89-1.5], p = 0.17). In a multivariate model adjusting for age, tumor size, and number of lymph nodes harvested T-stage (p < 0.001), adjuvant therapy (p < 0.0038), lymphatic invasion (p = 0.023), and LNR (p = 0.038) were significantly associated with RFS, but not L level or tumor location. CONCLUSION: The anatomical location of invaded LN does not significantly correlate with RFS in CC, after adjusting for potential confounders. Central LN are infrequently invaded and confer a worse RFS.
INTRODUCTION: Debate persists on the ideal extent of lymphadenectomy for colon cancer (CC). Specifically, it is unknown whether the anatomical location of positive lymph nodes (LN) has any independent prognostic significance. We assessed the prognostic value of positive LN location in stage III CC patients who underwent extensive (D3) lymphadenectomy. METHODS:Patients from Kanagawa Cancer Center, Japan, who underwent D3 dissection for CC from 2000 to 16 were analyzed. Mesenteric LN were classified according to location as paracolic (L1), intermediate (L2), or central (L3). Recurrence-free survival (RFS) and the corresponding hazard function were evaluated with their trends over the L groups. Multivariate Cox models were used to evaluate the association of LN location with RFS. RESULTS: Four hundred forty-six stage III patients were analyzed. The mean number of examined/positive nodes per patient was 42.5/2.6 in L1 (n = 310), 40.9/4.8 in L2 (n = 111), and 44.0/9.8 in L3 (n = 25). RFS was worse for L3 vs. L2 (HR: 2.00, 95%CI [1.05-3.75], p = 0.034) and for L3 vs. L1 (2.62 [1.45-4.71], p = 0.001), but not significantly different between L2 and L1 (1.32 [0.89-1.5], p = 0.17). In a multivariate model adjusting for age, tumor size, and number of lymph nodes harvested T-stage (p < 0.001), adjuvant therapy (p < 0.0038), lymphatic invasion (p = 0.023), and LNR (p = 0.038) were significantly associated with RFS, but not L level or tumor location. CONCLUSION: The anatomical location of invaded LN does not significantly correlate with RFS in CC, after adjusting for potential confounders. Central LN are infrequently invaded and confer a worse RFS.