BACKGROUND: Lymph node metastasis (LNM) has been regarded as one of the prognostic factors in patients with ampulla of Vater carcinoma (AC). However, the consensus about an optimal cutoff value of the number of LNMs and the definition of the regional lymph nodes (RLNs) has not been achieved. METHODS: This study included 114 consecutive patients who underwent pancreatoduodenectomy for AC between January 2002 and March 2019. RESULTS: The minimum p value approach for the greatest difference in the overall survival classified the number of LNM into none (N0, n = 66), from 1 to 2 (N1, n = 32), and ≥3 LNM (N2, n = 11) (p = 0.004). Distant LNM was defined as M1 (n = 5). Significant differences in relapse-free survival (RFS) were found between N0 and N1 (p < 0.001), N1 and N2 (p = 0.047), and N1 and M1 (p = 0.044) but not between N2 and M1 (p = 0.683). Moreover, the patients with regional LNM were classified into two groups: Np group (n = 35, LNM only in pancreatic head region) and Nd group (n = 8, LNM in other regional location). Significant differences in the RFS were found between N0 and Np (p < 0.001), Np and Nd (p = 0.004), and Np and M1 (p = 0.033) but not between Nd and M1 (p = 0.883). A Cox proportional hazards analysis for RFS revealed that ≥ 3 LNMs (hazards ratio [HR], 3.22) and LNM except for pancreatic head region (HR, 4.27) were individually independent worse prognostic factors. CONCLUSIONS: ≥3 LNMs and regional LNM except for pancreatic head region were associated with poor prognosis comparable to that of the patients with M1.
BACKGROUND: Lymph node metastasis (LNM) has been regarded as one of the prognostic factors in patients with ampulla of Vater carcinoma (AC). However, the consensus about an optimal cutoff value of the number of LNMs and the definition of the regional lymph nodes (RLNs) has not been achieved. METHODS: This study included 114 consecutive patients who underwent pancreatoduodenectomy for AC between January 2002 and March 2019. RESULTS: The minimum p value approach for the greatest difference in the overall survival classified the number of LNM into none (N0, n = 66), from 1 to 2 (N1, n = 32), and ≥3 LNM (N2, n = 11) (p = 0.004). Distant LNM was defined as M1 (n = 5). Significant differences in relapse-free survival (RFS) were found between N0 and N1 (p < 0.001), N1 and N2 (p = 0.047), and N1 and M1 (p = 0.044) but not between N2 and M1 (p = 0.683). Moreover, the patients with regional LNM were classified into two groups: Np group (n = 35, LNM only in pancreatic head region) and Nd group (n = 8, LNM in other regional location). Significant differences in the RFS were found between N0 and Np (p < 0.001), Np and Nd (p = 0.004), and Np and M1 (p = 0.033) but not between Nd and M1 (p = 0.883). A Cox proportional hazards analysis for RFS revealed that ≥ 3 LNMs (hazards ratio [HR], 3.22) and LNM except for pancreatic head region (HR, 4.27) were individually independent worse prognostic factors. CONCLUSIONS: ≥3 LNMs and regional LNM except for pancreatic head region were associated with poor prognosis comparable to that of the patients with M1.
Authors: R Tang; J Y Wang; J S Chen; C R Chang-Chien; S Tang; S E Lin; Y T You; K C Hsu; Y S Ho; H A Fan Journal: J Am Coll Surg Date: 1995-06 Impact factor: 6.113