BACKGROUND: There is little information to determine prognosis or to guide clinical care for patients with duodenal adenocarcinoma. We have hypothesized that survival following resection of duodenal cancer is similar to survival following resection of distal gastric cancer. We tested the utility of a nomogram created for determining disease-specific survival (DSS) after R0 resection of gastric cancer in estimating DSS for patients with resected duodenal cancer. METHODS: Review of a prospective database identified 106 patients who underwent R0 resection of duodenal cancer. Comparison was made to 459 patients with distal gastric cancer. The Student t test, Fisher exact test, Pearson chi-square test, and log-rank test were used to assess statistical significance. Concordance probabilities and calibration plots were used for nomogram validation. RESULTS: Duodenal cancers were more deeply invasive than gastric cancer (P < .01). The rate of lymph node positivity was not statistically different between the two tumors; however, there were differences in the rate of nodal positivity for certain depths of penetration. Younger age (P = .002), negative regional lymph nodes (P = .03), and tumors confined to the bowel wall or subserosa (P = .03) were associated with improved DSS for duodenal cancer. When applied to patients with duodenal cancer, the nomogram had a concordance probability of 0.70, and calibration appeared to be accurate. CONCLUSIONS: A nomogram created for determining DSS after resection of gastric cancer predicts outcome for duodenal cancer patients and may prove to be useful for research and in guiding clinical care.
BACKGROUND: There is little information to determine prognosis or to guide clinical care for patients with duodenal adenocarcinoma. We have hypothesized that survival following resection of duodenal cancer is similar to survival following resection of distal gastric cancer. We tested the utility of a nomogram created for determining disease-specific survival (DSS) after R0 resection of gastric cancer in estimating DSS for patients with resected duodenal cancer. METHODS: Review of a prospective database identified 106 patients who underwent R0 resection of duodenal cancer. Comparison was made to 459 patients with distal gastric cancer. The Student t test, Fisher exact test, Pearson chi-square test, and log-rank test were used to assess statistical significance. Concordance probabilities and calibration plots were used for nomogram validation. RESULTS:Duodenal cancers were more deeply invasive than gastric cancer (P < .01). The rate of lymph node positivity was not statistically different between the two tumors; however, there were differences in the rate of nodal positivity for certain depths of penetration. Younger age (P = .002), negative regional lymph nodes (P = .03), and tumors confined to the bowel wall or subserosa (P = .03) were associated with improved DSS for duodenal cancer. When applied to patients with duodenal cancer, the nomogram had a concordance probability of 0.70, and calibration appeared to be accurate. CONCLUSIONS: A nomogram created for determining DSS after resection of gastric cancer predicts outcome for duodenal cancerpatients and may prove to be useful for research and in guiding clinical care.
Authors: Michael David Johnson; Richard Mackey; Nancy Brown; James Church; Carol Burke; R Matthew Walsh Journal: J Gastrointest Surg Date: 2010-02 Impact factor: 3.452