Development and validation of a prognostic nomogram for predicting early recurrence after curative resection of stage II/III gastric cancer.

摘要: The biological behavior of early recurrence is more invasive and the prognosis worse in gastric cancer (GC). risk (ER) for GC stage II/III has not been reported which majority patients are China. Therefore, it necessary to analyze ER II/III. medical records 1511 consecutive who received resections were retrospectively reviewed. They randomly classified into either a development or validation group at ratio 7:3. nomogram was constructed based on prognostic factors using logistic regression analysis validated by bootstrap resampling dataset, respectively. Concordance index (C-index) values calibration curves used evaluate predictive accuracy discriminatory capability. Three hundred eleven experienced ER, accounting 20.58% investigated. Multivariate identified tumors located upper, middle third, mixed, positive lymph node ≥ 0.335, pTNM III, lymphocyte count < 1.5 × 109/L, postoperative infection complications adjuvant chemotherapy 6 cycles all independent predictors after curative resection GC. C-index value obtained model 0.780 (95% CI, 0.747–0.813), yielded 0.739 0.684–0.794), suggesting practicability model. developed predicting surgery had good verified through both internal external validation. established can assist clinicians determining optimal therapy strategies counseling, treatments, subsequent follow-up planning.