Anzeige
Erschienen in:
03.05.2018 | Gastrointestinal Oncology
Is the 8th Edition of the AJCC TNM Staging System Sufficiently Reasonable for All Patients with Noncardia Gastric Cancer? A 12,549-Patient International Database Study
Erschienen in: Annals of Surgical Oncology | Ausgabe 7/2018
Einloggen, um Zugang zu erhaltenAbstract
Background
The aim of this work is to compare the prognostic ability between the 7th and 8th editions of the American Joint Committee on Cancer (AJCC) tumor–node–metastasis (TNM) classification for gastric cancer (GC).
Methods
A total of 10,194 noncardia GC patients were identified from the Surveillance, Epidemiology, and End Results database from 1988 to 2008. Concordance index (C-index), bayesian information criterion (BIC), and time-dependent receiver operating characteristic (t-ROC) analyses were used. External validation was performed using a dataset (n = 2355) derived from Fujian Medical University Union Hospital.
Results
Overall survival for all five AJCC N categories differed significantly when patients were subgrouped into ≤ 15 versus >15 examined lymph nodes (eLNs). The prognostic ability of the 8th edition (C-index 0.716) was not improved over the 7th edition (C-index 0.716). Subgroup analysis showed superior performance of the 8th over the 7th edition in patients with > 15 eLNs (C-index 0.742 vs. 0.735); however, the two editions showed similar performance for patients with ≤ 15 eLNs (C-index 0.713 vs. 0.713). The BIC and t-ROC analyses were consistent. To better predict the prognosis of patients with ≤ 15 eLNs, we established a novel prognostic model based on independent prognostic factors (C-index 0.735). BIC analysis showed that this new model was better than the 7th and 8th editions. Similar results were obtained from the validation set.
Conclusions
The 8th edition of the AJCC TNM classification shows better prognostic ability than the 7th edition in noncardia GC patients with > 15 eLNs, but no improvement was found in patients with ≤ 15 eLNs; therefore, a novel prognostic model is proposed.