Erschienen in:
01.10.2011 | Original Article
A prognostic model in patients who receive chemotherapy for metastatic or recurrent gastric cancer: validation and comparison with previous models
verfasst von:
Dong Hoe Koo, Baek-Yeol Ryoo, Hwa Jung Kim, Min-Hee Ryu, Sung-Sook Lee, Jung-Hwa Moon, Heung-Moon Chang, Jae-Lyun Lee, Tae Won Kim, Yoon-Koo Kang
Erschienen in:
Cancer Chemotherapy and Pharmacology
|
Ausgabe 4/2011
Einloggen, um Zugang zu erhalten
Abstract
Purpose
To make up for the limitations of previous prognostic models, we developed and validated a model in patients with metastatic or recurrent gastric adenocarcinoma (AGC), and to compare with previous models.
Methods
A total of 2,805 patients received chemotherapy for AGC in Asan Medical Center between January 2000 and December 2008 and were randomly split into training and validation sets of 1,870 and 935 patients, respectively. A prognostic model was developed from the training set.
Results
The median follow-up duration was 26.5 months (range, 10.8–116.3), during which time 2,495 patients (88.9%) died. Eight factors associated with poor prognosis were identified by multivariate analysis: ECOG performance status ≥2 (2 points), no gastrectomy, peritoneal metastasis, bone metastasis (2 points), lung metastasis, alkaline phosphatase > 120 IU/l, albumin < 3.3 g/dL, and total bilirubin > 1.2 mg/dL. A prognostic model was developed by dividing patients into good (0–1 points), moderate (2–3), and poor (≥4) risk groups. The overall survival (OS) curves for three risk groups differed significantly for both the training and the validation sets (P < 0.001 each). In the training set, the median OS for the three risk groups was 14.0, 9.4, and 5.1 months, respectively. Application of three previous prognostic models to our validation set showed that the four models, including ours, had similar ability to predict survival outcomes (c-statistics, 0.5520–0.5836).
Conclusion
Validation and comparison of prognostic models indicated that our model was as effective as the previous models to stratify the patients with AGC.