PREOPERATIVE PREDICTION OF WALL INVASION DEPTH OF PRIMARY TUMOR (PT) IN GASTRIC CANCER PATIENTS
The limited effectiveness of traditional examination methods for patients with locally advanced gastric cancer (LAGC) leads to inadequate clinical staging and an irrational approach to treatment planning. Therefore, developing and utilizing prognostic models to accurately assess the extent of local tumor spread in patients with LAGC is essential. Objective: To create a prognostic nomogram that evaluates the likelihood of serosal invasion of the primary tumor in the stomach wall (рТ4) by analyzing preoperative data. This model aims to optimize treatment plans and enhance the effectiveness of treatment for LAGC. Methods: A retrospective analysis was conducted on the treatment outcomes of 1,054 patients who underwent radical surgery for metastatic gastric cancer (mGC). The study examined the relationship between the depth of primary tumor invasion into the gastric wall (GVI) and various preoperative clinical, morphological, and laboratory parameters. Significant risk factors identified in the univariate analysis were further assessed as independent variables in a multivariate logistic regression analysis. These variables were then incorporated into a nomogram. The clinical validation of the model was performed by evaluating the long-term treatment outcomes. Results: It has been established that the following factors are prognostically significant for assessing the likelihood of invasion of the serosa of the stomach by the primary tumor (рТ4) preoperatively: 1. Size of the primary tumor (natural logarithm) – odds ratio (OR) 5.5 (95% CI 3.8-8.3), p<0.001.2. Infiltrative variant of the macroscopic growth form – OR 1.9 (95% CI 1.3-3.0), p=0.002. 3. Total or subtotal involvement of the stomach – OR 1.8 (95% CI 1.1-3.0), p=0.029. 4. Non-cohesive adenocarcinoma (high grade, GIII) – OR 1.7 (95% CI 1.1-2.9), p=0.029. 5. Fibrinogen level increase of 1 g/l – OR 1.5 (95% CI 1.2-1.8), p<0.001. Based on a linear combination of these predictors, a prognostic model in the form of a nomogram was developed. This model demonstrates high predictive accuracy, with a concordance index of 0.826 (95% CI 0.78-0.86). Conclusion: Careful consideration of the clinical and morphological features of the tumor process within the framework of applying the developed prognostic model increases the accuracy of determining the preoperative T category, bringing the predicted clinical T-stage (prT) as close as possible to the proper рТ stage determined by histopathological examination (HPE). Keywords: Locally advanced gastric cancer, preoperative pT4 prediction, nomogram.
Сведения
- Журнал
- Avicenna Bulletin
- Опубликовано
- 2025 · Т. 27, № 2, С. 327-339
- Язык
- Английский
- ISSN
- 2074-0581 · 2959-6327
- Издатель
- Avicenna Tajik State Medical University