MRI radiomics-based machine learning model for predicting tumor-infiltrating CD 8+ T cells and prognosis of patients with pancreatic cancer
10.3760/cma.j.cn115667-20230109-00005
- VernacularTitle:基于MRI影像组学特征的机器学习模型预测胰腺癌CD 8+T细胞浸润及预后的研究
- Author:
Mingzhi LU
1
;
Fang LIU
;
Xu FANG
;
Yun BIAN
;
Chengwei SHAO
;
Jianping LU
;
Jing LI
Author Information
1. 海军军医大学第一附属医院放射治疗科，上海 200433
- Keywords:
Pancreatic cancer;
CD 8 positive T lymphocytes;
Magnetic resonance imaging;
Radiomics;
Prognosis
- From:
Chinese Journal of Pancreatology
2023;23(5):344-352
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To investigate the value of machine learning model based on MRI in predicting the abundance of tumor infiltrating CD 8+ T cell and prognosis of pancreatic cancer patients. Methods:The clinical data of 156 patients with pathological confirmed pancreatic cancer who underwent pre-operative MRI within 7 days before surgery in the First Affiliated Hospital of Naval Medical University from January 2017 to April 2018 was retrospectively analyzed. According to the international consensus on the predictive model, a total of 116 patients from January to December 2017 were included in the training set, and a total of 40 patients from January to April 2018 were included in the validation set. With the overall survival of patients as the outcome variable, X-Tile software was used to obtain cut-off values of the percentage of CD 8+ T cells, and all patients were divided into CD 8+ T-high and -low groups. The clinical, pathological and radiological features were compared between two groups. 3D slicer software was used to draw the region of interest in each layer of the primary MR T 1- and T 2-weighted imaging, arterial phase, portal venous phase, and delayed phase images for tumor segmentation. Python package was applied to extract the radiomics features of pancreatic tumors after segmentation and the extracted features were reduced and chosen using the least absolute shrinkage and selection operator (Lasso) logistic regression algorithm. Lasso logistic regression formula was applied to calculate the rad-score. The extreme gradient boosting (XGBoost) were used to construct the machine learning predicted model. The models′ performances were determined by area under the ROC curve (AUC), sensitivity, specificity, accuracy, positive predictive value, and negative predictive value. Results:The cut-off value of the CD 8+ T-cell level was 19.09% as determined by the X-tile program. Patients in the high CD 8+ T cell group had a longer median survival than those in the low CD 8+ T cell group (25.51 month vs 22.92 month, P=0.007). The T stage in the training set and tumor size in the validation set significantly differed between the groups (all P value <0.05). A total of 1 409 radiomics features were obtained, and 19-selected features associated with the level of CD 8+ T cell were determined after being reduced by the Lasso logistic regression algorithm. The rad-score was significantly lower in the CD 8- high group (median: -0.43; range: -1.55 to 0.65) than the CD 8- low group (median: 0.22; range: -0.68 to 2.54, P<0.001). The prediction model combined the radiomics features and tumor size. In the training set, the AUC, sensitivity, specificity, accuracy, and positive and negative predictive value were 0.90 (95% CI 0.85-0.95), 75.47%, 90.48%, 0.84, 0.87, and 0.81. In the validation set, the AUC, sensitivity, specificity, accuracy, and positive and negative predictive value were 0.79 (95% CI 0.63-0.96), 90.00%, 80.00%, 0.85, 0.82, and 0.89. The predictive model can accurately distinguish patients with high and low CD 8+ T cells in pancreatic cancer. Conclusions:The radiomics-based machine learning model is valuable in predicting the CD 8+ T cells infiltrating level in pancreatic cancer patients, which could be useful in identifying potential patients who can benefit from immunotherapies.
