Survival Prediction in Pancreatic Ductal Adenocarcinoma by Quantitative Computed Tomography Image Analysis

Abstract

Background: Pancreatic cancer is a highly lethal cancer with no established a priori markers of survival. Existing nomograms rely mainly on post-resection data and are of limited utility in directing surgical management. This study investigated the use of quantitative computed tomography (CT) features to preoperatively assess survival for pancreatic ductal adenocarcinoma (PDAC) patients.

Methods: A prospectively maintained database identified consecutive chemotherapy-naive patients with CT angiography and resected PDAC between 2009 and 2012. Variation in CT enhancement patterns was extracted from the tumor region using texture analysis, a quantitative image analysis tool previously described in the literature. Two continuous survival models were constructed, with 70% of the data (training set) using Cox regression, first based only on preoperative serum cancer antigen (CA) 19-9 levels and image features (model A), and then on CA19-9, image features, and the Brennan score (composite pathology score; model B). The remaining 30% of the data (test set) were reserved for independent validation.

Results: A total of 161 patients were included in the analysis. Training and test sets contained 113 and 48 patients, respectively. Quantitative image features combined with CA19-9 achieved a c-index of 0.69 [integrated Brier score (IBS) 0.224] on the test data, while combining CA19-9, imaging, and the Brennan score achieved a c-index of 0.74 (IBS 0.200) on the test data.

Conclusion: We present two continuous survival prediction models for resected PDAC patients. Quantitative analysis of CT texture features is associated with overall survival. Further work includes applying the model to an external dataset to increase the sample size for training and to determine its applicability.

FIG. 1

Left Extraction of the tumor from preoperative CT angiography of the pancreas. Middle Appearance of the segmented tumor with magnified area with grayscale pixels (black box). Right List of image features extracted from the tumor region. GLCM gray-level co-occurrence matrix, RLM run-length matrix, LBP local binary pattern, FD fractal dimension, IH intensity histogram, ACM angle co-occurrence matrix, CT computed tomography

FIG. 2

Model building and evaluation. The cohort was randomly split into training and testing groups. Univariate analysis was performed on the training group to select features significant for overall survival. Two multivariate continuous survival models were built on the training data and assessed using c-indices and IBS. The two models were then validated on the testing group and assessed similarly using c-indices and IBS. IBS integrated Brier scores, pre-op preoperative, post-op postoperative, CA cancer antigen, c-indices Concordance indices

FIG. 3

Study cohort. PDAC pancreatic ductal adenocarcinoma, pre-op preoperative, CT computed tomography

FIG. 4

Representative images of patients with good and poor overall survival. The magnified tumor region supports our hypothesis that heterogeneously hypoattenuating tumors are prognostic of poor survival