CT prediction of resectability and prognosis in patients with pancreatic ductal adenocarcinoma after neoadjuvant treatment using image findings and texture analysis
To assess utility of CT findings and texture analysis for predicting the resectability and prognosis in patients after neoadjuvant therapy for pancreatic ductal adenocarcinoma (PDAC).
Materials and methods
Among 308 patients, 45 with PDAC underwent neoadjuvant therapy (concurrent-chemoradiation-therapy, CCRT, n = 27 and chemotherapy, ChoT, n = 18) before surgery were included. All underwent baseline and preoperative CT. Two reviewers assessed CT findings and resectability. We analyzed relationship between CT resectability and residual tumor. CT texture values obtained by subtracting preoperative from baseline CT were analyzed using multivariate Cox/logistic regression analysis to identify significant parameters predicting resectability and prognosis.
There were 30 patients without residual tumor (CCRT, n = 20; ChoT, n = 10) and 15 with residual tumor (CCRT, n = 7; ChoT, n = 8). Considering borderline as resectable was more accurate for R0 resectability than considering borderline as unresectable (68.9% vs 55.6% and 51.1%, p < 0.001). Particularly, neoadjuvant CCRT provided better accuracy than that in (p < 0.001). In CT texture analysis, higher subtracted entropy (cut-off: 0.03, HR 0.159, p = 0.005) and lower subtracted GLCM entropy (cut-off: –0.35, HR 10.235, p = 0.036) are important parameters for prediction of longer overall survival.
CT findings with texture analysis can be useful for predicting a patient’s outcome, including resectability and prognosis, after neoadjuvant therapy for PDAC.
• Considering borderline resectable tumor as resectable have better accuracy for resectability.
• Considering borderline as resectable, CCRT-patients have better resectability accuracy than chemotherapy-patients.
• Higher subtracted entropy and lower subtracted GLCM entropy are predictors of favorable outcome.
KeywordsPancreatic neoplasm Neoadjuvant therapy Pancreatectomy Prognosis Diagnosis
Angular second moment
Concurrent chemoradiation therapy
Grey level co-occurrence matrices
Inverse difference moment
Superior mesenteric vein
We also thank Bonnie Hami, M.A. (USA) for her editorial assistance in the preparation of this manuscript.
The authors state that this work has not received any funding.
Compliance with ethical standards
The scientific guarantor of this publication is Joon Koo Han, M.D.
Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
Statistics and biometry
Seo-Youn Choi MD has significant statistical expertise and no complex statistical methods were necessary for this paper.
Written informed consent was waived by the institutional review board.
Institutional review board approval was obtained.
Study subjects or cohorts overlap
Among 45 patients who were enrolled in our study, 13 patients have been previously reported in our previous paper (AJR Am J Roentgenol 2018, 210(5):1059–1065). However, the study purposes of these two studies were different. The previously published paper was a comparison of the diagnostic performance of CT in assessing tumor resectability pancreatic cancers after receiving neoadjuvant chemoradiation in comparison with those undergoing up-front surgery. The purpose of this study was to assess the utility of CT findings and texture analysis for predicting the resectability and prognosis in patients after neoadjuvant therapy for pancreatic cancer.
• Diagnostic or prognostic study
• Performed at one institution
- 1.Society AC (2017) Cancer facts & figures 2017. American Cancer Society, AtlantaGoogle Scholar
- 9.Soriano A, Castells A, Ayuso C et al (2004) Preoperative staging and tumor resectability assessment of pancreatic cancer: prospective study comparing endoscopic ultrasonography, helical computed tomography, magnetic resonance imaging, and angiography. Am J Gastroenterol 99:492–501CrossRefGoogle Scholar
- 23.Sobin L, Gospodarowicz MK, Wittekind C (2010) International Union against Cancer TNM classification of malignant tumors, 7th ed. 2009 edn. Wiley-Blackwell, ChichesterGoogle Scholar
- 30.Chee CG, Kim YH (2017) CT texture analysis in patients with locally advanced rectal cancer treated with neoadjuvant chemoradiotherapy: A potential imaging biomarker for treatment response and prognosis. 12:e0182883Google Scholar
- 37.Albregtsen F (2008) Statistical texture measures computed from gray level coocurrence matrices. Image Processing Laboratory, Department of Informatics, University of Oslo Web site; Available via http://www.uio.no/studier/emner/matnat/ifi/INF4300/h08/undervisningsmateriale/glcm.pdf
- 38.Partio M, Cramariuc B, Gabbouj M, Visa A (2002) Rock texture retrieval using gray level co-occurrence matrix Proc of 5th Nordic Signal Processing SymposiumGoogle Scholar