CT Textural Analysis (CTTA) of Metastatic Treatment‐Resistant Pancreatic Adenocarcinoma (PDAC): Identifying Biomarkers for Genetic Instability and Overall Survival

Persistent Link:
http://hdl.handle.net/10150/603564
Title:
CT Textural Analysis (CTTA) of Metastatic Treatment‐Resistant Pancreatic Adenocarcinoma (PDAC): Identifying Biomarkers for Genetic Instability and Overall Survival
Author:
Campbell, David
Affiliation:
The University of Arizona College of Medicine - Phoenix
Issue Date:
23-Mar-2016
Rights:
Copyright © is held by the author. Digital access to this material is made possible by the College of Medicine - Phoenix, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Collection Information:
This item is part of the College of Medicine - Phoenix Scholarly Projects 2016 collection. For more information, contact the Phoenix Biomedical Campus Library at pbc-library@email.arizona.edu.
Publisher:
The University of Arizona.
Abstract:
Significance: Metastatic, treatment‐resistant pancreatic ductal adenocarcinoma (PDAC) is a rapidly fatal disease that typically carries a bleak prognosis. Contrast‐enhanced CT is the current standard of care tool for imaging evaluation, and repeat imaging is routinely performed in clinical trials. The availability of these imaging data render them exploitable for further analysis. CT texural analysis (CTTA), a quantitative tool for examining a region of interest on CT and generating statistical parameters based on gray‐level pixel data, is powerful technique that has been studied in other cancers and shown to correlate with features such as tumor grade, stage, and prognosis. However, the application of CTTA to PDAC has not been studied. Given the paucity of diagnostic tests to guide therapy, validated CTTA biomarkers could be immensely useful. Identifying PDAC variants that have a relative deficit in DNA repair might allow these cancers to be treated with targeted cytotoxic regimens sooner. Additionally, identifying prognostic CTTA parameters would be useful in gauging the severity of disease. Objective: We sought to perform quantitative textural analysis on CT imaging from a clinical trial cohort of patients with metastatic, treatment‐resistant PDAC. We aimed to correlate CTTA features to molecular profiling results (copy number variations obtained by array CGH) and clinical features (overall survival). Design: Metastatic tumor sites from patients with treatment‐resistant PDAC were biopsied and molecularly profiled. Intrachromosal copy number were assessed by CGH in tumor specimens, and patients were treated based on these individual molecular profiling results. Pre‐biopsy portal‐venous phase and non‐contrast CT scans were obtained for retrospective analysis (n=15). CTTA was performed by drawing regions of interest around the primary pancreas adenocarcinoma and the normal pancreas tissue. CTTA parameters including mean positive pixels, entropy, kurtosis, and skewness were derived using the TexRAD platform at texture filtering densities of 0, 2, 3, 4, 5, and 6 pixels. CTTA values were then compared to intrachromosomal copy number variation (CNV) per tumor and overall survival (OS) posttreatment using a Spearman’s rank correlation coefficient. Additional linear regression analysis was performed for positive correlations, and a Kaplan‐Meier statistic was generated for OS using median CTTA entropy. Multivariate analyses for CNV and OS were also performed. Results: CNV were negatively correlated with the kurtosis value of the primary tumor mass using medium texture filtering (p=0.034, n=15). Linear regression revealed a significant negative correlation between kurtosis and CNV (p=0.038). Secondary analysis of the normal pancreas using coarse texture filtering revealed that increasing entropy was associated with decreased OS (p=0.0014, n=12). Using median entropy as a cutoff value (median: 4.165), median OS was greater in the entropy < 4.165 group versus the entropy > 4.165 group (179 days v 43 days; 95% CI 73.137 – 166.87; p=0.004, n=12). Conclusions: This exploratory study with admittedly limited sample size raises interesting questions about the use of CTTA parameters as diagnostic tools and/or biopsy adjuncts in assessing PDAC susceptibility to commercially available cytotoxics. Secondarily, entropy, a potential marker of heterogeneity and inflammation in the normal pancreas, represents an intriguing possibility for gauging prognosis.
MeSH Subjects:
Adenocarcinoma; Pancreatic Neoplasms; Biomarkers; Tomography, X-Ray Computed
Description:
A Thesis submitted to The University of Arizona College of Medicine - Phoenix in partial fulfillment of the requirements for the Degree of Doctor of Medicine.
Mentor:
Korn, Ronald MD, PhD

Full metadata record

DC FieldValue Language
dc.language.isoen_USen
dc.titleCT Textural Analysis (CTTA) of Metastatic Treatment‐Resistant Pancreatic Adenocarcinoma (PDAC): Identifying Biomarkers for Genetic Instability and Overall Survivalen_US
dc.contributor.authorCampbell, Daviden
dc.contributor.departmentThe University of Arizona College of Medicine - Phoenixen
dc.date.issued2016-03-23en
dc.rightsCopyright © is held by the author. Digital access to this material is made possible by the College of Medicine - Phoenix, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.en_US
dc.description.collectioninformationThis item is part of the College of Medicine - Phoenix Scholarly Projects 2016 collection. For more information, contact the Phoenix Biomedical Campus Library at pbc-library@email.arizona.edu.en_US
dc.publisherThe University of Arizona.en
dc.description.abstractSignificance: Metastatic, treatment‐resistant pancreatic ductal adenocarcinoma (PDAC) is a rapidly fatal disease that typically carries a bleak prognosis. Contrast‐enhanced CT is the current standard of care tool for imaging evaluation, and repeat imaging is routinely performed in clinical trials. The availability of these imaging data render them exploitable for further analysis. CT texural analysis (CTTA), a quantitative tool for examining a region of interest on CT and generating statistical parameters based on gray‐level pixel data, is powerful technique that has been studied in other cancers and shown to correlate with features such as tumor grade, stage, and prognosis. However, the application of CTTA to PDAC has not been studied. Given the paucity of diagnostic tests to guide therapy, validated CTTA biomarkers could be immensely useful. Identifying PDAC variants that have a relative deficit in DNA repair might allow these cancers to be treated with targeted cytotoxic regimens sooner. Additionally, identifying prognostic CTTA parameters would be useful in gauging the severity of disease. Objective: We sought to perform quantitative textural analysis on CT imaging from a clinical trial cohort of patients with metastatic, treatment‐resistant PDAC. We aimed to correlate CTTA features to molecular profiling results (copy number variations obtained by array CGH) and clinical features (overall survival). Design: Metastatic tumor sites from patients with treatment‐resistant PDAC were biopsied and molecularly profiled. Intrachromosal copy number were assessed by CGH in tumor specimens, and patients were treated based on these individual molecular profiling results. Pre‐biopsy portal‐venous phase and non‐contrast CT scans were obtained for retrospective analysis (n=15). CTTA was performed by drawing regions of interest around the primary pancreas adenocarcinoma and the normal pancreas tissue. CTTA parameters including mean positive pixels, entropy, kurtosis, and skewness were derived using the TexRAD platform at texture filtering densities of 0, 2, 3, 4, 5, and 6 pixels. CTTA values were then compared to intrachromosomal copy number variation (CNV) per tumor and overall survival (OS) posttreatment using a Spearman’s rank correlation coefficient. Additional linear regression analysis was performed for positive correlations, and a Kaplan‐Meier statistic was generated for OS using median CTTA entropy. Multivariate analyses for CNV and OS were also performed. Results: CNV were negatively correlated with the kurtosis value of the primary tumor mass using medium texture filtering (p=0.034, n=15). Linear regression revealed a significant negative correlation between kurtosis and CNV (p=0.038). Secondary analysis of the normal pancreas using coarse texture filtering revealed that increasing entropy was associated with decreased OS (p=0.0014, n=12). Using median entropy as a cutoff value (median: 4.165), median OS was greater in the entropy < 4.165 group versus the entropy > 4.165 group (179 days v 43 days; 95% CI 73.137 – 166.87; p=0.004, n=12). Conclusions: This exploratory study with admittedly limited sample size raises interesting questions about the use of CTTA parameters as diagnostic tools and/or biopsy adjuncts in assessing PDAC susceptibility to commercially available cytotoxics. Secondarily, entropy, a potential marker of heterogeneity and inflammation in the normal pancreas, represents an intriguing possibility for gauging prognosis.en
dc.typeThesisen
dc.subject.meshAdenocarcinomaen
dc.subject.meshPancreatic Neoplasmsen
dc.subject.meshBiomarkersen
dc.subject.meshTomography, X-Ray Computeden
dc.descriptionA Thesis submitted to The University of Arizona College of Medicine - Phoenix in partial fulfillment of the requirements for the Degree of Doctor of Medicine.en
dc.contributor.mentorKorn, Ronald MD, PhDen
All Items in UA Campus Repository are protected by copyright, with all rights reserved, unless otherwise indicated.