Back to Search Start Over

Advanced Hepatocellular Carcinoma: Pretreatment Contrast-enhanced CT Texture Parameters as Predictive Biomarkers of Survival in Patients Treated with Sorafenib

Authors :
Sébastien Mulé
Alain Luciani
Charlotte Costentin
Christine Hoeffel
Gérard Thiéfin
Carole Durot
Alain Rahmouni
Centre de Recherche en Sciences et Technologies de l'Information et de la Communication - EA 3804 (CRESTIC)
Université de Reims Champagne-Ardenne (URCA)
Source :
Radiology, Radiology, Radiological Society of North America, 2018, 288 (2), pp.445-455. ⟨10.1148/radiol.2018171320⟩
Publication Year :
2018

Abstract

Purpose To determine whether texture features on pretreatment contrast material-enhanced computed tomographic (CT) images can help predict overall survival (OS) and time to progression (TTP) in patients with advanced hepatocellular carcinoma (HCC) treated with sorafenib. Materials and Methods This retrospective study included 92 patients with advanced HCC treated with sorafenib between January 2009 and April 2015 at two independent university hospitals. Sixty-four of the 92 patients (70%) (six women, 58 men; median age, 66 years) were included from institution 1 and constituted a training cohort; 28 patients (30%) (five women, 23 men; median age, 64 years) were included from institution 2 and constituted a validation cohort. Pretreatment CT texture analysis was performed on late arterial and portal venous phase HCC images. Mean gray-level intensity, entropy, kurtosis, skewness, and standard deviation values were derived from the pixel distribution histogram before and after spatial filtration at different anatomic scales ranging from fine to coarse texture. Lesion heterogeneity was also visually graded on a 4-point scale. Correlations between visual analysis and texture parameters were assessed with the Spearman rank correlation. Univariate Kaplan-Meier and multivariate Cox proportional hazards regression analyses were performed in the training cohort to identify independent predictors of OS and TTP. Their predictive capacity was tested on the validation cohort by using Kaplan-Meier analysis. Results Visual analysis of tumor heterogeneity correlated with entropy at both arterial (P = .012) and portal venous (P = .038) phases. Portal phase-derived entropy at fine (hazard ratio [HR], 5.08; P = .0033), medium (HR, 2.23; P = .019), and coarse (HR, 2.26; P = .0032) texture scales was identified as an independent predictor of OS and confirmed in the validation cohort (P.05). The difference in median survival between patients in the validation cohort with entropy values below and above the identified threshold was 272 days (with fine texture) and 741 days (with medium and coarse textures). Arterial phase-derived texture parameters (P.085) and visual analysis (P.11) were not associated with changes in survival. Conclusion Pretreatment portal venous phase-derived tumor entropy may be a predictor of survival in patients with advanced HCC treated with sorafenib.

Details

ISSN :
15271315 and 00338419
Volume :
288
Issue :
2
Database :
OpenAIRE
Journal :
Radiology
Accession number :
edsair.doi.dedup.....c8a1913ac0b81a139e86a88dc7003a45