Steatosis in patients with nonalcoholic fatty liver disease (NAFLD) is often benign, but may progress to fibrosis. The accurate diagnosis of hepatic steatosis is therefore important for clinical decision-making and prognostic assessments. The controlled attenuation parameter (CAP), a noninvasive measurement obtained with Fibro-Scan, has been developed for liver steatosis assessment. CAP performs poorly in patients with high BMI. The XL probe was initially developed for measuring liver stiffness in overweight patients. We assessed the diagnostic value of CAP in candidates for bariatric surgery with suspected NAFLD examined with the XL probe.
For the retrospective group, raw ultrasonic radiofrequency signals were stored prospectively in the Fibro-Scan examination file for offline CAP calculation in 194 consecutive obese patients undergoing liver stiffness measurement in the 15 days before liver biopsy. For the prospective group, CAP was calculated automatically and prospectively from the XL probe in 123 obese patients.
In the retrospective group, the diagnostic accuracy of CAP was satisfactory for differentiating S3 from S0–S1–S2 (0.79±0.03; 95% confidence interval: 0.71–0.84) and S3 from S0 (0.85±0.05; 95% confidence interval: 0.73–0.92). The Obuchowski measure demonstrated a very good discriminatory performance: 0.87±0.02 in the retrospective group and 0.91±0.02 in the prospective group.
CAP calculations from XL probe measurements efficiently detected severe steatosis in morbidly obese patients with suspected NAFLD. However, the cutoff values should now be confirmed in a larger prospective cohort.
Departments of aHepato-Gastroenterology and Nutrition
cDigestive Minimally Invasive Surgery, Antoine-Beclere Hospital, APHP
dINSERM, U996, Clamart
eFaculty of Medicine Paris-South, University Paris-South, Orsay, France
Correspondence to Sylvie Naveau, MD, Department of Hepato-Gastroenterology and Nutrition, Antoine-Beclere Hospital, 157 rue de la Porte de Trivaux, 92141 Clamart cedex, France Tel: +33 145 374 372; fax: +33 140 940 656; e-mail: email@example.com
Received January 24, 2017
Accepted May 9, 2017