Diagnostic accuracy of controlled attenuation parameter (CAP) as a non-invasive test for steatosis in suspected non-alcoholic fatty liver disease: a systematic review and meta-analysis

Abstract

Background: Controlled attenuation parameter (CAP) is a non-invasive method for diagnosing hepatic steatosis. Despite good diagnostic performance, clinical application of CAP is limited due to the influences of covariates. Here, a systematic review on the performance of CAP in the diagnosis and staging of hepatic steatosis in NAFLD patients was performed.

Methods: The sensitivity, specificity, diagnostic odds ratio (DOR) and area under receiver operating characteristics (AUROC) curves of the pooled data for CAP in diagnosing and staging the mild (Stage 1), moderate (Stage 2) and severe (Stage 3) steatosis in NAFLD patients were assessed. The clinical utility of CAP was evaluated by Fagan plot. Heterogeneity was explored using subgroup analysis.

Results: Nine studies involving 1297 patients with liver biopsy-proven NAFLD were analyzed. The pooled sensitivity of CAP in detecting mild hepatic steatosis was 87% with a specificity of 91% and a DOR of 84.35. The pooled sensitivity of CAP in detecting moderate hepatic steatosis was 85% with a specificity of 74% and a DOR of 21.28. For severe steatosis, the pooled sensitivity was 76% with a specificity of 58% and a DOR of 4.70. The mean AUROC value for CAP in the diagnosis of mild, moderate, and severe steatosis was 0.96, 0.82 and 0.70, respectively. A subgroup analysis indicated that variation in the geographic regions, cutoffs, age and body mass index (BMI) could be the potential sources of heterogeneity in the diagnosis of moderate to severe steatosis.

Conclusions: CAP should be cautiously considered as a non-invasive substitute for liver biopsy in clinical practice.

Keywords: Controlled attenuation parameter (CAP); Hepatic steatosis, diagnostic accuracy; Non-alcoholic fatty liver disease (NAFLD); Transient elastography.

Fig. 1

Article selection process

Fig. 2

Quality assessment of the included studies by methodological quality graph (a) and Cochrane Handbook (b)

Fig. 3

Forest plots and meta-analyses of studies showing pooled sensitivity (a) and specificity (b) of CAP for detection of ≥S1 steatosis (Stage 0 vs Stage 1–3) in NAFLD patients. c Summary of AUROC of CAP for the diagnosis of ≥S1 steatosis (Stage 0 vs Stage 1–3) in NAFLD patients

Fig. 4

Forest plots and meta-analyses of studies showing pooled sensitivity (a) and specificity (b) of CAP for detection of ≥S2 steatosis (Stage 0–1 vs Stage 2–3) in NAFLD patients. c Summary of AUROC of CAP for the diagnosis of ≥S2 steatosis (Stage 0–1 vs Stage 2–3) in NAFLD patients

Fig. 5

Forest plots and meta-analyses of studies showing pooled sensitivity (a) and specificity (b) of CAP for detection of ≥S3 steatosis (Stage 0–2 vs Stage 3) in NAFLD patients. c Summary of AUROC of CAP for the diagnosis of ≥S3 steatosis (Stage 0–2 vs Stage 3) in NAFLD patients