In the JournalsPerspective

FibroTest yielded suboptimal accuracy in detecting significant fibrosis, cirrhosis

FibroTest showed low accuracy in detecting significant fibrosis and cirrhosis among patients with chronic hepatitis B-related cirrhosis, according to results from a meta-analysis.

Nermin N. Salkic, MD, PhD, department of gastroenterology and hepatology, University Clinical Center Tuzla, Bosnia and Herzegovina, and colleagues included 16 studies with 2,494 patients after searching Medline, Embase and Cochrane Library databases. Studies were used in a hierarchical summary receiver operating curve model to determine the presence of liver fibrosis, while 13 of the studies (n=1,754) were included in a different hierarchical summary receiver operating curve model to assess liver cirrhosis.

Nermin N. Salkic

The AUC for significant liver fibrosis was 0.84 (95% CI, 0.78-0.88) and was 0.87 (95% CI, 0.85-0.9) in the studies for significant liver cirrhosis. At a FibroTest threshold of 0.74, the sensitivity, specificity and diagnostic odds ratio (DOR) for cirrhosis were 62 (47-75%), 91 (88-93%) and 15.7% (8.6-28.8), respectively. With a threshold of 0.48, the sensitivity, specificity and DOR of the FibroTest for significant fibrosis were 61 (48-72%), 80 (72-86%) and 6.2% (3.3-11.9), respectively.

In the fibrosis curve, strong heterogeneity among the studies was observed. Heterogeneity also was present among the cirrhosis curve, but in studies that used a FibroTest threshold less than 0.74, reduced heterogeneity occurred.

“FibroTest is of excellent utility for excluding cirrhosis in patients with chronic hepatitis B,” Salkic told Healio.com/Hepatology. “The diagnostic performance of FibroTest in detection of significant fibrosis and cirrhosis and exclusion of significant fibrosis is suboptimal. It is important for clinicians to adhere to the recommended thresholds of FibroTest until better ones are derived.”

Disclosure: The researchers report no relevant financial disclosures.

FibroTest showed low accuracy in detecting significant fibrosis and cirrhosis among patients with chronic hepatitis B-related cirrhosis, according to results from a meta-analysis.

Nermin N. Salkic, MD, PhD, department of gastroenterology and hepatology, University Clinical Center Tuzla, Bosnia and Herzegovina, and colleagues included 16 studies with 2,494 patients after searching Medline, Embase and Cochrane Library databases. Studies were used in a hierarchical summary receiver operating curve model to determine the presence of liver fibrosis, while 13 of the studies (n=1,754) were included in a different hierarchical summary receiver operating curve model to assess liver cirrhosis.

Nermin N. Salkic

The AUC for significant liver fibrosis was 0.84 (95% CI, 0.78-0.88) and was 0.87 (95% CI, 0.85-0.9) in the studies for significant liver cirrhosis. At a FibroTest threshold of 0.74, the sensitivity, specificity and diagnostic odds ratio (DOR) for cirrhosis were 62 (47-75%), 91 (88-93%) and 15.7% (8.6-28.8), respectively. With a threshold of 0.48, the sensitivity, specificity and DOR of the FibroTest for significant fibrosis were 61 (48-72%), 80 (72-86%) and 6.2% (3.3-11.9), respectively.

In the fibrosis curve, strong heterogeneity among the studies was observed. Heterogeneity also was present among the cirrhosis curve, but in studies that used a FibroTest threshold less than 0.74, reduced heterogeneity occurred.

“FibroTest is of excellent utility for excluding cirrhosis in patients with chronic hepatitis B,” Salkic told Healio.com/Hepatology. “The diagnostic performance of FibroTest in detection of significant fibrosis and cirrhosis and exclusion of significant fibrosis is suboptimal. It is important for clinicians to adhere to the recommended thresholds of FibroTest until better ones are derived.”

Disclosure: The researchers report no relevant financial disclosures.

    Perspective
    William Carey

    William Carey

    Liver biopsy is referred to as the “gold standard” in assessing both the activity and degree of fibrosis in many chronic liver diseases including hepatitis B. It is not likely to retain this lofty status much longer. Liver biopsy has several important drawbacks. Among them are cost, risk for complications, need for additional health care resources, patient and physician aversion to the procedure, inadequate specimen size and the lack of specific findings.

    Liver biopsy adds between $2,500 to $3,500 to the cost of an evaluation (even higher for transvenous liver biopsy). Approximately 20% of patients will experience significant pain following percutaneous liver biopsy. More severe complications include pneumothorax, major bleeding, inadvertent biopsy of the kidney or colon, and perforation of the gallbladder. Death, most often due to uncontrolled bleeding, may occur in up to 1 in 1,000 biopsies. Underappreciated is the risk of no-representative sampling, either because of the small size of biopsy specimen or patchy distribution of fibrosis.

    Noninvasive measures to assess hepatic fibrosis have been around for a generation and are increasingly used as a substitute for liver biopsy. The 2014 medley of noninvasive estimates of hepatic fibrosis includes FibroTest/FibroSure, APRI, FIB-4, other serum based test combinations, and elastography (either ultrasound- or MRI-based). Noninvasive tests have potential both for determination of current liver damage and for monitoring disease progression. They can be done at a fraction of the cost of a liver biopsy. Salkic and colleagues have reported the results of an exquisitely performed meta-analysis of peer reviewed published reports and confirmed the value of FibroTest/FibroSure in hepatitis B — mainly in excluding the diagnosis of cirrhosis. The findings of this review are restricted to hepatitis B, but others have shown similar findings in hepatitis C and alcoholic liver disease.

    While there is growing consensus that noninvasive markers provide valuable information, allowing the clinician to make important decisions about treatment, screening for varices and hepatocellular carcinoma, it is essential to understand limitations of FibroTest, including distortions in results in individuals with Gilbert’s syndrome and in those with hemolysis. This study reiterates the relative insensitivity of noninvasive tests in discriminating between lesser degrees of fibrosis (F0, F1, and F2).

    Data are accumulating to suggest noninvasive markers (combining a noninvasive test plus ultrasound-based elastrography, for example) are powerful tools in assessing natural history of individuals with many chronic liver diseases including hepatitis B, providing indices of disease activity, progression and fibrosis regression after treatment. Convenience, lower cost, and ease of repeated measurements over time favor widespread acceptance of these tools in clinical practice.

    • William Carey, MD
    • Division of Gastroenterology and Hepatology Cleveland Clinic

    Disclosures: Carey reports no relevant financial disclosures.