Article first published online: 27 FEB 2004
Copyright © 2004 American Association for the Study of Liver Diseases
Volume 39, Issue 3, page 863, March 2004
How to Cite
Lok, A. S. and Wai, C.-T. (2004), Reply:. Hepatology, 39: 863. doi: 10.1002/hep.20145
- Issue published online: 27 FEB 2004
- Article first published online: 27 FEB 2004
We appreciate the interests expressed by Calvez and colleagues on the use of the aspartate aminotransferase to platelet ratio index (APRI) in predicting fibrosis in patients with chronic hepatitis C.1
We welcome external validation of APRI, because the ultimate utility of any noninvasive model for prediction of hepatic fibrosis depends on its practicality and validation by other investigators in a wide range of patients. Several factors may account for the lower accuracy of APRI in the cohort of French patients studied. First, we used an Ishak fibrosis score, not a Metavir score. Second, there may have been differences between the patient populations in the two studies. Our study included consecutive treatment-naïve chronic hepatitis C patients who underwent liver biopsy at our center. Also, 26% of our patients were non-Caucasians. The French study did not specify if any of the patients had prior antiviral therapy, how many patients declined to participate, or the ethnicity of their patients.2 Third, the proportion of patients biopsied who had normal aminotransferases may be different. Patients in our study had higher aspartate aminotransferase than those in the study by Calvez and colleagues (mean: 2.3 vs. 1.7 times the upper limit of normal). At our center, we recognize the risks of liver biopsy and the generally benign course of patients with persistently normal aminotransferases; consequently, liver biopsy is performed only on selected chronic hepatitis C patients who have normal aminotransferases. Fifty-one (19%) patients in our study had normal aminotransferases at the time of biopsy. We are surprised that Calvez and colleagues routinely perform liver biopsies in patients with normal aminotransferases when their group emphasizes the risks of liver biopsies and recommends that liver biopsy should not be mandatory anymore.3
We recognize that the FibroTest has been validated in various patient populations by Poynard and colleagues,2, 3 but a recent study in Australia using the FibroTest score computed from the same biochemical analytes, age, and gender found a lower area under receiver operating characteristic curve (AUROC) of 0.739 and lower positive and negative predictive values4 compared with other studies reported by the French investigators. This highlights the importance of external validation by other investigators in different patient populations with test results from other laboratories.
We do not claim that APRI is the most accurate model for prediction of significant fibrosis or cirrhosis in patients with chronic hepatitis C. In fact, our study showed that APRI was slightly less accurate than other models using complex formulas with more variables.1 The advantages of APRI include the use of a simple formula that is amenable to mental calculation in the clinic, and the use of laboratory test results that are routinely available to every clinician managing chronic hepatitis C patients. Thus, no additional blood collection or costs are needed. We acknowledge that one limitation of APRI is the uncertainty regarding the appropriate definition of the upper limit of normal for aspartate aminotransferase. We agree that this is an important question that needs to be resolved, not only for the application of APRI but also for the evaluation of all forms of liver diseases.
Anna S. Lok*, Chun-Tao Wai*, * University of Michigan Medical Center, Ann Arbor, MI.