Noninvasive Tests of Fibrosis: To Get the Right Answers, We Need to Ask the Right Questions

We read with interest the work by yimaz et al. entitled ‘noninvasive assessment of liver fibrosis with the aspartate transaminase to platelet ratio index (apri): usefulness in patients with chronic liver disease’ in th journal of hepatity monthly (1). this study adds to the growing body of literature showing the utility of the apri as a noninvasive method to assess liver fibrosis. this retrospective study included 455 patients with different liver diseases: 207 patients with chronic hepatitis b (chb), 108 with chronic hepatitis c (chc), and 140 patients with nonalcoholic fatty liver disease (nafld). apri scores were calculated and compared to liver biopsy findings in all patients. roc curves were created to assess the accuracy of the apri as a predictor of the absence or presence of fibrosis (f0 vs. f1-4). the authors concluded that apri was significantly associated with fibrosis scores in patients with chc and nafld, but not in those with chb. this is a well-designed study with adequate sample size that addresses an important clinical question. we agree with the authors that the apri is a good tool for the assessment of liver fibrosis, with some notable caveats. the authors chose to use the roc curves to assess the utility of the apri as a predictor of absence or presence of fibrosis. alternatively, the roc curves could have been created with various apri cutoffs to predict the absence or presence of cirrhosis. using cirrhosis rather than any fibrosis may be more clinically relevant because the presence of cirrhosis determines the urgency of treatment, serves as a prognostic marker for a therapeutic outcomes and establishes the need to monitor of complications. perhaps more important is that, the performance of apri may actually be much better for determining if cirrhosis has an effect than for dentifying the presence of any fibrosis. the two variables that constitute apri are ast and platelet count, both of witch are markers of advanced fibrosis and cirrhosis (2, 3). ast is a mitochondrial enzyme that increases disproportionately more than alt and thus increases the ast/alt ratio in cirrhosis with progressive liver injury (4). furthermore, the platelet count declines as liver disease progresses due to splenic sequestration from portal hypertension and eventually from thrombopoietin deficiency with very impaired liver synthetic function. both of these parameters may be entirely normal in patients with early-stage fibrosis, and therefore it is not surprising that apri was useful but far from foolproof in excluding fibrosis.