Role of simple biomarkers in predicting fibrosis progression in HCV infection

Hepatic fibrosis biomarkers have been found to be variably accurate in the estimation of fibrosis in chronic Hepatitis C hepatitis (HCV). There is little information whether they are useful longitudinal predictors of changes of fibrosis stage; i.e., progression of fibrosis to cirrhosis. The aim of the study is to examine the accuracy of biomarker panels AST to Platelet Ration Index (APRI ) and FIB-4 in predicting changes in liver histology in patients with chronic HCV.The study is a retrospective cohort study. Data was extracted from a HCV data base that contained clinical and lab information about patients presenting for HCV evaluation to a tertiary academic health-care center (University of Texas Medical Branch at Galveston- UTMB) from 1999-2007. \r\n\r\nThe data base was screened for patients with repeat liver biopsies separated at least by a year. Patients having a sustained viral response (SVR) with treatment, alcohol intake of >15 grams/day ,hepatitis B infection (HBV), metabolic liver disease,, hepatocellular carcinoma (HCC), human inmunodeficieny virus infection (HIV), and stage 4 liver disease on the first biopsy were excluded. All biopsies were read by a single pathologist using Batts Ludwig criteria (F0-F4). Results: A total of 36 patients met the inclusion criteria. The median age at the time of initial biopsies was 47 (Range 25-68). The median interval between biopsies was 4 yrs (IQR 3-5 yrs).The majority of the patients were Stage 1 on initial biopsy (50%) followed by Stage 2( 25%) and Stage 3 (22%).One patient had Stage 0 fibrosis. Out of a total of 36 patients 19 patients (53%) had progression of fibrosis on repeat biopsies, 16 patients (44%) showed no change in stage and 1 patient (3%) showed improvement in fibrosis on repeat biopsy. Among the patient factors examined (Table 3), the group that showed progression had significantly higher ALT and AST values when compared to the group that did not show progression .There was no statistically significant difference between baseline APRI, FIB-4 index , stage at initial biopsy, .age or length of time between biopsies in progressors compared to non progressors. The correlation between change in stage and change in APRI is 0.627 (r) (r² = 0.393377) (p value=0.00001).The correlation between change in stage and change in FIB-4 is 0.56 ( r) ( r² = 0.313431) p=.00004. A change in APRI ( delta APRI ) of 0.18 had 80%Positive Predictive Value( PPV) and 67 %Negative Predictive Value( NPV )in predicting progression of fibrosis . An analysis of ROC curve revealed that change in FIB-4 (delta Fib-4) of 0.39 had 75% PPV AND 75 % NPV for predicting progression of fibrosis.\r\n\r\n\r\n APRI and Fib-4 at the time of second biopsy had excellent predictive value for progression in fibrosis. The Area Under ROC Curve (AUC) of APRI at the time of second biopsy was 0.824 (95 % CI =0.660 to 0.930)(p value=0.0001). APRI >.0.69 had a 79 % PPV and APRI <0.69 had a 78% NPV for progression of fibrosis. FIB-4 index at the time of second biopsy had an AUC of 0.78 (95% CI 0.608 to .899) (p value of 0.0004).FIB-4 of >1.65 had a 81 % PPV and 70% NPV for prediction of fibrosis by any stage. APRI and FIB-4 at the time of repeat biopsy had an ideal predictive value for two stage progression. APRI had an AUC of 1.0 (95% CI 0.85 to 1.00).APRI of greater than 1.94 had a 100% PPV and 100% NPV for prediction of two stage change in fibrosis.FIB-4 had an AUC of 0.911(95% CI 0.717 to 0.986) (p value =0.0001).FIB-4 of greater than 3.01 had a 100% NPV and 62 % PPV for 2 stage change in fibrosis. The limitations of the study are 1)Retrospective study 2) Relatively small sale size\r\n Conclusions: 1) Changes in APRI and FIB-4 between first and second biopsies show significant correlation with change in fibrosis stage. 2) APRI and FIB-4 have good accuracy in predicting fibrosis progression on biopsies 3) Clinicians are justified in using the APRI and FIB-4 to follow hepatic histologic progression to cirrhosis in chronic HCV patients\r\n