Validation of Hepascore, Compared With Simple Indices of Fibrosis, in Patients With Chronic Hepatitis C Virus Infection in United States

Transcription

1 CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2009;7: Validation of Hepascore, Compared With Simple Indices of Fibrosis, in Patients With Chronic Hepatitis C Virus Infection in United States LAREN BECKER,* WAEL SALAMEH, ANTHONY SFERRUZZA, KE ZHANG, RONG CHEN, RAZA MALIK,* RICHARD REITZ, IMAD NASSER,* and NEZAM H. AFDHAL* *Beth Israel Deaconess Medical Center, Boston, Massachusetts; and Quest Diagnostics, Nichols Institute, San Juan Capistrano, California Background & Aims:: Biomarkers are being developed as alternatives to liver biopsy for predicting liver fibrosis in patients with chronic hepatitis C. Hepascore uses noninvasive serum markers and has been validated in Australian and European populations for predicting different degrees of fibrosis. This study validated this test in a U.S. population. Methods:: Patients with chronic hepatitis C virus infection were assigned to training (n 203) or validation (n 188) sets. Liver fibrosis was staged according to the META- VIR scoring system. The Hepascore algorithm uses data on age, sex, as well as total bilirubin, -glutamyl transferase, 2-macroglobulin, and hyaluronic acid levels. Results:: The ability of Hepascore to predict significant fibrosis (F2 4) as determined by the area under the receiver operating curve was similar in training (0.83) and validation sets (0.81) and was comparable to results seen in previous studies. A cutoff score of >0.55 was best for predicting significant fibrosis, with a sensitivity and specificity of 82% and 65% and positive and negative predictive values of 70% and 78%. When compared with 2 simple indices, FIB-4 (age, platelets, AST, and ALT) and APRI (AST/platelet ratio index), Hepascore performed better at excluding advanced fibrosis by using a low cutoff score but worse at predicting fibrosis by using a high cutoff score. An algorithm with Hepascore followed by FIB-4 or APRI spared 103 of 391 individuals a liver biopsy and missed advanced fibrosis in only 1 patient. Conclusions:: Hepascore accurately predicted likelihood of developing fibrosis and could alleviate the need for liver biopsy in a subset of patients. Chronic hepatitis C has an estimated worldwide prevalence of 3%. 1 Management of the disease is dependent on the extent of liver fibrosis. Currently liver biopsy is the gold standard for determining the degree of fibrosis. Unfortunately, liver biopsy has many drawbacks including complications such as pain, bleeding and rarely death, cost, interobserver and intraobserver variability in histologic scoring, and sampling error. 2 7 Furthermore, it offers only a static view of a dynamic disease, necessitating multiple biopsies to determine reversal or progression of fibrosis. These factors have sparked intense interest in developing noninvasive alternatives. Significant progress has been made in identifying both nonspecific and specific biomarkers of fibrosis. Nonspecific markers include age, gender, laboratory markers of liver injury or dysfunction including ALT, AST, -glutamyl transferase (GGT), bilirubin, haptoglobin, platelet count, prothrombin time, and metabolic markers including cholesterol, apoprotein A1, and 2-macroglobulin (A2M). To date, the AST-platelet ratio index (APRI) is the simplest test with nonspecific markers that has documented value in predicting fibrosis. 8,9 Other tests that use combinations of nonspecific markers include the Fibrotest (A2M, haptoglobin, GGT, bilirubin, and apoprotein A1), 10 Forns (age, cholesterol, platelet count, and GGT), 11 and FIB-4 (age, platelet count, AST, and ALT). 12,13 There has also been some success in identifying biomarkers that are specific for fibrosis, including hyaluronic acid (HA), matrix metalloproteinase-2, tissue inhibitor of matrix metalloproteinase-1, and amino-terminal peptide of type III procollagen These fibrosis markers have improved value when used in combination, as is the case for the European liver fibrosis score. 16 More recently, attempts have been made to improve nonspecific marker indices by including fibrosis-specific markers. The Fibrometer, which includes HA with prothrombin time, platelet count, AST, A2M, urea, and age, has been shown to have some efficacy Adams et al 20 proposed the Hepascore, which combines HA, total bilirubin, GGT, A2M, age, and sex. Hepascore was validated in an Australian HCV population and subsequently by 2 other groups in European populations (France) and was found to have an area under the receiver operating curve (AUROC) of between 0.76 and 0.9 for predicting different degrees of fibrosis The aim of the present study was to validate the Hepascore in a U.S. population with chronic HCV infection and to compare it with nonspecific marker indices, APRI and FIB-4. Materials and Methods Determination of Normal Hepascore Reference Range in Healthy Volunteers Serum was collected and analyzed from 214 healthy volunteer donors, including 106 women and 108 men years of age. The inclusion criteria were that individuals were apparently healthy, ambulatory, community-dwelling, and nonmedicated adults. Individuals with a history of liver disease, any other major illness, or abnormal GGT, total bilirubin, A2M, and/or HA levels were excluded. Abbreviations used in this paper: APRI, AST/platelet ratio index; AUROC, area under the receiver operating curve; A2M, 2-macroglobulin; CI, confidence interval; FIB-4, age, platelets, AST, and ALT; GGT, -glutamyltransferase; HA, hyaluronic acid; NPV, negative predictive value; PPV, positive predictive value; SD, standard deviation by the AGA Institute /09/$36.00 doi: /j.cgh

2 June 2009 VALIDATION OF HEPASCORE 697 Patients Three hundred ninety-one patients with chronic HCV infection undergoing liver biopsy were enrolled from the Liver Center at Beth Israel Deaconess Medical Center, a tertiary referral center in Boston, MA. All patients had chronic HCV infection as confirmed by HCV-RNA polymerase chain reaction analysis in serum, and none were on active antiviral treatment for HCV at time of biopsy. Coexisting liver diseases attributable to alcohol, hepatitis B, autoimmune hepatitis, primary biliary cirrhosis, hemochromatosis, 1-antitrypsin deficiency, or Wilson s disease were reasonably excluded by history and standard clinical, laboratory, imaging, and histologic studies. Human immunodeficiency virus coinfection and post-transplant patients were also excluded. The study was approved by the institutional review board at Beth Israel Deaconess Medical Center. All patients gave informed consent. Liver biopsies were performed as part of routine clinical care in the staging and grading of liver disease. Age and sex were recorded at the time of liver biopsy. The subjects were split into training and validation sets on the basis of time of sample collection. The demographic and biochemical characteristics of the training (n 203) and validation set (n 188) are outlined in Table 1. Laboratory Testing Serum was collected either at the time of liver biopsy or within 3 months of the biopsy. Standard laboratory assessments including complete blood cell count with platelets, serum chemistry panels, ALT, AST, GGT, and total bilirubin were performed by a licensed clinical laboratory on fresh serum within 36 hours of collection. Other analyses were performed on sera stored frozen at 80 o C. Total bilirubin and GGT were measured on an AU640 e instrument (Olympus Diagnostica, Center Valley, PA) by using the GGT reagent pack and the total bilirubin reagent pack (Olympus Diagnostica System), respectively. A2M was measured on a BNII nephelometer (Dade Behring Inc, Marburg, Germany). HA was measured with an enzyme-linked protein-binding assay (Corgenix Inc, Denver, CO). All analyses were performed at Quest Diagnostics Nichols Institute, San Juan Capistrano, CA. Liver Biopsies Liver biopsy samples were obtained under ultrasound guidance or marking with either a 16-gauge TruCut or 18- gauge Menghini needle. All biopsies were deemed adequate on the basis of either specimen size ( 10 mm) or the number of portal tracts ( 8). A single expert pathologist blinded to all clinical and serologic results evaluated all slides. Biopsies were interpreted according to the scoring schema developed by the METAVIR group. 21 Fibrosis was scored on a 5-point scale: F0, no fibrosis; F1, portal fibrosis alone; F2, portal fibrosis with rare septae; F3, portal fibrosis with many septae; F4, cirrhosis. The presence of stage F2, F3, or F4 was termed significant fibrosis, whereas F3 or F4 was considered advanced fibrosis. Table 1. Clinical and Laboratory Features of Training and Validation Cohorts Present cohort Australian cohort Training (n 203) Validation (n 188) P value Training (n 117) Validation (n 104) Mean (SD) age, y 50 (9) 50 (8).4 40 (9) 41 (9) Female, n (%) 61 (30) 56 (30) 1 38 (32) 28 (27) Genotype, a n (%) (75) 142 (75) 67 (61) 50 (48) 2 19 (9) 13 (7) 45 (39) 48 (47) 3 9 (4) 10 (5) 4 4 (2) 4 (2) (5) 6 4 (2) 0 Mean (SD) ALT, b U/L 81 (74) 86 (86) (90) 131 (99) Mean (SD) bilirubin, mol/l 11 (7) 12 (7).3 12 (8) 12 (5) Mean (SD) albumin, g/l 44 (3.1) 44 (2.9).3 42 (4) 42 (3) Stage, c n (%).7 F0 30 (15) 32 (17) 23 (19) 17 (16) F1 75 (37) 58 (31) 43 (37) 28 (27) F2 28 (14) 29 (15) 29 (25) 35 (34) F3 30 (15) 34 (18) 15 (13) 7 (7) F4 40 (20) 35 (19) 7 (6) 17 (16) Significant fibrosis (F2 F4) 98 (48) 98 (52).4 51 (44) 59 (57) Advanced fibrosis (F3, F4) 70 (34) 69 (37).6 22 (19) 24 (23) Necroinflammatory activity, n (%).02 Grade A0 5 (2) 1 (1) 20 (17) 8 (8) Grade A1 97 (48) 69 (37) 75 (64) 63 (60) Grade A2 85 (42) 105 (56) 13 (11) 32 (31) Grade A3 16 (7) 13 (7) 9 (8) 1 (1) Significant activity (A2, A3) 101 (50%) 118 (63%) (19%) 33 (32%) a Genotype not available for 14 patients in the training set and 19 patients in the validation set. b ALT might be measured in different units between Quest data and Australia data. c Histology scored according to METVAIR.

3 698 BECKER ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 6 Statistical Analysis The Hepascore model was computed from the results obtained for the following 4 biochemical markers: total bilirubin, GGT, A2M, and HA along with age and sex. The Hepascore value for each patient was calculated by using the model first published by Adams et al. 20 The Hepascore is derived by using 2 equations: I. Logistic regression model: y exp [ ( X age) ( X sex) ( X A2M) ( X HA) ( X bilirubin) ( X GGT)] With age provided in years, male sex 1, female sex 0, A2M in g/l, HA in g/l, bilirubin in mol/l, and GGT in U/L. II. Hepascore calculation: y 1 y Clinical and demographic characteristics of the training and validation sets were compared by the Student t test for continuous variables and 2 tests or Fisher exact tests for categorical variables. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were evaluated to predict significant fibrosis (F2 4) for various Hepascore points from 0 1, with an increment of 0.05 for the training set. The 95% confidence intervals (CIs) were calculated for the sensitivity and specificity by using binomial distribution. The performance at the best cutoff was then evaluated for all subjects and subjects with more than 1 elevated analyte in the training set, validation set, and entire population (combined training and validation sets) and compared with the Australian group data. Training and validation sets were combined to build a reference patient population to estimate the probabilistic likelihood for each fibrosis stage at any Hepascore value of x. For a significant fibrosis-positive patient with a Hepascore of x, a patient population with Hepascore equal to or greater than that value was retrieved from our data set. The percentages of subjects at each fibrosis stage were calculated to estimate the probability of each fibrosis stage within this subset of the patient population. Similarly, for a significant fibrosis-negative patient with a Hepascore of x, a patient population with Hepascore equal to or lower than this value was retrieved to estimate the percentages of each fibrosis stage. This is used for clinical reporting purposes, because sensitivity, specificity, PPV, and NPV vary across the range of Hepascore values and are more informative for clinical decision making than the values calculated for the negative/positive cutoff of With this approach, the PPV for significant fibrosis at any given Hepascore value percentage (F2) percentage (F3) percentage (F4), whereas the NPV percentage (F0 1). FIB-4 is derived from the markers age, AST, ALT, and platelets by using the following formula 12,13 : Age (years) X AST [U/L]/(platelets [10 9 /L] X (ALT [U/L]) 1/2 ). APRI is derived from the markers AST and platelets by using the formula 8,9 : (AST [U/L]/platelets [10 9 /L]) X 100. The sensitivity, specificity, PPV, and NPV were evaluated to predict advanced fibrosis (F3 4) for cutoffs 1.45 and 3.25 for FIB-4 and 0.5 and 1.5 for APRI. All statistical analyses were performed with SAS 9.1 (SAS, Cary, NC). Results Reference Range Study A reference range for a negative Hepascore was determined from the 214 normal volunteers described above. There was both age and sex dependence. The data were Gaussian with logarithmic transformation. With this transformation, the mean 2 standard deviations (SDs) for the normal male and female population single reference range were The upper boundary of the mean 2 SDs in the male population was slightly higher at 0.48, with a CI of Patient Characteristics The demographic and biochemical characteristics of the training (n 203) and validation set (n 188) did not differ significantly (Table 1), except that the validation set had a higher proportion of patients with necroinflammatory activity of grade A2 (56% vs 42%) and less grade A1 (37% vs 48%). The combined population had an average age of 50 and was predominantly male (70%). The majority (75%) had HCV genotype 1. The median length of liver biopsies was 16 mm; 11.2% had liver biopsy lengths of less than 10 mm, consistent with biopsy lengths seen in clinical practice. Overall, 195 (49.9%) patients had no/mild fibrosis (F0 1), and 139 (35.5%) had advanced fibrosis or cirrhosis (F3 4). In comparison with the Australian cohort, there were several notable clinical differences; our subjects were a decade older, predominantly genotype 1, had more active inflammation, and had more advanced fibrosis, 35.5% vs 21% (Table 1). Test Performance The AUROC for discriminating significant fibrosis (F2 4) was 0.83 (95% CI, ) for the training set and 0.81 (95% CI, ) for the validation set (Figure 1). This was similar to previous studies, which found the equivalent AUROC to range from (Table 2). For the entire population (n 391), the diagnostic performance score for Hepascore, as determined by AUROC, was 0.81 for significant fibrosis, 0.83 for advanced fibrosis (F3 4), and 0.88 for cirrhosis. The AUROC did not differ substantially when individuals with liver biopsy lengths less than 10 mm were excluded (0.811 Figure 1. Hepascore ROC curves for predicting significant fibrosis in patients with chronic hepatitis C: training (n 203) and validation sets (n 188).

4 June 2009 VALIDATION OF HEPASCORE 699 Table 2. Comparison of Hepascore Performance for Significant Fibrosis (F2 4) Across Several Studies Entire patient population Subjects with more than 1 analyte elevation Training set Australian cohort, Adams et al 20 Validation set French cohort, Leroy et al 19 French cohort, Halfon et al 18 Population Cutoff AUROC Sensitivity 82% 88% 92% 89% 54% 77% Specificity 65% 69% 67% 63% 84% 63% PPV 70% 76% N/A N/A 78% 59% NPV 78% 83% N/A N/A 63.5% 80% vs 0.815). We found 0.55 to be the best cutoff for predicting significant fibrosis. This value is higher than that chosen by the Australian group (0.5) but is supported by our reference range studies that indicate the 2.5% upper boundary of the CI can extend to 0.55 in normal male subjects. Using a cutoff of 0.55 in the training set predicted significant fibrosis, with a sensitivity of 79 (95% CI, 69 86), a specificity of 71% (95% CI, 61 79), a PPV of 71%, and an NPV of 78%. For the validation set, this cutoff yielded a sensitivity of 86% (95% CI, 77 92), a specificity of 59% (95% CI, 48 69), a PPV of 69%, and an NPV of 79% (Table 3). Within our total population (n 391), 102 individuals (26%) had only a single elevated index analyte. The majority (65%) had an isolated elevation of A2M, with GGT, HA, and total bilirubin elevations making up 18%, 15%, and 3%, respectively. When these patients were removed and the data reanalyzed, the predictive value of the Hepascore improved. The AUROC for significant fibrosis rose from 0.81 (95% CI, ) to 0.85 (95% CI, ) (Figure 2). There was noticeable improvement of performance measures at a cutoff value of For both the training and validation sets, sensitivity rose to 88% (from 79% and 86%, respectively), and specificity rose to 74% and 63% (from 71% and 59%, respectively). For the entire population, the sensitivity rose from 82% to 88%, the specificity from 65% to 69%, the PPV from 70% to 76%, and the NPV from 78% to 83% (Table 3). We then looked at the probability of fibrosis at particular Hepascore values for either the entire population or subjects with more than 1 elevated analyte. The probabilistic likelihoods for each fibrosis stage for individuals with high or low Hepascore values are shown in Table 4. A Hepascore value 0.2 has a NPV of 98.3% for excluding advanced fibrosis (F3 4), with a sensitivity of 99.2%. Values 0.8 have a PPV of 61.7% for predicting advanced fibrosis, with a specificity of 76.8%. Removing patients with a single elevated analyte did not substantially affect performance for these high and low cutoffs. Overall, 55% of the entire patient population had Hepascore values 0.2 or 0.8, and among these individuals, 71.8% were correctly classified. We compared the use of high and low cutoffs for predicting advanced fibrosis for Hepascore with the simple nonspecific marker indices FIB-4 and APRI. High and low cutoffs of 1.45 and 3.25 have been previously used for FIB-4 12,13 and 0.5 and 1.5 for APRI. 8,9 FIB-4 values 1.45 have an NPV for excluding advanced fibrosis of 81.6%, with a sensitivity of 72.7%. Values 3.25 have a PPV for predicting fibrosis of 91.3% and a specificity of 98.4%. Overall, 65% of our population fell within this range, and 83.4% of these were correctly classified with FIB-4. For APRI, values 0.5 excluded advanced fibrosis, with an NPV of 82.6% and a sensitivity of 77.0%, and values 1.5 predicted it, with a PPV of 84.4% and specificity of 97.2%. Also 58.3% of the population fell within this range, and of these, 83% were correctly classified with APRI. We evaluated how performing FIB-4 and APRI subsequent to Hepascore might improve its value for predicting fibrosis. In individuals with Hepascore values 0.8, a FIB-4 value 3.25 increases the PPV from 61.7% to 90.9%, and an APRI value 1.5 increases it to 82.5%. Discussion Finding a noninvasive means to assess fibrosis and replace liver biopsy has been the holy grail of hepatology. Many indices consisting of both nonspecific and specific markers have been proposed, but none have reached the threshold of sensitivity and specificity needed to become accepted as the new Table 3. Performance Summary After Removal of Individuals With Single Analyte Elevation Our cohort (cutoff 0.55) Training set Validation set Training validation All subjects 1 analyte elevation All subjects 1 analyte elevation All subjects 1 analyte elevation Samples Sensitivity 79% 88% 86% 88% 82% 88% Specificity 71% 74% 59% 63% 65% 69% PPV 71% 78% 69% 75% 70% 76% NPV 78% 85% 79% 80% 78% 83%

5 700 BECKER ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 6 Figure 2. Hepascore ROC curves for predicting significant fibrosis in patients with chronic hepatitis C: entire patient population (n 391) vs subjects with more than 1 elevated analyte (n 289). gold standard. A review comparing many of these indices with the exception of Hepascore in the HCV population found the median AUROCs for predicting significant fibrosis (F2 4) to be 0.82 and 0.78 in training and validation sets, respectively. 22 Depending on the cutoff that was selected, high sensitivities could be achieved but at the expense of lower specificities, or vice versa. Hepascore combines a biomarker of fibrosis (HA) with several nonspecific markers. It has been validated in Australian and European populations with chronic HCV. We have shown that Hepascore maintains a similar predictive value in a U.S. population of chronic HCV patients, with AUROCs for significant fibrosis of 0.81 or 0.85 in the combined data set for all individuals or for only those subjects with more than 1 elevated analyte (Table 2). However, the optimal cutoffs differed between Europe, Australia, and the United States. In our study, we also evaluated a control population, which is important to validate the cutoffs with respect to the normal population. Because indices with fibrosis markers will likely cost more than those derived from routine nonspecific markers, it is important to assess whether they perform better than their simpler counterparts. A recent study found FIB-4 to perform well in predicting advanced fibrosis (F3 4) in patients with HCV. With high and low cutoffs of 1.45 and 3.25, FIB-4 correctly classified 72.8% of liver biopsies in their population of With these cutoffs, FIB-4 correctly classified only 54% of our population. Compared with their study, we found FIB-4 to perform better at the high cutoff of 3.25 (PPV of 91.3% versus 82.1%) and worse at the lower cutoff of 1.45 (NPV of 81.6% versus 94.7%). An explanation for these discrepancies is the difference in advanced fibrosis (F3 4) between the 2 studies, Table 4. Probabilities of Each Fibrosis Stage at High and Low Hepascore Values (n 391) Hepascore Population F0 1 95% 86% 84% 22% 17% F2 5% 12% 9% 14% 12% F3 0 2% 6% 21% 21% F4 0 0% 1% 43% 50% Figure 3. Algorithm for managing patient with chronic HCV infection. with our population having 35.5% compared with 17.2% in theirs. Like FIB-4, we found APRI, the simplest of nonspecific marker indices, to perform better at predicting advanced fibrosis at the high cutoff ( 1.5) than excluding it at the low cutoff ( 0.5). With this range, 48.3% of liver biopsies were correctly classified. When compared with FIB-4 and APRI, Hepascore performed favorably at excluding advanced fibrosis but worse at predicting it. Whereas Hepascore is unlikely to entirely replace liver biopsy as a staging test for liver fibrosis, one can envision an algorithm with Hepascore in the management of chronic HCV patients, particularly in the tertiary center where prevalence of fibrosis is greatest (Figure 3). For individuals with a Hepascore value 0.2, significant fibrosis is unlikely, and continued observation on an annual basis is sufficient. For individuals with a Hepascore 0.8, FIB-4 and/or APRI should be performed. A FIB-4 value 3.25 raises the PPV of Hepascore for predicting advanced fibrosis from 61.7% to 90.9%, and an APRI value 1.5 raises it to 82.5%. For those individuals with elevated FIB-4 or APRI values, initiation of HCV therapy should be strongly considered, along with other screening modalities to evaluate for varices and hepatocellular carcinoma. With this algorithm, 103 of the 391 patients would have been spared a liver biopsy, with advanced fibrosis missed in only 1 patient and incorrectly diagnosed in only 4 patients. For individuals with a Hepascore value that falls between 0.2 and 0.8, 0.55 can be used as a cutoff. Liver biopsy should be recommended for those individuals above 0.55 and for those with high Hepascores but lower FIB-4/APRI scores. For individuals below 0.55, alternative noninvasive testing (ie, transient elastography) should be considered along with repeating Hepascore annually. Future studies should prospectively evaluate multimodalities of fibrosis assessment to determine the optimal clinical strategy. References 1. Global surveillance and control of hepatitis C: Report of a WHO Consultation organized in collaboration with the Viral Hepatitis Prevention Board, Antwerp, Belgium. J Viral Hepat 1999;6: Bedossa P, Dargere D, Paradis V. Sampling variability of liver fibrosis in chronic hepatitis C. Hepatology 2003;38: Bravo AA, Sheth SG, Chopra S. Liver biopsy. N Engl J Med 2001;344: Cadranel JF, Rufat P, Degos F. Practices of liver biopsy in France: results of a prospective nationwide survey for the Group of Epi-

6 June 2009 VALIDATION OF HEPASCORE 701 demiology of the French Association for the Study of the Liver (AFEF). Hepatology 2000;32: Gilmore IT, Burroughs A, Murray-Lyon IM, et al. Indications, methods, and outcomes of percutaneous liver biopsy in England and Wales: an audit by the British Society of Gastroenterology and the Royal College of Physicians of London. Gut 1995;36: Regev A, Berho M, Jeffers LJ, et al. Sampling error and intraobserver variation in liver biopsy in patients with chronic HCV infection. Am J Gastroenterol 2002;97: Rousselet MC, Michalak S, Dupre F, et al. Sources of variability in histological scoring of chronic viral hepatitis. Hepatology 2005;41: Lackner C, Struber G, Liegl B, et al. Comparison and validation of simple noninvasive tests for prediction of fibrosis in chronic hepatitis C. Hepatology 2005;41: Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003;38: Imbert-Bismut F, Ratziu V, Pieroni L, et al. Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study. Lancet 2001;357: Forns X, Ampurdanes S, Llovet JM, et al. Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 2002;36: Sterling RK, Lissen E, Clumeck N, et al. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology 2006;43: Vallet-Pichard A, Mallet V, Nalpas B, et al. FIB-4: an inexpensive and accurate marker of fibrosis in HCV infection. comparison with liver biopsy and fibrotest. Hepatology 2007;46: Kasahara A, Hayashi N, Mochizuki K, et al. Circulating matrix metalloproteinase-2 and tissue inhibitor of metalloproteinase-1 as serum markers of fibrosis in patients with chronic hepatitis C: relationship to interferon response. J Hepatol 1997;26: McHutchison JG, Blatt LM, de Medina M, et al. Measurement of serum hyaluronic acid in patients with chronic hepatitis C and its relationship to liver histology: Consensus Interferon Study Group. J Gastroenterol Hepatol 2000;15: Rosenberg WM, Voelker M, Thiel R, et al. Serum markers detect the presence of liver fibrosis: a cohort study. Gastroenterology 2004;127: Cales P, Oberti F, Michalak S, et al. A novel panel of blood markers to assess the degree of liver fibrosis. Hepatology 2005; 42: Halfon P, Bacq Y, De Muret A, et al. Comparison of test performance profile for blood tests of liver fibrosis in chronic hepatitis C. J Hepatol 2007;46: Leroy V, Hilleret MN, Sturm N, et al. Prospective comparison of six non-invasive scores for the diagnosis of liver fibrosis in chronic hepatitis C. J Hepatol 2007;46: Adams LA, Bulsara M, Rossi E, et al. Hepascore: an accurate validated predictor of liver fibrosis in chronic hepatitis C infection. Clin Chem 2005;51: Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C: the METAVIR Cooperative Study Group. Hepatology 1996;24: Parkes J, Guha IN, Roderick P, et al. Performance of serum marker panels for liver fibrosis in chronic hepatitis C. J Hepatol 2006;44: Reprint requests Address requests for reprints to: Nezam H. Afdhal, MD, Liver Center, 110 Francis Street, Boston, Massachusetts Conflicts of Interest The authors disclose the following: Dr Afdhal is a consultant and has received grant support from Prometheus, Quest Diagnostics, and Echosens. The remaining authors disclose no conflicts. Funding Funded by Quest Diagnostics.