INTRODUCTION

Several biochemical indices are being proposed for the noninvasive evaluation of liver fibrosis. The aim comes from the need for a timed course follow-up evaluation of fibrotic tissue deposition in the liver to evaluate progression/regression of chronic liver disease and eventual benefits of therapy. Liver biopsy is limited as a diagnostic procedure not only by the ongoing debate over the usefulness of the impact of histological findings on decision making about treatment (^1,2,3), but also because of major complications in 0.5% of patients and the need for hospitalization. Moreover, ethical and scientific limitations of performing repeated liver biopsies for the follow-up of the disease have promoted the search for surrogate markers to distinguish cirrhosis or severe fibrosis from milder stages of the disease.

Many investigators have attempted to identify several laboratory indices of noninvasive evaluation of liver fibrosis by using biochemical and clinical parameters, individual or in different combinations (^{4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26}). Poynard and colleagues have focused on five biochemical markers (haptoglobin, apolipoprotein A1, -2 macroglobulin, and total bilirubin, gamma glutamyl transpeptidase) combined with patient's age and gender (¹⁵); these indices correlate well with Metavir fibrosis stage, distinguishing between stage 0–1 versus 2–4, with a positive predictive value exceeding 90% (index 0.6) and a negative predictive value of 100% (index 0.1). Other studies have focused on specific markers of liver fibrosis, as hyaluronic acid and YKL-40, both well correlated with the stage of liver fibrosis (^{22,23,24,25,26}). The disadvantage of the indices reported above is that the component assays are not routinely utilized and some scores can be obtained through commercial sources only. The ability of routine biochemical tests to predict fibrosis at histology has been researched in numerous studies, whose findings are summarized in Table 1.

Table 1 - Sensitivity, Specificity, Positive (PPV), and Negative Predictive Value (NPV) of Evaluated Parameters in Detecting Significant (Bridging Fibrosis and Cirrhosis) Fibrosis or Cirrhosis.

Full table

The reported figures show a consistent variation from one study to the other; for instance, sensitivity for the AST/ALT ratio in detecting cirrhosis was reported to vary from 39% to 77.8%. Failure to replicate initial successful studies occur because of the included number of patients, in particular power limitations resulting from low sample size (more than 90% of the studies included fewer than 200 patients), different cut-offs used for significant fibrosis, and differences in prevalence (pretest probability) of patients with significant fibrosis, which may affect the clinical value (posttest probability) of proposed indices. More importantly, the suggested indices have received limited or no external validation from centers different from the original. Therefore, although an individual index may be valid for a particular study patient population, the results may not be generalized to all patients having the diagnostic test.

The aim of this study was to compare each one of the proposed, noninvasive, routinely performed, biochemical tests for noninvasive evaluation of liver fibrosis, by using a large sample of patients, all affected by chronic hepatitis C.

PATIENTS AND METHODS

Patients

We reviewed charts of all consecutive patients with HCV infection evaluated at our Liver Unit between 1995 and 2003. All patients were at their first evaluation after the incidental discovery of positive markers of HCV infection in the blood. Chronic liver disease, secondary to HCV infection, was diagnosed on the basis of the following: elevated aminotransferase levels for >6 months (measured on at least two separate occasions), detectable levels of HCV RNA, and compatible hepatic histology. Other causes of liver disease had been ruled out with appropriate serological and instrumental investigations, and patients affected by those were excluded from the study. None of the patients had received or had been treated with antiviral therapy at the time of the biopsy or blood testing. Alcohol consumption of each patient was recorded, and patients with ingestion of 50 g of alcohol per day for more than 5 yr preceding evaluation were considered as alcoholics. Demographics and laboratory data recorded from each patient at the time of liver biopsy were as follows: body mass index, gender, age, fasting glucose, ALT, AST, gamma glutamyl transpeptidase, alkaline phosphatase, bilirubin, cholesterol, albumin, serum globulins, leucocyte, platelet counts, and prothrombin time (measured as a percentage of the daily internal control).

Histological Staging

All pathological specimens were reviewed by an experienced liver pathologist. The liver sample length and the number of portal areas were coded, and a minimum of five portal tracts in the specimen were required for the appropriate assessment of histological results. The degree of fibrosis was staged according to Scheuer et al. (²⁷), as follow: 0 = no fibrosis; 1 = mild fibrosis (portal fibrosis without fibrous septum formation); 2 = moderate fibrosis (fibrous septa extending into lobules, but not reaching terminal hepatic (venules and other portal tracts); 3 = severe fibrosis (fibrous septa extending to adjacent portal tracts and terminal hepatic venules); and 4 = cirrhosis.

Statistical Analysis

The diagnostic performance of the several proposed indices for the assessment of either significant fibrosis (S_3–4 vs S_0–2) or cirrhosis (S₄ vs S_0–3) was assessed by calculating sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV), accuracy, and positive and negative likelihood ratios (LR+ and LR-, respectively) of each index (^28,29). The considered indices were platelet count < 140.000/l; AST/ALT ratio of 1 alone or in combination with platelet count < 140.000/l and/or globulin albumin ratio > 1; globulin/albumin ratio > 1 alone or in combination with platelet count < 140.000/l; AST/ALT 1 and platelet count < 150.000/l; AST to platelet ratio index (APRI) > 1.5, and finally Forns' score higher than 6.9 points.

RESULTS

A total of 1,502 patients affected by chronic hepatitis C were reviewed. Two hundred and fifty patients were excluded; 184 patients had evidence of coinfection with hepatitis B virus (HbsAg-), and 66 patients had serological markers of auto-immune liver disease. The study group consisted of the remaining 1,252 patients, whose baseline characteristics are reported in Table 2. There were 710 (56.7%) men and 542 (43.3%) women. One hundred and thirty-eight patients were alcoholics (11%). At histology, 164 patients (13.1%) were classified as stage 0; 361 patients as stage 1 (28.8%); and 484 patients as stage 2 (38.6%). The overall prevalence of significant fibrosis was 19.4%, 165 patients being scored in stage 3 (13.2%), and 78 patients (6.2%) were cirrhotics.

Table 2 - Demographic, Biochemical, and Histological Characteristics of 1,252 Patients with Chronic Hepatitis C.

Full table

Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and positive (LR+) and negative (LR-) likelihood ratios of all noninvasive indices are reported in Table 3 for the detection of significant fibrosis (S₃ and S₄), and in Table 4 for the detection of cirrhosis (S₄).

Table 3 - Sensitivity, Specificity, Positive (PPV), and Negative Predictive Value (NPV), Accuracy, and Likelihood Ratios (LR) of Evaluated Parameters in Detecting Significant (S₃ and S₄) Fibrosis.

Full table

Table 4 - Sensitivity, Specificity, Positive (PPV), and Negative Predictive Value (NPV), Accuracy, and Likelihood Ratios (LR) of Evaluated Parameters in Detecting Cirrhosis.

Full table

In the 243 patients with documented significant fibrosis at histology (19.4% of included patients), all indices had specificity rates exceeding the 82% value for the exclusion of S₃ or S₄, whereas the sensitivity for ruling in these fibrotic stages was acceptable for only two indices: platelet count < 140.000/l (specificity rate of 70.6%), and the Forns' score > 6.9 (specificity rate of 79.3%). At exception of the AST/ALT ratio, all the other indices have excellent performances in excluding significant fibrosis (NPV > 90%), whereas a similar excellent performance in identifying the condition was provided only by the combined assessment of platelet count < 140.000/l and globulin/albumin ratio > 1 (91% PPV).

The performance of the considered indices in identifying liver cirrhosis is given in Table 4. In the 78 cirrhotic patients, specificity rates were optimal with values higher than 85% observed for all tests, while sensitivity was excellent for platelet count < 140.000/l. All tests performed well in excluding cirrhosis in an individual patient (NPV higher than 90%), whereas for ruling in cirrhosis all tests had an unacceptable low performance (PPV less than 50% at best).

Since a low platelet count, either alone or in combination with some other biochemical parameters, appears to enter in all the best performing tests for either significant fibrosis or cirrhosis, the distribution of their values within the different fibrosis classes is shown in Figure 1; a mean value of 201.8 44.9 was found in patients with no fibrosis, with values decreasing with the progression of the liver fibrosis. The decline was mild until the fibrotic deposition was moderate (stage 2), and was abrupt in more severe stages. Patients in stages from 0 to 2 of fibrosis had platelet count less than 140.000/l respectively in 4.2%, 3.3%, and 3.7% of cases, whereas in stages 3 and 4 platelet count was higher than 140.000/l, in 22.4% and 7.6% of cases, respectively.

Figure 1.

Comparison of platelet count with the stage of fibrosis.

Full figure and legend (22K)

DISCUSSION

In patients with liver diseases, the assessment of fibrosis is currently considered of paramount relevance for prognosis, implementation of follow-up, and need of therapy (^{27,30,31,32,33,34,35,36,37}). How to obtain this information, whether by invasive or noninvasive means, is currently debated. Our work emphasizes that some of the information provided by histology may also be obtained by several of the proposed indices for the noninvasive assessment of liver fibrosis. From a clinical point of view, it might be more important to determine whether or not significant fibrosis is present, rather than to detail the exact amount of liver fibrosis as patients with severe fibrosis have more adverse prognoses, need to be followed up more closely, and are the ideal candidates for antiviral therapies, whereas those without significant fibrosis may be handled with less stringency.

Ruling out either significant fibrosis or cirrhosis can be easily obtained by evaluating almost all proposed indices. As 1,009 of 1,252 patients showed up at histology with null or nonsignificant fibrosis, a liver biopsy could have been spared in these patients by a careful evaluation of these noninvasive tests. Among the several considered indices, our preference is addressed to platelet counts, as they are routinely available and do not require a mathematical computation. Ruling in the condition of significant fibrosis is more problematic with currently available tests. While the PPV were acceptable for some indices with rates > 82% for ascertaining significant fibrosis, all of them performed poorly when cirrhosis had to be diagnosed. The highest positive likelihood ratios were found for platelets <140.000/l alone or in combination with globulin/albumin ratio >1 or AST/ALT ratio 1. Once again, we would suggest considering platelet counts, as the performance of the index was comparable to the other, more cumbersome estimations. The individual distribution of platelet counts among the different stages of fibrosis (Fig. 1) revealed either low values in patients with null or nonsignificant fibrosis, either high counts in those with severe fibrosis or even cirrhosis. The first situation, documented in 37 of 1,009 patients (3.7%), could be secondary to the clearance of immune complexes or increased platelet sequestration driven from an HCV-induced antigen-antibody immune response (^38,39). High values in spite of advanced fibrosis were found especially in stages 2 and 3, and might be due to two mutually-exclusive biases: the variation of histology (sampling error/intraobserver variations) or the inherent deficiency of the index by itself.

It should be considered that the performance of the selected indices has been checked in our patients with a 19.4% pretest probability of significant fibrosis. This low value is of no surprise since all included patients were naïve to both liver biopsy and antiviral treatment. It is commonly understood that both positive and NPV vary considerably with changes in the pretest probability of the target condition, and this might explain some results at variance from previous reports (^{4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26}). However, likelihood ratios do not vary with changes in prevalue of a disorder, and can be used to derive a posttest probability. With the 19.4% pretest probability of significant fibrosis, and positive likelihood ratios 5, 5.2, and 5.7, respectively, for low platelet count, APRI >1.5, and Forns' score >6.9, a HCV infected outpatient presenting for the first time at clinical evaluation will have a 60% posttest probability of showing significant fibrosis at histology by the noninvasive evaluation of these indices.

Similarly, with a very low prevalence of cirrhosis (6.2%) in our population of 1,252 HCV carriers, and positive likelihood ratio of 8–9, he will have only a 25% posttest probability of liver cirrhosis at histology, by relying uniquely on noninvasive indices. In order to improve these figures, one might either select a group of patients with a higher pretest probability of cirrhosis or develop new noninvasive tests. Indeed, where cirrhotics represent not less than 50% of total cases, similar to patients usually admitted to a liver unit ward, the posttest probability of liver cirrhosis can be as high as 90% by using currently available noninvasive tests. Alternatively, when the prevalence of patients with significant fibrosis cannot be expected to increase, as in the outpatient clinic, where the naïve patients with chronic HCV infection seek evaluation for the first time, we need to develop future indices that will likely result in better likelihood ratios.

In conclusion, more should be done in ameliorating the performance of noninvasive tests for ruling it severe liver fibrosis by adding new parameters highly specific for liver fibrosis. This will hopefully improve the positive predictive value of actual indices that, up to now, have shown to be reliable only in excluding severe fibrosis or cirrhosis. Among the currently developed indices, serum platelets counts qualify as the simplest test to be evaluated.

References

1. Andriulli, A, Festa, V, Leandro, G, et al. AIGO members. Usefulness of a liver biopsy in the evaluation of patients with elevated ALT values and serological markers of hepatitis viral infection. Dig Dis Sci 2001;46: 1409–1415.
2. Heintges, T, Mohr, L, Hensel, F, et al. Value of liver biopsy prior to interferon therapy for chronic viral hepatitis. Dig Dis Sci 1998;43: 1562–1565.
3. Dienstag, JL. The role of liver biopsy in chronic hepatitis C. Hepatology 2002;36: S152–S160.
4. Sheth, SG, Flamm, SL, Gordon, FD, et al. AST/ALT ratio predicts cirrhosis in patients with chronic hepatitis C virus infection. Am J Gastroenterol 1998;93: 44–48. | Article | PubMed | ChemPort |
5. Pohl, A, Behling, C, Oliver, D, et al. Serum aminotransferase levels and platelet counts as predictors of degree of fibrosis in chronic hepatitis C virus infection. Am J Gastroenterol 2001;96: 3142–3146. | Article | PubMed | ChemPort |
6. Reedy, DW, Loo, AT, Levine, RA. AST/ALT is not diagnostic of cirrhosis in patients with chronic hepatitis C. Dig Dis Sci 1998;43: 2156–2159.
7. Forns, X, Ampurdanès, S, Llovet, JM, et al. Identification of chronic hepatitis C patients without hepatic fibrosis by a simple predictive model. Hepatology 2002;36: 986–992. | PubMed |
8. Pradat, P, Alberti, A, Poynard, T, et al. Predictive value of ALT levels for histologic findings in chronic hepatitis C: A European collaborative study. Hepatology 2002;36: 973–977. | PubMed |
9. Saadeh, S, Cammell, G, Carey, WD, et al. The role of liver biopsy in chronic hepatitis C. Hepatology 2001;33: 196–200. | Article | PubMed |
10. Oberti, F, Valsesia, E, Pilette, C, et al. Noninvasive diagnosis of hepatic fibrosis or cirrhosis. Gastroenterology 1997;113: 1609–1616.
11. Wai, CT, Greenson, JK, Fontana, RJ, et al. A simple non-invasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003;38: 518–526. | Article | PubMed |
12. Luo, JC, Hwang, SJ, Chang, FY, et al. Simple blood tests can predict compensated liver cirrhosis in patients with chronic hepatitis C. Hepatogastroenterology 2002;49: 478–481.
13. Giannini, E, Risso, D, Botta, F, et al. Validity and clinical utility of the aspartate aminotransferase-alanine aminotransferase ratio in assessing disease severity and prognosis in patients with hepatitis C virus related chronic liver disease. Arch Intern Med 2003;163: 218–228.
14. Park, GJH, Lin, BPC, Ngu, MC, et al. Aspartate aminotransferase ratio in chronic hepatitis C infection: Is it a useful predictor of cirrhosis? J Gastroenterol Hepatol 2000;15: 386–390. | Article | PubMed | ChemPort |
15. Bismuth, FI, Ratziu, V, Pieroni, L, et al. Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: A prospective study. Lancet 2001;357: 1069–1075. | Article | PubMed | ISI | ChemPort |
16. Kaul, V, Friedenberg, FK, Braitman, LE, et al. Development and validation of a model to diagnose cirrhosis in patients with hepatitis C. Am J Gastroenterol 2002;97: 2623–2628.
17. Anderson, FH, Zeng, L, Rock, NR, et al. An assessment of the clinical utility of serum ALT and AST in chronic hepatitis C. Hepatol Res 2000;18: 63–71.
18. Imperiale, TF, Said, AT, Cummings, OW, et al. Need for validation of clinical decision aids: Use of the AST/ALT ratio in predicting cirrhosis in chronic hepatitis C. Am J Gastroenterol 2000;95: 2328–2332.
19. Williams, AL, Hoofnagle, JH. Ratio of aspartate to alanine aminotransferase in chronic hepatitis. Relationship to cirrhosis. Gastroenterology 1988;95: 734–739.
20. Assy, N, Minuk, GY. Serum aspartate but not alanine aminotransferase levels help to predict the histological features of chronic hepatitis C viral infections in adults. Am J Gastroenterol 2000;95: 1545–1550.
21. Fortunato, G, Castaldo, G, Oriani, G, et al. Multivariate discriminant function based on six biochemical markers in blood can predict the cirrhotic evolution of chronic hepatitis. Clin Chem 2001;47: 1696–1700.
22. Johansen, JS, Christoffersen, P, Moller, S, et al. Serum YKL-40 is increased in patients with hepatic fibrosis. J Hepatol 2000;32: 911–920.
23. Kamal, SM, Turner, B, Koziel, MJ, et al. YKL-40 and PIIINP correlate with the progression of fibrosis in chronic hepatitis C. Gastroenterology 2001;120: 1895A.
24. McHutchinson, 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;15945–15951.
25. Murawaki, Y, Koda, M, Okamoto, K, et al. Diagnostic value of serum type IV collagen test in comparison with platelet count for predicting the fibrotic stage in patients with chronic hepatitis C. J Gastroenterol Hepatol 2001;16: 777–781.
26. Verbaan, H, Bondeson, L, Eriksson, S. Non invasive assessment of inflammatory activity and fibrosis (grade and stage) in chronic hepatitis C infection. Scand J Gastroenterol 1997;32: 494–499.
27. Perrillo, RP. The role of liver biopsy in hepatitis C. Hepatology 1997;26: 57S–61S.
28. Black, ER, Panzer, RJ, Mayewski, RJ, et al. Characteristics of diagnostic tests and principles for their use in quantitative decision making. In: Black ER, Bordley DR, Tape TG, Panzer RJ, eds. Diagnostic strategies for common medical problems. 2nd Ed. American College of Physicians: Philadelphia, PA 1999;1–17.
29. Suchman, AL, Dolan, JG, Odds and likelihood ratios. In: Black ER, Bordley DR, Tape TG, Panzer RJ, eds. Diagnostic strategies for common medical problems. 2nd Ed. American College of Physicians: Philadelphia, PA 1999;31–36.
30. Jouet, P, Roudot-Thovaral, F, Dhumeaux, D, et al. Comparative efficacy of interferon alfa in cirrhotic and non cirrhotic patients with non A, non B, C hepatitis. Le Groupe Francaise puor l'etude du traitement des hepatites chroniques. NANB/C. Gastroenterology 1994;106: 686–690.
31. Van Leeuwen, DJ, Wilson, L, Crowe, DR. Liver biopsy in the mid-1990s: Questions and answers. Semin Liver Dis 1995;15: 340–359.
32. Afdhal, NH. Diagnosing fibrosis in hepatitis C: Is the pendulum swinging from biopsy to blood tests? Hepatology 2003;37: 972–974. | Article | PubMed |
33. Poynard, T, Ratziu, V, Charlotte, F, et al. Rates and risk factors of liver fibrosis progression in patients with chronic hepatitis C. J Hepatol 2001;34: 730–739. | Article | PubMed | ISI | ChemPort |
34. Shiratori, F, Imazeki, F, Moriyama, M. Histologic improvement of fibrosis in patients with hepatitis C who have sustained response to interferon therapy. Ann Intern Med 2000;132: 517–524. | PubMed | ISI | ChemPort |
35. Caballero, T, Perez-Milena, A, Masseroli, M, et al. Liver fibrosis assessment with semiquantitative indexes and image analysis quantification in sustained-responder and non responder interferon-treated patients with chronic hepatitis C. J Hepatol 2001;34: 740–747.
36. Masseroli, M, Caballero, T, O'Valle, F, et al. Automatic quantification of liver fibrosis: Design and validation of a new image analysis method: Comparison with semi-quantitative indexes of fibrosis. J Hepatol 2000;32: 453–464.
37. Duchatelle, V, Marcellin, P, Giostra, E, et al. Changes in liver fibrosis at the end of alpha interferon therapy and 6 to 18 months later in patients with chronic hepatitis C: Quantitative assessment by a morphometric method. J Hepatol 1998;29: 20–28.
38. Panzer, S, Seel, E. Is there an increased frequency of autoimmune thrombocytopenia in hepatitis C infection? Wien Med Wochenschr 2003;153: 417–420.
39. De Almeida, AJ, Campos-De-Magalhaes, M, De Melo Marcal, OP, et al. Hepatitis C virus-associated thrombocytopenia: A controlled prospective, virological study. Ann Hematol 2004 Feb 13 (Epub ahead of print).