Comments on “BIOIDENTICAL HORMONES” - March, 2009 Below is an article that was published in the Wall Street Journal on Mar 16th, 2009. Two of my patients have been kind enough to give me a copy of the article, to warn me of what’s out there. My comments are dispersed throughout the article. This type of article is exactly why there is so much “confusion, ignorance and misinformation”
Australian people can buy antibiotics in Australia online here: http://buyantibioticsaustralia.com/ No prescription required and cheap price!
Hepatitiscentral.net2A Simple Noninvasive Index Can Predict Both Significant Fibrosis and Cirrhosis in Patients With Chun-Tao Wai,1 Joel K. Greenson,2 Robert J. Fontana,1 John D. Kalbfleisch,3 Jorge A. Marrero,1 Hari S. Conjeevaram,1 and Anna S.-F. Lok1 Information on the stage of liver fibrosis is essential in managing chronic hepatitis C (CHC)
patients. However, most models for predicting liver fibrosis are complicated and separate
formulas are needed to predict significant fibrosis and cirrhosis. The aim of our study was to
construct one simple model consisting of routine laboratory data to predict both significant
fibrosis and cirrhosis among patients with CHC. Consecutive treatment-naive CHC patients
who underwent liver biopsy over a 25-month period were divided into 2 sequential cohorts:
training set (n ؍ 192) and validation set (n ؍ 78). The best model for predicting both
significant fibrosis (Ishak score > 3) and cirrhosis in the training set included platelets,
aspartate aminotransferase (AST), and alkaline phosphatase with an area under ROC curves
(AUC) of 0.82 and 0.92, respectively. A novel index, AST to platelet ratio index (APRI), was
developed to amplify the opposing effects of liver fibrosis on AST and platelet count. The
AUC of APRI for predicting significant fibrosis and cirrhosis were 0.80 and 0.89, respec-
tively, in the training set. Using optimized cut-off values, significant fibrosis could be
predicted accurately in 51% and cirrhosis in 81% of patients. The AUC of APRI for pre-
dicting significant fibrosis and cirrhosis in the validation set were 0.88 and 0.94, respectively.
In conclusion, our study showed that a simple index using readily available laboratory results
can identify CHC patients with significant fibrosis and cirrhosis with a high degree of
accuracy. Application of this index may decrease the need for staging liver biopsy specimens
among CHC patients. (HEPATOLOGY 2003;38:518-526.)
Histologicexaminationoftheliverisanintegral treatmentpossiblycouldbedelayedorwithheld.Onthe part of the evaluation of patients with chronic other hand, patients with significant fibrosis (i.e., septal or hepatitis C (CHC).1,2 Knowledge of the stage of bridging fibrosis) progress almost invariably to cirrhosis liver fibrosis is essential for prognostication and decisions over a 10- to 20- year period so antiviral treatment should on antiviral treatment.3,4 CHC patients with no or min- be strongly considered.5 For patients with cirrhosis, sur- imal fibrosis at presentation appear to progress slowly and veillance for hepatocellular carcinoma and gastroesopha-geal varices should be considered also.6,7 Liver biopsy is currently the gold standard in assessing Abbreviations: CHC, chronic hepatitis C; AST, aspartate aminotransferase; liver histology. Although percutaneous liver biopsy is in gen- ALT, alanine aminotransferase; HCV, hepatitis C virus; IDU, injection drug use; eral a safe procedure, it is costly and does carry a small risk for ALP, alkaline phosphatase; ULN, upper limit of normal; ROC, receiver operatingcharacteristics; AUC, area under receiver operating curves; CI, confidence interval; complication.8 In addition, there could be sampling error APRI, aspartate aminotransferase to platelet count ratio index. because only 1/50,000 of the organ is sampled. Furthermore, From the 1Division of Gastroenterology, 2Department of Pathology, 3Department inter- and intraobserver discrepancies of 10% to 20% in of Biostatistics, University of Michigan Medical School, Ann Arbor, MI. Received March 11, 2003, accepted May 20, 2003. assessing hepatic fibrosis have been reported, which may lead Supported by the Singapore HMDP Fellowship (C.T.W.) and by National In- to understaging of cirrhosis.9-11 Hence, there is a need to stitutes of Health contract N01-DK-9-2323, and grants U01-DK-57577, U01- develop accurate and reliable noninvasive means to assess the DK-62498, and R43-AI-51919 (A.S.-F.L.). Address reprint requests to: Anna S.-F. Lok, M.D., Division of Gastroenterology, University of Michigan Medical Center, 3912 Taubman Center, Box 0362, Ann Noninvasive approaches to assess histology in CHC Arbor, MI 48109-0362. E-mail: firstname.lastname@example.org; fax: 734-936-7392. patients include clinical symptoms and signs, routine lab- Copyright 2003 by the American Association for the Study of Liver Diseases.
0270-9139/03/3802-0030$30.00/0 oratory tests, serum markers of fibrosis and inflammation, quantitative assays of liver function, and radiologic imag- ing studies.12-15 However, at present, none of these tests or Patients were divided into 2 sets: consecutive patients markers alone is accurate or reliable in predicting histol- who were biopsied between January 2001 and July 2002 ogy, in particular, liver fibrosis. An ideal noninvasive di- constituted the training set, whereas those patients who agnostic test for hepatic fibrosis should be simple, readily were biopsied between August 2002 and January 2003 available, inexpensive, and accurate.16 An index compris- constituted the validation set. All study subjects gave in- ing routinely available laboratory tests would meet these formed consent for the liver biopsy. This study was ap- proved by the Institutional Review Board.
Many studies have been performed to evaluate the use Methods. A list of consecutive CHC patients who un-
of readily available laboratory test results to predict signif- derwent percutaneous liver biopsy at the University of icant fibrosis or cirrhosis in patients with CHC.17-21 Michigan Medical Center was generated from the De- However, sensitivity was generally poor, and most studies partment of Pathology. Clinical information about these did not validate their results in a separate group of pa- patients obtained from electronic medical record and tients. A recent study by Forns et al.21 performed internal hard-copy clinical charts were reviewed by one investiga- validation using a randomly chosen cohort from the study tor (C.T.W.) to assess eligibility for the study. Demo- patients and found that absence of significant fibrosis graphics and laboratory variables were recorded. Other could be predicted in 39% of patients. However, only clinical variables were extracted from the medical records 24% of their patients had significant fibrosis so it is un- according to a set of predetermined criteria.
certain if the results could be extrapolated to patients with Patients on diabetic medications or patients who had a past history of diabetes mellitus were considered to have For the prediction of cirrhosis, most studies examined diabetes mellitus. Patients who had been drinking more the usefulness of predetermined formulas such as aspar- than an average of 7 drinks per week, for more than 4 tate aminotransferase (AST)/alanine aminotransferase weeks in a row before the liver biopsy, were considered (ALT) ratio or the cirrhosis discriminant score, without current drinkers. Patients who drank less than 7 drinks analyzing other confounding factors or validating their per week for the past 4 weeks in a row were considered results.22-26 Kaul et al.27 performed univariate and multi- nondrinkers. Patients who had stopped drinking com- variate analysis on 351 patients and derived a model con- pletely for more than 1 year before the biopsy were con- sisting of gender, AST, platelet count, and spider nevi.
sidered ex-drinkers. Patients with no explicitly mentioned This model was validated internally and externally with amount or duration of drinking were considered to have good accuracy but it included one subjective variable.
We aimed to develop one single model consisting of Patients with a history of blood transfusion before readily available, objective laboratory data to predict both 1992 were considered to have acquired CHC through significant fibrosis and cirrhosis in treatment-naive CHC transfusion.28,29 For those with multiple transfusions, patients. To accomplish this, a training set of clinical and the date of the first transfusion was considered to be the laboratory data from 192 consecutive CHC patients were time of infection. Patients with a history of injection used to formulate predictive models, which were vali- drug use (IDU) were considered to have acquired CHC through IDU and the year in which IDU began wasconsidered to be the time of infection. Patients with no Patients and Methods
history of transfusion or IDU, but who had othermodes of percutaneous exposure such as a tattoo, oc- Patients. This retrospective cohort study included
cupational exposure, and so forth, were considered to 579 consecutive adult patients with CHC who had un- have acquired CHC through others means and the year dergone percutaneous liver biopsy at the University of of first percutaneous exposure was considered as the Michigan Medical Center from January 2001 to January time of infection. Patients without parenteral risk fac- 2003. The diagnosis of CHC was established by the pres- tors were considered to be unknown regarding both the ence of hepatitis C virus (HCV) RNA using polymerase chain reaction assays. Patients with the following condi- Except for HCV genotype, only laboratory results per- tions were excluded from the study: presence of other formed within 4 months from the date of the liver biopsy causes of liver disease, hepatocellular carcinoma, prior were used for this study. If more than one set of laboratory liver transplantation, prior interferon therapy, immuno- test results were available, the results closest to the time of suppressive therapy, insufficient liver tissue for staging of biopsy were used. Results of serum aminotransferase and fibrosis, and incomplete data on complete blood counts alkaline phosphatase (ALP) levels were expressed as ratios of the upper limit of normal (ULN). HCV-RNA level was expressed as log10 IU/mL. Abdominal ultrasound reports Table 1. Comparison of Patients in the Training
within 6 months from the time of biopsy were reviewed.
and Validation Sets
Patients with splenomegaly, enlarged spleen, or spleen Training Set
size of more than 14 cm were considered to have spleno- Variable
Histologic slides of all eligible patients were retrieved.
All liver biopsies were reviewed by one pathologist (J.K.G.), who had no knowledge of the clinical character- istics of the study subjects. Hepatic fibrosis was assessed using the Ishak fibrosis score.31 Significant fibrosis was defined as Ishak score of 3 or more (presence of bridging fibrosis) and cirrhosis as Ishak score of 5 or 6.
Statistical Analysis. Data were expressed as mean Ϯ
SEM unless otherwise stated. Statistical analysis was performed by SPSS software version 9.0 (SPSS Inc., Chicago, IL). There were 2 endpoints in this study: presence of significant fibrosis and cirrhosis. The fol- lowing variables were included in the univariate analy- sis: demographics (age, sex, ethnicity), alcohol intake, viral factors (mode of HCV infection, age at infection, duration of infection, HCV-RNA level, genotype), and other test results (white cell count, platelet count, in- ternational normalized ratio, bilirubin, albumin, ALT, AST, and ALP). All continuous variables were analyzed after logarithmic transformation for normality of dis- tribution. Categoric variables were compared by 2 or Fisher exact tests, whereas continuous variables were compared with the Student’s t test. Correlation was evaluated by the Spearman correlation coefficient. A 2-sided P value of less than .05 was considered statisti- NOTE. Values are expressed as mean Ϯ SEM.
For the formulation of predictive models, univariate analysis was performed on variables between patients withand without the study endpoints in the training set. Sig- Characteristics of the Patients in the Training Set.
nificant variables from the univariate analysis (P Ͻ .05), From January 2001 to July 2002, 428 percutaneous liver together with age at biopsy, were then subjected to mul- biopsies were performed on patients with CHC at our tivariate analysis by forward logistic regression to identify institution. A total of 236 patients were excluded from the independent factors associated with either endpoint.
study: 102 had prior interferon therapy, 82 had prior liver Variables with missing values in more than 20% of the transplants, 23 had concomitant liver diseases, 9 were on patients (i.e., splenomegaly on ultrasonography, body immunosuppressive therapy, 4 had insufficient liver tis- mass index, age at infection, and duration of infection) sues for staging of fibrosis, and 16 had incomplete data on were not included in the regression analysis.
complete blood count and/or liver panel.
Formulas with risk scores that could best predict the The mean age of the 192 patients included in the train- study endpoints (significant fibrosis and cirrhosis) were ing set was 46.8 Ϯ 0.6 years, 123 (64%) were men, 151 constructed by entering different sets of independent vari- (79%) were Caucasians, and 16 (8%) were African Amer- ables into the regression model. The diagnostic value of icans. Thirteen (7%) patients had diabetes mellitus (Ta- each formula was assessed by the area under the receiver ble 1). The age at infection and duration of infection, operating characteristic (ROC) curves.
available in 65% of the patients, were 21.1 Ϯ 0.7 years The best model derived from the training set then was and 26.7 Ϯ 0.7 years, respectively. Of the 98 patients who applied to the validation set to test for accuracy by mea- had ultrasound results, 18 (18%) had splenomegaly. The suring the areas under the ROC curves.
mean Ishak fibrosis score was 2.83 Ϯ 0.10. Ninety-one Table 2. Univariate Analysis of Variables Associated With the Presence of Significant Fibrosis
and Cirrhosis in the Training Set
No Significant Fibrosis
Ishak Score 0-2
Ishak Score 3-6
Ishak score 0-4
Ishak Score 5-6
(n ؍ 101)
(n ؍ 164)
(47%) patients had significant fibrosis and 28 (15%) had Regression formula for prediction of significant fibrosis: Predictors of Significant Fibrosis and Cirrhosis
From the Training Set. Variables associated with the
presence of significant fibrosis and cirrhosis were first as- Ϫ 0.375⅐ln (platelet count [109/L]).
sessed by univariate analysis (Table 2). Subsequent mul-tivariate analysis showed that platelet count (P Ͻ .001), Regression formula for prediction of cirrhosis: AST level (P Ͻ .001), and ALP level (P ϭ .029) were theindependent predictors of significant fibrosis whereas platelet count (P Ͻ .001), AST level (P ϭ .017), white cell count (P ϭ .01), ALP level (P ϭ .019), and AST/ALT Ϫ 0.436⅐ln (platelet count [109/L]).
ratio (P ϭ .001) were the independent predictors of cir-rhosis.
Although both histologic endpoints could be predicted by Variables in the best models for prediction of signifi- the same variables, 2 separate formulas were required and cant fibrosis included platelet count, AST levels, and ALP levels, and for prediction of cirrhosis platelet count, white Validation Set. From August 2002 to January 2003,
cell count, AST level, ALP level, and AST/ALT ratio (Ta- 151 liver biopsies were performed on adult patients with ble 3). Models with only platelet count and AST level CHC. Seventy-three patients were excluded from the were more simple and had accuracies comparable with study: 39 had prior interferon therapy, 23 had prior liver those with 3 or more variables in prediction of both end- transplant, 5 had concomitant liver diseases, 2 were on immunosuppressive therapy, and 4 had incomplete re- Table 3. Models With Different Combination of Variables for Predicting Significant Fibrosis and Cirrhosis
in the Training Set and the Validation Set
Prediction of Cirrhosis
Prediction of Cirrhosis
Variables in the Model
AUC (95% CI)
AUC (95% CI)
AUC (95% CI)
AUC (95% CI)
NOTE. NA, not applicable because not all the variables were significant in the regression model.
sults on blood count or liver panel. Seventy-eight patients ROC curves of APRI for predicting significant fibrosis fulfilled the study entry criteria and comprised the valida- and cirrhosis in the training set were plotted in Fig. 2A tion set. Characteristics of the validation set were similar with AUC of 0.80 and 0.89, respectively (Table 3). Based to that of the training set, in particular, there was no on the ROC, 2 cut-off points were chosen to predict the difference in the mean fibrosis score and the proportion absence (coordinate A: APRI Յ 0.50) or presence (coor- with significant fibrosis and cirrhosis. The 2 groups also dinate B: APRI Ͼ 1.50) of significant fibrosis (Fig. 2A).
were comparable in platelet count and AST value. How- For patients with APRI of 0.50 or less, 47 of 55 (85%) ever, there were more African Americans, a higher pro- would not have significant fibrosis. Among the 91 pa- portion with acquisition of hepatitis C through other tients who had significant fibrosis, only 8 (9%) would means besides transfusion and IDU, a higher viral load, have APRI of 0.50 or less, 7 of whom had an Ishak score and a higher ALP level in the validation set (Table 1).
of 3 and 1 had an Ishak score of 4. For patients with APRI Models comprising platelet count and AST level for greater than 1.50, 37 of 42 (88%) would have significant prediction of significant fibrosis and cirrhosis were ap- fibrosis, and only 5 of 101 (5%) without significant fibro- plied to the validation set. The area under ROCs (AUC) sis would be classified incorrectly. Together, using APRI for prediction of significant fibrosis and cirrhosis were below the lower cut-off value (0.50) and above the higher 0.87 (95% confidence interval [CI], 0.79-0.95) and 0.93 cut-off value (1.50), 51% of the patients could be identi- (95% CI, 0.85-1.0), respectively. Formulas with more fied correctly as either without or with significant fibrosis variables did not improve the AUC for either significant fibrosis or cirrhosis in the validation set (Table 3).
Similarly, 2 cut-off points were chosen to predict the Novel Index in Predicting Liver Fibrosis. Because
absence (coordinate C: APRI Յ 1.00) or presence (coor- platelet count and AST level were the most important dinate D: APRI Ͼ 2.00) of cirrhosis (Fig. 2A). For pa- predictors of both significant fibrosis and cirrhosis, we tients with APRI of 1.00 or less, 123 of 126 (98%) would further analyzed the relationship between these 2 factors not have cirrhosis. Only 3 of 28 (11%) with cirrhosis and the stage of hepatic fibrosis. Figure 1A and B showed would be classified falsely. On the other hand, for patients that severity of liver fibrosis was correlated significantly with APRI greater than 2.00, 16 of 28 (57%) had cirrho- with a gradual increase in AST level (r ϭ .50, P Ͻ .001) as sis, and only 12 of 164 (7%) without cirrhosis would be well as a decrease in platelet count (r ϭ Ϫ.46, P Ͻ .001).
identified falsely. Among the 12 patients with APRI However, there was significant overlap in AST and plate- greater than 2.00 but who did not have cirrhosis, 1 had an let among patients with different stages of fibrosis. To Ishak score of 2, 6 had an Ishak score of 3, and 5 had an amplify the difference in AST and platelet values among Ishak score of 4. Using the cut-off values of 1.00 and 2.00, patients with different fibrosis stages, we devised a novel the absence or presence of cirrhosis can be identified in index, called the AST to platelet ratio index (APRI): Applying APRI to the validation set, AUC for predic- tion of significant fibrosis and cirrhosis were 0.88 (95% CI, 0.80-0.96) and 0.94 (95% CI, 0.89-1.0), respectively APRI was correlated significantly with the stage of fibro- (Fig. 2B). Accuracy of using APRI for prediction of sig- sis, with a higher correlation coefficient than platelet nificant fibrosis and cirrhosis in the validation set is com- count, or AST level alone (r ϭ .60, P Ͻ .001) (Fig. 1C).
parable with models with a formula comprising more Finally, we applied the models to the 270 patients from the training and validation sets combined. For the formu-las comprising platelet count and AST level, the AUCwere 0.82 (95% CI, 0.77-0.87) and 0.92 (95% CI, 0.87-0.96) for prediction of significant fibrosis and cirrhosis,respectively. For APRI, the AUC were 0.83 (95% CI,0.78-0.88) and 0.90 (95% CI, 0.86-0.94) for predictionof significant fibrosis and cirrhosis, respectively.
To show the use of APRI in predicting fibrosis, for a hypothetical patient with CHC who has a platelet countof 120 ϫ 109/L and an AST level of 90 IU/L (ULN ϭ45), the APRI could be calculated as follows: This APRI value is more than 1.5 (the higher cut-off valuefor significant fibrosis), so the positive predictive value forsignificant fibrosis is 0.88. The APRI value is less than 2.0 Fig. 1. Box plot of (A) AST, (B) platelet count, and (C) AST platelet ratio index in relation to the Ishak fibrosis score. The box represents the
interquartile range. The whiskers indicate the highest and lowest values,
and the asterisks represent outliers. The line across the box indicates
the median value.
variables (Table 3). Predictive values of the APRI in thevalidation set were similar to that in the training set. Forthe prediction of significant fibrosis in the validation set,the positive predictive value and negative predictive valueof an APRI of 0.50 were 64% and 90%, and the corre-sponding values for an APRI of 1.50 were 91% and 65%,respectively. For the prediction of cirrhosis in the valida-tion set, the positive and negative predictive value of anAPRI of 1.00 were 35% and 100%, and the correspond- Fig. 2. ROC curves of APRI in the prediction of significant fibrosis and cirrhosis in the (A) training set and (B) validation set. An AUC of 1.0 is ing values for APRI of 2.00 were 65% and 95%, respec- characteristic of an ideal test, whereas an AUC of 0.5 or less indicates Table 4. Accuracy of APRI in Predicting Significant Fibrosis and Cirrhosis in the Training Set
(n ؍ 192)
(n ؍ 101)
(n ؍ 164)
Abbreviations: PPV, positive predictive value; NPV, negative predictive value.
(the higher cut-off level for cirrhosis), so the negative pre- naive patients only because several studies have shown dictive value for cirrhosis is 0.93. Hence, this patient is that liver histology may improve even among nonre- likely to have significant fibrosis but not cirrhosis.
sponders to interferon-based therapy.33-35 Secondly, our study included a sufficient proportion of Discussion
patients with significant fibrosis (47%) and cirrhosis In this study, we attempted to develop a single model (15%), thus allowing us to study variables that could pre- using routinely available laboratory test results to predict dict both of the study endpoints within the same patient significant fibrosis and cirrhosis in a consecutive series of population. Although the overall study population only treatment-naive CHC patients. We found that platelet included 270 patients, and differences in race and mode count, AST level, and ALP level were the independent of infection were present between the training and valida- predictors for significant fibrosis, whereas platelet and tion sets, the accuracy of APRI was validated in a sequen- white cell count, AST and ALP levels, as well as AST/ALT tial cohort of CHC patients undergoing a liver biopsy at ratio, were the independent predictors for cirrhosis. Our our institution. This suggests that the model is robust and findings echoed results from many previous studies, which showed that platelet count, AST level, and AST/ALT Most importantly, our predictive model consists of ob- ratio were important predictors of either significant fibrosis jective and readily available laboratory variables. Both or cirrhosis.17-27 To amplify the opposite relationship be- platelet count and AST level are routine tests performed tween the stage of fibrosis and AST level and platelet count, in CHC patients in clinical practice, so no additional tests we devised a novel index, the APRI, which was simple to use are needed. The finding of decreased platelet count and and had comparable accuracy with models that comprised 3 increased AST level with progression of liver fibrosis has or more variables in predicting both significant fibrosis and been reported in many studies. With increasing fibrosis cirrhosis. The performance of APRI in predicting significant and worsening portal hypertension, there is increased se- fibrosis and cirrhosis was validated in a subsequent set of questration and destruction of platelets in the enlarging spleen.36 In addition, studies in liver transplant patients Many studies on prediction of significant fibrosis and showed that progression of liver fibrosis is associated with cirrhosis among CHC patients have been published in the decreased production of thrombopoietin by hepatocytes, past few years.13-27 Our study has several unique features.
and hence reduced platelet production.37,38 Progression First, we recruited consecutive patients undergoing per- of liver fibrosis may reduce the clearance of AST,39 lead- cutaneous liver biopsies at our medical center who met ing to increased serum AST levels. In addition, advanced eligibility criteria. Many prior studies have recruited only liver disease may be associated with mitochondrial injury, patients enrolled in treatment trials,18,32 which may have resulting in more marked release of AST, which is present introduced selection bias. Our study included treatment- in mitochondria and cytoplasm, relative to ALT.40,41 To amplify the difference in AST and platelet values were objective laboratory results, most of which were among patients with different stages of fibrosis, we de- available in the hospital computer system. All histologic vised a novel index, the APRI. The concept of a ratio of 2 slides were retrieved and re-read by one liver pathologist important variables in prediction of significant fibrosis is (J.K.G.) to avoid interobserver discrepancy. In addition, not new. In the study by Williams and Hoofnagle,32 the all the slides were re-read over a 12-week period to mini- investigators observed that as patients with chronic liver disease progressed, AST levels increased more than ALT We acknowledge that there are limitations to our levels. The investigators exploited the difference between study. Our study included patients from a university these 2 factors and devised the AST/ALT ratio for predic- hepatology clinic, half of whom had significant fibrosis on tion of cirrhosis. Although several investigators have con- histology and none had prior antiviral treatment.
firmed the value of AST/ALT ratio in predicting Whether our results can be generalized to community- cirrhosis,22-25 its accuracy varies widely among studies, based practice in which patients may have milder disease, with positive predictive values ranging from 0.64 to 1.00, or to patients who failed prior antiviral therapy remain to and negative predictive values ranging from 0.72 to 0.88, be determined. Despite the simplicity and accuracy of the respectively. In this study, although AST/ALT ratio was 1 APRI, there was overlap among patients with different of the 5 independent predictors of cirrhosis, it alone was stages of fibrosis. Thus, the use of APRI in the prediction insufficient for accurate prediction of cirrhosis. In addi- of fibrosis in individual patients with CHC must be con- tion, AST/ALT ratio alone has not been shown to be firmed in prospective studies. Finally, our study is based useful in predicting significant fibrosis.17-21 on the premise that liver biopsy is the gold standard for The APRI was accurate in predicting both significant assessing hepatic fibrosis, but sampling error as well as fibrosis and cirrhosis, with area under ROC of 0.80 and intra- and interobserver variability can complicate the 0.89 in the training set, and 0.88 and 0.94 in the valida- correlations between histology and noninvasive markers tion set, respectively. Although we could not define one single cut-off value to predict either study endpoint, using In conclusion, we showed that a simple index, the values below the lower cut-off level or above the higher APRI, consisting of 2 readily available laboratory results cut-off level, a prediction of absence or presence of cirrho- (AST level and platelet count), can predict significant sis could be made in 81% of patients. Similarly, a predic- fibrosis and cirrhosis in treatment-naive CHC patients tion of absence or presence of significant fibrosis could be with a very high degree of accuracy. Our results were made in 51% of patients. Our index compared favorably validated in a subsequent cohort of CHC patients at our with results from other studies. Forns et al.21 could predict institute. The APRI can be determined in the clinic or at significant fibrosis in 51% of patients using 4 factors the bedside. Using one simple formula, significant fibrosis (platelet count, ␥-glutamyltransferase level, age, and cho- and cirrhosis can be predicted accurately in 51% and 81% lesterol), with an AUC of 0.94. The fibrosis index from of treatment-naive CHC patients, respectively, potentially the MULTIVIRC group could predict significant fibrosis avoiding the need for liver biopsies in these patients. Further in 46% of patients by using a combination of 6 markers prospective studies are needed to validate the APRI in a larger (␣2 macroglobulin, haptoglobin, ␥ globulin, apolipop- number of CHC patients in other institutes, in particular, totein A1, ␥ glutamyl-transpeptidase, and total bilirubin), community-based practices where the prevalence of signifi- with an AUC of 0.84.15 Although the value of the index of cant fibrosis and cirrhosis may be lower, and in patients who Forns et al.21 in predicting the absence of significant fi- had received antiviral therapy previously.
brosis was better than the APRI, it involved a complicatedformula. The major advantage of the APRI is its simplic- References
ity. APRI can be determined in the clinic or bedside with-out the help of a calculator. Moreover, the APRI allows 1. Gebo KA, Herlong HF, Torbenson MS, Jenckes MW, Chander G, Ghanem KG, El-Kamary SS, et al. Role of liver biopsy in management of chronic clinicians to use one formula to predict significant fibrosis hepatitis C: a systematic review. HEPATOLOGY 2002;36:S161-S172.
2. National Institutes of Health Consensus Development Conference State- Although our study was retrospective in design, we ment: Management of hepatitis C: 2002—June 10-12, 2002. HEPATOL- took all the necessary measures to maximize the accuracy 3. Dienstage JL. The role of liver biopsy in chronic hepatitis C. HEPATOLOGY of our data collection. To ensure consistency in data ex- traction, a predetermined set of criteria for all subjective 4. Lauer GM, Walker BD. Hepatitis C virus infection. N Engl J Med 2001; variables was established before the medical records were 5. Yano M, Kumada H, Hage M, Ikeda K, Shimamatsu K, Inoue O, Hashi- reviewed, and all data extraction was performed by one moto E, et al. The long-term pathological evolution of chronic hepatitis C.
investigator (C.T.W.). The key variables in our study 6. Lok A, McMahon BJ. Chronic hepatitis B. HEPATOLOGY 2001;34:1225- 24. Park GJ, Lin BP, Ngu MC, Jones DB, Katelaris PH. Apartate aminotrans- ferase: alanine aminotransferase ratio in chronic hepatitis C infection: is it 7. Jalan R, Hayes PC. UK guidelines on the management of variceal haem- a useful predictor of cirrhosis? J Gastroenterol Hepatol 2000;15:386-390.
orrhage in cirrhotic patients. Gut 2000;46:III1-III15.
25. Imperiale TF, Said AT, Cummings OW, Born LJ. Need for validation of 8. Cadranel JF, Rufat P, Degos F. Practices of liver biopsy in France: results of clinical decision aids: use of the AST/ALT ratio in predicting cirrhosis in a prospective nationwide survey. HEPATOLOGY 2000;32:477-481.
chronic hepatitis C. Am J Gastroenterol 2000;95:2328-2332.
9. Bedossa P, Poynard T, and the Metavir Cooperative Group. Intraobserver 26. Saddeh S, Cammell G, Carey WD, Younossi Z, Barnes D, Easley K. The role and interobserver variations in liver biopsy interpretation in patients with of liver biopsy in chronic hepatitis C. HEPATOLOGY 2001;133:196-200.
chronic hepatitis C. HEPATOLOGY 1994;20:15-20.
27. Kaul V, Friedenberg FK, Braitman LE, Anis U, Zaeri N, Fazili J, Herrine 10. Westin J, Lagging LM, Wejstal R, Norkans G, Dhillon AP. Interobserver SK, et al. Development and validation of a model to diagnose cirrhosis in study of liver histology using the Ishak score in patients with chronic patients with hepatitis C. Am J Gastroenterol 2002;97:2623-2628.
hepatitis C virus infection. Liver 1999;19:183-187.
28. Alter MJ. Prevention of spread of hepatitis C. HEPATOLOGY 2002;36:S93-S98.
11. Regev A, Berho M, Jeffers LJ, Milikowski C, Molina EG, Pyrsopoulos NT, 29. Centers for Disease Control. Recommendation for prevention and control Feng ZZ, et al. Sampling error and intraobserver variation in liver biopsy in of hepatitis C virus (HCV) infection and HCV-related chronic disease.
patients with chronic HCV infection. Am J Gastroenterol 2002;97:2614- MMWR Morb Mortal Wkly Rep 1998;47:RR-19:RR-39.
30. Dardenne AN. The spleen. In: Cosgrove DO, Meire HB, Dewbury KCD, 12. Geroge J. Biochemical markers of hepatic fibrogenesis: single measure- eds. Clinical Ultrasound. Edinburgh: Churchill Livingstone, 1993;353-365.
ments are not reliable enough to replace liver biopsy. J Gastroenterol Hepa- 31. Ishak K, Baptista A, Bianchi L, Callea F, De Groote J, Gudat J, Denk H, et al. Histological grading and staging of chronic hepatitis. J Hepatol 13. McHutchison JG, Blatt LM, de Medina M, Craig JR, Conrad A, Schiff ER, Tong MJ. Measurement of serum hyaluronic acid in patients with 32. Williams AL, Hoofnagle JH. Ratio of serum aspartate to alanine amino- chronic hepatitis C and its relationship to liver histology. Consensus In- transferase in chronic hepatitis. Relationship to cirrhosis. Gastroenterology terferon Study Group. J Gastroenterol Hepatol 2000;15:945-951.
14. Guechot J, Laudat A, Loria A, Serfaty L, Poupon R, Giboudeau J. Diag- 33. de Ledinghen V, Trimoulet P, Winnock M, Bernard PH, Bourliere M, nostic accuracy of hyaluronan and type III procollagen amino-terminal Portal I, Remy AJ, et al. Daily or three times per week interferon alpha-2b peptide serum assays as markers of liver fibrosis in chronic viral hepatitis C in combination with ribavirin or interferon alone for the treatment of evaluated by ROC curve analysis. Clin Chem 1996;42:558-563.
patients with chronic hepatitis C not responding to previous interferon 15. Imbert-Bismut F, Ratziu V, Pieroni L, Charlotte F, Benhamou Y, Poynard T, for the MULTIVIRC group. Biochemical markers of liver fibrosis in 34. Teuber G, Berg T, Naumann U, Raedle J, Brinkmann S, Hopf U, Zeuzem patients with hepatitis C virus infection: a prospective study. Lancet 2001; S. Randomized, placebo-controlled, double-blind trial with interferon- alpha with and without amantadine sulphate in primary interferon-alpha 16. Fontana RJ, Lok AS. Noninvasive monitoring of patients with chronic nonresponders with chronic hepatitis C. J Viral Hepat 2001;8:276-283.
hepatitis C. HEPATOLOGY 2002;36:S57-S64.
35. Di Bisceglie AM, Thompson J, Smith-Wilkaitis N, Brunt EM, Bacon BR.
17. Bonacini M, Hadi G, Govindarajan S, Lindsay KL. Utility of a discriminant score for diagnosing advanced fibrosis or cirrhosis in patients with chronic Combination of interferon and ribavirin in chronic hepatitis C: re-treat- hepatitis C virus infection. Am J Gastroenterol 1997;92:1302-1304.
ment of nonresponders to interferon. HEPATOLOGY 2001;33:704-707.
18. Pohl A, Behling C, Oliver D, Kilani M, Monson P, Hassanein T. Serum 36. Aster R. Pooling of platelets in the spleen: role in the pathogenesis of aminotransferase levels and platelet counts as predictors of degree of fibro- “hypersplenic” thrombocytopenia. J Clin Invest 1996;45:645-657.
sis in chronic hepatitis C virus infection. Am J Gastroenterol 2001;96: 37. Kawasaki T, Takeshita A, Souda K, Kobayashi Y, Kikuyama M, Suzuki F, Kageyama F, et al. Serum thrombopoietin levels in patients with chronic 19. Poynard T, Bedossa P, METAVIR and CLINIVIR cooperative study hepatitis and liver cirrhosis. Am J Gastroenterol 1999;94:1918-1922.
groups. Age and platelet count: a simple index for predicting the presence 38. Adinolfi LE, Giordano MG, Andreana A, Tripodi MF, Utili R, Cesaro G, of histological lesions in patients with antibodies to hepatitis C virus. J Ragone E, et al. Hepatic fibrosis plays a central role in the pathogenesis of thrombocytopenia in patients with chronic viral hepatitis. Br J Haematol 20. Wong V, Caronia S, Wight D, Almer CR, Petrik J, Britton P, Alexander GJM. Importance of age in chronic hepatitis C virus infection. J Viral 39. Kamimoto Y, Horiuchi S, Tanase S, Morino Y. Plasma clearance of intra- venously injected aspartate aminotransferase isozymes: evidence for pref- 21. Forns X, Ampurdanes S, Llovet JM, Aponte J, Quinto L, Martinez-Bauer E, erential uptake by sinusoidal liver cells. HEPATOLOGY 1985;5:367-375.
Bruguera M, et al. Identification of chronic hepatitis C patients without he- 40. Okuda M, Li K, Beard MR, Showalter LA, Scholle F, Lemon SM, Wein- patic fibrosis by a simple predictive model. HEPATOLOGY 2002;36:986-992.
man SA. Mitochondrial injury, oxidative stress, and antioxidant gene ex- 22. Reedy DW, Loo AT, Levine RA. AST/ALT ratio Ͼ or ϭ 1 is not diagnos- pression are induced by hepatitis C virus core protein. Gastroenterology tic of cirrhosis in patients with chronic hepatitis C. Dig Dis Sci 1998;43: 41. Nalpas B, Vassault A, Le Guillou A, Lesgourgues B, Ferry N, Lacour B, 23. Sheth SG, Flamm SL, Gordon FD, Chopra S. AST/ALT ratio predicts Berthelot P. Serum activity of mitochondrial aspartate aminotransferase: a cirrhosis in patients with chronic hepatitis C virus infection. Am J Gastro- sensitive marker of alcoholism with or without alcoholic hepatitis. HEPA-
Inflammatory Bowel Disease: Crohn’s Disease ● Endoscopy and biopsies of the upper and lower intestine Inflammatory Bowel Disease (IBD)? How is Crohn’s disease treated? I BD refers to a chronic inflam- The aim of treatment is to decrease the inflammation causingthe damage to the intestines. Even though a cure is not yetpossible, control of symptoms can be very effective in