Orlistat for Overweight Subjects with Nonalcoholic Steatohepatitis

To the Editor:

We read with interest the article by Harrison et al.1 on the use ofOrlistat in overweight patients with nonalcoholic steatohepatitis(NASH). Their prospective randomized trial of 41 patients concludedthat improvement in liver histology was not significantly different after36 weeks of treatment with Orlistat/vitamin E compared to vitamin Ealone. The mean weight loss between the two groups was higher in theOrlistat group but did not reach statistical significance (8.3% versus6.0%). The authors therefore reanalyzed the data to compare thosewho lost � or � 5% body weight or � or � 9% body weight,respectively. Using these stratifications, they demonstrated that weightloss of 5% was associated with improvement in steatosis but notNAFLD activity score (NAS), whereas weight loss of 9% was alsoassociated with an improvement in NAS.

Regarding design, there are no power calculations presented in themethods section, so there is no way of assessing if the trial was adequatelypowered to detect differences between the treatment groups. Also, threeprimary endpoints are listed; does this imply the study was powered for allthree primary outcomes? We estimate that to detect a 10% reduction inweight could require up to 60 patients per group and question whether thestudy presented was adequately powered. The authors’ data demonstratesgreater loss of weight in the Orlistat group, but this did not reach statisticalsignificance and could be due to power. With respect to the subgroupanalysis, this does not appear to have been powered at the outset, and assuch, descriptive statistics only should be presented.

Regarding analysis, it would be helpful if the authors could clarify whythe percent weight loss is categorized into two further analyses: �5%,�5% and �9%, �9% and if these were preplanned or data-driven cut-points. As a continuous measure, the percent weight loss should be ana-lyzed as a continuous variable since loss of information and bias areintroduced by dichotomizing variables. The statistical analyses presenteduse a mix of parametric and nonparametric approaches; given the smallsize of the study, then nonparametric statistics are more appropriate(Spearman’s correlation, Mann-Whitney tests).

No significant difference was observed in the histological improve-ments seen between the two groups at 36 weeks; however, a detailedhistopathological description was only provided for liver biopsies at thestart of the study. We would be interested to see a more detaileddescription of the histopathological changes at follow-up of the twotreatment groups.

We know that the degree of weight loss achieved with Orlistat isvariable. Orlistat may provide additional histological improvementcompared with a similar degree of diet-induced weight loss, but only inthe subset of patients who achieve �5%-10% weight loss. Indeed,previous studies have suggested that Orlistat can improve hepatic ste-atosis beyond its effect on weight reduction,2 and exerts additionalbeneficial effects on inflammation and fibrosis.3

We feel it may be premature at this stage to discount a role forOrlistat in NASH. Given the magnitude of NASH in the West, thereremains a pressing need for larger studies which are powered to detectdifferences in weight loss/NAS score.

JOANNA K. DOWMAN, M.D.1

DEBORAH STOCKEN, C.STAT.2

JEREMY W. TOMLINSON, M.D., PH.D.3

PHILIP NEWSOME, M.D., PH.D.41Wellcome Clinical Research Fellow in Hepatology and 4Senior Lecturer

in Hepatology and Honorary Consultant Physician, Centre for LiverResearch, Institute of Biomedical Research; 2Senior Biostatistician,Cancer Research UK Clinical Trials Unit; and 3Wellcome ClinicianScientist and Honorary Consultant Physician, School of Clinical andExperimental Medicine, Institute of Biomedical Research, Universityof Birmingham, Edgbaston, Birmingham, UK

References1. Harrison SA, Fecht W, Brunt EM, Neuschwander-Tetri BA. Orlistat for

overweight subjects with nonalcoholic steatohepatitis: A randomized, pro-spective trial. HEPATOLOGY 2009;49:80-86.

2. Zelber-Sagi S, Kessler A, Brazowsky E, Webb M, Lurie Y, Santo M, et al. Adouble-blind randomized placebo-controlled trial of orlistat for the treatment ofnonalcoholic fatty liver disease. Clin Gastroenterol Hepatol 2006;4:639-644.

3. Hussein O, Grosovski M, Schlesinger S, Szvalb S, Assy N. Orlistat reversefatty infiltration and improves hepatic fibrosis in obese patients with nonal-coholic steatohepatitis (NASH). Dig Dis Sci 2007;52:2512-2519.

Copyright © 2009 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.23079Potential conflict of interest: Dr. Tomilson is a consultant for Bristol-Myers

Squibb. Dr. Newsome is an investigator on trial for Astellas.

Reply:

We appreciate the comments of Dowman and colleagues in referenceto our study with orlistat in nonalcoholic steatohepatitis.1 Regarding studydesign, specifically with respect to the lack of a calculation for power, thisstudy was designed as a pilot trial and not necessarily intended to defini-tively answer the question of whether orlistat therapy alone is efficacious.However, we have subsequently performed a retrospective power analysisusing the values for weight loss reported in the article. The independentvariables are treatment (control, orlistat) and time (before, after treat-ment). The dependent variable is body weight. The mean � standarddeviation (SD) of body weight before treatment is 225 � 42 pounds. A10% weight loss will be 22.5 pounds, an effect size of 0.54 SD. Fourtwo-tailed, post hoc tests (two between and two within groups) are appro-priate for this analysis, so a Bonferroni correction of P � 0.05/4 � 0.01was used. With SPSS Sample Power, version 2.0, we obtained a samplesize estimate of 82 subjects per group for a power of 80% with a level ofconfidence of 95%. We concur that a larger sample size than was used inthe article is needed to detect a 10% difference in weight loss betweengroups. With 23 subjects in one group and 18 in the other, we had thepower to detect a 1.13 SD effect size or 25.4 pound difference in weightloss between groups. Presented as descriptive statistics, body weight inpounds was 226 (95% confidence interval [CI]: 153-299) in the orlistatgroup before treatment, 224 (95% CI: 132-316) in the control groupbefore treatment, 208 (95% CI: 132-284) in the orlistat group after treat-ment, and 210 (95% CI: 120-300) in the control group after treatment.

Regarding analysis, as previously noted in our article, there were no signif-icantdifferences inhistopathologybetweenthe twogroups, so further case-by-case analysis was not thought to be helpful in this pilot trial. Subsequently,when no treatment effect was observed, the data were pooled and categorizedbyweight loss.Theobservedweight losseswere6.0%inthecontrolgroupand8.3% in the orlistat group. The 5% and 9% cut-points bracket the observedweight loss values and split the group sizes by a 2:3 ratio. Power analyses werenot performed. Power is not an issue when statistically significant differencesare found. However, increasing the number of hypothesis tests increases theprobability of a Type I error. The results of Spearman correlation analysisbetween weight change � 5% and changes in dependent variables were thequantitative insulin-sensitivity check index (QUICKI; r � �0.327, P �0.039, n�40) and steatosis (r��0.388, P�0.013, n�40). The results ofSpearman correlation analysis for weight change � 9% were QUICKI (r ��0.385, P � 0.014, n � 40), steatosis (r � �0.453, P � 0.003, n � 40),ballooning (r � �0.326, P � 0.040, n � 40), inflammation (r � �0.321,P � 0.044, n � 40), adiponectin (r � 0.461, P � 0.002, n � 41), and theNAFLDactivity score (NAS; r��0.419,P�0.007,n�40).Weagree thatlarger studies are needed to fully delineate the efficacy of orlistat in the treat-ment of nonalcoholic steatohepatitis. However, quantification of a specificweight loss goal that is associated with clinical and histopathologic improve-

HEPATOLOGY, Vol. 49, No. 0, 2009 CORRESPONDENCE 321

ment is also important. We thank Dowman and colleagues for giving us theopportunity to present the descriptive and nonparametric results.

STEPHEN HARRISON1

ELIZABETH BRUNT2

BRENT NEUSCHWANDER-TETRI3

1Gastroenterology Service, Department of Gastroenterology, BrookeArmy Medical Center, Fort Sam Houston, TX

2Department of Pathology and Immunology, Washington UniversitySchool of Medicine, St. Louis, MO

3Division of Gastroenterology and Hepatology, Saint Louis University,St. Louis, MO

Reference1. Harrison SA, Fecht W, Brunt EM, Neuschwander-Tetri BA. Orlistat for

overweight subjects with nonalcoholic steatohepatitis: a randomized, pro-spective trial. HEPATOLOGY 2009;49:80-86.

Copyright © 2009 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.23017Potential conflict of interest: Dr. Neuschwander-Tetri is a consultant for Gilead,

Amylin Pharmaceuticals, and Centofor.

Diagnosis, Management, and Treatment of Hepatitis C

To the Editor:

We read with great interest the updated American Association forthe Study of Liver Diseases practice guideline on the diagnosis, man-agement, and treatment of hepatitis C.1 In particular, the authorsshould be warmly congratulated for their practical and exhaustive ap-proach to patients with persistently normal alanine aminotransferase(PNALT) levels.

However, we have some concerns about the definition and man-agement of hepatitis C virus (HCV) carriers with PNALT proposed byGhany et al.1 We believe that the definition given in this article (“anALT value of less than 40 IU/L on 2 to 3 occasions separated by at leasta month over a period of six months”), although commonly used inclinical practice, could be misleading, as this observation period is tooshort and thus not adequate to discriminate between true HCV carrierswith PNALT levels and patients with only transient biochemical re-mission.2 Two or three normal alanine aminotransferase (ALT) valuesover a short-term period may not be representative of the true patternof ALT levels for a particular patient.3 Indeed, several studies haveshown that many subjects called HCV carriers with PNALT on thebasis of a 6-month observation period did suffer from ALT flaresduring the follow-up, showing histological worsening and fibrosis pro-gression after these flares.4,5

In light of these data, the Italian Association for the Study of theLiver6 suggested that the definition of HCV carriers with PNALTshould be made on the basis of at least nine normal ALT values 2months apart over an 18-month period. We have to be very cautiousbefore we define these persons as subjects with PNALT, given the risksof sudden exacerbation of the disease and a less benign natural historyin many of these apparently healthy carriers.

Furthermore, we have some comments to offer regarding the sug-gested management of these patients.1 The authors state that HCV-infected persons with normal ALT values do warrant treatment if liverbiopsy shows significant fibrosis. This approach seems to be too restric-tive, as it excludes from treatment many patients who might requiretherapy.

An International Clinical Workshop7 suggested that highly moti-vated, young patients with HCV 2 or 3 might have an excellent re-sponse to treatment and thus, in the absence of any contraindication,should receive treatment with pegylated interferon plus ribavirin ther-apy without the need for liver biopsy. On the contrary, in patientsharboring HCV type 1 or 4 (regardless of age) or in older patients(regardless of HCV type), liver biopsy might be invariably offered todecide the need for therapy, with treatment recommended only forpatients with evidence of liver disease (�F2).

This approach allows more tailored therapy for HCV carriers withPNALT,8 avoiding unnecessary biopsies in many patients and provid-ing the possibility of safe and highly effective treatment of HCV in-fection for selected patients.9,10

CLAUDIO PUOTI

RICCARDO GUARISCO

LIA BELLIS

LUCIA SPILABOTTI

Department of Internal Medicine and Liver UnitMarino General HospitalRome, Italy

References1. Ghany MG, Strader DB, Thomas DL, Seef LB. Diagnosis, management,

and treatment of hepatitis C: an update. HEPATOLOGY 2009;49:1335-1373.

2. Puoti C, Castellacci R, Montagnese F. Hepatitis C virus carriers withpersistently normal aminotransferase levels: healthy people or true pa-tients? Dig Liver Dis 2000;32:634-643.

3. Puoti C. HCV carriers with persistently normal aminotransferase levels:normal does not always mean healthy. J Hepatol 2003;38:529-532.

4. Puoti C, Castellacci R, Montagnese F, Zaltron S, Stornaiuolo M, Pace M,et al. Histological and virological features and follow-up of HCV carrierswith normal aminotransferase levels: the Italian Study of the Asymptom-atic C Carriers (ISACC). J Hepatol 2002;37:117-123.

5. Martinot-Peignoux M, Boyer N, Cazals-Hatem D, Bach-Nga P, GervaisA, Le Breton A, et al. Perspective study of anti hepatitis C virus positivepatients with persistently normal serum ALT with or without detectableserum HCV RNA. HEPATOLOGY 2001;34:1000-1005.

6. Puoti C, Guido M, Mangia A, Persico M, Prati D. Clinical management ofHCV carriers with normal aminotransferase levels. Dig Liver Dis 2003;35:362-369.

7. Zeuzem S, Alberti A, Rosenberg W, Marcellin P, Diago M, Negro F, et al.Review article: management of patients with chronic hepatitis C virusinfection and ‘normal’ alanine aminotransferase activity. Aliment Pharma-col Ther 2006;24:1133-1149.

8. Snoeck E, Hadziyannis SG, Puoti C, Swain MG, Berg T, Marcellin P, et al.Predicting efficacy and safety outcomes in patients with hepatitis C virusgenotype1and persistently ‘normal’ alanine aminotransferase levels treatedwith peginterferon alpha-2a (40KD) plus ribavirin. Liver Int 2008;28:61-71.

9. Puoti C, Pellicelli AM, Romano M, Mecenate F, Guarisco R, Barbarini G,et al. Treatment of HCV carriers with persistently normal alanine amino-transferase levels with peginterferon alfa-2a and ribavirin: a multicentricstudy. Liver Int. In press.

10. Puoti C. Hepatitis C virus with normal transaminase levels. Dig Dis 2007;25:277-278.

Copyright © 2009 by the American Association for the Study of Liver Diseases.Published online in Wiley InterScience (www.interscience.wiley.com).DOI 10.1002/hep.23015Potential conflict of interest: Nothing to report.

322 CORRESPONDENCE HEPATOLOGY, July 2009