Year : 2006 | Volume
: 12 | Issue : 2 | Page : 73--76
Clinical presentation, response to therapy, and predictors of fibrosis in patients with autoimmune hepatitis in Saudi Arabia
Ayman A Abdo
Division of Gastroenterology College of Medicine, King Saud University, Saudi Arabia
Ayman A Abdo
P.O. Bob 2925(59) Riyadh 11461
Background: Autoimmune hepatitis (AIH) is a relapsing inflammatory disease of the liver of unknown cause. Little is known about AIH in Saudi Arabia.
Objectives: Our aim is to  identify the special clinical or histological features in our patients, describe the initial response to immunosuppressive therapy and long term relapse, identify clinical and laboratory predictors of response to therapy, and  examine the utility of laboratory markers (platelet count, AST/ALT ratio, and the AST to platelet ratio index (APRI score) in predicting the presence of advanced fibrosis.
Methods: Patients were identified using a computer database. Patients responding to initial therapy were compared with patients who did not respond in terms of laboratory and histology parameters. The utility of three fibrosis markers/models were then examined.
Results: Thirty-nine patients with AIH were included in this analysis. The mean age was 45.4; 65% of patients were females. Mean ALT at presentation was 268 U/L and AST was 277 U/ L. GGT level was found to be the only statistically significant laboratory or histopathological parameter difference between responders and non-responders. Platelet count and AST/ALT ratio were found to be the best predictors of advanced fibrosis.
Conclusion: In Saudi patients with AIH, we found that the GGT level at presentation may serve as a useful predictor for response to therapy. Platelet count and AST/ALT ratio may be used to predict advanced fibrosis.
|How to cite this article:|
Abdo AA. Clinical presentation, response to therapy, and predictors of fibrosis in patients with autoimmune hepatitis in Saudi Arabia.Saudi J Gastroenterol 2006;12:73-76
|How to cite this URL:|
Abdo AA. Clinical presentation, response to therapy, and predictors of fibrosis in patients with autoimmune hepatitis in Saudi Arabia. Saudi J Gastroenterol [serial online] 2006 [cited 2023 Feb 2 ];12:73-76
Available from: https://www.saudijgastro.com/text.asp?2006/12/2/73/27849
Autoimmune hepatitis (AIH) is a relapsing inflammatorydisease of the liver of unknown cause. It is characterizedby the presence of hypergammaglobulinemia, serum autoantibodies,and liver infiltration with lymphocytes and plasmacells.
Unfortunately, little is known about the epidemiology,clinical course, and outcome of autoimmune hepatitis inSaudi Arabia.
In this study, we review all the patients diagnosed withautoimmune hepatitis in our gastroenterology unit at KingKhalid university hospital in Riyadh.
Our aim is to (1) identify any special clinical or histologicalfeatures in our patients that are different compared to theinternational literature (2), describe the initial response toimmunosuppressive therapy and long term relapse (3), identifyclinical and laboratory predictors of response to therapy, and(4) examine the utility of laboratory markers (platelet count,AST/ALT ratio, and the AST to platelet ratio index (APRIscore) in predicting the presence of advanced fibrosis definedas a fibrosis score of 3 or 4.
From a database of all liver biopsies performed in our unitfrom 1996-2004 we identified the biopsies with the finaldiagnosis of AIH. We then reviewed all the files and verifiedthe pathological diagnosis with the clinical diagnosis andincluded in this study only patients with the pathological andclinical diagnosis of AIH. History and physical examinationfindings were collected from the patients' charts. Laboratorytests were obtained from the hospital computer database. Liverbiopsy reports were obtained and reviewed for histologicalfeatures of AIH.
We then reviewed the initial treatment given to the patientand evaluated the initial response to therapy. Initial responsewas defined as the clinical and biochemical data three monthsafter starting therapy. Complete response was defined asnormalization of the ALT and GGT after three months oftherapy. Partial response was defined as 50% or more reductionin ALT or GGT values from pretreatment values. No responsewas defined as less that 50% reduction or any increase in theALT or GGT values. These definitions are in keeping with theAmerican Association for the study of the liver guidelines onthe diagnosis and management of AIH.
We then recorded the maintenance therapy that the patientsreceived and the outcome of this maintenance therapy. In thisregard, a sustained remission was defined as totally normalliver enzymes and liver function tests 6 months after startingmaintenance therapy.
We then compared between the patients who respondedto therapy and patients who did not in terms of all thelaboratory and histology parameters in an attempt to identifylaboratory or histological predictors of response to therapy.Subsequently, we looked at the sensitivities, specificities,positive and negative predictive values, and receiver operatorcharacteristic (ROC) curves to evaluate the accuracy of threewell known laboratory parameters (platelet count, AST/ALTratio, and APRI score) in predicting the presence of advancedfibrosis defined as a fibrosis score of 3 or 4 on the METAVIRliver histology classification system. In this particularanalysis, we used the data from forty patients as there weredata available on one more patient.
For all statistical analyses performed, the data was enteredin MS Excel software. SPSS PC+ statistical software wasused for the statistical analyses. The association between thecategorical variables was assessed by using the Chi-square testor Fisher exact test. The sensitivity, specificity, and positiveand negative predictive values were calculated to assess thevariables (platelet count, AST/ALT ratio, and the APRI score)in prediction of the presence /absence of advanced fibrosisand cirrhosis.
Of a database containing five hundred and twenty liverbiopsies, fifty-seven patients were found to have AIH as thefinal diagnosis from their liver biopsy reports. After reviewingthe patient's files, forty-eight patients were found to have bothhistological and clinical diagnosis of AIH. Of these fortyeightpatients, nine patients were either lost to follow up afterthe liver biopsy or were found to have incomplete data, andso they were excluded. The remaining thirty-nine patientswere included in this analysis. The mean age was 45.4 withthe oldest patient being eighty years old and the youngesteighteen. Females constituted 65% of the patients.
Ten patients presented with abnormal liver enzymes thatwere incidentally found to be raised but without any symptoms,eight patients presented with elevated liver enzymes and nonspecificsymptoms, eighteen patients presented with jaundice,and three patients presented with severe acute hepatitis.Twenty-three patients had an associated autoimmune disease(the most common of which was thyroid disease).
Initial laboratory data are summarized in tabular form[Table 1]. Mean ALT at presentation was 268 U/L and AST was277 U/L. A negative anti-nuclear antibody (ANA) was seenin 35% of patients, 18% had ANA between 1:20 and 1:120,and 47% had an ANA which was higher than 1:120. Antismoothmuscle antibody (ASMA) was positive in 75%, 45%had ASMA between 1:20 and 1:120, and 30% had an ASMAhigher than 1:120. Both ANA and ASMA were negative in17% of the patients, 13% had both positive above 1:120,while all the remaining patients had either ANA or ASMApositive. Important histological data is also summarized intabular form [Table 2].
Initial treatment and response
Of the thirty-nine patients who were treated in our center,twenty received prednisone 60 mg as their initial therapy whilenineteen patients received a combination of prednisone 30mg and azathioprine 50 mg. In both regimens, steroids weretapered quickly over a month. Eighteen patients (46%) hadcomplete response, twelve (30%) had partial response, and nine(23%) had no response.
Maintenance therapy and follow up
Patients were followed for an average duration of forty-fivemonths. In regards to maintenance therapy, three patients (7.6%)were maintained on low dose prednisone therapy alone, eight(20.5%) were maintained on azathioproine and prednisone,while nineteen (48.7%) were maintained on azathioprine alone.One patient with sever acute hepatitis died and one refusedto take maintenance therapy, while seven patients (18%) didnot require maintenance therapy because they were totallyasymptomatic and had normal liver enzymes after treatment.Within the duration of follow up, eight patients (20%) had atleast one flare on maintenance therapy requiring reintroductionor increased dosage of prednisone. Two patients, who continuedto have elevated liver enzymes on 30 mg of prednisone and 150mg of azathioprime, were switched to mycophenolate mofetil(starting dose of 500 mg twice daily up to 1 gm twice daily) withsubsequent normalization of their liver enzymes and withoutany serious side effects. Those two patients were successfullytapered from steroids completely. The remainder of the patientsachieved sustained remission on maintenance therapy.
Differences between initial responders and nonresponders
We compared patients who had a complete or partialresponse to initial therapy with those who had no response [Table 1][Table 2]. We found that only the level of GGT wasstatistically significantly different between the two groups(173 vs. 499, p=0.002). GGT was also the only significantlydifferent parameter found when the analysis was done withpatients achieving complete response was compared topatients with partial response and no response combined.
Patients with advanced fibrosis at presentation
Fifteen patients (38%) had advanced fibrosis at presentation.When these patients were compared to patients presentingwithout advanced fibrosis, no statistically significantdifferences were found except for the albumin level. Similarly,there were no significant differences between patients withadvanced fibrosis and those without in terms of initial responseto therapy or need for maintenance therapy.
Validation of fibrosis markers
The results for different parameters predicting fibrosis havebeen tabulated [Table 3]. It was found that a platelet countabove 100 had a sensitivity of 100% in predicting absence ofadvanced fibrosis but with a specificity of about 10%. AST/ALT ratio more than 0.70 was also found to be predictive ofadvanced fibrosis with a sensitivity of 100%, but with a poorspecificity of 50%. The APRI score was less sensitive andspecific.
Similarly, using the ROC curves to compare between thesepredictive models, we found that both platelet count and AST/ALT ratio gave reasonable area under the curve of 0.783 and0.793 respectively, while the APRI score gave an area underthe curve of 0.690.
Our study of all patients with autoimmune hepatitis in ourcenter showed that the clinical features of AIH at presentation,specifically age, sex distribution, associated diseases, liverenzymes elevations, and auto-antibodies are similar to whatis described in the international literature,. Similarly, thepercentage of patients with advanced fibrosis and cirrhosis(about 35%), the initial response to steroid therapy (around80%), and the need for long term maintenance therapy (around20%) were all compatible with the international literature.
It has been noticed in many studies that racial, regional,and genetic factors may affect the presentation and outcomeof patients with AIH. For example, cirrhosis is present morecommonly in black North American patients with AIH than inwhite. In addition, Blacks tend to be younger at presentationthan their counterparts. Other studies in Alaskan natives,Turks, Japanese, South Americans, and non-European, nonwhitepatients with AIH have shown marked differences inprevalence and prognosis.
Similar to the study by Roberts et al, we found thatpatients with cirrhosis had a similar response to therapy anda similar outcome compared to patients without cirrhosis.These findings are in contrast to the recently published datafrom Canada showing a less favorable outcome of patientspresenting with cirrhosis.
Two of our patients who did not respond to the combinationof high dose steroid and azathioprine therapy were started onMycophenolate mofetil with total response to therapy with noside effects. There has been increasing reports on the use ofthis agent in treating steroid resistant and steroid dependantpatients,.
When we looked at predictors of response to therapy wefound that the level of GGT may be a useful laboratory basednon-invasive predictor of response to therapy as it was the onlylaboratory parameter difference between responders and nonresponders,a finding not reported before to our knowledge.We found no histological parameters that may predict responseto therapy. Liver biopsy has been shown to be useful priorto termination of treatment as interface hepatitis is found in55% of patients with normal serum aminotransferases duringtherapy and these patients were shown to invariably relapseafter cessation of treatment.10 On the other hand, there ishardly any evidence in the literature regarding laboratory orhistologic predictors of response to initial therapy.
This study is one of the first studies to validate the utility ofplatelet count, AST/ALT ratio, and APRI score in predictingadvanced fibrosis in patients with AIH. These markers havebeen validated and found to be fairly sensitive and specific inhepatitis C mainly in international studies,. Our group hasalso validated the utility of these tests in local Saudi patientswith hepatitis B and C. In the present study, we found thatthe best utility of these markers in patients with autoimmunehepatitis is when a platelet count above 100 is used to predictthe absence of advanced fibrosis and ASL/ALT ratio of morethan 0.7 is used to predict the presence of advanced fibrosis[Table 3]. The APRI score was less useful in this study inpatients with AIH while being more useful in our previousstudy on local hepatitis B and C patients. These noninvasivemarkers should prove very useful in assessment and followupof patients with AIH.
Our study suffers from the usual shortcomings ofretrospective analyses. In addition, the number of patients istoo small to be able to offer concrete conclusions. Nevertheless,this study offers very useful information (some of which isoriginal) regarding the management of patients with AIH.
In conclusion, by reviewing all the patients with AIH in ourcenter we found that there are no peculiar features in our patientsin terms of epidemiological factors, clinical presentation,laboratory parameters, and response to immunosuppressivetherapy. We also found that the GGT level at presentationmay serve as a useful predictor of response to therapy. Plateletcount and AST/ALT ratio may be used to predict advancedfibrosis.
|1||Czaja A, Freese K. Diagnosis and treatment of autoimmune hepatitis. Hepatology 2002; 36:479-97.|
|2||Bedossa P, Poynard T. An algorithm for grading activity in chronic hepatitis C. Hepatology 1996; 24:289-93.|
|3||Czaja AJ. Current concepts in autoimmune hepatitis. Am Hepatol 2005; 4:6-24.|
|4||Lim K, Casanova R, Boyer T, Bruno C. Autoimmune hepatitis in African Americans: presenting features and responses to therapy. Am J Gastroenterol 2001; 96:3390-4.|
|5||Czaja A, Bianchi F, Carpenter HA, Krawitt EL, Lohse AW, Manns MP et al. Treatment challenges and investigational opportunities in autoimmune hepatitis. Hepatology 2005; 41:207-15. |
|6||Roberts S, Therneau T, Czaja A. Prognosis of histological cirrhosis in type 1 autoimmune hepatitis. Gastroenterology 1996; 110:848-57.|
|7||Feld J, Dinh H, Arenovich T, Marcus V, Wanless I, Heathcote J. Autoimmune hepatitis: Effect of symptoms and cirrhosis on natural history and outcome. Hepatology 2005; 42:53-62.|
|8||Devlin S, Swain M, Urbanski S, Burak K. Mycophenolate mofetil for the treatment of autoimmune hepatitis in patients refractory to standard therapy. Can J Gastroenterol 2004; 18:321-6.|
|9||Richardson P, James P, Ryder S. Mycophenolate mofetil for maintenance of remission in autoimmune hepatitis in patients resistant to or intelerant of azathioprine. J Hepatol 2000; 33:371-5.|
|10||Czaja A, Davis G, Ludwig J, Taswell H. Complete resolution of inflammatory activity following corticosteroid treatment of HBsAg-negative chronic active hepatitis. Hepatology 1984; 4:622-7.|
|11||Wai C, Greenson J, Fontana R, Kalbfleisch J, Marrero J, Conjeevaram H et al. A simple non-invasive index can predict both significant fibrosis and cirrhosis in patinets with chronic hepatitis C. Hepatology 2003; 38:518-26.|
|12||Fontana R, Lok A. Noninvasive monitoring of patients with chronic hepatitis C. Hepatology 2002; 36:S57-64.|
|13||Abdo A, Al Swat K, Ahmad S. Validation of predictive models for advanced fibrosis in 239 Saudi patients undergoing liver biopsy for chronic hepatitis B and C. Saudi J Gastroenterol 2005; 2:A122-3.|