If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Amsterdam UMC, Vrije Universiteit Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology and Metabolism, de Boelelaan 1117, Amsterdam, the Netherlands
Amsterdam UMC, Vrije Universiteit Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology and Metabolism, de Boelelaan 1117, Amsterdam, the NetherlandsDepartment of Gastroenterology and Hepatology, Radboudumc, Nijmegen, the Netherlands
Amsterdam UMC, Vrije Universiteit Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology and Metabolism, de Boelelaan 1117, Amsterdam, the Netherlands
Amsterdam UMC, Vrije Universiteit Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology and Metabolism, de Boelelaan 1117, Amsterdam, the Netherlands
Amsterdam UMC, Vrije Universiteit Amsterdam, Gastroenterology and Hepatology, Amsterdam Gastroenterology Endocrinology and Metabolism, de Boelelaan 1117, Amsterdam, the Netherlands
Recommendations for drug withdrawal in patients with autoimmune hepatitis (AIH) in longstanding remission are conflicting and rely on retrospective data. We prospectively investigated the predictive value of histological normalisation for successful treatment withdrawal in AIH patients.
Methods
Non-cirrhotic patients with established AIH and complete biochemical remission (normalisation of serum alanine aminotransferase [ALT] or aspartate aminotransferase [AST] and immunoglobulin G [IgG]) of at least 2 years were biopsied. Immunosuppressive therapy was only withdrawn in patients with histological normalisation (histological activity index [HAI] ≤3) with a minimum follow-up of 12 months.
Results
A total of 17 patients in biochemical remission for at least 2 years were included. Persistent histological inflammatory activity (HAI >3) precluded drug withdrawal in five patients. These had higher values of ALT (25 vs. 16 U/L; p = 0.01) and AST (26 vs. 22 U/L; p = 0.01) compared with patients in histological remission. Immunosuppressive medication was withdrawn in 12 patients; eight (67%, C.I. 40-93% p = 0.4) remained in remission during a median follow-up of 62 months (range: 13-75 months); and four (33%, C.I. 7-60% p = 0.4) required reinstitution of therapy after 1, 6, 11, and 40 months, all without clinical signs of disease progression or hepatic decompensation. No predictors of relapse were identified.
Conclusion
Two-thirds of the patients who prove to have histological normalisation after at least 2 years of biochemical remission achieve treatment-free remission. Although patient numbers were small and results should be interpreted with caution, these findings support a liver biopsy prior to drug withdrawal.
Long-term immunosuppressive therapy increases the risk of treatment-related side effects impacting quality of life in patients with autoimmune hepatitis (AIH). [
] while the vast majority of patients require long-term if not life-long maintenance therapy to prevent disease progression to cirrhosis and/or decompensation. [
] In fact, relapses after treatment withdrawal occur in 25-89% of patients, mostly within 12 months, yet relapses after decades have been documented. [
Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy.
] Both the European Association for the Study of the Liver (EASL) and American Association for the Study of Liver Diseases (AASLD) guidelines recommend repeatedly normal levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and immunoglobulin G (IgG) for at least 24 months before treatment withdrawal. [
Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American association for the study of liver diseases.
] In addition, a liver biopsy prior to treatment withdrawal is recommended in both guidelines, yet the recent AASLD guideline states that a biopsy does not seem to be necessary in all adults. [
Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American association for the study of liver diseases.
] When tested in clinical practice, the application of strategies without liver biopsy before withdrawal results in relapse in almost 50% of patients. [
] Nevertheless, other studies have reported relapse rates of 70-79% after treatment withdrawal, even among patients with AIH with established histological remission. [
Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy.
] Available data are contradictory and rely on retrospective observations. No studies have prospectively evaluated the guideline recommendation to perform a liver biopsy prior to treatment withdrawal. [
Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American association for the study of liver diseases.
In this prospective study, we investigated the success rate after treatment withdrawal in patients with AIH after stringent selection based both on biochemical and histological remission.
2. Patients and methods
2.1 Study population
Patients were identified and recruited by gastroenterologists and hepatologists in hospitals that collaborate with the Dutch Autoimmune Hepatitis Study Group. Patients were included from two general and two tertiary referral hospitals in the Netherlands. All patients provided written informed consent prior to participating in this study. The protocol (number 2012/035) was approved by the Institutional Review Board of the VU University Medical Center. We included patients with a definite or probable diagnosis of AIH, [
] who were in complete biochemical remission on monotherapy for at least 2 years, wished to discontinue treatment, and were willing to undergo a liver biopsy. Patients with cirrhosis (F=6) at the time of diagnosis were excluded. Viral, alcoholic, and drug-induced hepatitis had been excluded in all patients. Due to a high INR level, a diagnostic biopsy was not performed in one patient. Complete biochemical remission was defined as complete normalisation of both serum ALT or AST and IgG. Data on anti-nuclear (ANA) and/or anti-smooth muscle (SMA) antibodies were collected. Anti-liver/kidney microsomal antibodies type 1 (LKM-1) and IgG were not routinely performed in all centers. In some patients, one of both transaminases were measured.
2.2 Withdrawal and primary outcome
All patients were subjected to a liver biopsy before withdrawal; immunosuppressive therapy was only withdrawn in patients with histological normalisation/remission defined as a HAI ≤3 and without progression to fibrosis stage 5 or 6 between diagnosis and treatment. [
] The primary outcome was sustained complete biochemical remission without immunosuppressive treatment after a follow-up of at least 12 months. The secondary outcomes were relapse defined as serum ALT three times the upper limit of normal (ULN) and/or raised serum IgG levels over 2g/dl. Loss of remission (LOR) was defined by an increase in serum ALT levels above ULN on at least three occasions with an interval of 3 weeks and reinstitution of drug therapy. Hereinafter, we use the term ‘relapse’ for both LOR and relapse.
2.3 Liver histology
Liver biopsies were revised and scored according to the Ishak system by an expert liver pathologist (E.B.), who was blinded to clinical information (Table 1). [
] The HAI was the sum of grades for portal inflammation (range: 0-4), periportal or periseptal interface hepatitis (range: 0-4), focal inflammation (range: 0-4) and confluent necrosis (range: 0-6). Fibrosis was staged from 0-6. The presence of plasma cells in the portal and intra-acinar areas, presence of steatosis, and sinusoidal dilatation were assessed on a semi-quantitative scale (0-3) (Table 1). The presence of bile duct injury, ductopenia, sclerosis, rosetting of liver cells, and emperipolesis was assessed on a dichotomous scale (Table 1). Biopsies were also evaluated for signs of nodular regenerative hyperplasia with specific attention paid to signs of circulatory disturbances, including nodularity, phlebosclerosis, and sinusoidal dilatation. Patients who were in histological remission (defined as no active inflammation; HAI ≤3/18) started with a withdrawal schedule. For patients with inflammatory activity in the biopsy (HAI >3/18), therapy was continued.
Table 1Scoring sheet for inflammation and fibrosis as well as histopathological features Inflammation and fibrosis were assessed with the Ishak score.
Normally distributed variables are denoted as mean with standard deviation (S.D.), whereas non-normally distributed variables are denoted as median with range (min-max) if stated. Categorical data are denoted as number and percentage. The follow-up time was defined as months from date of treatment withdrawal to the occurrence of relapse/LOR or the end of the study (April 1st, 2019) if no LOR or relapse occurred. For differences in relapse between two or more groups, we used the Log-rank test. Ordinal and continuous variables were compared between the two groups with the Mann-Whitney test (non-normal). Proportional differences between two or more groups were tested with the X2-test or Fisher's exact test. Differences between the observed and hypothesized proportion (hypothesized proportion set at 0,5) of binary outcomes (remission or relapse) within one group were tested with the z-test and reported with 95% confidence interval (C.I.). The level of significance was set at p < 0.05. All statistical analyses were performed with Statistical Package for Social Sciences, version 22.0.
3. Results
3.1 Patient characteristics
Patient characteristics at the time of diagnosis of AIH are presented in Table 2. All patients had a minimum diagnostic post-treatment AIH score of 12 points, indicating a diagnosis of AIH. [
A total of 18 patients were biopsied. One patient was excluded for analysis due to a failed biopsy; thus, a total of 17 patients were included for analysis (Fig. 1). Five patients (29%) were not eligible for withdrawal due to persistent active inflammation (HAI >3) and continued therapy. One of these patients had progressed from no fibrosis (stage 0) at diagnosis to incomplete cirrhosis (stage 5) with a HAI of 10 at evaluation biopsy. Immunosuppressive therapy was withdrawn in the 12 patients (71%) with histological remission. Patients who were in histological remission had significantly lower median levels of ALT (16 vs. 25 U/L; p = 0.01) and AST (22 vs. 26 U/L; p = 0.01) compared with patients who were not eligible for withdrawal (Table 3). IgG levels were similar between patients with and without histological remission (12.2 vs. 13.7 g/dL; p = 0.15).
Fig. 1Flow chart and outcome of included patients.
A total of eight (67%, C.I. 40-93% p = 0.4) patients remained in treatment-free remission during a median follow-up of 62 months (range: 13-75) (Fig. 2). A relapse occurred in four patients (33% C.I. 7-60%, p = 0.4) after 1, 6, 11, and 40 months after discontinuation of therapy with a median ALT of 152 U/L (range: 63-208). A brief summary of treatment-free remission seen in other studies is presented in Table 4.
Fig. 2Relapse-free survival after treatment withdrawal.
Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy.
Remission group 47 ± 16; relapse group 40 ± 14. 32% cirrhosis
Repeatedly normal ALT, AST and IgG
35 (88%) in histological remission
Duration of treatment: 60 (24–132) months
Mycophenolate mofetil (n = 40)
75%, n = 30, no adverse outcomes reported
5 (2–24) months
Relapse: AST and ALT >3x ULN and/or IgG > 2g/dL; loss of remission: AST, ALT and IgG >1x ULN
Shorter duration of Mycophenolate mofetil treatment, lower ALT levels at baseline, no normalisation of IgG at 6 months, no significant improvement of inflammatory activity with at least stable fibrosis in liver biopsy before drug withdrawal
The first patient who relapsed after 40 months was retreated with prednisone induction therapy and achieved complete biochemical remission after 28 days. Prednisone was fully tapered and the patient was eventually treated with 100 mg of azathioprine, the same regimen as prior to withdrawal. The second patient relapsed after 6 months and was retreated with prednisone induction therapy, achieving complete biochemical remission after 21 days. Histology at the time of withdrawal indicated severe sinusoidal dilatation, which has been described in relation to thiopurine use; hence, this patient was not restarted on thiopurine therapy but received prednisone maintenance therapy. [
] The third patient relapsed 11 months after the withdrawal of tioguanine therapy. In this patient, complete biochemical remission was both achieved (after 12 days) and maintained with budesonide. The last patient was retreated with prednisolone and achieved remission after 133 days. This patient underwent prophylactic mastectomy during retreatment, which may have led to a longer time to normalisation of transaminases. One patient developed arthralgia arising after induction therapy, but no other adverse effects were reported. None of the patients exhibited signs or had clinical symptoms of disease progression or hepatic decompensation.
3.5 Predictors of relapse
We could not identify significant differences at the time of diagnosis or withdrawal between patients with treatment-free remission and those who had a relapse (Table 3). In addition, there were no histological features predictive of relapse (Table 1). Noteworthily, liver cell rosettes were only observed in patients who were not eligible for withdrawal (n = 2) or those who relapsed after withdrawal (n = 2).
3.6 Nodular regenerative hyperplasia
In 17 patients with AIH, one had severe sinusoidal dilation at the time of withdrawal. This patient had been treated with azathioprine (75 mg/day) during the 58 months before withdrawal. A follow-up biopsy 6 years after retreatment with prednisone was consistent with nodular regenerative hyperplasia. The other biopsies (n = 16) did not reveal signs of nodular regenerative hyperplasia in the treatment evaluation biopsy after treatment with azathioprine (n = 14) or mercaptopurine (n = 1), or tioguanine (n = 1) for a median of 5 years (range 2-17).
3.7 Results in patients excluded for analysis
The patient who was excluded due to a failed biopsy chose to withdraw from immunosuppressive therapy. After 16 months of follow-up, this patient had sustained complete biochemical remission.
4. Discussion
We found that 67% of patients with AIH with >2 years biochemical (ALAT or ASAT and IgG <1ULN) and histological remission (HAI ≤3) remained in remission at least 12 months after the withdrawal of immunosuppression. This proportion is among the highest treatment-free success rates documented. [
Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy.
Previously, we reported a rate of relapse or LOR of 89% in 131 patients with AIH who were withdrawn from immunosuppressive therapy after complete biochemical remission for 2 or more years. [
] Liver biopsy confirmed remission in 24 patients (18%). The majority of these patients were on combination therapy at the start of drug withdrawal (n = 15), or had histologically proven cirrhosis (n = 1), which were exclusion criteria in our cohort. The remaining eight patients where on monotherapy without cirrhosis. Two of these patients remained in treatment-free remission until the last follow-up, 98 and 110 months. Interestingly, in another five patients of the remaining eight patients, drug withdrawal was attempted after a history of one or two relapses during prior withdrawal attempts. When patients have relapses after drug withdrawal, subsequent attempts of drug withdrawal treatment are expected to fail, and are therefore not recommended. In addition, patients who experience more than one relapse, are at an increased risk of death from liver failure or progression to cirrhosis. [
] Even though a history of one or more relapses was not a strict exclusion criterium in our study, no relapses have been observed prior to inclusion. Summarizing, only three of the 24 patients in our study would be suitable for withdrawal, which explains the discrepancy between the current study and the previous retrospective study. A total of two (68%) patients remained in treatment-free remission, which is similar to the success percentage in our prospective study. Since that study, we have prospectively biopsied patients on monotherapy to improve patient selection before withdrawal.
Guidelines recommend reconsidering the diagnosis of AIH in patients who do not respond to therapy. Similarly, in patients who do not relapse after therapy withdrawal, one might question the diagnosis of AIH. In our cohort, approximately half of the patients were seronegative for both ANA and SMA autoantibodies, which is more frequent than the 20% reported in recent guidelines. [
Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American association for the study of liver diseases.
] In addition, all patients had an initial response to corticosteroid therapy, consistent with a clinical diagnosis of AIH.
Five of 17 (29%) patients could not be withdrawn from medication due to ongoing histological activity despite more than 2 years of complete biochemical remission. This is in line with findings of Dhaliwal et al., [
] who found that a significant proportion (55 out of 120; 46%) of patients in biochemical remission were not in histological remission. By contrast, two studies performed by the same group revealed that relevant histological disease activity was not quite prevalent (4 out of 22 patients; 18%, and 2 out of 13 patients; 15%) in patients in long-term remission. [
] Despite the wide range of percentages of patients with ongoing histological activity identified in these studies, these patients are at increased risk of progression to cirrhosis and require intensified therapy. In our cohort, one patient progressed from no fibrosis to incomplete cirrhosis with an HAI of 10 at the evaluation biopsy, despite long-term biochemical remission. This emphasises the need for a liver biopsy as well as additional or alternative drug treatment rather than a trial of drug withdrawal in some patients.
Patients who were in histological remission had significantly lower transaminases compared with those patients with ongoing histological activity despite complete biochemical remission. This is in line with findings of Dhaliwal et al. [
] Hence, interpreting the transaminase level within the normal range might aid in the identification of patients who are in complete biochemical but not histological remission. Our study and another study [
], both elevated serum transaminases and IgG predicted histological activity (HAI >3). The negative predictive value of these serum parameters was 21-33%. Hartl et al. found variations within the normal range of ALT and IgG levels in patients who relapsed, and advised that the aim of serum ALT should be less than half the ULN as well as IgG levels not higher than 12 g/L. [
] It is reassuring that in this small group of patients, those who lost remission achieved complete biochemical remission quickly after the re-institution of therapy without adverse events. We found no features that predicted a LOR after withdrawal of therapy, contrary to other studies. [
] Factors that have been associated with relapse are an unknown (viral or drug) precipitant of AIH, a longer time to remission, combination therapy, and a high AIH diagnostic score (1999). [
] Multiple plausible explanations exist for this discrepancy with the literature, including the following: first, we applied a highly stringent threshold to identify patients who were eligible for withdrawal; and second, our study was not designed nor powered to identify such a difference.
A recent review found that the severity of histological activity at the time of diagnosis was not associated with a relapse after treatment withdrawal. [
] In our cohort, 5 out of 17 patients exhibited ongoing histological activity, and one of these patients progressed from no fibrosis at diagnosis to incomplete cirrhosis at the follow-up biopsy (with an HAI of 10). Interestingly, in our cohort, rosettes in the follow-up biopsy were only found in those patients who could either not be withdrawn due to continued histological activity or who relapsed after withdrawal. However, this difference failed to reach statistical significance.
Future research into predictors of outcomes after withdrawal may need to focus on other parameters. For example, using serial measurements of transient elastography as a marker of fibrosis progression, could be a more valuable medium-term endpoint than a relapse or biochemical parameters. [
A limitation of the study is the sample size, it was not designed to compare groups of patients with or without a biopsy prior to withdrawal. Therefore, we were unable to determine whether we prevented relapses or clinical deterioration. However, in our opinion, such a randomised study design would not be ethical because prior studies have suggested the benefit of a biopsy prior to the withdrawal of immunosuppressive treatment. [
] In addition, we identified five patients who required an intensification of therapy instead of a withdrawal attempt. Hence, in our study a liver biopsy prior to drug withdrawal improved the selection of patients who could maintain remission without immunosuppressive therapy, and also revealed ongoing histological inflammation in 29% of patients and progression to cirrhosis in one patient.
The results of our study confirm that a strict selection of the patients with AIH who want to withdraw medication is necessary. First, patients were in complete biochemical remission for at least 2 years. Second, we only included patients who were on monotherapy since previous studies conclude that sustained remission on monotherapy can improve the success rate. [
] Third, we excluded patients with advanced liver disease to minimize the risk of clinical deterioration towards acute hepatic failure. Applying our selection criteria, we are aware that only a highly selected population are eligible for an attempt of drug withdrawal.
In this prospective study, we demonstrated that a small, properly selected group of patients with AIH without cirrhosis at diagnosis can benefit from treatment withdrawal. Taken together, we propose the discontinuation of treatment only after long-term (≥2 years) complete biochemical remission on monotherapy in patients without advanced liver disease and with histological remission (HAI ≤3).
Guarantor or the article
Prof. dr. G. Bouma.
Specific author contributions
GB had the original idea and supervised the study. RS, FB, YB, BV, CN, EB, SK, JD collected the data. RS and FB analyzed the data. FB and RS wrote the manuscript. All authors critically reviewed the manuscript and approved the final version of the manuscript.
Financial support
None.
Declaration of Competing Interest
No conflict of interest.
References
Janik M.K.
Wunsch E.
Raszeja-Wyszomirska J.
et al.
Autoimmune hepatitis exerts a profound, negative effect on health-related quality of life: a prospective, single-centre study.
Factors predicting relapse and poor outcome in type I autoimmune hepatitis: role of cirrhosis development, patterns of transaminases during remission and plasma cell activity in the liver biopsy.
Diagnosis and management of autoimmune hepatitis in adults and children: 2019 practice guidance and guidelines from the American association for the study of liver diseases.
00088-1&id=gr1.jpg){kind=link}
00088-1&id=gr2.jpg){kind=link}