Development of autoimmune hepatitis in patients with typical primary biliary cirrhosis


  • Potential conflict of interest: Nothing to report.


Primary biliary cirrhosis (PBC)–autoimmune hepatitis (AIH) overlap syndrome is a clinical entity characterized by the occurence of both conditions at the same time in the same patient. In addition to PBC-AIH overlap syndrome, transitions from one autoimmune disease to another have been reported, but no systematic series have been published. We report a series of 12 patients with consecutive occurrence of PBC and AIH (i.e., PBC followed by AIH). Among 282 PBC patients, 39 were identified who fulfilled criteria for probable or definitive AIH. AIH developed in 12 patients (4.3%). The baseline characteristics of the patients were similar to those of patients with classical PBC. Time elapsed between the diagnosis of PBC and the diagnosis of AIH varied from 6 months to 13 years. Patients with multiple flares of hepatitis at the time of diagnosis of AIH had cirrhosis on liver biopsy. Ten patients were given prednisone ± azathioprine; short-term as well as sustained remissions were obtained in 8 of these, while two had multiple relapses and eventually died 8 and 7 years after diagnosis of AIH. In conclusion, the development of superimposed AIH could not be predicted from baseline characteristics and initial response to UDCA therapy. If not detected early, superimposed AIH can result in rapid progression toward cirrhosis and liver failure in PBC patients. (HEPATOLOGY 2006;44:85–90.)

After the identification of the hepatitis C virus in 1989, it became evident that chronic hepatitis C was the most frequent inflammatory liver disorder associated with autoantibodies. A better definition and delineation of the spectrum of autoimmune liver diseases—in particular autoimmune hepatitis (AIH)—was subsequently achieved. Several variants of AIH as well as of primary biliary cirrhosis (PBC) and primary sclerosing cirrhosis were thus identified, confirming early reports describing cases of chronic inflammatory liver diseases exhibiting both cholestatic and hepatitic features. During the last decade, the so-called “PBC-AIH overlap syndrome” has been defined as the concurrent main characteristics of the two conditions at the same time in the same patient.1–3 Its prevalence in PBC patients has been estimated to be approximately 10%.4 Whether it constitutes a distinct entity or a variant of AIH or PBC remains unclear. In addition to PBC-AIH overlap syndrome, there have been some reports of patients with typical features of PBC or AIH who switched from one disease to another over time.4–10 However, no systematic series have been published. We reviewed all the patients referred to our center with the diagnosis of typical PBC that evolved into AIH. The clinical course and outcome of these patients are described herein. Development of AIH occurred in 4.3% of our PBC patients and was unpredictable. Importantly, our cases emphasize the possible role of AIH in the deterioration of liver function in PBC patients unless diagnosis is recognized early and steroid therapy is administered at the right time. Furthermore, analysis of the cases suggest that these patients may have two coincident autoimmune diseases rather than a variant of PBC or AIH.


PBC, primary biliary cirrhosis; AIH, autoimmune hepatitis; UDCA, ursodeoxycholic acid; SMA, smooth muscle antibody.

Patients and Methods

Study Population

We reviewed the database of PBC patients attending the Service d'hépatologie, Hôpital Saint-Antoine, AP-Hopitaux de Paris, Faculté PM Curie, Paris, since 1990. Up to the end of 2004, 39 of 282 PBC patients fulfilled the criteria for probable or definite AIH as defined by the International Autoimmune Hepatitis Group11 as well as our own criteria of PBC-AIH overlap syndrome.4 Twenty-two patients presented with the features of both AIH and PBC simultaneously at diagnosis. Five patients with an initial diagnosis of AIH developed typical PBC during their follow-up. AIH occurred in 12 patients with an initial diagnosis of typical PBC while they were being treated with ursodeoxycholic acid (UDCA).

Clinical and Laboratory Assessments

Data from clinical histories, laboratory examinations, and other diagnostic investigations were obtained at initial and follow-up visits after initiation of therapy. None of the patients exhibited excessive alcohol consumption (≥20 g/d); there was no evidence of exposure to hepato- or cholangio-toxic drugs or bile duct obstruction. All patients were negative for exposure to hepatitis A and hepatitis B virus at diagnosis of PBC and AIH. At the time of hepatitis flares, some patients had additional comprehensive viral investigations (e.g., Epstein-Barr virus, cytomegalovirus, Herpes virus) that were negative.


Antimitochondrial antibodies, anti–smooth muscle antibodies (SMA), antinuclear antibodies, and liver–kidney microsomal antibodies were detected by indirect immunofluorescence on rat tissue sections. Positivity was defined by a titer >1/80. Antimitochondrial antibody–negative samples were tested via immunoblotting against liver mitochondrial antigens as previously described.4 Immunofluorescence on colchicin-treated Hep-2 cell monolayers was used to detect anti-actin antibodies in SMA-positive samples. Soluble liver antigen/liver–pancreas antibodies were determined via inhibition capture enzyme-linked immunosorbent assay.

Liver Biopsy Specimens.

Liver biopsy specimens were fixed in Bouin's liquid and embedded in paraffin, and sections were stained with hematoxylin-eosin and Masson's trichrome. Classical histological features of PBC were assessed, and ductopenia was defined by the absence of bile ducts in at least 50% of the portal tracts. Activity of chronic hepatitis was assessed by the intensity of periportal and lobular necroinflammatory lesions. Fibrosis and activity were graded as previously described.4, 12 Briefly, the stage of fibrosis was assessed on a 5-point scale: 0= no fibrosis; 1= portal fibrosis without septa; 2= few septa; 3= numerous septa without cirrhosis; 4= cirrhosis. The grade of activity was assessed on a 4-point scale: A0= no histological activity; A1= mild activity; A2= moderate activity; A3= severe activity.4, 12


Frequency and Baseline Characteristics at the Time of Diagnosis of PBC and Response to UDCA.

Among 282 patients with a diagnosis of PBC, 39 (13.8%) fulfilled the criteria of simultaneous or sequential PBC-AIH overlap syndrome (see Appendix). Among these, 12 (4.3%) developed AIH during follow-up for PBC, 22 (7.8 %) presented with the features of both PBC and AIH simultaneously, and 5 (1.7 %) with AIH developed PBC during their follow-up.

The baseline characteristics of these patients at the time of PBC diagnosis are shown in Table 1. Data from 78 patients with “pure” PBC matched for sex and age with the 39 patients having both PBC and AIH are also provided. All the baseline clinical, biochemical, and histopathological features of the 12 patients who developed AIH during the follow-up of PBC were those usually encountered in classical PBC. Except for positivity of SMA antibodies and severity of the interface hepatitis, there were no features suggesting occult AIH. Liver–kidney microsomal antibodies were always absent. All the patients were given UDCA (12-15 mg/kg/d). A significant short-term response to UDCA (i.e., a >50% decline in alkaline phosphatase and aminotransferase levels) was obtained in 8 patients. One patient did not respond at all to UDCA therapy.

Table 1. Baseline Demographic, Clinical, and Laboratory Features at Diagnosis of PBC in Patients With PBC and AIH Either Simultaneously or Consecutively and PBC Without AIH
Patient CharacteristicsPBC and AIH Consecutively (n = 12)PBC-AIH Simultaneously (n = 22)AIH and PBC Consecutively (n = 5)PBC Without AIH (n = 78)
  1. NOTE. Normal serum bilirubin level: ≤17 μmol/L.

  2. Abbreviations: ULN, upper limit of normal; ALT, alanine aminotransferase; IgG, immunoglobulin G; AMA, antimitochondrial antibodies; ANA, antinuclear antibodies; ASMA, anti–smooth muscle antibodies.

Age in years, mean (range)46 (23–65)43 (21–71)38 (22–62)53 (23–72)
Pruritus or fatigue33%55%0%55%
Associated autoimmune disorders16%18%20%24%
Serum bilirubin (μmol/L), mean (range)14 (7–23)25 (8–170)10 (7–16)19 (5–87)
Alkaline phosphatase (ULN), mean (range)2.6 (1.9–6.1)3.2 (0.8–21)1.4 (1–2.2)4.0 (0.6–21)
ALT (ULN), mean (range)2.1 (1.1–3.7)5.8 (3.1–16)1.2 (0.6–1.6)2.7 (0.5–12)
IgG (ULN), mean (range)1.2 (0.9–1.8)1.6 (0.9–2.5)1.2 (0.8–1.4)1.2 (0.6–1.4)
AMA (>1/80e)100%95%100%95%
ANA (>1/80e)67%77%75%67%
ASMA (>1/80e)25%68%75%17%
Liver histology    
  Absent or portal and periportal septa66%59%75%71%
  Numerous septa or cirrhosis34%41%25%29%
 Interlobular bile duct lesions    
  Paucity (>50%)25%18%0%31%
 Lymphocytic piecemeal necrosis    
  Absent or mild58%9%75%78%

Development of AIH.

Five patients had one flare of hepatitis leading to a diagnosis of AIH. Seven patients had 2 or more flares of hepatitis before the diagnosis was established. In these 7 patients, liver biopsy obtained at the time of diagnosis of AIH demonstrated extensive fibrosis or cirrhosis. The amount of time between diagnosis of PBC and diagnosis of AIH varied from 6 months to 13 years. Eight patients were SMA-positive, all of the patients were antiactin-positive, and 4 patients were SMA-negative at the time of AIH diagnosis. When these 4 patients were tested for soluble liver antigen/liver–pancreas, 3 were positive. All of the patients remained antimitochondrial antibody–positive. All of the patients had marked liver inflammation on the liver biopsy specimen obtained at the time of diagnosis of AIH. Five patients had unequivocal bridging necrosis.

Outcome and Results of Steroid and UDCA Therapies.

In addition to UDCA treatment, 10 patients (Tables 2 and 3) were given conventional initial steroid therapy (0.5 mg/kg/d) and then maintenance therapy consisting of prednisone ± azathioprine (10-15 mg/d ± 1.5 mg/kg/d). All the patients had short-term biochemical remission (ALT level ≤2 upper limit of normal; serum bilirubin < 20 μmol/L) after introduction of prednisone therapy. Two patients with spontaneous and sustained remission under UDCA therapy were not given corticosteroids. Two patients had multiple relapses during the maintenance treatment (prednisone and azathioprine). They eventually died 12 and 20 years after diagnosis of PBC and 8 and 7 years after diagnosis of AIH. Both had liver failure and exhibited hepatocellular carcinoma at necropsy. By comparison, 9 of the 242 (3.7 %) with only PBC and 2 of the 22 patients (9%) with simultaneous PBC-AIH overlap developed hepatocellular carcinoma during their follow-up. In all the patients except 2, PBC remained asymptomatic with biochemical indices (alkaline phosphatase, serum bilirubin) within near normal range. Until the time of the last visit, liver biopsy specimens were available 2 to 4 years after the diagnosis in five patients. Despite frank improvement in liver inflammation and fibrosis, none of the patients had normal liver architecture or were free of inflammation.

Table 2. Details of the 12 Patients With the Diagnosis of PBC and AIH Consecutively
PatientAge at and Year of Diagnosis of PBCAge at Diagnosis of AIHNumber of Hepatitis Flare-ups Before Diagnosis of AIHBiochemistries at Diagnosis of AIHAMA/SMA/ANA at Diagnosis of AIHLiver Histology at Diagnosis of AIHAIH Score*
S. bilirubin (μmol/L)ALT (× N)
  • Abbreviations: AMA, antimitochondrial antibodies; ANA, antinuclear antibodies; ALT, alanine aminotransferase; × N, times the upper limit of normal value; CH, bile duct inflammation and destruction.

  • *

    According to Alvarez et al.11

135 (1992)4112815+/+/+CH + A3 F113
248 (1992)4911824+/−/+CH + A3 F214
342 (1995)4712512+/+/+CH + A3 F118
454 (1999)5512212+/+/−CH + A2 F216
550 (1996)5237014+/+/+CH + A3 F416
623 (1997)2444210+/+/−A3 F419
766 (1986)70325025+/+/+A3 F423
850 (1981)63310815+/+/+A3 F423
942 (1996)4526217+/−/+A2 F422
1058 (1991)67218070+/+/−A3 F320
1137 (1995)43321040+/−/+CH + A3 F418
1248 (1999)4913715+/−/−CH + A3 F219
Table 3. Outcome and Results of Immunosuppressive Therapy
PatientPrednisonePrednisone + AzathioprineRemission/RelapseSustained RemissionDeath/Alive
  1. Abbreviation: HCC, hepatocellular carcinoma.

5+++/−+Alive, reversal of cirrhosis
7+++/+Died (liver failure, HCC, 1998)
8+++/+Died (liver failure, HCC, 2001)


It is now recognized that simultaneous or consecutive occurrence of AIH and PBC in the same patient is not a fortuitous event. A review of the literature via Medline revealed 21 case reports and 11 reviews dealing with this topic since the first series was published in 1998.4, 7, 13 Our data show that 14% of our population of PBC patients exhibited the key features of both AIH and PBC either simultaneously or consecutively. This prevalence falls within the range (2%-20%) reported in case series published since 1998.14–18 To date, consecutive forms of PBC-AIH overlap syndrome are thought to be rare. However, our data suggest that this variant is not uncommon, because 4% of our PBC patients developed superimposed AIH. Bias in patient referral could be an explanation. However, we believe that this event in the course of PBC could be easily missed if lifelong close monitoring is not performed. The occurrence of AIH is unpredictable, because it often occurs late after the initial diagnosis of PBC. One of the patients of the present series was referred for liver transplantation because of liver failure complicating long-standing PBC. The diagnosis of AIH might be questionable in 2 of our 12 patients, mainly because of “spontaneous” resolving features of AIH under UDCA therapy. Spontaneous recovery of a flare in patients with severe AIH is not uncommon.19 Furthermore, partial and complete remission of AIH could occur during UDCA therapy.20, 21 We reported patients having severe flares-up of hepatitis. It cannot be excluded that other PBC patients may also present with minor forms of AIH that, together with the cholestatic process, could contribute to the deterioration of liver function.

The baseline characteristics of the patients with an initial diagnosis of PBC were very similar to those of patients with classical PBC. However, it should be noted that 25% of them had SMA with antiactin specificity at the time of diagnosis of PBC, a frequency that tends to be higher than that found in classical PBC. However, the low number of patients precludes any firm statistical conclusion. The same remark could be made regarding the prevalence of the severity of portal inflammation and interface hepatitis, which again was more frequently found in this series of 12 patients than in our patients with classical PBC. Soluble liver antigen/liver–pancreas autoantibodies are thought to be good markers of AIH.22 It has been suggested that they may help to identify PBC patients with secondary AIH. Unfortunately, we were unable to screen our patients for these autoantibodies when the initial diagnosis of PBC was established, precluding its assessment as a predictor of secondary development of AIH.

AIH is a disease of unknown cause that occurs in persons with genetic predisposition. Lohse et al.13 found that the majority of their patients with PBC-AIH overlap syndrome have the characteristic HLA haplotypes of AIH, namely HLA-B8, DR3, and DR4. HLA typing was not available in the majority of our patients. Further studies are needed to determine if HLA typing might help to predict secondary AIH.

One of our patients had her first flare-up of hepatitis within 3 months following vaccination against hepatitis B virus. Ohba et al.9 reported a similar case of AIH flare-up that was presumably secondary to an immune rebound mechanism after a pregnancy.

A role for viruses in AIH has been repeatedly proposed; however, convincing evidence links only two viruses—hepatitis A and Epstein-Barr virus—to AIH in cases where a genetic predisposition exists.23, 24 We had no evidence for hepatitis A and Epstein-Barr virus infection as a triggering agent in our patients. The role of drugs is also well known as a possible trigger of AIH. None of our patients received drugs suspected of inducing AIH.

Whether steroid therapy is truly warranted in patients with PBC-AIH overlap syndrome has been questioned.15 The short-term observation of these patients who where included in UDCA trials for PBC had similar improvement in biochemistries under UDCA in patients with or without features of PBC-AIH overlap syndrome. Consistent with this observation, two of our patients had spontaneously resolving flare-up of AIH and, at the time of this writing, had normal biochemical liver tests. Nonetheless, because of the severity of hepatitis, we estimated that the conventional therapeutic regimen for AIH was mandatory. We believe that treatment is justified in all patients with flare-ups of aminotransferase if they fulfill the criteria for AIH, even in the presence of PBC. This notion is supported by a case of reversal of cirrhosis in a patient with PBC-AIH overlap syndrome treated with a combination of UDCA and prednisone.25

It remains unclear whether PBC-AIH overlap syndrome forms a distinct entity or a variant of PBC or AIH. The consecutive occurrence of the two diseases seen in our cases support the notion that the patients have two coincident autoimmune diseases. This is consistent with the fact that autoimmune diseases are associated with one another in 5% to 10% of cases.

In conclusion, this retrospective study shows that the development of AIH in long-standing PBC is not uncommon. Because of the severity of AIH and the good response to steroid therapy shown in this series, this unusual condition should not be misdiagnosed.


Table  . Revised Scoring System for Diagnosis of AIH According to Alvarez et al.11
  1. NOTE. Diagnosis of PBC-AIH according to our own criteria is as follows. PBC criteria: (1) serum alkaline phosphatase levels at least 2 times the upper limit of normal or serum γ-glutamyltransferase levels at least 5 times the upper of limit of normal; (2) positive test for antimitochondrial antibodies (>1/80); and (3) a liver biopsy specimen showing florid bile duct lesions. AIH criteria: (1) serum alanine aminotransferase levels at least 5 times the upper limit of normal; (2) serum IgG levels at least 1.5 times the upper limit of normal or a positive test for anti–smooth muscle antibodies (>1/80e); and (3) a liver biopsy showing moderate or severe periportal or periseptal lymphocytic piecemeal necrosis.

  2. Abbreviations: ALP, alkaline phosphatase; AST, aspartate aminotransferase; ALT, alanine aminotransferase; IgG, immunoglobulin G; ANA, antinuclear antibodies; LKM, liver–kidney microsomal; AMA, antimitochondrial antibodies.

ALP:AST (or ALT) ratio 
Serum globulins or IgG above normal 
ANA, SMA, or LKM-1 
Hepatitis viral markers 
Drug history 
Average alcohol intake 
 <25 g/d+2
 >60 g/d−2
Liver histology 
 Interface hepatitis+3
 Predominantly lymphoplasmacytic infiltrate+1
 Resetting of liver cells+1
 None of the above−5
 Biliary changes−3
Other autoimmune disease(s)+2
Optional additional parameters 
 Seropositivity for other defined autoantibodies+2
 HLA, DR3, or DR4+1
Response to therapy 
Interpretation of aggregate scores 
  Definite AIH>15
  Probable AIH10–15
  Definite AIH>17
  Probable AIH12–17