Autoimmune Liver Disease — Clinical Case

DCE Long Case

Patient profile

Mrs K is a 54-year-old woman who presents with a 12-month history of progressive fatigue and pruritus, and a 3-month history of painless jaundice and dark urine. She has hypothyroidism but no other significant past medical history. [1]

Presenting concern: Over the past year she has noticed worsening tiredness (she now needs a nap most afternoons), and generalised itch that is worse at night and spares her face. Over the past three months she has noticed yellowing of her eyes and skin, dark tea-coloured urine, and pale stools. She has lost 4 kg. She presented to her GP, who found abnormal liver function tests and referred her to the hepatology clinic. [1]

Past medical history: Hypothyroidism for 8 years (on thyroxine 100 micrograms daily). No prior liver disease. No surgeries. Menopause at 50. No history of inflammatory bowel disease. [1]

Medications: Thyroxine 100 micrograms daily. No herbal supplements, no over-the-counter medications, no recent antibiotics, no recreational drugs. Occasional paracetamol. Does not drink alcohol (never a heavy drinker). No blood transfusions, no tattoos, no recent travel. [1]

Family history: Mother with rheumatoid arthritis and hypothyroidism; a maternal aunt with "liver trouble" (details unclear). No known family history of PBC, PSC or AIH. [1]

Social: Married, two adult children. Works part-time as a teacher. Non-smoker, minimal alcohol. [1]

Examination:

• Icteric sclerae and diffuse jaundice. Excoriations on the forearms and shins, with prurigo nodularis. Hyperpigmentation of the extensor surfaces.
• Bilateral xanthelasma around the eyes.
• Palmar erythema. Three small spider naevi on the anterior chest.
• Pulse 78 regular. Blood pressure 128/78. Temperature 36.8. Respiratory rate 16. SpO2 98 per cent room air.
• Abdomen: a firm, smooth liver edge palpable 3 cm below the costal margin; a palpable spleen tip. No ascites. No caput medusae.
• No asterixis. No fetor hepaticus. Alert and oriented. [1]

Investigations:

• ALP 520 U/L; gamma-GT 390 U/L; bilirubin 75 (conjugated 52); albumin 33 g/L; ALT 180 U/L; AST 160 U/L.
• INR 1.2. Platelets 145. Haemoglobin 118.
• IgM 4.8 g/L (raised, upper limit 2.8). IgG 28 g/L (raised, upper limit 16). IgA normal. IgG4 normal.
• AMA positive at 1:640. ANA positive at 1:5120 (homogeneous pattern). Anti-smooth muscle antibody positive at 1:160. Anti-LKM1 negative. Anti-SLA/LP not yet sent.
• HBsAg, anti-HBc, anti-HCV, anti-HEV all negative. HIV negative.
• Caeruloplasmin normal (Wilson excluded). Alpha-1-antitrypsin normal. Ferritin and transferrin saturation normal (haemochromatosis excluded).
• Thyroid function: TSH 3.2 (within range on her thyroxine).
• Ultrasound: normal biliary ducts, no gallstones, no focal liver lesion, patent portal vein, splenomegaly (14 cm).
• Liver biopsy: florid duct lesion AND interface hepatitis with a plasma-cell-rich infiltrate; fibrosis stage 3 (bridging fibrosis).
• FibroScan: 14 kPa (consistent with stage 4 fibrosis / cirrhosis). [1]

Candidate's opening statement (SASPOP)

"This is Mrs K, a 54-year-old teacher presenting with a 12-month history of fatigue and pruritus and a 3-month history of painless jaundice, found to have an AIH-PBC overlap syndrome. She meets the Paris criteria for both diseases: for PBC, a cholestatic LFT pattern with a positive AMA at 1:640, a raised IgM of 4.8, and a florid duct lesion on biopsy; and for AIH, a hepatocellular component with an ALT of 180, a raised IgG of 28, positive ANA and SMA, and interface hepatitis with a plasma-cell infiltrate. Her FibroScan of 14 kPa indicates established cirrhosis. Her main problems are the overlap syndrome, which needs a combined UDCA and immunosuppressive strategy; the cirrhosis with its surveillance requirements; the pruritus; the osteoporosis risk from both the cholestasis and the corticosteroids I will use; the fat-soluble vitamin deficiency likely present given her cholestasis; her hypothyroidism (a common associated autoimmune condition); and the psychosocial impact of a complex chronic diagnosis. My priorities are to start ursodeoxycholic acid for the PBC component, check the TPMT activity and then start prednisolone plus azathioprine for the AIH component, stage and surveil the cirrhosis, manage the pruritus and the bone health, and involve a multidisciplinary team." [1]

Structured problem list (numbered, prioritised)

1. AIH-PBC overlap syndrome — the central problem; needs combined UDCA and immunosuppression.
2. Cirrhosis (FibroScan 14 kPa, biopsy stage 3, splenomegaly) — requires variceal and HCC surveillance.
3. Pruritus with excoriations and prurigo nodularis — a major quality-of-life issue.
4. Osteoporosis risk — from the cholestasis, the menopause, and the corticosteroids to come.
5. Fat-soluble vitamin deficiency — likely given the cholestasis and the jaundice.
6. Hypothyroidism — a common associated autoimmune condition; ensure it is well replaced.
7. Anaemia (haemoglobin 118) — likely chronic disease; exclude gastrointestinal blood loss given the cirrhosis.
8. Psychosocial impact and communication — complex chronic diagnosis, lifelong therapy, work implications. [1]

Integrated management plan

Step 1 — Confirm the diagnosis and exclude mimics: [1]

The diagnosis of AIH-PBC overlap is secure on the Paris criteria. I have already excluded the mimics: viral hepatitis (negative serology), Wilson disease (normal caeruloplasmin), alpha-1-antitrypsin deficiency (normal), haemochromatosis (normal iron studies), drug-induced liver injury (no culprit drug), and IgG4-sclerosing cholangitis (normal IgG4, no pancreatic abnormality, the classic PBC features). I would send anti-SLA/LP to complete the AIH antibody panel — it is highly specific for AIH and identifies a relapse-prone subset [5]. No further diagnostic investigations are needed.

Step 2 — Ursodeoxycholic acid for the PBC component: [1]

Start UDCA at 13 to 15 mg/kg/day in divided doses — first-line for all stages of PBC. The 1994 Lindor NEJM trial showed it slows histological progression and reduces the need for transplant [2]. This is lifelong therapy. It is well tolerated; diarrhoea is the main side effect. I would assess the biochemical response at 6 to 12 months using the Barcelona or Paris response criteria (a normalisation or sufficient improvement in the ALP and bilirubin).

Step 3 — Immunosuppression for the AIH component: [1]

Check the thiopurine methyltransferase (TPMT) activity BEFORE starting azathioprine — TPMT deficiency causes fatal myelosuppression [5]. Once the TPMT is confirmed normal, start prednisolone 0.5 to 1 mg/kg/day (around 40 mg daily for a 60 kg woman) plus azathioprine 1 to 2 mg/kg/day (around 100 to 150 mg daily). Because she has established cirrhosis, I would use prednisolone rather than budesonide — budesonide is ineffective in cirrhosis because of portosystemic shunting. I would monitor the full blood count weekly for the first month, then monthly, then three-monthly; the liver function monthly; and the blood glucose and blood pressure at every visit. I would add bone protection (calcium, vitamin D, a DEXA, and a bisphosphonate as needed), a PPI for gastroprotection, and sun protection advice (azathioprine increases skin cancer risk).

Step 4 — Cirrhosis surveillance: [1]

Upper GI endoscopy for variceal surveillance at diagnosis and every one to three years depending on findings. Six-monthly ultrasound plus alpha-fetoprotein for hepatocellular carcinoma (her cirrhosis carries an HCC risk regardless of the underlying cause). A baseline and biennial DEXA for osteoporosis. Annual fat-soluble vitamin levels (A, D, E, K) given her cholestasis, with replacement as needed. [1]

Step 5 — Symptom control for the pruritus: [1]

Cholestyramine first-line (4 g before and after breakfast, given separately from UDCA by at least 2 to 4 hours so it does not bind the UDCA). If inadequate, add rifampicin 150 to 300 mg twice daily with liver function monitoring. Third-line options include naltrexone (an opioid antagonist, 50 mg daily) and sertraline. The pruritus should improve with the UDCA and the immunosuppression as the cholestasis improves, but it may be persistent and is a major quality-of-life issue. I would warn her that if I add obeticholic acid in the future, it can WORSEN the pruritus — the commonest reason to reduce the dose or stop the drug [3].

Step 6 — Bone and nutritional health: [1]

Calcium and vitamin D supplementation. A bisphosphonate (alendronate 70 mg weekly) if the DEXA shows osteoporosis or if she is on long-term prednisolone. Replace vitamins A, D, E and K as needed (the prothrombin time may normalise with vitamin K replacement). A dietitian review for general nutrition and weight management. [1]

Step 7 — Second-line and transplant consideration: [1]

If her biochemistry does not normalise on UDCA plus immunosuppression, I would consider adding obeticholic acid for the PBC component — but ONLY if her cirrhosis is compensated (Child-Pugh A). OCA is contraindicated in decompensated cirrhosis because it can precipitate liver failure and death [3]. Her bilirubin of 75, albumin of 33 and INR of 1.2 put her at the boundary of compensated and decompensated; I would reassess after the initial therapy. If she develops any feature of decompensation — refractory ascites, encephalopathy, variceal bleed, a rising bilirubin despite therapy — I would refer her urgently for liver transplant assessment. Five-year post-transplant survival exceeds 80 percent for PBC and AIH; both diseases can recur in the graft (PBC in 11 to 42 percent, AIH in 20 to 30 percent), though recurrence rarely causes graft loss in PBC.

Step 8 — Communication and the multidisciplinary team: [1]

I would explain the diagnosis in plain language, the lifelong UDCA, the 24-month-plus immunosuppression, the surveillance burden, and the corticosteroid side effects. I would involve the GP, the dietitian, the physiotherapist, the hepatology nurse specialist, and (if she develops decompensation) the transplant coordinator. I would document the shared decisions and review them as the biochemistry responds. [1]

Probing questions the examiner would ask

Q: How do you apply the Paris criteria for AIH-PBC overlap in this patient? [1]

A: "The Paris criteria require two of three AIH features AND two of three PBC features. For AIH: interface hepatitis; ALT two to five times the upper limit; IgG two times the upper limit OR positive ANA/SMA. She meets all three — interface hepatitis on biopsy, an ALT of 180, a raised IgG, and positive ANA and SMA. For PBC: AMA positive; compatible biopsy; raised IgM. She meets all three — AMA positive, florid duct lesion, raised IgM. So she has definite overlap. The IAIHG position statement (Boberg 2011) cautions that overlap is partly a research concept and that patients should be treated for their predominant feature, but here both diseases are clearly active and a combined strategy is justified [7]."

Q: Why did you check the IgG4, and what would a raised IgG4 have meant? [1]

A: "I checked the IgG4 to exclude IgG4-related sclerosing cholangitis, the steroid-responsive mimic of PSC. A raised IgG4 in a cholestatic patient, especially an older man with a strictured cholangiogram and autoimmune pancreatitis, would indicate IgG4-sclerosing cholangitis, whose treatment is systemic corticosteroids — the opposite of PSC, which is not steroid-responsive. In this patient the normal IgG4 and the classic PBC features make IgG4-disease unlikely, but the check is part of the standard cholestatic workup and a normal value is reassuring." [1]

Q: Why prednisolone rather than budesonide for the AIH component? [1]

A: "Budesonide is a first-pass corticosteroid that achieves high hepatic concentrations with minimal systemic exposure, making it an attractive steroid-sparing option in non-cirrhotic AIH. However, in cirrhosis the portosystemic shunting bypasses the first-pass metabolism, so budesonide behaves like systemic prednisolone without the steroid-sparing benefit — and it is contraindicated in cirrhosis. This patient has established cirrhosis (FibroScan 14 kPa, splenomegaly, platelets 145), so prednisolone is the correct choice, with diligent attention to bone density, blood glucose and blood pressure." [1]

Q: She develops worsening pruritus after you start her therapy. How do you manage it, and what does it imply for the obeticholic acid decision? [1]

A: "I would step up the pruritus management: cholestyramine first (given separately from UDCA), then rifampicin 150 to 300 mg twice daily with liver function monitoring, then naltrexone or sertraline. The key implication for OCA is that OCA commonly WORSENS pruritus — it is the commonest reason to reduce the dose or stop the drug. So if I were considering adding OCA for her PBC component and her pruritus is already a problem, I would think very carefully: OCA is contraindicated in decompensated cirrhosis in any case, and the pruritus-worsening effect may outweigh the biochemical benefit. I would prioritise optimising the UDCA and the immunosuppression and managing the pruritus before reaching for OCA [3]."

Q: What surveillance does she need, and how does the cirrhosis change it? [1]

A: "For the cirrhosis: upper GI endoscopy for variceal surveillance at diagnosis and every one to three years; six-monthly ultrasound plus alpha-fetoprotein for hepatocellular carcinoma; a baseline and biennial DEXA; and annual fat-soluble vitamin levels. For the overlap syndrome itself: biochemical monitoring every one to three months until remission, then six-monthly; and a follow-up biopsy at 12 to 24 months to confirm histological remission of the AIH component before considering immunosuppression withdrawal. The cirrhosis changes the management in three ways — it rules out budesonide, it constrains the use of OCA, and it mandates the HCC and variceal surveillance. If she decompensates, she needs a transplant assessment." [1]

Q: How would you counsel her about pregnancy, given she is 54 and post-menopausal? [1]

A: "At 54 and post-menopausal, pregnancy is unlikely, but if she were younger and wished to conceive, I would plan pregnancy for when the disease was in biochemical remission. UDCA is safe in pregnancy. Prednisolone and azathioprine are considered compatible with pregnancy. Mycophenolate (if she were on it) is teratogenic and must be stopped at least six weeks before conception. The cirrhosis complicates pregnancy — there is a risk of variceal bleeding (especially in the second and third trimesters when the splanchnic blood flow rises) and decompensation, so I would involve a maternal-fetal medicine specialist and a hepatologist in a joint clinic, and I would screen for and treat varices before conception. Postpartum, I would monitor closely for an AIH flare within three to six months." [1]

Communication and shared decision-making

"I would explain the diagnosis in plain language: her immune system is attacking both the bile ducts and the liver cells, which is why she has features of two diseases, and the treatment needs to address both. I would explain that the UDCA is for the bile duct disease (lifelong), and the prednisolone and azathioprine are for the immune attack on the liver cells (at least 24 months before we consider withdrawal). I would be honest about the chronicity, the surveillance burden, the corticosteroid side effects, and the transplant possibility if the disease progresses. I would address the fatigue specifically — it is the most disabling symptom for many patients, it is under-recognised, and there is no specific therapy, but I would exclude other contributors (anaemia, hypothyroidism, vitamin D deficiency, depression) and offer a graded exercise and sleep programme. I would involve the GP, the dietitian, the physiotherapist, and a hepatology nurse specialist, and I would put her in touch with a patient support group. I would document the shared decisions and review them as the biochemistry responds." [1]

Outcome and follow-up

Mrs K is started on UDCA 750 mg twice daily (15 mg/kg/day). Her TPMT activity is normal, so azathioprine 125 mg daily is added with prednisolone 40 mg daily. The prednisolone is tapered over 8 weeks to a maintenance of 10 mg daily, with the plan to taper further as the azathioprine takes effect. Her pruritus is managed with cholestyramine (with a good response) and her bone health is protected with calcium, vitamin D, and alendronate (the baseline DEXA shows osteoporosis, T-score minus 2.7 at the lumbar spine). Her upper GI endoscopy shows small oesophageal varices (no prophylactic banding needed; she starts carvedilol for primary prophylaxis). Her six-monthly HCC surveillance begins. [1]

At six months, her ALT has normalised (25 U/L), her IgG has fallen to 17 g/L (near normal), her ALP has fallen to 240 U/L (improved but not normal), and her bilirubin has normalised. She remains clinically well. At 12 months, a follow-up biopsy shows no active interface hepatitis (histological remission of the AIH component) but persistent bile duct changes of PBC. The immunosuppression is continued (she is not yet a candidate for withdrawal), and the UDCA is continued lifelong. She does not develop decompensation over the next two years, and the question of obeticholic acid for the residual PBC biochemistry is deferred (her pruritus is well controlled, and the small benefit does not justify the risk in her cirrhosis). She returns to part-time teaching. She remains under annual hepatology review with the surveillance programme in place. [1]