Primary biliary cholangitis presenting with Fanconi syndrome: an important phenotype

Background

Primary biliary cholangitis (PBC) is a chronic, progressive autoimmune disease that eventually leads to liver failure.1 It is characterised by non-suppurative cholangitis and destruction of interlobular bile ducts.2 It is rare with an estimated incidence of 3.9–15 cases per million population per year, affecting predominantly middle-aged women.1 Its hallmark features include presence of antimitochondrial antibodies (AMA) in serum, found in 90%–95% of patients, and cholestatic liver blood tests with raised levels of alkaline phosphatase (ALP) and gamma-glutamyl transferase (GGT).1 2 Raised bilirubin level and evidence of liver synthetic dysfunction are markers of poor prognosis.3 Liver biopsy is not required for routine diagnosis due to the highly disease-specific nature of AMA but is still used where there is diagnostic uncertainty.1 National4 and European5 guidelines emphasise three key diagnostic features. These are sustained elevation (>6 months) above the upper limit of normal for serum ALP activity; characteristic changes on liver biopsy and typical pattern of serum and secretory autoantibodies (AMA or reactivity with primary-biliary-cholangitis-specific antinuclear antibodies). The presence of all three criteria (cholestatic liver biochemistry, AMA or other primary-biliary-cholangitis-specific autoantibody at a titre of >1/40, and diagnostic or supportive liver histology) indicates definite PBC and two out of three criteria is referred to as ‘probable PBC’. Serum gamma-glutamyl transferase is not part of the diagnostic criteria. The UK guidelines state that in clinical practice, the vast majority of patients are appropriately and confidently diagnosed without a liver biopsy, and in clinical practice the term ‘probable PBC’ should not be used with patients and a liver biopsy is no longer recommended to diagnose PBC when the other criteria are fulfilled. Typical symptoms are fatigue and cholestatic pruritus that can have a significant impact on quality of life for some patients.6

Immunosuppression has not been shown to have survival benefit in patients with PBC.6 Ursodeoxycholic acid (UDCA), a hydrophilic bile acid, is the recommended first-line therapy with evidence of improving liver biochemistry and delaying time to death or liver transplantation.6 With early diagnosis and effective treatment, the prognosis of PBC has significantly improved.7 Of note, the nomenclature has changed from ‘primary biliary cirrhosis’ to ‘PBC’ during the last decade to reflect the fact that many patients are not cirrhotic.7

Approximately 70% of patients with PBC have extrahepatic manifestations, mostly autoimmune phenomena, such as Sjogren’s syndrome, thyroid dysfunction and systemic sclerosis.8 Kidney involvement typically presents as distal renal tubular acidosis (dRTA), which is usually clinically silent.9 10 We present a patient with PBC who had features of tubulointerstitial nephritis and Fanconi syndrome. Recognition of this disease association is important as it is easily missed and may respond to steroid therapy.

Case presentation

A woman in her 50s was referred to nephrology clinic due to a progressive decline in kidney function. She had previously documented chronic kidney disease and type 2 diabetes mellitus (for 13 years) with proteinuria. In addition, she previously had a transaminitis and had been found to have AMA. These two criteria (cholestatic liver biochemistry and presence of AMA) indicated a likely diagnosis of PBC. Her liver biochemistry and kidney function were being regularly monitored in primary care.

When the patient was reviewed in nephrology clinic, she reported no major symptoms other than a few episodes of urinary tract infection in the preceding year. Her diabetes had been well-controlled, only requiring treatment with metformin for the previous 4 months. Her haemoglobin A1c (HbA1c)level had never been greater than 55 mmol/mol. She had no other significant medical history.

Investigations

Urine dipstick showed pH of 6, blood+, glucose 3+, protein 3+ (table 1). The severity of glycosuria was inconsistent with her good glycaemic control. She was not taking a sodium-glucose cotransporter 2 (SGLT2) inhibitor, which would have been an obvious cause of glycosuria. When her proteinuria was quantified, we found a disparity between urine albumin/creatinine ratio (ACR) and protein/creatinine ratio (PCR): 18.8 mg/mmol and 124 mg/mmol, respectively, suggesting a non-glomerular protein leak. Indeed, the urinary albumin to total urinary protein ratio was 0.15, where values <0.4 are highly suggestive of primary tubulointerstitial disorders of the kidney. Serum protein electrophoresis and serum-free light chains levels were normal. A spot urine sample was sent for further analysis, which revealed generalised aminoaciduria. Furthermore, she had intermittent hypouricaemia and hypophosphataemia, and normal anion gap metabolic acidosis. All these features are in keeping with a proximal renal tubulopathy known as Fanconi syndrome. However, she had an ‘incomplete’ Fanconi syndrome phenotype as the associated serum electrolyte deficiencies were milder and intermittent in nature. A screen for autoantibodies was negative including anti-SS-A, anti-SS-B and anti-Scl-70.

Her estimated glomerular filtration rate (eGFR) was 21 mL/min/1.73 m², demonstrating a significant decline from the historical values, prompting a kidney biopsy. This showed normal glomeruli, with normal mesangial cellularity and matrix appearances and normal basement membranes but with features typical of a tubulointerstitial nephritis with chronic tubulointerstitial damage (figure 1). The glomeruli appeared unremarkable and there were no features of diabetic nephropathy. The interstitium showed mild fibrosis. Electron microscopy did not show any other specific features. The chronic kidney disease was likely to be secondary to chronic tubulointerstitial nephritis. There was no history of infection, recent antibiotic use or significant drug to suggest an alternative cause for the tubulointerstitial nephritis. From the liver perspective, her ALP levels remained elevated with a persistently elevated serum immunoglobulin M level and circulating AMA with titre of greater than 1/640 (table 1). A liver fibroscan suggested moderate fibrosis. A diagnosis of PBC, complicated by tubulointerstitial nephritis and Fanconi syndrome, was made.

Differential diagnosis

Diabetes mellitus is the leading cause of chronic kidney disease. When a patient with diabetes presents with chronic kidney disease and proteinuria, diabetic nephropathy is often presumed to be the diagnosis. Our patient had a similar presentation but her severe glycosuria, despite good glycaemic control (and normal HbA1c levels) and the absence of treatment with an SGLT2 inhibitor (such as dapagliflozin) prompted further investigations, which confirmed a proximal renal tubulopathy accounting for the glycosuria and tubulointerstitial nephritis leading to progressive chronic kidney disease. There were no other underlying causes for the tubulointerstitial nephritis including drugs (such as proton pump inhibitors and non-steroidal anti-inflammatory drugs), or recent infections requiring antibiotics. There were no features of diabetic nephropathy on the renal biopsy. The elevated liver enzymes levels and presence of AMA were in keeping with a diagnosis of PBC with extrahepatic complications. Sjogren’s syndrome, which may be seen in association with PBC and may cause Fanconi syndrome, was excluded with negative immunology and a lack of supportive clinical features.

Treatment

Given the unifying diagnosis of PBC with associated tubulointerstitial nephritis and Fanconi syndrome, we commenced treatments targeted to each disease process. UDCA 900 mg (13–15 mg/kg) daily was initiated to treat the cholestasis. To correct the metabolic acidosis, sodium bicarbonate 500 mg three times a day was given. Oral prednisolone 30 mg daily was started to treat the tubulointerstitial nephritis and continued for 6 weeks. It was then tapered down to 5 mg daily over 6 weeks and weaned off over another 6 weeks.

Outcome and follow-up

The corticosteroid therapy led to an initial improvement in of eGFR to 28 mL/min/1.73 m², however, this was not sustained (figure 2). Further corticosteroid therapy was considered but its potential benefit would be at the expense of worsening glycaemic control and weight gain. Moreover, COVID-19 became a global concern at this time. In view of these factors, further immunosuppressive therapy was not given. UDCA was well tolerated and titrated further (according to weight) to 1250 mg daily and liver enzymes levels improved.