Nodular regenerative hyperplasia of the liver: causes, diagnosis and management

The liver normally has an orderly architecture. In NRH the liver tissue is diffusely transformed into small regenerative nodules, distributed evenly through the organ, with minimal or no fibrosis. A published review describes it as a benign transformation and notes that the absence of fibrous septa between nodules is what separates it from cirrhosis. It belongs to a group of conditions that cause portal hypertension without cirrhosis.

It is rare, and our knowledge of it is limited. The review says there are no published systematic population studies and that understanding rests on case reports and case series, and it quotes autopsy studies in which the condition was found in between 0.72% and 2.6% of cases, and that NRH comprises about 27% of non-cirrhotic portal hypertension in Europe.

The cause is not fully established. The review suggests the underlying process involves abnormal blood flow in the small liver veins, with an adaptive overgrowth of liver cells in response. It says NRH may develop as a result of underlying autoimmune, inflammatory, neoplastic or idiopathic disease, and lists conditions reported alongside it, including rheumatoid arthritis, lupus, HIV infection, inflammatory bowel disease, coeliac disease, scleroderma, antiphospholipid syndrome, sarcoidosis, lymphomas, myeloproliferative disorders and congenital thrombophilias.

Medicines are another recognised link. The review notes that immunosuppressive medications may induce NRH by damaging the endothelial cells of the small hepatic veins, and it describes about 70 reported cases related to the chemotherapy drug oxaliplatin. A separate case report describes a man with long-undiagnosed coeliac disease in whom NRH was found while investigating portal hypertension, and notes that an association with azathioprine has also been described. These links do not mean that everyone taking the medicine or with the condition will develop NRH.

Many people have no symptoms. The review says the disease is usually asymptomatic and slowly or non-progressive unless complications of portal hypertension develop. When it does, the features are those of portal hypertension: an enlarged spleen, swollen veins in the gullet (oesophageal varices) that can bleed, and fluid in the abdomen (ascites). Liver blood tests are often normal or only mildly abnormal. The review says mildly increased liver enzymes, usually alkaline phosphatase, occur in about 11% to 25% of patients.

Because liver function is usually well preserved, someone with NRH may look very different from a person with cirrhosis, without the typical skin signs of chronic liver disease. The case report makes the same point, noting that liver function is usually well preserved and that portal hypertension may be present in some patients.

The review says NRH should be considered in anyone with clinical features of portal hypertension (enlarged spleen, oesophageal varices, ascites) but normal transaminases and no signs of cirrhosis. Accurate diagnosis is by histopathology, which shows the micronodular transformation without fibrous septa.

Investigation also looks for the underlying disease. That can include tests for autoimmune and rheumatological conditions, blood disorders, infections and clotting tendencies, and a review of medicines. In the case report, a thrombophilia screen and autoantibody tests were negative and the portal hypertension was traced to NRH on biopsy, with coeliac disease found on a duodenal biopsy.

The review says treatment is directed towards eliminating the causative factor once established, treating concomitant diseases appropriately and minimising drug toxicity, with long-term anticoagulation usually indicated in the thrombophilias. In patients with NRH, the mainstay of management is the prevention and treatment of complications of portal hypertension, particularly variceal bleeding, which is the main source of mortality.

Treatment of portal hypertension follows standard approaches: a low sodium diet, diuretics, and endoscopic banding of oesophageal varices. A portosystemic surgical shunt or transjugular intrahepatic portosystemic shunt may offer significant benefit in severe recurrent variceal bleeding. The review says liver transplantation is rarely necessary and is reserved for patients with hepatic failure.

The review states that the prognosis of NRH is generally better than that of chronic liver disease and relates to the complications of portal hypertension and the severity of any associated disease. In most cases it is slowly progressive, though the long-term outlook is uncertain and depends on the underlying process. The review also mentions case reports of NRH occurring alongside hepatocellular carcinoma without cirrhosis, although a causal link has not been proven.

Follow-up is usually with a hepatologist, including screening for varices when portal hypertension is present. Anyone with known portal hypertension who vomits blood, passes black stools or becomes confused should seek emergency help through 999 or A&E.