Statistics from Altmetric.com
See article on page 576
Acute liver failure, defined as hepatic encephalopathy developing within 12 weeks of jaundice in the absence of a previous history of symptomatic liver disease,1 is a medical emergency. Accurate recognition of the underlying disease is crucial to enable specific therapy, when available, to be instituted. Rowbothamet al (see page 576), in the most comprehensive examination of this topic to date, discuss the importance of malignant hepatic infiltration as a cause of acute liver failure. This is, however, an undoubtedly rare cause for this presentation in the UK, responsible for only 0.44% of all cases of liver failure admitted to the King’s Unit since 1978.
This syndrome can occur in a number of scenarios. Firstly and most importantly, patients with no previous history of malignancy may develop fulminant liver failure and present as a diagnostic problem. Secondly, acute liver failure may develop in a patient with previously diagnosed malignancy but without known liver involvement. Finally, fulminant liver failure may be seen as a terminal event in a patient already known to have primary or secondary hepatic malignancy. Such patients, although not fulfilling the strict diagnostic criteria for acute liver failure, may provide further insight into the likely mechanisms of liver failure in malignancy.
In a patient presenting with acute liver failure with no cause apparent, the diagnosis of hepatic infiltration should be considered but may be surprisingly difficult to make. Certain clinical and biochemical parameters may heighten the suspicion of infiltrative disease. All patients in the King’s series had a non-specific prodromal illness for 1–8 weeks prior to presentation. Jaundice tends to be milder compared with other causes of acute liver failure whereas hepatomegaly is more pronounced (indeed in many cases of acute liver failure the liver is small). In contrast to liver failure resulting from paracetemol overdose, signs of cerebral oedema are rare in malignancy with only one case report2 describing the occurrence of pupillary abnormalities and decerebrate posturing as a terminal event in a patient with non-Hodgkin’s lymphoma. The finding of lymphadenopathy and splenomegaly clinically or on imaging is helpful in recognising cases owing to lymphoma, but is not helpful in the diagnosis of other malignancies. Abnormal liver texture was noted on imaging in less than half the cases described by Rowbotham et al. Massive increases in aminotransferases (up to 100-fold elevated) can occur with malignant infiltration3 4 5 and would help to exclude acute alcoholic hepatitis, which may occasionally be misdiagnosed in this setting.6 Lactic acidosis and a disproportionately raised lactate dehydrogenase are suggestive of haematological malignancy.7 Sugano and Suzuki recently reported the finding of a notable increase of the brain isozyme of creatine kinase (CK-BB) in a patient with hepatic failure and non-Hodgkin’s lymphoma and suggest it may be useful as a marker of ischaemia of hepatic sinusoidal cells.8 The specificity of this test needs to be evaluated. No clinical or biochemical features are sufficiently specific to obviate the importance of a tissue diagnosis. Reassuringly, in the King’s series, there were no complications reported from liver biopsy (percutaneous, transjugular or via laparotomy) in the eight patients in whom it was performed. Frequently there may be only a narrow window of opportunity for safe biopsy in these rapidly deteriorating patients.3
Alternate sources of tissue for diagnostic purposes include bone marrow and lymph nodes.
In the second group of patients a history of previous malignancy will alert the clinician to the possible infiltrative cause of liver failure. Haematological malignancies are implicated most commonly, in particular Hodgkin’s and non-Hodgkin’s lymphoma, as well as malignant histiocytosis, acute leukaemia and, recently described, Richter’s transformation of chronic lymphocytic leukaemia.3
Other less common causes are breast, colon, gastric, and small cell carcinoma.4 9 Occasionally, the history of malignancy may be remote and the patient believed to be tumour free.10Alternately, the patient may have a predisposing condition to malignancy only, such as coeliac disease2 or autoimmune haemolytic anaemia (Rowbotham et al). The differential diagnosis of acute liver failure in patients with malignancy should also include the possibility of chemotherapy or radiotherapy induced liver failure, or of reactivation of underlying chronic hepatitis B liver disease, described as a complication of chemotherapy.11
The third group of patients is those where hepatic failure occurs as a terminal event in the setting of advanced malignant disease. This may be under-recognised as a clinical entity for in such patients hepatic coma may be mistakenly attributed to severe constitutional disease, cerebral metastases, drugs or sepsis.9 Eras and Sherlock found in a retrospective study that 7.2% of hospitalised patients with hepatic metastases had evidence, albeit only clinical, of hepatic encephalopathy but in over half this was not recognised or treated.9
A number of mechanisms are proposed to explain the pathogenesis of infiltrative liver failure and these are clearly discussed by Rowbothamet al. Massive cytokine release has been implicated as a cause of liver failure and this theory is particularly attractive given the fact that responsible cytokines, such as interleukin 2, are released by lymphomas, the commonest malignancy responsible for acute liver failure. This mechanism explains why liver failure can occur despite a relatively low, or even absent tumour burden in the liver.12 Cytokine release can cause liver failure by damaging bile ducts both directly and via recruitment of effector cells, and by activation of leucocytes and hepatic sinusoidal cells, thus impeding hepatic sinusoidal microcirculation. Ultimately, liver failure occurs as a result of ischaemia, probably as a consequence both of cytokine release and on occasion from direct physical compression of hepatic sinusoids by the expanding tumour mass.10 Acute liver failure can occasionally reverse very quickly with appropriate chemotherapy,3 consistent with the mechanisms outlined above, whereas if liver failure occurred as a result of the loss of a critical mass of hepatocytes due to replacement with malignant cells, rapid improvement would not be anticipated. The direct effect of tumour infiltration with replacement of hepatocytes is probably more important as a mechanism in non-haematological malignancies such as small cell carcinoma.4 13 Liver failure may also occur due to ischaemia produced by tumour emboli compromising the portal venous circulation, particularly in primary hepatocellular carcinoma.14 A further mechanism proposed is liver failure due to hepatic venous obstruction (secondary to emboli, hypercoagulability or a compressive mass).
Unfortunately, the prognosis of patients with acute liver failure resulting from malignant infiltration is dismal. The few survivors are those with haematological malignancies3 7 in whom, crucially, the diagnosis was made promptly and treatment instituted. The only survivor in the King’s series (Rowbotham et al) was a patient where chemotherapy was initiated by the referring hospital prior to transfer. Unfortunately, survival is not reported for patients with non-haematological malignancies, such as in our own personal experience with two patients with small cell and breast carcinoma. Despite being unable to offer potentially curative treatment in this group, recognition of the diagnosis is important to ensure that these patients are not inappropriately listed for liver transplantation.
See article on page 576
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.