Mildy deranged LFTs in the GP setting

A man in his mid 30s was invited to make a routine appointment to discuss his LFTs.

6 months ago Now Reference range
Alanine Transferase  67 132 < 40 IU/L
Aspartate Transferase  Not done Not done < 48 IU/L
Alkaline Phosphatase  85 91 30 – 130 IU/L 
Gamma GT  43 38 < 50 IU/L in men< 35 IU/L in women
Bilirubin  16 18 < 21 µmol/L
Albumin  42 44 35 – 50 g/L


He used to drink quite heavily as a student, but was now drinking about a pint of beer every two weeks since he graduated.

He had lived in the Far East for five years, and returned to the UK two years ago. He used to use heroin and amphetamines IV whilst in the Far East. He had never received a blood transfusion.

He had never been jaundiced.

He was overweight, with a BMI of 33. His fasting glucose was 4.5. His total cholesterol was 6.6 mmol/L with HDL 0.7 mmol/L.

There was no family history of liver disease, but his father did have a coronary stent put in at the age of 40.

On examination, there were no signs of chronic liver disease, and his liver was not enlarged nor tender.

He had an US abdomen 6 months ago, which showed a diffusely fatty liver.

Old me: Bloody hell. Treat the patient, not the number. It’s just a slightly abnormal isolated LFT; these things are common.

New me: We’re probably underestimating the extent of liver disease in our general population. Most of the time (90-95%) patients with abnormal LFTs do actually have an underlying liver disease.

Of those who don’t, some have a non-liver sources for their enzymes (commonly muscle/heart for AST and bone for alkaline phosphatase).

And to make things worse, our reference range probably misses some liver disease. Lots of slow burning chronic liver disease can occur without dramatic hepatocyte death, so the intracellular enzymes from the liver that we use as markers of liver injury (AST and ALT) don’t rise so dramatically.  In South Korea, a prospective cohort study (wonderful Wikipedia diagram of how a cohort study works) involving 94 533 people found that having your AST or ALT in the upper end of the normal range was associated with having increased mortality from liver disease. The authors suggested that reducing the reference range to around 30 IU/L would pick up liver disease earlier. Does this apply in the UK? The prevalence of hepatitis B (UK: 0.3%, S. Korea 4%) and C (UK: 0.5%, S.Korea 1.3%) is lower here, but our burden of liver disease is thought to be significantly underestimated.

Old me: OK, so we should probably take it seriously. Full liver screen, see him in two weeks with the results.

New me: In this patient’s case, there is a clear risk factor for viral hepatitis. IVDU in the Far East is a dangerous game. However, if this risk factor were not present, would we still send the panel?

You should first check the patient does not have any evidence of decompensation. This means no ascites, variceal bleeding or encephalopathy. These patients may need admission.

You should also consider any acute problems leading to markedly deranged LFTs. In particular, drug-induced, ischaemia and infection can cause an acute rise in the transaminases > 1000 IU/L.

If there’s nothing too dramatic going in, you could split the possible causes into those initially treatable in GP world and those that need a referral to a Liver God. The main initially treatable ones in GP world ones are drug induced, alcohol and NASH. So if this is the first occasion of an LFT derangement, the patient is asymptomatic and the derangement is not too marked (<2x upper limit of normal), you could repeat the LFTs at 3-6 months having stopped any offending drugs (including over the counter culprits), cut down the alcohol, got the BMI below 25 and reassessed. If the LFTs have not resolved after 6 months, or if the derangement was initially marked (>2x upper limit of normal), you should proceed to a liver screen.

Old me: Drug induced, eh? This means you would stop a statin/not start him on a statin because of his LFTs.

New me: NICE (1.4.33) still recommend a baseline LFT and stopping if LFTs reach more than three times the upper limit of normal at 3 months. This will probably change eventually, and most hepatologists do not consider the transient elevation of transaminases that occurs when starting statins to be linked with liver failure. There is even evidence that NASH improves with statin therapy. This patient would benefit from a statin given his lipid profile and his family history of CV disease.

Old me: And to assess severity, its not about the degree of derangement of the liver enzymes. It’s about the functional state of the liver.

New me: Absolutely. It’s like amylase in acute pancreatitis – you get the initial organ damage and this releases an intracellular enzyme that we use for diagnosis, but not prognosis. You’ll notice that the King’s College criteria for paracetamol induced and non-paracetamol induced liver failure both ignore the transaminases and alkaline phosphatase levels. They care about the functional bits – pH and bilurubin (metabolic function), encephalopathy (detox function) prothrombin time (synthetic function) etc. The prognosis (in terms of mortality) can be estimated using the MELD score or Child Pugh score.

Old me: You do an FBC to look for a low platelet count, which goes with alcohol.

New me: Yes it does go with alcohol, for many reasons. But in terms of liver things, any chronic fibrosis can lead to portal hypertension and hypersplenism. As a result of the hypersplenism you get thromboycytopenia.

Old me: You do a liver screen to look for autoimmune, metabolic and infectious etiologies. If this is negative and the LFTs are still deranged, refer.

New me: There is information in the history and examination that cannot be obtained in the blood tests. Alcohol use is the main feature on the history. The presence of heart failure on history or examination is important too, as this can cause a congestive hepatopathy. If the liver screen is negative despite all this, an ultrasound of the abdomen can help distinguish some biliary problems and provide some information about the structure of the liver. Referral is usually done after US of the abdomen and a liver screen unless there is a clinical urgency.

According to the BSG, the liver screen should consist of:

Hepatitis B surface antigen

Hepatitis C antibody

Anti-mitochondrial antibody

Anti-smooth muscle antibody

Anti-nuclear antibody

Serum immunoglobulins (IgG can be markedly raised in autoimmune hepatitis)

Serum ferritin

Transferrin saturation

Copper and caeruloplasmin (under 40 years)

a1 antitrypsin level