Students read liver function tests by looking for whatever is highest. That finds the obvious case and misses the useful information, because in LFTs the diagnosis is in the proportions rather than in any one number.
Train yourself to ask one question first: is this pattern hepatocellular or cholestatic?
## The two patterns
Hepatocellular injury means the liver cells themselves are damaged and releasing their contents. ALT and AST rise, and they rise disproportionately to ALP.
Cholestatic injury means bile flow is obstructed, somewhere between the canaliculus and the duodenum. ALP and GGT rise disproportionately to the transaminases.
Most liver disease declares itself as one or the other, and knowing which redirects the entire workup — hepatocellular sends you toward viral serology, drugs and alcohol, while cholestatic sends you toward imaging and the biliary tree.
Look at the relative rise, not the absolute numbers. A patient with an ALT of 300 and an ALP of 400 is not cholestatic just because the ALP is numerically higher, because the reference ranges differ. Compare each against the upper limit of its own range.
## What individual values add
**ALT** is relatively liver-specific. **AST** is present in muscle, heart and red cells too, so an isolated AST rise should prompt a thought about rhabdomyolysis or haemolysis before the liver.
The **AST:ALT ratio** is worth knowing. Greater than 2, particularly with a raised GGT, points toward alcohol. Most other hepatocellular disease has ALT above AST.
Very high transaminases — into the thousands — mean acute hepatocellular necrosis. The differential narrows sharply: viral hepatitis, drug toxicity including paracetamol, ischaemic hepatitis. Chronic conditions rarely produce those figures.
**ALP** comes from bone as well as liver. A raised ALP with a normal GGT is probably bone, and in an adolescent it is probably normal growth. Reporting a growing teenager's ALP against an adult reference range generates unnecessary referrals, which is a laboratory failure rather than a clinical one.
**GGT** is sensitive for biliary disease and induced by alcohol and several drugs. Sensitive, not specific — a raised GGT alone means very little.
## The tests that are not tests of injury
This is the distinction most worth teaching.
ALT, AST, ALP and GGT are markers of damage. They tell you the liver is being injured. They do not tell you whether it is working.
**Albumin** and the **prothrombin time** measure synthetic function — what the liver is actually producing. A patient with modestly raised transaminases, a low albumin and a prolonged PT is in far more trouble than one with transaminases in the thousands and normal synthetic function.
Albumin has a long half-life, so it reflects chronic disease. The PT responds within days, so it reflects acute change. That is why the PT is the better acute marker and albumin the better chronic one.
Note that a prolonged PT in liver disease does not mean the patient is anticoagulated. The liver makes the natural anticoagulants too, and they fall alongside. The PT measures only one arm of a balance that has moved on both sides.
**Bilirubin** rises in both patterns and is not a synthetic marker. Split it: a raised conjugated fraction points to hepatic or post-hepatic causes, unconjugated to haemolysis or Gilbert's.
## Reading it as a whole
A middle-aged man with AST twice ALT, raised GGT, normal ALP and a low albumin tells a coherent story before you have taken a history.
A young patient with ALT of 2,000, normal ALP and a normal PT is acutely injured but compensating.
A patient with isolated ALP elevation and a normal GGT probably has a bone problem and does not need a liver ultrasound.
Standard chemistry platforms such as the Agappe TBA-2000FR report the whole panel together in one run, which is precisely why it should be read together rather than scanned for outliers.