AFP Blood Test (Alpha-Fetoprotein): Adults and Liver Health

AFP Blood Test (Alpha-Fetoprotein): Adults and Liver Health

This page is educational and not medical advice, and every result belongs in a conversation with a clinician or licensed provider who knows your history.

What is alpha-fetoprotein, and where does it come from?

Alpha-fetoprotein is a protein produced by the developing liver and the yolk sac before birth, circulating at high concentration in fetal blood and falling after delivery so that adult blood holds only a small amount 1.

It sits in the same protein family as albumin and acts as a carrier protein during development, binding and moving substances through fetal circulation 1. Mature liver cells make very little of it. Production can switch back up in two broad settings, and both matter for how the test is read: when liver cells divide and regenerate after injury, and when a tumour arises from a cell line that keeps the fetal production pattern, which includes liver cells and germ cells 1.

That single piece of biology is why one protein turns up in three unrelated corners of medicine. In liver care it is a marker of what liver tissue is doing. In cancer care it is a marker some tumours release. In prenatal care it is measured for an entirely separate purpose. Same molecule, three contexts, three different reasons to look at it.

On a report the value appears in nanograms per millilitre (ng/mL) or micrograms per litre (µg/L). The two units give the same number, so a value printed one way can be compared with a value printed the other way without conversion.

The presence of this protein in adult blood is not an abnormality in itself. The useful questions are how much there is, how that compares with the range printed on the report, and which direction the number moves when the test is repeated. Because the level shifts over days rather than hours, a repeat test is spaced weeks apart rather than ordered the same week.

Why do adults have this protein measured?

Adults have it measured for a small, defined set of clinical questions, not as a routine wellness number. Those questions include surveillance for liver cancer in people with cirrhosis, assessment of a suspected germ cell tumour, working up an unexplained liver abnormality, and following a known marker-producing tumour over time.

The liver use is the one adults meet most readily. Practice guidance from the American Association for the Study of Liver Diseases describes surveillance for hepatocellular carcinoma in people with cirrhosis using abdominal ultrasound, with or without AFP, repeated at roughly six-month intervals 2. Surveillance in that sense is not the same as screening the general population. It applies to people whose clinician has already identified a raised baseline risk, and the decision to start it, pause it or stop it belongs in a conversation with that clinician.

The second use is in germ cell tumours of the testis or ovary, where the protein can be released by tumour tissue and measured in blood 1.

The third use is diagnostic groundwork. An above-range value has a broad differential that includes hepatocellular carcinoma and a list of other disorders, which is why the number is read in context rather than alone 3. On its own, in someone with no liver disease and no other reason for the test, a single value answers very little and can open questions that only imaging and a full history can close.

How is the blood draw done, and is any preparation needed?

It is an ordinary venous blood draw, the same as any other serum test, and no special preparation is needed for this marker itself.

A phlebotomist takes blood from a vein in the arm, the sample is spun down, and the serum is run on an automated immunoassay platform. AFP alone does not call for fasting. Other tests drawn at the same appointment can, so follow whatever instructions appear on your requisition rather than assuming the whole draw is fasting or non-fasting.

Turnaround depends on the laboratory and on whether the sample is run in-house or sent out. Two practical points matter more than speed.

First, laboratories use different assay platforms, calibrators and reporting ranges. A value is read against the range printed on the report that carries it, not against a number remembered from a different lab. Second, when a result is being tracked over months, serial values are easier to interpret when the same laboratory runs them, because a change in platform can shift the number without anything changing in the person.

Some laboratories can also report a specialised fraction of the protein, measured by a separate method, and that version has to be requested specifically rather than assumed to come with the standard test. A clinician decides whether that extra measurement adds anything to the question being asked.

What range is the result compared against?

A result is compared against the reference range printed on your own lab report, because that range belongs to the assay the laboratory used.

Marker Units on the report Lab's own reference range FitScript optimal range
Alpha-fetoprotein (AFP) ng/mL or µg/L (the same number in either unit) As printed on your lab report Read against the lab's printed range
ALT, men U/L As printed on your lab report 10 to 25 U/L
ALT, women U/L As printed on your lab report 8 to 19 U/L

Reference ranges vary by laboratory, age and sex, so the comparison that counts is the one on the page in front of you. FitScript's optimal ranges are optimization targets, not diagnostic cut-offs, and a value sitting outside an optimal range is a reason to look further, never a label.

Two things follow from that. A value inside the printed range is reassurance about one measurement at one moment, not a statement about the whole liver. A value outside the printed range is information that needs a clinician to place it, because the same number carries different weight in someone with established liver disease than in someone without it 3.

What can a result above the printed range reflect?

A value above the printed range means the laboratory measured more of the protein than the range it uses for comparison, and nothing more specific than that until a clinician interprets it.

The result does not identify or confirm any condition. What it does is change the next question. Three features shape how a clinician reads it: the size of the elevation relative to the printed range, the direction of travel across repeated tests, and the clinical setting, particularly whether chronic liver disease is already known.

Direction of travel carries real weight. A value that sits a little above the range and stays flat over repeat testing reads differently from one that climbs across successive draws, and a clinician uses that pattern alongside imaging rather than treating any single number as a verdict.

Context carries just as much weight. An abrupt rise in someone living with chronic hepatitis B can accompany a flare of liver inflammation rather than a tumour, a pitfall described in a reported case of exactly that pattern 4.

The practical sequence is simple. Observation: the number is above the printed range. Meaning: several explanations are possible, and they differ in how urgent they are. Next step: bring the report to a clinician, who reviews your history, decides whether to repeat the test, and decides whether imaging belongs in the workup. A value outside the printed range is worth discussing with a clinician promptly, and never worth filing away unread.

Which conditions other than liver cancer can push the number up?

Several conditions besides hepatocellular carcinoma raise this protein, which is exactly why the result is never read in isolation. Listed without any ranking:

The point of the list is not to generate worry, it is to show why one number cannot carry a diagnosis. Each line on it leads somewhere different, and the thing that tells them apart is the rest of the picture: symptoms, history, the liver panel, imaging, and in some cases a repeat test weeks later.

Can the number sit inside the printed range when something is still wrong?

Yes. An in-range value does not rule out liver disease or a tumour, because not every tumour arising from liver cells produces enough of the protein to push the number past the printed range.

This is the single most useful thing to understand about the marker. It is not a yes-or-no gate. Research suggests that, in people with cirrhosis, surveillance using ultrasound together with AFP detects a greater proportion of early-stage hepatocellular carcinoma than ultrasound alone, which is also a plain statement that neither element of that pairing catches everything by itself 7.

So an in-range result is read as one piece of information gathered on one day. If symptoms, imaging or the rest of the liver panel point somewhere, an in-range value does not cancel them out, and a clinician weighs the whole set.

For anyone in an active surveillance arrangement, this is why the schedule exists. A programme repeated at intervals is built around the fact that any single test can miss, and that the pattern over time carries more information than one draw.

How does this marker work alongside ultrasound in liver cancer surveillance?

It works as an add-on to imaging, not as a replacement for it. Practice guidance from the American Association for the Study of Liver Diseases describes abdominal ultrasound, with or without AFP, repeated at roughly six-month intervals for people with cirrhosis 2.

The two tests fail in different ways, which is the argument for pairing them. Ultrasound looks for a lesion and depends on image quality, body habitus and operator experience. The blood marker looks for a protein signal that some tumours release and some do not. Research suggests that combining the two detects a greater proportion of early-stage hepatocellular carcinoma than ultrasound alone in people with cirrhosis 7.

What that means in a clinic room is a repeating cycle rather than a one-off answer: an ultrasound, a blood draw, a review of both against the previous round, and a decision about whether anything needs a closer look with cross-sectional imaging. Scan preparations and scheduling are set by the imaging service.

Who enters a surveillance arrangement, how long it continues, and what happens when one element of it is abnormal are all clinical decisions. They depend on the cause and stage of the underlying liver disease and on the person's overall health, and they belong in a conversation with the clinician managing that liver disease rather than in a self-directed testing schedule.

What are AFP-L3 and PIVKA-II, and why are they measured?

They are two refinements that exist because the standard measurement alone cannot always separate a tumour signal from benign liver disease.

AFP-L3 is a fraction of the same protein defined by how it binds to a particular plant lectin. Different liver conditions produce versions of the protein with different sugar structures, and measuring that lectin-reactive fraction was developed as a way to tell liver disease patterns apart in serum 5. It is reported as a percentage of the total rather than as a separate concentration, and it requires a specific assay that not every laboratory offers.

PIVKA-II is a different protein altogether, produced by some liver tumours, and it is measured as a complementary marker rather than a substitute. Research suggests that combining PIVKA-II with AFP has utility in the surveillance and monitoring of hepatocellular carcinoma 8.

Neither addition turns a blood test into a diagnosis. Both exist for the same reason the surveillance pairing exists, which is that markers overlap imperfectly and a combination narrows the gap. Whether either is worth ordering depends entirely on the clinical question, and a clinician decides that, not a general rule.

Can serial results be used to follow someone over time?

Yes, and tracking a trend is where this marker earns its place. A sequence of values from the same laboratory says more than any single number.

In people with a known marker-producing tumour, the value can be followed as part of monitoring. Studies indicate that serum AFP has utility for monitoring for relapse in marker-positive hepatoblastoma, a liver tumour of childhood, where serial measurement is used to watch for a rise after initial care 9. The same logic of combined serial markers has been described for hepatocellular carcinoma monitoring in adults 8.

For serial values to mean anything, the conditions have to be stable: same laboratory, same assay, consistent intervals set by the clinician rather than by the person being tested. A one-off value obtained outside that structure is hard to place next to an older number from elsewhere.

Monitoring schedules and any decisions that follow from them, including what happens when a tracked value moves, are clinical judgements. Treatment decisions belong to a licensed provider who has reviewed the person, the imaging and the full history.

Why is this protein raised in some inherited ataxias?

Because a group of inherited neurological conditions carries a persistently elevated level of this protein as part of their biochemical signature, which makes a simple blood test a useful signpost in an otherwise difficult workup.

An elevated value is associated with a set of autosomal recessive cerebellar ataxias, described as uncommon conditions sharing an accessible biomarker 6. These conditions present neurologically rather than through the liver, with progressive problems of balance, coordination and eye movement that develop over years.

The diagnostic path in that setting is not straightforward, and an elevated value in someone with progressive ataxia has been described as a finding that raises specific diagnostic questions and directs further assessment 10.

For an adult reader, the practical takeaway is narrow. An elevated value does not suggest a neurological condition in someone with no neurological symptoms. But in someone already under assessment for progressive problems with balance and coordination, a neurologist may measure this protein as part of that workup, and the result contributes to a picture built from examination, imaging and genetic testing. A value outside the printed range in that setting is worth discussing with the clinician running the assessment, who decides what it reflects.

Where does ALT fit alongside it?

ALT is a liver enzyme that leaks into blood when liver cells are injured, so it answers a different question entirely: not whether a tumour-associated protein is circulating, but whether liver cells are under stress right now. The two are read together because each covers a gap the other leaves open. A full account of what the enzyme is, how it is tested, what its ranges are and what a result can mean sits on the ALT page. In practice a clinician looks at both alongside imaging and the rest of the liver panel, and a value outside the printed range on either marker is worth discussing with a clinician, who decides what it reflects.

Which other markers are worth running at the same time?

The ones that describe liver function, liver injury and liver structure, because this protein on its own describes none of them.

A standard liver panel covers enzymes that reflect cell injury and bile flow, along with bilirubin, albumin and total protein, which say something about how well the liver is doing its synthetic and clearance work. If you want the mechanics of that panel, see what do liver function tests show.

Alongside it, a clinician may look at:

This is the argument for reading a panel rather than a number. Within FitScript's membership, 150+ biomarkers tested per year are interpreted by Atlas, its AI health advisor, so a single marker is read next to the rest of the panel rather than in isolation, and any treatment that follows is decided by a licensed provider after reviewing the member.

When is a result worth discussing with a clinician?

Any value outside the range printed on your lab report is worth discussing with a clinician, and the conversation should happen promptly rather than at some vague later date.

Book that conversation when:

Bring the printed report itself, including the laboratory's own range and the units, along with any previous values. That context is what allows a clinician to read a trend rather than a snapshot.

Some signs call for emergency care rather than a scheduled appointment: vomiting blood, black or tarry stools, new confusion or drowsiness, severe abdominal pain, or yellowing of the skin or eyes with fever. For any of those, call 911 or go to the emergency department.

And the reverse situation deserves saying plainly. A value inside the printed range does not close a question that symptoms or imaging have opened. If something feels wrong, that is worth raising with a clinician whatever the number on the report says.

Key takeaways

References

  1. Alpha-fetoprotein: a review. Critical Reviews in Clinical Laboratory Sciences, 1981.
  2. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology, 2023.
  3. Elevated alpha-fetoprotein: differential diagnosis - hepatocellular carcinoma and other disorders. Clinics in Liver Disease, 2015.
  4. Gastrointestinal: Pitfalls of an abruptly elevated alpha-fetoprotein in a chronic hepatitis B carrier. Journal of Gastroenterology and Hepatology, 2023.
  5. Serum alpha-fetoprotein and its lectin reactivity in liver diseases: a review. Annals of Clinical and Laboratory Science, 1990.
  6. Autosomal Recessive Cerebellar Ataxias With Elevated Alpha-Fetoprotein: Uncommon Diseases, Common Biomarker. Movement Disorders, 2020.
  7. Surveillance Imaging and Alpha Fetoprotein for Early Detection of Hepatocellular Carcinoma in Patients With Cirrhosis: A Meta-analysis. Gastroenterology, 2018.
  8. Utility of combining PIVKA-II and AFP in the surveillance and monitoring of hepatocellular carcinoma in the Asia-Pacific region. Clinical and Molecular Hepatology, 2023.
  9. The Utility of Serum Alpha-fetoprotein for Monitoring for Relapse of Alpha-fetoprotein-Positive Hepatoblastoma. Journal of Pediatric Hematology/Oncology, 2024.
  10. Progressive Ataxia with Elevated Alpha-Fetoprotein: Diagnostic Issues and Review of the Literature. Tremor and Other Hyperkinetic Movements, 2019.

Common questions

What does an AFP blood test actually measure?

It measures the amount of alpha-fetoprotein circulating in serum, a protein made in large amounts before birth and in small amounts in adult life. The result is reported in ng/mL or µg/L, which are the same number in either unit. On its own the value describes one measurement on one day, not a diagnosis.

Do you need to fast before the blood draw?

No special preparation is needed for this marker itself. Other tests drawn at the same appointment may require fasting, so follow the instructions printed on your requisition rather than assuming the whole draw is one or the other. If you are unsure, ask the ordering clinician or the laboratory before you go.

What does it mean if the result comes back above the printed range?

It means the laboratory measured more of the protein than the range it uses for comparison, and nothing more specific than that until a clinician interprets it. Several different conditions can raise the value, and the size of the elevation, the trend across repeat tests and your medical history all change how it is read. Take the printed report to a clinician, who decides what it reflects and whether imaging or a repeat test is the next step.

Can the result be inside the range and still miss a liver problem?

Yes. Not every tumour arising from liver cells produces enough of this protein to push the number past the printed range, so an in-range value does not rule out disease. That is why the blood marker is paired with imaging rather than used as a stand-alone answer. If symptoms or a scan point somewhere, an in-range number does not cancel them out.

Is this a general screening test anyone should ask for?

No. It is used for specific clinical questions, including surveillance in people with known liver disease and the assessment of certain tumours, rather than as a routine wellness number. Measured without a clear reason, a single value can raise questions that then require imaging to settle. Whether it belongs on your panel is a decision for your clinician.

Why is it measured every six months in some people?

Because surveillance works as a repeating cycle rather than a one-off check, pairing an ultrasound with a blood draw so that each round is compared with the last. Any single test can miss something, and a pattern over time carries more information than one value. Who enters a surveillance schedule and how long it continues are decisions made with the clinician managing the underlying liver condition.

Can results from two different laboratories be compared directly?

Not reliably. Laboratories use different assay platforms and print different reference ranges, so a change in laboratory can shift the number without anything changing in you. When a value is being tracked, keeping the testing in one laboratory makes the trend far easier to interpret. Always read a value against the range printed on the report that carries it.

What should you do with an abnormal result while you wait for an appointment?

Keep the printed report with its units and reference range, gather any previous values, and book the conversation promptly rather than leaving it. Do not start or stop anything on your own, and do not ignore the result. If you develop vomiting of blood, black or tarry stools, new confusion, severe abdominal pain, or yellowing of the skin or eyes with fever, call 911 or go to the emergency department.

Related reading