AFP (Alpha-Fetoprotein): Liver Cancer Surveillance, Germ Cell Tumors, and Pregnancy Screening

Reviewed by AskAnything Clinical Team, MD-reviewedLast updated 2026-04-26

AFP is one of the few labs in medicine where the same number can mean three completely different things depending on who is being tested. A 25,000 in a fetus is normal. A 25,000 in a 55-year-old with cirrhosis is liver cancer until proven otherwise. A mildly elevated AFP in a non-pregnant adult with a healthy liver is usually nothing at all. The number is meaningless without the patient.

If you are looking at this page, you are probably in one of three situations:

  • You have cirrhosis, and AFP is part of your every-six-months surveillance for hepatocellular carcinoma alongside an ultrasound. AASLD recommends both. Neither alone is enough.
  • A young person, often a young man, has a testicular mass (or a young woman has an ovarian mass), and AFP was drawn alongside beta-hCG and LDH. Non-seminomatous germ cell tumors produce AFP. Pure seminomas and dysgerminomas do not.
  • You are pregnant, and AFP is part of second-trimester screening for neural tube defects, reported as multiples of the median (MoM) for gestational age.

Outside those contexts, AFP makes a poor general cancer screen. It rises in chronic hepatitis, cirrhosis flares, pregnancy, hereditary tyrosinemia, ataxia-telangiectasia, and a long tail of benign liver insults. An AFP ordered without a clear question behind it tends to generate confusion rather than answers. The right framing for any AFP result is to ask which of those three stories it belongs to first, and then read the number through that lens.

What AFP actually is

AFP is the major fetal serum protein, made by the yolk sac and fetal liver during gestation. Levels in the fetus and newborn are extraordinarily high (often above 50,000 ng/mL) and decline rapidly after birth, reaching adult levels (below 10 ng/mL) by about a year of age. In healthy adults it has no major function and is barely detectable.

AFP rises again whenever hepatocytes are regenerating in bulk or have been transformed:

  • Hepatocellular carcinoma (HCC), the classic adult oncologic association. Roughly 60 to 70 percent of HCCs produce AFP. Small early-stage HCCs are far less likely to.
  • Non-seminomatous germ cell tumors (yolk sac, embryonal, mixed). These cancers recapitulate fetal yolk sac biology, including AFP production.
  • Active hepatic regeneration during chronic hepatitis flares, cirrhosis, or recovery from acute hepatitis.
  • Pregnancy, from placental and fetal sources.
  • Rare hepatic conditions: hereditary tyrosinemia type 1, ataxia-telangiectasia.

AFP cutoffs

DemographicLowHighUnit
Adult (non-pregnant)010ng/mL
Mild elevation1020ng/mL
Moderate elevation20200ng/mL
High suspicion HCC in cirrhosis200400ng/mL
Diagnostic of HCC with mass in cirrhosis400100000ng/mL
Maternal serum (pregnancy)00reported as MoM

For non-pregnant adults outside infancy:

  • Below 10 ng/mL: normal in non-pregnant adults.
  • 10 to 20 ng/mL: mild elevation. Often seen in chronic hepatitis or cirrhosis without cancer. Warrants context, not panic.
  • 20 to 200 ng/mL: moderate elevation. In a cirrhotic patient, raises HCC suspicion. Combined with imaging, often confirms or excludes.
  • 200 to 400 ng/mL in cirrhosis: highly suggestive of HCC unless proven otherwise.
  • Above 400 ng/mL in cirrhosis with a liver mass: diagnostic of HCC under most international criteria, often without need for biopsy.
  • Above 1000 ng/mL: typically advanced HCC or a non-seminomatous germ cell tumor.

Maternal serum AFP in pregnancy is reported as multiples of the median (MoM) for gestational age, not as raw ng/mL. That is a separate clinical workflow and should not be conflated with oncology AFP.

Why AFP goes up, and how to tell what it means

The interpretive question is always the clinical setting. The same AFP value lands in different worlds depending on the patient.

In a cirrhotic patient on HCC surveillance:

  • A stable mild elevation (10 to 20) over time often reflects ongoing chronic hepatitis activity rather than cancer.
  • A rising AFP, even within the "normal" range, is meaningful. Trends matter more than absolute thresholds.
  • An AFP above 200 with a liver lesion on imaging is highly suggestive of HCC.
  • An AFP above 400 with a liver mass meeting imaging criteria is generally accepted as diagnostic of HCC without biopsy.

In a young person with a testicular or ovarian mass:

  • Elevated AFP suggests a non-seminomatous germ cell tumor (yolk sac, embryonal, mixed types).
  • Pure seminomas and pure dysgerminomas typically do not produce AFP. An elevated AFP excludes "pure" classification.
  • Combined with beta-hCG and LDH, AFP is part of the IGCCCG risk stratification used to guide treatment.

Common benign causes of elevated AFP:

  • Acute or chronic hepatitis (viral, alcoholic, autoimmune).
  • Cirrhosis without HCC, especially during regenerative flares.
  • Recovery from acute liver injury.
  • Pregnancy.
  • Hereditary tyrosinemia type 1, ataxia-telangiectasia, hereditary persistence of AFP (rare).

Other malignancies that can raise AFP: hepatoid variants of gastric, pancreatic, and lung adenocarcinoma; rarely, biliary tract tumors. Uncommon, but worth remembering when a markedly elevated AFP shows up without a clear hepatocellular or germ cell explanation.

Low AFP

Low AFP is reassuring in two specific contexts: it is consistent with no detectable HCC in a cirrhotic surveillance population, and in a treated germ cell tumor patient it is consistent with remission. There is no clinically meaningful "too low" AFP in a non-pregnant adult.

A normal AFP does not rule out HCC. Sensitivity of AFP alone for HCC is roughly 40 to 65 percent at standard cutoffs, and small early-stage tumors often produce little or no AFP. This is why AASLD pairs AFP with abdominal ultrasound every 6 months. Neither test alone is sufficient.

Pure seminomas and pure dysgerminomas do not produce AFP. A normal AFP in someone with a testicular or ovarian mass narrows the histology but does not rule out cancer.

Trend interpretation

AFP trend matters more than any single value, especially in three settings:

  • HCC surveillance. A doubling of AFP from a stable cirrhotic baseline, especially a rise that crosses 100 to 200 ng/mL, prompts cross-sectional imaging (multiphase CT or MRI) even when ultrasound is unremarkable.
  • After HCC treatment. AFP should fall toward normal after curative resection, transplant, or ablation. A rise from nadir is a recurrence signal, often preceding imaging changes.
  • During germ cell tumor treatment. AFP has a known half-life of 5 to 7 days. Failure to fall at the expected rate during chemotherapy is a recognized signal of inadequate response and may prompt treatment intensification per IGCCCG guidance.

In benign chronic liver disease, AFP can wax and wane with disease activity. A persistently rising trend, especially in cirrhosis, is the worrying pattern. Stable mildly elevated values in known chronic hepatitis are usually fine but warrant tighter imaging surveillance.

Track this biomarker over time in AskAnything.health — upload your lab results and see trends at a glance.

When AFP actually warrants action

  • AFP above 200 ng/mL in a cirrhotic patient. Multiphase CT or MRI of the liver is indicated.
  • AFP above 400 ng/mL plus a liver mass meeting imaging criteria in cirrhosis. Diagnostic of HCC under most international guidelines.
  • Rising AFP from a stable cirrhotic baseline, even within the "normal" range. Image, do not wait for the absolute cutoff.
  • Elevated AFP in a young person with a testicular or ovarian mass. Proceed urgently to oncologic workup. Non-seminomatous germ cell tumors are often curable but treatment is time-sensitive.
  • Failure of AFP to fall at the expected half-life during germ cell tumor chemotherapy. Discuss with oncology.
  • Markedly elevated AFP without a clear cause. Workup for hepatoid variants of GI/pulmonary adenocarcinoma after pregnancy and chronic liver disease are excluded.

What does not warrant action: a borderline AFP of 12 to 18 ng/mL in a patient with known chronic hepatitis and stable disease activity, or an isolated mild elevation ordered as part of an undirected "tumor marker panel" in someone with no cirrhosis, no testicular or ovarian mass, and no pregnancy. Repeat once if uncertain. Do not chase.

This information is for educational purposes only and is not a substitute for professional medical advice. Always consult your healthcare provider about your lab results.

Tests that complete the picture

  • Abdominal ultrasound. Paired with AFP every 6 months in HCC surveillance per AASLD. Neither test alone is adequate.
  • Multiphase CT or MRI of the liver. Diagnostic imaging when AFP rises or surveillance ultrasound is concerning.
  • Beta-hCG and LDH. Paired with AFP in germ cell tumor diagnosis, staging, and monitoring (IGCCCG criteria).
  • ALT, AST, alkaline phosphatase, bilirubin, albumin, INR. Characterize the underlying liver disease that determines AFP interpretation.
  • Hepatitis B and C serologies. Most cirrhotic surveillance candidates are defined by chronic viral hepatitis status.
  • PIVKA-II (DCP). Alternative HCC marker complementary to AFP, used in some surveillance protocols (especially in Asia) to improve sensitivity.

Patterns to recognize

Combinations of values that together point at a specific clinical picture. One number rarely tells the whole story.

HCC suspicion in cirrhosis

  • Cirrhosis on imaging or biopsy
  • AFP >20 ng/mL (especially >200)
  • New liver lesion on US or MRI

In cirrhosis, an elevated AFP plus a new lesion meets non-invasive HCC criteria with classic imaging.

Next: Multiphase contrast MRI or CT; hepatology and HCC tumor-board referral.

Testicular germ cell tumor (non-seminoma)

  • AFP elevated
  • Beta-hCG elevated
  • Young man with testicular mass
  • Possible LDH elevation

Combined AFP and beta-hCG elevation in a man with a testicular mass strongly suggests non-seminomatous germ cell tumor.

Next: Scrotal ultrasound; urology referral; CT chest/abdomen/pelvis for staging before orchiectomy.

HCC surveillance in cirrhosis (routine)

  • AFP every 6 months
  • Abdominal ultrasound every 6 months
  • Trend matters more than single value

Paired AFP plus US is the AASLD-recommended surveillance combination; neither alone is sufficient.

Next: Continue 6-month surveillance; investigate any rising trend or new lesion.

Maternal serum AFP screening abnormality

  • Maternal serum AFP outside reference (typically >2.5 MoM or <0.4 MoM)
  • At 16–18 weeks gestation
  • Interpreted with hCG, estriol, inhibin A

High MoM raises concern for neural tube or abdominal wall defect; low MoM raises chromosomal anomaly risk.

Next: Detailed anatomy ultrasound; maternal-fetal medicine referral; offer cell-free DNA or amniocentesis.

Normal AFP does not rule out HCC

  • AFP <20 ng/mL
  • Cirrhotic patient with new liver lesion
  • Suspicious imaging features

Sensitivity of AFP alone is 40–65% and small early HCC often produces little AFP.

Next: Multiphase contrast imaging; biopsy if imaging non-diagnostic; do not rely on AFP alone.

Frequently Asked Questions

Below 10 ng/mL in non-pregnant adults. Mild elevations (10 to 20) are common in chronic hepatitis and cirrhosis without cancer. AFP is interpreted very differently in pregnancy (reported as multiples of the median for gestational age) and in infants (sky-high physiologic levels that decline through the first year of life).

In the general population, no. Sensitivity for early HCC is too low. AFP is recommended specifically as part of HCC surveillance in patients with cirrhosis, paired with abdominal ultrasound every 6 months per AASLD guidelines. Outside cirrhosis, AFP is not a useful screening test.

The most common benign causes are chronic viral, alcoholic, or autoimmune hepatitis; cirrhosis without HCC; recovery from acute liver injury; and pregnancy. Hereditary tyrosinemia, ataxia-telangiectasia, and hereditary persistence of AFP are rare causes. In someone with chronic liver disease, mild stable elevations often reflect ongoing hepatic regeneration rather than cancer.

In a cirrhotic patient with a liver mass meeting imaging criteria (typical arterial enhancement and washout on multiphase CT or MRI), AFP above 400 ng/mL is generally accepted as diagnostic of HCC under international guidelines, often without a need for biopsy. Outside this setting, without cirrhosis or imaging findings, very high AFP can also reflect non-seminomatous germ cell tumors or rare hepatoid variants of other adenocarcinomas.

AFP is part of the diagnostic, staging, and monitoring workup for testicular cancer alongside beta-hCG and LDH. Elevated AFP indicates a non-seminomatous germ cell component (yolk sac, embryonal, or mixed). Pure seminomas do not produce AFP. The half-life is 5 to 7 days. Failure of AFP to fall at the expected rate during chemotherapy is a recognized signal of inadequate response.

No. Sensitivity of AFP alone for HCC is roughly 40 to 65 percent at standard cutoffs, and small early-stage tumors often produce little or no AFP. This is why HCC surveillance in cirrhosis pairs AFP with abdominal ultrasound every 6 months. Neither test alone is sufficient. A normal AFP in a patient with cirrhosis and a new liver lesion does not rule out HCC.

AASLD recommends AFP every 6 months alongside abdominal ultrasound for all patients with cirrhosis, regardless of cause. Some centers use shorter intervals (3 to 4 months) in higher-risk populations such as hepatitis B carriers without cirrhosis but with risk factors, or patients on the transplant list. The trend matters more than any single value.

Maternal serum AFP is part of second-trimester screening for neural tube defects (open spina bifida, anencephaly) and certain abdominal wall defects. It is reported as multiples of the median (MoM) for gestational age and is interpreted alongside other markers (hCG, estriol, inhibin A) in quad-marker screening. This is a completely different clinical workflow from oncology AFP and should not be conflated.

Track your lab results over time

Upload your blood work and see trends, reference ranges, and AI-powered insights — all in one place.

Get Started

Not medical advice. AskAnything.health is an AI-powered second-opinion tool designed to help you understand your health data. It does not diagnose, treat, or replace professional medical care. Always consult a qualified healthcare provider before making medical decisions. Your data is processed securely and never shared with third parties — see our Privacy Policy.