Autoantibodies and tumour markers

A useful way to think about this page is that blood tests for autoantibodies and tumour markers answer a different question from most laboratory tests. A full blood count or a urea and electrolytes panel asks "what is happening right now"; an autoantibody panel asks "is the immune system attacking a particular tissue"; a tumour marker asks "is a known or suspected cancer producing a measurable substance". Both are adjuncts: they refine probabilities set by the history, examination and first-line investigations, and they are only as good as the question being asked.

This topic has no heading in the standard syllabus, so it is arranged around how the laboratory reports the results: pattern, number and direction. The reference ranges are worked examples, not a formulary. The reading rule for every page of a result sheet applies here: the laboratory's own reference interval, stated on the report, always beats the number printed in any textbook.

Three habits matter at each result.

• Look at the pattern first, the number second. An antinuclear antibody (ANA) titre of 1 to 320 with a centromere pattern means something different from the same titre with a dense fine speckled pattern. A raised CA 125 in a 55-year-old woman with ascites and a pelvic mass is a different result from the same CA 125 after she has had surgery for ovarian cancer.
• Take the result together with the organ it comes from. Each antibody points to a target antigen and a disease family: anti-double-stranded DNA (anti-dsDNA) to systemic lupus erythematosus, anti-tissue transglutaminase IgA (anti-tTG IgA) to coeliac disease, proteinase 3 ANCA to granulomatosis with polyangiitis. None is an organ-specific diagnosis on its own.
• Expect false positives. Autoantibodies are common in healthy people and in other diseases; tumour markers are elevated by many non-cancer conditions. Working out why a test is positive is most of the job.

The panels named throughout are those a student meets before finals: the ANA with anti-dsDNA and extractable nuclear antigen (ENA) add-ons, the ANCA screen, rheumatoid factor (RF) with anti-cyclic citrullinated peptide (anti-CCP), antimitochondrial antibody (AMA) and anti-smooth muscle antibody (ASMA), tissue transglutaminase with IgA, complement C3 and C4, immunoglobulins and serum protein electrophoresis, and the seven tumour markers prostate-specific antigen (PSA), CA 125, carcinoembryonic antigen (CEA), alpha-fetoprotein (AFP), beta-human chorionic gonadotrophin (beta-hCG), CA 19-9 and calcitonin.

Reference points for the editions used: the 2010 American College of Rheumatology/European Alliance of Rheumatology Associations (ACR/EULAR) classification criteria for rheumatoid arthritis, the 2019 EULAR/ACR classification criteria for systemic lupus erythematosus, NICE guideline NG12 (suspected cancer: recognition and referral, last updated 2026), and the ESMO Clinical Practice Guidelines programme for the marker-by-marker indications with calcitonin for medullary thyroid carcinoma in particular. StatPearls on the NCBI Bookshelf remains the standard free review for the laboratory background and is the source of the reference-interval lists used for the immunoglobulins.

Antinuclear antibody (ANA). ANA is a screening test for antibodies that react with components of the nucleus. It is reported in two parts: a titre (for example 1 to 160 or 1 to 320) and a pattern (homogeneous, speckled, centromere, nucleolar, nuclear dots, cytoplasmic). A higher titre in a healthy laboratory pattern matters less than a lower titre with a clearly disease-linked pattern, and the pattern determines which add-on test is run. Antigen-specific tests (anti-dsDNA, ENA) are the confirmatory layer ^[1].

Anti-double-stranded DNA (anti-dsDNA). This is an IgG antibody against native double-stranded DNA. It is highly specific for systemic lupus erythematosus (about 96 percent) but only moderately sensitive (about 52 to 70 percent), so a negative result does not exclude lupus and a positive result strongly supports it ^[2]. Antibody levels track disease activity, especially renal disease, and are used with complement for monitoring rather than for the initial diagnosis alone ^[2].

Extractable nuclear antigens (ENA). ENA is a panel rather than one test. Most laboratory menus include SSA/Ro (Sjogren's syndrome antigen A), SSB/La, Sm (Smith), RNP, Scl-70/topoisomerase I, Jo-1 and chromatin. They are more specific for a particular connective tissue disease but less sensitive than ANA ^[9]. Anti-Sm in particular is 98 to 99 percent specific for lupus ^[9]. The panel that a given laboratory runs is set locally, so the same request may produce different answers in different centres.

ANCA. Antineutrophil cytoplasmic antibodies are requested mainly when a small-vessel vasculitis is suspected. Current guidance is to run assays for the two antigen-specific antibodies first, anti-proteinase 3 (PR3, usually associated with granulomatosis with polyangiitis) and anti-myeloperoxidase (MPO, usually associated with microscopic polyangiitis), because they are more specific and better standardised than the indirect immunofluorescence pattern ^[10]. c-ANCA (cytoplasmic, PR3) and p-ANCA (perinuclear, MPO) remain the pattern names written on reports.

Rheumatoid factor and anti-CCP. RF is an IgM antibody against the Fc portion of IgG. It is modestly sensitive and modestly specific for rheumatoid arthritis, and is also positive in other connective tissue diseases, in persistent infection and in a proportion of healthy older adults. Anti-CCP (anti-cyclic citrullinated peptide) is more specific: it appears before clinical disease and predicts erosive joint damage. In the ACR/EULAR 2010 classification criteria both contribute points according to level (low-positive: above the manufacturer cut-off but below three times it; high-positive: above three times the cut-off) and both carry the same weight ^[3].

Anti-tissue transglutaminase IgA (anti-tTG IgA). This is the first-line serological test for coeliac disease in a patient with normal IgA. It must be run with a total IgA level: in selective IgA deficiency the IgA-based assay gives a false-negative result, and IgG-based serology (anti-tTG IgG or anti-deamidated gliadin peptide) is used instead.

Antimitochondrial and anti-smooth muscle antibodies. AMA is the serological signature of primary biliary cholangitis, directed mainly against the M2 antigen of the pyruvate dehydrogenase complex, with sensitivity that depends on the assay method used. ASMA points to autoimmune hepatitis, particularly when accompanied by hypergammaglobulinaemia and raised transaminases, but is also seen in chronic viral hepatitis and other liver disease, so a positive ASMA is not a diagnosis on its own.

Complement C3 and C4. Complement testing quantifies the consumption and production of C3 and C4 proteins. In immune-complex disease, especially lupus nephritis, C3 and C4 fall as complexes form and activate the classical pathway, and rise again with remission. Values are also lowered by genetic deficiency, loss through renal or gut disease, and some treatments, and raised as acute-phase reactants. Reading complement alongside anti-dsDNA is part of every lupus review visit ^[2].

Immunoglobulins and plasma protein electrophoresis. Immunoglobulins are measured in two ways: quantitative IgG, IgA and IgM, which give a circulating amount, and serum protein electrophoresis, which gives a pattern. Electrophoresis separates plasma proteins into albumin, alpha1, alpha2, beta and gamma fractions. A polyclonal rise in the gamma fraction (a broad-based elevation) is seen in chronic infection, liver disease, connective tissue disease and some immunodeficiencies. A narrow monoclonal band in the gamma or beta region is the hallmark of a monoclonal gammopathy and needs further study with immunofixation and free light-chain assays ^[6].

Tumour markers. The seven markers below are used to monitor a known cancer or to add information during work-up; none of them is a screening test for cancer in the general population.

• PSA (prostate-specific antigen). A 28 to 34 kDa glycoprotein secreted by prostatic epithelial cells. A serum level above 4.0 ng/mL is commonly used as a raised cut-off in clinical practice, though the age- and race-specific normal ranges and the debate about over-treatment of screen-detected disease are well described ^[8].
• CA 125 (cancer antigen 125, MUC16). A glycoprotein associated with epithelial ovarian cancer, with a physiological source in coelomic epithelium (peritoneum, pleura, pericardium). NICE uses age-based thresholds from 35 IU/mL (ages 40 to 49) down to 24 IU/mL (ages 60 to 69) in the referral pathway for suspected ovarian cancer ^[12].
• CEA (carcinoembryonic antigen). An oncofetal glycoprotein raised in many adenocarcinomas, above all colorectal. Manufacturer inserts often quote 2.5 ng/mL in nonsmokers and 5.0 ng/mL in smokers; levels also rise in smoking alone, in inflammatory bowel disease, in cirrhosis and in pancreatitis ^[7].
• AFP (alpha-fetoprotein). An oncofetal protein produced in liver and yolk sac. It is used in the surveillance and diagnosis of hepatocellular carcinoma, in the follow-up of germ cell tumours, and in prenatal screening for neural tube defects and aneuploidy.
• Beta-hCG. The beta subunit of human chorionic gonadotrophin, produced by syncytiotrophoblast and some germ cell tumours. It is the standard marker for gestational trophoblastic disease and for the monitoring of non-seminomatous germ cell tumours, and is raised physiologically in pregnancy.
• CA 19-9 (carbohydrate antigen 19-9). A sialylated Lewis A antigen, raised in pancreatic, biliary and gastric cancer but also in biliary obstruction of any cause and in inflammatory or benign pancreatic disease. It is most useful for monitoring rather than diagnosis.
• Calcitonin. A hormone produced by the parafollicular C cells of the thyroid. Calcitonin (together with CEA) is the working marker for medullary thyroid carcinoma; C-cell hyperplasia and renal failure are among its other causes of elevation. A calcitonin above 100 to 200 pg/mL after surgery is the range in which imaging is recommended to look for residual or recurrent structural disease ^[11].

Reference intervals are assay and method dependent. They come from the literature ranges quoted on page 1 of the report, the laboratory's own in-house verification, and population studies; no two laboratories agree exactly. The principle is that the reference interval printed on your report is the one that applies.

The unit convention: international units are used (g/L for immunoglobulins and complement, ng/mL for PSA and CEA, U or IU per mL for most enzymes and glycoprotein tumour markers), with conventional units in brackets where the US literature and laboratory reports still use them. A student switching between laboratories should not re-memorise the numbers; instead, note that the report explains how the value was measured and against which interval.

The lysing point of the tumour marker section is the inverse of the autoimmune one: tumour markers are easy to raise with non-cancer disease, whereas autoantibodies are easy to raise in healthy people. The direction of the error matters in the clinic: a false-positive autoantibody usually leads to over-investigation and over-treatment, while a false-negative tumour marker in a person already known to have cancer can delay the detection of recurrence.

Named combinations do much of the work on a real result sheet.

• ANA positive with negative anti-dsDNA and ENA, in an otherwise well person. This pattern is common and is rarely clinically relevant. ANA can be positive in infection, in older age and in low titre in a proportion of healthy people; it should not be re-run indefinitely ^[1].
• High ANA titre with positive anti-dsDNA and low C3/C4. This is the lupus pattern. It suggests active systemic lupus erythematosus and is used for monitoring, especially renal activity ^[2].
• ANA positive, ENA positive for SSA with or without SSB. This supports Sjogren's syndrome or cutaneous lupus, and a subset of mothers with SSA pass IgG across the placenta, so neonatal lupus and congenital heart block need to be considered in pregnancy.
• Anti-Sm with anti-dsDNA and anti-RNP. Anti-Sm (98 to 99 percent specific for SLE) together with anti-dsDNA is strongly in favour of lupus; anti-RNP is associated with mixed connective tissue disease when high-titre U1-RNP occurs with high-titre ANA and clinical overlap ^[9].
• Positive ANA with a Scl-70 pattern. A nucleolar ANA with anti-Scl-70 (topoisomerase I) is associated with diffuse systemic sclerosis.
• Centromere pattern. A discrete speckled pattern with centromere staining is associated with limited systemic sclerosis (formerly CREST syndrome).
• PR3 ANCA positive. Typically associated with granulomatosis with polyangiitis; a positive result should be followed up with a clinical and renal assessment and tissue diagnosis where indicated ^[10].
• MPO ANCA positive. Typically associated with microscopic polyangiitis and eosinophilic granulomatosis with polyangiitis, often in people with asthma and eosinophilia.
• RF and anti-CCP together. Double positivity in a person with an inflammatory arthritis strongly supports rheumatoid arthritis and predicts an erosive course. ACPA-positive disease carries the same weight as RF-positive disease in the ACR/EULAR score, although ACPA is more specific ^[3].
• AMA positive in a middle-aged woman with a raised alkaline phosphatase. AMA is the serological marker of primary biliary cholangitis, directed at the M2 antigen of the pyruvate dehydrogenase complex.
• ASMA positive with high globulins. Suggests autoimmune hepatitis, provided viral hepatitis and drug causes have been excluded.
• Anti-tTG IgA positive with normal or low total IgA. High-titre anti-tTG IgA points to coeliac disease in a gluten-eating patient; with a low total IgA, IgG-based testing is needed; with a previously negative gluten-free-diet patient, serology can be negative in a well-treated person ^[11].
• C3 and C4 both low after starting treatment for lupus. Expected normalisation follows disease activity falling; persistently low complement raises the question of a flare, of additional infection, or of a genetic deficiency.
• High IgG with a broad gamma band on electrophoresis. A polyclonal rise, seen in chronic infection, liver disease and connective tissue disease.
• A narrow monoclonal band with a suppressed background. Needs immunofixation and a free light-chain assay; the minimum work-up for a new monoclonal gammopathy.
• Rising PSA after radical prostatectomy. Any rise is a recurrence until proven otherwise; thresholds for salvage radiotherapy are given by prostate-specific guidelines.
• Rising CEA with a normal CT colonography. Look for hepatic metastasis too small to see, or recurrence at the anastomosis; CEA detects rather than diagnoses.
• Rising CA 125 after first-line treatment for ovarian cancer. Suggests recurrence; the decision to re-image and treat is clinical, because CA 125 rises earlier than symptoms and is used in the relapse definitions of the ESMO ovarian cancer guidance.
• Rising AFP after surgery for hepatocellular carcinoma. Suggests residual or recurrent disease, with imaging to follow.
• Rising beta-hCG after evacuation of a molar pregnancy. The standard evidence of persistent or metastatic gestational trophoblastic disease.
• Rising calcitonin after thyroidectomy for medullary carcinoma. Residual disease; localise with imaging, particularly if above about 100 to 200 pg/mL ^[11].

Every value below is invented for teaching and is labelled as an illustrative sample. They are not real patient data, and the reference interval should be taken from the laboratory's own report.

Numbered interpretation:

1. Each individual result is within the reference interval of the reporting laboratory. ^[1]
2. The autoantibody screen is negative and complement is normal, so there is no evidence of active systemic autoimmunity on this panel. ^[2]
3. Total IgA is adequate, so an anti-tTG IgA result can be trusted; a low total IgA would require re-testing with IgG-based serology. ^[3]
4. No further autoantibody testing is indicated without a new clinical question: repeating all of these annually is not part of good practice. ^[1]
5. The panel shows 14 IU/mL RF, which sits close to the cut-off, a reminder that cut-offs are not cliffs and that serology without clinical context is misleading ^[9]. Clinical conclusion: no systemic autoimmune activity is evident from this panel, and the next step is clinical follow-up rather than more antibody testing.

Numbered interpretation:

1. A high-titre homogeneous ANA with a positive anti-dsDNA and low complement is the lupus serological pattern. ^[2]
2. Both C3 and C4 are below their intervals, which argues for classical-pathway consumption from immune complexes rather than an isolated low component. ^[5]
3. Anti-Sm supports the specificity of the SLE pattern: it is 98 to 99 percent specific for lupus. ^[9]
4. A raised total IgG fits an active immune response; it is not specificity evidence on its own.
5. The RF of 42 IU/mL is above the manufacturer cut-off: in lupus this can be a non-erosive polyarthritis pattern rather than rheumatoid arthritis, so it is not a diagnosis. ^[4]
6. The raised urine protein-creatinine ratio turns this into active lupus with renal involvement until proven otherwise; it is the reason the result sheet exists as a clinical prompt. ^[2]
7. High titre anti-dsDNA with low complement plus proteinuria is a flare picture under the lupus monitoring indices, and needs renal function, blood pressure and, if indicated, a kidney biopsy. ^[2] Clinical conclusion: the panel is consistent with active systemic lupus erythematosus with probable renal involvement; the next step is a rheumatology or renal review with a kidney biopsy considered.

Numbered interpretation:

1. A CEA of 4.8 ng/mL exceeds the usual nonsmoker cut-off of 2.5 ng/mL and lies just under the smoker cut-off of 5.0 ng/mL. The first question to ask is the smoking status and the test history. ^[7]
2. A single borderline CEA is not a diagnosis of recurrence: sensitivity and specificity for CEA alone are poor, and most centres require a rising trend on at least two occasions before acting ^[7].
3. The dose-response shape matters: in a previous post-surgical surveillance series, serial CEA detected recurrent colorectal cancer before imaging, with variable lead times, but acting on one value led to unnecessary imaging ^[7].
4. CA 19-9, PSA, AFP and calcitonin in this fictionally combined panel are all within their laboratory's own interval, so they do not contribute to the interpretation despite their presence on the sheet.
5. A CA 125 of 14 IU/mL is comfortably below the NICE age-based thresholds and does not support ovarian malignancy ^[12].
6. The next step for the raised CEA with a normal cross-sectional imaging is a clinically driven re-test in six to eight weeks with the same laboratory, together with a review of symptoms and signs, before any invasive investigation. ^[7] Clinical conclusion: the patient has a borderline, non-diagnostic elevation of CEA with otherwise normal markers; the next step is a validated rising trend or imaging or clinical evidence of recurrence, not treatment.

• Over-testing. A blanket autoantibody panel requested without a clinical question produces more false positives than useful information; an ANA strategy of stepwise reflex reduces wasted ENA and anti-dsDNA tests without losing cases that matter ^[1].
• Forgetting total IgA with anti-tTG IgA. Selective IgA deficiency is common enough to matter and produces a false-negative coeliac serology if it is not checked.
• Reporting a single tumour marker as a screening test. PSA, CA 125 and CEA are not screening tests in unselected populations; using them as such over-calls disease and under-calls it because sensitivity is modest.
• Ignoring smoker status for CEA. A smoker's own cut-off is 5.0 ng/mL, so a value of 3.5 ng/mL or 4.0 ng/mL means something different depending on whether the patient smokes.
• Treating reference ranges as universal. Each laboratory uses its own assay, its own population and its own statistical cut-off; numbers from a different laboratory cannot be dropped into another report.
• Assuming a rising marker needs immediate treatment. Doubling time and the pattern of change matter more than a single rise, and early biochemical-only treatment has not improved survival in all tumour types.
• Misreading complement direction. A high C3 does not mean the patient is cured: complement rises as an acute-phase reactant even when immune activity is ongoing.
• Confusing PSA units with mass. PSA is a secreted glycoprotein measured in ng per mL; clinical risk reclassification is based on PSA density, PSA velocity and the clinical context, not the mass alone.
• Reading electrophoresis without immunofixation. A broad gamma rise can mask a small monoclonal band, and a band can be over-called on a low-quality track.
• Overlooking non-cancer causes of AMA. Although most AMA-positive patients have primary biliary cholangitis, the result comes after, not before, clinical and biochemical assessment.
• Believing an ANA pattern is disease-specific. ANA patterns suggest, rather than diagnose, a disease and are subject to assay and reader variation.
• Forgetting that beta-hCG exists in a successful pregnancy. A tumour marker result in a woman of reproductive age has to be framed against pregnancy first.

• Autoantibodies point at targets and at disease families; they are probability-modifiers, not diagnoses on their own.
• ANA is a screen (pattern plus titre); anti-dsDNA and ENA are the confirmation.
• ANCA is now requested as PR3 and MPO antigen-specific tests first, with the immunofluorescence pattern as backup.
• RF is a screening antibody; anti-CCP is specific and predicts erosive disease.
• Anti-tTG IgA is the first-line test for coeliac disease but only works with a total IgA.
• AMA marks primary biliary cholangitis; ASMA marks autoimmune hepatitis; both need clinical correlation.
• Low C3 and C4 with active disease are the read-out of immune-complex consumption.
• Immunoglobulins and electrophoresis distinguish a polyclonal rise (broad gamma) from a monoclonal band.
• Tumour markers are used for monitoring known cancer and for context in work-up; they are not general screening tests.
• PSA above 4 ng/mL is the commonly quoted raised cut-off; age-specific ranges and over-treatment are important caveats.
• CA 125 uses age-specific thresholds in NICE guidance, from 35 IU/mL down to 24 IU/mL.
• CEA is 2.5 ng/mL in nonsmokers and 5.0 ng/mL in smokers as the insert-based cut-off, and rises with many non-cancer conditions.
• CA 19-9 is misleading as a diagnostic test (Lewis-negative patients produce none at all) and is used to monitor.
• Calcitonin is the marker of medullary thyroid carcinoma; above 100 to 200 pg/mL after surgery it directs imaging.
• The laboratory's own reference interval, method and units always have priority over the numbers on a page.