Autoantibodies Before Symptoms: What to Order and What to Ignore

Drew, I ordered an ANA on a 42-year-old with fatigue, no physical findings, kind of as a screen. It came back positive at 1:160. Now what?

First — ANA at 1:160 is positive in about 5% of healthy adults. Without specific symptoms or signs, that titer alone tells you very little. The question is what's driving the immune system to produce it.

So walk me through what's supposed to stop that from happening in the first place.

Two checkpoints. Central tolerance happens in the thymus — T cells that bind strongly to the body's own proteins are deleted there. The gene AIRE drives that process by forcing the thymus to display proteins from distant organs, like insulin and thyroid antigens. B cells get a parallel audit in bone marrow.

And the ones that slip through?

Peripheral tolerance. Regulatory T cells — a suppressive CD4 subset that requires the transcription factor FOXP3 to develop — actively restrain self-reactive cells in tissues. Lose AIRE and you get multi-organ autoimmunity. Lose FOXP3 and you get IPEX: early type 1 diabetes, severe enteropathy, worse and faster. Both monogenic, both prove the architecture.

What about the HLA variants patients bring in from consumer genomics reports?

HLA — the human leukocyte antigen loci, which encode the proteins that present peptides to T cells — contributes the largest single genetic signal in most autoimmune diseases. HLA-DQ2 and DQ8 in celiac, HLA-DR4 in rheumatoid, HLA-B27 in ankylosing spondylitis. But carrying a risk allele is not a prediction. Most carriers never develop disease. Do not act on an HLA result alone.

So back to my ANA patient. She doesn't have a rash, no joint swelling, normal complement. Do I reflex to anti-dsDNA, anti-Sm?

Yes, with intent. Anti-dsDNA and anti-Sm are highly specific for lupus — guideline-endorsed by EULAR and ACR. If those are negative and she has no clinical criteria, you monitor, not diagnose. Pre-clinical antibody windows can precede SLE by years, but an antibody alone without criteria is not SLE.

What about the morning-stiffness patient with normal rheumatoid factor?

Order anti-CCP — anti-cyclic citrullinated peptide antibody. It's positive in roughly 70% of RA patients, has specificity above 95%, and ACR and EULAR include it in the 2010 classification criteria. RF alone misses a significant fraction of early RA. Normal RF is not reassuring enough to stop there.

And for the celiac family — does the child need HLA testing or antibody testing?

HLA-DQ2 or DQ8 negativity effectively rules out celiac — negative predictive value near 99%, guideline-supported by ACG and BSG. If negative, stop. If positive, test anti-tissue transglutaminase IgA on a gluten-containing diet. Sequence matters.

Is there anything to do with a pre-clinical positive beyond monitoring?

For type 1 diabetes specifically, TrialNet data show that two or more islet autoantibodies — against GAD65, IA-2, or ZnT8 — confer roughly 50% five-year progression risk. Teplizumab is FDA-approved to delay Stage 3 onset in Stage 2 disease, meaning antibody-positive with dysglycemia. That's the one scenario where pre-clinical detection changes a drug decision today.

What's the single thing you'd want me to carry out of this?

An autoantibody in an asymptomatic person is a risk marker, not a diagnosis. Reflex to the specific, high-predictive-value antibody for the disease you're actually worried about, apply classification criteria before you name a disease, and only intervene where approved therapy exists for that pre-clinical stage — which right now means teplizumab in Stage 2 type 1 diabetes and essentially nothing else.