Early axSpA diagnosis showed less spinal damage — but only after ten years

Design

Longitudinal registry analysis with two-reader adjudicated radiographic scoring and four-parameter logistic progression modelling

Population

376 patients with axial spondyloarthritis and 1,181 radiograph sets, Swiss Clinical Quality Management registry 2004-2024; 138 early, 238 established; mean follow-up 11.5 years

Primary outcome

Long-term modified Stoke Ankylosing Spondylitis Spinal Score from time of diagnosis

Effect

-3.85 mSASSS units for early versus established (95% CI -7.28 to -0.42); -5.70 units after adjustment for baseline damage (-9.31 to -2.09); difference evident at 10-15 years

Short-term studies have found no difference in outcome between early and established axial spondyloarthritis once biologic treatment starts, which has quietly weakened the argument for chasing early diagnosis. This longitudinal registry analysis explains why those studies found nothing: they did not look for long enough.

Patients in the Swiss Clinical Quality Management registry with at least two sets of spinal radiographs between 2004 and 2024 were scored with the modified Stoke Ankylosing Spondylitis Spinal Score by two readers with an adjudicator for disagreement — 376 patients and 1,181 radiograph sets, 138 diagnosed within two years of axial symptom onset and 238 later. Baseline characteristics were similar, 61% male and 69% HLA-B27 positive, with mean follow-up of 11.5 years (SD 8.6). Progression was modelled with a four-parameter logistic curve to accommodate the non-linear way mSASSS accumulates.

Early-diagnosed patients had 3.85 fewer mSASSS units of long-term spinal damage (95% CI -7.28 to -0.42), and 5.70 units fewer after adjusting for baseline radiographic damage (95% CI -9.31 to -2.09). The difference became evident ten to fifteen years after diagnosis.

That timeline is the actionable content. A referral pathway justified on a two-year outcome will never show a benefit, and a delay of two or three years in a thirty-year-old feels absorbable in clinic. This puts a structural cost on it that arrives in their forties — which is the argument to make to the primary care pathway that holds most of that delay.

• Make the case for early referral on a ten-to-fifteen-year horizon, not a two-year one
• Adjusting for baseline damage widened the gap rather than narrowing it
• Early is defined as two years or less of axial symptoms, by the ASAS definition
• This is registry data under standard care, not a randomised comparison of referral timing
• The delay is usually in primary care; that is where the pathway change has to happen

Why it matters

It answers why short-term studies found no benefit from early diagnosis — they stopped before the difference appeared.

Don't overread it

Registry data under standard care; early diagnosis is associated with less damage but was not randomly assigned.

The statistics, in plain English

The unadjusted difference of 3.85 mSASSS units has a confidence interval reaching -0.42, so it only just clears no effect; the baseline-adjusted estimate of 5.70 units (-9.31 to -2.09) is more convincing. mSASSS runs from 0 to 72, so several units is a real but not dramatic difference in radiographic terms. And this is observational: patients diagnosed early may differ in ways the registry did not capture, including how obviously inflammatory their presentation was.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.