Inflammation markers predict early biologic treatment in spondyloarthritis, age shapes psoriatic arthritis care

Patients newly diagnosed with spondyloarthritis who showed clear signs of objective inflammation at their first assessment were substantially more likely to start biologic therapy, while among patients with psoriatic arthritis the strongest predictor was a surprising one: older age made treatment less likely. Those are the central findings of a retrospective cohort study conducted at Skåne University Hospital in Malmö, Sweden, published in BMC Rheumatology, which followed hundreds of patients from diagnosis through the end of 2021 to identify what drives the decision to begin biologic or targeted synthetic disease-modifying antirheumatic drugs, collectively known as bDMARDs and tsDMARDs.

The research team, led by Mohaned A. Hameed and Carl Turesson, identified all consecutive adult patients receiving a new registered diagnosis of either spondyloarthritis or psoriatic arthritis at the Malmö outpatient rheumatology clinic between 2012 and 2016. From an initial pool of 188 psoriatic arthritis cases and 148 spondyloarthritis cases, expert review of medical records by a senior rheumatologist excluded 69 and 28 patients respectively whose diagnoses were not supported by the available data. That left 119 validated psoriatic arthritis patients and 120 validated spondyloarthritis patients whose clinical characteristics at diagnosis could be linked to subsequent treatment starts recorded in the Swedish National Quality Register and complementary case record reviews through December 31, 2021.

The two cohorts differed markedly in their presentation. The psoriatic arthritis patients had a mean age at diagnosis of 49 years, were 49 percent male, and had a median symptom duration of 17 months before diagnosis. Polyarthritis was the predominant phenotype, affecting 69 patients, with asymmetric oligoarthritis in 31 and a predominantly axial phenotype in only about 5 percent. The spondyloarthritis patients were younger, with a mean age of 38 years, and 64 percent were male, but their median symptom duration was far longer at 44 months. Classification criteria were fulfilled by 79 percent of the psoriatic arthritis patients under CASPAR and 73 percent of the spondyloarthritis patients under ASAS, with most of the latter meeting criteria for the axial form of the disease.

Treatment patterns diverged in an instructive way. Roughly 42 percent of each cohort eventually started a first biologic or targeted synthetic drug, but the timing differed sharply. Among the 119 psoriatic arthritis patients, 48, or 41.7 percent, initiated treatment after a median of 15 months from diagnosis, with most starting a tumor necrosis factor inhibitor. Among the 120 spondyloarthritis patients, 49, or 41.5 percent, began a first bDMARD after a median of only 7 months, and every one of them started with a TNF inhibitor. The authors note that the limited effectiveness of conventional synthetic DMARDs against axial inflammation helps explain the faster escalation in spondyloarthritis, since methotrexate and similar drugs often fail to address the needs of patients with axial involvement.

To identify predictors, the researchers applied Cox proportional hazards regression, modeling time to first bDMARD or tsDMARD start in unadjusted, age-sex-adjusted, and multivariable analyses, run separately for each disease. For spondyloarthritis, two features of active inflammation stood out. Patients with elevated CRP levels at diagnosis were more than twice as likely to receive biologic treatment later, with a hazard ratio of 2.68 and a 95 percent confidence interval of 1.44 to 5.01. Those with bone marrow edema on magnetic resonance imaging of the sacroiliac joints had a hazard ratio of 1.94, with a confidence interval of 1.02 to 3.70. In the final multivariable model, both remained independent predictors, and a sensitivity analysis found no statistically significant interaction between them.

The psoriatic arthritis results told a different story. The only significant predictor was age at diagnosis, which showed an inverse relationship with treatment initiation: for each standard deviation increase in age, the hazard ratio for starting a bDMARD or tsDMARD was 0.74, with a 95 percent confidence interval of 0.55 to 1.00. Exploratory analyses dividing patients into age quartiles sharpened the picture. Those in the highest quartile, aged 62 or older at diagnosis, had the lowest probability of ever starting biologic therapy, with a hazard ratio of 0.34 compared with the youngest quartile of 35 years or under. Notably, this age effect was not explained by comorbidity burden, since the Charlson Comorbidity Index showed no significant association with treatment initiation, and no differences emerged by sex, psoriatic arthritis subtype, education level, family background, or country of birth.

The authors suggest several explanations for the age pattern. Prior studies indicate that multimorbidity, common in older populations, may reduce biologic initiation in rheumatoid arthritis, but that mechanism did not hold here. Apprehension about adverse events, particularly infection risk in elderly patients, may play a role, a concern documented across arthritis diseases generally. The authors also point out that prognostication in early spondyloarthritis may be more straightforward than in psoriatic arthritis, with standard biomarkers considered useful for treatment decisions, whereas psoriatic arthritis can present with joint tenderness before obvious signs of arthritis appear and lacks a comparably structured diagnostic process.

The findings align with current management guidance. The authors note that their spondyloarthritis results are consistent with the latest ASAS-EULAR recommendations, which advise that biologic or targeted synthetic therapy should be considered in patients with elevated CRP and or positive sacroiliac joint MRI who have high disease activity and have failed standard treatment. Importantly, however, the study could not measure clinical disease activity scores directly, so the authors caution that the underlying treatment decisions may have been driven by clinical disease activity that was not captured in their data, and the observed associations between inflammatory markers and treatment reflect the decisions actually made in routine practice rather than a tested treatment rule.

The study carries important limitations. The sample was relatively small, which may have limited the ability to detect weaker associations. Different physicians assessed patients, comorbidities relied on registered diagnostic codes without validation, and imaging findings came from radiologists’ routine clinical reports rather than standardized scoring methods. Twelve percent of sacroiliac joint MRIs fell outside a six-month window around diagnosis. Data on disease activity scores, psoriasis severity, body mass index, and other potential predictors were unavailable. In sensitivity analyses restricted to patients with complete clinical data, the age association in psoriatic arthritis did not reach statistical significance, and the bone marrow edema association in spondyloarthritis also lost significance in that subset. The authors emphasize that because multiple comparison correction was not applied, the analyses should be regarded as exploratory, and the single-center setting within Sweden’s equal-access healthcare system may limit generalizability to countries with different healthcare structures.

On one point the results were consistent across both diseases: no associations were found between treatment initiation and sex, education level, family background, country of birth, or comorbidity categories in either cohort. The authors interpret this as suggesting equitable access to biologic therapy within this Swedish patient group, with objective measures of inflammation driving decisions in early spondyloarthritis while the basis for prognostication in early psoriatic arthritis remains more challenging to define.

Subject of Research: Predictors of biologic and targeted synthetic DMARD initiation in newly diagnosed spondyloarthritis and psoriatic arthritis

Article Title: Predictors at diagnosis for initiation of biologic or targeted synthetic disease-modifying anti-rheumatic drugs in patients with spondyloarthritis and psoriatic arthritis

Article References: Hameed, M. A., Exarchou, S., Sharma, A., Bergström, U., Einarsson, J. T., & Turesson, C. (2026). Predictors at diagnosis for initiation of biologic or targeted synthetic disease-modifying anti-rheumatic drugs in patients with spondyloarthritis and psoriatic arthritis. BMC Rheumatology, 10(1), Article 75. https://doi.org/10.1186/s41927-026-00698-w

Image Credits: AI Generated

DOI: 10.1186/s41927-026-00698-w

Keywords: spondyloarthritis, psoriatic arthritis, biologic DMARDs, CRP, bone marrow edema, MRI, Cox regression, treatment equity, rheumatology, Sweden, cohort study, age at diagnosis