No Blood Biomarker Is Ready to Pick Psoriasis Therapy

Choosing a systemic therapy for psoriasis still runs on trial and error, and patients often cycle through more than 1 treatment before something works. A systematic review and meta-analysis of 26 studies published in Journal of Clinical Medicine looked for a blood test that could shorten that process and did not find one ready to use.¹

Why a Predictive Test Would Matter

The stakes are visible in how differently drugs perform head to head. In the phase 4 IMMpactful trial, 57.3% of biologic-naive adults with moderate psoriasis reached PASI 90 on risankizumab at week 16 against 22.9% on deucravacitinib, a gap of more than 34 percentage points between 2 approved first-line systemic options.² A test that identified in advance which patients fall on which side of that gap would change the sequencing question entirely, the study authors explained.

Psoriasis is driven principally by the interleukin (IL)-23/T helper (Th) 17 pathway, which gives biomarker research an obvious place to look. The review searched MEDLINE and Web of Science in August 2025, registered its protocol with PROSPERO, and limited inclusion to original studies from the prior decade evaluating circulating blood-derived biomarkers against cutaneous response to approved systemic therapies.¹

From 1118 records, the review narrowed to 926 after deduplication, assessed 412 full texts, and included 26 studies. Half evaluated tumor necrosis factor inhibitors, 9 evaluated IL-17 inhibitors, 8 evaluated IL-23 inhibitors, and 2 each covered apremilast and methotrexate, with some studies spanning more than 1 class. Response was assessed anywhere from 4 to 100 weeks after initiation, against thresholds ranging from PASI 50 to PASI 100.

That heterogeneity shapes everything downstream. Most studies carried moderate to high risk of bias, reflecting small samples, exploratory or post hoc analyses, limited confounding adjustment, multiple testing without correction, and absent external validation. The strongest methodology sat in biomarker analyses nested inside large randomized trials; the weakest in small single-center observational work.

Downstream Markers Recurred, Upstream Ones Did Not

The formal meta-analysis was the thinnest part of the paper. Pooled baseline levels of IL-17A, tumor necrosis factor, IL-6, IL-12, and IL-23 rested on 2 studies and 71 patients per marker and showed no significant difference between responders and nonresponders. The authors are explicit that this is insufficient power rather than evidence of no association, with IL-6 showing a nonsignificant trend toward higher levels in responders (standardized mean difference, –1.03; 95% CI, –2.15 to 0.10; P = .074).

An exploratory pooled-value analysis across more studies produced the recurring signals. Beta-defensin 2 returned the largest (z = 4.63; P < .001), followed by IL-17A (z = 3.27; P = .001) and IL-17F (z = 2.92; P = .003), all downstream effectors of the IL-23/Th17 axis. The upstream cytokine IL-23 itself did not reach significance (z = 1.88; P = .060). One biologically plausible reading is that psoriatic inflammation is compartmentalized in skin, so circulating upstream cytokines may not reflect lesional activity while amplified downstream products do.

The authors caution against reading those z-scores as effect sizes. The synthesis combined therapies with distinct mechanisms, different PASI thresholds, and assessment points spanning induction through long-term follow-up, and the weighting method favored larger studies over larger effects. Studies with nonconcordant directions could not be folded in at all, which may bias the pooled signal toward markers with directionally consistent published results.

Not Ready for the Clinic

“At present, soluble circulating biomarker-guided treatment selection in psoriasis cannot be recommended outside a research setting,” the authors wrote.

Their reasoning is specific rather than reflexive. No marker in the review satisfied all 3 requirements for clinical adoption: analytic validity, clinical validity, and clinical utility. Standardized assays do not exist, clinically meaningful cutoffs have not been established, prospective validation cohorts are absent, cost-effectiveness has not been analyzed, and no evidence shows that biomarker-guided prescribing improves outcomes. Publication bias and selective reporting cannot be excluded given the predominance of small studies and subgroup analyses.

Requests to cover biomarker panels for psoriasis treatment selection now have a clear answer, at least for the moment: beta-defensin 2 and the IL-17 isoforms are the candidates worth watching, and the validation work that would justify paying for them has not been done. Sequencing, therefore, stays a clinical judgment, made on disease severity, comorbidity, route preference, and access rather than on a laboratory value.

References

1. Villa-Gonzalez JM, Arevalo-Ortega I, Gonzalez-Hermosa MR, et al. Potential blood biomarkers predicting treatment response in psoriasis: a systematic review and meta-analysis. J Clin Med. 2026;15(15):6096. doi:10.3390/jcm15156096

2. Magnolo N, Soung J, Frew J, et al. Risankizumab versus deucravacitinib in adults with moderate plaque psoriasis: 16-week results from the phase 4 IMMpactful trial. Dermatol Ther (Heidelb). 2026;16(7):3415-3429. doi:10.1007/s13555-026-01779-x