Evaluation of Psoriasis Area and Severity Index as a Proxy for Biomarkers of Systemic Disease under Treatment with Tumour Necrosis Factor-alpha and Interleukin 12/23 Antagonists in Patients with Psoriasis: A Retrospective Cohort Study of 186 Treatment Cycles

Jochen H.O. Hoffmann, Christian Knoop, Knut Schäkel, Alexander H. Enk, Eva N. Hadaschik
DOI: 10.2340/00015555-3814

Abstract

The efficacy of psoriasis treatments is usually evaluated using the Psoriasis Area and Severity Index (PASI). However, there is a lack of systematic statistical assessments of PASI as a proxy for systemic disease in individual patients. Therefore, a retrospective study of 186 treatments with adalimumab, etanercept, and ustekinumab for psoriasis (341 patient-years) was performed. While PASI significantly and independently correlated with biomarkers of systemic inflammation (especially neutrophil-to-lymphocyte ratio, C-reactive protein), the strengths were only weak-to-moderate and varied considerably inter-individually. A decrease in PASI indicated a neutrophil-to-lymphocyte ratio decrease and a C-reactive protein decrease or stable low margin C-reactive protein in ≥ 80%. Sensitivity, specificity, and positive predictive value of PASI 0 and PASI 2.75 (optimal Youden Index) for low cardiovascular risk C-reactive protein were 24%, 92%, 85%, and 62%, 61%, 76%, respectively. Performance was similar using absolute thresholds and PASI 100 or PASI 75, and overall worse for low cardiovascular risk neutrophil-to-lymphocyte ratio and if psoriasis arthritis was present. In conclusion, PASI allows robust low-order estimates of systemic inflammation, but cannot substitute for laboratory biomarkers for more precise assessments.

Significance

The efficacy of psoriasis treatments is usually evaluated by cutaneous response (Psoriasis Area and Severity Index; PASI). However, it is unclear how precise PASI improvements reflect systemic manifestations in individual patients. This study retrospectively analysed PASI and inflammatory biomarkers during 186 treatments (mean 22 months) of patients with psoriasis with biologics. While PASI decrease and complete cutaneous clearance indicated a decrease in inflammatory biomarkers and low cardiovascular risk levels of 1 of 2 inflammatory biomarkers, respectively, no single PASI threshold was both sensitive and specific enough for more precise assessments of systemic inflammation. Thus, further research into laboratory biomarkers as valid additional treatment goals would be desirable.