Evaluation of the Physician Global Assessment and Body Surface Area Composite Tool for Assessing Psoriasis Response to Apremilast Therapy: Results from ESTEEM 1 and ESTEEM 2

psoriasis and a history of ≥5% BSA.13,15 In this study, PGAxBSA and PASI were highly correlated with each other (r =0.92; 95% confidence interval [CI] 0.91-0.93), and both had similar moderate correlations with the DLQI (r =0.48; 95% CI 0.44-0.51 vs r =0.47; 95% CI 0.43-0.5).¹³ Consistent with these findings from these 2 analyses of clinical registry data, we found that PGAx- BSA and PASI exhibited similar high correlation at baseline and week 16 and moderate correlations with DLQI at week 16 in the ESTEEM patient population. PGAxBSA was shown to be well correlated with PASI with respect to assessing responsiveness to therapy in patients with moderate to severe plaque psoriasis.¹⁴ In a large phase 3, randomized, placebo-controlled study comparing etancercept or tofacitinib, a high concordance was observed bewteen PGAxBSA and PASI for 50%, 75%, and 90% improvement from baseline.¹⁴ Consistent with these findings, high levels of concordance between PGAxBSA and PASI were noted at the same response cutoff values for apremilast in the ESTEEM patient population. It should be noted that different static PGA scales have been used in reported comparisons of PGAxBSA and PASI. Analyses conducted in the registry settings have used a 6-point (0 [clear] to 5 [severe]) PGA scale to assess erythema, induration, and desquamation.^5,13,15 The PGA used in the ESTEEM study consisted of a 5-point scale (Table 2). Similarly, a 5-point PGA scale was used in the analysis of responsiveness of PGAxBSA and PASI from the phase 3 study of etanercept or tofacitinib in patients with moderate to severe psoriasis.¹⁴ Despite this difference in static PGA scales used, the PGAxBSA score has consistently been shown to correlate well with PASI.^5,13,14 The ability of the PGAxBSA composite tool to provide a simple and consistent measure of global psoriasis disease severity has made it an attractive alternative to PASI. The PASI assessment does not account for incremental changes in disease severity. For example, when assessing disease severity using PASI, a BSA involvement of 1% to 9% will result in a score of 1 whereas a BSA involvement of 10% to 29% will be assigned a score of 2, which on the legs correspond to a BSA variability of 1% to 3% and 4% to 11%, respectively. One limitation of the PASI is its use of a non-linear scale for scoring for area of involvement. As a result, the PASI is insensitive to detecting changes in disease severity at the lower end of the scale, particularly when BSA involvement is <10% because a BSA of 1% to 9% is assigned a single value of 1.⁴ Thus, any change within that category would not be re ected in the PASI score. The PGAxBSA overcomes the limitations of the PASI in detecting changes at the lower end of the scale and thus may be particularly useful in assessing response to therapy patients with milder disease (ie, BSA involvement <10%). Initial analysis of PGAxBSA and PASI in patients with low (0.1%-2.9%), moderate (3.0%-9.9%), and high (≥10%) BSA involvement found that PGAxBSA correlated well with PASI and was sensitive in patients with mild disease (BSA <10%).⁵ The ease of calculating the PGAxBSA score may make this tool useful in the clinical practice as well as the clinical trial setting. Future studies should aim to validate the scale in clinical practice.

This post hoc analysis was limited to patients enrolled in the ESTEEM 1 and ESTEEM 2 studies, which selected individuals with moderate to severe plaque psoriasis with static PGA ≥3 and BSA involvement ≥10%, and thus may not generalize to patients with milder psoriasis. In conclusion, the findings of this post hoc analysis suggest that the PGAxBSA composite tool may be a simpler, alternative tool for the assessment of disease severity and response to therapy with apremilast, compared with the PASI. Based on