Answers are grounded in this paper and cite their sources.
Topical Janus kinase (JAK) inhibitors reduce systemic exposure compared with oral dosing, but they do not eliminate it. Ruxolitinib cream, the current topical benchmark in atopic dermatitis (AD), achieves an Eczema Area and Severity Index 75% improvement (EASI-75) rate of 70.0% at week 8 in TRuE-AD4 and itch relief within 15 minutes, yet its plasma concentrations must still be interpreted against a 281 nM threshold linked in vitro to JAK-mediated myelosuppression. A soft drug — active in skin, designed to be rapidly hydrolysed to an inactive metabolite on entering blood — offers a mechanistically different way to widen that margin, and the approach has been clinically explored with ATI-1777, a topical soft JAK1/3 inhibitor. This study asks what a next candidate would have to look like to be worth making, and how it should be tested before any efficacy claim is entertained. Three analyses were combined. First, a simplified physiologically based pharmacokinetic (PBPK) skin–plasma model (1.5% cream, 200 cm2, 12 h, Potts–Guy permeability) compared ruxolitinib, upadacitinib and a hypothetical soft-drug target profile (molecular weight 420, logD 3.2, plasma esterase clearance 300 L/h). Second, 16 known JAK inhibitors with public structures from PubChem were clustered with RDKit using Murcko scaffolds and Morgan fingerprints (Butina clustering, distance cutoff 0.55). Third, five benchmark drugs — ATI-1777, topical brepocitinib, oral soficitinib, tapinarof and ruxolitinib — were compared on published efficacy, plasma exposure and development status, with three commonly repeated figures corrected against primary sources. In the PBPK model the target profile gave a skin partitioning index of 174, about six times that of ruxolitinib (30), and a plasma area under the curve (AUC) of 20 µg·h/L, about one-fourteenth that of ruxolitinib (273 µg·h/L); upadacitinib was intermediate (index 78, AUC 150 µg·h/L). The 16 inhibitors fell into 13 clusters. The pyrrolo[2,3-d]pyrimidine family was the crowded zone, while deucravacitinib (Tanimoto similarity to ruxolitinib 0.10), filgotinib (0.12), upadacitinib (0.18) and peficitinib (0.22) were the most distant chemotypes. The benchmark analysis shows that novelty can no longer rest on the soft-drug idea alone: it must rest on the combination of soft inactivation, a non-crowded scaffold, JAK1/TYK2 selectivity and measured paraoxonase 1 (PON1) clearance that still holds in low-PON1 donors. A five-step validation chain with a pre-specified PON1 gate is proposed. These are paper-validation and ranking results only; they do not demonstrate that any molecule exists, works, or is safe.
No takes yet. Share an insight, caveat, or question.
Yubo Du (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: