Three Verdicts on the Same New Psoriasis Pill

On August 8, the Journal of the American Academy of Dermatology published a paper promising an answer to a question dermatologists have asked since March: how does icotrokinra, the first oral pill built to block the IL-23 receptor, compare to the injectable IL-23 inhibitors already on the market. It is titled a comparative meta-analysis. Before a single number in it is read, one fact matters: no randomized trial has ever put icotrokinra head to head against an injectable IL-23 drug.

Icotrokinra, sold as ICOTYDE, won FDA approval in March 2026 as a once-daily tablet, developed by Protagonist Therapeutics and Janssen. Its pivotal ICONIC trials tested it against placebo, and in two studies, against deucravacitinib, an oral drug that works through a different pathway. None of the ICONIC trials enrolled a comparison arm using risankizumab, guselkumab, or any other injectable IL-23 blocker. Every claim about how icotrokinra stacks up against those drugs is built after the fact, by lining up separate trials and adjusting statistically for differences between their patients.

Two analyses, two different answers

That indirect method has already produced conflicting verdicts. A network meta-analysis presented at the Maui Derm Hawaii conference, published through Janssen's own medical information channel, described icotrokinra's rate of completely clear skin as comparable to risankizumab, guselkumab, secukinumab, and ixekizumab, and better than ustekinumab, adalimumab, and deucravacitinib.

A separate matching-adjusted indirect comparison, published in Dermatology and Therapy with AbbVie's involvement, the maker of risankizumab, reached the opposite impression. Reweighting risankizumab's individual patient data to resemble icotrokinra's trial population, the authors reported risankizumab ahead at week 16 on every measure: placebo-adjusted PASI 75 of 80.6 percent versus 61.7 percent, PASI 90 of 71.4 percent versus 49.9 percent, and complete clearance of 42.3 percent versus 29.4 percent, with the gap continuing through week 52 in a later, unanchored analysis.

Neither paper is a trial. Both are statistical reconstructions of trials never run against each other, built or reviewed by the company whose drug came out ahead.

The new JAAD paper, from Sana Gupta, Brendan Amoyaw, Grace Xiong, and Mohannad Abu-Hilal of McMaster University, is at least the sixth meta-analysis or systematic review of icotrokinra's trial data to appear in a journal in 2026, after similar papers in Inflammopharmacology, Frontiers in Immunology, and the International Journal of Dermatology. Most compared icotrokinra only to placebo. This one returns to the harder question the industry-linked analyses already answered two different ways.

What the method can settle, and what it cannot

A network meta-analysis and a matching-adjusted comparison are legitimate tools in a market where head-to-head trials are slow and expensive. Neither substitutes for one. No patient has been randomly assigned to icotrokinra or an injectable IL-23 inhibitor and followed under identical conditions. The reported percentages are reconstructions, sensitive to which trials are pooled and who funded the reweighting.

For a reader weighing a pill against an injection, the useful question is not which paper to believe. It is what trials fed the numbers being quoted, who paid for the comparison, and whether a direct trial is actually planned before the choice gets called settled.