A Psoriasis Drug Comparison Gets a Published Rebuttal, Then a Reply, in the Same Journal

The Journal of the American Academy of Dermatology posted a new item online on October 5, 2026: a reply from Christopher Willy Schwarz, Nikolai Loft, and Lone Skov to a critique of their own earlier paper. The posting is the third piece in a chain that started with a study, moved to a rebuttal from Amélie Spangenberg and John Koo, and ends, for now, with this response. Readers who find the reply alone, without the two documents behind it, are missing most of the argument.

The paper that started the exchange ran online in April 2026 and in print that August, in JAAD's volume 95, issue 2, at pages 553 to 555. Schwarz, Loft, and Skov used a change in Danish clinical guidelines as what they called a quasi-randomized natural experiment. From January 2021 to April 2022, Danish guidelines set brodalumab as the default second-line biologic for psoriasis patients without joint involvement after adalimumab failure. From November 2022 on, the default switched to bimekizumab. The authors modeled outcomes as if every patient in the first period had received brodalumab and every patient in the second had received bimekizumab, then compared two-year drug survival between the two groups.

Their result: a 26 percentage point advantage for bimekizumab, with a 95 percent confidence interval of 12 to 41 percent and a p-value under .001. The authors flagged their own limitation in the original text, writing that comparisons built this way might be biased because of differences in populations, design, and analysis plans. A calendar-time split isn't a randomized assignment. Patients treated in 2021 differ from patients treated in 2023: referral patterns change, physician comfort with newer biologics changes, and the mix of patients reaching second-line treatment changes too.

That gap between a natural experiment and a controlled one is the likely terrain of Spangenberg and Koo's critique, published online in July 2026 as DOI 10.1016/j.jaad.2026.07.143. The full text of their letter sits behind JAAD's paywall, so the exact statistical objection they raised isn't confirmed here. The shape of the exchange is confirmed: a study with a self-disclosed design limitation, a published letter contesting it, and a reply from the original authors four months later, on October 4, 2026, under DOI S0190-9622(26)03667-4.

Other drug-survival data on the same two biologics don't line up neatly with a clean 26-point gap. A Danish nationwide register study covering 2018 to 2023 found bimekizumab, the most recently introduced drug in that cohort, holding a 19 percent drug-survival rate, with a maximum follow-up of only 349 days and a sample of 140 bimekizumab patients against 262 on brodalumab. A January 2026 JAMA Dermatology analysis of the Danish DERMBIO cohort, restricted to bio-experienced patients, found bimekizumab, guselkumab, and risankizumab each had a significantly lower two-year discontinuation risk than ustekinumab. Brodalumab's figure in that same analysis wasn't flagged as significantly different from ustekinumab at all, a result that sits oddly next to a 26-point survival gap reported elsewhere.

The two drugs work differently enough that treating one as a stand-in for the other invites its own scrutiny. Brodalumab blocks the interleukin-17 receptor A, a broader downstream target with a boxed warning history tied to suicidality. Bimekizumab inhibits both interleukin-17A and interleukin-17F. A drug-survival model that swaps one for the other at a single calendar date compares two different mechanisms across two different patient eras, not two arms of one trial.

What the October 5 posting settles is narrower than its title suggests. It confirms that the original authors answered Spangenberg and Koo in the same journal, within the same publication year, and that the full argument now runs three documents deep: a study published in April, a rebuttal filed in July, and a response dated four days before this listing went live.