EBM Focus: The Dupilumab Dilemma – Trading Eczema for Psoriasis?

Practice Point: Clinical resolution of moderate-to-severe eczema with dupilumab comes at a cost: an increased risk of developing psoriasis.

EBM Pearl: Don’t let yourself (or your patients) be led astray by mistaking a relative measure of effect for absolute risk. In the case of dupilumab for eczema, the 60 percent increased risk of developing psoriasis equates to an absolute risk increase of one percent.

Dermatologists, close your ears. In primary care, rashes without a slam-dunk clinical diagnosis often follow the treatment sequence of a) topical steroid followed by b) topical antifungal followed by c) referral to Dermatology if the rash isn’t gone by then. Eczema, also known as atopic dermatitis, is no exception. 

Eczema is commonly seen in primary care offices, but many patients with moderate-to-severe eczema still end up seeing a dermatologist who will prescribe more powerful agents such as biologics, corticosteroids, methotrexate, conventional immunosuppressants, and/or ultraviolet light therapy. The biologic dupilumab has been FDA approved to treat moderate-to-severe eczema since 2017 and is strongly recommended for this indication because of its high potential for complete symptomatic resolution (NNT of 3-‌5 at 16 weeks). The same cannot be said for other systemic treatments, which on average offer moderate clinical improvement (but not resolution) in about 60 percent of people treated.

However, new cases of psoriasis seemed to be popping up in patients taking dupilumab for eczema as well as other indications -- an adverse event not observed during clinical trials. Because eczema and psoriasis are driven by distinct, and at times opposing, immune pathways, it was thought that the conditions would not coexist in the same person. However, stranger things have happened, and both conditions have been documented in a small subset of patients. To sort this out, investigators designed a population-based retrospective cohort study.

This recent study published in JAMA Dermatology used data from the TriNetX Global Collaborative Network to identify over 200,000 patients diagnosed with moderate-to-severe eczema, who were then followed for three years. Approximately 90,000 patients were eligible for propensity-score matching based on age, sex, race, comorbidities, laboratory findings, and prior antihistamine use. Following propensity matching, two well-balanced cohorts were created, each consisting of around 10,000 patients: One group received dupilumab, and the other group was treated with nonbiologic systemic agents, including corticosteroids, methotrexate, cyclosporine, azathioprine, or mycophenolate mofetil. This design aimed to mimic the conditions of a randomized controlled trial as best it could and allowed for a quasi-causal inference.

Results demonstrated a three-year cumulative incidence of psoriasis in 2.9 percent of patients taking dupilumab vs. 1.8 percent of those treated with other systemic agents (hazard ratio 1.58, 95% CI 1.25-‌1.99). This translates to an NNH of 94. The risk of developing a new, potentially worse, condition is kind of a big deal. But how big? Given an NNT of 3-5, for every 25 patients who achieve resolution of eczema with dupilumab, one patient will develop psoriasis. That’s quite a trade-off. On a population level, a high NNH paired with a low NNT gets a green light for continued use. But for the patients who develop iatrogenic psoriasis, the likelihood of harm is 100 percent. That’s the tricky part. And without more long-term data, we don’t know whether the psoriasis will recede with cessation of the drug or become a chronic condition with all its known morbidities.

While we can’t say with complete certainty that dupilumab causes psoriasis in people who take it for moderate-to-severe eczema, the propensity-matching executed in this study gets us close to being able to draw that conclusion. And while close usually only counts in horseshoes and hand grenades, sometimes it’s the best we get in clinical medicine. In this case, the key is to discuss the potential for harm with patients before prescribing this medication, and to speak specifically in terms of absolute risk. From a patient’s perspective, hearing about a nearly 60 percent increased risk of psoriasis hits differently than being told your risk goes up from two percent to three percent. Some may be willing to take that risk, but we better be sure our patients understand the actual risk before agreeing to the treatment, particularly when other options are available.

Reference: JAMA Dermatol. 2025 Jun 18:e251578 early online

Log in to DynaMedex to read more about Atopic Dermatitis.