Adquey (Difamilast): The Gentlest PDE4 Inhibitor Gets Its FDA Moment

A Third PDE4 Inhibitor. But This One Has Receipts.

On February 13, 2026, the FDA approved Adquey (difamilast 1% ointment) for the treatment of mild-to-moderate atopic dermatitis in patients aged 2 years and older.¹ It is the third topical PDE4 inhibitor to reach the U.S. market, following crisaborole (Eucrisa) in 2017 and roflumilast (Zoryve) in 2024.

If your first reaction is "do we really need a third one?" — fair question. But difamilast isn't just another molecule in the same class. It's a drug that has already been used in Japan for over three years under the brand name Moizerto, giving us something rare in dermatology: a substantial body of real-world data before U.S. launch.²

The manufacturer is Acrotech Biopharma, a subsidiary of the Indian pharmaceutical company Aurobindo Pharma, working in collaboration with Otsuka Pharmaceutical, who developed the original compound in Japan.³ This is Aurobindo's first innovative NDA approval in dermatology — and they've entered the market with a drug that has more supporting data than most new topicals arrive with.

The indication is specific: mild-to-moderate atopic dermatitis. Not severe. Not psoriasis. This is a targeted approval for a targeted drug, and the data backing it is worth examining in detail.

PDE4B Selectivity: The Reason This One Might Not Burn

All three topical PDE4 inhibitors work by the same general principle: inhibiting the enzyme phosphodiesterase-4, which normally breaks down cyclic AMP (cAMP) inside immune cells. When you block PDE4, cAMP levels rise, and the production of pro-inflammatory cytokines — IL-4, IL-13, IL-31, TNF-α — drops. These are the exact cytokines that drive atopic dermatitis inflammation.⁴

But PDE4 isn't one enzyme. It comes in four subtypes: PDE4A, PDE4B, PDE4C, and PDE4D. And this is where difamilast's story gets interesting.

Difamilast selectively inhibits PDE4B over PDE4D, with IC50 values of 0.0112 μM for PDE4B versus 0.0738 μM for PDE4D — a 6.6-fold selectivity ratio.⁴ This matters because PDE4B is primarily responsible for the anti-inflammatory effects in skin, while PDE4D is the subtype most associated with gastrointestinal side effects (nausea, emesis) and, potentially, the local irritation that has plagued earlier PDE4 inhibitors.

Crisaborole, by contrast, inhibits PDE4 more broadly. And if you've ever prescribed crisaborole — or used it — you know the complaint: it stings. Application site pain was one of the most commonly reported adverse events in crisaborole trials, and in clinical practice, it's the reason a meaningful number of patients discontinue treatment.⁵

Recent research has also identified additional mechanisms beyond simple cAMP elevation. Difamilast appears to inhibit ERK phosphorylation in activated basophils, reducing IL-4 production through a pathway distinct from general PDE4 inhibition. It also stimulates production of soluble ST2 via the aryl hydrocarbon receptor (AHR)-NRF2 axis, which helps inhibit IL-33 activity — a key driver of the itch-scratch cycle in atopic dermatitis.⁶

The pharmacology here is elegant. The question is whether it translates to meaningful clinical differences. The trials say yes.

Multiple Phase 3 RCTs, a Meta-Analysis, and 52-Week Safety Data

This is where Adquey separates itself from many new dermatology launches. The clinical program is substantial: a Phase 2 trial across 30 centers, two Phase 3 randomized controlled trials (adults and pediatric), a 52-week long-term safety study, an infant safety study, and a meta-analysis pooling data from five RCTs covering 1,009 patients.^7,8,9,10,11

Three PDE4 Inhibitors, Two Calcineurin Inhibitors, One JAK Inhibitor, and Steroids

The mild-to-moderate atopic dermatitis market is now crowded with non-steroidal options — which is a good thing for patients, even if it creates prescribing complexity. Here's where difamilast fits.

Versus crisaborole (Eucrisa): The most direct comparison. Both are topical PDE4 inhibitors, both approved for ages 2+, both twice daily. The key difference is tolerability. Crisaborole's application site pain (burning, stinging) has been its Achilles heel since launch. Difamilast's PDE4B selectivity appears to meaningfully reduce this problem. In the Phase 3 trials, application site reactions with difamilast were minimal and rarely led to discontinuation.^5,7 For the substantial population of patients who tried crisaborole and quit because it hurt, difamilast is the obvious next step.

Versus roflumilast (Zoryve): The newest PDE4 inhibitor before difamilast. Roflumilast has a key advantage — once-daily dosing versus difamilast's twice daily — which matters for adherence. However, difamilast has a broader age indication (2+ versus roflumilast's initial 6+ for the 0.15% cream), and the longer real-world track record from Japan.¹²

Versus topical corticosteroids: Steroids remain first-line for acute flares, and nothing here changes that. Difamilast's value is in the spaces where steroids are problematic: long-term maintenance therapy, sensitive skin areas (face, eyelids, skin folds), and pediatric patients where parents are understandably anxious about steroid side effects. Difamilast showed zero skin atrophy, zero telangiectasia, and zero striae over 52 weeks of continuous use — which is the promise that non-steroidal options are supposed to deliver.¹⁰

Versus topical calcineurin inhibitors (tacrolimus, pimecrolimus): These have occupied the non-steroidal niche for two decades. Difamilast offers comparable efficacy with potentially better local tolerability — calcineurin inhibitors also cause burning and pruritus at the application site in many patients.¹³

Versus ruxolitinib (Opzelura): The topical JAK inhibitor shows numerically higher efficacy in trials (IGA success around 50–54% at 8 weeks), but it's approved only for ages 12+ and represents a different mechanism class with different safety monitoring considerations. Difamilast and ruxolitinib are not interchangeable — they occupy different tiers of the treatment algorithm.¹⁴

Three Years of Moizerto: What the Real World Shows

This is the part of the story that makes difamilast unusual. Most new drugs arrive in the U.S. with Phase 3 trial data and a prayer. Difamilast arrives with over three years of real-world clinical use in Japan, where it launched as Moizerto in June 2022 after receiving approval in September 2021.²

Otsuka Pharmaceutical developed the compound and holds the Japanese rights. In Japan, both the 1% and 0.3% ointment strengths are available — the lower strength used primarily in younger pediatric patients. The clinical experience has been extensive enough to generate long-term extension studies that most U.S. approvals can only promise.

The 52-week Japanese data is particularly important because it answers the question that Phase 3 trials can't: does this drug keep working? The answer appears to be yes. EASI-75 cumulative success rates climbed steadily over 52 weeks — 55.4% in adults and an impressive 73.5% in children — with no evidence of tachyphylaxis or accumulating adverse events.¹⁰

Perhaps most remarkably, a Phase 3 study in Japanese infants aged 3–24 months (n=41) showed the drug was well-tolerated even in this youngest population, with a 100% completion rate for the primary evaluation period and a mean treatment duration of 344 days.¹⁵ Only one infant (2.4%) discontinued due to an adverse event. This data, while not part of the U.S. approval (which starts at age 2), suggests the safety profile extends younger than the current label.

No Black Box Warning. Minimal Stinging. No Skin Atrophy.

The safety data for difamilast is, by any reasonable standard, reassuring.

No black box warning. Unlike the historical debate around topical calcineurin inhibitors, difamilast carries no FDA boxed warning and no special REMS requirements.¹

Application site reactions are minimal. This is the headline for clinicians who've watched patients quit crisaborole. In the Phase 3 trials, burning and stinging were rarely reported with difamilast and almost never led to discontinuation. The most common adverse event across trials was nasopharyngitis — which is about as generic a finding as you can get in a clinical trial.^7,8

No corticosteroid-class side effects. Over 52 weeks of continuous use, there was zero skin atrophy, zero telangiectasia, and zero striae. The plasma concentrations remained in the nanogram range, indicating minimal systemic absorption. No HPA axis suppression was observed.¹⁰

In the 52-week long-term study, 72.3% of adults and 89.0% of pediatric patients experienced at least one treatment-emergent adverse event — but these were overwhelmingly mild-to-moderate, and the high pediatric rate largely reflects the expected frequency of common childhood illnesses (upper respiratory infections, nasopharyngitis) over a year of observation, not drug-related toxicity.¹⁰

An FDA Approval That Earned Its Rating