New Topical and Systemic Treatments for Atopic Dermatitis

ABSTRACT

Atopic dermatitis is an inflammatory skin condition characterised by pruritus and a chronic relapsing–remitting course. Previous topical and systemic treatments for atopic dermatitis were broadly immunosuppressive, which limited their long‐term use. Recently, more targeted therapies for atopic dermatitis have been developed. In this review, we discuss the efficacy and safety of new therapies, including topical PDE4 (phosphodiesterase 4) inhibitors, a topical aryl hydrocarbon receptor agonist, topical and systemic JAK (Janus kinase) inhibitors and biologics.

This review summarises new topical and systemic treatments for atopic dermatitis, their relative effectiveness, and important safety data.

Summary.

• There are many new safe and effective treatments available for atopic dermatitis.

• While there is minimal head‐to‐head data, network meta‐analyses allow comparisons of treatments' short‐term efficacy.

• Long‐term comparative efficacy and safety data are notable gaps in the literature.

1. Introduction

Atopic dermatitis is an inflammatory skin condition characterised by pruritus and a chronic relapsing–remitting course. The pathogenesis of atopic dermatitis is complex and involves local epidermal barrier dysfunction and immune dysregulation driven by genetics and environmental factors. Historically, topical and systemic treatments for atopic dermatitis were broadly immunosuppressive and, particularly for systemic treatments, were associated with harmful long‐term side effects [1, 2]. More effective and safe therapies have been developed recently as our mechanistic understanding of atopic dermatitis has improved. In this review, we summarise new topical and systemic treatments for atopic dermatitis, their relative effectiveness, and important safety information.

2. Topical Treatments

Topical treatment options for atopic dermatitis have greatly expanded over the past 25 years. Topical corticosteroids have been the mainstay of therapy for decades and remain important first‐line treatments for most children and adults with atopic dermatitis. However, their clinical benefit is limited in some instances by the potential for local adverse events (AEs), particularly in sensitive anatomic locations such as the face and folds [2]. Additionally, while topical corticosteroids are safe and effective for most people when used properly, patient concerns over AEs may negatively impact adherence [3].

The introduction of the topical calcineurin inhibitors (TCIs), tacrolimus and pimecrolimus, in the early 2000s was a major development for the treatment of atopic dermatitis, giving patients the first non‐corticosteroid topical anti‐inflammatory treatment option. TCIs work by blocking both Th(T helper cell)1 and Th2 immune responses and reduce inflammation and improve the skin barrier without the risk of atrophy [4, 5]. TCIs are not effective for all patients, particularly for more severe lesions of atopic dermatitis, and are limited by application‐site discomfort. Fortunately, the development of TCIs paved the way for other non‐steroidal targeted topical therapies.

2.1. Roflumilast and Crisaborole

In 2016, crisaborole became the first US Food and Drug Administration (FDA) approved topical phosphodiesterase 4 (PDE4) inhibitor for the treatment of mild‐to‐moderate atopic dermatitis (Figure 1, Table 1). Its development was preceded by preclinical studies that showed increased phosphodiesterase activity in lymphocytes cultured from patients with atopic dermatitis [6, 7]. Crisaborole contains a boron atom that binds to the active domain of the PDE4 enzyme, resulting in increased cyclic adenosine monophosphate levels and downregulation of pro‐inflammatory cytokines [8]. Clinically, crisaborole's use has been limited, in part, by application‐site reactions and modest efficacy [9, 10].

FIGURE 1.

(A) Mechanism of action of PDE4 inhibitors. A stimulatory ligand activates a G‐protein coupled receptor, which activates adenylyl cyclase. This results in the production of cyclic adenosine monophosphate (cAMP) from adenosine triphosphate (ATP). Increased levels of cAMP activate protein kinase A (PKA) and the production of anti‐inflammatory cytokines. Phosphodiesterase 4 (PDE‐4) promotes the hydrolysis of cAMP to 5′‐adenosine monophosphate (5‐AMP) and transcription of pro‐inflammatory cytokines. PDE4 inhibitors block the production of 5‐AMP and promote increased cAMP levels. (B) Mechanism of action of Tapinarof. Tapinarof binds and activates the cytosolic aryl hydrocarbon receptor (AhR) and then translocates to the nucleus and heterodimerizes with AhR nuclear translocator (ARNT). This complex promotes expression of anti‐inflammatory cytokines.

TABLE 1.

Comparison of different topical medications used for the treatment of atopic dermatitis.

Topical	Target	Age	Formulation	Frequency of use	Other uses
Crisaborole	PDE4	≥ 3 months	Ointment 2%	Twice daily
Roflumilast	PDE4	≥ 6 years	Cream 0.15%	Once daily	Plaque psoriasis (cream 0.3%) Seborrheic dermatitis (foam 0.3%)
Ruxolitinib	JAK1 JAK2	≥ 12 years	Cream 1.5%	Twice daily	Nonsegmental vitiligo
Delgocitinib	JAK1 JAK2 JAK3	≥ 2 years	Ointment 0.25% and 0.5%	Twice daily	Chronic hand eczema (cream 2%)
Tapinarof	AhR	≥ 2 years	Cream 1%	Once daily	Plaque psoriasis

In 2024, a second topical PDE4 inhibitor, roflumilast, was FDA approved for the treatment of mild to moderate eczema, following its initial approval for plaque psoriasis in 2022 (Table 1). Compared to crisaborole, roflumilast is a more potent and selective competitive PDE4 inhibitor [11]. In randomised controlled trials (RCTs) of patients with atopic dermatitis, roflumilast was used once daily, whereas crisaborole was used twice daily [12, 13]. There were low rates of nausea and diarrhoea among those treated with roflumilast compared to vehicle [12]. The use of systemic PDE4 inhibitors, such as apremilast and oral roflumilast, for other indications has been limited by gastrointestinal side effects [14, 15]. Observational safety data will be important in determining whether topical roflumilast demonstrates increased gastrointestinal AEs with prolonged use and high body surface area (BSA). Application site AEs were not prominent in the roflumilast trial, including among patients who previously stopped crisaborole because of poor tolerability [12]. However, a network meta‐analysis showed that both crisaborole and roflumilast were among the least effective topical treatments for atopic dermatitis compared to a vehicle control [16].

2.2. Ruxolitinib

Ruxolitinib cream 1.5% is a potent JAK(Janus kinase)1 and JAK2 inhibitor (Table 1). It was FDA approved for the treatment of mild to moderate atopic dermatitis in 2021. In preclinical models, oral ruxolitinib inhibits downstream STAT(signal transducer and activation of transcription)5 and extracellular signal‐regulated kinase 1/2 phosphorylation. Topical ruxolitinib was initially investigated in mouse models of psoriasis, where it was shown to inhibit JAK/STAT signalling and reduce cutaneous inflammation [17]. JAK and IL(interleukin)‐4Rα signalling were also found to be important in chronic itch [18]. Phase III RCTs showed that compared to a vehicle control, a greater proportion of patients achieved IGA‐TS (investigator global assessment treatment success) with 1.5% topical ruxolitinib compared to vehicle (53.8% vs. 15.1%) [19]. Long‐term safety analyses showed low plasma ruxolitinib concentrations. The only serious AE which was considered related to treatment was a hydatidiform mole, and more data is needed to determine topical ruxolitinib's safety in pregnancy [20]. However, these studies only included patients with an affected BSA of 3%–20%. A maximum use study of patients with a BSA involvement of ≥ 25% found that higher plasma concentrations of ruxolitinib were associated with a higher affected BSA and/or the amount of cream applied, but were not associated with hematologic abnormalities [21].

In a network meta‐analysis of topical treatments for atopic dermatitis, ruxolitinib was among the most effective compared to a vehicle control (IGA odds ratio [OR] 9.34, 95% confidence interval [CI] [4.8–18.18]) [16]. Topical ruxolitinib should not be used in combination with systemic immunosuppressive therapy [22]. It may worsen acne and should be used with caution on sebaceous sites. As ruxolitinib is a JAK inhibitor, it contains a black box warning for serious infections, all‐cause mortality, malignancy, major adverse cardiac events (MACE), and thrombosis. There has been no clear signal for these AEs in patients with atopic dermatitis up to 1 year; but given that there is systemic absorption of topical ruxolitinib, particularly among those with more widespread disease, patients should be counselled on these risks and advised to limit their use to < 20% BSA as directed [20, 21, 22, 23].

2.3. Delgocitinib

Delgocitinib ointment 0.25% and 0.5% was approved in Japan in 2020 for the treatment of atopic dermatitis (Table 1). Delgocitinib is a competitive inhibitor of JAK1, JAK2, JAK3 and TYK(tyrosine kinase)2. In vitro assays showed that delgocitinib inhibited similar cytokine pathways compared to tofacitinib, while ruxolitinib was more potent at blocking IL‐23, granulocyte macrophage‐colony stimulating factor, and interferon α signalling pathways [24].

Delgocitinib ointment 0.5% twice daily was shown to be effective in patients with atopic dermatitis ≥ 16 years of age in a phase III RCT [25]. The primary endpoint was the percent change from baseline in the modified Eczema Area and Severity Index (mEASI) score at the end of treatment. A greater proportion of those treated with delgocitinib met the primary endpoint compared to vehicle (percent change in mEASI −44.3% vs. 1.7%). The most common treatment‐related AE was impetigo. Another phase III RCT in paediatric patients aged 2–15 years allowed the use of delgocitinib 0.25% or 0.5% twice daily, depending on disease severity [26]. The primary endpoint was yet again the percent change from baseline in the modified mEASI score at the end of treatment. A greater proportion of those in the delgocitinib treated group met the primary endpoint compared to vehicle (percent change in mEASI −39.3% vs. 10.9%). The most common treatment‐related AEs were application‐site folliculitis and acne. In both trials, a proportion of patients had detectable plasma levels of delgocitinib throughout different timepoints (trial 1 8.9%–17.0%, trial 2 4.9%–16.4%) [25, 26]. Both trials were followed by an optional open‐label extension study (trial 1 28 weeks, trial 2 56 weeks), which showed sustained efficacy and no treatment‐related severe AEs [25, 26].

Delgocitinib cream 2% was approved in 2024 in the United Kingdom for the treatment of adult patients with moderate to severe chronic hand eczema and is currently under review with the FDA. Delgocitinib cream 2% showed promising results for the treatment of chronic hand eczema in two phase III RCTs with a vehicle control and in a phase III RCT versus oral alitretinoin [27, 28]. Currently, there are no published clinical trials of delgocitinib cream 2% in atopic dermatitis.

A network meta‐analysis found that, compared to a vehicle control, delgocitinib ointment 0.5% was among the most effective topical therapies when assessing clinician‐reported outcomes (OR 7.61, 95% CI [3.72–15.58]), slightly below topical ruxolitinib 1.5% (9.34 [4.8–18.18]) [16]. The RCTs of delgocitinib in atopic dermatitis limited use to ≤ 30% BSA and a maximum dose of 5 g per application; it is currently unclear if there are safety concerns beyond these parameters [25, 26]. Moreover, some patients had detectable plasma levels of delgocitinib, and it will be important to determine if there are any associated long‐term implications. While delgocitinib cream 2% has shown to be a promising alternative to oral alitretinoin in chronic hand dermatitis, its efficacy and safety have not yet been reported in atopic dermatitis [28].

2.4. Tapinarof

Tapinarof cream 1% was FDA approved in 2024 for the treatment of moderate to severe atopic dermatitis in patients ≥ 2 years old. It has a unique mechanism of action as a topical aryl hydrocarbon receptor agonist (Figure 1, Table 1). Aryl hydrocarbon receptor is a cytosolic ligand‐activated transcription factor that plays a role in both the innate and adaptive immune response. In vitro studies suggest that tapinarof induces barrier gene expression in keratinocytes and modulates the Th17 cytokine response [29]. While it primarily binds to the aryl hydrocarbon receptor, it may also promote the antioxidant response programme by binding to Nrf2 (nuclear factor‐erythroid 2‐related factor‐2) [29].

Two vehicle‐controlled RCTs (ADORING 1 and ADORING 2) assessed the efficacy and safety of tapinarof once daily in patients ≥ 2 years old. 80% of participants were children [30]. The primary endpoint was a score of 0 or 1 and ≥ 2 improvement from baseline at Week 8 on the validated IGA for atopic dermatitis (vIGA‐AD). Significantly more patients in the tapinarof group achieved the primary endpoint of a vIGA‐AD response in both trials (ADORING 1 45.4% vs. 13.9%; ADORING 2 46.4% vs. 18.0%). The most common AEs were follicular events (ADORING 1 10.0% tapinarof vs. 0.7% placebo; ADORING 2 8.9% tapinarof vs. 1.5% placebo). Rates of contact dermatitis were lower in the tapinarof group compared to placebo (ADORING 1 1.5% vs. 2.2%; ADORING 2 1.1% vs. 1.5%). Tapinarof maintained its efficacy and favourable safety profile up to 1 year in a long‐term extension study [31].

Tapinarof provides an additional non‐steroidal topical option for atopic dermatitis patients with severe atopic dermatitis, particularly paediatric patients who may not be eligible for other topical treatments. Unlike topical JAK inhibitors, it does not have body surface area restrictions; there are no concerns currently regarding its use in patients with concurrent systemic immunosuppression. Tapinarof's once daily regimen, as opposed to many other topicals used twice daily, may improve patient adherence [32]. Compared to a vehicle control, tapinarof 1% is among the least effective non‐steroidal topical therapies (clinician‐reported signs of eczema OR 2.45, 95% CI [1, 6.02]) [16].

3. Biologics

The role of Th2 cytokines IL‐4 and IL‐13 in driving atopic disease has been well known for several decades. They play important roles in inflammation, epidermal barrier dysfunction and itch in atopic dermatitis [33, 34]. Both cytokines promote IgE (immunoglobulin E) production from plasma cells. IL‐4 specifically promotes Th2 differentiation and eosinophil recruitment. IL‐4 receptor‐α (IL4Rα) binding both IL‐4 and IL‐13 is expressed on many inflammatory cells involved in atopic dermatitis, such as keratinocytes, T cells, dendritic cells, and eosinophils [35]. People with atopic dermatitis have polymorphisms in multiple immune‐related pathways, including IL4Rα, IL‐4 and IL‐13 [33, 34]. As such, biologics targeting Th2 cytokines are an inherently appealing strategy to manage atopic dermatitis.

3.1. Dupilumab

Dupilumab is a fully human IgG4 monoclonal antibody that binds to and inhibits IL4Rα, which is the shared alpha chain subunit of IL‐4 and IL‐13 receptors (Figure 2, Table 2). Skin biopsy samples from patients who underwent a 4‐week treatment of dupilumab showed reduced expression of Th2‐associated cytokines without any reciprocal increase in Th1 pathway expression [35]. Of note, dupilumab treatment response did not correlate with serum IgE levels, suggesting that inhibition of IL4Rα can be beneficial in patients with either intrinsic or extrinsic atopic dermatitis [35].

FIGURE 2.

Mechanism of action of biologics used for the treatment of atopic dermatitis. Dupilumab binds IL4Rα, blocking IL‐4 binding and consequently both IL‐4 and IL‐13 intracellular JAK/STAT signalling. Both lebrikizumab and tralokinumab bind IL‐13 directly and inhibit IL‐13 mediated signalling. Tralokinumab also prevents IL‐13 binding to the decoy receptor, IL13Rα2. Nemolizumab binds to IL31Rα and inhibits IL‐31 mediated JAK/STAT signalling.

TABLE 2.

Comparison of different biologics used for the treatment of atopic dermatitis.

Biologic	Target	Age	Adult dosing	Other uses
Dupilumab	IL4Rα	≥ 6 months	600 mg and then 300 mg q4w	Asthma Chronic obstructive pulmonary disease Chronic rhinosinusitis with nasal polyposis Chronic spontaneous urticaria Eosinophilic esophagitis Prurigo nodularis Bullous pemphigoid (off‐label)
Tralokinumab	IL13	≥ 12 years	600 mg and then 300 mg q2w
Lebrikizumab	IL13	≥ 12 years	> 40 kg: 500 mg Week 0 and 2 and then 250 mg q2w until Week 16, and then 250 mg q4w
Nemolizumab	IL31Rα	≥ 12 years	60 mg and then 30 mg q4w. At 16 weeks, if clear or almost clear, 30 mg q8w	Prurigo nodularis

SOLO‐1 and SOLO‐2 were RCTs comparing dupilumab to a placebo control in adults with moderate‐to‐severe atopic dermatitis [36]. The primary endpoint was the proportion of patients with an IGA of 0 or 1 and a reduction of ≥ 2 points from baseline at week 16. In SOLO‐1, the primary outcome occurred in 38% of patients receiving dupilumab every other week and in 37% receiving weekly dupilumab, compared to 10% receiving placebo. SOLO‐2 was similar; the primary outcome occurred in 36% of patients receiving dupilumab every other week and in 36% receiving weekly dupilumab, compared to 10% receiving placebo. Compared to placebo, a greater proportion of patients on dupilumab reported an improvement of at least 75% on the EASI (EASI‐75) and patient‐reported outcomes [36]. A 5‐year prospective, multi‐center, observational cohort study found that dupilumab maintained its effectiveness and favourable safety profile; the most common AE was conjunctivitis, affecting 33.7% of patients, of which 3.4% stopped treatment [37].

In addition to atopic dermatitis, dupilumab is also FDA approved for prurigo nodularis, asthma, chronic rhinosinusitis with nasal polyps, eosinophilic esophagitis, and chronic spontaneous urticaria—five other Th2‐driven diseases (Table 1) [38, 39, 40, 41, 42]. It is also FDA approved for chronic obstructive pulmonary disease [41, 43, 44]. Patients treated with dupilumab for atopic dermatitis noted improvement in their comorbid asthma and/or chronic sinonasal conditions [45]. Dupilumab has also shown promising results in bullous pemphigoid, another dermatological condition with a Th2 cytokine profile and elevated levels of IL‐4 and IL‐5 [46, 47, 48, 49]. Similar to atopic dermatitis, mycosis fungoides and Sezary syndrome have a Th2 cytokine profile, and dupilumab was initially proposed as an off‐label treatment for these conditions [50]. However, dupilumab may actually drive the progression of cutaneous T‐cell lymphomas [51, 52]. There are reports of patients developing psoriasis and seronegative arthritis during dupilumab treatment, and this may be due to a Th17 shift upon blockade of IL‐4 and IL‐13 [35, 53, 54, 55, 56].

Dupilumab can be considered in patients ≥ 6 months of age with moderate to severe atopic dermatitis [57, 58]. In indirect comparisons against JAK inhibitors and other systemic therapies for atopic dermatitis, dupilumab is a highly effective treatment, only being surpassed by higher doses of JAK inhibitors abrocitinib (200 mg daily) and upadacitinib (30 mg daily) (change in EASI vs. abrocitinib 200 mg 2.4, 95% credible interval [CrI] [0.4, 4.2]); (change in EASI vs. upadacitinib 30 mg 3.0 [1, 5]) [59, 60, 61]. These findings are supported by direct evidence from head‐to‐head RCTs, in which upadacitinib 30 mg daily and abrocitinib 200 mg daily were superior to dupilumab [57, 58, 62]. However, unlike JAK inhibitors, dupilumab is not broadly immunosuppressive and has not been associated with higher infection rates. Patients with atopic dermatitis and comorbid conditions (i.e., asthma or nasal polyps) may experience improvement in multiple conditions with dupilumab. Dupilumab may be less effective in patients with obesity [37].

3.2. Tralokinumab

Tralokinumab is a fully human IgG4 monoclonal antibody that binds to IL‐13 and prevents it from binding to both IL‐13Rα1 and IL13Rα2 (Figure 2, Table 2). Downstream, it prevents IL4Rα/IL13Rα1 heterodimerisation. IL13Rα2 is thought to be a decoy receptor and it is unclear if disrupting IL‐13/IL‐13Rα2 binding has any downstream consequences [63, 64]. IL13Rα2 has previously been shown to internalise IL‐13 without activating STAT signalling [65].

ECZTRA 1 and ECZTRA 2 were two placebo‐controlled RCTs that evaluated the efficacy and safety of tralokinumab in adults with moderate‐to‐severe atopic dermatitis who did not respond to topical treatments. The primary outcome was an IGA score of 0 or 1 at week 16 and ≥ 75% improvement in EASI‐75 at week 16. In both ECZTRA 1 and ECZTRA 2, a greater proportion of patients in the tralokinumab group met the primary outcomes compared to placebo ECZTRA 1 IGA difference 8.6%, 95% CI [4.1–13.1] and EASI‐75 difference 12.1% [6.5–17.1]; ECZTRA 2 IGA difference 11.1% [5.8–16.4] and EASI‐75 difference 21.6% [15.8–27.3]. Much like dupilumab, the most common AEs were URTI and conjunctivitis [66]. However, unlike dupilumab, which has been associated with facial and neck erythema, a similar signal was not seen with the use of tralokinumab up to 4 years [67, 68].

A network meta‐analysis found tralokinumab to be less efficacious compared to dupilumab (change in EASI between tralokinumab vs. dupilumab 4.2, 95% CrI [2.3, 6.2]) [59, 61, 69]. Among biologics and JAK inhibitors, tralokinumab is one of the least effective treatments in achieving EASI‐75, only slightly ahead of nemolizumab (OR 1.2 [0.8, 1.7]) [60, 61].

3.3. Lebrikizumab

Like tralokinumab, lebrikizumab binds IL‐13 and prevents downstream IL4Rα/IL‐13Rα1 heterodimerization (Figure 2, Table 2). Unlike tralokinumab, it has no impact on IL‐13/IL‐13Rα2 binding.

Two phase III RTCs, ADvocate1 and ADvocate2, demonstrated the efficacy and safety of lebrikizumab in patients ≥ 12 years of age with moderate‐to‐severe atopic dermatitis. The primary outcome in both trials was an IGA score of 0 or 1 with a reduction of at least 2 points from baseline at week 16. The primary outcome was met in a greater proportion of patients treated with lebrikizumab compared to placebo (ADvocate1 43.1% vs. 12.7%; ADvocate2 33.2% vs. 10.8%) [70]. Like all inhibitors targeting IL4Rα/IL‐13Rα1 heterodimerisation, there was a higher incidence of conjunctivitis among patients treated with lebrikizumab compared to placebo [70]. A long‐term extension study showed maintained response and no new safety signals over 68 weeks [71].

There are no head‐to‐head trials comparing lebrikizumab and dupilumab. However, indirect comparisons via network meta analyses found lebrikizumab to be similar or slightly less efficacious compared to dupilumab (change in EASI with lebrikizumab vs. dupilumab 2.1, 95% CrI [0.1–4.3]) [59, 60, 72]. However, it may be more effective than tralokinumab (change in EASI −2.1 [−4.3, 0.2]) and nemolizumab (change in EASI −3.9 [−6.5, −1.2]).

3.4. Nemolizumab

Nemolizumab has a distinct mechanism of action compared to other biologics for atopic dermatitis. It is a fully human IgG2 monoclonal antibody that binds to IL31Rα (Figure 2, Table 2). IL‐31 is a neuroimmune cytokine that is secreted by Th2 cells and immature dendritic cells. Binding of IL‐31 to IL31Rα on sensory nerves links peripheral itch signalling to the central nervous system [73].

Nemolizumab was initially FDA approved for the treatment of prurigo nodularis in August 2024; its approval for atopic dermatitis followed 4 months later and was based on two phase III placebo‐controlled trials, ARCADIA 1 and ARCADIA 2, which evaluated the safety of nemolizumab with background use of topical steroids with or without TCIs [74]. A greater proportion of patients in the nemolizumab treated group met the primary outcomes of an IGA of 0 or 1 and EASI‐75 (ARCADIA 1 IGA difference 11.5% [97.5% CI 4.7–18.3]; ARCADIA 1 EASI‐75 response difference 14.9% [7.8–22.0]; ARCADIA 2 IGA difference 12.2% [4.6–19.8]; ARCADIA 2 EASI‐75 difference 12.5% [4.6–20.3]) [74]. The most common adverse events were worsening of atopic dermatitis, urticaria, and asthma. A long‐term extension study showed that nemolizumab maintained its effectiveness with a consistent safety profile up to 68 weeks [75].

A living network meta‐analysis found that nemolizumab was less effective compared to dupilumab with respect to changes in EASI scores (change in EASI 6, 95% CrI [3.7, 8.5]) [61]. However, nemolizumab had similar efficacy to dupilumab with regard to change in peak pruritus numeric rating scale (0.1 [−0.4, 0.6]), in keeping with its putative role in suppressing itch. Further, given that nemolizumab works through a different receptor compared to other biologics for atopic dermatitis, identifying predictive biomarkers may help with patient selection.

4. JAK Inhibitors

The JAK/STAT signalling pathway is upstream of various major biological pathways, including haematopoiesis, inflammation, and apoptosis. JAK proteins bind to the intracellular portions of cytokine receptors. Upon binding of a cytokine to its receptor, intracellular JAK proteins mediate recruitment and tyrosine phosphorylation of STAT proteins. Phosphorylated STAT proteins dimerize and translocate to the nucleus, where they regulate different genes. The JAK family of kinases consists of JAK1, JAK2, JAK3 and TYK2 (Figure 3). The JAK/STAT pathway has been implicated in Th2 differentiation, inflammation, and pruritus in atopic dermatitis [2, 76, 77]. JAK inhibitors are FDA approved for multiple indications beyond atopic dermatitis, which should be considered when selecting a treatment in a patient with any immune‐mediated comorbidities (Table 3).

FIGURE 3.

Mechanism of action of JAK inhibitors used for the treatment of atopic dermatitis. JAK inhibitors inhibit different combinations of JAK intracellular proteins. Upon binding of a cytokine to its receptor, intracellular JAK proteins mediate recruitment and tyrosine phosphorylation of STAT proteins. Phosphorylated STAT proteins dimerise and translocate to the nucleus, where they regulate different genes.

TABLE 3.

Comparison of different oral JAK inhibitors used for the treatment of atopic dermatitis.

JAK inhibitor	Target	Age (years)	Dosing	Drug interactions	Other uses
Baricitinib	JAK1 JAK2	≥ 18	2 mg once daily 4 mg once daily	OAT3 inhibitors	Alopecia areata Rheumatoid arthritis COVID‐19
Upadacitinib	JAK1	≥ 12	15 mg once daily 30 mg once daily	CYP3A4 inhibitors	Ankylosing spondylitis Axial spondyloarthritis Crohn's disease Psoriatic arthritis Rheumatoid arthritis Ulcerative colitis
Abrocitinib	JAK1	≥ 12	100 mg once daily 200 mg once daily	CYP2C19 inhibitors/inducers CYP2C9 inhibitors/inducers

All JAK inhibitors carry warnings for thromboembolism, malignancy and MACE. These warnings are largely based on an FDA review of clinical trial data of patients with rheumatoid arthritis treated with the pan‐JAK inhibitor tofacitinib [78]. A head‐to‐head safety trial of patients with rheumatoid arthritis showed a higher risk of malignancy and MACE among those treated with tofacitinib compared to TNF inhibitors [79]. As of yet, a clear signal for these serious AEs has not been identified in clinical trials or observational studies of JAK inhibitors among people with atopic dermatitis [78, 79, 80, 81, 82]. However, compared to dupilumab, JAK inhibitors as a class are associated with higher risks of infection (particularly herpesviruses), acneiform eruptions, cytopenias and dyslipidaemia in patients with atopic dermatitis [83]. JAK inhibitors should be used cautiously in patients with a history of cardiovascular risk factors, thromboembolism, and malignancy, and in older adults whose baseline risk for serious harms is higher [83].

4.1. Baricitinib

Baricitinib is a selective inhibitor of JAK1 and JAK2 (Figure 3, Table 3). It was initially characterised in rat models of rheumatoid arthritis [84]. It was later shown to be effective in clinical trials of rheumatoid arthritis and was ultimately FDA approved for rheumatoid arthritis at a daily dose of 2 mg as higher doses were associated with severe side effects [84, 85, 86]. It is approved in Europe, but not the US, for atopic dermatitis.

BREEZE‐AD1 and BREEZE‐AD2 were two phase III RCTs that studied the efficacy and safety of patients with moderate‐to‐severe atopic dermatitis who did not respond to topical therapies [82]. Patients were randomised to four groups: placebo or baricitinib 1 mg, 2 mg or 4 mg. Compared to placebo, baricitinib 2 mg and 4 mg showed significant improvement in the primary endpoint of a vIGA‐AD (validated Investigator's Global Assessment of Atopic Dermatitis) score of 0 or 1 (BREEZE‐AD1 placebo 4.8%, baricitinib 2 mg 11.4%, baricitinib 4 mg 16.8%; BREEZE‐AD2 placebo 4.5%, baricitinib 2 mg 10.6%, baricitinib 4 mg 13.8%). Safety analysis showed no differences between the frequency of AEs and treatment group, even in the baricitinib 4 mg group [82]. However, subsequent placebo‐controlled studies of baricitinib found higher rates of nasopharyngitis, upper respiratory tract infection, and herpes simplex in the baricitinib treated group [80, 81].

Like all JAK inhibitors, baricitinib is a pill and can be considered in patients with moderate‐to‐severe atopic dermatitis who cannot tolerate subcutaneous injections. Baricitinib is not as effective as other JAK inhibitors and is slightly less effective than dupilumab (baricitinib 4 mg change in EASI compared to dupilumab 3, 95% CrI [0.7, 5.2]) [61, 69, 87]. It is unclear if baricitinib is better tolerated compared to other JAK inhibitors as there are no head‐to‐head trials and the incidence of AEs is too low in individual trials to create a well‐powered indirect comparison. Post‐marketing surveillance studies will be important in determining if baricitinib at the 4 mg daily dose shows similar AEs in the atopic dermatitis population as it did in patients with rheumatoid arthritis [86].

4.2. Upadacitinib

Upadacitinib is a selective JAK1 inhibitor that was developed in light of tofacitinib's dose limiting side effects (Figure 3, Table 3). Avoiding pan‐JAK inhibition could theoretically reduce the risk of anaemia (JAK2 drives erythropoiesis) and infection (JAK3 is important for NK [natural killer] cell survival). Structural hypotheses were used to identify and exploit differences between JAK1 and JAK's active site. Upadacitinib showed JAK1 selectivity in cellular assays and in vivo rat studies showed less reticulocyte and NK cell inhibition with upadacitinib compared to tofacitinib [88].

Measure Up 1 and Measure Up 2 were two phase III RCTs that evaluated placebo vs. upadacitinib (15 mg daily and 30 mg daily) in patients ≥ 12 years of age with moderate‐to‐severe atopic dermatitis [89]. The coprimary end points were a vIGA of 0 or 1 with improvement of ≥ 2 grades and EASI‐75 at week 16. A greater proportion of patients treated with upadacitinib met both coprimary endpoints (Measure Up 1 adjusted difference in EASI‐75 response rate vs. placebo, upadacitinib 15 mg 53.3%, 95% CI [46.4–60.2], upadacitinib 30 mg 63.4% [57.1–69.8], adjusted difference in vIGA‐AD, upadacitinib 15 mg 39.8%, 95% CI [33.2–46.4], upadacitinib 30 mg 53.6% [47.2–60.0]; Measure Up 2 adjusted difference in EASI‐75 response rate vs. placebo, upadacitinib 15 mg 46.9% [39.9–53.9], upadacitinib 30 mg 59.6% [53.1–66.2], adjusted difference in vIGA‐AD, upadacitinib 15 mg 34.0% [27.8–40.2], upadacitinib 30 mg 47.4% [41.0–53.7]). The most frequent treatment‐related AEs were acne, upper respiratory tract infections, and nasopharyngitis [89]. Long‐term non‐comparative efficacy and safety analyses up to 1 year were consistent with shorter trials [90, 91].

Upadacitinib is among the most effective treatments for atopic dermatitis. In both head‐to‐head comparisons and network meta‐analyses, upadacitinib showed superior efficacy compared to dupilumab [57, 58, 59, 60, 69]. Patients treated with upadacitinib have a more rapid clinical response compared to those treated with dupilumab [57, 58]. In indirect comparisons, upadacitinib 30 mg daily is the most effective treatment among all biologics and JAK inhibitors (upadacitinib 30 mg change in EASI compared to dupilumab −3, 95% CrI [−5, −1]) [59, 60, 61, 69, 72, 87]. A recently published trial compared a dose‐escalation strategy for upadacitinib versus dupilumab, with participants in the upadacitinib arm starting at 15 mg daily and increasing to 30 mg daily if they had not had an adequate response. In that trial, upadacitinib was superior to dupilumab, suggesting that starting at the 15 mg dose and escalating to 30 mg if necessary is a reasonable strategy [57].

While one study found high‐dose upadacitinib to be associated with a higher frequency of AEs, another study found that withdrawals due to an AE did not differ significantly between treatment groups [59, 72]. Of note, unlike baricitinib and abrocitinib, upadacitinib can be used in patients with severe renal impairment [86, 92, 93].

4.3. Abrocitinib

Abrocitinib is a JAK1 inhibitor (Figure 3, Table 3). It was initially developed by screening compounds that modified tofacitinib's 3‐aminopiperidine group. It was the most selective JAK1 inhibitor identified in a screen; abrocitinib showed 28‐fold selectivity over JAK2 and > 340‐fold selectivity over JAK3 [94].

JADE MONO‐1 and JADE MONO‐2 were two replicate phase III RCTs of abrocitinib (100 mg daily or 200 mg daily) or placebo in patients 12 years or older with moderate‐to‐severe atopic dermatitis [95, 96]. The coprimary end points were an IGA of 0 or 1 with improvement of ≥ 2 grades and EASI‐75 at week 12. In both trials, significantly more patients treated with abrocitinib met both primary endpoints (JADE MONO‐1: IGA; placebo 8%, abrocitinib 100 mg 24%, abrocitinib 200 mg 44%; EASI‐75 placebo 12%, abrocitinib 100 mg 40%, abrocitinib 200 mg 63%. JADE MONO‐2: IGA; placebo 9.1%, abrocitinib 100 mg 28.4%, abrocitinib 200 mg 38.1%; EASI‐75 placebo 10.4%, abrocitinib 100 mg 44.5%, abrocitinib 200 mg 61.0%). A long‐term extension trial found that abrocitinib showed sustained improvements in patient‐reported outcomes up to 48 weeks [97].

Compared to other JAK inhibitors, abrocitinib was associated with higher rates of nausea. Other common side effects were nasopharyngitis and headache. Treatment‐related serious AEs that led to discontinuation of abrocitinib were reported in four patients across both trials and included herpangina, pneumonia, inflammatory bowel disease, and pancreatitis. No cases of treatment‐related venous thromboembolism, MACE, or malignancies were observed [95, 96]. Long‐term safety data up to 4 years showed no new safety signals. The most frequent serious infections were herpes zoster (abrocitinib 100 mg 0.2%, abrocitinib 200 mg 0.5%), pneumonia (0.2% with either dose), and herpes simplex (0.1% with either dose) [98].

In a head‐to‐head RCT of placebo versus dupilumab versus abrocitinib, abrocitinib 200 mg daily, but not 100 mg daily, was superior to dupilumab in improving itch response at week 2 (95% CI [13.5–30.7]), a secondary endpoint [62]. Among the primary endpoints (IGA of 0 or 1 with improvement of ≥ 2 grades and EASI‐75), there was no difference in EASI‐75 and no conclusion could be made with respect to IGA. A separate head‐to‐head trial of dupilumab versus abrocitinib 200 mg daily found that abrocitinib was associated with earlier improvements in itch and EASI‐90 response [99]. A network meta‐analysis found abrocitinib 200 mg daily to be associated with somewhat better improvement in EASI scores compared to dupiumab [69].

Abrocitinib at the 200 mg daily dose is among the most effective JAK inhibitors [59, 60, 69]. In indirect analyses, abrocitinib 200 mg is superior to almost all systemic therapies, with the exception of upadacitinib 30 mg, which has similar efficacy (change in EASI compared to upadacitinib 0.7, 95% Crl [−1.6, 3]). While a dose escalation strategy for abrocitinib has not been specifically studied in an RCT, it is inherently reasonable to start at 100 mg and escalate to 200 mg daily as needed.

Abrocitinib can be an effective alternative for patients who are unable to tolerate biologics (i.e., injection site reactions, conjunctivitis, etc.) or who prefer an oral medication. Unlike other JAK inhibitors, abrocitinib is associated with reduced platelet counts and should be used cautiously in patients with a history of thrombocytopenia [62, 93, 95, 96].

5. Discussion

Over the last several years, there has been a tremendous expansion in the repertoire of treatments for atopic dermatitis. Most of these medications are based on efficacy in short‐term clinical studies. Longer clinical trials are often non‐comparative, making it difficult to draw strong conclusions. Whether treatment response holds over time is yet to be determined. Similarly, long‐term safety data is needed to detect rare side effects that would not have been captured in clinical trials with relatively small populations and short durations. In particular, more data comparing the long‐term safety profile of JAK inhibitors to biologics for atopic dermatitis is needed [100, 101]. Warnings for JAK inhibitors used for atopic dermatitis have been applied based on data from rheumatoid arthritis, a very different disease—atopic dermatitis‐specific data is essential.

Randomised clinical trials for systemic treatment can either be monotherapy trials, in which the use of topical anti‐inflammatory medications is prohibited, or combination therapy trials, where topical anti‐inflammatory medications are permitted alongside systemic treatment. The latter more closely resembles clinical practice, where most patients will continue topical therapy while beginning a new systemic therapy. NMAs summarised in this review include both monotherapy and combination therapy trials [59, 72]. Comparisons between systemic medications in the NMA are similar when pooling and separating monotherapy and combination therapy trials because topical anti‐inflammatory medications increase both placebo and active treatment response rates [59].

There are even more treatment advances on the horizon. Rocatinlimab and amlitelimab are two monoclonal antibodies that target OX40 signalling, a driver of multiple effector and memory T‐cell populations [102, 103]. Another monoclonal antibody, GSK1070806, targets IL‐18, a cytokine that is overexpressed in atopic dermatitis. IL‐18 knockout mice were less likely to develop atopic dermatitis compared to wild type [104, 105, 106]. New drugs against established targets are also in the pipeline. These include AK120, an anti‐IL4Rα antibody, and LNK01001, a JAK1 inhibitor [107, 108].

The rapidly evolving therapeutic landscape of atopic dermatitis is a testament to bench‐to‐bedside discovery. As new targeted therapies enter the market, the next challenge will be determining whether they stand the test of time and how to best translate their use into clinical practice.

Author Contributions

S.M. drafted the manuscript, tables, and figures. A.M.D. provided supervision and critical revisions.

Conflicts of Interest

Dr. Mirali has no conflicts of interest to declare. Dr. Drucker has received compensation from the British Journal of Dermatology (reviewer and Editor), American Academy of Dermatology (guidelines writer), Canadian Dermatology Today (manuscript writer) and National Eczema Association (consultant) Canada's Drug Agency (consultant). Dr. Drucker has received research grants to his institution from the National Eczema Association, Canadian Dermatology Foundation, Canadian Institutes for Health Research, US National Institutes of Health and Physicians Services Incorporated Foundation.

Data Availability Statement

The authors have nothing to report.