In AD, epithelial cell–derived cytokines such as IL-25, IL-33, and TSLP promote the overexpression of Th2 cytokines, including IL-4, IL-5, IL-13, and IL-31, either directly through Th2 cytokine-secreting cells or indirectly through dendritic cell-mediated polarization (38). IL-4 performs its biological function by binding to an IL-4 receptor (IL-4R), which is a heterodimeric receptor complex on the cellular surface. IL-4R has two subtypes. Type I IL-4R comprises an IL-4 receptor alpha (IL-4Rα) chain and a common gamma (γ) chain, whereas type II IL-4R comprises an IL-4Rα and an IL-13 receptor alpha 1 (IL-13Rα1) chain (39, 40). IL-13Rα1, along with IL-13 receptor alpha 2 (IL-13Rα2), is a subtype of the IL-13 receptor alpha (IL-13Rα) chain. Unlike the ubiquitously presented IL-4R, IL-13Rα1 chains are expressed on limited cell surfaces including those of B cells, monocytes, macrophages, basophils, mast cells, and endothelial cells (39). Nonhematopoietic structural tissue cells such as keratinocytes also express relatively high amounts of IL-13Rα1, facilitating type II IL-4 receptor formation even with low levels of or absence of common γ chain (γc). Considering IL-4Rα as a functional component responsible for intercellular signal transduction, IL-13 may deliver its signal by binding type II IL-4R with the coexpression of the IL-4Rα and IL-13Rα1 chains (41). Different JAKs bind to different subunits of the receptors, accounting for the variability of JAK–STAT signaling pathways. For example, JAK1, JAK3, and TYK2 bind to IL-4Rα, γc chain, and IL-13Rα subunits of the IL-4R receptors, respectively.

In the IL-4 mediated immune response, IL-4 binds to type I IL-4R, stimulating the phosphorylation of JAK1 and JAK3, which in turn activate and phosphorylate IL-4Rα and STAT6 ( Figure 2 ) (42). The phosphorylated STAT6 subsequently dimerizes and acts as a transcription factor by binding to specific DNA sequences of IL-4-responsive genes. Activated JAK1 can also phosphorylate Fes kinase, which then phosphorylates insulin receptor substrate-1/2 (IRS-1/2), a protein on the IL-4Rα chain linked to several downstream signaling pathways. Phosphorylated IRS-1/2 recruits p85 regulatory subunits, which activate the phosphatidylinositol 3-kinase (PI3K)–protein kinase B (AKT) pathway. The PI3K–AKT pathway then promotes the proliferation, regeneration, and apoptosis of keratinocytes (43, 44). Conversely, IL-4 and IL-13 can bind to type II IL-4R, inducing the robust phosphorylation of JAK1 and TYK2, followed by the activation and phosphorylation of STAT3 and STAT6. This results in the downregulation of FLG expression and skin barrier dysfunction, and the increased production of TSLP, IL-25, and IL-33 in keratinocytes (45). The IL-4-induced JAK–STAT signal cascade can be inhibited by suppressors of cytokine signaling-1 (SOCS-1), which directly prevent JAK1/STAT6 phosphorylation and arrest IL-4-induced gene expression.

TSLP, a pro-Th2 cytokine, plays a key role in the activation and migration of dendritic cells, which drive the differentiation of Th2 cells to produce IL-4, IL-5, and IL-13 (46). TSLP is highly associated with IL-7, another stromal cell-derived cytokine. TSLP signals by binding to a heterodimeric receptor comprising IL-7 receptor alpha (IL-7Rα) and TSLPR, which interact with JAK1 and JAK2, respectively (20). Accordingly, the TSLP-mediated response activates STAT5 through the phosphorylation of JAK1 and JAK2 (20, 47).

The IL-5 receptor (IL-5R) comprises an alpha subunit (IL-5Rα), which specifically recognizes only IL-5 and a common beta (βc) heterodimeric chain, which can nonspecifically conjugate with not only IL-5 but also IL-3 and granulocyte-macrophage colony stimulating factor (GM-CSF) (27, 48, 49). In the JAK–STAT signaling pathway, the conjugation of IL-5 and the IL-5R triggers the phosphorylation of JAK1 and JAK2, resulting in the activation of STAT1, STAT3, and STAT5 (50).

IL-31 is a helical cytokine belonging to the glycoprotein 130 (gp130)/IL-6 family, a group of receptor proteins that share a common gp130 subunit in their receptor complex (51). IL-31 constitutively engages with the IL-31 receptor (IL-31R), a heterodimeric receptor comprising an IL-31 receptor alpha chain (IL-31Rα) coupled with an oncostatin M receptor beta chain (OSMRβ) (51). The intercellular signals of IL-31 are transmitted by the JAK–STAT pathway, mitogen-activated protein kinases (MAPKs), and the PI3K–AKT pathway. By engaging with the IL-31Rα–OSMRβ complex, IL-31 triggers the activation of JAK1 and JAK2, thereby stimulating STAT3, STAT5 and, to a lesser degree, STAT1 transcriptional activity (52).

During the chronic phase of AD, Th17 and Th22 cytokines cooperatively modulate local inflammation through upregulation of proinflammatory cytokines that stimulate epidermal hyperplasia (53). IL-17A and IL-17F are eponymous cytokines of the Th17 cell lineage, and IL-17A is believed to have a greater influence on driving autoimmunity than IL-17F because of its more potent signaling ability (54). The IL-17 receptor comprises IL-17RA and IL-17RC, which present different affinities to IL-17A and IL-17F, resulting in different ligand preferences depending on the ratios of IL-17RA and IL-17RC in the receptor complexes (54). In contrast with the classic Th1 and Th2 immune response through the JAK–STAT signaling pathway, Th17-derived IL-17 recruits ACT1, an IL-17 receptor complex adaptor protein, thus engaging with tumor necrosis factor receptor-associated factors (TRAF6) and activating nuclear factor (NF)-κB, CCAAT/enhancer binding protein (C/EBP), and MAPK pathways ( Figure 3 ) (54). Generally, it seems that JAK–STAT pathway is not directly involved in the Th17 signaling and may only play a limited role. Nevertheless, STAT3 is found to orchestrate diverse aspects of Th17 differentiation and function (55). Via STAT3, a combination of Th17-promoting cytokines including IL-6, IL-21, and IL-23 would stimulate STAT3-dependent Th17 differentiation and enhance the production of IL-17 (56). STAT3 also mediate multiple epigenetic modification of Th17 gene transcription and is essential for Th17 proliferation and survival (57). As a result, selective JAK inhibitors may show the impact on the inhibition of Th17 immune response in AD via targeting upstream STAT3 activation.

IL-22 was originally named IL-10-related T-cell-derived inducible factor (IL-TIF) and is identified as a key mediator of epidermal hyperplasia (58). Studies have primarily attributed IL-22 production to Th17- and Th22-cell subsets (59). IL-22 heterodimeric receptor complex comprises the IL-22 receptor (IL-22R1) and IL-10 receptor 2 (IL-10R2) subunits, which are expressed in small amounts by keratinocytes of the skin, gut, and lungs. IL-10R2 is also a subunit of the receptors for IL-10, IL-26, IL-28, and IL-29, whereas IL-22R1 is a component of IL-20 receptor 2 (IL-20R2) for IL-20 and IL-24 (58). After binding to IL-22R1, IL-22 induces the phosphorylation of JAK1 and TYK2. Although STAT3 is considered to be the primary mediator in this cascade, the activation of STAT1 and STAT5 has also been observed in IL-22 signaling (60). Apart from the JAK–STAT pathway, MAPK and p38 pathways are also engaged in IL-22 signaling (58).

Although Th1 cells are more predominant in the pathogenesis of psoriasis, Th1 skewed immune response also plays a role in the chronic phase of AD. Higher levels of Th1 cytokines including interferon-gamma (IFN- γ), IL-12, and granulocyte colony stimulating factor (G-CSF) are found in chronic AD lesions (61, 62). In the pathogenesis of AD, IL-12 secreted by dendritic cells can polarize Th1 cells from naïve T-helper cells and inhibit the activation of Th2 cells. The IL-12 signaling pathway is initiated by binding to a distinct heterodimeric receptor comprising interleukin 12 receptor beta (IL-12Rβ)1 and IL-12Rβ2 subunits (63). Subsequently, JAK2 and TYK2 are polarized, resulting in the mediation of signal transduction through the activation of STAT4, and to a lesser degree STAT1, STAT3, and STAT5 ( Figure 4 ) (64).

IFN-γ contributes to epidermal dysfunction by reducing the expression levels of ceramides and long-chain fatty acids (65). JAK1 and JAK2 are activated by the conjugation of IFN-γ and its receptor, which comprises IFN-gamma receptor (IFN-γR)1 and IFN-γR2. This phosphorylates STAT1 to transduce transcription signals into the nucleus and bind to the γ-IFN-activated sequence, affecting IFN-γ-induced gene expression (66).

JAK1 is involved in the pathogenesis of AD through the signaling pathways of IL-4, IL-5, IL-13, IL-22, TSLP, and IFN-γ. Hyperactivation of JAK1 results in the overexpression of cutaneous serine protease, causing skin barrier dysfunction (67). Furthermore, the STAT3 signal is a key transcriptional factor that modulates keratinocyte differentiation and maintains skin integrity. In a murine dry skin model, delgocitinib, a JAK inhibitor blocking JAK1, JAK2, and JAK3, inhibited the activation of STAT3 and improved skin barrier functionality by increasing epidermis-associated proteins such as FLG, loricrin, and natural moisturizing factors (68).

The etiology of pruritus in AD involves the interplay between hyperinnervation of the epidermis; increased pruritogens such as histamine, 5 hydroxytryptamine, nerve growth factor (NGF), and substance P; TSLP; and IL-31 (69, 70). The sensation of itchiness is transmitted by cutaneous unmyelinated C-fiber and myelinated Aδ fiber afferents originating from the dorsal root ganglion. The activation of STAT3 mediates astrogliosis in the spinal dorsal horn, acting as an amplifier of itchiness and leading to chronic pruritus (71). In this signaling pathway, lipocalin-2 (LCN2) was identified as the distinct upregulated factor that enhances pruritus signals and triggers the vicious itch-scratch cycle (72).

Notably, IL-31 was observed to stimulate nerve elongation and directly communicate with the primary cutaneous afferent neurons required by pruritogens to transduce itch signals (73, 74). In an AD-like murine model, topical delgocitinib was implicated to suppress IL-31-promoted nerve elongation in vivo and morphological changes of peripheral nerves from the dorsal root ganglion in vitro (75). The topical application of delgocitinib also resulted in substantially reduced scratching behavior in AD mouse models.

Patients with AD may also experience skin pain, demonstrating that the AD symptom burden extends beyond itchiness. The pain inflicted by AD can cause sleep disturbances and decrease quality of life. Evidence supports that the underlying mechanism for the pain response in AD is inflammation induced by proinflammatory cytokines and neuropathic pain of the peripheral nervous system (76). Cytokines involved in the JAK–STAT pathway also contribute to pain modulation. GM-CSF can induce hyperalgesia (77), whereas IL-17 may result in allodynia (78). TNF-α in the spinal cord is believed to modulate neuropathic pain sensations and sensitize pain-receptors for both mechanical and thermal hyperalgesia (79). Hence, patients receiving JAK inhibitors such as baricitinib reported reduced skin pain according to a numerical rating scale (NRS) (80).

Tofacitinib is a first-generation small molecules designed to selectively block JAK1 and JAK3, and to a lesser extent, JAK2 and TYK2. Tofacitinib has oral and topical formulations, but only the oral form of tofacitinib is commercially available as of writing this paper. Oral tofacitinib was approved by the U.S. FDA for the treatment of rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, polyarticular juvenile idiopathic arthritis, and ulcerative colitis (82). Although the oral administration of tofacitinib has demonstrated clinical efficacy in the treatment of recalcitrant AD without significant AEs (83, 84), clinical application remains limited because of the paucity of randomized controlled trials (RCTs). Nevertheless, the relevant literature has reported that topical tofacitinib therapy for mild-to-moderate AD resulted in a significantly greater mean percentage change in Eczema Area and Severity Index (EASI) score, Physician’s Global Assessment (PGA) response, and body surface area than the vehicle treatment ( Table 2 ) (85). Notably, topical tofacitinib demonstrated rapid efficacy during the first week of treatment and continuous improvement throughout a 4-week treatment course. The patients were reported to be well-tolerated to topical tofacitinib without any serious adverse events nor death (85).

Ruxolitinib is a selective JAK1 and JAK2 inhibitor. Its oral formulation has been approved by the U.S. FDA for the treatment of myelofibrosis, polycythemia vera, and graft-versus-host disease (94). Topical 1.5% ruxolitinib cream has been approved by the U.S. FDA for the short-term and noncontinuous chronic treatment of mild-to-moderate AD in patients aged ≥12 years (95). Compared with the vehicle, ruxolitinib therapy resulted in significant improvement of EASI score and Investigator’s Global Assessment (IGA) scores among patients with AD (86). Notably, the rapid activation of ruxolitinib accounted for a significant reduction in itching within the first 12 hours after application, with sustainable efficacy throughout an 8-week treatment period (87). No differences in application site reactions were observed between ruxolitinib cream and the vehicle, indicating the favorable skin tolerability of ruxolitinib cream (87).

As a pan-JAK inhibitor, topical delgocitinib (Corectim) has been approved by Japan Tobacco for the treatment of AD; however, this drug has yet to be approved by the U.S. FDA (96). Patients aged 2–15 and ≥16 years who were treated with 0.25% and 0.5% formulations of topical delgocitinib, respectively, had substantially higher modified EASI (mEASI) scores than the vehicle group. The majority of AEs were mild and well tolerated, such as nasopharyngitis, Kaposi varicelliform eruption, and acne (89–91). Application site reactions were rare and mostly occurred within the first 2 weeks of treatment, indicating high skin tolerability of topical delgocitinib (97). Furthermore, blood tests revealed no plasma concentration of delgocitinib in most patients, indicating minimal systemic absorption of delgocitinib (91).

Brepocitinib (PF-06700841) is a selective JAK1 and TYK2 inhibitor. Oral brepocitinib for the treatment of plaque psoriasis, alopecia areata, inflammatory bowel disease, and other inflammatory diseases remains under investigation (98–100). TYK2 is phosphorylated in the IL-4, IL-12, IL-13, and IL-22 cytokine signaling pathways; thus, brepocitinib is believed to be a potentially effective treatment for AD. The efficacy of topical brepocitinib in treating patients between the ages of 12–75 years with AD was evaluated by a dose-ranging, phase 2b RCT (92). In the RCT, brepocitinib met the primary endpoint of a mean change in EASI score at Week 6 and demonstrated superiority over the vehicle treatment. The rate of AEs in the brepocitinib group was generally lower than that in the vehicle group.

ATI-502, a topical JAK1 and JAK3 inhibitor, was investigated for efficacy and safety in the treatment of AD in a phase 2 open-label study (93). ATI-502 demonstrated high efficacy through improved EASI score, IGA score, and itch-NRS. ATI-502 was also demonstrated to be well tolerated, with primary safety endpoints met (93, 101).

Upadacitinib, an oral reversible JAK inhibitor with greater inhibitory potency for JAK1 than JAK2, JAK3, and TYK2, has been approved by the U.S. FDA for treating AD, rheumatoid arthritis, and psoriatic arthritis (102). According to the FDA recommendation, the strategy for upadacitinib treatment in patients with AD aged 12-64 years is initiation from 15 mg/day and titration to 30 mg/day if necessary ( Table 3 ). The same guideline proposed a 15 mg/day dose for patients aged ≥65 years (102). In two consecutive phase 3 RCTs (Measure Up 1 and Measure Up 2), both the 30 and 15 mg doses of upadacitinib presented significantly higher EASI and IGA responses than the placebo (104). Furthermore, upadacitinib combined with TCS demonstrated high tolerability and sustainable efficacy over a 52-week treatment period (123).

Dupilumab was designed as a fully human monoclonal antibody targeting IL-4 and IL-13 receptors and was considered a promising biologics for AD. Nevertheless, less than 50% of patients achieved clear or almost clear skin during a 16-week course of dupilumab monotherapy (13), leading to the search for more effective therapeutic options. Compared with the 50% success rate for dupilumab, patients treated with upadacitinib presented substantially better results, with ≥75% of patients achieving an improvement in EASI score from baseline (EASI75) (106). Upadacitinib was also superior to dupilumab in terms of several secondary endpoints, including the percentage of patients achieving ≥90% and 100% improvement in EASI score (EASI90 and EASI100) and the percentage change of Worst Pruritus NRS at week 16 (106). In contrast, patients receiving upadacitinib presented higher rates of serious infection, eczema herpeticum, herpes zoster, and laboratory-related AEs than patients receiving dupilumab (106).

Overall, the rate of AEs associated with upadacitinib was relatively low, and the frequency of AEs was reported to be similar between the 15 and 30 mg doses. The 30 mg dose was associated with a slightly higher rate of treatment discontinuation because of AEs and serious infection than the 15 mg dose, which was considered to be attributed to the dosage effect (124). The most common treatment-emergent AEs (TEAEs) of upadacitinib were acne, upper respiratory tract infection, nasopharyngitis, and headache (104). Nevertheless, long-term use of upadacitinib demonstrated a favorable risk–benefit profile in treating patients with moderate-to-severe AD (124). Therefore, upadacitinib remains a promising orally systemic therapeutic alternative to conventional therapies and subcutaneous biologics for patients with moderate-to-severe AD.

Baricitinib was approved by the U.S. FDA as a selective JAK1 and JAK2 inhibitor for the indications of rheumatoid arthritis and Coronavirus Disease 2019 (COVID-19) but not yet for atopic dermatitis (125). Nevertheless, it has been approved in European union and Japan for adult patients with moderate to severe AD (126). Dosages of 4, 2, and 1 mg baricitinib were used in various RTCs to evaluate efficacy. In a phase 2 study, baricitinib 4 and 2 mg both effectively improved EASI scores when combined with TCS treatment (107). Among the patients with moderate-to-severe AD who did not respond to TCS therapy, monotherapy of baricitinib provided rapid itch reduction and demonstrated significant amelioration of AD severity relative to the placebo. Additionally, baricitinib improved the patients’ quality of life with fair tolerability (80). Furthermore, baricitinib demonstrated sustained long-term efficacy of up to 68 weeks and greater amelioration of itch-related symptoms throughout long-term administration (108). The frequency of serious infections and opportunistic infections was similar to that of the placebo (111). Notably, the 4-mg dose of baricitinib presented a higher risk of herpes simplex than the 2-mg dose, which was considered to be attributed to the dosage effect (111). Overall, baricitinib alone or in combination with TCS is an effective and well-tolerated systemic therapy for patients with moderate-to-severe AD.

Abrocitinib, a selective JAK1 inhibitor, is U.S. FDA approved for treating refractory, moderate-to-severe AD, with a recommended dosage of 100 mg/day (127). A dose-ranging phase 2b study reported that only abrocitinib 200 and 100 mg/day demonstrated significant efficacy and tolerability compared with the placebo, whereas the outcomes for the 30 and 10 mg doses of abrocitinib were statistically nonsignificant (113). Several phase 3 RCTs have demonstrated that the 200 and 100 mg doses of abrocitinib as monotherapy or concomitant with topical agents were highly beneficial treatments for patients aged ≥12 with moderate-to-severe AD (113–116). Compared with dupilumab, the 200 mg dose of abrocitinib showed a better IGA 0/1 or EASI75 response during the 16-week treatment period but did not reach statistical significance (117). The 200-mg dose of abrocitinib was superior to dupilumab with respect to itch response at week 2 (117). In a recent published phase 3 trial, abrocitinib 200 mg once per day provided higher rates of early itch reduction than dupilumab 300 mg (a 4 point or higher improvement in Peak Pruritus Numerical Rating Scale [PP-NRS4] response at week 2) which complements the above findings (120). Notably, a statistical superiority of abrocitinib versus dupilumab on the primary endpoints with faster onset of high-level improvement of disease signs (EASI-90 response at week 4), and the difference in EASI-90 response remained significant at week 16 (120). Furthermore, abrocitinib demonstrated sustainable long-term efficacy through 52 weeks (118). The results of these trials are supporting the potential important role of JAK1 inhibition in patients who need fast relief of pruritus and skin inflammation in atopic dermatitis (113–118, 120).

The most common TEAEs associated with abrocitinib were nausea, upper respiratory tract infection, nasopharyngitis, and headache, which were similar to the AEs reported for other JAK inhibitors (114). Nausea, the most frequent gastrointestinal AE, occurred only in patients with fasted status, indicating that the gastrointestinal symptoms may be caused by the local gastric concentration of abrocitinib (128). Accordingly, abrocitinib is an effective and safe small molecules for treating patients aged ≥12 years with moderate-to-severe AD.

ASN002 is a potent dual inhibitor, which simultaneously blocks both spleen tyrosine kinase (SYK) and JAKs, but is yet to receive U.S. FDA approval. SYK is a nonreceptor tyrosine kinase involved in the regulation of B cells, the differentiation of dendritic cells, the inhibition of keratinocyte differentiation, and Th17 signaling (129).

Patients with AD who were treated with ASN002 exhibited robust tissue improvement in cellular infiltration (AD lesions), inflammatory pathogenesis, and skin barrier dysfunction (129). Clinically, ASN002 is a safe and well-tolerated therapy for patients with moderate-to-severe AD. Mild AEs were observed, including headache, nausea, and diarrhea. However, only the highest dose (80 mg, as opposed to 40 or 20 mg) of ASN002 demonstrated significantly greater amelioration of pruritus than the placebo. A phase 2b RCT was completed but the data were unavailable at the time of publication.

SHR0302 is a highly selective JAK1 inhibitor that also exercises a mild JAK2 inhibiting effect. In laboratory studies, SHR0302 was revealed to minimize the risk of AEs such as neutropenia and anemia (36). SHR0302 is easily absorbed through oral administration, and pharmacodynamic evidence supports a once-daily treatment strategy. A phase 2 study demonstrated that both 8 and 4 mg doses of SHR0302 were superior to a placebo in achieving IGA0/1, EASI75, and an improvement of ≥3 points on the NRS (NRS-3) (122). The oral therapy was also well tolerated, without clinically significant changes in hematology parameters. Further studies evaluating the efficacy and safety of SHR0302 are warranted.