JAK Inhibitor Safety in Atopic Dermatitis and Alopecia Areata: A 4.5-Year Real-World Retrospective Cohort Study

Abstract

Introduction

JAK inhibitors (JAKi) represent a novel therapeutic approach for treating the immune-mediated dermatologic diseases alopecia areata (AA) and atopic dermatitis (AD). JAKi have shown a favorable safety profile in controlled clinical trials; however, real-world safety data in routine dermatologic practice remain limited. The objective of this study is to evaluate the safety of JAKis (abrocitinib, baricitinib, upadacitinib, and ritlecitinib) in patients with AA or AD in a real-world setting.

Methods

We conducted a single-center retrospective observational study based on electronic medical record data from December 2020 to June 2025. Adult patients with a confirmed diagnosis of atopic dermatitis or alopecia areata who were treated as per routine clinical practice with a dermatologic JAKi were included. Adverse events (AE) were retrospectively collected.

Results

A total of 376 adult patients were included. AE events were observed across all agents and were predominantly mild to moderate. Mild total cholesterol increase, weight gain, and upper respiratory tract infection were the most common AE. Discontinuation was mainly driven by loss of efficacy, while AE-related discontinuations were less frequent and heterogeneous in nature. Importantly, only one case of venous thrombo-embolism was reported in a patient with other thrombotic risk factors, and no cases of major adverse cardiovascular events (MACEs) were recorded. Four cases of malignancies were reported. Notably, sex-specific differences emerged in metabolic adverse events, with weight gain occurring more frequently among female patients. Menstrual alteration, although rare, was a newly described safety event and was the leading cause of treatment discontinuation due to safety issues.

Conclusion

While no safety signals regarding MACEs or cancer emerged, other metabolic events, such as cholesterol increase, weight gain and menstrual alterations, were identified. These findings highlight that routine clinical practice represents a less controlled and more heterogeneous setting than randomized trials, underscoring the importance of continuous real-world safety monitoring to fully characterize treatment-associated risks.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s13555-026-01867-y.

Plain Language Summary

Janus kinase inhibitors (JAKi) are a group of oral medications increasingly used to treat severe forms of inflammatory skin diseases such as atopic dermatitis and alopecia areata. While clinical trials have shown that these drugs are effective and generally safe, patients included in trials are carefully selected and closely monitored. For this reason, it is important to understand how JAKi perform in everyday clinical practice, where patients are more diverse and monitoring is less standardized. This study examined the real-world safety of JAKi in patients treated at a single dermatology center in Italy over 4.5 years. Overall, the treatments were well tolerated. Serious safety problems, such as major cardiovascular events, blood clots, or aggressive cancers, were extremely rare and comparable to the general population. Most side effects were mild or moderate and manageable in routine care. Some side effects occurred more frequently, including increases in cholesterol levels, weight gain, mild infections such as common respiratory infections and acne. Weight gain was more common in female patients, and menstrual changes, although rare, were reported in women of childbearing age and frequently led to treatment discontinuation. Most patients discontinued therapy because the drug became less effective over time rather than because of safety concerns. In summary, this real-world study supports the overall safety of JAKi for dermatologic conditions, while highlighting the importance of regular follow-up to monitor especially metabolic and sex-specific side effects.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s13555-026-01867-y.

Key Summary Points

Why carry out this study?

Janus kinase inhibitors (JAKi) have emerged as effective treatments for immune-mediated dermatologic conditions such as alopecia areata and atopic dermatitis. Randomized controlled trials have demonstrated a generally favorable safety profile; however, evidence on their safety in real-world clinical practice remains limited.

Given the high prevalence of atopic dermatitis and alopecia areata, robust real-world safety data remain a major unmet need to define the benefit–risk profile of JAK inhibitors in the heterogeneous populations encountered in routine clinical practice, which are not fully represented in randomized clinical trials.

This study was conducted to provide a comprehensive characterizaction of the real-world safety profile of JAK inhibitors in patients with atopic dermatitis and alopecia areata including treatment-emergent adverse events, treatment discontinuation patterns adverse events of special interest, and early or low-grade clinical and laboratory abnormalities.

What was learned from the study?

This study provides real-world safety data on dermatologic JAK inhibitors in a cohort of 376 patients treated in routine clinical practice. It confirms the absence of major safety signals regarding major adverse cardiovascular events (MACEs) and malignancies, while identifying metabolic adverse events, including increases in cholesterol levels and weight gain. Additionally, it highlights potential sex-specific differences in adverse events and reports menstrual alterations as a novel safety signal that may impact treatment discontinuation.

These findings support the overall safety of JAK inhibitors in routine dermatologic practice but emphasize the need for ongoing monitoring, particularly of metabolic parameters. Clinicians should be aware of potential sex-specific adverse events, including menstrual alterations. The results underscore the importance of real-world evidence to complement clinical trial data and to better inform long-term risk–benefit assessment in diverse patient populations.

Introduction

Janus kinase (JAK) inhibitors (JAKi) represent a novel therapeutic approach for treating immune-mediated dermatologic diseases. Atopic dermatitis (AD) and alopecia areata (AA) are inflammatory diseases characterized by dysregulated cytokine signaling. In AD, type 2 inflammation driven by interleukin (IL)−4, IL-13, and IL-31 plays a central pathogenic role, whereas AA is primarily mediated by cytotoxic T-cell responses and interferon-γ signaling, leading to the collapse of hair follicle immune privilege. Many of these cytokines signal through the -signal transducer and activator of transcription (STAT) pathway, making JAK inhibition an effective therapeutic strategy for both conditions.

Abrocitinib, baricitinib, and upadacitinib are approved in the European Union for the treatment of AD, whereas baricitinib and ritlecitinib are approved for AA. Randomized controlled trials have consistently demonstrated significant improvements in disease severity and quality of life across these indications [1–5], together with a favorable safety profile [1, 5–15]. However, despite the increasing use of JAKi in clinical practice, real-world evidence on their long-term safety remains scarce and limited by small sample sizes [6, 12, 16–30]. This study aimed to assess the incidence, nature, and clinical impact of adverse events associated with JAKi in patients with AD and AA in real-world clinical practice.

In 2023, the Pharmacovigilance Risk Assessment Committee (PRAC) of the European Medicines Agency (EMA) issued updated safety recommendations for JAKi, advising caution in patients at increased risk of major cardiovascular events, venous thromboembolism, malignancy, or serious infections. These recommendations highlighted the need for real-world data in dermatologic populations [31]. Accordingly, this study evaluates adverse events in patients with atopic dermatitis or alopecia areata treated with JAKi, with a focus on PRAC-relevant outcomes.

Adverse event definitions in real-world studies may differ from those used in randomized clinical trials, which generally rely on protocol-driven definitions and CTCAE (Common Terminology Criteria for Adverse Events)-based grading systems. In this study, adverse events were identified using clinically relevant thresholds commonly applied in routine practice and derived from established clinical guidelines, allowing the capture of mild laboratory and clinical abnormalities that may be overlooked in trial settings. For example, hypercholesterolemia was defined according to NCEP ATP III criteria and categorized as mild (total cholesterol 200–239 mg/dL) or severe (≥240 mg/dL), whereas clinical trials typically focus on higher-grade laboratory abnormalities. Similarly, weight changes were recorded and categorized according to percentage variation from baseline body weight, including changes that may not meet conventional adverse event severity thresholds used in clinical trials. This approach aimed to provide a more patient-centered assessment of treatment-related effects relevant to routine clinical care.

Methods

This is a single-center retrospective observational study based on medical record data collected at IRCCS Humanitas Research Hospital, in Rozzano, Milan, Italy. Eligible participants were adults with severe AA or AD who were treated with abrocitinib, baricitinib, ritlecitinib, or upadacitinib for at least 4 weeks during the study period (December 2020–June 2025). Patients with incomplete demographic data were excluded. Safety data related to JAKi therapy were retrospectively extracted from electronic medical records. Laboratory monitoring was performed according to routine clinical practice rather than a predefined study protocol. Frequency and timing of laboratory assessments varied according to treating physician judgment, patient characteristics, and evolving clinical recommendations during the study period. This was a non-interventional observational study. According to local regulations, formal ethics committee approval and written informed consent were not required. The study was notified to the Italian Medicines Agency (AIFA, Study ID 2805, study code JAKiSafe), and the competent Ethics Committee acknowledged the study. All procedures were conducted in accordance with the Declaration of Helsinki and applicable national regulations.

Statistical Analysis

Demographic data are described as number and percentage if categorical, or mean and standard deviation, if continuous and approximately Gaussian, or median and range otherwise.

Adverse events (AE) rates are presented as exposure-adjusted event (E) rates (EAERs) or exposure-adjusted incidence rates (EAIRs) per 100 PY (PY) for the entire treatment period to adjust for potentially different durations of follow-up. To calculate EAERs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; a single patient could contribute more than 1 event to the numerator. EAIRs were calculated taking the number of patients experiencing a specific AE divided by the total exposure time for patients within that treatment group at risk of an initial occurrence of that event; patients could only contribute once to the numerator of EAIR.

Because most safety outcomes were rare and several treatment groups had very low or zero event counts, statistical models could not be reliably estimated. Therefore, only descriptive statistics (event counts and EAERs/EAIRs) were reported, without formal hypothesis testing. Confidence intervals were not calculated, as the sparse nature of the data would have resulted in wide and unstable estimates that could be misleadingly interpreted as comparative measures of precision between treatment groups. This approach is consistent with methodological recommendations for the descriptive reporting of sparse safety data.

Results

Baseline Characteristics and JAKi Exposure

A total of 416 patients were treated with a JAKi inhibitors for dermatologic conditions since December 2020 to June 2025. 376 patients were treated for at least 4 weeks and had follow-up data at the time of analysis and were included in the study. Total years of exposure was 521,90 years (mean of 1.39 years per patient), specifically 249.78 years for upadacitinib, 184.38 years for baricitinib, 45.47 years for ritlecitinib, and 42.27 years for abrocitinib. A more detailed description of treatment frequencies is available in Supplementary material and in Suppl. Table 1.

Baseline demographics are reported in Table 1. The overall mean age was 42.56 years (SD 16.16), with comparable age distributions across treatment groups and 60% of patients were female. Body mass index was predominantly within the normal weight range, with 76.3% of patients having a BMI <25 kg/m2, while overweight and obesity were observed in 18.6% and 5.0% of patients, respectively. Hypercholesterolemia, thyroid disease, and hypertension were the most common comorbidities. Smoking was reported in a minority of patients (active smokers were 12.2%, previous smokers 6.7%). A more detailed description of comorbidities is available in Supplementary material.

Table 1.

Demographics and comorbidities of the study population

Characteristic	Variable	TOTAL (n=376)	UPA (n=150)	BARI (n= 145)	RITLE (n=71)	ABRO (n=44)
		N (% or SD)	N (% or SD)	N (% or SD)	N (% or SD)	N (% or SD)
Sex	Male	151 (40.2%)	85 (56.7%)	42 (29.0%)	19 (26.8%)	22 (50.0%)
	Female	225 (59.8%)	65 (43.3%)	103 (71.0%)	52 (73.2%)	22 (50.0%)
Age		42.6 (SD 16.2)	42.4 (SD 17.2)	44.6 (SD 13.8)	39.4 (SD 16.2)	40.5 (SD 16.4)
BMI		23.59 (SD 4.16)	24.3 (SD 3.9)	23.15 (SD 3.7)	22.0 (SD 4.8)	24.31 (SD 3.9)
	<25 kg/m2	287 (76.3%)	104 (69.3%)	114 (78.6%)	56 (78.9%)	33 (75.0%)
	>25 and < 30 kg/m2	70 (18.6%)	37 (24.7%)	25 (17.2%)	12 (16.9%)	8 (18.2%)
	>30 kg/m2	19 (5.1%)	9 (6.0%)	6 (4.1%)	3 (4.2%)	3 (6.8%)
Previous clinical history and comorbidities	High cholesterol	64 (17.0%)	12 (8.0%)	38 (26.2%)	16 (22.5%)	6 (13.6%)
	Thyroid disease	52 (13.8%)	5 (3.3%)	37 (25.5%)	13 (18.3%)	1 (2.3%)
	HTN	49 (13.0%)	24 (16.0%)	20 (13.8%)	6 (8.5%)	8 (18.2%)
	Cancer history	5 (1.3%)	3 (2.0%)	2 (1.4%)	0 (0.0%)	0 (0.0%)
	Diabetes	4 (1.1%)	2 (1.3%)	1 (0.7%)	2 (2.8%)	0 (0.0%)
	MACE*	1 (0.3%)	1 (0.7%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
	VTE+	1 (0.3%)	1 (0.7%)	0 (0.0%)	0 (0.0%)	0 (0.0%)
	Previous abortions	1 (0.3%)	0 (0.0%)	1 (0.7%)	0 (0.0%)	0 (0.0%)
Smoke	Yes	46 (12.2%)	22 (14.7%)	20 (13.8%)	4 (5.6%)	3 (6.8%)
	No	179 (48.6%)	104 (69.3%)	48 (33.1%)	38 (53.5%)	2 (27.3%)
	Previous smoker	25 (6.7%)	12 (8.0%)	11 (7.6%)	1 (1.4%)	1 (2.3%)
	UNK	126 (33.5%)	12 (8.0%)	66 (45.5%)	28 (39.4%)	28 (63.6%)

^*MACE is defined as nonfatal myocardial infarction and nonfatal stroke and includes acute myocardial infarction, myocardial infarction, cerebellar stroke, and ischemic stroke

⁺VTE includes deep vein thrombosis and pulmonary embolism

ABRO abrocitinib; BARI barictinib; BMI Body mass index; HTN hypertension; MACE major adverse cardiovascular events; SD standard deviation; RITLE ritlecitinib; UNK unknown; UPA Upadacitinib, VTE venous thromboembolism

Regarding major adverse cardiovascular events (MACEs, defined as nonfatal myocardial infarction and nonfatal stroke), only one patient had a previous history of myocardial infarction (treated with upadacitinib 15 mg), whereas three other patients had a history of angina and therefore did not qualify as MACE. Two patients, one treated with ritlecitinib 50 mg and one with upadacitinib 15 mg had a previous history of stable angina, and one patient, treated with baricitinib 2 mg, had coronary diseases that required coronary artery stents. Only one patient treated with upadacitinib 15 mg had a previous history of pulmonary embolism related to a motor-vehicle accident. Five patients with a history of cancer (diagnosed at least five years prior to initiation of JAKi therapy) were identified. One patient treated with upadacitinib 30 mg had a history of breast cancer, two patients (one patient treated with upadacitinib 15 mg and one with baricitinib 4 mg) had a previous melanoma, one patient treated with upadacitinib 15 mg had a history of hairy cell leukemia, one patient treated with baricitinib 4 mg had a history of cervical CIN2.

General Safety Overview

Across a total exposure of 521.9 patient-years, 583 treatment-emergent adverse events were recorded, corresponding to an overall incidence rate of 111.7 events per 100 PY. AEs occurred with variable frequency among individual agents (Table 2). The most frequent adverse events were mild total cholesterol increase (defined as total cholesterol 200–240 mg/dL, 18.01 per 100 PY), weight gain (defined as increase in at least 5% of body weight, 8.62 per 100 PY), and upper respiratory tract infection (8.05 per 100 PY). Increased cholesterol levels and weight gain are detailed reported in paragraph 3.5 and 3.6, respectively.

Table 2.

Most common adverse events during treatment with JAKi

AE	Total JAKi (n=376) (PY=521,90)		UPA (n=150) (PY=249.78)		BARI (n=145) (PY=184.38)		ABRO (n=44) (PY=42.27)		RITLE (n=71) (PY=45.47)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
Any AE	583	111.70	315	126.11	192	104.13	16	37.85	61	134.15
AE leading to discontinuation	36	6.89	18	7.02	10	5.42	6	14.19	2	4.40
Serious AE	10	1.92	4	1.60	4	2.17	2	4.73	0	0.00
SAE leading to discontinuation	7	1.34	4	1.60	2	1.80	1	2.31	0	0.00
Most frequent AE
New onset mild hypercholesterolemia 200–240 mg/dL	94	18.01	49	19.62	34	19.53	0	0.00	11	24.19
Weight Gain≥5%	45	8.62	21	8.41	20	10.85	2	4.73	2	4.40
URI	42	8.05	28	11.21	11	5.97	1	2.37	2	4.40
Acne	38	7.28	28	11.21	8	4.34	0	0.00	2	4.40
Herpes zoster reactivation	17	3.26	17	6.81	0	0.00	0	0.00	0	0.00
Oral herpes simplex	13	2.49	6	2.40	5	2.71	0	0.00	2	4.40
Fatigue	8	1.53	3	1.20	3	1.08	0	0.00	2	4.40
Fever	6	1.15	3	1.20	2	1.08	0	0.00	1	2.20
Nausea	6	1.15	3	1.20	1	0.54	1	2.37	1	2.20
Impetigo	5	0.96	3	1.20	2	1.08	0	0.00	0	0.00
UTI	5	0.96	2	0.80	3	1.63	0	0.00	0	0.00
Infrequent AE
Urticaria	4	0.77	2	0.80	1	0.54	1	2.37	0	0.00
Muscle cramps	4	0.77	1	0.40	1	0.54	0	0.00	2	4.40
Pneumonitis or Bronchitis	3	0.57	2	0.80	1	0.54	0	0.00	0	0.00
Diarrhea	3	0.57	1	0.40	0	0.00	0	0.00	3	6.60
Common warts	3	0.57	3	1.20	0	0.00	0	0.00	0	0.00
Headache	3	0.57	3	1.20	0	0.00	0	0.00	0	0.00
Joint pains	2	0.38	2	0.80	0	0.00	0	0.00	0	0.00
Increased hunger	2	0.38	1	0.40	0	0.00	0	0.00	1	2.20
New onset hypertension	2	0.38	1	0.40	0	0.00	0	0.00	1	2.20
Worsening hypertesion	2	0.38	2	0.80	0	0.00	0	0.00	0	0.00
Urolithiasis	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
Gengivitis	1	0.19	0	0.00	0	0.00	0	0.00	1	2.20
Pancreatitis	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
Face hirsutism	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
Vertigo	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
COPD worsening	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00

Data are reported as EAERs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; a single patient could contribute more than 1 event to the numerator

ABRO abrocitinib; AE adverse events; BARI barictinib; N number of events; PY patient/year; RITLE ritlecitinib; UPA Upadacitinib; URI upper respiratory tract infections; UTI lower urinary tract infections.

Upper respiratory tract infections represented the most common infectious AE (8.05 per 100 PY), with the highest incidence observed in patients treated with upadacitinib (11.21 per 100 PY; upadacitinib 30 mg, 12.03 per 100 PY; upadacitinib 15 mg, 10.80 per 100 PY), followed by baricitinib (5.97 per 100 PY; baricitinib 4 mg, 6.32 per 100 PY; baricitinib 2 mg, 0.00 per 100 PY) and ritlecitinib (4.40 per 100 PY). Lower urinary infections were infrequent overall (0.96 per 100 PY) but occurred more frequently with baricitinib (1.63 per 100 PY, exclusively in the baricitinib 4 mg group) than with upadacitinib (0.80 per 100 PY), while no cases were reported with abrocitinib or ritlecitinib. Infective events affecting the skin were also common. Herpes zoster reactivation was recorded exclusively in the upadacitinib group (6.81 per 100 PY), both in the 30 mg and in the 15 mg groups (9.62 per 100 PY and 5.40 PY per 100 PY, respectively), whereas oral herpes simplex infections occurred across multiple agents and doses, with rates ranging from 2.40 to 4.40 per 100 PY (Table 2).

Dermatologic AEs were also common, particularly acne (7.28 per 100 PY), which was the fourth most frequent adverse event, and which showed the highest incidence with upadacitinib (11.21 per 100 PY). Other infrequent adverse events included urticaria, gastrointestinal symptoms (diarrhea), musculoskeletal complaints (muscle cramps and arthralgia), fatigue, fever, new-onset or worsening hypertension, gingivitis, increased hunger, pancreatitis, facial hirsutism, vertigo, and worsening of chronic obstructive pulmonary disease. All occurred at low frequencies and without a consistent treatment-specific pattern (Supplementary Materials).

Adverse Event of Special Interest (Serious Infections, MACEs, VTE and Malignancies)

Serious infections, MACEs, VTE, and malignancies were rare (Table 3). Excluding non-melanoma skin cancer (NMSC), malignancies included one stage IA melanoma (abrocitinib 100 mg), one mantle cell lymphoma (upadacitinib 15 mg), and two cases under baricitinib 4 mg (ductal breast carcinoma and a grade 1 pancreatic neuroendocrine tumor). Three basal cell carcinomas (NMSC) were reported and fully excised (occurring in patients under upadacitinib, baricitinib and abrocitinib). One case of Pneumocystis jirovecii pneumonia occurred under upadacitinib 30 mg, requiring hospitalization and resolving with full recovery. Thromboembolic events included one asymptomatic pulmonary embolism under upadacitinib 30 mg in a high-risk patient (heavy smoker and alcohol consumer with a history of JAK2-negative polycythemia) and one case of Mondor syndrome (a rare, benign, self-limited superficial thrombophlebitis of the thoracoepigastric vein) under upadacitinib 15 mg, both resolving without complications. No MACEs were recorded.

Table 3.

AE of special interest

AE of special interest	Total JAKi (n=376) (PY=521.90)		UPA (n=150) (PY=249.78)		BARI (n=145) (PY=184.38)		ABRO (n=44) (PY=42.27)		RITLE (n=71) (PY=45.47)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
Malignancies	4	0.77	1	0.40	2	1.08	1	2.37	0	0.00
NMSC	3	0.57	1	0.40	1	0.54	1	2.37	0	0.00
Serious infections	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
VTE	1	0.19	1	0.40	0	0.00	0	0.00	0	0.00
Thromboflebitis	1	0.19	0	0.00	1	0.54	0	0.00	0	0.00
MACE	0	0.00	0	0.00	0	0.00	0	0.00	0	0.00

Data are reported as EAERs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; a single patient could contribute more than 1 event to the numerator

^*MACEs is defined as nonfatal myocardial infarction and nonfatal stroke and includes acute myocardial infarction, myocardial infarction, cerebellar stroke, and ischemic stroke

⁺VTE includes deep vein thrombosis and pulmonary embolism preferred terms

ABRO abrocitinib; AE adverse events; BARI barictinib; MACEs major adverse cardiovascular events; N number of events; NMSC non-melanoma skin cancer; PY patient/year; RITLE ritlecitinib; UPA Upadacitinib; VTE venous thromboembolism

Laboratory Parameters Alterations

Across the JAK inhibitor cohort, hematologic and biochemical abnormalities were generally infrequent (Table 4). Hemoglobin reductions (3.26 per 100 PY) and leukocyte abnormalities (2.11 per 100 PY) were observed mainly with upadacitinib, while neutropenia (2.49 per 100 PY) occurred across agents, with higher rates for abrocitinib and ritlecitinib, despite likely influenced by small sample sizes and exposure time. Lymphocyte reductions were rare (0.96 per 100 PY), and leukocyte differential inversion (a reversal of the normal relative proportions of leukocyte subpopulations in the peripheral blood differential count) was reported only with upadacitinib. Thrombocytopenia was uncommon, occurring mainly with upadacitinib and abrocitinib. Elevated creatine phosphokinase (CPK) levels were always asymptomatic and were reported exclusively with upadacitinib, although incidence may be underestimated due to reduced monitoring in clinical practice over time [32]. Liver enzyme alterations were heterogeneous, with γ-GT, ALT, and AST elevations observed primarily under baricitinib and, to a lesser extent, upadacitinib, while no hepatic abnormalities were reported with abrocitinib.

Table 4.

Laboratory alterations during treatment with JAKi

Laboratory AE	Total JAKi (n=376) (PY=521,90)		UPA (n=150) (PY=249.78)		BARI (n=145) (PY=184.38)		ABRO (n=44) (PY=42.27)		RITLE (n=71) (PY=45.47)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
Hemoglobin decrease	17	3.26	11	4.40	2	1.08	1	2.37	3	6.60
CPK elevation	14	2.68	14	5.60	0	0.00	0	0.00	0	0.00
Decreased Neutrophils count	13	2.49	6	2.40	3	1.63	2	4.73	2	4.40
Decreased Leucocytes count	11	2.11	7	2.80	4	2.17	0	0.00	0	0.00
ALT elevation	10	1.92	1	0.40	8	4.34	0	0.00	1	2.20
γ-GT elevation	9	1.72	4	1.60	5	2.71	0	0.00	0	0.00
LF inversion	8	1.53	8	3.20	0	0.00	0	0.00	0	0.00
Decreased Platelets count	8	1.53	6	2.40	0	0.00	2	4.73	0	0.00
AST elevation	8	1.53	1	0.40	6	3.25	0	0.00	1	2.20
Decreased Lymphocytes count	5	0.96	4	1.60	1	0.54	0	0.00	0	0.00

Data are reported as EAERs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; a single patient could contribute more than 1 event to the numerator. The data in the table are presented in descending order of frequency per patient-year (PY), with higher frequencies listed first.

ABRO abrocitinib; AE, adverse events; ALT alanine aminotransferase; AST aspartate aminotransferase; BARI baricitinib; CPK creatine phosphokinase; G-GT gamma-glutamyl transferase; TLF leucocyte formula; N, number of events; PY patient/year; RITLE ritlecitinib; UPA upadacitinib

Lipid Elevation

Hypercholesterolemia was defined according to NCEP ATP III criteria, distinguishing between mild (200–239 mg/dL) and severe (≥240 mg/dL), and between new-onset and worsening of pre-existing conditions; hypertriglyceridemia was similarly categorized using the thresholds of mild (150–300 mg/dL) and severe hypertriglyceridemia (≥300 mg/dL). Due to the smaller sample size and shorter exposure with ritlecitinib, PY-based estimates should be interpreted with caution, and absolute event counts may be more reliable, particularly given the more recent implementation of lipid monitoring in routine clinical practice. Overall (Table 5), new-onset hypercholesterolemia was common (18.01 per 100 PY), particularly with upadacitinib (49 events; 19.96 per 100 PY) and baricitinib (34 events; 19.53 per 100 PY), while ritlecitinib showed fewer events (11) but higher incidence rates (24.19 per 100 PY); no cases were observed with abrocitinib. Severe hypercholesterolemia (≥240 mg/dL; 9.39 per 100 PY) followed a similar pattern, with baricitinib (11.49 per 100 PY) and upadacitinib (9.21 per 100 PY) showing the highest burden. Worsening of pre-existing hypercholesterolemia was also observed across treatments, particularly with ritlecitinib (13.20 per 100 PY) and baricitinib (5.74 per 100 PY). Hypertriglyceridemia was less frequent, with new-onset mild cases reported in 5.56 per 100 PY (upadacitinib 5.60 per 100 PY; baricitinib 5.17 per 100 PY; ritlecitinib 13.20 per 100 PY), while severe elevations were rare (1.34 per 100 PY), mainly under upadacitinib (2.00 per 100 PY) and baricitinib (1.15 per 100 PY).

Table 5.

Lipid elevation

Lipid elevation	Total JAKi (n=376) (PY=521,90)		UPA (n=150) (PY=249.78)		BARI (n=145) (PY=184.38)		ABRO (n=44) (PY=42.27)		RITLE (n=71) (PY=45.47)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
New onset mild hypercholesterolemia 200–240 mg/dL	94	18.01	49	19.62	34	19.53	0	0.00	11	24.19
New onset severe hypercholesterolemia ≥240 mg/dL	49	9.39	23	9.21	20	11.49	0	0.00	6	13.20
Worsening hypercholesterolemia 200–240 mg/dL (mild)	19	3.64	3	1.20	10	5.74	0	0.00	6	13.20
Worsening hypercholesterolemia ≥240 mg/dL (severe)	15	2.87	4	1.60	8	4.59	2	4.73	1	2.20
New onset mild hypertriglyceridemia 150–300 mg/dL	29	5.56	14	5.60	9	5.17	0	0.00	6	13.20
New onset severe hypertriglyceridemia ≥300 mg/dL	7	1.34	5	2.00	2	1.15	0	0.00	0	0.00
Worsening hypertriglyceridemia150-300 mg/dL (mild)	4	0.77	2	0.80	2	1.15	0	0.00	0	0.00
Worsening hypertriglyceridemia≥300 mg/dL (severe)	1	0.19	0	0.00	0	0.00	0	0.00	1	2.19

Data are reported as EAIRs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; patients could only contribute once to the numerator

ABRO abrocitinib; BARI baricitinib; N number of events; PY patient/year; RITLE ritlecitinib; UPA upadacitinib

Weight Gain

Weight gain emerged as a relevant adverse event during JAKi therapy, as described in a recent report [33]. In the overall cohort (Table 6), mild weight gain (<5% of body weight, BW) occurred in 4.60 per 100 PY, while clinically significant increases (>5% of BW) were more common (8.62 per 100 PY), particularly with baricitinib (10.85 per 100 PY) and upadacitinib (8.01 per 100 PY); abrocitinib and ritlecitinib showed lower rates. Severe weight gain (>10% of BW) was observed in 3.07 per 100 PY, mainly under upadacitinib. Interestingly weight gain occurred more frequently during treatment with upadacitinib 15 mg compared to the higher 30 mg dose (Suppl. Table 2). All weight gain cases recorded in the baricitinib group occurred in the 4 mg subjects. In the abrocitinib group, one case of weight gain (7–10% of BW) occurred in the 100 mg group, while the other (>10% of BW) occurred in the 200 mg group.

Table 6.

Weight gain during treatment with JAKi

Weight gain	Total JAKi (n=376) (PY=521.90)		UPA (n=150) (PY=249.78)		BARI (n=145) (PY=184.38)		ABRO (n=44) (PY=42.17)		RITLE (n=71) (PY=45.47)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
<5% of BW	24	4.60	8	3.20	13	7.05	0	0	3	6.60
>5% of BW	45	8.62	21	8.41	20	10.85	2	4.73	2	4.40
5%−7% of BW	16	3.07	7	2.80	9	4.88	0	0.00	0	0.00
7%−10% of BW	13	2.49	5	2.00	6	3.25	1	2.37	1	2.20
>10% of BW	16	3.07	9	3.60	5	2.71	1	2.37	1	2.20

Data are reported as EAIRs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; patients could only contribute once to the numerator. Weight increase is reported as percent increase of body weight (BW)

ABRO abrocitinib; BARI baricitinib; BW body weight; N number of events; PY patient/year; RITLE ritlecitinib; UPA upadacitinib

Notably, a clear sex-specific pattern was observed (Suppl. Table 3): although mild increases were comparable between sexes, clinically relevant weight gain (>5% of BW) was markedly higher in females compared to males (11.15 vs 5.31 per 100 PY), with severe increases (>10% of BW) also more frequent among women (4.39 vs 1.33 per 100 PY).

Fertile Age Group: Oral Contraceptive Pills Use and Menstrual Alterations

The role of oral contraceptive pills (OCPs) as an independent thrombotic risk factor during JAK inhibitor therapy remains debated [34]. Although some meta-analyses of randomized controlled trials do not show a statistically significant increase in VTE risk with JAKi use in immune-mediated inflammatory disease populations, real-world data and regulatory warnings emphasize the importance of individualized risk stratification [21, 35, 36]. In our cohort, 34% of patients were females of fertile age (defined as female subjects in the fertile age between 15 and 49 years old, as per World Health Organization definition [37]), among whom OCP use was reported in 18.9% (Table 7). OCP use was observed in 16.2% of fertile age patients receiving upadacitinib, 25.5% of those treated with baricitinib, 6.7% of patients on abrocitinib, and 18.2% of those receiving ritlecitinib. No VTE events were observed among OCP users; as described in paragraph 3.3, the only VTE occurred in a male patient, while a single case of Mondor syndrome was reported in a non-OCP user.

Table 7.

OCP use frequencies in the fertile age population

OCP use	Total (n=127)	UPA (n=37)	BARI (n=51)	ABRO (n=15)	RITLE (n=33)
Yes	24 (18.9%)	6 (16.2%)	13 (25.5%)	1 (6.7%)	6 (18.2%)
No	67 (52.8%)	9 (24.3%)	31 (60.8%)	5 (33.3%)	26 (78.8%)
UNK	36 (28.3%)	22 (59.5%)	7 (13.7%)	9 (60.0%)	1 (3.0%)

Data are reported as number and percentage (in brackets)

ABRO abrocitinib; BARI baricitinib; N number of events; RITLE ritlecitinib; UNK unknown; UPA upadacitinib

Sex-specific adverse events, particularly menstrual disturbances (Table 8), emerged as clinically relevant findings. Overall, 18 cases of menstrual disturbances (including amenorrhea, hyper/hypomenorrhea, poly/oligomenorrhea, and dysmenorrhea) were observed (12.37 per 100 PY), most frequently with baricitinib (9 cases among 51 patients, incidence of 13.41 per 100 PY, all in the 4 mg group) and upadacitinib (7 events among 37 patients, incidence of 10.67 per 100 PY, with 5 patients receiving 15 mg and 2 patients receiving the 30 mg dosage), while only 2 cases occurred with abrocitinib and none with ritlecitinib. Notably, menstrual alterations led to treatment discontinuation in 10 patients, representing a major cause of treatment interruption due to safety as described in paragraph 3.8.

Table 8.

Menstrual alterations in the fertile age population

Events	Total (n=127) (PY=145.48)		UPA (n=37) (PY=65.62)		BARI (n=51) (PY=67.12)		ABRO (n=15) (PY=13.19)		RITLE (n=33) (PY=25.30)
	N	P/Y	N	P/Y	N	P/Y	N	P/Y	N	P/Y
Menstrual Disturbances	18	12.37	7	10.67	9	13.41	2	15.16	0	0.00

Data are reported as EAIRs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; patients could only contribute once to the numerator

ABRO abrocitinib; BARI baricitinib; N number of events; PY patient/year; RITLE ritlecitinib; UPA upadacitinib

Discontinuation

Treatment discontinuation (Table 9) occurred in 135 patients over 521.9 patient-years (25.87 per 100 PY), with rates highest for abrocitinib (48.53 per 100 PY) and lowest for ritlecitinib (17.59 per 100 PY), while upadacitinib and baricitinib showed intermediate values (21.62 and 28.20 per 100 PY). The higher discontinuation rate observed for abrocitinib and ritlecitinib may be influenced by their lower cumulative exposure, which can amplify incidence estimates in smaller patient-year denominators. Secondary inefficacy was the leading cause of discontinuation (9.2 per 100 PY), followed by adverse events and primary inefficacy, particularly under abrocitinib. Dose-related differences were observed, with higher discontinuation rates at upadacitinib 30 mg and baricitinib 4 mg (Suppl. Table 4). Discontinuations attributed to clinical remission reflected dose de-escalation rather than true cessation of therapy and it was observed exclusively among patients treated with upadacitinib and baricitinib at the highest doses. Less frequent reasons included loss to follow-up (0.57 per 100 PY) and pregnancy planning (0.38 per 100 PY).

Table 9.

Reasons for drug discontinuation across JAKi

Discontinuation	Total JAKi (n=376) (PY=521.90)		UPA (n=150) (PY=249.78)			BARI (n=145) PY=184.38)			ABRO (n=44) (PY=42.27)		RITLE (n=71) (PY=45.47)
	N	P/Y	N		P/Y	N		P/Y	N	P/Y	N		P/Y
Total	135	25.87	54		21.62	52		28.20	21	48.53	8		17.59
Secondary inefficacy	48	9.20	17		6.81	23		12.47	6	13.87	2		4.40
Adverse events	40	7.66	19		7.61	12		6.51	7	16.18	2		4.40
Primary inefficacy	25	4.79	9		3.60	5		2.71	8	18.49	3		6.60
Clinical remission	16	3.07	7		2.80	9		4.88	0	0.00	0		0.00
Lost to FU	3	0.57	1		0.40	1		0.54	0	0.00	1		2.20
Desire of pregnancy	2	0.38	0		0.00	2		1.08	0	0.00	0		0.00
Other	1	0.19	1		0.40	0		0.00	0	0.00	0		0.00

Data are reported as number and EAERs, the number of AEs reported was divided by the total exposure time for patients within that treatment group multiplied by 100; a single patient could contribute more than 1 event to the numerator

ABRO abrocitinib; BARI baricitinib; N number of events; PY patient/year; RITLE ritlecitinib; UPA upadacitinib

Adverse events leading to discontinuation (Table 9, Table 10 and Supplementary Material) were reported in 36 patients (40 events), most commonly under abrocitinib (16.18 per 100 PY), followed by upadacitinib (7.61 per 100 PY), baricitinib (6.51 per 100 PY), and ritlecitinib (4.40 per 100 PY), although rates for abrocitinib and ritlecitinib may be overestimated. Menstrual disturbances were the leading cause (n=10), mainly with baricitinib (n=5) and upadacitinib (n=3). Infectious events were also prominent, including recurrent upper respiratory infections (baricitinib) and herpes zoster (upadacitinib). All the events reported in paragraphs 3.3 (malignancies, serious infections, VTE and thrombophlebitis), except for NMSC lead to treatment discontinuation.

Table 10.

AE causing JAKi discontination

	TOTAL	UPA (n=150)			BARI (n=145)			ABRO (n=44)			RITLE (n=71)
Dose		Tot	30 mg	15 mg	Tot	4 mg	2 mg	Tot	200 mg	100 mg	50 mg
N subjects	36	18	9	9	10	10	0	6	3	3	2
N of side effects	40	19	9	10	12	12	0	7	3	4	2
Menstrual cycle alteration	10	3	1	2	5	5	0	2	0	2	0
Recurrent URI	4	1	1	0	3	3	0	0	0	0	0
H. zoster	4	4	4	0	0	0	0	0	0	0	0
Malignancies	4	1	0	1	2	2	0	1	0	1	0
Nausea	3	1	0	1	0	0	0	2	2	0	0
Acne	3	2	1	1	0	0	0	0	0	0	1
Neutropenia	2	0	0	0	1	1	0	1	0	1	0
Weight gain	2	0	0	0	1	1	0	1	1	0	0
VTE	1	1	1	0	0	0	0	0	0	0	0
Thromboflebitis	1	1	0	0	0	0	0	0	0	0	0
Serious infections	1	1	1	0	0	0	0	0	0	0	0
Punctate cheratitis	1	0	0	0	0	0	0	0	0	0	1
Hypercholesterolemia	1	1	0	1	0	0	0	0	0	0	0
Headache	1	1	0	1	0	0	0	0	0	0	0
Fatigue	1	1	0	1	0	0	0	0	0	0	0
Autonomous suspension by patient	1	1	0	1	0	0	0	0	0	0	0

Data are reported as number

ABRO abrocitinib; BARI barictinib; N number of events; RITLE ritlecitinib; UPA upadacitinib; VTE venous thromboembolism

Other causes of discontinuation were heterogeneous and infrequent, including nausea (n=3), acne (n=3), neutropenia (n=2), weight gain (n=2), hypercholesterolemia (n=1), and one case of punctate keratitis.

Discussion

Real-world evidence is essential to complement data derived from randomized clinical trials, as it reflects treatment performance in broader, less selected patient populations and under routine clinical conditions, where comorbidities, concomitant medications, and variable adherence may influence both efficacy and safety outcomes. In this large single-center cohort with more than 520 patient-years of cumulative exposure, JAKi demonstrated an overall favorable safety profile in patients with AD and AA, without the emergence of unexpected serious adverse events.

The relatively high number of adverse events observed should be interpreted in light of the inclusive reporting strategy adopted, which captured minor laboratory abnormalities and low-grade clinical events; thus, the overall incidence of adverse events apparently exceeded that commonly reported in randomized clinical trials, where only events meeting predefined severity thresholds are typically recorded. Despite this, serious adverse events and safety-related discontinuations were infrequent and occurred at lower rates than those reported in clinical trials, suggesting that the higher event burden reflects increased sensitivity in detection rather than reduced safety.

Hypercholesterolemia was the most frequent adverse event, particularly among patients treated with upadacitinib and baricitinib. According to the European Society of Cardiology (ESC) guidelines, hypercholesterolemia is primarily defined based on low-density lipoprotein cholesterol (LDL-C) levels, with recommended thresholds varying according to the individual’s cardiovascular risk profile [38]. However, in routine clinical practice, LDL-C values were not consistently available for all patients in our cohort. Therefore, for the purposes of this real-world analysis, total cholesterol levels were used as a surrogate marker to identify cases of hypercholesterolemia. Lipid elevations were generally mild and manageable, with only one discontinuation attributable to hypercholesterolemia.

Surprisingly, weight gain exceeding 5% of BW represented the second most frequently reported adverse event. Weight gain is increasingly recognized as a relevant real-world metabolic complication [33] and, in our cohort, occurred with an incidence comparable to that of infectious events, being mainly associated with upadacitinib and baricitinib treatment. Notably, this adverse event showed a clear sex-specific pattern, as it was more frequent and more pronounced among female patients, leading to treatment discontinuation in two cases.

In this context, menstrual alterations emerged as a particularly important and underrecognized sex-specific adverse event in women of reproductive age, especially among those treated with upadacitinib and baricitinib, as recently reported in the literature [39]. Importantly, menstrual cycle disturbances represented the leading cause of treatment discontinuation due to safety concerns in our cohort. These findings emphasize not only their clinical relevance but also the need for proactive assessment and counseling on reproductive and hormonal symptoms, which remain insufficiently addressed in current clinical practice and literature. Nevertheless, the observational design of the study and the absence of a control group preclude definitive conclusions. In addition, both atopic dermatitis and alopecia areata may be associated with psychological stress and systemic inflammation, factors that could independently contribute to menstrual irregularities. Therefore, the association observed in our cohort should be considered hypothesis-generating rather than evidence of a causal relationship. Future prospective studies specifically designed to evaluate reproductive health outcomes are needed, ideally including comparator groups treated with alternative systemic therapies for severe atopic dermatitis and alopecia areata, in order to determine whether the incidence of menstrual disturbances differs according to treatment exposure.

Infections were among the most frequently reported adverse events, consistent with clinical trial data. Most were mild to moderate, predominantly involving upper respiratory tract infections, while opportunistic or severe infections were uncommon. Herpes zoster occurred sporadically and rarely led to treatment discontinuation.

Acne was also a commonly observed dermatologic adverse event, particularly among patients treated with upadacitinib and baricitinib. Despite its relatively high frequency, acne was typically mild to moderate in severity and only one case led to treatment interruption. These findings suggest that acne, although clinically relevant from a quality-of-life perspective, has a limited impact on treatment persistence and can be effectively managed within routine dermatologic care.

Regarding events of special interest, no MACEs were observed and only one thromboembolic event was reported. These findings should be interpreted cautiously because the study population was relatively young and had a low burden of baseline cardiovascular risk factors. This characteristic likely reflects the progressive implementation of the 2023 PRAC recommendations in routine dermatologic practice, leading to more stringent selection of patients eligible for JAK inhibitor therapy. Consequently, our findings may not be generalizable to populations at higher cardiovascular risk. Four cases of malignancy and three cases of NMSC were observed, with incidence rates comparable to those reported in the general population. Overall, in light of the study limitations, adverse events of special interest occurred at rates comparable to or lower than those observed in untreated patients with atopic dermatitis [32, 40–43].

Secondary inefficacy was the leading cause of treatment discontinuation across all JAKi, whereas adverse-event-related discontinuations were less frequent. Among these, menstrual disturbances and infections were the main drivers of treatment interruption. Although discontinuation rates appeared higher for abrocitinib, this finding should be interpreted cautiously due to shorter exposure. Overall, these data suggest that treatment persistence in real-world practice is primarily driven by sustained effectiveness rather than tolerability.

This study has several limitations. Differences in baseline demographic and clinical characteristics across treatment groups, including sex distribution, may have influenced the observed adverse event rates and limit the interpretation of treatment-specific findings. The retrospective design may have introduced information and surveillance bias. Laboratory assessments were not performed according to a standardized schedule, and monitoring practices evolved during the study period, particularly for lipid parameters following increasing awareness of JAK inhibitor-associated metabolic effects. Consequently, some laboratory abnormalities may have been underdetected in earlier years, whereas enhanced monitoring may have increased event detection later in the study. Missing laboratory data and variability in follow-up intensity may have also influenced the reported incidence of laboratory adverse events. Additionally, as previously stated, the shorter exposure for ritlecitinib and abrocitinib may have led to overestimation of certain outcomes, particularly incidence rates in lipid alterations and treatment discontinuations. Information on concomitant medications was not systematically available for all patients and was therefore not included in the analysis. As concomitant treatments may influence both laboratory parameters and adverse event occurrence, residual confounding cannot be excluded. These limitations should be considered when interpreting the results and their generalizability.

Conclusions

In conclusion, this real-world study provides additional safety data on JAK inhibitors in patients with AD and AA treated in routine clinical practice. Treatment discontinuation was predominantly driven by secondary inefficacy rather than adverse events. While no major cardiovascular safety signals emerged during follow-up, the relatively young age and low cardiovascular risk profile of the study population warrant cautious interpretation of these findings. Metabolic alterations, including lipid changes and weight gain, were frequently observed and support the need for regular monitoring. Menstrual disturbances emerged as a potentially relevant sex-specific finding, although causality cannot be established given the observational nature of this study and the absence of a control group. These results underscore the need for standardized monitoring and proactive management of sex-specific side effects in routine clinical practice.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization: Paola Facheris, Alessandra Narcisi, Antonio Costanzo; methodology: Paola Facheris, Alessandra Narcisi, Antonio Costanzo; software: Paola Facheris; formal analysis: Paola Facheris; investigation: Paola Facheris, Luigi Gargiulo, Luciano Ibba, Francesco D’Oria, Costanza Falcidia, Guilio Foggi, Sara Di Giulio, Carlo A. Vignoli, Alessandra Narcisi, Mario Valenti, Antonio Costanzo; resources: Paola Facheris, Luigi Gargiulo, Luciano Ibba, Francesco D’Oria, Costanza Falcidia, Guilio Foggi, Sara Di Giulio, Carlo A. Vignoli, Alessandra Narcisi, Mario Valenti, Antonio Costanzo; data curation: Paoal Facheris, Francesco D’Oria, Costanza Falcidia, Giulio Foggi, Sara Di Giulio, Alessandra Narcisi; writing-original draft: Paola Facheris, writing-review and editing: Paola Facheris, Luigi Gargiulo, Luciano Ibba, Mario Valenti, Francesco D’Oria, Costanza Falcidia, Giulio Foggi, Sara Di Giulio, Carlo A. Vignoli Alessandra Narcisi, Antonio Costanzo; visualization: Paola Facheris; supervision: Paola Facheris, Alessandra Narcisi, Antonio Costanzo; project administration: Paola Facheris, Alessandra Narcisi, Antonio Costanzo.

Funding

No funding or sponsorship was received for this study or publication of this article. The Rapid Service Fee was funded by the authors’ Institution.

Data Availability

The datasets generated during and/or analyzed during the current study are not publicly available due to ethical and privacy reasons.

Declarations

Conflict of Interest

Paola Facheris has served as a consultant for Eli Lilly and as a speaker for UCB, Abbvie, and Pfizer. Luigi Gargiulo has been a consultant and/or speaker and has participated in advisory boards for AbbVie, Almirall, Amgen, BMS, Eli Lilly, Incyte, Galderma, Pierre Fabre, Johnson and Johnson, Novartis, Pfizer, Sanofi, and UCB Pharma. Luciano Ibba served as consultants for Almirall. Mario Valenti has served as consultant/speaker for Novartis, Almirall, Eli-Lilly, Abbvie, Johnson&Johnson, UCB, Boehringer-Ingelheim, Difa Cooper, Sun-Pharma, Leo Pharma, and Bristol Myers Squibb. Mario Valenti is an Editorial Board member of Dermatology and Therapy. Mario Valenti was not involved in the selection of peer reviewers for the manuscript nor any of the subsequent editorial decisions. Francesco D’Oria, Giulio Foggi, Costanza Falcidia and Carlo A. Vignoli have no conflict of interest to disclose. Alessandra Narcisi has served on advisory boards, received honoraria for lectures and research grants from Almirall, Abbvie, Leo Pharma, Celgene, Eli Lilly, Janssen, Novartis, Sanofi Genzyme, Amgen and Boehringer Ingelheim. Antonio Costanzo has served as an advisory board member, consultant and has received fees and speaker’s honoraria or has participated in clinical trials for Abbvie, Almirall, Biogen, LEO Pharma, Lilly, Janssen, Novartis, Pfizer, Sanofi Genzyme and UCB.

Ethical Approval

This was a non-interventional observational study. According to local regulations, formal ethics committee approval and written informed consent were not required. The study was notified to the Italian Medicines Agency (AIFA, Study ID 2805, study code JAKiSafe), and the competent Ethics Committee acknowledged the study. All procedures were conducted in accordance with the Declaration of Helsinki and applicable national regulations.