Appropriate Injection Intervals of Dupilumab in Patients With Adult Atopic Dermatitis: A Step Toward Developing Guidelines for Daily Practice

Abstract

Background

There is a growing demand for extending dosing intervals of dupilumab injections in patients with atopic dermatitis (AD) due to treatment burden and side effects. However, studies on successful dose reduction in real-world settings are lacking.

Objective

To assess the efficacy of a patient-centered dupilumab tapering regimen and to propose guidelines for target patients, appropriate intervals, and timing for tapering.

Methods

This single-center retrospective study included moderate to severe adult AD patients who underwent at least 16 weeks of dupilumab treatment. Interval prolongation was considered in controlled patients assessed by Eczema Area and Severity Index (EASI) score and serum inflammatory markers after at least 40 weeks of treatment with a standard regimen. Logistic regression model with generalized estimating equations was used to compare repetitive measurements over time between the two groups.

Results

A total of 52 patients were included with 11 patients extending intervals to 3–4 weeks without flare-ups. The mean duration of dupilumab treatment before tapering was 53.27 weeks. The tapering group exhibited significantly lower body mass index. All patients of the tapering group showed EASI scores under 4 and immunoglobulin E (IgE) levels under 1,000 IU/mL at week 40. EASI scores and IgE levels remained consistently low after dose reduction, with a mean follow-up time of 14.36 months.

Conclusion

Patients with extended dosing intervals demonstrated sustained effectiveness. Dose tapering might be a valuable option for non-obese patients with positive clinical response characterized by an EASI score under 4 and IgE levels under 1,000 at week 40.

INTRODUCTION

Dupilumab is a human monoclonal antibody targeting interleukin (IL)-4 receptor that can inhibit IL-4 and IL-13 signaling. Due to its long-term efficacy and safety on atopic dermatitis (AD) patients, it has been approved and used widely with a treatment regimen of one 300 mg injection every two weeks for adult patients after a starting dose of 600 mg¹.

Recently, a flexible regimen of dupilumab injection with a longer dosing interval has been suggested for patients with a good clinical response (GCR) or adverse events²,³,⁴. High costs and strict regulations for insurance coverage pose great economical demand on patients. Frequent hospital visits are also burdensome for patients managing chronic conditions. Moreover, various side effects including injection site reactions, allergic conjunctivitis, facial redness, and eosinophilia are quite prevalent. In some cases, hypereosinophilia (more than 1,500/mm³) might induce severe clinical complications such as eosinophilic pneumonia, vasculitis, and massive stroke⁵. Previous studies have reported that dose tapering can enhance patient's adherence, reduce costs, and improve dupilumab-induced conjunctivitis²,³,⁴.

Such disease activity-guided dose reduction of biologics is widely implemented in clinical practice for well-controlled rheumatoid arthritis and psoriasis⁶,⁷. However, real-world data on the efficacy of dupilumab tapering in AD patients are limited²,³,⁴. Moreover, in Korea, the stringent criteria for dupilumab insurance poses challenges in extending the interval, as additional insurance coverage is not granted if clinical conditions worsen after dose tapering⁸.

Our study aimed to evaluate an efficacy of a patient-centered dupilumab tapering regimen in controlling adult AD in daily practice and to identify patient characteristics and biomarkers for whom it might be advisable.

MATERIALS AND METHODS

Patients and inclusion criteria

We retrospectively collected data from adult AD patients aged over 18 years who visited our dermatology center between January 2020 and December 2023 and received a minimum of 16 weeks of dupilumab treatment. Patients who irregularly received dupilumab injections at intervals exceeding four weeks and those with a history of switching to a Janus kinase inhibitor during treatment were excluded from this study. The research protocol was approved by the Institutional Review Board (SMC 2023-12-125). Patients in this manuscript have given written informed consent to the publication of their case details.

Study design

Initially, all patients had received a subcutaneous loading dose of dupilumab at 600 mg, followed by subsequent administrations of 300 mg of dupilumab every two weeks until week 40. Interval prolongation was considered in patients with stable low disease activity, which was determined by a dermatology specialist in consideration with Eczema Area and Severity Index (EASI) score, subjective symptoms, and serum inflammatory markers including immunoglobulin E (IgE), eosinophil count, lactate dehydrogenase (LDH), eosinophil cationic protein (ECP), and uric acid. The decision of interval prolongation was reached collaboratively through shared decision-making between the patient and the physician. All patients received concurrent topical steroid therapy to maintain disease control. There was no concurrent administration of conventional treatments with cyclosporine or methotrexate throughout the treatment period with dupilumab.

Outcome measures of EASI score and biomarker results were analyzed at four distinct time points: baseline (T0), 16 weeks (T1), 40 weeks (T2), and 68 weeks (T3) after treatment initiation. These analyzed time points were selected in accordance with a regulation of National Health Insurance in Korea that permits the approval of an additional 6 months of treatment if EASI score shows 75% or greater reduction at week 16. It approves subsequent continuous administration if the initial response is sustained at regular 6-month evaluations. Due to a high cost of the medication, most patients initiated dupilumab treatment when satisfying insurance regulations and typically visited the clinic for examinations at these specified time points.

Statistical analysis

Two-sample t-test for continuous variables and χ² test for categorical variables were used for statistical analysis and comparison of demographics and biomarker results at each timepoint. Logistic regression estimation employing generalized estimating equations (GEE) was utilized to compare mean estimated values of EASI score, IgE, and eosinophil count at each timepoint. This approach was chosen to address main and interaction effects between time and tapering, considering the correlation structure which could occur due to multiple measurements over time within patients. Findings are presented as means along with 95% confidence intervals (CIs). Data after tapering are presented through a line graph of mean estimated values via GEE for EASI scores. A p-value of less than 0.05 was considered statistically significant. All data were analyzed using R version 4.3.1 (R Core Team 2023, Vienna, Austria).

RESULTS

A total of 52 adult AD patients were included. There were 41 individuals (group 1) persisting on the standard interval and 11 patients (group 2) undergoing interval prolongation of 3 to 4 weeks. In group 2, the mean duration of treatment preceding the tapering phase was 53.27 weeks (range, 42–64 weeks), which corresponded to the timepoint between T2 and T3. The mean EASI score at the tapering baseline was 2.5 (range, 0.6–4.3) (Table 1). Maintenance of controlled disease was the main reason for dose tapering in group 2, with only one attributing to an elevated eosinophil count.

Table 1. Patient characteristics.

Characteristic		Overall (n=52)	Group 1* (n=41)	Group 2† (n=11)	p-value	Missing value (%)
Sex (male)		36 (69.2)	28 (68.3)	8 (72.7)	1	0
Age (yr)		31.65±13.86	31.71±14.70	31.45±10.73	0.958	0
AD onset					0.295	0
	Childhood	26 (50.0)	22 (53.7)	4 (36.4)
	Adolescence	12 (23.1)	10 (24.4)	2 (18.2)
	Adulthood	14 (26.9)	9 (22.0)	5 (45.5)
Allergic comorbidity
	Allergic asthma	9 (17.3)	6 (14.6)	3 (27.3)	0.593	0
	Allergic rhinitis	13 (25.0)	9 (22.0)	4 (36.4)	0.556	0
	Food allergy	2 (3.8)	2 (4.9)	0 (0.0)	1	0
	Allergic conjunctivitis	5 (9.6)	4 (9.8)	1 (9.1)	1	0
Weight		69.06±15.45	71.05±15.30	61.99±14.48	0.086	3.8
Height		169.43±7.93	169.52±7.50	169.12±9.72	0.885	3.8
BMI		23.91±4.59	24.60±4.53	21.45±4.08	0.043	3.8
EASI score
	Baseline (T0)	28.23±7.61	28.28±8.29	28.04±4.54	0.926	0
	16 wk (T1)	6.33±5.27	6.90±5.73	4.23±2.04	0.137	0
	16 wk + 6 mo (T2)	3.78±2.78	4.22±2.97	2.57±1.76	0.091	21.2
	16 wk + 12 mo (T3)	2.67±1.95	2.71±2.16	2.60±1.63	0.885	44.2
Duration of dupilumab at time of dose spacing (wk)		-	-	53.27 (42–64)	-	-
EASI score at the tapering baseline		-	-	2.5 (0.6–4.3)	-	-
Reason for interval switch
	Maintenance of controlled disease	-	-	10 (90.9)	-	-
	Side effects‡	-	-	1 (9.1)	-	-

Values are presented as mean ± standard deviation or mean (interquartile range), or number (%).

AD: atopic dermatitis, BMI: body mass index, EASI: Eczema Area and Severity Index.

^*Patients with standard interval; ^†Patients with interval extension to 3–4 weeks; ^‡Elevated eosinophil count.

Regarding demographics, gender proportion, age at initiation of dupilumab, frequency of allergic comorbidities, and baseline EASI score showed no statistically significant differences between the two groups (Table 1). Notably, body mass index (BMI) showed a significant difference between the two groups (p<0.05), with a mean value of 24.60 for group 1, classified as overweight according to Korean standards, while group 2 had a BMI mean value of 21.45, indicating normal weight (Table 1). The analysis of biomarker characteristics at each timepoint between the two groups showed significantly lower IgE levels in group 2 at T1 (week 16) and T3 (week 68) (Table 2). LDH, ECP, and uric acid were excluded from further analyses due to an absence of significant difference between the two groups.

Table 2. Biomarker results per study group at each timepoint.

Timepoint		Group 1* (n=41)	Group 2† (n=11)	p-value
T0 (baseline)
	IgE	3,477.67±1,921.10	2,037.90±1,991.55	0.062
	Eosinophil count	660.46±551.63	887.75±509.95	0.313
	ECP	55.20±61.98	38.45±39.62	0.508
	LDH	236.33±24.54	N/A	N/A
	Uric acid	6.35±1.79	5.82±1.65	0.457
T1 (16 wk)
	IgE	2,398.06±2,046.91	836.94±768.06	0.039
	Eosinophil count	424.36±399.72	437.78±263.83	0.93
	ECP	42.01±24.47	31.29±29.11	0.455
	LDH	235.71±41.56	187.50±21.92	0.17
	Uric acid	6.18±1.85	5.40±1.45	0.408
T2 (40 wk)
	IgE	1,363.79±1,829.65	622.11±1,000.78	0.248
	Eosinophil count	434.00±238.92	363.90±188.33	0.432
	ECP	45.33±35.96	32.64±29.05	0.47
	LDH	203.50±36.19	209.67±71.07	0.885
	Uric acid	5.53±1.27	4.54±0.98	0.131
T3 (68 wk)
	IgE	2,004.12±1,848.27	280.24±334.66	0.031
	Eosinophil count	187.44±90.30	318.57±194.20	0.093
	ECP	18.20±16.12	24.90±16.51	0.647
	LDH	180.00±24.04	177.00±1.41	0.876
	Uric acid	5.68±1.72	4.93±1.30	0.544

All values are presented as mean ± standard deviation.

IgE: immunoglobulin E, ECP: eosinophil cationic protein, LDH: lactate dehydrogenase, N/A: not available.

^*Patients with standard interval; ^†Patients with interval extension to 3–4 weeks.

In the GEE analysis of continuous variables, a statistically significant lower predicted value of the EASI score at T2 (week 40) was observed for group 2 compared to group 1, with the highest estimated mean of 3.06 (95% CI, 1.36–3.06). At T3, no significant difference in EASI score was observed between the two groups, showing maintained efficacy after tapering (Table 3 and Fig. 1A). For IgE, group 1 consistently demonstrated higher mean values than group 2 across all time points. Particularly noteworthy was the observation at T2 immediately preceding the tapering phase, where the highest estimated mean IgE value in group 2 reached 671.83, consistently remaining below the 1,000 cutoff value (Table 3 and Fig. 1B). The eosinophil count did not show a significant difference between the two groups. All predicted values remained below the normal range after T1 (Table 3 and Fig. 1C). Following tapering in group 2, EASI scores and IgE levels were consistently sustained below 4 and 1,000, respectively, while eosinophil count showed variable results, with some values spiking to the normal limit (Table 3). The mean follow-up time after tapering was 14.36 months (range, 7–24 months).

Table 3. Comparison of clinical and laboratory results at each timepoint using generalized estimating equations analysis.

Result		Group 1	Group 2	Difference
				p-value	Ratio (95% CI)
EASI score
	T0	26.58 (23.10–30.27)	27.66 (25.28–30.57)	0.588	0.96 (0.81–1.13)
	T1	5.58 (4.53–6.89)	3.56 (2.41–5.26)	0.045	1.57 (1.01–2.44)
	T2	3.46 (2.64–4.53)	2.05 (1.36–3.06)	0.038	1.68 (1.03–2.75)
	T3	2.16 (1.45–3.25)	1.79 (0.92–3.42)	0.615	1.22 (0.57–2.62)
IgE
	T0	2,591.52 (1,510.20–4,491.76)	1,603.59 (445.86–5,825.50)	0.5	1.62 (0.40–6.54)
	T1	1,211.97 (317.35–4,628.55)	607.89 (149.90–2,489.91)	0.492	1.98 (0.28–13.88)
	T2	720.54 (139.77–3,751.83)	237.46 (83.93–671.83)	0.266	3.03 (0.43–21.37)
	T3	518.01 (172.43–1,571.84)	146.94 (56.26–383.75)	0.095	3.53 (0.80–15.53)
Eosinophil count
	T0	487.85 (333.62–706.27)	742.48 (445.86–1,248.88)	0.189	0.65 (0.35–1.23)
	T1	244.69 (152.93–391.51)	347.23 (214.86–555.57)	0.313	0.71 (0.36–1.38)
	T2	376.15 (295.89–478.19)	314.19 (219.20–450.34)	0.416	1.20 (0.78–1.85)
	T3	183.09 (130.32–254.68)	273.14 (159.17–473.43)	0.211	0.67 (0.35–1.26)

Values are presented as mean (interquartile range).

Mean estimated values are demonstrated, accompanied by 95% CIs. The logistic regression estimation with generalized estimating equations was used for analysis to account for the interaction effect between time and tapering.

CI: confidence interval, EASI: Eczema Area and Severity Index, IgE: immunoglobulin E.

Fig. 1. Comparison of mean estimated values of EASI score (A), IgE (B), and eosinophil count (C) between group 1 (black, standard interval) and group 2 (red, interval extension to 3–4 weeks) at each timepoint. The logistic regression estimation using generalized estimating equations was used for analysis to account for the interaction effect between time and tapering. Dot lines represent normal range, IgE under 200, and eosinophil under 500.

EASI: Eczema Area and Severity Index, IgE: immunoglobulin E.

DISCUSSION

Our analyses showed that adult AD patients who were treated with dupilumab in an extended interval sustained effectiveness, up to a maximum of 24 months during the follow-up period.

In our study, patients included in the tapering group showed statistically significantly lower BMI than the non-tapering group. This finding aligns with a prior study by Olydam et al.⁹ showing that patients with an extended interval exhibit lower BMI than those in the standard interval, although their differences are not statistically significant. One plausible explanation is the impact of obesity on the pharmacokinetics of IgG monoclonal antibody. Animal model studies have indicated a lower area under the curve for human IgG and a higher clearance rate in obese groups, which may induce subtherapeutic drug concentrations¹⁰. Similarly, research on atopic patients has revealed a significant influence of body weight on linear clearance and central volume of dupilumab¹¹. However, despite these findings, dose adjustment was not recommended, given the wide therapeutic index of dupilumab and the minimal impact of weight variability on inter-individual variability of the efficacy¹². Another possibility is that obese patients might not exhibit a quick, substantial reduction in inflammation markers, rendering them less eligible for inclusion in the tapering group. While exact molecular mechanisms remain unclear, proinflammatory cytokines and hormones secreted by excess adipocytes may aggravate atopic inflammation¹³. Recent studies have revealed that obesity can increase JAK-STAT signaling and mTOR1 expression which can decrease filaggrin expression¹⁴. Some studies have also shown that high BMI is strongly associated with increased transepidermal water loss and dysregulation of skin pH, which may worsen atopic inflammation¹⁵. These enhanced inflammation status and immune dysregulation might require higher doses of dupilumab for obese patients.

Due to the high costs and prevalent side effects of newly emerging biologics for AD, there is a growing consensus among both patients and physicians in South Korea regarding the necessity to extend dosing intervals. In case of tralokinumab, a new anti-IL-13 monoclonal antibody for AD, the regimen of 300 mg every 4 weeks is recommended for patients below 100 kg who have achieved clear or almost clear skin after 16 weeks of 2-week interval therapy¹⁶. Although scarce, a few recent daily practice studies have proposed that dose reduction of dupilumab might be successfully applied to patients with controlled AD⁴,⁵,⁶,⁷. When extending the injection interval, physicians may consider three factors: the condition of the target patient, the maximum interval for maintaining the efficacy, and the appropriate timing for tapering.

Identifying suitable candidates for dose tapering is pivotal but challenging. Jendoubi et al.¹⁷ have found that tapering is more effective in older patients, those with lower topical corticosteroid usage, and those in whom this strategy is initiated for clinical good response rather than in response to adverse events. No other studies have identified predictive values for successful tapering. Based on the results, we are introducing a new proposal, suggesting BMI under 23, EASI score below 4, and IgE level under 1,000 at week 40 as specific criteria for initiating dose tapering, although further validation is required through prospective studies in the future.

In consideration of an appropriate extended interval of dupilumab, several regimens have been attempted in real-world studies. In our study, we prolonged the interval to 3 to 4 weeks without any observed flare-ups or notable exacerbation of the disease. In line with our results, Patruno et al.³ showed that patients with interval extended to 3 to 4 weeks due to GCR maintained mean disease severity scores up to week 32 after tapering. Spekhorst et al.⁴ adopted a gradual tapering approach, starting with a dosage interval of every 3 weeks and progressing sequentially to Q4W, Q6W, and Q8W, contingent on a state of controlled disease (EASI score ≤7). Throughout the tapering phase, both disease severity scores (EASI, Numeric Rating Scale) and serum severity markers (pulmonary and activation-regulated chemokine [PARC/CCL18], thymus and activation-regulated chemokine [TARC/CCL17]) consistently maintained low levels⁴. However, in accordance with the recommendation of Dekkers et al.¹⁸, it appears advisable to maintain an extended interval of less than 40 days as there is a risk of detectable IL-4Ra reemergence and a subsequent increase in EASI scores beyond this 40-day threshold.

Another crucial issue is determining the optimal timing for dose spacing. Most clinical trials for dupilumab followed a treatment protocol of a 16-week induction phase, but only about 50% of patients exhibited sufficient efficacy by week 16¹²,¹⁹. Additionally, the SOLO-continue clinical trial revealed diminished efficacy and no safety advantage when extending intervals to 4 or 8 weeks at week 16 in a good-response patients²⁰. This suggests the necessity for a longer induction phase and an establishment of a distinct definition of “controlled disease” in a real-world practice. Immunologically, Bangert et al.²¹ revealed that specific immune cells such as mature dendritic cells and memory T cell populations remained in clinically healed skin of AD patients until a year after dupilumab treatment, suggesting sufficient length of induction phase before extending the interval. Previous daily practice studies prolonged interval after at least 1 year of treatment with the standard dose of dupilumab, regarding the maintenance of an EASI ≤7 for over 6 months as an indicator of disease stability²,³,⁴. In our study, we succeeded extending intervals between week 42 and week 64, which was earlier than the previous studies, and determined the control of disease through stabilization of IgE levels in addition to the EASI score. Our research was focused on serum IgE level due to a previously known correlation with the severity of AD, along with convenience and cost-effectiveness²². However, some studies noted that implications of IgE levels differed between intrinsic and extrinsic types of AD, often failing to entirely reflect clinical activity²³. Further studies on serum biomarkers suitable for subjective evaluation of disease stability and treatment prediction with more substantial real-world data are needed for establishing a standardized tapering timing.

This study has some limitations stemming from its retrospective and observational design. Verifying a direct causal effect between patient criteria we proposed for dose tapering and the efficacy of tapering is challenging. Prospective studies with well-defined inclusion criteria for the tapering group would be helpful for validating the data from our analysis. Another limitation was the small size of the tapering group, which was influenced by the complex reimbursement criteria for dupilumab in Korea. Only patients self-funding their dupilumab injections were eligible to attempt interval spacing. A multicenter study might be needed to address these limitations. Furthermore, we anticipate that a relaxation of insurance regulations, driven by the widespread acceptance of the efficacy of lengthening the interval, may lead to the collection of more extensive data.

In conclusion, our study demonstrated that extending dupilumab dose intervals to 3 to 4weeks after at least 42 weeks of 2-week interval treatment, based on a shared-decision making process between patients and physicians, seemed to be effective in daily practice. Additionally, dose tapering might be a valuable option for non-obese patients showing a positive clinical response characterized by an EASI score under 4 and immunoglobulin level under 1,000 at week 40. Ongoing daily practice and prospective clinical studies are essential for establishing guidelines on dupilumab interval prolongation, thereby advancing personalized treatment for severe AD.