How long should dupilumab maintenance continue?

30-second key takeaways

Why this study matters

Since the 2017 FDA approval, dupilumab has rapidly become the moderate-to-severe AD standard of care. In real-world practice, however, clinicians face a question the pivotal trials never answered: "After reaching EASI-75 and clinical clearance, should I continue? At what interval? When can I stop?"

In practice, patients discontinue (or are forced to taper) for reasons such as relocation, NHI continuation thresholds, out-of-pocket cost, satisfaction with current response, and pregnancy planning. These unplanned discontinuations leave the real-world maintenance phase under-characterized. Qiu et al's May 2026 JAAD single-center retrospective is one of few studies that directly compares continuous q2w, interval extension, and stop strategies over 52 weeks in real patients.

Study design

Main efficacy results

At T1 (week 16), all three groups improved significantly — SCORAD, EASI, and pruritus NRS each fell with p<0.001. EASI-50 was 72.15%, EASI-75 44.30%, consistent with multiple prior real-world reports (pivotal RCT EASI-75 ~38-44%; real-world is often slightly higher). No significant difference between Han Chinese and minority subgroups.

Relapse analysis (the most clinically actionable finding)

The most clinically actionable comparison in this study is relapse-free duration across the three groups:

The paradox: why does tapering relapse earlier than stopping?

The authors propose two possible explanations, both requiring confirmation in larger RCTs:

1. Pharmacokinetic hypothesis: Tapering creates sub-therapeutic plasma levels. Low, fluctuating IL-4/IL-13 receptor blockade may be insufficient to fully suppress underlying Th2 inflammation, leaving the disease in an unstable equilibrium that's actually less predictable than a clean stop — yielding earlier relapse.
2. Selection bias: With shared decision-making (non-randomized), clinicians may have steered patients with less-than-ideal response or more relapse-prone disease toward the tapering arm. The tapering cohort then carries higher baseline relapse risk independent of the dosing strategy itself.

Clinical implication: Until a larger RCT replicates this, the intuition that "tapering is the gentle compromise" deserves caution. Patients with a good dupilumab response who want to extend dosing intervals should be told: gradual extension is not clearly safer than a clean stop with re-treatment on relapse.

Relapse risk factors

Combining multivariate and univariate analyses, the study identified the following relapse predictors:

Safety and the first-reported optic-nerve-atrophy case

Across 52 weeks of follow-up, 7 adverse events were recorded — all self-limiting or resolving with treatment. No patient discontinued for adverse events.

Biomarker changes (IgE and EOS)

The study measured serum total IgE and peripheral eosinophil (EOS) counts at T0, T1, and T2. Findings align with prior reports:

• Both long-term and tapering groups showed significant IgE and EOS reductions at weeks 16 and 52, reflecting type-2 inflammation suppression.
• Higher baseline IgE correlated with relapse risk — consistent with the "greater type-2 burden, greater rebound potential" principle.
• Effects may vary by comorbidity (asthma), AD subtype (intrinsic vs extrinsic), and treatment regimen (monotherapy vs combination); larger samples are needed.

Take-home messages for clinicians

1. Do not use tapering as a cost-saving strategy after target achievement. This study shows extended-interval tapering relapsed earlier than abrupt stop (48 vs 66 days).
2. If true relapse delay is the goal → continue q2w. The long-term q2w group's mean relapse-free duration was 159 days, far above the others; some patients exceeded 300 days.
3. High-relapse-risk subgroups (high baseline IgE, disease duration > 5 yrs, non-classic phenotype) → lean toward continued therapy rather than cessation.
4. If discontinuation is attempted, abrupt stop appears more stable than gradual tapering. The short-term stop group's relapse window (66 days) was longer than tapering's.
5. Use IgE / EOS trends as relapse-risk markers. High baseline IgE correlated with relapse — useful for risk stratification.
6. The optic-nerve-atrophy signal is a flag, not a verdict. One case, one year post-discontinuation, with cerebral ischemia comorbidity — causality unconfirmed. Routine ocular symptom inquiry suffices; no routine neuroimaging screening warranted.

Taiwan NHI reimbursement criteria (Dupixent®)

Prescribing specialty and indication

• ≥ 18 years: dermatology or rheumatology/immunology specialist only
• 12–18 years: dermatology specialist or pediatric allergy/immunology/rheumatology subspecialist
• Indication: moderate-to-severe chronic systemic AD that failed both phototherapy and systemic therapy, or has medical contraindications to both

Objective definition of moderate-to-severe AD

• Lesions persisting ≥ 6 months
• EASI ≥ 16
• Body surface ≥ 30% inflamed (only inflamed eczematous areas count; dryness, scaling, excoriation alone don't qualify)
• IGA 3–4 (moderate to severe)

Required prior therapies (pre-authorization evidence)

Applicants must complete a "full prior course" totaling 6 months (may combine records from other hospitals):

1. Phototherapy (PUVA or nb-UVB), at least twice weekly, 12-week course, with detailed dose log
2. At least two of three traditional immunosuppressants, each used ≥ 12 weeks without response (adult full doses: methotrexate 15 mg/wk / azathioprine 2 mg/kg/d / cyclosporin 5 mg/kg/d)
3. Medical contraindications (photosensitivity, albinism, lupus, skin cancer history, etc.) waive phototherapy

Dosing

Continuation and pause rules

• Initial application: 6-month course, pre-authorization required
• Week-16 review: failure to meet EASI 50 ends continuation
• Continuation: re-authorization every 6 months with before/after photos
• Pause continuation: biologics (dupilumab) — after 1 year, if EASI ≤ 16, treatment must pause. JAK inhibitors (upadacitinib / abrocitinib) — per the 2025/6/1 revision, pause now triggered after 2 years if EASI ≤ 16. May re-apply on relapse (≥ 50% recurrence or EASI ≥ 16)
• Upa and Abro dose restrictions: Upadacitinib — 15 mg/day only if EASI 16–20; 30 mg/day reserved for ≥ 18 yr with EASI ≥ 20. Abrocitinib — ≥ 12 yr may use either 100 mg/day or 200 mg/day; week-16 EASI 50 required to continue. Upa and Abro can only be used one at a time; switching allowed only for intolerable AE
• Reapplying after > 3 months off-therapy: treated as a new case
• No combination: dupilumab, upadacitinib, abrocitinib — only one at a time; upadacitinib / abrocitinib cannot combine with cyclosporin

Contraindications and stop conditions

• Pregnancy or lactation
• Parasitic (helminth) infection
• Failure to achieve EASI 50 at 6 months → discontinue
• New malignancy → discontinue

Brief patient-facing summary

If you're being treated with Dupixent® for moderate-to-severe AD, the practical takeaways:

• People who continued for the full year experienced slower relapse; some went 10+ months without relapse.
• Self-extending the interval (e.g. q4w, q6w, q8w) is not necessarily safer than stopping outright — extended-interval patients in this study relapsed earlier than those who stopped completely.
• Discuss any pause with your dermatologist first. Taiwan NHI has its own pause rules (e.g. EASI ≤ 16 pauses continuation).
• Any vision change (blurred vision, visual field defect) — during or after treatment — see an ophthalmologist promptly.
• This is a 79-patient retrospective from Xinjiang, China — not directly applicable to Taiwanese patients. Taiwan's NHI rules and population differ; discuss your individual case with your physician.

Conclusion

Qiu et al's 2026 JAAD real-world cohort of 79 moderate-to-severe AD patients delivers three clinically meaningful signals: (1) continuous q2w dupilumab through year 1 significantly delays relapse; (2) interval extension (tapering) paradoxically relapses earlier than abrupt stop, possibly via sub-therapeutic concentration or selection bias; (3) high baseline IgE, long disease duration, and non-classic AD phenotype mark relapse-risk subgroups. If the clinical goal is true relapse delay, continuous q2w is supported over interval reduction. As a single-center retrospective with n=79, results don't directly transfer to Taiwan but provide a real-world signal worth confirming in larger RCTs.

Related reading on this site: "Atopic dermatitis complete guide," "Systemic therapy for atopic dermatitis," and "Biologics and small-molecule targeted therapy overview."

References

1. Qiu Y, Ding Y, Hu W, Wang H, Lei Z, Kang X. Long-term dupilumab efficacy and safety in moderate-to-severe atopic dermatitis: A real-world single-center retrospective study. J Am Acad Dermatol. 2026;94(5):1454–1460. DOI: 10.1016/j.jaad.2025.12.088
2. Spekhorst LS, de Graaf M, Zuithoff N, et al. Dupilumab drug survival and associated predictors in patients with moderate to severe atopic dermatitis: long-term results from the daily practice BioDay registry. JAMA Dermatol. 2022;158(9):1048–1056. DOI
3. Paller AS, Simpson EL, Siegfried EC, et al. Dupilumab in children aged 6 months to younger than 6 years with uncontrolled atopic dermatitis: a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2022;400:908–919. DOI
4. De Bruin-Weller M, Biedermann T, Bissonnette R, et al. Treat-to-target in atopic dermatitis: an international consensus on a set of core decision points for systemic therapies. Acta Derm Venereol. 2021;101(2):adv00402. DOI
5. Barei F, Calzari P, Pezzolo E, et al. Effectiveness of tralokinumab across atopic dermatitis phenotypes. J Clin Med. 2025;14(6):2077. DOI
6. Yeung J, Gooderham MJ, Hong HC, et al. Treat-to-target in the management of moderate-to-severe atopic dermatitis in adults: A Canadian perspective. J Am Acad Dermatol. 2023;89(2):372–375. DOI
7. Olesen CM, Holm JG, Norreslet LB, et al. Treatment of atopic dermatitis with dupilumab: experience from a tertiary referral centre. J Eur Acad Dermatol Venereol. 2019;33(8):1562–1568. DOI
8. Faiz S, Giovannelli J, Podevin C, et al. Effectiveness and safety of dupilumab for the treatment of atopic dermatitis in a real-life French multicenter adult cohort. J Am Acad Dermatol. 2019;81(1):143–151. DOI
9. Saeki H, Fujita H, Suzuki K, Arima K. Safety and effectiveness of dupilumab in the real-world treatment of atopic dermatitis in Japan: 1-year interim analysis from a post-marketing surveillance. J Cutan Immunol Allergy. 2023;3:1–10. DOI
10. Francesca B, Martina Z, Simona T, et al. Assessment of patient-reported outcomes at 48 months of treatment with dupilumab for severe atopic dermatitis: a single-center real-life experience with 126 patients. Pharmaceuticals. 2024;17:117. DOI
11. Hideyuki N, Masahiro K, Yoshiki O, et al. Real-world experience of 3-year treatment with dupilumab: significant decrease in circulating neutrophils and eosinophils in Japanese patients with atopic dermatitis. Exp Dermatol. 2024;33(11):e70010. DOI
12. National Health Insurance Administration, Ministry of Health and Welfare, Taiwan. NHI drug reimbursement regulation 13.17.1 (dupilumab / upadacitinib / abrocitinib for atopic dermatitis). Effective 23 April 2025. https://www.nhi.gov.tw/
13. Taiwan Food and Drug Administration, Ministry of Health and Welfare. Dupixent® (dupilumab injection) prescribing information. https://www.fda.gov.tw/