INTRODUCTION
Atopic dermatitis (AD) is an exceedingly common inflammatory skin disorder that typically presents with pruritic eczematous lesions with a chronic and relapsing course.1 The prevalence of AD in the adult population is not fully characterized, but some authors suggested a lifetime prevalence that ranged from 3.0% to 17.7%.2 Moderate to severe AD usually requires long-term systemic therapy with classic agents such as cyclosporine, azathioprine, methotrexate or systemic corticosteroids, often associated with severe side-effects and variable efficacy.
AD is deeply associated with other diseases within the atopy spectrum3 and increased awareness about the importance of Th2 inflammatory cells in these conditions4allowed dupilumab, a fully human monoclonal antibody that targets the shared IL-4Rα subunit of heterodimeric IL-4 and IL-13 receptors, to be approved as the first biologic treatment for adult patients with inadequately controlled moderate-to-severe AD. In phase 3 clinical trials, 16-week dupilumab treatment significantly improved clinical parameters and symptoms of AD, while maintaining an acceptable safety profile.5-7
Although dupilumab has emerged in these trials as a breakthrough therapy,8 continuous data collection in the postmarketing phase is needed in order to validate its efficacy and safety performance in the real-world clinical setting. As such, we aim to describe our daily-practice experience with dupilumab, providing further evidence for its clinical use.
MATERIAL AND METHODS
In this observational cohort study, we prospectively included all adult patients with moderate to severe AD who initiated dupilumab in our portuguese tertiary care hospital, from July 2019 to April 2020.
Our dermatology team was instructed to initially register the patient’s epidemiologic data; comorbidities; previous treatments; and Eczema Area and Severity Index (EASI).
A 600 mg loading dose of dupilumab was injected subcutaneously at baseline, followed by an injection of 300 mg dupilumab every other week. The need for other ongoing systemic treatments was clinically assessed and recorded. Topical anti-inflammatory agents and moisturizers usage was recommended but not systematically described.
Patients were re-evaluated after 16 weeks of treatment and EASI, adverse events and treatment interruptions were assessed. Appointments between week 12 and 20 of treatment - motivated by physicians or patient agenda constraints - were also considered suitable for inclusion. At the end of the study, data was reviewed by the authors and patients with non-compliant clinical records were excluded.
Statistical analysis was performed using SPSS 24.0 (IBM, Armonk, NY, U.S.A.).
RESULTS
Our study analysed 32 patients but 7 were excluded due to incomplete clinical records. Of the 25 included patients (Table 1), most were male (64%; 16 of 21) with a mean age of 32 years (20-60). Asthma (36%) and allergic rhinitis (20%) were the most frequent comorbidities. All patients had been previously treated with oral corticosteroids; 96% (24 of 25) with oral cyclosporine; 60% with phototherapy (n=15); 36% with methotrexate (n=9); 24% with mycophenolate mofetil (n=6); and 20% with azathioprine (n=5).
Epidemiological and clinical characteristics at baseline | Number of patients (%) |
Sex • Male, n (%) • Female, n (%) | 16 (64%) 9 (36%) |
Age at the start of dupilumab treatment (years) • Mean (range) | 32 (20 - 60) |
Previous use of conventional systemic immunosuppressants, n (%) • Oral corticosteroids • Cyclosporine • Methotrexate • Mycophenolate mofetil • Azathioprine | 25 (100%) 24 (96%) 9 (36%) 6 (24%) 5 (20%) |
Previous use of phototherapy | 15 (60 %) |
Atopic/allergic conditions, n (%) • Asthma • Allergic rhinitis • Food allergy | 9 (36%) 5 (20%) 1 (4%) |
EASI at week 0 • Mean (range) | 27.8 (11 - 51) |
The mean EASI score at week 0 was 27.8, ranging from 11 to 51. Concomitant treatment was initially maintained in 32% of patients (n=8), with three patients medicated with cyclosporine and others with mycophenolate mofetil; oral prednisolone; narrow-band UVB; azathioprine with oral prednisolone; or cyclosporine with oral prednisolone.
At week 16 (+/- 4 weeks), one patient missed the revaluation assessment, voluntarily abandoned therapy against clinical decision and was excluded from the study. The mean EASI score at week 16 was 8.8 (68.6% reduction), ranging from 0 to 50 (Table 2). EASI 50 (defined by an EASI score improvement of at least 50%) was achieved in 87.5% of patients (n=21); 58.3% (n=14) reached EASI 75; 29.1% (n=7) EASI 90; and 16.6% (n=4) EASI 100. However, 8.3% (n=2) of patients did not respond to dupilumab therapy.
Efficacy and safety outcomes (week 16) | Number of patients (%) |
---|---|
EASI • Mean score (range) • Mean reduction • EASI 50, % (n) • EASI 75, % (n) • EASI 90, % (n) • EASI 100, % (n) • Non-responders, % (n) | 8.8 (0-50) 68.6% 87.5% (21) 58.3% (14) 29.1% (7) 16.6% (4). 8.3% (2) |
Adverse-events, % (n) • Conjunctivitis, % (n) • Eyelid eczema | 25% (6) 20.8% (5) 4.2% (1) |
Outcomes are a comparison between baseline and follow-up at week 16 (+/- 4 weeks). EASI, Eczema Area and Severity Index; EASI 50, EASI score improvement of at least 50%; EASI 75, EASI score improvement of at least 75%; EASI 90, EASI score improvement of at least 90%; EASI 100, EASI score improvement of 100%.
When evaluating patients with no concomitant systemic treatment (n=16), the mean initial EASI was 25.9 and EASI at week 16 was 6.8 (73.8% reduction; EASI 50 in 93.7% of patients (n=15); EASI 75 in 68.7% of patients (n=11); EASI 90 in 31.2% of patients (n=5); EASI 100 in 18.7% of patients (n=3); 6.2% of patients with no response (n=1)).
Dupilumab-associated conjunctivitis (DAC) was the main reported side-effect, affecting 20.8% (n=5) of patients. Another patient presented with eyelid eczema (4.2%). No patient had to discontinue dupilumab due to an adverse-event.
DISCUSSION
In AD, barrier-disrupted keratinocytes produce immunoregulatory cytokines (alarmins) such as thymic stromal lymphopoietin or IL-33, activating group 2 innate lymphoid cells (ILC2s).9 These activated ILC2s produce type 2 cytokines, which cause further skin barrier disruption and allow the entry of various antigens into the skin, leading to the differentiation of antigen-specific naive T cells into effector Th2 cells. These cells produce IL-4 and IL-13, known to be involved in several proinflammatory pathways of AD, such as down-regulation of filaggrin expression in keratinocytes (further increasing epidermal barrier dysfunction); amplification of IL-31-induced and histamine-induced pruritus; stimulation of B cells to produce immunoglobulin E (IgE) which binds to mast cells and induces their degranulation upon binding to allergens; or increased production of CCL17, CCL22 and CCL26, that together with IL-5 can further recruit Th2 cells and eosinophils.10,11
These key-functions of IL-4 and IL-13 in AD immune response, together with evidence of a common component shared by their receptors, were used in dupilumab development and explain the significant clinical improvement of AD patients treated with this drug. In phase 3 clinical trials, 16-week treatment with dupilumab (300 mg q2w) lead to a 70.07% mean EASI reduction, with 61% of patients reaching EASI 50; 50.2% EASI 75; and 31.8% EASI 90.5
Our study showed a similar mean EASI reduction and a significant increase of EASI 50 responses, reinforcing what has already been suggested by some early “real-life” data12,13: dupilumab is effective in most AD adult patients treated under daily-practice conditions.
However, our results have also highlighted that there is a considerable individual variability in the effectiveness of dupilumab, with some patients displaying an extraordinary treatment response (29.1% reaching EASI 90), while others failed to respond (8.3%).
Based on these data, predicting treatment response with dupilumab seems to be of the utmost importance. However, practical predictors of its effectiveness are still under investigation. While initial studies5,12-15suggested baseline EASI, IgE, lactate dehydrogenase (LDH), eosinophilia, allergic comorbidities or early-onset AD as possible predictive markers of treatment response, there are still no validated guidelines for treatment eligibility based on these possible predictors.
We believe that future studies should be focused on proper validation of these predictive biomarkers, in order to allow a better patient selection, a realistic setting of treatment goals and an improved management of our patient’s expectations.
Regarding the safety profile of dupilumab, phase 3 clinical trials showed that the overall incidence of adverse events was similar between dupilumab and placebo groups.5 In fact, these trials underlined that placebo-treated patients had a higher-risk for a serious adverse event (mainly AD exacerbation), although conjunctivitis (9.7%) and injection-site reactions (16.7%) were more common in the dupilumab-treated group.
Daily-practice early data, however, showed that DAC’s incidence rate was significantly higher.5,16While the mechanism for this adverse effect is still unknown, some authors proposed that ocular comorbidities are dependent on disease severity, prior conjunctivitis history or certain biomarkers such as thymus and activation-regulated chemokine (TARC) or IgE.17 Hence, we tend to agree that intrinsic differences in the analysed cohorts - namely regarding conjuntivitis’ proposed risk-factors - are a likely explanation for the discrepancy between clinical trials and real-life data, and that the true incidence of conjunctivitis-induced by dupilumab was initially underestimated.
Currently, there is no standard treatment to prevent and manage DAC, although topical corticosteroids, topical calcineurin inhibitors, cyclosporin eye drops, hyaluronic acid eye drops or artificial tears have been successfully used in several patients.18 In our study, some of our physicians used artificial tear drops in the beginning of treatment.
However, this clinical intervention was not systematically recorded and therefore we cannot evaluate its efficacy on preventing DAC, nor recommend its usage just based on our data. As such, we believe that future studies should properly address this question in order to produce solid evidence that can support a clinical orientation guideline for the prevention of DAC.
Finally, we acknowledge that our study has some limitations. Dupilumab’s efficacy on pruritus reduction and its overall impact on patients’ quality of life are among other important key-metrics that should be considered when dupilumab is prescribed, and they were not included in our study. Besides, by narrowing dupilumab’s efficacy and safety assessment to a solo 16-week EASI revaluation - and by not including an earlier week 4-8 observation - we missed the opportunity to determine its onset of action (or of its complications) in a real-life setting.
CONCLUSION
Our study corroborates dupilumab as an effective treatment for AD in the real-world clinical setting, with some individual variances that should be further explored in a near future. No serious events were reported but conjunctivitis is a common side-effect that should be specifically mentioned to all patients..