Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2021 Dec 1;157(12):1414-1424.
doi: 10.1001/jamadermatol.2021.3668.

Machine Learning-Based Deep Phenotyping of Atopic Dermatitis: Severity-Associated Factors in Adolescent and Adult Patients

Laura Maintz  1   2 , Thomas Welchowski  2   3 , Nadine Herrmann  1   2 , Juliette Brauer  1   2 , Anna Sophie Kläschen  1   2 , Rolf Fimmers  3 , Matthias Schmid  3 , Thomas Bieber  1   2 ; CK-CARE Study Group; Peter Schmid-Grendelmeier  2   4 , Claudia Traidl-Hoffmann  2   5   6 , Cezmi Akdis  2   7 , Roger Lauener  2   8 , Marie-Charlotte Brüggen  2   4   9 , Claudio Rhyner  2 , Eugen Bersuch  4 , Ellen Renner  10   11 , Matthias Reiger  2   5   6 , Anita Dreher  2 , Gertrud Hammel  2   5   6 , Daria Luschkova  2   5   6 , Claudia Lang  4
Affiliations

Machine Learning-Based Deep Phenotyping of Atopic Dermatitis: Severity-Associated Factors in Adolescent and Adult Patients

Laura Maintz et al. JAMA Dermatol. .

Abstract

Importance: Atopic dermatitis (AD) is the most common chronic inflammatory skin disease and is driven by a complex pathophysiology underlying highly heterogeneous phenotypes. Current advances in precision medicine emphasize the need for stratification.

Objective: To perform deep phenotyping and identification of severity-associated factors in adolescent and adult patients with AD.

Design, setting, and participants: Cross-sectional data from the baseline visit of a prospective longitudinal study investigating the phenotype among inpatients and outpatients with AD from the Department of Dermatology and Allergy of the University Hospital Bonn enrolled between November 2016 and February 2020.

Main outcomes and measures: Patients were stratified by severity groups using the Eczema Area and Severity Index (EASI). The associations of 130 factors with AD severity were analyzed applying a machine learning-gradient boosting approach with cross-validation-based tuning as well as multinomial logistic regression.

Results: A total of 367 patients (157 male [42.8%]; mean [SD] age, 39 [17] years; 94% adults) were analyzed. Among the participants, 177 (48.2%) had mild disease (EASI ≤7), 120 (32.7%) had moderate disease (EASI >7 and ≤ 21), and 70 (19.1%) had severe disease (EASI >21). Atopic stigmata (cheilitis: odds ratio [OR], 8.10; 95% CI, 3.35-10.59; white dermographism: OR, 4.42; 95% CI, 1.68-11.64; Hertoghe sign: OR, 2.75; 95% CI, 1.27-5.93; nipple eczema: OR, 4.97; 95% CI, 1.56-15.78) was associated with increased probability of severe AD, while female sex was associated with reduced probability (OR, 0.30; 95% CI, 0.13-0.66). The probability of severe AD was associated with total serum immunoglobulin E levels greater than 1708 IU/mL and eosinophil values greater than 6.8%. Patients aged 12 to 21 years or older than 52 years had an elevated probability of severe AD; patients aged 22 to 51 years had an elevated probability of mild AD. Age at AD onset older than 12 years was associated with increased probability of severe AD up to a peak at 30 years; age at onset older than 33 years was associated with moderate to severe AD; and childhood onset was associated with mild AD (peak, 7 years). Lifestyle factors associated with severe AD were physical activity less than once per week and (former) smoking. Alopecia areata was associated with moderate (OR, 5.23; 95% CI, 1.53-17.88) and severe (OR, 4.67; 95% CI, 1.01-21.56) AD. Predictive performance of machine learning-gradient boosting vs multinomial logistic regression differed only slightly (mean multiclass area under the curve value: 0.71 [95% CI, 0.69-0.72] vs 0.68 [0.66-0.70], respectively).

Conclusions and relevance: The associations found in this cross-sectional study among patients with AD might contribute to a deeper disease understanding, closer monitoring of predisposed patients, and personalized prevention and therapy.

PubMed Disclaimer

Conflict of interest statement

Conflict of Interest Disclosures: Dr Maintz reported grants and personal fees from CK-CARE during the conduct of the study; and being an investigator for LEO Pharma, AbbVie, Galderma, Pfizer, and Eli Lilly outside the submitted work. Dr Herrmann reported grants and personal fees from CK-CARE during the conduct of the study. Mrs Brauer reported personal fees from CK-CARE during the conduct of the study. Drs Schmid, Hammel, and Lang reported grants from Christine Kühne-Center for Allergy Research and Education (CK-CARE) during the conduct of the study. Prof Bieber reported personal fees from LEO during the conduct of the study; personal fees from Allmiral, AnaptysBio, Arena, Asana Biosciences, Bayer Health, Boehringer Ingelheim, BMS, Domain Therapeutics, Galapagos/Morphosys, Galderma, Glenmark, Incyte, IQVIA, Jansen, Kymab, LG Chem, Lilly, Novartis, Vifor Pharma, Pfizer, Pierre Fabre, and UCB outside the submitted work; and being the founder of the nonprofit biotech company Davos Biosciences within the International Kühne-Foundation. Prof Schmid-Grendelmeier reported honoraria from AbbVie, GlaxoSmithKline, Lilly, LEO, Pfizer, Sanofi, and Novartis outside the submitted work. Prof Lauener reported grants from Kühne Foundation during the conduct of the study. Dr Rhyner reported grants from LEO Pharma during the conduct of the study. Dr Reiger reported grants from CLR, Germany, and Beiersdorf, Germany; and personal fees from Bencard, Germany; La Roche-Posay, Germany; Galderma, Germany; Sebapharma, Germany; and LEO Pharma, Germany, outside the submitted work. No other disclosures were reported.

Figures

Figure 1.
Figure 1.. Total Serum IgE (tIgE) and Eosinophil Values as Predictive Parameters for Atopic Dermatitis (AD) Severity
A, The tIgE level correlates with AD severity (Eczema Area and Severity Index [EASI]). Orange dots: patients with increased tIgE levels (75.1% of patients, n = 263; age-dependent cutoff points, in patients older than 15 y: 100 kU/L, patients 12 to 15 years: 200 kU/L). Blue dots: patients with normal tIgE levels (24.9% of patients, n = 87). rs = Spearman rank correlation coefficient. B, Higher tIgE levels in patients with severe AD (EASI >21, n = 68) compared with moderate AD (EASI >7 and ≤ 21, n = 114) or mild AD (EASI <7, n = 168). C-E, The estimated probability of severe AD increased from tIgE greater than 1708 IU/mL, while the probability of mild AD decreased from tIgE less than 467 IU/mL. F-H, Patients with less than 3.79% eosinophils had a high probability of mild AD. The probability of moderate AD rose from 3.7% on, while patients with greater than 6.8% eosinophils had a rising probability of severe AD (reference range eosinophils: 0.5%-5.5%). C-H, Accumulated local effect plots of the tIgE levels visualize the expected change in probability for the grouping into mild, moderate, or severe AD dependent on tIgE levels (C-E) and eosinophil values (D-H) compared with the average prediction in the complete data set.
Figure 2.
Figure 2.. Association of Atopic Dermatitis (AD) Severity With Age at AD Onset
A, AD severity in patients stratified by age at AD onset. Boxplots visualize the Eczema Area and Severity Index (EASI) of patients with AD stratified by age at disease onset. Boxes display the 25th to 75th percentile (inner horizontal line: median) of the EASI scores; whiskers: minimum, maximum (excluding outliers). B, Childhood onset was associated with increased probability of mild AD; age at AD onset older than 12 years, with severe AD; and adulthood onset at older than 33 years, with moderate to severe AD. Accumulated local effect plots visualize the expected change in probability of grouping into mild, moderate, or severe AD dependent on age at onset of AD compared with the mean prediction in the complete data set.
Figure 3.
Figure 3.. Association of Atopic Dermatitis (AD) Severity With Atopic Stigmata
Higher severity scores and probability of moderate to severe AD in patients displaying atopic stigmata. Boxplots (leftmost graphs) visualize the Eczema Area and Severity Index (EASI) of patients with AD stratified by atopic stigmata. Boxes display the 25th to 75th percentile (inner horizontal line: median) of the respective EASI scores; whiskers: minimum, maximum (excluding outliers). Accumulated local effect plots (other graphs) visualize the expected change in probability of grouping into mild (EASI ≤7), moderate (EASI >7 and ≤21), or severe (EASI >21) AD dependent on respective atopic stigma compared with the mean prediction in the complete data set.
Figure 4.
Figure 4.. Association of Atopic Dermatitis (AD) Severity With Sex and Sports Frequency
A, Male patients (n = 157 [42.8%]) had higher severity scores compared with female patients (n = 210 [57.2%]), measured by Eczema Area and Severity Index (EASI). B, Male patients had a lower estimated probability of mild AD (male, 37.6% vs female, 56.2%) and an increased probability of moderate AD (male, 37.6% vs female, 29.0%) or severe AD (male, 24.8% vs female, 14.8%). Accumulated local effect plots visualize the expected change in probability for the grouping into mild, moderate, or severe AD dependent on sex compared with the average prediction in the complete data set. C, Male patients had higher total serum immunoglobulin E (tIgE) levels compared with female patients. D, AD severity measured by EASI dependent on frequency of sports per week. Never sport: n = 68, less than once per week: n = 58, once per week: n = 68, 2 to 3 times per week: n = 107, 4 to 7 times per week: n = 60, more than 7 times per week: n = 4. EASI scores (A, D) and tIgE levels (C) are depicted in boxplots. Boxes display the 25th to 75th percentile (inner horizontal line: median) of EASI scores (A, D) and tIgE (C); whiskers: respective minimum, maximum (excluding outliers).
See this image and copyright information in PMC

References

    1. Weidinger S, Beck LA, Bieber T, Kabashima K, Irvine AD. Atopic dermatitis. Nat Rev Dis Primers. 2018;4(1):1. doi:10.1038/s41572-018-0001-z - DOI - PubMed
    1. Abuabara K, Ye M, McCulloch CE, et al. . Clinical onset of atopic eczema: results from 2 nationally representative British birth cohorts followed through midlife. J Allergy Clin Immunol. 2019;144(3):710-719. doi:10.1016/j.jaci.2019.05.040 - DOI - PMC - PubMed
    1. Czarnowicki T, He H, Krueger JG, Guttman-Yassky E. Atopic dermatitis endotypes and implications for targeted therapeutics. J Allergy Clin Immunol. 2019;143(1):1-11. doi:10.1016/j.jaci.2018.10.032 - DOI - PubMed
    1. Thorsteinsdottir S, Stokholm J, Thyssen JP, et al. . Genetic, clinical, and environmental factors associated with persistent atopic dermatitis in childhood. JAMA Dermatol. 2019;155(1):50-57. doi:10.1001/jamadermatol.2018.4061 - DOI - PMC - PubMed
    1. Reiger M, Traidl-Hoffmann C, Neumann AU. The skin microbiome as a clinical biomarker in atopic eczema: promises, navigation, and pitfalls. J Allergy Clin Immunol. 2020;145(1):93-96. doi:10.1016/j.jaci.2019.11.004 - DOI - PubMed

Publication types

LinkOut - more resources