Multivariable Predictive Models to Identify the Optimal Biologic Therapy for Treatment of Patients With Psoriasis at the Individual Level
Research output: Contribution to journal › Journal article › Research › peer-review
Standard
Multivariable Predictive Models to Identify the Optimal Biologic Therapy for Treatment of Patients With Psoriasis at the Individual Level. / Nielsen, Mia-Louise; Petersen, Troels Christian; Maul, Julia-Tatjana; Wu, Jashin J; Rasmussen, Mads Kirchheiner; Bertelsen, Trine; Ajgeiy, Kawa Khaled; Skov, Lone; Thomsen, Simon Francis; Thyssen, Jacob Pontoppidan; Egeberg, Alexander.
In: JAMA Dermatology, Vol. 158, No. 10, 2022, p. 1149-1156.Research output: Contribution to journal › Journal article › Research › peer-review
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Multivariable Predictive Models to Identify the Optimal Biologic Therapy for Treatment of Patients With Psoriasis at the Individual Level
AU - Nielsen, Mia-Louise
AU - Petersen, Troels Christian
AU - Maul, Julia-Tatjana
AU - Wu, Jashin J
AU - Rasmussen, Mads Kirchheiner
AU - Bertelsen, Trine
AU - Ajgeiy, Kawa Khaled
AU - Skov, Lone
AU - Thomsen, Simon Francis
AU - Thyssen, Jacob Pontoppidan
AU - Egeberg, Alexander
PY - 2022
Y1 - 2022
N2 - IMPORTANCE Identifying the optimal long-term biologic therapy for patients with psoriasisis often done through trial and error.OBJECTIVE To identify the optimal biologic therapy for individual patients with psoriasisusing predictive statistical and machine learning models.DESIGN, SETTING, AND PARTICIPANTS This population-based cohort study used data fromDanish nationwide registries, primarily DERMBIO, and included adult patients treated formoderate-to-severe psoriasis with biologics. Data were processed and analyzed betweenspring 2021 and spring 2022.MAIN OUTCOMES AND MEASURES Patient clusters of clinical relevance were identified andtheir success rates estimated for each drug. Furthermore, predictive prognostic models toidentify optimal biologic treatment at the individual level based on data from nationwideregistries were evaluated.RESULTS Assuming a success criterion of 3 years of sustained treatment, this study included2034 patients with a total of 3452 treatment series. Most treatment series involved malepatients (2147 [62.2%]) originating from Denmark (3190 [92.4%]), and 2414 (69.9%) hadfinished an education longer than primary school. The average ages were 24.9 years atpsoriasis diagnosis and 45.5 years at initiation of biologic therapy. Gradient-boosted decisiontrees and logistic regression were able to predict a specific cytokine target (eg, interleukin-17inhibition) associated with a successful treatment with accuracies of 63.6% and 59.2%,and top 2 accuracies of 95.9% and 93.9%. When predicting specific drugs resulting insuccess, gradient boost and logistic regression had accuracies of 48.5% and 44.4%,top 2 accuracies of 77.6% and 75.9%, and top 3 accuracies of 89.9% and 89.0%.CONCLUSIONS AND RELEVANCE Of the treatment prediction models used in this cohort studyof patients with psoriasis, gradient-boosted decision trees performed significantly betterthan logistic regression when predicting specific biologic therapy (by drug as well as target)leading to a treatment duration of at least 3 years without discontinuation. Predicting theoptimal biologic could benefit patients and clinicians by minimizing the number of failedtreatment attempts.
AB - IMPORTANCE Identifying the optimal long-term biologic therapy for patients with psoriasisis often done through trial and error.OBJECTIVE To identify the optimal biologic therapy for individual patients with psoriasisusing predictive statistical and machine learning models.DESIGN, SETTING, AND PARTICIPANTS This population-based cohort study used data fromDanish nationwide registries, primarily DERMBIO, and included adult patients treated formoderate-to-severe psoriasis with biologics. Data were processed and analyzed betweenspring 2021 and spring 2022.MAIN OUTCOMES AND MEASURES Patient clusters of clinical relevance were identified andtheir success rates estimated for each drug. Furthermore, predictive prognostic models toidentify optimal biologic treatment at the individual level based on data from nationwideregistries were evaluated.RESULTS Assuming a success criterion of 3 years of sustained treatment, this study included2034 patients with a total of 3452 treatment series. Most treatment series involved malepatients (2147 [62.2%]) originating from Denmark (3190 [92.4%]), and 2414 (69.9%) hadfinished an education longer than primary school. The average ages were 24.9 years atpsoriasis diagnosis and 45.5 years at initiation of biologic therapy. Gradient-boosted decisiontrees and logistic regression were able to predict a specific cytokine target (eg, interleukin-17inhibition) associated with a successful treatment with accuracies of 63.6% and 59.2%,and top 2 accuracies of 95.9% and 93.9%. When predicting specific drugs resulting insuccess, gradient boost and logistic regression had accuracies of 48.5% and 44.4%,top 2 accuracies of 77.6% and 75.9%, and top 3 accuracies of 89.9% and 89.0%.CONCLUSIONS AND RELEVANCE Of the treatment prediction models used in this cohort studyof patients with psoriasis, gradient-boosted decision trees performed significantly betterthan logistic regression when predicting specific biologic therapy (by drug as well as target)leading to a treatment duration of at least 3 years without discontinuation. Predicting theoptimal biologic could benefit patients and clinicians by minimizing the number of failedtreatment attempts.
U2 - 10.1001/jamadermatol.2022.3171
DO - 10.1001/jamadermatol.2022.3171
M3 - Journal article
C2 - 35976663
VL - 158
SP - 1149
EP - 1156
JO - JAMA Dermatology
JF - JAMA Dermatology
SN - 2168-6068
IS - 10
ER -
ID: 316751840