Date: Monday, November 6, 2017
Session Type: ACR Concurrent Abstract Session
Session Time: 2:30PM-4:00PM
Background/Purpose: Highly effective, targeted DMARD therapies with different mechanisms of action are available for RA. Translating precision medicine into clinical practice requires treatment-specific predictive models, with a goal of individualized, targeted therapy.Therefore, we created separate predictive models for response to abatacept (ABA) or adalimumab (ADA), using baseline biomarker (BM) data from the head-to-head AMPLE study.1
Methods: Predictive models were built using demographic data, baseline disease characteristics and several BMs, including RF, cyclic citrullinated peptide-2 (CCP2) and BMs from the multi-biomarker disease activity test as predictor variables and ‘polar’ clinical responses as the response variables. The polar responses were defined as patients (pts) who, after 1 year of treatment, achieved an ACR70 response or failed to achieve an ACR20 response. The elastic net method2 was used to build separate predictive models for ABA and ADA responders using their respective clinical data with 6-fold cross-validation (CV) repeated 10 times. Parameter tuning for model selection was based on a fixed alpha of 0.95 and varying levels of lambda. The final model was selected based on the lambda with maximum mean area under the curve (AUC) across all folds of CV. Glmnet3 and caret4 packages in R were used for model-building purposes.
Results: Predictive models generated included 13 variables for ABA and 11 for ADA. Of all the variables, resistin, vascular cell adhesion molecule-1, sex, CCP2 and pt-reported disease activity were unique to the ABA predictor, whereas matrix metalloproteinase-1, physician-reported disease activity and disease duration were unique to the ADA predictor. The variables common to both models showed the same association (positive or negative) but differing magnitude with response. AUC by receiver operating characteristic curves were used to assess the performance of the predictive models. The performance (AUC) of the ABA model on the ABA and ADA arms was 0.855 and 0.530, respectively (Figure a, b). The performance (AUC) of the ADA model on the ADA and ABA arms was 0.860 and 0.631, respectively (Figure c, d). This indicates that the models are very specific for their respective treatment and not for the other treatment.
Conclusion: Response-to-treatment predictive models were generated using baseline data from AMPLE that were highly specific to their respective treatment. This suggests that treatment-specific response predictors could be developed and should be considered, and highlights the value of head-to-head studies in predictive biomarker generation. Further testing on validation datasets is warranted.
1. Weinblatt M, et al. Arthritis Rheum 2013;65:28–38.
2. Zou H, Hastie T. J R Stat Soc Series B Stat Methodol 2005;67:301–20.
3. Friedman J, et al. J Stat Softw 2010;33:1–22.
4. Kuhn M. caret: Classification and Regression Training. R package version 5.15-044. 2012.
To cite this abstract in AMA style:Bandyopadhyay S, Maldonado M, Ammar R, Schiff M, Weinblatt M, Fleischmann R, Connolly S. Development of Abatacept- and Adalimumab-Specific Predictive Models of Response to Therapy in RA Using Data from a Head-to-Head Study [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 10). https://acrabstracts.org/abstract/development-of-abatacept-and-adalimumab-specific-predictive-models-of-response-to-therapy-in-ra-using-data-from-a-head-to-head-study/. Accessed October 28, 2021.
« Back to 2017 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/development-of-abatacept-and-adalimumab-specific-predictive-models-of-response-to-therapy-in-ra-using-data-from-a-head-to-head-study/