Predictors of Persistency with TNFi in Biologic-Experienced Versus Biologic-Naive Psa Patients Enrolled in the Corrona Registry

CJ Etzel¹, Bradley S. Stolshek², Sabrina Rebello³, David Collier⁴, Alex Mutebi⁵, Sally W Wade⁶, Wendi Malley⁷, JD Greenberg⁷ and Leslie R Harrold⁸, ¹The University of Texas MD Anderson Cancer Center, Houston, TX, ²Amgen, Thousand Oaks, CA, ³Epidemiology, Corrona, LLC, Southborough, MA, ⁴Amgen Inc., Thousand Oaks, CA, ⁵Global Health Economics, Amgen, Thousand Oaks, CA, ⁶Wade Outcomes Research and Consulting, Salt Lake City, UT, ⁷Corrona, LLC, Southborough, MA, ⁸UMass Medical School, Worcester, MA

Meeting: 2016 ACR/ARHP Annual Meeting

Date of first publication: September 28, 2016

Keywords: Biologic drugs, psoriatic arthritis and treatment

Session Information

Date: Monday, November 14, 2016

Title: Spondylarthropathies and Psoriatic Arthritis – Clinical Aspects and Treatment - Poster II: Psoriatic Arthritis

Session Type: ACR Poster Session B

Session Time: 9:00AM-11:00AM

Background/Purpose: Little is known about factors associated with persistency of TNFi use among biologic-naïve and biologic-experienced Psoriatic Arthritis (PsA) patients in routine clinical practice in the US. The purpose of this analysis/study is to identify predictors of persistence with TNFi in biologic naïve (bio-naïve) and biologic experienced (bio-experienced) TNFi initiators in the US Corrona registry.

Methods: We identified all PsA patients initiating a TNFi with at least 1 follow-up rheumatology visit from October 2002 to March 2013. Patients were followed until the earliest of: discontinuation of the TNFi, initiation of another biologic/small molecule therapy or last follow-up visit; such that patients who remained on original TNFi were defined as persistent and patients that discontinued or switched TNFi were defined as non-persistent. Patient demographic and clinical characteristics at index date (TNFi initiation) were evaluated. Kaplan-Meier curves evaluated time to discontinuation/switch of TNFi, with censoring for all patients with ≥ 1year gap between visits. The log rank test assessed differences in persistency between bio-naïve versus bio-experienced patients. Cox proportional hazards models (allowing for factors to vary over time [time varying]) were performed to identify predictors of persistency, controlling for potential confounders; models were completed, separately, for the bio-naïve and bio-experienced subgroups.

Results: A total of 1241 patients initiated a TNFi during the study period (at initiation 549 [44%] bio-naïve; 692 [56%] bio-experienced). Median time to discontinuation was 32 months (95% CI 27-37) in bio-naïve patients and 23 (95% CI 18-29) in bio-experienced TNFi initiators which was significantly different based on the log rank test. Moderate or high disease activity based on Clinical Disease Activity Index (CDAI) was associated with being non-persistent within both subgroups, although the level of effect appeared stronger in bio-naïve PsA patients (Table). Bio-naïve patients with higher comorbidity burden had a higher, although not statistically significant, risk of being non-persistent. In contrast, bio-experienced patients with prior treatment with a conventional synthetic disease modifying anti-rheumatic drug (csDMARD) or higher degree of skin disease had higher risks of being non-persistent. Within both groups, patients with shorter disease duration were more likely to be persistent.

Covariates*	Cox Proportional Hazards Models**
	Biologic-Naïve HR (95% CI)	Biologic-Experienced HR (95% CI)
Modified Charlson comorbidity index***
Presence of 1 comorbid condition (vs. none)	1.32 [0.81; 2.15]	0.84 [0.60; 1.19]
Presence of 2 or more comorbid conditions (vs. none)	2.50 [0.99; 6.29]	0.85 [0.47; 1.57]
PsA characteristics
Disease duration	0.96 [0.92; 0.99]	0.99 [0.97; 1.00]
Prior treatment with a csDMARD	1.18 [0.78; 1.79]	1.40 [1.04; 1.89]
Concomitant prednisone use	1.42 [0.72; 2.83]	1.48 [0.99; 2.23]
Disease activity (time varying)
Patient reported pain	1.00 [1.00; 1.01]	1.00 [0.99; 1.01]
Moderate disease activity (vs. LDA)	2.50 [1.47; 4.27]	1.62 [1.13; 2.32]
High disease activity (vs. LDA)	2.15 [1.11; 4.13]	1.60 [1.01; 2.53]
Physician global skin assessment	1.18 [0.74; 1.87]	1.38 [1.01; 1.87]

*PsA – psoriatic arthritis, LDA-low disease activity **Adjusted for female gender, race, age, BMI, and smoking status. ***Excludes PsA; dementia, kidney disease, hemiplegia and AIDS not included

Conclusion: Predictors of persistency differed in those who were bio-naïve versus bio-experienced when initiating TNFi. Disease activity appeared to be a stronger predictor of non-persistence in those who were biologic-naïve as compared to those who were experienced.

Disclosure: C. Etzel, Corrona, LLC, 3,Merck, 9; B. S. Stolshek, Amgen, 1,Amgen, 3; S. Rebello, Corrona, LLC, 3; D. Collier, Amgen, 1,Amgen, 3; A. Mutebi, Amgen, 1,Amgen, 3; S. W. Wade, Amgen, 5; W. Malley, Corrona, LLC, 3; J. Greenberg, Corrona, LLC, 1,Corrona, LLC, 3,Genentech, Janssen, Novartis and Pfizer, Eli Lilly, 5; L. R. Harrold, Corrona, LLC, 3,Pfizer Inc, 2,Roche, 5,Corrona, LLC, 1.

To cite this abstract in AMA style:

Etzel C, Stolshek BS, Rebello S, Collier D, Mutebi A, Wade SW, Malley W, Greenberg J, Harrold LR. Predictors of Persistency with TNFi in Biologic-Experienced Versus Biologic-Naive Psa Patients Enrolled in the Corrona Registry [abstract]. Arthritis Rheumatol. 2016; 68 (suppl 10). https://acrabstracts.org/abstract/predictors-of-persistency-with-tnfi-in-biologic-experienced-versus-biologic-naive-psa-patients-enrolled-in-the-corrona-registry/. Accessed .

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-persistency-with-tnfi-in-biologic-experienced-versus-biologic-naive-psa-patients-enrolled-in-the-corrona-registry/