ACR Meeting Abstracts

ACR Meeting Abstracts

Abstract Number: 2891

Do Certain Dmards Increase Risk of New-Onset Type 2 Diabetes in RA Patients? a Disease Risk Score Analysis Using Administrative Databases

E Alemao, Z Guo and L Burns, Bristol-Myers Squibb, Princeton, NJ

Meeting: 2017 ACR/ARHP Annual Meeting

Date of first publication: September 18, 2017

Keywords: , ,

Session Information

Date: Wednesday, November 8, 2017

Title: Health Services Research II: Methods and Technology in Care and Research

Session Type: ACR Concurrent Abstract Session

Session Time: 9:00AM-10:30AM

Background/Purpose: Data on the association between RA and type 2 diabetes mellitus (T2DM) are inconsistent, suggesting RA treatments such as glucocorticoids (GCs) and hydroxychloroquine could impact T2DM risk.1–3 Recent randomized controlled trials demonstrated that biologic (b)DMARDs such as abatacept (ABA) improve pancreatic beta cell function in T1DM and have a lower risk of T2DM in RA patients (pts) in clinical practice.4,5 The objective of this analysis was to evaluate the development of T2DM in RA pts treated with ABA, other bDMARDs and conventional (c)DMARDs. Methods: Administrative claims data from Optum Clinformatics Data Mart (database A) and QuintilesIMSTM PharMetrics Plus (database B) from 2006 to 2016 were used. Study inclusion criteria were: 2 diagnosis codes for RA plus 1 DMARD prescription; aged ≥18 yrs; and ≥3 mths baseline (pre-index date) and 3 mths of follow-up (post-index date). Mutually exclusive treatment groups were created based on the first prescription (index date) using hierarchy of ABA, other bDMARD and cDMARD. Also, an RA group without record of DMARD (NoDMARD) was identified. The index date for NoDMARD was first diagnosis date for NoDMARD. Incident T2DM was identified using 1 ICD-9 or ICD-10 diagnosis code for T2DM. Assessment of T2DM risk between treatment groups was based on traditional regression and a disease risk score (DRS) approach. DRS is the probability of developing T2DM estimated using a Cox model with a pre-specified list of 27 covariates. Adjusted incidence rate for T2DM was based on Cox model (stratified by DRS groups categorized into 4 equal groups based on quartile scores) with treatment as independent variable. Results: From database A, 105,683 (72.8% female, 58.3 [16.0] yrs) and from database B, 266,842 (73.1% female, mean [SD] age 51.6 [13.3] yrs) RA pts were included. Respectively, 2.9%, 10.4%, 34.8% and 51.8% were prescribed ABA, bDMARD, cDMARD and NoDMARD in database A; and 2.7%, 13.0%, 35.4% and 48.9% in database B. Based on the overall pooled sample size of 372,525 RA pts, the incidence rate per 1000 pt-yrs for new-onset T2DM was 37.8 (95% CI 37.4, 38.1). Pooled hazard ratios (HRs) for T2DM were significantly higher for bDMARDs vs ABA in both DRS-based and regression-based approaches; HRs for cDMARDs were also higher (vs ABA) but were significant only in the regression approach. NoDMARD group (vs ABA) had the highest HRs for incidence of T2DM (Table).  

Conclusion: We observed a lower incident risk of T2DM in RA pts treated with abatacept (vs risk in pts treated with other bDMARDs and cDMARDs ) in two large real-world databases, warranting further comparisons. NoDMARD group should be interpreted with caution as it comprises a heterogeneous pt population.

1.  Simard JF, Mittleman MA. J Rheumatol 2007;34:469–73.

2.  Solomon DH, et al. Arthritis Rheum 2004;50:3444–9.

3.  Hakala M, et al. J Rheumatol 1992;19:856–8.

4.  Orban T, et al. Lancet 2011;378:412–19.

5.  Ozen G, et al. Ann Rheum Dis 2017;76:848–54.

 

Table. HRs for Incidence of New-Onset Type 2 Diabetes Mellitus in Patients with RA by DMARD Treatment 
  HR (95% CI) p value
Pooled data of databases A and B – disease risk score method
bDMARD (vs abatacept) 1.12 (1.04, 1.20) 0.003
cDMARD (vs abatacept) 1.06 (0.99, 1.14) 0.079
NoDMARD (vs abatacept) 1.29 (1.21, 1.38) <0.001
Pooled data of databases A and B – regression method
bDMARD (vs abatacept)

1.14 (1.05, 1.22)

<0.001

cDMARD (vs abatacept)

1.08 (1.01, 1.16)

0.028

NoDMARD (vs abatacept)

1.33 (1.24, 1.42)

<0.001

bDMARD=biologic DMARD; cDMARD=conventional DMARD; HR=hazard ratio; NoDMARD=no record of DMARD prescription

 
Disclosure: E. Alemao, Bristol-Myers Squibb, 1,Bristol-Myers Squibb, 3; Z. Guo, Bristol-Myers Squibb, 3; L. Burns, Bristol-Myers Squibb, 1,Bristol-Myers Squibb, 3.

To cite this abstract in AMA style:

Alemao E, Guo Z, Burns L. Do Certain Dmards Increase Risk of New-Onset Type 2 Diabetes in RA Patients? a Disease Risk Score Analysis Using Administrative Databases [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 10). https://acrabstracts.org/abstract/do-certain-dmards-increase-risk-of-new-onset-type-2-diabetes-in-ra-patients-a-disease-risk-score-analysis-using-administrative-databases/. Accessed .

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/do-certain-dmards-increase-risk-of-new-onset-type-2-diabetes-in-ra-patients-a-disease-risk-score-analysis-using-administrative-databases/