Session Information
Date: Tuesday, November 10, 2015
Title: Epidemiology and Public Health IV: Risk Factors and Comorbidities for SLE and Psoriatic Arthritis
Session Type: ACR Concurrent Abstract Session
Session Time: 4:30PM-6:00PM
Background/Purpose: Patients with systemic
lupus erythematosus (SLE) are at high risk of
infections partly from use of immunosuppressives
(IS). Clinical trials are powered to detect differences in treatment efficacy
and therefore the comparative risks of infections are understudied. We compared
the rates of hospitalized infection and all-cause mortality among SLE patients
receiving mycophenolate mofetil
(MMF) vs. azathioprine (AZA), two
commonly and often interchangeably used IS. We hypothesized based on prior
small studies that MMF users may be at increased risk of infection and death.
Methods: We used Medicaid data
from 47 states and Washington, DC, 2000-2006, to identify adults (18-65 years)
with ≥2 SLE ICD-9 codes separated by ≥30 days, who were new
users of MMF or AZA with no use of either drug in the prior 183 days of
continuous Medicaid enrollment. The index date was the first MMF or AZA
prescription fill date. We excluded patients with HIV/AIDS or prior organ
transplant. We used multivariable logistic regression to calculate propensity
scores (PS) for receipt of MMF vs. AZA. Covariates included sociodemographics,
year, comorbidities, lupus nephritis, vaccines, healthcare utilization, hydroxychloroquine and other IS use during the 183 days
prior to the index date, and cumulative corticosteroid dose during the 60 days
prior to the index date. We matched new users of MMF and AZA by PS (1:1) and
conducted as-treated (AT) analyses, censoring at drug discontinuation, gaps
>30 days, switch to the comparator drug, death, disenrollment or end of
follow-up. In secondary intention-to-treat (ITT) analyses, patients were
censored at death, disenrollment or end of follow-up. We identified
hospitalized infections using a validated algorithm (PPV 80%) and used Cox
proportional hazard regression models to estimate hazard ratios (HR) of first
infection and death comparing PS-matched new users of MMF and AZA (referent).
Results: We identified 926
PS-matched pairs of MMF and AZA new users with SLE. Characteristics were well
balanced: mean age was 34 (SD 11) years, 92% were female, 45% were Black, 25%
White and 20% Hispanic. In AT analyses, the incidence rate (IR) of first
hospitalized infection per 100 person-years was 13.95 among MMF users and 11.91
among AZA users (Table). The hazard
ratio of first infection was similar for new MMF vs. AZA users (HR 1.18, 95% CI
0.80-1.75). Mortality rates per 100 person-years were 2.39 among MMF users and
1.74 among AZA users (HR 1.39, 95% CI 0.53-3.65). Findings of secondary ITT
analyses were similar to the AT analyses with no significant differences
between groups.
Conclusion: In this large, diverse
cohort of SLE patients, while the rates of hospitalized infection and all-cause
mortality were marginally higher among MMF vs. AZA new users, there were no
significant differences. Further studies are needed to determine whether there
are differences in other relevant safety outcomes.
|
Table. Propensity Score-Matched* Analyses of First Hospitalized Infection and All-Cause Mortality among SLE New Users of Mycophenolate mofetil (MMF) vs. Azathioprine (AZA)
|
|||||||||
|
|
Mycophenolate Mofetil (N=926)
|
Azathioprine (N=926)
|
|||||||
|
Analysis**
|
Outcome
|
Events
|
Person-years
|
IR+ (95% CI)
|
HR++ (95% CI)
|
Events
|
Person-years
|
IR+ (95% CI)
|
HR++
|
|
As treated (AT)
|
First infection |
56 |
401.5 |
13.95 (13.83-14.07) |
1.18 (0.80-1.75) |
46 |
386.3 |
11.91 (11.80-12.02) |
Ref |
|
Death |
10 |
417.8 |
2.39 (2.35-2.44) |
1.39 (0.53-3.65) |
7 |
402.8 |
1.74 (1.70-1.78) |
Ref |
|
|
Intention-to-treat (ITT, 183 days)
|
First infection |
79 |
401.7 |
19.67 (19.52-19.81) |
1.13 (0.82-1.56) |
70 |
402.1 |
17.41 (17.27-17.54) |
Ref |
|
Death |
15 |
421.6 |
3.56 (3.50-3.62) |
1.06 (0.51-2.2) |
14 |
418.9 |
3.34 (3.29-3.40) |
Ref |
|
|
Intention-to-treat (ITT, 365 days)
|
First infection |
116 |
690.8 |
16.79 (16.69-16.89) |
1.25 (0.96-1.64) |
93 |
694.8 |
13.39 (13.30-13.48) |
Ref |
|
Death |
18 |
748.3 |
2.41 (2.37-2.44) |
0.94 (0.49-1.79) |
19 |
742.6 |
2.56 (2.52-2.60) |
Ref |
|
|
*Propensity scores included age, sex, race/ethnicity, US region, SLE comorbidity index (Ward MM, 2000), comorbidities, lupus nephritis, number of SLE and renal-related laboratory tests, number of medications, immunosuppressive use, 60-day cumulative corticosteroid dose, NSAID use, healthcare utilization, vaccinations, and calendar year **Mean follow-up for AT infection analysis was 0.43 (SD 0.53) person-years, for ITT (183 days) 0.43 (SD 0.13) person-years and for ITT (365 days) 0.75 (SD 0.33) person-years + IR is the incidence rate per 100 for first hospitalized infection and for all-cause death ++ HR is the hazard ratio calculated using Cox proportional hazard regression models of propensity score-matched MMF vs. AZA new users |
|||||||||
To cite this abstract in AMA style:
Feldman CH, Marty FM, Winkelmayer WC, Guan H, Franklin JM, Solomon DH, Kim SC, Costenbader KH. Comparative Rates of Serious Infections and All-Cause Mortality Among Systemic Lupus Erythematosus Patients Receiving Mycophenolate Mofetil Versus Azathioprine [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/comparative-rates-of-serious-infections-and-all-cause-mortality-among-systemic-lupus-erythematosus-patients-receiving-mycophenolate-mofetil-versus-azathioprine/. Accessed .« Back to 2015 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/comparative-rates-of-serious-infections-and-all-cause-mortality-among-systemic-lupus-erythematosus-patients-receiving-mycophenolate-mofetil-versus-azathioprine/