Dimitrios A Pappas1, Michele M Hooper2, Joel M Kremer3, George Reed4, Ying Shan5, Deborah Wenkert2, Jeffrey D Greenberg6, Jeffrey R Curtis7. 1. Columbia University College of Physicians and Surgeons, New York, New York, and Corrona, Southborough, Massachusetts. 2. Amgen, Thousand Oaks, California. 3. Albany Medical Center, Albany, New York, and Corrona, Southborough, Massachusetts. 4. Corrona, Southborough, and University of Massachusetts, Worcester, Massachusetts. 5. Corrona, Southborough, Massachusetts. 6. Corrona, Southborough, Massachusetts, and New York University School of Medicine, New York. 7. University of Alabama at Birmingham.
Abstract
OBJECTIVE: To identify the rheumatoid arthritis (RA) characteristics associated with increased herpes zoster (HZ) risk in the Corrona registry RA patients, and to evaluate the risk in initiators of tumor necrosis factor inhibitors (TNFi) or non-TNFi biologic agents or (among those who were currently on or had been previously treated with methotrexate [MTX]) conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) other than MTX. METHODS: Cox regression modeling estimated the association between first HZ incidence and selected RA characteristics, including disease activity. Medication-related risk for HZ in RA patients taking current or past MTX (to exclude milder RA disease) were categorized by treatment initiation (TNFi versus non-TNFi versus csDMARD). Hazard ratios (HRs) estimated HZ risk of each treatment initiation category after stratification on trimmed propensity score (PS) quintiles to control for potential confounders. RESULTS: A total of 28,852 patients contributed 95,287 person-years. Seven hundred twenty-nine observed HZ cases yielded a 7.7 (95% confidence interval [95% CI] 7.1-8.2) per 1,000 patient-years crude incidence rate, lower than found in prior RA cohorts. However, consistent with prior studies, increasing age (HR 1.14, 95% CI 1.09-1.19 per 5 years) and prednisone therapy ≥7.5 mg/day (HR 1.81, 95% CI 1.23-2.67) were associated with a higher HZ risk. Referent to TNFi exposure, PS-stratified analysis showed an HR for csDMARDs of 1.36 (95% CI 0.82-2.25) and for non-TNFi of 0.83 (95% CI 0.51-1.38). CONCLUSION: In the Corrona registry, the HZ risk in RA patients taking prior or current MTX increased with older age and higher prednisone dose. The HZ risk among these patients with RA was comparable after initiation of TNFi versus non-TNFi versus csDMARDs.
OBJECTIVE: To identify the rheumatoid arthritis (RA) characteristics associated with increased herpes zoster (HZ) risk in the Corrona registry RApatients, and to evaluate the risk in initiators of tumor necrosis factor inhibitors (TNFi) or non-TNFi biologic agents or (among those who were currently on or had been previously treated with methotrexate [MTX]) conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) other than MTX. METHODS: Cox regression modeling estimated the association between first HZ incidence and selected RA characteristics, including disease activity. Medication-related risk for HZ in RApatients taking current or past MTX (to exclude milder RA disease) were categorized by treatment initiation (TNFi versus non-TNFi versus csDMARD). Hazard ratios (HRs) estimated HZ risk of each treatment initiation category after stratification on trimmed propensity score (PS) quintiles to control for potential confounders. RESULTS: A total of 28,852 patients contributed 95,287 person-years. Seven hundred twenty-nine observed HZ cases yielded a 7.7 (95% confidence interval [95% CI] 7.1-8.2) per 1,000 patient-years crude incidence rate, lower than found in prior RA cohorts. However, consistent with prior studies, increasing age (HR 1.14, 95% CI 1.09-1.19 per 5 years) and prednisone therapy ≥7.5 mg/day (HR 1.81, 95% CI 1.23-2.67) were associated with a higher HZ risk. Referent to TNFi exposure, PS-stratified analysis showed an HR for csDMARDs of 1.36 (95% CI 0.82-2.25) and for non-TNFi of 0.83 (95% CI 0.51-1.38). CONCLUSION: In the Corrona registry, the HZ risk in RApatients taking prior or current MTX increased with older age and higher prednisone dose. The HZ risk among these patients with RA was comparable after initiation of TNFi versus non-TNFi versus csDMARDs.
Authors: Christien Rondaan; Kornelis S M van der Geest; Elisabeth Eelsing; Annemieke M H Boots; Nicolaas A Bos; Johanna Westra; Elisabeth Brouwer Journal: Front Immunol Date: 2017-10-24 Impact factor: 7.561