Background/Purpose: RANKL is a cytokine member of the tumor necrosis family that mediates osteoclastic bone resorption.  Denosumab prevents RANKL from activating RANK on the cell surface of osteoclasts, resulting in decreased bone resorption.  RANK/RANKL also plays a role in the immune system, with RANK receptors found on macrophages and dendritic cells; as such, denosumab may interfere with normal immune pathways. Infections (including serious events such as endocarditis) have been reported in clinical trials of denosumab. Data on the potential infectious complications of denosumab in patients with autoimmune disease is lacking, despite higher baseline rates of infection among this group. We sought to identify the rate of infection among patients with autoimmune disease and recent initiation of denosumab.

Methods: We utilized the CORRONA registry to calculate the infection rate among patients with rheumatoid arthritis (RA) and psoriatic arthritis (PsA) following initiation of denosumab. Eligible patients had visit data ≥ 6 months prior to the start of denosumab and at least 1 follow-up visit post drug initiation. Data between 8/2009 and 2/2013 were included. Age, sex, disease activity, prior infection history, comorbid disease, and disease specific medication use (biologic DMARD: bDMARD, non-biologic DMARD: nbDMARD and prednisone) were recorded. Overall incident and serious infections (requiring hospitalization or IV antibiotics) were recorded. Rates of incident and serious infections per 100 person-years were calculated. Cox proportional hazard analysis was performed to identify factors associated with infection risk.

Results: 33,288 patients had ≥ 2 study visits (28,852  RA + 4436 PsA); 96 were eligible for this analysis. Demographic data is presented in Table 1. Data regarding incident and serious infections are presented in Table 2. Prednisone use was associated with an increased incidence of infection (HR 3.73; [1.29-10.77]) in univariate analysis. No other variable demonstrated a significant association.

Table   1. Baseline characteristics among RA and PsA patients on denosumab
	All   Patients N=96	RA   Patients N=90   (93.8)	PsA Patients N=6   (6.3)
Age (years, SD)	73.3±10.7	73.3±10.7	73.2±10.5
Female (%)	94 (97.9)	88 (97.8)	6 (100.0)
Caucasian (%)	88 (91.7)	82 (91.1)	6 (100.0)
Disease Duration (years, SD)	17.0±12.8	16.7±12.3	21.5±20.0
Disease Activity CDAI (mean, SD) Low Disease activity1 (%)	7.7±8.4 71 (74.0)	7.8±8.5 67 (74.0)	6.7±7.5 4 (66.7)
Drug Therapy Denosumab alone (%) nbDMARD2 (%) bDMARD3 ± nbDMARD (%)	7 (7.3) 46 (47.9) 43 (44.8)	7 (7.8) 42 (46.7) 41 (45.5)	0 (0.0) 4 (66.7) 2 (33.3)
Prednisone Use Current (%) Prior (%)	30 (31.3) 56 (58.3)	28 (31.1) 51 (56.7)	2 (33.3) 5 (83.3)
Prior Serious Infections (%)	10 (10.4)	10 (11.1)	0 (0.0)
Abbreviations: SE: standard error. CDAI: clinical disease activity index 1  Low disease activity defined as CDAI ≤ 10 2 nbDMARD: leflunomide, methotrexate, cyclosporine, azathioprine 3   bDMARD: certolizumab, anakinra, etanercept, adalimumab, golimumab, infliximab,   abatacept, rituximab, or tocilizumab

 

Table   2. Infection Rates among Patients on Denosumab*
	Events   (%)	Person-Years At Risk	Infection Rate per 100 Person-Years** (95% CI)
Incident Infections+ (N=96) Denosumab alone (N=7) nbDMARD (N=46) bDMARD (N=10) bDMARD and nbDMARD (N=33)	14 (14.6) 0 (0) 6 (13.0) 1(10.0) 7 (21.2)	58.9	24.0   (14-40) NC NC NC NC
Serious   Infections++ (N=96)	3 (3.1)	58.9	5   (2-16)
Incident infections: septic arthritis/bursitis, cellulitis, sinusitis, diverticulitis, sepsis, pneumonia bronchitis, gastroenteritis, meningitis, UTI, URI or other specified site *Table includes 90 patients with RA and 6 with PsA. Among the PsA cohort, 2 total infections (1 serious) were recorded **Historical rates of infection among RA patients in CORRONA (Au et. al; Ann Rheum Dis 2011;70:785–791): 31.2 outpatient infections per 100 person-yrs; 0.8 hospitalized infections per 100 person-yrs + Kaplan-Meier estimate of incident infection 3 months post denosumab start: 2.5% (95% CI 0.6-9.3); at 6 months: 12.0% (95% CI 6.4-22) ++ Kaplan-Meier estimate of serious infection 3 months post denosumab start: 1.2% (95% CI 0.1-8.3); at 6 months 2.5% (95% CI 0.6-9.7)

Conclusion: The infection rate appears to be low among patients on denosumab and concurrent DMARD therapy. Compared to historical incidence rates for infection within the larger CORRONA database (and other historical cohorts), the rates observed herein appear comparable. Univariate analysis was limited by cohort size, but did not identify an association with nbDMARD or bDMARD use and subsequent infection. Additional study will be necessary to further delineate the risk for infection in patients exposed to denosumab.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/infection-risk-among-patients-receiving-concurrent-denosumab-and-biologic-or-non-biologic-dmard-therapy-an-analysis-of-the-consortium-of-rheumatology-researchers-of-north-america-corrona-registry/