Background/Purpose: There is little information on Systemic Lupus Erythematosus (SLE) patients undergoing total joint arthroplasty (TJA).  The purposes of this study were to determine the patterns of TJA in a contemporary SLE cohort and to determine if SLE is a risk factor for worse outcomes compared with similar osteoarthritis (OA) patients.

Methods:   Patients with SLE who underwent hip or knee arthroplasty between May 2007 and June 2011 and enrolled in our institution’s arthroplasty registry were eligible for this study.  SLE cases were identified by ICD-9 code 710 and confirmed if they met 3/14 ACR criteria, were on immunosuppressant therapy, or had the diagnosis confirmed by a rheumatologist. Validated SLE cases were matched to two OA controls by age (+/- 2.5 years), gender, procedure, and presence of osteonecrosis (ON).  ON was confirmed in cases and controls by pathology or radiology review.  Pain and function were measured by WOMAC and the validated Lower Extremity Activity Scale (LEAS). Administrative and self-report data were collected at baseline and 2 years. Standard univariate comparisons were performed to compare SLE with matched OA cases at baseline and at two years post-op.  Multivariate regressions were performed to analyze the relationship between diagnosis, presence of ON, and WOMAC pain and function at 2 years.  TKA patients’ pre-operative expectations were measured with the validated TKR Expectations Survey.

Results: 101 SLE cases were identified, 56 hip (THR) and 45 knee (TKR).  5 cases could only be matched to 1 control.  Pre-operatively, SLE THR cases had statistically and clinically significantly worse WOMAC pain, stiffness, and function than matched OA hips, (see Table).  Renal failure, hypertension, pulmonary, and valvular disease were also more common in SLE THA patients.  Both SLE THA and TKA has statistically significantly worse SF-36 PCS pre-operatively, and, despite significant improvements, they remained statistically significantly worse compared to matched OA controls post-operatively. Two-years post-operatively, there were no differences in pain and function scores between SLE and OA controls.  In multivariate regressions controlling for type of surgery, disease type, and ON, neither SLE nor ON predicted worse pain or function at 2 years.  There were no differences in expectations of surgery between SLE TKA and OA TKA.

Conclusion: SLE THA and TKA patients have similar pain and functional outcomes at 2 years compared with matched OA controls.   Although PCS scores improved after arthroplasty, they remained lower in SLE patients than OA controls.  To our knowledge, this is the first study to demonstrate that neither SLE nor ON should be considered risk factors for poor post-operative outcomes.