2267 - Predictors of Outcome in Peripheral Nervous System Manifestations of Systemic Lupus Erythematosus

Tuesday, November 8, 2011: 9:00 AM-6:00 PM
Hall F2 - Poster Hall (McCormick Place West)
Brandusa Florica1, Jiandong Su2 and Paul R. Fortin1, 1Toronto Western Hospital, Toronto, ON, 2The Toronto Western Hospital, Toronto, ON
Presentation Number: 2267

Background/Purpose: 

Peripheral neuropathies (PN) in patients with systemic lupus erythematosus (SLE) are diverse and may cause significant functional impairment. The identification of factors that influence the outcome of PN is essential in decision regarding the timely treatment of potentially reversible nerve degeneration.

Method: 

Data collected prospectively from the University of Toronto Lupus Clinic cohort was used to identify factors associated with improved neurological deficit (no need for assistance in self care). Potential predictors of improvement in PN were considered in univariate and multivariate logistic regression models. Also, the univariate and multivariate Cox proportional regression models were used to ascertain the results. The analysis included two groups of patients:  those with SLE-related and those with PN not thought to be related to SLE.  The statistical software used was SAS 9.13 

Result:   

From 1533 SLE patients, 184 patients had PN diagnosed within 6 months of their first visit, with 105 patients with SLE-related PN and 79 with non SLE-related PN. Mean duration of follow up after PN diagnosis was 18.8 ± 11.0 years for SLE-related PN and 21.2 ± 11.0 years for non SLE-related PN. Patients with SLE-related PN had significantly higher SLEDAI at time of PN diagnosis (p<0.0001) and used more often steroids in the first year after PN diagnosis (p=0.023). SLE duration at PN onset and SLICC score were positively associated with improved outcome in patients with SLE-related PN, in the univariate logistic regression model, with OR (95%CI) of 0.949(0.909, 0.991) p =0.017 and  of 0.730 (0.546, 0.977). Using multivariable Cox proportional model, the use of steroids in the first year after PN onset was the only factor that was independently associated with improved outcome in SLE-related [HR (95%CI) of 4.181 (1.254, 13.940) p=0.019] and with negative association for non-SLE related PN [HR (95%CI) 0.306 (0.127, 0.739) p=0.008]. Similar results were replicated using the multivariate logistic regression for non-SLE related PN [OR (95%CI) 0.137 (0.035, 0.538) p =0.004].   

Conclusion: 

Clinical and laboratory characteristics of SLE patients at the time of PN diagnostic are not helpful to predict the outcome of the peripheral neurological manifestations. Steroid treatment in the first year after PN onset was associated with more patients improving in SLE related PN group and inversely in non-SLE related PN. Correct delineation of PN etiology is critical for timely intervention on nerve damage.  


Keywords: neurologic involvement, prognostic factors and systemic lupus erythematosus (SLE)

Disclosure: B. Florica, None; J. Su, None; P. R. Fortin, grants, 2, consulting, 5, advisory board members, 6 .