Cluster analysis of antinuclear autoantibodies in the prognosis of SLE nephropathy: are anti-extractable nuclear antibodies protective?

To investigate the possible role of anti-ENA autoantibodies in the pathogenesis of SLE nephropathy, we performed a cross sectional clustering study of 91 SLE patients using 75 clinical and laboratory variables examining the presence of anti-dsDNA and ENA autoantibodies by ELISA and Western blot. We applied principal component, hierarchical cluster, multiple correspondence and logistical regression analysis. Two polar forms of SLE nephropathy and five clinical groups were identified: group 1 without overt nephropathy (n = 37), group 2 with nephropathy and only proteinuria (n = 19), group 3 nephropathy and only hematuria (n = 11), group 4 with hematuria and proteinuria (n = 14) and group 5 on renal failure (n = 10). When analyzed individually, levels of anti-dsDNA and single anti-ENA antibodies did not allow us to differentiate between renal and non-renal groups. However, when the anti-ENA autoantibodies were analyzed as a cluster, a high predictive value for clinical nephropathy was obtained. Thus, the absence of ENA antibodies (ENA ve or Venezuelan cluster) increased eleven-fold the odds ratio to develop SLE nephropathy. We suggested that the ENA ve cluster may predict development of the most severe forms of renal lupus while the ENA Sm/RNP and the ENA Ro/La/Sm/RNP clusters could be associated with the absence and the most benign form of SLE nephropathy. It must be interesting to apply similar cluster methodology in an SLE population with different ethnic background.