Efficacy of the Lupus-Cruces Nephritis protocol in different prognostic groups: a propensity score and cluster analysis.

Abstract

To assess the efficacy of the Lupus-Cruces Nephritis (LCN) protocol vs standard of care (SOC) in achieving complete vrenal response (CRR) among different subpopulations and clinical clusters of patients with LN. This was an observational comparative study with clinical care data of patients from the Lupus-Cruces, historic Lupus-Cruces and Lupus-Bordeaux cohorts. Different prognostic factors were obtained from previous studies and analysed in both LCN and SOC groups. Three different clinical clusters of patients were identified using a two-way clustering analysis. A multivariate propensity score (PS) adjusted analysis and Cox proportional hazards analysis were performed comparing both therapeutic schemes in the different identified subpopulations and clusters. Within the cohort of 147 patients, those receiving the LCN protocol (40/47, 85%) were more likely to achieve CRR at 12 months than patients receiving the SOC (44/100, 44%). This was consistent across all prognostic subpopulations. The PS-adjusted logistic regression by subgroups confirmed these results. Three different clusters (named C1, C2 and C3) were identified. Both the univariate and the PS-adjusted Cox regression analysis for CRR at any time confirmed the higher likelihood of achieving CRR of patients treated with the LCN protocol in the three clusters: C1, hazard ratio (HR) 10.7 (95% CI 1.65-69.6, P = 0.013); C2, HR 2.2 (95% CI 1.16-4.1, P = 0.016); C3, HR 6.1 (95% CI 2.99-12.51, P < 0.001). The LCN protocol was superior to the SOC schemes in achieving CRR in patients with LN across different subpopulations and clinical clusters.