A Longitudinal Analysis of Outcomes of Lupus Nephritis in an International Inception Cohort Using a Multistate Model Approach

John G Hanly, Li Su, Murray B Urowitz, Juanita Romero-Diaz, Caroline Gordon, Sang-Cheol Bae, Sasha Bernatsky, Ann E Clarke, Daniel J Wallace, Joan T Merrill, David A Isenberg, Anisur Rahman, Ellen M Ginzler, Michelle Petri, Ian N Bruce, M A Dooley, Paul Fortin, Dafna D Gladman, Jorge Sanchez-Guerrero, Kristjan Steinsson, Rosalind Ramsey-Goldman, Munther A Khamashta, Cynthia Aranow, Graciela S Alarcón, Barri J Fessler, Susan Manzi, Ola Nived, Gunnar K Sturfelt, Asad A Zoma, Ronald F van Vollenhoven, Manuel Ramos-Casals, Guillermo Ruiz-Irastorza, S Sam Lim, Kenneth C Kalunian, Murat Inanc, Diane L Kamen, Christine A Peschken, Soren Jacobsen, Anca Askanase, Chris Theriault, Vernon Farewell

To study bidirectional change and predictors of change in estimated glomerular filtration rate (eGFR) and proteinuria (ePrU) in lupus nephritis (LN) using a multistate modelling approach. In the SLICC inception cohort we determined annual eGFR state 1 (eGFR: >60 ml/min), 2 (eGFR: 30-60 mL/min), and 3 (eGFR: <30 ml/min); ePrU state 1 (ePrU: <0.25 gr/day), 2 (ePrU: 0.25-3.0 gr/day), and 3 (ePrU: >3.0 gr/day); End stage renal disease (ESRD) and death. Using multistate modelling, relative transition rates between states indicated improvement and deterioration. In 700/1,826 (38.3%) patients with LN, and mean (SD followup 5.2(3.5) years the likelihood of improvement in eGFR and ePrU was greater than deterioration. After 5 years, estimated transition to ESRD was 62% of patients initially in eGFR state 3 and 11% from ePrU state 3. The probability of remaining in initial eGFR states 1, 2 and 3 was 85%, 11%, 3% and for ePrU was 62%, 29%, 4%. Male sex predicted improvement in eGFR states; older age, race/ethnicity, higher ePrU state and higher renal biopsy chronicity scores predicted deterioration. For ePrU, race/ethnicity, earlier calendar years, damage scores without renal variables and higher renal biopsy chronicity scores predicted deterioration; male sex, positive lupus anticoagulant, class V nephritis and mycophenolic acid use predicted less improvement. In LN, the expected improvement and deterioration in renal outcomes can be estimated by multistate modelling and is predicated by identifiable risk factors. New therapeutic interventions for LN should meet or exceed these expectations. This article is protected by copyright. All rights reserved.