Kidney Week

Abstract: SA-OR063

Using Long Term Outcome Data to Redefine Proteinuria and eGFR End Points for Lupus Nephritis Trials

Session Information

• IgA Nephropathy and Lupus Nephritis
  October 27, 2018 | Location: 6E, San Diego Convention Center
  Abstract Time: 05:30 PM - 05:42 PM

Category: Glomerular Diseases

• 1203 Glomerular Diseases: Clinical, Outcomes, and Trials

Authors

• Wilson, Hannah R., Imperial College London, Hammersmith Hospital, London, United Kingdom

Hannah R. Wilson, MBBS

• Turner-Stokes, Tabitha, Imperial College London, Hammersmith Hospital, London, United Kingdom

Tabitha Turner-Stokes,

• Dasilva Santos, Iara, Fundacio Puigvert, Barcelona, Spain

Iara Dasilva Santos,

• Griffith, Megan, Imperial College London, Hammersmith Hospital, London, United Kingdom

Megan Griffith,

• Levy, Jeremy B., Imperial College London, Hammersmith Hospital, London, United Kingdom

Jeremy B. Levy,

• Cairns, Tom, Imperial College London, Hammersmith Hospital, London, United Kingdom

Tom Cairns,

• Cook, H. Terence, Imperial College London, Hammersmith Hospital, London, United Kingdom

H. Terence Cook,

• Lightstone, Liz, Imperial College London, Hammersmith Hospital, London, United Kingdom

Liz Lightstone,

Background

Lupus Nephritis (LN) is a common & severe manifestation of systemic lupus erythematosus (SLE) that can lead to renal failure (ESRF). Reviews of Euro-Lupus Nephritis & MAINTAIN cohorts suggest a less stringent cut off proteinuria at 1yr (<0.8g/d & <0.7g/d respectively) better predicted good renal outcome at 7yrs than usual complete remission (CR) criteria (<0.5g/d). This cut-off has yet to be validated using urinary protein creatinine ratio (uPCR) or in a larger real world cohort.

Methods

Data were reviewed for LN biopsies 1/1/1996 to 1/1/2016. Definition CR:uPCR (mg/mmol) <50 & estimated glomerular filtration rate (eGFR mls/min/1.73m²) ≥60, or if b/l <60, no fall >20%; Partial remission (PR):uPCR <300 with ≥50% improvement & eGFR as CR; Non-remission (NR):no PR by 1yr. Factors predicting good outcome, defined as 1-46yr survival with eGFR >60, assessed by multiple logistic regression (LR) with correction & receiver operating characteristic (ROC) curves.

Results

476 patients had 819 biopsies. Median age diagnosis:SLE 29yrs; LN 32yrs. Female:82%. Ethnicity:32% S Asian, 27% Black, 26% White& 3% SE Asian.
At latest f/up since diagnosis LN (median 9yrs(0-46): majority, 293(62%) had good outcome; ESRF: 68(14%), median 5yrs(0-43); died: 31(7%), 7yrs(0-19).
LR identified 1yr eGFR and uPCR as predictors of good outcome: eGFR odds ratio (OR) 2.186 (95% CI 1.529-3.126) p<0.001 for each rise 10; uPCR OR 0.547 (0.356-0.841) p=0.006 for increase between ranges (0-50,50-100,100-300,300-500,500-1000,>1000).
ROC curves identified 1yr uPCR <68 (AUC0.70,p<0.0001,sens 67%,spec 65%) & eGFR >80.5 (AUC 0.83,p<0.0001,sens 77%,spec 76%) predictive of good outcome.
Standard CR&NR definitions fail to predict good outcome; in contrast, achieving uPCR & eGFR identified by ROC curves strongly predicted good outcome OR 5.68 (95% CI 2.367-13.635, p<0.001); OR good outcome if not achieved 0.42 (95% CI 0.188-0.926, p=0.032).

Conclusion

Our “real world” multi-ethnic cohort, the largest to date, support Eurolupus and MAINTAIN data: less stringent proteinuria thresholds (<68mg/mmol) and excellent renal function (>80.5ml/min/m²) at 1yr better predict long term outcome than current trial endpoints. As the key outcome that matters to patients, we argue it is time to change definitions of success of treatments in LN trials.