Lupus nephritis study finds proteinuria drop at 12 months best predicts survival

For patients whose immune systems turn against their own kidneys, the single most important question is deceptively simple: how will they respond to treatment in the first year? A new 28-year cohort study from Romania now offers one of the clearest answers yet. Researchers at Fundeni Clinical Institute in Bucharest followed 147 patients with biopsy-proven lupus nephritis and found that dropping urinary protein below 0.7 grams per day within 12 months was the strongest predictor of long-term kidney and patient survival — more powerful than any baseline measurement taken at the time of diagnosis.

Lupus nephritis is the kidney manifestation of systemic lupus erythematosus, an autoimmune disease in which immune complexes deposit in the glomeruli, the microscopic filters of the kidney, triggering inflammation, scarring, and progressive loss of renal function. It affects up to 60 percent of lupus patients. Despite decades of therapeutic progress, outcomes remain stubbornly suboptimal: in developed countries, up to a third of patients with the most aggressive proliferative form of the disease still progress to end-stage kidney disease within 10 years of diagnosis. Once patients reach dialysis or transplant, their mortality risk rises 26-fold compared with lupus patients whose kidneys survive.

The study, published in Frontiers in Nephrology, retrospectively analyzed all patients with systemic lupus erythematosus and a first biopsy-proven episode of lupus nephritis followed at the Fundeni Nephrology Department between 1997 and 2025. Kidney biopsies were classified according to the International Society of Nephrology/Renal Pathology Society criteria and scored using the modified National Institutes of Health activity and chronicity indices, which quantify reversible inflammatory damage and permanent scarring, respectively. The researchers tracked a composite outcome — doubling of serum creatinine, progression to end-stage kidney disease, or death from any cause — over a median follow-up of 39.5 months.

The cohort was typical of lupus nephritis populations: a median age of 33 years, 80.3 percent female, with a median time of two years between lupus diagnosis and kidney involvement. Most patients presented with heavy proteinuria — a median of 1.98 grams per day — and nearly 38 percent exceeded 3 grams daily, the nephrotic range. Histologically, 68 percent had proliferative class III or IV disease, the most damaging forms. Induction immunosuppression relied mainly on cyclophosphamide, used in 40 percent of patients, followed by mycophenolate mofetil, with treatment patterns shifting noticeably after 2020 toward Euro-Lupus protocols, combination regimens, and increased rituximab and corticosteroid use.

Over follow-up, 10.3 percent of patients experienced a doubling of serum creatinine, 11 percent progressed to end-stage kidney disease, and 6.8 percent died, with 17.1 percent reaching the combined endpoint. Combined kidney and general survival stood at 91 percent at one year, 88 percent at two years, 82 percent at five years, and 69 percent at ten years. Patients with heavy baseline proteinuria fared worst: among those exceeding 3 grams per day, 31.4 percent reached the combined event, compared with 19.2 percent of patients with proliferative disease overall.

Yet the most striking finding concerned treatment response rather than initial severity. Roughly half of the cohort — 49.1 percent at 6 months and 54.4 percent at 12 months — achieved proteinuria below 0.7 grams per day, the threshold used to define renal response in pivotal trials such as BLISS-LN. In multivariate Cox regression, baseline predictors of the composite outcome included age over 50 years, proteinuria of 3 grams per day or more, and hemoglobin below 10 grams per deciliter. But when 12-month proteinuria was added to the model, baseline proteinuria lost its statistical significance entirely, suggesting that the trajectory of protein loss matters far more than its starting point.

Achieving proteinuria below 0.7 grams per day at 12 months emerged as a strong protective factor, with a hazard ratio of 0.14 in the entire cohort — an 86 percent reduction in the risk of the combined event — and an even more dramatic 0.09 among patients with proliferative lupus nephritis. The finding echoes the post hoc analyses of the Euro-Lupus and MAINTAIN Nephritis trials, where a proteinuria cut-off of 0.7 to 0.8 grams per day at 12 months proved the best short-term predictor of good long-term renal outcome, with sensitivities of 71 to 81 percent and positive predictive values reaching 88 to 94 percent.

The study also yielded a seemingly paradoxical histological insight. Patients with a high activity index — a score of 9 or above, indicating heavy reversible inflammation — showed a protective trend rather than increased risk, while those with a high chronicity index, marking permanent scarring, had significantly worse survival. The explanation lies in the biology of the lesions: active inflammatory damage is potentially reversible with aggressive immunosuppression, whereas chronic lesions — interstitial fibrosis and tubular atrophy — represent irreversible injury that portends progressive renal decline. Chronicity, rather than activity, appears to drive long-term kidney loss.

Complement levels added another layer of complexity. Low baseline C3 tended to predict adverse outcomes, consistent with its role as a marker of sustained alternative-pathway overactivation, while low C4 showed a protective trend — an observation never reported before in lupus nephritis cohorts. The researchers speculate that C4 hypocomplementemia may reflect heightened classical-pathway immunologic activity that responds well to immunosuppression, rather than irreversible tissue damage, though they caution this warrants further investigation.

Building on these predictors, the team constructed a prognostic score assigning one point for each risk factor and subtracting one for each protective factor, stratifying patients into low, intermediate, and high risk categories. Using only baseline variables, the score achieved an area under the curve of 0.82; adding 12-month proteinuria raised it to 0.86 and made the differences between risk categories statistically significant. The authors emphasize that the score requires external validation before clinical use, and acknowledge limitations including the retrospective single-center design, treatment heterogeneity over nearly three decades, and the exclusion of comorbidities such as hypertension and diabetes. Still, the message is clear: in lupus nephritis, prognosis should be dynamic, reassessed at one year based on treatment response — because how the kidney responds matters more than how sick it was at the start.

Subject of Research: Predictors of renal and patient survival in lupus nephritis: a 28-year single-center cohort study

Article Title: Predictors of renal and patient survival in lupus nephritis: a 28-year single-center cohort study

Article References: Vrabie, A., Obrişcă, B., Sorohan, B., Jurubiţă, R., Achim, C., Zilişteanu, D. S., Andronesi, A. G., Lupuşoru, G., Rusu, E. E., Bobeică, R., Micu, E. G., Bălănică, S., Berechet, A., Mocanu, V., Lujinschi, Ş., Caceaune, N., & Ismail, G. (2026). Predictors of renal and patient survival in lupus nephritis: a 28-year single-center cohort study. Frontiers in Nephrology, 6, Article 1945616. https://doi.org/10.3389/fneph.2026.1945616

Image Credits: AI Generated

DOI: 10.3389/fneph.2026.1945616

Keywords: lupus nephritis, proteinuria, renal survival, prognostic score, lupus, kidney disease, Predictors, renal, patient, survival, nephritis, year