ABSTRACT
Proliferative lupus nephritis (PLN) is a serious organ-threatening manifestation of systemic lupus erythematosus (SLE) that is associated with high mortality and renal failure. Here, we analyzed data from 1287 SLE patients with renal manifestations, including 780 of which were confirmed as proliferative or non-proliferative LN patients by renal biopsy, divided into a training cohort (547 patients) and a validation cohort (233 patients). By applying a least absolute shrinkage and selection operator (LASSO) regression approach combined with multivariate logistic regression analysis to build a nomogram for prediction of PLN that was then assessed by receiver operating characteristic (ROC) curves, calibration curves, and clinical decision curves (DCA) in both the training and validation cohorts. The area under the ROC curve (AUC) of the model in the training cohort was 0.921 (95% confidence interval (CI): 0.895-0.946), the AUC of internal validation in the training cohort was 0.909 and the AUC of external validation was 0.848 (95% CI: 0.796-0.900). The nomogram showed good performance as evaluated using calibration and DCA curves. Taken together, our results indicate that our nomogram that comprises 12 significantly relevant variables could be clinically valuable to prognosticate on the risk of PLN in SLE, so as to improve patient prognoses.
Subject(s)
Lupus Erythematosus, Systemic , Lupus Nephritis , Nomograms , Humans , Female , Male , Adult , Lupus Erythematosus, Systemic/complications , Kidney/pathology , ROC Curve , Middle Aged , Prognosis , Young Adult , Cohort Studies , Risk FactorsABSTRACT
OBJECTIVES: To compare proliferative (PLN) and membranous (MLN) lupus nephritis (LN) regarding clinical and laboratory presentation and long-term outcomes; To investigate predictors of progression to chronic kidney disease (CKD). METHODS: Multicentre observational study, with retrospective analysis of a prospective cohort, using data from the Rheumatic Diseases Portuguese Registry-Reuma.pt. Patients with biopsy-proven PLN, MLN and mixed LN were included. Cox regression survival analysis was used to investigate predictors of CKD. RESULTS: 260 patients were included. Median follow-up was 8 years (IQR 11; minimum 1, maximum 35 years). MLN patients presented with significantly lower serum creatinine (0.70 (IQR 0.20; minimum 0.50, maximum 1.30) mg/dl vs 0.80 (IQR 0.31; minimum 0.26, maximum 2.60) in PLN, p= 0.003). Proteinuria levels did not differ between groups (p= 0.641). Levels of complement were reduced in PLN but nearly normal in MLN patients, and there were fewer patients with positive anti-dsDNA antibodies in the MLN group (p< 0.001). One year after the beginning of treatment, 62% of the patients achieved EULAR/ERA-EDTA complete response, with further 5% achieving partial response. Patients with lower proteinuria at diagnosis were more likely to achieve a complete renal response at one year, however, proteinuria at diagnosis or at one year did not predict long term CKD. Estimated glomerular filtration rate (eGFR) ≤75 mL/min/1.73 m2 at one year was the strongest predictor of progression to CKD (HR 23 [95% CI 8-62], p< 0.001). Other possible predictors included the use of azathioprine for induction of remission, older age at diagnosis and male sex. CONCLUSION: Proteinuria levels did not predict LN histologic class in our cohort. eGFR cutoff of 75 mL/min/1.73 m2 after one year of treatment was strongly predictive of progression to CKD.
ABSTRACT
Objective: This study aims to develop and validate machine learning models to predict proliferative lupus nephritis (PLN) occurrence, offering a reliable diagnostic alternative when renal biopsy is not feasible or safe. Methods: This study retrospectively analyzed clinical and laboratory data from patients diagnosed with SLE and renal involvement who underwent renal biopsy at West China Hospital of Sichuan University between 2011 and 2021. We randomly assigned 70% of the patients to a training cohort and the remaining 30% to a test cohort. Various machine learning models were constructed on the training cohort, including generalized linear models (e.g., logistic regression, least absolute shrinkage and selection operator, ridge regression, and elastic net), support vector machines (linear and radial basis kernel functions), and decision tree models (e.g., classical decision tree, conditional inference tree, and random forest). Diagnostic performance was evaluated using ROC curves, calibration curves, and DCA for both cohorts. Furthermore, different machine learning models were compared to identify key and shared features, aiming to screen for potential PLN diagnostic markers. Results: Involving 1312 LN patients, with 780 PLN/NPLN cases analyzed. They were randomly divided into a training group (547 cases) and a testing group (233 cases). we developed nine machine learning models in the training group. Seven models demonstrated excellent discriminatory abilities in the testing cohort, random forest model showed the highest discriminatory ability (AUC: 0.880, 95% confidence interval(CI): 0.835-0.926). Logistic regression had the best calibration, while random forest exhibited the greatest clinical net benefit. By comparing features across various models, we confirmed the efficacy of traditional indicators like anti-dsDNA antibodies, complement levels, serum creatinine, and urinary red and white blood cells in predicting and distinguishing PLN. Additionally, we uncovered the potential value of previously controversial or underutilized indicators such as serum chloride, neutrophil percentage, serum cystatin C, hematocrit, urinary pH, blood routine red blood cells, and immunoglobulin M in predicting PLN. Conclusion: This study provides a comprehensive perspective on incorporating a broader range of biomarkers for diagnosing and predicting PLN. Additionally, it offers an ideal non-invasive diagnostic tool for SLE patients unable to undergo renal biopsy.
Subject(s)
Lupus Nephritis , Machine Learning , Humans , Lupus Nephritis/diagnosis , Lupus Nephritis/pathology , Female , Male , Adult , Retrospective Studies , Middle Aged , Biomarkers , Young AdultABSTRACT
INTRODUCCIÓN: La nefritis lúpica proliferativa (NPL) con proliferación extra capilar focal (PECF) es una entidad frecuente en las biopsias renales, pero poco se conoce acerca de su evolución. Este estudio se realizó para investigar la correlación clínica, histológica y de laboratorio en la evolución de estos pacientes. OBJETIVO: Analizar las biopsias de nefritis lúpicas proliferativas focales o difusas (clases III y IV de la OMS), y su evolución de acuerdo al grado de proliferación extracapilar en un período de 10 años, en el Hospital de Clínicas "José de San Martín". MATERIAL Y MÉTODOS: Estudio observacional retrospectivo, se incluyeron 147 pacientes con GN clase III o IV, estudiados en un período de 10 años. Se subdividieron en base al compromiso extracapilar en: grupo 1 (5- <25% de semilunas, n=33), grupo 2 (25- 50% semilunas, n=34) y grupo 3 (>50% semilunas, n= 20) o proliferación extracapilar difusa (PECD). Los pacientes con nefritis lúpica III y IV, sin compromiso extra capilar, se denominó grupo control (n=60); las características clínicas, laboratorios y evolución en cuanto alcanzar remisión completa al año fueron comparadas entre los subgrupos y el grupo de control. Remisión completa se definió según normas KDIGO 2012. Se realizó un análisis univariado y posteriormente un multivariado, ambos con regresión logística. RESULTADOS: Hubo 124 mujeres (84,35%) y 23 hombres (15,65%) con una edad media de 35±10 años. Edad media en el grupo control, 32±10 años. El univariado mostro mayor remisión completa en el grupo control con odds ratio 0.05 (0.02.-0.12) p=0.001, y menor remisión completa en los 3 subgrupos con proliferación extra capilar, inclusive el grupo I con un odds ratio 7.61 (2,42-23.9) y valor de p=0.001 en dicho grupo. El análisis multivariado evidenció que la persistencia de la proteinuria a los 6 meses tenía significación estadística para no alcanzar la remisión completa, con un odd ratio 2.81 (1.58-5.0), valor de p=0.001. CONCLUSIÓN: La presencia de distintos grados de semilunas en la histología de las NL proliferativas (NPL), clase III y IV, y la persistencia de la proteinuria a los 6 meses, son marcadores independiente para no alcanzar la remisión completa al año
IINTRODUCTION: Although focal lupus nephritis (FLN) with extracapillary proliferation (ECP) is usually found in renal biopsies, little is known about its course. This study was carried out to investigate the correlation of clinical, histological and laboratory findings in these patients' progress. OBJECTIVE: To analyze focal or diffuse proliferative lupus nephritis biopsies (class III and IV, according to WHO) and their development regarding the level of extracapillary proliferation throughout a period of 10 years at Hospital de Clínicas 'José de San Martín'. METHODS: In this retrospective observational study, 147 patients with class III or IV GN were observed for 10 years. According to the degree of extracapillary proliferation, they were divided into the following groups: group 1 (5% <25% of crescents, n=33), group 2 (25% 50% of crescents, n=34) and group 3 [>50% of crescents, i.e. diffuse proliferative nephritis (DPN), n=20]. Patients suffering from class III and IV lupus nephritis with no extracapillary proliferation formed the control group (n=60). After one year, clinical and laboratory findings as well as the development of the condition among the four groups were compared to assess their level of remission. Complete remission was considered as defined in KDIGO 2012 Guideline. A univariate logistic regression analysis and a subsequent multivariate one were performed. RESULTS: There were 124 women (84.35%) and 23 men (15.65%) with an average age of 35±10. The average age in the control group was 32±10. The univariate analysis showed a greater number of patients who achieved complete remission in the control group with an odds ratio of 0.05 (0.02 0.12) p=0.001, whereas in the 3 groups with extracapillary proliferation fewer patients reached this state, even in group 1, which had an odds ratio of 7.61 (2.42 23.9) and p=0.001. The multivariate analysis revealed that the persistent protein in the urine after 6 months was statistically significant for not achieving complete remission, the odds ratio being 2.81 (1.58 5.0) and p=0.001. CONCLUSION: The presence of various degrees of crescent formation in class III and IV proliferative lupus nephritis (PLN) shown in the histological findings and the persistent protein in the urine at 6 months constitute independent factors for not achieving complete remission after one year