Garcia Bañol, Diego FernandoArias Choles, Adrianny MahelisAldana Peréz, SilviaAroca Martínez, Gustavo J.Guido Musso, CarlosNavarro Quiroz, RobertoDominguez Vargas, AlexGonzalez Torres, Henry J.2026-01-282026-01-282025Garcia-Bañol DF, Arias-Choles AM, Aldana-Peréz S, Aroca-Martínez GJ, Musso CG, Navarro-Quiroz R, Dominguez-Vargas A and Gonzalez-Torres HJ (2025) Machine learning in lupus nephritis: bridging prediction models and clinical decision-making towards personalized nephrology. Front. Med. 12:1686057. doi: 10.3389/fmed.2025.16860572296-858X  (Electrónico)https://hdl.handle.net/20.500.12442/17312Lupus nephritis (LN) is one of the most severe manifestations of systemic lupus erythematosus (SLE), affecting up to 65% of patients during the disease (1, 2). Its clinical course is heterogeneous, characterized by alternating periods of exacerbation and remission, and influenced by a complex interplay of immunological, endocrine, genetic, and environmental factors (3–5). Renal involvement ranges from subclinical disease to end-stage renal disease (ESRD), in which a generalized pro-inflammatory state accelerates renal function decline and significantly worsens patient survival (6). There is currently no definitive cure for SLE or LN. Since the 1950s, standard treatment has aimed to induce remission, suppress disease activity, reduce symptoms, preserve renal function, and maintain remission (7). Although therapeutic regimens have evolved over time (induction vs. maintenance strategies), they typically combine an immunosuppressant with an intermediate-acting glucocorticoid to prevent persistent inflammation, irreversible renal damage, and progression to ESRD (8). Multiple factors influence LN progression, including dysregulation of autoantibody production, poor adherence to therapy, excessive sun exposure (9), and socioeconomic disadvantages (10). However, these variables alone have limited predictive value for anticipating disease flares or renal deterioration (5). In this regard, machine learning (ML) algorithms offer the ability to incorporate multiple clinical and biological variables simultaneously, detect hidden patterns, and generate predictive models with greater accuracy (2). The application of ML to LN monitoring provides several potential benefits. These include timely interventions to prevent disease progression and complications (11–15), the development of personalized follow-up strategies based on patient-specific characteristics and trajectories (14–17), and the ability to identify high-risk patients who may require closer surveillance. Moreover, ML models can predict the likelihood of flares by analyzing historical and longitudinal data, enabling clinicians to implement preventive measures such as therapy adjustments or lifestyle modifications.pdfengAttribution-NonCommercial-NoDerivatives 4.0 Internationalinfo:eu-repo/semantics/openAccessinfo:eu-repo/semantics/otherhttps://doi.org/10.3389/fmed.2025.1686057https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1686057/fullLupus nephritisMachine learningArtificial intelligenceDisease progressionPredictive modelsPersonalized nephrology