Early Prediction of Tacrolimus-Induced Tubular Toxicity in Pediatric Refractory Nephrotic Syndrome Using Machine Learning

Abstract

Background and Aims: Tacrolimus(TAC)-induced nephrotoxicity, which has a large individual variation, may lead to treatment failure or even the end-stage renal disease. However, there is still a lack of effective models for the early prediction of TAC-induced nephrotoxicity, especially in nephrotic syndrome(NS). We aimed to develop and validate a predictive model of TAC-induced tubular toxicity in children with NS using machine learning based on comprehensive clinical and genetic variables.Materials and Methods: A retrospective cohort of 218 children with NS admitted between June 2013 and December 2018 was used to establish the models, and 11 children were prospectively enrolled for external validation. We screened 47 clinical features and 244 genetic variables. The changes in urine N- acetyl- β-D- glucosaminidase(NAG) levels before and after administration was used as an indicator of renal tubular toxicity.Results: Five machine learning algorithms, including extreme gradient boosting (XGBoost), gradient boosting decision tree (GBDT), extremely random trees (ET), random forest (RF), and logistic regression (LR) were used for model generation and validation. Four genetic variables, including TRPC6 rs3824934_GG, HSD11B1 rs846910_AG, MAP2K6 rs17823202_GG, and SCARB2 rs6823680_CC were incorporated into the final model. The XGBoost model has the best performance: sensitivity 75%, specificity 77.8%, accuracy 77.3%, and AUC 78.9%.Conclusion: A pre-administration model with good performance for predicting TAC-induced nephrotoxicity in NS was developed and validated using machine learning based on genetic factors. Physicians can estimate the possibility of nephrotoxicity in NS patients using this simple and accurate model to optimize treatment regimen before administration or to intervene in time after administration to avoid kidney damage.

Development of a predictive model for nephrotoxicity during tacrolimus treatment using machine learning methods

“…Mo et al developed a model to predict TACinduced nephrotoxicity using genetic variables. 33 For more accurate predictions, it may be necessary to consider genetic features. However, the features required in our model are available in daily medical practice, and it is remarkable that patients do not require additional examinations.…”

“…This was because genetic information was not available in the medical records. Mo et al developed a model to predict TAC‐induced nephrotoxicity using genetic variables 33 . For more accurate predictions, it may be necessary to consider genetic features.…”

Early prediction of growth patterns after pediatric kidney transplantation based on height-related single-nucleotide polymorphisms

“…This finding is consistent with the results of previous studies showing that optimized machine learning algorithms have better prediction efficacy than traditional statistical methods. [13,44] The variables in the final model included age and 19 SNPs corresponding to the genes NUCB2, PSMC5, PCSK1, SRRM2, etc. We calculated that the patients with genotypes NUCB2 rs757081_GG, PCSK1 rs6235_GG, SRRM2 rs3094775_GG, PSMC5 rs13030_TT, etc.…”

“…[ 11 ] Paré et al [ 12 ] developed a machine learning model based on GWAS, which showed significant efficacy at predicting human height. Mo et al [ 13 ] established an SNP model for early prediction of tacrolimus-induced renal tubular toxicity in pediatric refractory nephrotic syndrome through comparison of various machine learning algorithms. These models can provide scientific and individualized decision-making for clinical management.…”

Predicting the effect of sirolimus on disease activity in patients with systemic lupus erythematosus using machine learning

“…Machine learning was an important branch of computer science and artificial intelligence, which could learn from data based on computational modelling and be applied to complex medical problems with better behaviour than traditional statistical analysis, especially when analysing big medical data. 25,26 So far, machine learning was widely used in medical researches, among which Mo et al 27 reported early prediction of tacrolimus-induced tubular toxicity in paediatric refractory nephrotic syndrome using machine learning. Casiraghi et al 28 reported explainable machine learning for early assessment of COVID-19 risk prediction in emergency departments.…”