Prediction of nephrotoxicity associated with cisplatin-based chemotherapy in testicular cancer patients
Menée à partir de données cliniques portant sur 433 patients atteints d'un cancer du testicule et à partir d'échantillons d'ADN extraits de la salive, cette étude évalue la performance d'un algorithme d'apprentissage automatique, basé sur des critères cliniques et la présence de polymorphismes à simple nucléotide, pour prédire la néphrotoxicité d'une chimiothérapie à base de cisplatine
Background : Cisplatin-based chemotherapy may induce nephrotoxicity. This study presents a random forest predictive model that identifies testicular cancer patients at risk of nephrotoxicity before treatment.
Methods : Clinical data and DNA from saliva samples were collected on 433 patients. These were genotyped on Illumina HumanOmniExpressExome-8 v1.2 (964,193 markers). Clinical and genomics-based random forest models generated a risk score for each individual to develop nephrotoxicity defined as a 20% drop in isotopic glomerular filtration rate during chemotherapy. Area under the receiver operating characteristic curve (ROC-AUC) was the primary measure to evaluate models. Sensitivity, specificity and positive and negative predictive values were used to discuss model clinical utility.
Results : Out of 433 patients assessed in this study, 26.8% developed nephrotoxicity after bleomycin-etoposide-cisplatin treatment. Genomic markers found to be associated with nephrotoxicity were located at NAT1, NAT2, and intergenic region of CNTN6 and CNTN4. These in addition to previously associated markers located at ERCC1, ERCC2, SLC22A2, were found to improve predictions in a clinical-feature trained random forest model. Using only clinical data for training the model, a ROC-AUC of 0.635 (95% CI 0.629-0.640) was obtained. Re-training the classifier by adding genomics markers increased performance to 0.731 (95% CI 0.726-0.736), and 0.692 (95% CI 0.688-0.696) on the holdout set.
Conclusions : A clinical and genomics-based machine learning algorithm improved the ability to identify patients at risk of nephrotoxicity compared to using clinical variables alone. Novel genetics associations with cisplatin-induced nephrotoxicity were found for NAT1, NAT2, CNTN6 and CNTN4 that require replication in larger studies before application to clinical practice.
JNCI Cancer Spectrum , Article en libre accès, 2019