INVESTIGADORES
MONGE Maria Eugenia
congresos y reuniones científicas
Título:
A Coupled Lipidomics-Machine Learning Approach for Early Diagnosis of clear cell Renal Cell Carcinoma
Autor/es:
MARÍA EUGENIA MONGE; MALENA MANZI; MARTÍN PALAZZO; NICOLÁS ZABALEGUI; MARÍA ELENA KNOTT; PIERRE BEAUSEROY; PATRICIO YANKILEVICH; MARÍA ISABEL GIMÉNEZ
Lugar:
La Haya
Reunión:
Conferencia; Metabolomics 2019; 2019
Institución organizadora:
The Metabolomics Society
Resumen:
Clear cell (ccRCC) is the most common (75%) lethal subtype of RCC, and is considered a glycolytic and lipogenic tumor. More than 30% of patients, often incidentally diagnosed by imaging procedures, exhibit locally advanced or metastatic RCC at the time of diagnosis and the disease is inherently resistant to chemotherapy and radiotherapy. In this work, serum samples from a cohort that included patients with different ccRCC stages (SI, SII, SIII and, SIV; n=112) and controls (n=52) were interrogated with a discovery-based lipidomics approach using ultraperformance liquid chromatography coupled to quadrupole-time-of-flight mass spectrometry and machine learning methods. Support vector machine models and the least absolute shrinkage and selection operator (Lasso) variable selection method yielded two discriminant panels for ccRCC detection and early diagnosis. A first 18-feature panel allowed discriminating ccRCC patients from controls with 96.8% accuracy in a training set under cross-validation, and 81.4% accuracy in an independent test set. Fifteen features of the panel were significantly decreased in ccRCC after correcting for multiple testing. A second model was trained to discriminate early stages (I-II) from late stages (III-IV) ccRCC and provided a 26-feature panel that allowed sample classification with 84.5% accuracy in the training set under cross-validation, and 82.1% accuracy in the classification of stage I ccRCC patients from an independent test set. Current work involves feature identification in both panels by accurate mass, isotopic pattern, tandem MS experiments, and chemical standard analysis. Results are auspicious for early ccRCC diagnosis after validation of the panels in larger and different cohorts.
