Extent of brain tumor resection continues to be one of the central decisions taken during standard of care in glioma patients. Here, we aimed to evaluate the most essential molecular factors, such as IDH (isocitrate dehydrogenase) mutation in gliomas classification with patient-derived glioma organoids (PGOs) using differential mobility spectrometry (DMS). we prospectively recruited 12 glioma patients, 6 IDH-mutated and 6 IDH wild-type tumors, from which PGOs were generated ex-vivo. Altogether, 320 PGOs DMS spectra were analyzed with a classifier algorithm based on linear discriminant analysis (LDA). LDA model classification accuracy (CA) obtained between IDH-mutant and IDH wild-type PGOs was 90% (91% sensitivity and 89% specificity). Furthermore, 1p/19q codeletion classification within IDH mutant PGOs reached 98% CA (93% sensitivity and 99% specificity), while CDKN2A/B homozygous loss status had 86% CA (63% sensitivity 93% specificity). DMS suitability to differentiate IDH-mutated PGOs was thus validated in ex vivo cultured samples, PGOs. Preliminary results regarding 1p/19q codeleted PGOs and CDKN2A/B loss PGOs identification endorse testing in a prospective intraoperative glioma patient cohort. Our results reveal a sample classification set-up that is compatible with real-time intraoperative surgery guidance.
Read full abstract