Here we describe a new instrument (SpiderMass) designed for in vivo and real-time analysis. In this instrument ion production is performed remotely from the MS instrument and the generated ions are transported in real-time to the MS analyzer. Ion production is promoted by Resonant Infrared Laser Ablation (RIR-LA) based on the highly effective excitation of O-H bonds in water molecules naturally present in most biological samples. The retrieved molecular patterns are specific to the cell phenotypes and benign versus cancer regions of patient biopsies can be easily differentiated. We also demonstrate by analysis of human skin that SpiderMass can be used under in vivo conditions with minimal damage and pain. Furthermore SpiderMass can also be used for real-time drug metabolism and pharmacokinetic (DMPK) analysis or food safety topics. SpiderMass is thus the first MS based system designed for in vivo real-time analysis under minimally invasive conditions.

Graphical AbstractHighlights d SpiderMass technology collects molecular information ex vivo and in vivo d Tumor classification by typing and grading is obtained by the technology d Found markers are cross-validated with mass spectrometry imaging d The technology was assessed ex vivo at the vet surgery room SUMMARY Histopathological diagnosis of biopsy samples and margin assessment of surgical specimens are challenging aspects in sarcoma. Using dog patient tissues, we assessed the performance of a recently developed technology for fast ex vivo molecular lipid-based diagnosis of sarcomas. The instrument is based on mass spectrometry (MS) molecular analysis through a laser microprobe operating under ambient conditions using excitation of endogenous water molecules. Classification models based on cancer/normal/necrotic, tumor grade, and subtypes showed a minimum of 97.63% correct classification. Specific markers of normal, cancer, and necrotic regions were identified by tandem MS and validated by MS imaging. Real-time detection capabilities were demonstrated by ex vivo analysis with direct interrogation of classification models.