Surgical Tissue Profiling and Computational Pipeline Using the Liquid Micro-junction Surface Sampling Probe (LMJ-SSP) Mass Spectrometry; a Proof-of-Concept Study

Perioperative assessment of tissue enhances surgical efficacy by providing real-time tissue analysis. A liquid microjunction surface sampling probe (LMJ-SSP) coupled with electrospray ionization mass spectrometry is shown to differentiate neoplastic tissue from adjacent non-neoplastic parenchyma of kidney and liver specimens. The LMJ-SSP consists of two open-ended concentric tubes between which a desorption solvent is pumped; after tissue contact, the liquids are drawn through the inner tube connected to a mass spectrometer. Tissues are interrogated by contact with the desorption solvent at the probe opening and extracted analytes were delivered to the mass spectrometer within seconds. For each sample, both a mass spectrometry (MS) image and training spectra were acquired in positive and negative-ion modes. A computational pipeline, from non-negative matrix factorization (NMF) to supervised affine space formation, was used for MS image visualization. MS visualizations distinguished neoplastic tissue from adjacent non-neoplastic kidney and liver parenchyma. Visualizations showed strong correlation with annotated gross pathology and histology diagnosis. Affine spaces produced from both ion mode acquisitions identified different m/z bins for tissue-type distinction. Analytes gathered in positive-ion mode produced the highest quality visualizations for both tissue types. The proposed LMJ-SSP technology enabled rapid tissue-type distinction of neoplastic or neoplastic tissue from non-neoplastic parenchyma.