P59: Evaluation of the MasSpec Pen Technology for Molecular Identification of Sarcomas and Surgical Margin Assessment

Surgical resection with negative margins is the standard treatment of localized bone and soft tissue sarcomas. Intraoperative bone and soft tissue sarcoma tissue identification and surgical margin evaluation are challenging, resulting in subjectivity between centers. To address this problem, our goal is to evaluate the applicability of the MasSpec Pen (MSPen) in accurately identifying sarcoma subtypes based on metabolite and lipid profiles compared to normal tissues.

Methods:

The study was designed in two phases using retrospective sets of banked tissue samples and prospective analysis of tissues in the operating room. One hundred twenty-one frozen sarcoma tissues, including myxofibrosarcomas, well-differentiated liposarcomas, de-differentiated liposarcomas, osteosarcomas, chordomas, malignant peripheral nerve sheath tumors (MPNST), neurofibromas were identified. Controls were bone, cartilage, fat, nerve, skeletal, and smooth muscle tissues. Sarcoma pathologists confirmed each sample's diagnosis. The MSPen integrated into an Orbitrap Exploris mass spectrometer was used to analyze the tissues to define their molecular profiles. Performance using an Orbitrap and a linear ion trap mass spectrometer was tested. Statistical classification models are being generated using logistic regression with LASSO regulation and will be assessed using training, validation, and test sets. The classifier's sensitivity, specificity, and accuracy will be compared to histopathologic analysis.

Results:

Analysis of 121 sarcoma and normal tissues (described in Methods) was performed with the MSPen. We detected distinct metabolite and lipid profiles for the analyzed histologic sarcoma subtypes, which were used as a training set to create a statistical classification model for the sarcoma subtypes in comparison to normal tissue. Based on the number of samples, we obtained preliminary results for some subtypes. Please see Table 1. For example, the chondrosarcoma vs. normal classification model generated a sensitivity of 100.00%, specificity of 90.91%, and overall accuracy of 94.74%.

Conclusions:

The MSPen leverages its non-destructive nature for direct and gentle analysis of tissues. The MSPen and statistical analysis using logistic regression with LASSO regulation offer a promising technology for sarcoma diagnosis and discrimination from normal tissue. These data support further evaluation of the MSPen in the operating room for in vivo and ex vivo use in conjunction with standard surgical tools for margin assessment.