Abstract by Lauritz Brorsen

Keratinocyte carcinoma (KC), which includes basal cell carcinoma (BCC) and cutaneous squamous cell carcinoma (SCC), is the most common form of skin cancer globally, posing significant challenges to healthcare systems due to its high incidence and morbidity. Current treatment regimens primarily rely on histopathological examination for diagnosis, are time-consuming, resource-intensive, and subject to interobserver variation. Laser ablation treatment, although highly regarded for its precision, speed, and favorable cosmetic outcomes, is used off-label for KC treatment and is limited to low-risk, superficial tumors due to the destruction of tissue during the ablation process, which precludes the histopathological examination of tumor margins.

A deeper pathophysiological understanding of KC is crucial for developing more effective diagnostic and therapeutic strategies. Metabolomics, including lipidomics, is a growing field in oncology with the potential to uncover new biomarkers, improve patient stratification, enhance disease prognosis, and identify novel therapeutic targets.

This thesis aims to address these challenges by evaluating the potential of matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) for the metabolomic characterization and clinical diagnosis of KC, and by developing a laser-assisted rapid evaporative ionization mass spectrometry (LA-REIMS) system for imaging and real-time intraoperative tissue classification using an ablative laser.

MALDI-MSI was used to analyze tissue sections from mouse models of SCC and BCC. Logistic regression (LR) machine learning models were trained to identify distinct metabolic profiles differentiating tumor from non-tumor regions. Cross-validation was applied to evaluate model accuracy on pre-labeled data. Histological experts then assessed the models' performance by evaluating predictive accuracy, sensitivity, and specificity across full tissue areas. Additionally, a LA-REIMS system was developed by integrating an ablative CO₂ laser operating at 10.6 mm with a compact mass spectrometer, coupled with a REIMS ion source and an imaging stage, specifically designed for clinical implementation, enabling both metabolomic profiling and accurate diagnosis.

Results demonstrated that MALDI-MSI could reliably characterize the metabolic profiles of KC and accurately identify tumor regions based on spectral characteristics. Assessing the concordance with histology revealed an overall predictive accuracy of 99.4% for SCC and 99.9% for BCC with a notably high sensitivity. Metabolic profiles for each tumor type were identified and compared, contributing to a deeper understanding of KC pathogenesis. A complete LA-REIMS system was successfully developed, but not thoroughly tested yet. Preliminary results showed promise in real-time SCC recognition, revealing similarities with the metabolic profiles obtained from MALDI-MSI. These results support the potential for enhancing surgical outcomes and reducing operation times with LA-REIMS.

In conclusion, this thesis establishes a pre-clinical foundation for the clinical implementation of advanced mass spectrometry techniques in KC management. The findings suggest that LA-REIMS could provide a method for real-time diagnosis and treatment of KC in a single procedure, while also improving the pathophysiological understanding of the disease through metabolomics. Future research should focus on validating these techniques in larger, diverse datasets and exploring their applicability in human clinical settings.