To achieve successful surgical oncology outcomes, surgeons must completely extirpate the full extent of a patient’s cancer volume, both gross and microscopic. Currently, there is a lack of technology capable of providing surgeons with real-time feedback to quantify the extent of tumor; therefore, cancer may be missed, or surgeons may potentially compromise benign tissue. This research investigates whether Fluorescence Lifetime Imaging (FLIm) can optimize upper aerodigestive oncologic surgery through real-time surgical margin delineation. One hundred patients were evaluated, comprising N=45 patients with oral cavity cancer and N=55 patients with cancer of the oropharynx.

This work first demonstrates the series of methodologies, tools, and techniques developed to acquire intraoperative FLIm data and validate against gold-standard histopathology. Next, the ability of FLIm to demarcate primary tumors of the oropharynx was investigated for an initial N=10 patient cohort of tonsil tumors. In this study, FLIm demarcated the full extent of all patients’ cancers; however, the best benign vs. tumor contrast originated from multivariate analysis of all FLIm time-resolved and spectral intensity parameters. To this end, a random forest classifier method was used to integrate multivariate FLIm parameters from an N=55 patient dataset comprising conventional primary tumors and leveraged to demarcate occult primary tumors of the oropharynx for a cohort of N=7 patients. This work concludes by investigating sources of FLIm data variability across all patients, such as tumor characteristics (e.g., p16 status), patient age, lymphoid tissue, high-grade dysplasia, and local anesthetic use. A preliminary demonstration of FLIm’s ability to identify positive surgical margins in vivo within a patient’s surgical cavity is also illustrated.

Collectively, this original research demonstrates the feasibility for FLIm to intraoperatively demarcate carcinoma during upper aerodigestive oncology procedures, therefore motivating ongoing investigation to materialize this technology towards an intraoperative diagnostic modality.