Label-free, real-time detection of perineural invasion and cancer margins in a murine model of head and neck cancer surgery

Abstract

Surgical management of head and neck cancer requires a careful balance between complete resection of malignancy and preservation of function. Surgeons must also determine whether to resect important cranial nerves that harbor perineural invasion (PNI), as sacrificing nerves can result in significant morbidity including facial paralysis. Our group has previously reported that Dynamic Optical Contrast Imaging (DOCI), a novel non-invasive imaging system, can determine margins between malignant and healthy tissues. Herein, we use an in vivo murine model to demonstrate that DOCI can accurately identify cancer margins and perineural invasion, concordant with companion histology. Eight C3H/HeJ male mice were injected subcutaneously into the bilateral flanks with SCCVIISF, a murine head and neck cancer cell line. DOCI imaging was performed prior to resection to determine margins. Both tumor and margins were sent for histologic sectioning. After validating that DOCI can delineate HNSCC margins, we investigated whether DOCI can identify PNI. In six C3H/HeJ male mice, the left sciatic nerve was injected with PBS and the right with SCCVIISF. After DOCI imaging, the sciatic nerves were harvested for histologic analysis. All DOCI images were acquired intraoperatively and in real-time (10 s per channel), with an operatively relevant wide field of view. DOCI values distinguishing cancer from adjacent healthy tissue types were statistically significant (P < 0.05). DOCI imaging was also able to detect perineural invasion with 100% accuracy compared to control (P < 0.05). DOCI allows for intraoperative, real-time visualization of malignant and healthy tissue margins and perineural invasion to help guide tumor resection.

ICG-Functionalized Gold Nanostars As An Effective Contrast Agent For Real-time Tumor Localization with Dynamic Optical Contrast Imaging (DOCI) and Enhanced Radiation Therapy

“…We have previously designed Dynamic Optical Contrast Imaging (DOCI), a novel imaging modality that acquires temporally dependent measurements of tissue autofluorescence. 5-9 Furthermore, we demonstrated that DOCI can distinguish HNSCC from adjacent healthy tissue with a high degree of accuracy. 5 DOCI images are captured in real time, offering an operatively wide field of view.…”

“…The utility of GNS is shown via both tail vein injection and direct tumoral injections as shown in Figure 4, indicating that clinicians would have the ability of performing either form of GNS administration with adequate accumulation within tumors. As DOCI has already reliably differentiated head and neck SCC from surrounding normal tissue in prior experiments, 5-7,9 addition of GNSs in delineating deep margins as highlighted in Figures 2 and 3 will allow for improved cancer resection in real time, guiding surgeons to determine suspicious microscopic regions more accurately. Our murine model results indicate that implementation of GNS has the capability to revolutionize DOCI’s clinical promise in cancer care.…”

“…Our group has previously demonstrated that DOCI imaging with a 365nm light source provides valuable cancer margin delineation in vivo and intraoperatively. [5][6][7][8][9] All eukaryotic and prokaryotic cells exhibit an intrinsic natural autofluorescence (AF) easily excited at 365nm, due to the presence of various fluorescent cellular structural components and metabolites, such as flavins, nicotinamide-adenine dinucleotide (NAD), aromatic amino acids, lipofuscins, advanced glycation end products, and collagen. 23,24 This difference in AF cellular bioproducts between tissues gives distinctive lifetime values that provide natural contrast between cancer and noncancer tissue when imaged.…”

AI-Enabled Classification of Head and Neck Tumors using Convolutional Neural Networks with Dynamic Optical Contrast Imaging

“…As has already been demonstrated by our previous studies and in the current one, DOCI reliably differentiates head and neck cancer tissue from surrounding normal tissue on the basis of fluorescence decay information. [5][6][7][8][9][10] The performance of the DOCI system is enhanced by the CNN-based classification model for automated classification of tumor and nontumor tissue of DOCI-acquired images. As the CNN model continues to acquire further DOCIacquired image samples, we anticipate that the tumor region confidence will continue to improve above the current 76.02% confidence mark.…”

“…Our group has developed and previously described Dynamic Optical Contrast Imaging (DOCI), a novel imaging modality that acquires temporally-dependent measurements of tissue autofluorescence. [5][6][7][8][9] Additionally, in fresh ex-vivo experiments, our group showcased that the DOCI system reliably differentiates HNSCC from surrounding normal tissue and facilitates specific tumor localization [Figure 1]. 5 DOCI images are captured in real-time, do not require administration of contrast agents, and offer an operatively wide field of view.…”