Panitumumab-IRDye800 Improves Laryngeal Tumor Mapping During Transoral Laser Microsurgery

Introduction:

Laryngeal cancer is a relatively common entity, with nearly 13,000 new cases being diagnosed in the United States annually¹. Transoral laser microsurgery (TLM) is the primary surgical modality for early-staged laryngeal tumors (T1/T2) with oncologic outcomes similar to radiotherapy. However, local control failure rates, particularly in the setting of T2 and/or radio-recurrent disease, are reported as exceeding 25%.²

Currently, margin assessment following TLM is dependent on visual assessment of the resection bed, and small-volume marginal biopsies that are susceptible to inaccurate histopathology or geographic miss, particularly after ablative or piecemeal resections. A method to improve assessment of tumor geography is necessary to maximize the oncologic efficacy of TLM; to expand the range of endoscopically resectable tumors, including those with wider geographic footprints; and, to help to mitigate the impact of surgeon experience on outcomes.

The use of fluorescence guided surgery (FGS) in patients with head and neck squamous cell carcinoma (HNSCC) has empirically improved intraoperative tumor identification.³ Panitumumab-IRDye800 is an epidermal growth factor (EGFR) monoclonal antibody covalently linked to a near-infrared imaging dye that is administered as an infusion preoperatively; it has been investigated as part of an ongoing phase II clinical trial at the authors’ institution as a method to improve intraoperative margin assessment following resection of HNSCC. The antibody-fluorophore conjugate binds specifically to EGFR, which is relatively overexpressed among head and neck cancers, resulting in tumor-specific immunofluorescence detectable intraoperatively using a camera equipped for near-infrared fluorescence imaging. Of note, a recent appraisal of the technique found that panitumumab-IRDye800 changed the final HNSCC resection margin in 21% of cases.³ The ability of the dye to enhance margin assessment is contingent on its high sensitivity for neoplasia; in fact, in vitro studies have revealed visible uptake in tumor volumes as minimal as 450 cells.^4,5 The authors of this study predicted that fluorescence guided surgery using panitumumab-IRDye800 would be of particular value in the endoscopic management of laryngeal cancer, given the functionally sophisticated, three-dimensional structure of the larynx, and limitations of tumor visualization through a tubular endoscope. The following case details the use of a current Phase II open-label clinical trial protocol to evaluate the efficacy of FGS using panitumumab-IRDye800 during a potassium titanyl phosphate (KTP) laser-assisted TLM ablative resection of a T2N0M0 glottic squamous cell carcinoma.

Case Report:

The patient is a 78-year-old male who previously underwent TLM for a diagnosis of at least glottic carcinoma in situ. Laryngeal examination demonstrated sessile disease involving the bilateral musculomembranous vocal folds and the right vestibular fold, with significant perturbation of the underlying mucosal wave, raising suspicion for disease penetrating into superficial lamina propria. The patient received 50mg intravenous injection of panitumumab-IRDye800 three days prior to surgery. In vivo near-infrared fluorescence imaging was performed using both the Stryker PINPOINT Endoscopic Fluorescence Imaging System^® and the Intuitive da Vinci Advanced Firefly Imaging System^® (Figure 1). Both endoscopes were advanced through the lumen of the glottiscope prior to tumor resection, mid-resection, and post-resection to evaluate the wound bed incrementally (Figures 2, 3, & 4).

Figure 1:

A. Pre-operative endoscopic exam demonstrating disease originating from the bilateral musculomembranous folds and the right vestibular fold. B. Telescopic, intraoperative view of the larynx with disease footprint delineated by hashed white line. C. Larynx visualized intraoperatively through the da Vinci Xi robotic endoscope via Advance Firefly imaging. D. Intraoperative view of the larynx using the Stryker PINPOINT near-infrared fluorescence imaging system

Figure 2:

TBR ranges pre-ablation for the Stryker imaging system (D) were 1.5 to 1.9, which has less of a dynamic range compared to the Da Vinci pre ablation (B) which are 2.4 to 4.5. These ranges decreased dramatically post ablation (F) to background uptake levels.

Figure 3:

A: Pre-ablation, B: Mid-ablation, C: Post-ablation imaging via da Vinci Advanced Firefly and corresponding biopsy specimens (roman numerals) viewed with brightfield and back-table fluorescence imaging. D: Bar graph of mean fluorescence values (MFI) of biopsy samples. Final pathology for samples with at least severe dysplasia include I through VII. Marginal biopsies, represented by samples VIII through X, demonstrated chronic inflammation but no evidence of malignancy.

Figure 4:

A. Mid-ablation right vocal fold brightfield and corresponding da Vinci Advanced Firefly images. B. Post-ablation right vocal fold brightfield and corresponding da Vinci Advanced Firefly images C. Anterior commissure post-ablation brightfield and da Vinci Advanced Firefly images

The resection was paused mid-ablation in order to characterize the sensitivity of panitumumab-IRDye800 in the identification of extremely small-volume disease on the surface of the thyroarytenoid muscle (Figure 3C, 3E & 4A); differential uptake of the dye corresponded precisely to the footprint of the residual tumor. Ablation through this disease resulted in the elimination of visible, differential uptake of the dye in the wound bed. (Figure 2 & 4B). Quantitatively, we also assessed a decrease in fluorescent uptake by calculating the tumor-to-background ratio (TBR), a value normalizing fluorescent uptake values by using background values as a denominator (Figure 3).

Multiple representative lateral mucosal and deep marginal biopsies were obtained at the completion of the resection using microlaryngeal instrumentation in standard fashion. (Figure 3) Specimens were imaged ex vivo on the back table immediately in a closed-field near-infrared optical imaging system, before undergoing routine histopathologic assessment. Fluorescence uptake data was assessed against final histopathologic diagnoses. Nearly every biopsy obtained in regions wherein there was high relative uptake of the conjugate dye was interpreted histopathologically as at least severe dysplasia, except the right posterior vestibular fold that was read as squamous mucosa with cautery artifact (Figure 3). All post-resection marginal biopsies in regions without differential uptake of the conjugate dye were negative for dysplasia or neoplasia. The patient was ostensibly disease-free at most recent, 3-month post-operative follow-up.

Discussion:

We present for the first time in the extant literature the use of a tumor-specific intraoperative imaging agent to elucidate tumor geography during an endoscopic laryngeal cancer resection. There was close correlation between histopathologic diagnosis and differential uptake of panitumumab-IRDye800 assessed intraoperatively. The only non-correlated biopsy (posterior vestibular fold) in an area of higher, relative uptake was confounded by the presence of cautery artifact. Moreover, FGS offered the competitive technical advantage of visualizing extremely small volume disease in a location wherein a representative tissue biopsy would have cultivated a functionally significant soft tissue deficit. (Figure 4A) In fact, histopathologic assessment in this case failed to establish a diagnosis of cancer, which was clinically confirmed based on the presence of disease deep to the epithelium. The authors believe that improving tumor delineation via FGS has the potential to improve both oncologic and functional outcomes following TLM by more precisely defining resection margins. The authors plan for prospective follow-up and additional recruitment to fully assess the relative benefit of this technique to local control and disease-specific survival in the endoscopic management of laryngeal cancers, and to understand patterns of differential uptake in dysplasia versus neoplasia. Additional investigation into more sensitive NIR imaging systems to improve detection of fluorescence may also improve the efficacy of the technique.

Conclusion:

We describe for the first time the use of tumor-specific fluorescence-guided surgery in the endoscopic management of an early laryngeal cancer. We propose the potential of this technique to overcome empirical challenges associated with non-representative marginal biopsies and tumor identification within the complex three-dimension structure of the larynx, in order to radically optimize functional and oncologic outcomes following narrow-margin TLM.