Frozen Section Analysis in Head and Neck Surgical Pathology: A Narrative Review of the Past, Present, and Future of Intraoperative Pathologic Consultation

A review of histopathologic assessment for head and neck oncologists

GOAL OF REVIEW

With a deeper understanding of histopathologic assessment, head and neck oncology specialists (surgical oncologists, radiation oncologists, and medical oncologists) will be better equipped to address the increasing complexity encompassing head and neck cancer management.

INTRODUCTION

Histopathologic assessment of surgical specimens imparts crucial information that is essential for post-operative treatment planning and prognostication for patients with head and neck squamous cell carcinoma (HNSCC). Herein, we discuss the most current guidelines and recommendations to elucidate the clinically relevant histopathologic features in HNSCC. This review discusses the following pathology features: extranodal extension, margins, perineural invasion, histologic grade, dysplasia, depth of invasion, lymphovascular invasion, and other considerations such as p16 immunohistochemistry, HPV in situ hybridization and worst pattern of invasion.

DISCUSSION

Understanding histopathology in HNSCC is essential for accurate diagnosis, prognostication, understanding tumor behavior, and treatment management. This complexity of care has led to consensus guidelines from numerous authorities which this paper discusses and summarizes for readers.

CONCLUSION

The understanding of key histopathology elements in HNSCC will augment multidisciplinary discussions and improve patient care. The current variability in existing consensus guidelines highlights the need for improved standardization of histopathology reporting in HNSCC. Standardization will enhance diagnostic accuracy, guide clinical decision-making, and facilitate the development of more effective treatment strategies.

Feasibility of 3D ultrasound for intraoperative tumor margin assessment in transoral robotic surgery for oropharyngeal squamous cell carcinoma: A pilot study

INTRODUCTION

Close margins after transoral robotic surgery (TORS) in oropharyngeal squamous cell carcinoma (OPSCC) are common due to narrow anatomical boundaries, requiring additional radiotherapy treatment (RT). Ultrasound (US) can be used intraoperatively to distinguish tumors from healthy tissue. Our objective was to explore ex vivo US of surgical specimens from OPSCCs using a novel 3D US method to correlate tumor and margin measurements with histopathology.

METHODS

Patients with OPSCC undergoing TORS either primarily or as salvage surgery were included. Ex vivo US was performed immediately after resection in the operation room and 3D US models were obtained. The US images were then analyzed by four surgeons blinded to histopathology for correlation analyses. The accuracy to classify close or free margins using a 2 mm threshold was computed.

RESULTS

Nine patients with OPSCC were included (median age 63 years, six males, five with previous RT, and five were Human Papillomavirus-positive). US and histopathology had a high correlation for tumor (r = 0.84-0.85) and margin measurements (r = 0.76-0.78). US measured deep margins with a mean difference of 0.5 mm (SD: 1.2 mm) compared to histopathology and had 80% sensitivity to detect areas of the surgical specimens with close margins (<2 mm). US correctly categorized the deep margin status in 89% of the surgical specimens, compared to 44% for the lateral margins.

CONCLUSIONS

This proof-of-concept study shows that ex vivo 3D US is feasible for intraoperative evaluation of deep surgical margins during TORS of OPSCCs.

The Multidisciplinary Team in Head and Neck Cancer Reconstruction: A Reference Manual for the Plastic Surgeon

Head and neck cancers (HNCs) require a multidisciplinary team (MDT) approach to address their complex functional, aesthetic, and psychological impacts. This manuscript highlights the central role of plastic surgeons in the MDT, emphasizing their collaboration with other MDT members to align aesthetic and functional surgical outcomes with oncologic and rehabilitative goals. Our intention is for this to be used as a practical guide for plastic surgeons detailing the roles of key MDT members and their contributions across the preoperative, intraoperative, and postoperative phases. We will also highlight how MDTs improve survival, functional outcomes, and quality of life for HNC patients.

Intraoperative Fast Adaptive Focus Tracking Robotic OCT Enables Real-Time Tumor Grading and Large-Area Microvascular Imaging in Human Spinal Cord Surgery

Current surgical procedures for spinal cord tumors lack in vivo high-resolution multifunctional imaging systems, hindering precise tumor resection. This study introduces the fast adaptive focus tracking robotic optical coherence tomography (FACT-ROCT) system, which provides real-time, artifact-free imaging during surgery, addressing motion artifacts and resolution degradation. Imaging occurs in 22 patients, including 13 with gliomas (WHO grade I-IV). This represents the first in situ OCT imaging of human spinal cord tumors, enabling the differentiation of tumor types in real-time. The standard deviation of the attenuation coefficient serves as a physical marker, achieving 90.2%  ±  2.7% accuracy in distinguishing high- from low-grade gliomas intraoperatively at a threshold of 0.75  ±  0.01 mm-1. FACT-ROCT also enables microvascular imaging, covering areas of 70 mm × 13 mm × 10 mm within 2 min and revealing greater vascular tortuosity in higher-grade tumors. This extensive imaging capability provides critical information that guides surgical strategies, enhancing surgical outcomes. Overall, FACT-ROCT represents a significant advancement in intraoperative imaging, offering high-resolution, high-speed, and comprehensive insights into spinal cord tumor structure and vasculature.

Deep learning in surgical process modeling: A systematic review of workflow recognition

OBJECTIVE

The application of artificial intelligence (AI) in health care has led to a surge of interest in surgical process modeling (SPM). The objective of this study is to investigate the role of deep learning in recognizing surgical workflows and extracting reliable patterns from datasets used in minimally invasive surgery, thereby advancing the development of context-aware intelligent systems in endoscopic surgeries.

METHODS

We conducted a comprehensive search of articles related to SPM from 2018 to April 2024 in the PubMed, Web of Science, Google Scholar, and IEEE Xplore databases. We chose surgical videos with annotations to describe the article on surgical process modeling and focused on examining the specific methods and research results of each study.

RESULTS

The search initially yielded 2937 articles. After filtering on the basis of the relevance of titles, abstracts, and content, 59 articles were selected for full-text review. These studies highlight the widespread adoption of neural networks, and transformers for surgical workflow analysis (SWA). They focus on minimally invasive surgeries performed with laparoscopes and microscopes. However, the process of surgical annotation lacks detailed description, and there are significant differences in the annotation process for different surgical procedures.

CONCLUSION

Time and spatial sequences are key factors determining the identification of surgical phase. RNN, TCN, and transformer networks are commonly used to extract long-distance temporal relationships. Multimodal data input is beneficial, as it combines information from surgical instruments. However, publicly available datasets often lack clinical knowledge, and establishing large annotated datasets for surgery remains a challenge. To reduce annotation costs, methods such as semi supervised learning, self-supervised learning, contrastive learning, transfer learning, and active learning are commonly used.

Visual pathology reports for communication of final margin status in laryngeal cancer surgery

Background

Positive margins are frequently observed in total laryngectomy (TL) specimens. Effective communication of margin sampling sites and final margin status between surgeons and pathologists is crucial. In this study, we evaluate the utility of multimedia visual pathology reports to facilitate interdisciplinary discussion of margin status in laryngeal cancer surgery.

Methods

Ex vivo laryngeal cancer surgical specimens were three-dimensional (3D) scanned before standard of care pathological analysis. Using computer-aided design software, the 3D model was annotated to reflect inking, sectioning, and margin sampling sites, generating a visual pathology report. These reports were distributed to head and neck surgeons and pathologists postoperatively.

Results

Fifteen laryngeal cancer surgical specimens were 3D scanned and virtually annotated from January 2022 to December 2023. Most specimens (73.3%) were squamous cell carcinomas (SCCs). Among the cases, 26.7% had final positive surgical margins, whereas 13.3% had close margins, defined as <5 mm. The visual pathology report demonstrated sites of close or positive margins on the 3D specimens and was used to facilitate postoperative communication between surgeons and pathologists in 85.7% of these cases. Visual pathology reports were presented in multidisciplinary tumor board discussions (20%), email correspondences (13.3%), and teleconferences (6.7%), and were referenced in the final written pathology reports (26.7%).

Conclusions

3D scanning and virtual annotation of laryngeal cancer specimens for the creation of visual pathology reports is an innovative approach for postoperative pathology documentation, margin analysis, and surgeon-pathologist communication.

Virtual Tumor Mapping: A New Standard for Surgeon-Pathologist Collaboration in Treating Oral Squamous Cell Carcinoma

Background: The histopathological assessment is critical in the comprehensive treatment process for patients diagnosed with oral squamous cell carcinoma (OSCC). A detailed and precise specimen characterization is essential to facilitate effective surgeon-pathologist communication. Methods: In response to this need, a user-friendly virtual communication protocol utilizing a 3D scanner has been developed. This study involved 50 patients with OSCC, whose resected tumors were directly scanned in the operating room and subsequently annotated and characterized using available software. Results: The direct application of annotations and descriptions onto the virtual tumor specimens significantly enhanced the quantity and accuracy of information conveyed to the pathologist. Conclusions: The proposed solution's repeatability and standardized approach make integration into routine clinical practice feasible, thereby establishing a potential new standard in the field.

Trends and Future Directions in Margin Analysis for Head and Neck Cancers

Margin status in head and neck cancer has important prognostic implications. Currently, resection is based on manual palpation and gross visualization followed by intraoperative specimen or tumor bed-based margin analysis using frozen sections. While generally effective, this protocol has several limitations including margin sampling and close and positive margin re-localization. There is a lack of evidence on the association of use of frozen section analysis with improved survival in head and neck cancer. This article reviews novel technologies in head and neck margin analysis such as 3-dimensional scanning, augmented reality, molecular margins, optical imaging, spectroscopy, and artificial intelligence.

Illuminating the future of precision cancer surgery with fluorescence imaging and artificial intelligence convergence

Real-time and accurate guidance for tumor resection has long been anticipated by surgeons. In the past decade, the flourishing material science has made impressive progress in near-infrared fluorophores that may fulfill this purpose. Fluorescence imaging-guided surgery shows great promise for clinical application and has undergone widespread evaluations, though it still requires continuous improvements to transition this technique from bench to bedside. Concurrently, the rapid progress of artificial intelligence (AI) has revolutionized medicine, aiding in the screening, diagnosis, and treatment of human doctors. Incorporating AI helps enhance fluorescence imaging and is poised to bring major innovations to surgical guidance, thereby realizing precision cancer surgery. This review provides an overview of the principles and clinical evaluations of fluorescence-guided surgery. Furthermore, recent endeavors to synergize AI with fluorescence imaging were presented, and the benefits of this interdisciplinary convergence were discussed. Finally, several implementation strategies to overcome technical hurdles were proposed to encourage and inspire future research to expedite the clinical application of these revolutionary technologies.

Improving Interdisciplinary Communication and Pathology Reporting for Head and Neck Cancer Resections: 3D Visualizations and Margin Reconciliation

PURPOSE

Surgical pathology reports play an integral role in postoperative management of head and neck cancer patients. Pathology reports of complex head and neck resections must convey critical information to all involved clinicians. Previously, we demonstrated the utility of 3D specimen and defect scanning for communicating margin status and documenting the location of supplemental margins. We introduce a newly designed permanent pathology report which improves documentation of intraoperative margin mapping and extent of corresponding supplemental margins harvested.

METHODS

We test the hypothesis that gaps in understanding exist for head and neck resection pathology reports across providers. A cross-sectional exploratory study using human-centered design was implemented to evaluate the existing permanent pathology report with respect to understanding margin status. Pathologists, surgeons, radiation oncologists, and medical oncologists from United States-based medical institutions were surveyed. The results supported a redesign of our surgical pathology template, incorporating 3D specimen / defect scans and annotated radiographic images indicating the location of inadequate margins requiring supplemental margins, or indicating frankly positive margins discovered on permanent section.

RESULTS

Forty-seven physicians completed our survey. Analyzing surgical pathology reports, 28/47 (60%) respondents reported confusion whether re-excised supplemental margins reflected clear margins, 20/47 (43%) reported uncertainty regarding final margin status, and 20/47 (43%) reported the need for clarity regarding the extent of supplemental margins harvested intraoperatively. From this feedback, we designed a new pathology report template; 61 permanent pathology reports were compiled with this new template over a 12-month period.

CONCLUSION

Feedback from survey respondents led to a redesigned permanent pathology report that offers detailed visual anatomic information regarding intraoperative margin findings and exact location/size of harvested supplemental margins. This newly designed report reconciles frozen and permanent section results and includes annotated radiographic images such that clinicians can discern precise actions taken by surgeons to address inadequate margins, as well as to understand the location of areas of concern that may influence adjuvant radiation planning.

Association of the Specimen and Tumor Bed Margin Status with Local Recurrence and Survival in Open Partial Laryngectomy

Background/Objectives: Positive margins are associated with locoregional recurrence in early laryngeal cancer. The aim of this study was to evaluate the impacts of specimen-driven (ex vivo) positive margins on patients with early-stage laryngeal cancer whose tumor bed (defect-driven) margins had been negative. Methods: A retrospective study was performed on 60 consecutive T1b/T2 glottic cancer patients who underwent open frontolateral laryngectomy. The intraoperative margins were obtained from the tumor bed. Their recurrence and disease-free survival were evaluated. In all cases, negative margins were obtained from the surgical bed. The impact of positive margins from the specimen was evaluated in a paraffin study. Results: Among 10 patients with positive margins in the specimen, six experienced local relapse, and among 50 patients with negative margins in the specimen, three developed recurrence. The 5-year disease-free survival rates were 37.5% and 93.9%, respectively (p < 0.001; log-rank). Even with negative margins in the surgical bed, patients with positive margins in the specimen at the final histopathological examination had a 3.5-fold higher chance of developing local recurrence than those with negative margins (HR = 13.993; 95% CI: 3.479-56.281; p < 0.001; univariate Cox regression). Conclusions: Specimen-driven positive margins represent a significant risk factor for local recurrence, even under negative margins at the tumor bed.

Diagnosis of malignancy in oropharyngeal confocal laser endomicroscopy using GPT 4.0 with vision

PURPOSE

Confocal Laser Endomicroscopy (CLE) is an imaging tool, that has demonstrated potential for intraoperative, real-time, non-invasive, microscopical assessment of surgical margins of oropharyngeal squamous cell carcinoma (OPSCC). However, interpreting CLE images remains challenging. This study investigates the application of OpenAI's Generative Pretrained Transformer (GPT) 4.0 with Vision capabilities for automated classification of CLE images in OPSCC.

METHODS

CLE Images of histological confirmed SCC or healthy mucosa from a database of 12 809 CLE images from 5 patients with OPSCC were retrieved and anonymized. Using a training data set of 16 images, a validation set of 139 images, comprising SCC (83 images, 59.7%) and healthy normal mucosa (56 images, 40.3%) was classified using the application programming interface (API) of GPT4.0. The same set of images was also classified by CLE experts (two surgeons and one pathologist), who were blinded to the histology. Diagnostic metrics, the reliability of GPT and inter-rater reliability were assessed.

RESULTS

Overall accuracy of the GPT model was 71.2%, the intra-rater agreement was κ = 0.837, indicating an almost perfect agreement across the three runs of GPT-generated results. Human experts achieved an accuracy of 88.5% with a substantial level of agreement (κ = 0.773).

CONCLUSIONS

Though limited to a specific clinical framework, patient and image set, this study sheds light on some previously unexplored diagnostic capabilities of large language models using few-shot prompting. It suggests the model`s ability to extrapolate information and classify CLE images with minimal example data. Whether future versions of the model can achieve clinically relevant diagnostic accuracy, especially in uncurated data sets, remains to be investigated.