Reliability of Frozen Section Pathology in Transoral Laser Laryngectomy

Diagnosis of Early Glottic Cancer Using Laryngeal Image and Voice Based on Ensemble Learning of Convolutional Neural Network Classifiers

OBJECTIVES

The purpose of study is to improve the classification accuracy by comparing the results obtained by applying decision tree ensemble learning, which is one of the methods to increase the classification accuracy for a relatively small dataset, with the results obtained by the convolutional neural network (CNN) algorithm for the diagnosis of glottal cancer.

METHODS

Pusan National University Hospital (PNUH) dataset were used to establish classifiers and Pusan National University Yangsan Hospital (PNUYH) dataset were used to verify the classifier's performance in the generated model. For the diagnosis of glottic cancer, deep learning-based CNN models were established and classified using laryngeal image and voice data. Classification accuracy was obtained by performing decision tree ensemble learning using probability through CNN classification algorithm. In this process, the classification and regression tree (CART) method was used. Then, we compared the classification accuracy of decision tree ensemble learning with CNN individual classifiers by fusing the laryngeal image with the voice decision tree classifier.

RESULTS

We obtained classification accuracy of 81.03 % and 99.18 % in the established laryngeal image and voice classification models using PNUH training dataset, respectively. However, the classification accuracy of CNN classifiers decreased to 73.88 % in voice and 68.92 % in laryngeal image when using an external dataset of PNUYH. To solve this problem, decision tree ensemble learning of laryngeal image and voice was used, and the classification accuracy was improved by integrating data of laryngeal image and voice of the same person. The classification accuracy was 87.88 % and 89.06 % for the individualized laryngeal image and voice decision tree model respectively, and the fusion of the laryngeal image and voice decision tree results represented a classification accuracy of 95.31 %.

CONCLUSION

The results of our study suggest that decision tree ensemble learning aimed at training multiple classifiers is useful to obtain an increased classification accuracy despite a small dataset. Although a large data amount is essential for AI analysis, when an integrated approach is taken by combining various input data high diagnostic classification accuracy can be expected.

Predictive factors of frozen section in transoral microlaryngeal surgery for suspicious glottic lesions

OBJECTIVE

Frozen biopsy may guide surgical intraoperative decisions. We evaluated the accuracy of frozen biopsy for diagnosing benign, dysplastic and malignant laryngeal lesions, compared to paraffin section (gold standard).

METHODS

Retrospective review of the charts of all patients presenting with laryngeal lesions suspicious of malignancy, who underwent laryngeal microsurgery with frozen biopsy in our institution, between 2015 and 2020. Results of frozen biopsy and paraffin section examinations were compared.

RESULTS

Among 113 samples of 89 patients, paraffin section diagnosed 23 benign, 31 dysplastic and 59 malignant lesions. The accuracy of the frozen biopsy in identifying dysplasia or malignancy was 80.5% (91/113), and greater for lesions >5 mm (78.8% × 51.5%; p = 0.009). The positive and negative predictive values, sensitivity and specificity were 95.9%, 51.3%, 78.9% and 86.9%, respectively.

CONCLUSIONS

Frozen section is a reliable tool when malignancy is detected, but almost half of benign results exhibit dysplasia or malignancy in paraffin section. Other clinical parameters should be considered in intraoperative decisions to prevent undertreatment.

LEVEL OF EVIDENCE: 4

A Retrospective Study to Assess the Accuracy of Frozen Sections in Laser Laryngeal Surgery

OBJECTIVE

The basis of good vocal outcomes following Transoral Laser Microlaryngeal Cordectomy (TLMC) is the narrow margin that is oncologically accepted for the glottis. Our objective is to evaluate the reliability of frozen section (FS) compared to paraffin section (PS) during TLMC and during laser laryngeal surgery when an incisonal or excisional biopsy is being performed.

METHODS

Retrospectively, records of 159 sequential patients who underwent CO₂ laser laryngeal surgery with intraoperative FS were reviewed along with the final PS. Group A patients were TLMC patients where FS was utilized for free margin confirmation (42 specimens) and Group B patients were those in whom FS was performed for primary diagnosis (122 specimens).

RESULTS

A total of 164 samples were included where specimens submitted for FS were also processed for PS. Concordance was observed in 156 samples, discordance in 8, with 4 cases belonging to each group. FS was reported as false negative in 5 and false positive in 3 cases. In Group A where FS was utilized for free margin confirmation, the sensitivity was found to be 60%, specificity and positive predictive value (PPV) 100%, and negative predictive value (NPV) 88.9%. In Group B where FS was performed for primary diagnosis, the sensitivity was found to be 98.4%, specificity 95.1%, PPV 95.2%, and NPV 98.3%.

CONCLUSION

A 100% PPV of FS in group A suggests that positive FS margins during TLMC may be safely relied upon in making decisions to upgrade the type of cordectomy being performed. A NPV of 88.9% in group A suggests that despite a clear frozen margin report during TLMC, 11.1% of patients would need further treatment. A PPV of 95.2% in group B suggests that although FS analysis is important for guiding further management, decisions regarding major laryngeal surgery should not be undertaken based solely on FS.

Minimally Invasive Procedures for Laryngeal Carcinoma: Transoral Endoscopic Laser and Transoral Robotic Surgery

Open partial or total laryngectomies with or without radiotherapy (RT) or chemoradiotherapy (CRT) are the only mainstays in the treatment of laryngeal carcinomas. However, minimally invasive procedures such as transoral endoscopic carbondioxide laser microsurgery (TLM) or transoral robotic surgeries (TRS) are now being increasingly used in selected patients. The laryngeal framework is not disturbed in these procedures; therefore, the postoperative swallowing function improves more rapidly, and routine tracheotomy is not usually required. Moreover, they have oncological results comparable with open procedures.