Use of Transcervical Ultrasound for Identification of Primary Site in Patients With Oropharyngeal Cancer

Presentation: AHNS003
Topic: Pharynx / Larynx Cancer
Type: Oral
Date: Wednesday, April 18, 2018
Session: 9:05 AM - 10:00 AM Technology
Authors: C. Burton Wood, MD1, Sarah L Rohde, MD1, Robert Sinard, MD1, Kyle Mannion, MD1, Alexander Langerman, MD, SM1, Young J Kim, MD, PhD1, Joseph Aulino, MD, PhD1, Derek K Smith, DDS, PhD1, Arthur Fleischer, MD1, Carole Fakhry, MD, MPH2, James L Netterville, MD1, Krystle L Kuhs, PhD, MPH1
Institution(s): 1Vanderbilt University Medical Center, 2Johns Hopkins University

Background: HPV-driven oropharyngeal cancer (HPV-OPC) is increasing in incidence. Early identification of HPV-OPC is difficult, as patients are frequently asymptomatic until later stages of disease. Methods for early detection in asymptomatic patients at risk of developing HPV-OPC would be ideal. HPV16 E6 seropositivity is a promising early marker of HPV-OPC; yet, it is unclear if current imaging modalities are able to visualize early tumors within the oropharynx. Transcervical ultrasound has shown promise for detecting unknown primary lesions not visualized with traditional modalities such as CT or FDG-PET/CT. The combination of HPV16 E6 antibodies and ultrasound therefore represents a potential method for early detection of HPV-OPC.

Objective: This ongoing pilot study was designed to determine the sensitivity of HPV16 E6 antibodies and ultrasound for detecting HPV-OPC.

Methods: Patients with known or suspected OPC without prior treatment or cancer (other than non-melanoma skin cancer) were recruited upon their first visit to Head and Neck Surgery Clinic. Patients underwent ultrasound and biospecimen collection (blood and oral rinse). Using the Lumify portable ultrasound system and mobile application (Philips Healthcare, Bothwell, WA), 8 standard images (transverse/sagittal views of tonsils, transverse/coronal views of tongue base [BOT], and bilateral lateral BOT) were collected; tumors were measured if identified. Pathologic details, HPV status, final staging, and radiologic findings were recorded. The sensitivity of each imaging modality was compared to the “final clinical diagnosis” (gold-standard), which was determined after each patient completed a full diagnostic work-up as part of standard clinical care. The final clinical diagnosis was determined by a head and neck surgeon and was based on the combined findings from the clinical and radiologic workup (minus ultrasound) and biopsy of the primary site as required.  Oral HPV testing and multiplex serology will be performed once all patients are recruited.

Results: To date, 18 eligible patients were identified; 14 (78%) were enrolled (target enrollment=50).  Following diagnostic work-up, 8 (57%) BOT and 6 (43%) tonsil primaries were diagnosed. FNAs or primary biopsies from 13 patients were clinically tested for p16 immunohistochemistry; 100% were positive. All patients (N=14) underwent ultrasound, 13 patients had CT imaging, 8 had FDG-PET/CT imaging and 10 had a biopsy of the primary site. Lesions were correctly identified in 13 of 14 patients (93%) using ultrasound and 8 of 8 (100%) patients using FDG-PET/CT. In the one case where ultrasound incorrectly identified the lesion in the right tonsil, FDG-PET/CT correctly identified it in the right BOT. In contrast, 7 out of 13 lesions (54%) were correctly identified using CT; 2 primaries were misidentified and in 4 (31%) cases no tumor was visualized.  Ultrasound correctly identified tumors in all 4 cases where CT failed to visualize the primary and identified all 10 tumors with pathologic confirmation of the primary site. The smallest tumor identified was 1cm in greatest dimension; average size was 2.4cm.

Discussion: Our preliminary findings suggest that transcervical ultrasound is sensitive for detecting tumors within the oropharynx and may have greater accuracy than CT, even for tumors larger than 1cm.