Surgeon-performed Outpatient Transoral and Transcervical Ultrasound of the Oropharynx

A prospective study will be performed at the head \& neck departments at Copenhagen University Hospital - Rigshospitalet, Zealand University Hospital - Køge, and Aarhus University Hospital in Denmark.

Patients referred with a suspected cancerous lesion of the oropharynx will be invited to participate in this study. With written informed consent, a surgeon-performed outpatient transoral and transcervical examination of the tonsils and tongue base is conducted as an addition to the clinical workup at the head \& neck departments. Clinical work-up includes palpation of the oropharynx, flexible endoscopic exam with narrow-band imaging and neck ultrasound.

Ultrasound of the oropharynx will be performed using small-footprint, high-frequency (minimum 15MHz) transducers for transoral ultrasound. A lower-frequency (between 2-9 MHz) curved transducer will be used for transcervical ultrasound. If a suspected tumor is visualized, the location is noted and the surgeon performing clinical examination is informed.

MRIs will be performed on all patients with cancers, and an expert radiologist blinded to ultrasound results will assess all images for tumor detection, greatest tumor size and T-stage. PET/CTs will be performed in patients with suspected cancer of unknown primary tumor as well as patients with verified cancers. A nuclear medicine physician and an oncoradiologist will separately assess PET/CTs and CTs for tumor detection.

Surgeons performing initial clinical work-up, the ultrasound operators, neuroradiologist and nuclear medicine physician will register tumor detection (sub-location specific) and greatest tumor size immediately and prospectively, blinded to histopathology. Prospective gathered data will be registered on standardized data collection sheets or in REDCap using the same standardized variables. All scanning modalities will be evaluated blinded to each other but not to clinical findings.

Tumor detection: Suspected tumor locations from clinical evaluation with narrowband imaging and palpation, ultrasound, MRI, and PET-CT will be categorized according to the following criteria:

• Right tonsil
• Left tonsil
• Right tongue base
• Left tongue base
• Right overlapping tonsil and tongue base
• Left overlapping tonsil and tongue base
• Other oropharynx (soft palate, uvula, anterior and posterior pharyngeal arches, posterior oropharynx wall, and vallecula)
• Multiple tumor locations The examiners' tumor suspicion confidence will be rated with a 5-point Likert scale ranging from 0, very low suspicion - 5, very high suspicion.

The surgeon will also mark on a pictogram of the oropharynx with the biopsy location(s) numbered according to the histopathology registered biopsy locations.

Tumor size and staging: If a tumor is visualized or palpated, the greatest tumor diameter will be measured in three dimensions with clinical palpation, ultrasound, and MRI. T-staging will be estimated according to the Union for International Cancer Control 8th edition TNM-staging system:

• Tx: Unknown primary tumor.
• T1: Tumor <20 mm in greatest diameter.
• T2: Tumor 21-40 mm in greatest diameter.
• T3: Greatest size >40mm or extension to the lingual surface of the epiglottis
• T4: Invasion of the extrinsic tongue muscles, larynx, pterygoid muscles, hard palate, mandible, carotids nasopharynx or skull base (yes, no, indeterminate).

Additional variables: We will prospectively register the study center, date of outpatient ultrasound, ultrasound machine used, doctor performing the ultrasound scan, patient Mallampati 4-point score, Patient mouth opening measured in millimeters between incisors, patient self-reported discomfort from transoral ultrasound and palpation examinations (Supplementary Figure 2), ultrasound scan quality, MRI scan quality and PET-CT scan quality. Quality will be registered on a 5-point Likert scale (1, Very poor quality - 5, excellent quality).

The following data will be registered from medical charts: patient age, sex, tobacco smoking, alcohol consumption, date of first clinical contact, date of MRI scan performed, date of PET/CT scan performed, date of available histopathology results and date of treatment initiation.

Ultrasound video clips and B-mode images will be stored for retrospective descriptive evaluation of the following characteristics:

• Tumor shape (smooth/rounded, irregular/lobular).
• Tumor border delineation (well-circumscribed, unclear).
• Tumor heterogeneity (homogeneous, heterogeneous).
• Doppler flow compared to contralateral side (decreased, equivocal, increased).
• Doppler flow characteristics (peripheral/circumferential, internal organized, internal chaotic)

Reference standard: Final histopathology biopsy results will define the presence of cancer as binary results (no cancer vs. cancer present). An expert head and neck pathologist will re-evaluate biopsies results if inconclusive. The presence of a tumor in a correct sub-location assessed by final Multidisciplinary Team (MDT) conference will be used to define correct tumor location. Greatest tumor size and T-stage will also be determined by final MDT decision, considering the clinical and radiological findings. For patients with available complete tumor resections, a secondary analysis will be performed using the histopathologic reported greatest tumor diameter as reference standard.