Cervical vertebral and prevertebral pathologies – including retropharyngeal/prevertebral abscesses, spondylodiscitis and vertebral body lesions - are difficult to sample due to their deep location and proximity to vital neurovascular structures. Blood cultures are frequently negative, with a sensitivity ≤50% in deep neck or vertebral infections, making microbiological diagnosis uncertain and leading to frequent use of empirical antibiotics. CT-guided biopsy in the cervical region is technically challenging, offers limited real-time feedback, and carries procedural risks due to tight anatomical confines. Surgical biopsy is invasive and associated with higher morbidity. A safe, minimally invasive, real-time method for sampling cervical vertebral and prevertebral lesions—especially at high cervical levels (C2)—represents a major unmet clinical need.

We applied an innovative trans-pharyngeal and transesophageal EUS-guided sampling technique for cervical vertebral and prevertebral pathologies. Using a linear echoendoscope (FUJIFILM EG-580UT with ARIETTA system) introduced into the oropharynx and then the proximal esophagus under general anesthesia, the cervical vertebral column (C2–C7) was visualized in real time. Between March 2024 and July 2025, patients with suspected cervical vertebral or prevertebral pathologies difficult for interventional radiology sampling underwent EUS-guided sampling under general anesthesia with endotracheal intubation. In the left lateral position, a linear echoendoscope (FUJIFILM EG-580UT with ARIETTA 750) was introduced into the oropharynx and proximal esophagus, and lesions were aspirated or biopsied using 19G or 22G needles. Fluoroscopy was used in selected cases to confirm the location of the vertebral body. Samples were sent for bacterial culture (in aerobic blood culture bottles to maximize yield), cytology, histopathology, and GeneXpert MTB/RIF when tuberculosis was suspected. This technique extends the application of EUS—traditionally limited to thoracic and mediastinal structures—into the cervical spine, offering a minimally invasive diagnostic pathway for lesions previously inaccessible or unsafe to sample using radiology or surgery.

Nine patients underwent EUS-guided sampling (median age 60 years; range 37–77; 6 males, 3 females). All presented with neck pain; one had dysphagia and another stridor. Sampling was successfully achieved from C2 through C7, including the first documented EUS-guided sampling of a C2 vertebral body lesion. The following were the diagnostic yield. A) Tuberculosis (n=4): Caseating granulomas or MTB positivity on GeneXpert for TB or AFB staining. All patients responded to ATT alone with full clinical recovery. B) Pyogenic infections (n=3): Organisms included Acinetobacter baumannii, Enterobacter cloacae, and Klebsiella pneumoniae with Granulicatella adiacens. Two patients required surgical drainage; one was managed medically. C) Others (n=2): One culture-negative lesion with acute suppurative inflammation improved with prolonged antibiotics and drainage. Another non-diagnostic sample (only squamous cells) resolved with short oral antibiotics – likely self-resolving pathology. EUS confirmed TB in four and ruled it out in others. EUS-guided cultures enabled tailored antibiotic therapy. At least one patient avoided unnecessary surgery due to microbiological diagnosis via EUS. All patients improved clinically (follow-up range: 1.5–18 months). No procedural complications or adverse events occurred.

Transesophageal EUS-guided sampling is a safe, feasible, and clinically transformative technique for diagnosing cervical vertebral and prevertebral lesions. It significantly expands the diagnostic reach of EUS into the cervical spine—previously accessible only via invasive CT-guided or surgical methods—while providing high diagnostic yield with excellent safety. By enabling accurate pathogen identification, EUS supports evidence-based treatment decisions and can help avoid unnecessary surgery. This is the largest reported series of EUS-guided cervical vertebral sampling to date and includes the first documented C2 vertebral body case. Further multicenter studies are needed to validate and standardize this novel approach.