Background: Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) is an important modality to obtain tissue diagnosis from mediastinal, pancreatic, and intra-abdominal lesions in close proximity to the pulmonary and gastrointestinal tract. It is considered to be a relatively safe, rapid, and minimally invasive technique with low complication rates.
Aims: To determine the diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and outcome of EUS-FNA, with histological correlation where applicable.
Methods: Data of all 1059 consecutive patients who underwent EUS-FNA from 1 January 2005 to 31 December 2017 at Shaukat Khanum Memorial Cancer Hospital and Research Centre, Lahore was reviewed in this retrospective study. The major sites that were targeted for EUS-FNA were pancreatic (423), mediastinal (376), and other abdominal lesions (260).
Results: The average number of passes per patient was 2.22. Rapid on-site evaluation (ROSE) was adequate in 969 patients (91.4%). Concordance between ROSE and final cytology was 99.5%. Follow-up was available in 810 patients (76.4%). The overall diagnostic yield was 94.3%. Ancillary studies, including immunohistochemical stains and flow cytometry, helped to increase the diagnostic yield from 78.1% to 94.3%. The overall sensitivity, specificity, PPV, NPV, and diagnostic accuracy for EUS-FNA were 94.8%, 98.6%, 99.9%, 65.5%, and 95.1%, respectively. Seven of 1059 patients (0.6%) developed complications.
Conclusion: EUS-FNA is a very sensitive and specific diagnostic tool with a minimal complication rate. Ancillary studies helped to increase the sensitivity, as well as the diagnostic yield.
Keywords: ancillary studies; endoscopic ultrasound-guided fine-needle aspiration; rapid on-site evaluation.
© 2020 Wiley Periodicals LLC.