Dynamic telecytopathology is comparable to rapid onsite evaluation (ROSE) of endoscopic ultrasound fine needle aspirates: results of a prospective pilot study

Background: Endoscopic Ultrasound (EUS) with fine needle aspiration (FNA) is a highly accurate method to diagnose lesions in even the most remote aspects of the thorax and peritoneum. Nonetheless, it is invasive and requires continuous patient sedation, thus real time assessment of specimen quality is imperative. ROSE has been demonstrated to correlate highly with final cytologic interpretations and improves the diagnostic yield of EUS-FNA However, due to limited time and financial resources its availability is variable across centers. The objective of this prospective study was to evaluate whether remote telecytopathology can substitute for ROSE.

Methods: EUS fine needle aspirates were prospectively evaluated by three methods. ROSE was performed by a cytopathologist in the procedure suite; simultaneously dynamic telecytopathology was done by a different cytopathologist in a remote location at our institution. Subsequently, cytologic interpretation was made by the same or different cytopathologist in the laboratory which was the gold standard.

Telecytopathology was performed using an Olympus microscope system (BX) coupled with a digital camera (DP72 - 12.8 mega pixel cooled digital color camera). The transmitted image resolution is 800×600 mega pixel. The system broadcasts a streaming live image over the Internet via a static IP address and does not require special hardware to view the image by the remote viewer. It only requires an internet connection and a web browser. A pathologist can then view either a live image or a full resolution captured image. Communication can be facilitated by a phone or a web based communication system.

Results: Twenty-five samples were obtained from participants 40–87 years (mean age=64.5, 48% male). Seventeen (86%) of the samples were from the primary site of involvement including the pancreas (N=14), rectum (N=1), stomach (N=1), and bile duct (N=1). The remainder were from distant sites: mediastinum (N=2), adrenal gland (N=2), and lymph nodes (N=5). Final cytopathology confirmed that fifteen samples were malignant (60%), one was interpreted as atypical, and the rest were benign. There was 88% (22 of 25 cases) agreement between dynamic telecytology and final bench cytology, the gold standard (p<0.001) and 92% (23 of 25 cases) agreement between ROSE and bench cytology (p<0.001). There was no significant difference in agreement between dynamic telecytopathology and ROSE (p-value for McNemar's c2=1.0). Cohen's kappa for agreement for telecytopathology and ROSE cytology was 0.80 with a standard error of 0.11, confirming excellent and robust correlation between both methods.

Conclusion: This prospective study demonstrates that the accuracy of dynamic telecytopathology is comparable to that of ROSE in the assessment of EUS acquired fine needle aspirates. If confirmed by larger trials, this system will obviate the need for onsite interpretation of EUS-FNA specimens.

Acknowledgements: COI: We wish to thank Olympus America for providing the necessary equipments to make this project feasible.