Telecytopathology facilitates the use of rapid on-site evaluation in endoscopic ultraound fine needle aspiration of the pancreas to improve patient outcomes

Abstract

Telecytopathology (TCP) has a variety of different applications in clinical practice and is becoming more widely utilized. Potential uses are broad and include second opinion consultation, primary diagnosis, educational/tumor board, and immediate fine-needle aspiration (FNA) interpretation. More recently, TCP use has been increasingly applied to the immediate evaluation of FNA biopsy adequacy evaluation. This process is generally known as rapid on-site evaluation (ROSE). Khurana et al. reported their findings using TCP for ROSE in endoscopic ultrasound (EUS) FNA of the pancreas.[1] They separated 217 patients into two cohorts. One cohort had on-site evaluation performed by a cytopathologist utilizing TCP. The other cohort did not have on-site evaluation. The TCP system utilized a commercially available passive live video from the microscope, obtained at the point of care (EUS gastrointestinal suite) and delivered to the attending cytopathologist in their office at a remote location. A cytotechnologist at the point of care operated the microscope. This was a retrospective study and the decision for ROSE versus non-ROSE was made by the gastroenterologist performing the endoscopy. Based on the on-site evaluation, it was possible for the TCP ROSE cohort to have additional passes performed, as deemed necessary for diagnosis. The majority of solid lesions were selected for TCP ROSE, and the majority of cystic lesions were not selected for TCP ROSE. The authors examined a variety of individual data points, including the nature of the lesion being evaluated (solid vs. cystic) and the diagnostic rates. The four major diagnostic categories included: (1) Nondiagnostic, (2) negative/benign, (3) atypical/suspicious, and (4) malignant. In solid lesions, the percentages for negative/benign and atypical/suspicious were similar in each cohort. Use of TCP ROSE favored solid lesions and had a nondiagnostic rate of 3.7% in comparison to a 25.6% nondiagnostic rate in the solid lesions without immediate evaluation. There was no statistical difference in nondiagnostic rates for cystic lesions between the two groups, although most did not have TCP ROSE (79/93). For all patients, after using multivariate logistic regression, the odds of a nondiagnostic sample were 6.9 times greater without the use of TCP ROSE.