Dynamic telecytopathology of on site rapid cytology diagnoses for pancreatic carcinoma

BACKGROUND Diagnosis of pancreatic lesions can be accurately performed by endoscopic ultrasound guided fine needle aspiration (EUS-FNA) with onsite cytopathologists to assess specimen adequacy and to determine a preliminary diagnosis. Considerable time is needed to perform on-site assessments. This takes away work time of cytopathologists and prohibits them from serving remote locations. It is therefore logical to ask if real-time telecytopathology could be used to assess specimen adequacy and if telecytopathology diagnosis has the same level of agreement to the final diagnosis as that of onsite evaluation. In this study, we compare agreement between cytodiagnoses rendered using telecytopathology with onsite and final interpretations. METHOD 40 Diff-Quik-stained EUS-FNA were re-evaluated retrospectively (patient ages 31-62, 19:21 male:female, 15 non-malignant lesions, 25 malignant lesions as classified by final diagnosis). Each previously assessed by a cytopathologist and finally reviewed by the same or different cytopathologist. Blinded to the final diagnosis, a resident pathologist re-screened all slides for each case, selected a slide and marked the diagnostic cells most representative of the lesion. Blinded to the diagnosis, one cytopathologist assessed the marked cells through a real time remotely operated telecytopathology system (MedMicroscopy). Diagnosis and time spent were recorded. Kappa statistic was used to compare agreements between telecytopathology vs. original onsite vs. final diagnoses. RESULTS Time spent for prescreening ranged from 1 to 5 minutes (mean 2.6 +/- 1.3 minutes) and time spent for telecytopathology diagnosis ranged from 2-20 minutes (mean 7.5 +/- 4.5 minutes). Kappa statistics, K, was as follows: telecytopathology versus onsite diagnosis K, 95% CI = 0.65, 0.41-0.88, for telecytopathology versus final K, 95% CI = 0.61, 0.37-0.85 and for onsite diagnosis versus final K, 95% CI = 0.79, 0.61-0.98. There is no significant difference in agreement between onsite and telecytopathology diagnoses. Kappa values for telecytopathology were less than onsite evaluation when compared to the final diagnosis; however, the difference was not statistically significant. CONCLUSION This retrospective study demonstrates the potential use of telecytopathology as a valid substitute for onsite evaluation of pancreatic carcinoma by EUS-FNA.