Comparative Analysis of Telecytology Versus In-Person Rapid On-Site Evaluation for Fine Needle Aspiration (FNA) Specimens: Insights from a Single Institutional Study

Summary: Introduction: Fine needle aspiration (FNA) has become increasingly popular for diagnoses of various conditions as they are minimally invasive and cost-effective. Rapid on-site evaluation (ROSE) of FNA biopsies ensures appropriate specimen triage to guide the next steps for diagnosis (biomarkers, flow cytometry, or other ancillary studies). At our institution, with the increase in locations where FNAs are performed, there was an increase demand for cytopathologist at these sites for ROSE. This high demand brought the need for an alternative way of evaluating specimens remotely, by telecytology (TC). Remote ROSE of FNA specimens saves time and costs while increasing productivity. However, many centers are not optimizing its use despite their availability. This study aims to compare the accuracy of ROSE performed in-person versus TC at our institution. Materials and Methods: A retrospective study of FNA cytology conducted at our institution across three locations from Jan 1st, 2022, to Dec 31st, 2023, was performed (Figure 1). Data including the specimen source, ROSE in person or by TC, ROSE interpretation, final cytology diagnosis, and ancillary tests used for diagnosis were collected from electronic medical records. Cases excluded from analysis were those without ROSE or those with only adequacy assessment. The concordance between ROSE interpretations and final diagnoses across six categories was determined for both, ROSE in person and ROSE by TC, and the diagnostic accuracy between both was compared. Results: During the two-year study period, there were 2,343 FNA cytology cases across the three locations, totaling 2,836 specimens. ROSE was utilized in 1,731 cases (61%), with 1252 (72%), 300 (17%), and 179 (10%) occurring at locations 1, 2, and 3, respectively. In-person ROSE accounted for 1490 (86%) of these cases, while TC was used in 241 (14%) of the cases (Figure 1). Overall diagnostic accuracy for in person ROSE in 1459 cases was 71% across all diagnostic categories, with 99% accuracy for malignant neoplasms. Major discrepancies (considered as benign-malignant or vice versa) were very low in both in-person (3%) and TC (5%) (Tables 1 and 3). For the 238 cases with ROSE by TC, diagnostic accuracy was 56% across all diagnostic categories (Tables 2 and 3). There was no difference observed in TC performance based on the location. Most discrepancies involved lymph nodes, although they were also the most frequent specimen source. Conclusions: i. TC proves to be an effective tool for ROSE, exhibiting similar performance to on-site microscopic review by pathologists. ii. The differences observed between the two compared groups (in-person ROSE vs. TC) are most likely due to the vast difference in the number of cases (1459 vs. 238 respectively). Increasing the number of TC performed may result in an increased statistical power for TC. iii. Major discrepancies between ROSE (either TC or in-person), and the final diagnosis were very low despite the added advantage of having ancillary testing available including cell block for subsequent immunohistochemical stains, flow cytometry, and molecular testing. iv. Incorporating TC for ROSE in our main campus (location 3) is planned due to its cost-effectiveness and comparable results with the traditional method.