Diagnostic Concordance of Cytology and Histology in Samples Obtained via Endoscopic Ultrasound-Guided Fine-Needle Biopsy (EUS-FNB)

Introduction Endoscopic ultrasound (EUS)-guided fine-needle aspiration and biopsy (FNA/FNB) to obtain cytological aspirates and histological core samples, respectively, are the standard of care for diagnosing lesions in/adjacent to the upper/lower gastrointestinal tract. Due to the lack of standardization of tissue processing, it is unclear whether core samples should be sent only for histology (formalin) or cytology (CytoLyt), or both. The aim of this study was to investigate the diagnostic concordance rates between cytology and histology on EUS-FNB core samples. Methods A total of 227 patients underwent EUS-FNB between October-2017 and February-2019 by a single therapeutic endoscopist; 44 core-tissue samples (41 patients) were placed alternately in CytoLyt (cytology) and formalin (histology), with equal passes into each, to best achieve a proportionate sample amount. The patient's demographics, medical history, pertinent imaging, EUS indication/findings were reviewed. Main outcomes included concordance rates between cytology-histology and diagnostic accuracy for malignancy. Results Cytology and histology were discordant in five cases (11.5%); four with negative cytology but a definite diagnosis of malignancy achieved with histology. One case was suspected as neoplasm on cytology but further characterized as benign on histology. Cytology failed to sub-characterize an additional four mass-like pancreatic benign entities, due to inadequate tissue architecture assessment in the CytoLyt sample. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of cytology for diagnosis of malignancy were 87.88% (95%CI: 71.8-96.6), 90.91% (95%CI: 58.7-99.7), 96.67% (95%CI: 81.6-99.4), and 71.43% (95%CI: 49.4-86.4). Discussion We observed 11.5% diagnostic discordance between cytology and histology on EUS-FNB core samples, with histology being superior. Future multicenter prospective randomized studies are needed to establish an accurate and cost-effective diagnostic process.