Histology, cytology or both for sampling of solid pancreatic masses? Prospective evaluation of a new 22g biopsy needle

Introduction For solid pancreatic masses (SPM), EUS-fine needle aspiration cytology (FNA-C) has a variable diagnostic yield and sensitivity for diagnosis of malignancy, given the limited material, presence of inflammation and inherent difficulties in cytologic interpretation. Histology of formalin-fixed biopsies is simpler, cheaper and may allow extra marker studies more often. This study evaluated the ability of a new EUS biopsy needle (1) to obtain histological biopsies (FNB) with a limited number of passes; (2) to compare specimen adequacy and diagnostic sensitivity with FNA-C; (3) compare specimen quality between FNA-C and FNB. Methods Over 5 months, consecutive patients with SPM underwent EUS biopsy sampling using a standardised protocol. Patients in whom lesions were predominantly cystic, taking anticoagulants or in whom access with a 22 g needle was deemed impossible were excluded. Using a 22 g 'Sharkcore' needle (Medtronic, Dublin, Ireland), 4 passes were made, using slow stylet withdrawal and material expelled into Cytolyt cytology fixative (2 passes) or into formalin for histology processing. Sampling sequence was randomised (passes 1&3 into cytology with passes 2&4 into formalin, or vice versa). Pathological assessment was performed by one of 2 specialist GI pathologists using a structured proforma, to rate specimen adequacy, sample quality, diagnosis and subjective superiority. Results 43 patients were included (23 M, 20 F, mean age 66 years, range 46-83 y). One patient was excluded from analysis due to incomplete follow-up. Final diagnoses were pancreatic malignancy (35, 83.3%) and benign pancreaticobiliary disorders (7, 16.7%). Results (n = 42) are shown in Table 1: In 2 cases FNA-cytology was negative for malignancy but FNB histology was positive; in contrast there were no cases where cytology was positive but histology was negative. Subjectively, FNB histology samples were scored as superior in 28 cases (66.7%), compared to FNA-C samples (4, 9.5%). In the remainder, samples were judged to be of equivalent quality (10, 23.8%). Conclusion The 22 g Sharkcore needle provides histologicalquality samples in almost all cases of pancreatic or peripancreatic masses, with only 2 passes. Samples are more often superior to those obtained by cytology. This offers the prospect of [TABLE PRESENTED] shorter procedure times, easier sample processing and more reliable diagnosis.