Diagnostic challenges of intra-operative frozen consultation for gastrointestinal signet ring cell carcinoma

Aims Signet ring cell carcinoma (SRCC) is challenging to recognise on intra-operative frozen sections, with known high false-negative rates. The objective of this study was to investigate common factors contributing to discrepancies between intra-operative frozen diagnoses and those made upon review of permanent sections, and summarise our experiences gained and lessons learned on minimising errors on intra-operative frozen diagnoses of gastrointestinal SRCC. Methods and results We retrospectively examined our pathology database from 25 May 2000 to 1 January 2018 and re-reviewed intra-operative frozen sections and permanent haematoxylin and eosin (H&E) slides for specimens confirmed with SRCC on permanent sections. This study includes 83 specimens taken from 50 patients, with an accuracy of 85.5%. Main common factors causing discordance or deferral in recognising SRCC between intra-operative frozen procedures and permanent sections include: (i) resemblance of clusters of SRCC cells with a myxoid background; (ii) disguise as normal or reactive cells (histiocytes, macrophages, large reactive lymphocytes, plasma cells or adipocytes) due to their relatively clear or depleted cytoplasmic mucin; and (iii) histological sampling errors, leading to misses of small foci of SRCC on frozen section slides. Conclusions An accurate diagnosis of SRCC during intra-operative frozen consultations remains challenging. Based on our experiences and lessons, the most important strategies to reduce diagnostic errors are: (i) understanding the unusual histomorphological features of SRCC cells on frozen sections including, but not limited to, intracellular mucin depletion, absence of desmoplasia and no adjacent pre-cancer changes; and (ii) close attention to abrupt transition from normal architecture (e.g. glandular or submucosal loose connective tissue) to myxoid and/or inflammatory-like appearance, which potentially harbours SRCC.