ABSTRACT
OBJECTIVE: This study aims to validate the existence of a microbiome within intraductal papillary mucinous neoplasm (IPMN) that can be differentiated from the taxonomically diverse DNA background of next-generation sequencing procedures. DESIGN: We generated 16S rRNA amplicon sequencing data to analyse 338 cyst fluid samples from 190 patients and 19 negative controls, the latter collected directly from sterile syringes in the operating room. A subset of samples (n=20) and blanks (n=5) were spiked with known concentrations of bacterial cells alien to the human microbiome to infer absolute abundances of microbial traces. All cyst fluid samples were obtained intraoperatively and included IPMNs with various degrees of dysplasia as well as other cystic neoplasms. Follow-up culturing experiments were conducted to assess bacterial growth for microbiologically significant signals. RESULTS: Microbiome signatures of cyst fluid samples were inseparable from those of negative controls, with no difference in taxonomic diversity, and microbial community composition. In a patient subgroup that had recently undergone invasive procedures, a bacterial signal was evident. This outlier signal was not characterised by higher taxonomic diversity but by an increased dominance index of a gut-associated microbe, leading to lower taxonomic evenness compared with the background signal. CONCLUSION: The 'microbiome' of IPMNs and other pancreatic cystic neoplasms does not deviate from the background signature of negative controls, supporting the concept of a sterile environment. Outlier signals may appear in a small fraction of patients following recent invasive endoscopic procedures. No associations between microbial patterns and clinical or cyst parameters were apparent.