Characterizing and classifying neuroendocrine neoplasms through microRNA sequencing and data mining

Neuroendocrine neoplasms (NENs) are clinically diverse and incompletely characterized cancers that are challenging to classify. MicroRNAs (miRNAs) are small regulatory RNAs that can be used to classify cancers. Recently, a morphology-based classification framework for evaluating NENs from different anatomic sites was proposed by experts, with the requirement of improved molecular data integration. Here, we compiled 378 miRNA expression profiles to examine NEN classification through comprehensive miRNA profiling and data mining. Following data preprocessing, our final study cohort included 221 NEN and 114 non-NEN samples, representing 15 NEN pathological types and five site-matched non-NEN control groups. Unsupervised hierarchical clustering of miRNA expression profiles clearly separated NENs from non-NENs. Comparative analyses showed that miR-375 and miR-7 expression is substantially higher in NEN cases than non-NEN controls. Correlation analyses showed that NENs from diverse anatomic sites have convergent miRNA expression programs, likely reflecting morphologic and functional similarities. Using machine learning approaches, we identified 17 miRNAs to discriminate 15 NEN pathological types and subsequently constructed a multi-layer classifier, correctly identifying 217 (98%) of 221 samples and overturning one histologic diagnosis. Through our research, we have identified common and type-specific miRNA tissue markers and constructed an accurate miRNA-based classifier, advancing our understanding of NEN diversity.Diagnostic histopathology and small RNA cDNA library preparation information for all samples are presented in Table S1 of the associated manuscript. Funding provided by: Academic Health Sciences Center Alternative Funding Plan Innovation FundCrossref Funder Registry ID: Award Number: Funding provided by: Canada Foundation for Innovation John R Evans Leaders FundCrossref Funder Registry ID: Award Number: Funding provided by: Carcinoid and Neuroendocrine Tumor Society CanadaCrossref Funder Registry ID: Award Number: Funding provided by: Ontario Research Fund-Research InfrastructureCrossref Funder Registry ID: Award Number: Funding provided by: Southeastern Ontario Academic Medical OrganizationCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100010661Award Number: Funding provided by: Robertson Therapeutic Development*Crossref Funder Registry ID: Award Number: Funding provided by: Rockefeller UniversityCrossref Funder Registry ID: http://dx.doi.org/10.13039/100012007Award Number: UL1TR001866Funding provided by: Ontario Institute for Cancer ResearchCrossref Funder Registry ID: http://dx.doi.org/10.13039/100012118Award Number: Funding provided by: Robertson Therapeutic DevelopmentCrossref Funder Registry ID:Sequencing-based miRNA expression profiles from 378 clinical samples, comprising 239 neuroendocrine neoplasm (NEN) cases and 139 site-matched non-NEN controls, were used in this study. Expression profiles were either compiled from published studies (n=149) or generated through small RNA sequencing (n=229). Prior to sequencing, total RNA was isolated from formalin-fixed paraffin-embedded (FFPE) tissue blocks or fresh-frozen (FF) tissue samples. Small RNA cDNA libraries were sequenced on HiSeq 2500 Illumina platforms using an established small RNA sequencing (Hafner et al., 2012 Methods) and sequence annotation pipeline (Brown et al., 2013 Front Genet) to generate miRNA expression profiles. Scaling our existing approach to miRNA-based NEN classification (Panarelli et al., 2019 Endocr Relat Cancer; Ren et al., 2017 Oncotarget), we constructed and cross-validated a multi-layer classifier for discriminating NEN pathological types based on selected miRNAs