David Gentien1, Laure Piqueret-Stephan2,3, Emilie Henry1, Benoît Albaud1, Audrey Rapinat1, Serge Koscielny4, Jean-Yves Scoazec3,5, Philippe Vielh6,7,8. 1. Translational Research Department, Genomics Platform, Institut Curie, PSL Research University, Paris, France. 2. INSERM UMR 981, Villejuif, France. 3. Translational Research Laboratory, AMMICa (CNRS UMS3655, INSERM US23, Paris Sud University) Gustave Roussy, Villejuif, France. 4. Department of Biostatistics, Gustave Roussy, Villejuif, France. 5. Department of Medical Biology and Pathology, Gustave Roussy, Villejuif, France. 6. INSERM UMR 981, Villejuif, France, ph.vielh@outlook.com. 7. Translational Research Laboratory, AMMICa (CNRS UMS3655, INSERM US23, Paris Sud University) Gustave Roussy, Villejuif, France, ph.vielh@outlook.com. 8. Department of Medical Biology and Pathology, Gustave Roussy, Villejuif, France, ph.vielh@outlook.com.
Abstract
OBJECTIVE: Although transcriptomic assessments of small samples using high-throughput techniques are usually performed on fresh or frozen tissues, there is a growing demand for those performed on stained cellular specimens already used for diagnostic purposes. STUDY DESIGN: The possibility of detecting mRNAs and microRNAs (miRNAs) from routinely processed cytological samples using nCounter® technology was explored. Fresh samples from pleural and peritoneal effusions were analyzed using 2 parallel methods: samples were smeared and routinely stained using the May-Grünwald-Giemsa or Diff-Quik® method and mounted using conventional methods, and they were also studied following a snap freezing method, in which samples were maintained at -80°C until use. mRNAs and miRNAs were assessed and compared after total RNA extraction from both routinely processed samples and their matched frozen controls. RESULTS: A good concordance was found between the gene expression measured in routinely processed samples and their matched frozen controls for the majority of mRNAs and miRNAs tested. However, the standard deviation of low-expressed miRNA was high. CONCLUSIONS: Although nCounter® technology is a robust method to measure and characterize both mRNAs and miRNAs from routinely processed cytological samples, caution is recommended for the interpretation of low-expressed miRNA.
OBJECTIVE: Although transcriptomic assessments of small samples using high-throughput techniques are usually performed on fresh or frozen tissues, there is a growing demand for those performed on stained cellular specimens already used for diagnostic purposes. STUDY DESIGN: The possibility of detecting mRNAs and microRNAs (miRNAs) from routinely processed cytological samples using nCounter® technology was explored. Fresh samples from pleural and peritoneal effusions were analyzed using 2 parallel methods: samples were smeared and routinely stained using the May-Grünwald-Giemsa or Diff-Quik® method and mounted using conventional methods, and they were also studied following a snap freezing method, in which samples were maintained at -80°C until use. mRNAs and miRNAs were assessed and compared after total RNA extraction from both routinely processed samples and their matched frozen controls. RESULTS: A good concordance was found between the gene expression measured in routinely processed samples and their matched frozen controls for the majority of mRNAs and miRNAs tested. However, the standard deviation of low-expressed miRNA was high. CONCLUSIONS: Although nCounter® technology is a robust method to measure and characterize both mRNAs and miRNAs from routinely processed cytological samples, caution is recommended for the interpretation of low-expressed miRNA.