Comparison of a selection of rapid automated DNA and RNA extraction technologies for detection of somatic or constitutional gene abnormalities in cancer diagnosis

Molecular analyses of large numbers of patient samples are increasingly used for diagnostic applications as well as for understanding cancer biology. Their accuracy depends on the quality and quantity of nucleic acids extracted from human cells or tissues. To optimize these preanalytical steps, we evaluated several automated technologies for nucleic acid purification from clinical samples. Three automated platforms were compared. DNA was extracted from peripheral blood leukocytes from five normal individuals, and its quality was assessed by D-HPLC and sequencing after PCR. Clinical samples from acute leukemia patients were used for automated RNA extractions; results were compared to our standard manual technique. RNA qualification was done using capillary gel electrophoresis and analysis of the Abelson gene transcript by real-time PCR. One robot produced higher total output for both DNA and RNA. While the quality of DNAs obtained from the three workstations allowed implementation of their analysis for detection of germinal mutations, important differences were observed in the quality of RNAs. One robot isolated RNA with similar quality and quantity to the manual technique, but the resulting products displayed low concentrations. A robust technique had therefore to be evaluated and validated to allow the implementation of this workstation within the daily diagnostic practices. This study is of interest at a time when hospital-based laboratories are bringing molecular signatures to the clinic.