Implementation of microarray analysis for oncology samples: Effectiveness for detection of both copy number changes and copy-neutral loss of heterozygosity. S. Schwartz1, R. Burnside1, I. Gadi1, V. Jaswaney1, E. Keitges2, A. Penton1, K. Phillips1, H. Risheg2, J. Schleede1, J. Tepperberg1, P. Papenhausen1 1) Laboratory Corporation of America, Research Triangle, NC; 2) DynaCare/Laboratory Corporation of America, Seattle, WA.

   Over the past six years cytogenetic analysis has undergone a renaissance with microarray analysis utilized for the delineation of constitutional abnormalities. However, in oncology the acceptance of microarray analysis has been slower. We have studied over 4,000 oncology patients (including MDS, CLL, AML, ALL and MM), utilizing reporting cut-offs of 1 Mb for deletions and 2 Mb for duplications. However, changes as low as 50 kb were reported when significant pathogenic genes were involved. This work has not only proven the importance and efficacy of utilizing a SNP/oligo platform, but has yielded some illuminating results including: (a) The importance of establishing clear reporting cut-offs based on 70,000 previous pediatric studies; (b) The consideration that each hematologic disorder needs to be approached individually, some as an adjunct procedure (AML, ALL) and some as a primary tool (MDS, CLL); (c) The effectiveness in delineating abnormalities in all hematologic groups when the karyotype was normal, ranging from ~18.1% of MDS patients to ~78.6% of ALL patients; (d) The usefulness in delineating additional important abnormalities, even after FISH/cytogenetics revealed an abnormality [~50% (MDS and CLL) to ~88% (myeloma and ALL)]; (e) The importance and high incidence of copy-neutral loss of heterozygosity (CN-LOH) in all hematologic disorders [~10.2% of patients (MDS) to ~40% in CMML]. The CN-LOH was seen most often as a single finding, but multiple CN-LOH findings as well as temporal expansions were also seen; (f) The efficacy of its use as a heme-diagnostic tool in MDS, in which array abnormalities were detected in ~18% of patients in which Flow/hematopathology were not diagnostic and in ~36% of patients where Flow/hematopathology findings were only suspicious; (g) The prognostic utility for all groups, but specifically in ALL, where the genotyping array differentiates good prognosis hyperdiploidy from a poor prognostic doubled hyperhaploid karyotype; (h) The frequency of variants of unknown significance, due to detection cut-offs was only ~3.6%; (i) The ability to resolve structurally abnormal chromosomes in all cases, in the ability to detect chromothripsis (associated with a poor prognosis) and for the discovery of oncogenic fusion genes; and (h) Lastly, this work showed that a firm knowledge of pediatric array findings is necessary as constitutional pathogenic syndromes and consanguinity have been detected at a higher frequency than expected.

You may contact the first author (during and after the meeting) at