1. Variability in interpreting and reporting copy number changes detected by array-based techology in clinical laboratories K.D. Tsuchiya 1, L.G. Shaffer 2, S. Aradhya 3, J.M. Gastier-Foster 4, A. Patel 5, K. Rudd 6, J. Sanford Biggerstaff 7, W.G. Sanger 8, S. Schwartz 9, J.H. Tepperberg 9, E.C. Thorland 10, B.A. Torchia 2, A.R. Brothman 11 1 Dept of Laboratories, Seattle Children’s Hosp and Dept of Lab Med, Univ of Washington School of Med, Seattle, WA; 2 Signature Genomic Laboratories, Spokane, WA; 3 GeneDx, Gaithersberg, MD; 4 Dept of Path and Lab Med, Nationwide Children’s Hosp, Columbus, OH; 5 Dept of Molec and Hum Genet, Baylor College of Med, Houston, TX; 6 Dept of Hum Genet, Emory Univ School of Med, Atlanta, GA; 7 Idaho Cytogenet Diag Lab, Boise, ID and PAML, Spokane, WA; 8 Dept of Peds, Munroe-Meyer Inst, Univ of Nebraska Med Ctr, Omaha, NE; 9 Lab Corp of America, Res Triangle Park, NC; 10 Dept of Lab Med and Path, Mayo Clinic, Rochester, MN; 11 Depts of Peds, Hum Genet and Path, Univ of Utah, and ARUP Laboratories, Salt Lake City, UT INTRODUCTION The use of microarrays for the assessment of copy number changes (CNCs) in individuals with constitutional disorders is becoming widespread in clinical cytogenetics laboratories. Although this technology has significantly increased the detection of chromosome imbalances that are known to be associated with abnormal phenotypes, the finding of CNCs in healthy individuals has created new challenges for laboratorians and clinicians who must interpret the clinical significance of previously undescribed CNCs. The purpose of this study was to assess the variability between different laboratories in the interpretation and reporting of CNCs detected by array-based technology. METHODS RESULTS Thirteen CNCs that were detected by array comparative genomic hybridization (six by BAC array and seven by oligonucleotide array) in one clinical laboratory were evaluated by directors from ten other laboratories to determine how they would interpret and report the findings. The CNCs that were chosen for the study were not associated with a known abnormal phenotype in the literature. The CNCs were chosen because they potentially posed an interpretive challenge in terms of their clinical significance. Participants were provided with the clinical indication for testing, the results of other genetic testing, the nucleotide coordinates of the CNC, any adjacent gaps in probe coverage, and results of FISH confirmation. Two representative CNCs used in this survey are shown in Table 1 and Fig. 1. For each CNC, participants were asked to fill out a questionnaire. Questionnaire for each case Assuming no known clinically significant CNCs were detected in this case, how would you report the finding? Normal Abnormal Uncertain clinical significance (UCS) Other Which of the following criteria were used to determine how you chose to report this finding? Check all that apply Size of gain/lossSize of gaps in coverage adjacent to gain/loss Presence or absence of gene(s) in region of gain/loss Presence of specific gene(s) in region of gain/loss - which one(s)________ Presence of genes within gaps in coverage adjacent to gain/loss Clinical indication Gain/loss was confirmed by FISHGain/loss was not confirmed by FISH Presence of benign CNCs in personal/public databases Lack of benign CNCs in personal/public databases CONCLUSIONS There was not complete concordance of classification for any of the 13 CNCs (Table 2). Responses ranged from normal to abnormal for some CNCs. Some laboratories appear to use a binary reporting system (normal vs. abnormal), whereas other laboratories use three or more categories for reporting. Duplications were just as likely to be categorized as UCS or abnormal when compared with deletions. Small imbalances containing multiple genes were just as likely to be categorized as UCS compared to larger imbalances. CNCs with negative or absent FISH results were just as likely to be classified as UCS or abnormal as those with FISH confirmation. Table 1. Information provided to participants from two representative cases Array Clinical indication CNC Gaps FISH BAC Omphalocele, mega- cisterna magna, hypoglycemia 2 overlapping clone gain, 3p26.3: 955368-1149577 (0.19 Mb) Tel: 0.44 Mb Cen: 0.24 Mb Dup visualized by interphase OligoLanguage delay, dysmorphic, dwarfism 9 adjacent probe loss, 16p13.3: 1942265-2015940 (0.07 Mb) Tel: 0.01 Mb Cen: 0.012 Mb Del visualized Figure 1. Array CGH and FISH results from the two representative cases in Table 1. A. Partial chromosome 3 array plot (left panel), interphase (middle) and metaphase (right) FISH images from the BAC case in Table 1. B. Partial chromosome 16 array plot (left) and metaphase FISH image (right) from the oligonucleotide case in Table 1. Table 2. Participant responses for the two representative cases in Table 1 Participants 1234567891011 3p26.3 gain 16p13.3 loss NL UCSABNNL UCSNLUCS NL LB NL = normal, UCS = uncertain clinical significance, ABN = abnormal, LB = likely benign There is variability between clinical laboratories in the interpretation of CNCs detected by array-based technology There is a need for better guidelines for interpreting and reporting CNCs to clearly and more consistently communicate the clinical significance of these findings to ordering providers