Our Current Knowledge of Genetics and Genomics Unlocking the Diagnostic Potential of Genomic Medicine The Beaumont DNA Array Experience Mark Micale, Ph.D., F.A.C.M.G. Medical Director, Clinical Cytogenomics Laboratory Department of Pathology and Laboratory Medicine Beaumont Health System Associate Professor of Pathology and Laboratory Medicine Oakland University William Beaumont School of Medicine The Anatomy Lesson of Dr. Nicolaes Tulp Rembrandt 1632 Karyotyping Has Been the Gold Standard for Identifying Chromosome Abnormalities The era of modern cytogenetics began in 1956 with the discovery by Tijo and Levan that the human diploid chromosome number was 46 In 1959, Lejeune reported on the chromosomal basis of Down syndrome In 1960, Nowell and Hungerford demonstrated an association between chromosomes and cancer with the discovery of the Philadelphia chromosome in chronic myeloid leukemia 1959-1966: Additional associations between chromosome abnormalities and specific phenotypes were made (Turner, Klinefelter, Cri-duchat, Patau, Edward, Bloom, Fanconi anemia) Conventional cytogenetics (karyotyping) represents a whole-genome test but has somewhat limited resolution (5-10Mb) Karyotyping is great at identifying larger chromosome abnormalities, but is not good at identifying smaller rearrangements associated with intellectual disability, multiple congenital anomalies, and autism/asd Molecular Cytogenetics (FISH) Provides High-Resolution Targeted Genomic Analysis FISH was introduced into the clinical lab in the late 80s, and prenatal FISH aneuploidy detection became routine in the mid 1990s FISH is far more sensitive than karyotyping, but only interrogates a specific genomic region of interest FISH is a great test in those clinical settings whereby the phenotype provides some differential diagnosis: - Common microdeletion/microduplications syndromes - Fetal chromosome aneuploidy - FISH panels for hematolymphoid disorders; detection of specific rearrangements in AML, CML, etc. Resolution of Conventional and FISH Chromosome Analysis Increasing the Sensitivity of Genomic Assays Permits a Higher Resolution Study of the Human Genome Resolution Coverage Karyotyping >5-10Mb Complete FISH (interphase) >20kb Probe specific FISH (metaphase) >100kb Probe specific Chromosome 50kb/SNP 1 bp Complete microarray/single nucleotide polymorphism (SNP) array analysis Chromosome Microarray (CMA) or Single Nucleotide Polymorphism (SNP) Analysis Other highly-sensitive molecular assays: Massively parallel sequencing, exome or whole-genome sequencing (NextGen sequencing), mutation detection 1
Chromosome Microarray Analysis (CMA) Molecular Karyotyping A Normal CMA Study Analyzes multiple genetic loci at higher resolution than available by conventional cytogenetics (100kb vs. 5-10Mb) Compares DNA content between two differentially labeled chromosomes patient and reference DNA CMA Results - Characterization of Chromosome 18q21.2 Deletion CMA in the Diagnosis of Constitutional Genetic Syndromes DiGeorge Syndrome 2.47 Mb 22q11.2 deletion Prader Willi/Angelman Syndrome 5.81Mb 15q11.2q13.1 deletion 18q21.2 2.5Mb deletion 73 probes 10 genes Potocki-Lupski Syndrome 3.39Mb 17p11.2 duplication My Very Own CMA Abnormality - Chromosome 18q23 Duplication CMA Sensitivity is Based on Array Element Distribution 18q23 copy gain ~210.51kb duplication 24 probes 2 genes CTDP1 gene KCNG2 gene 2
CMA Characterization of 2.4Mb Chromosome 10 Deletion [del(10)(q25.1q25.1)] Low-resolution array High-resolution array Aston E et al (2008): J Med Genet 45, 268-274 SCIENCE 305(23): JULY 2004, 525 Copy Number Variants (CNVs) Genomic deletions or duplications that may represent rare or common benign variants or pathologic variants (1kb - several Mb in size) May encompass as much as 12% of the genome CNV may be clinically significant in child but have no phenotype in carrier parent - Addl environmental an genomic factors needed - CNV may be variably penetrant By constructing arrays that eliminate highly variable regions, frequency of benign variants can be minimized Currently more of a problem with oligo arrays Most resolved by parental studies Databases are needed for proper interpretation of CNVs Proprietary, Open access Database of Genomic Variants This is a problem for prenatal CMA Is a CNV Benign or Pathogenic? Size Location Genomic content Comparison with other cases Inherited or de novo Example of a Benign CNV Is a CNV Benign or Pathogenic? Benign 518.66kb chromosome 9p23 deletion. Notice that there are no genes in the deleted region. Miller DT et al (2010), AJHG 86, 749-764 3
CMA Has Been Transformational to the Practice of Clinical Genetics Child presents with developmental delay and microcephaly Am J Hum Genet 86, 749-764, 2010. Genet Med 12(11), 2010 CMA Identifies Three Genomic Abnormalities Resulting in Gain of 50 Genes 42.95kb Chromosome 22q Deletion Involving the SHANK3 Gene Derivative chromosome 4 with a 2.15 Mb single copy gain involving chromosome 4q21.23 with the likely formation of a neocentromere Derivative chromosome 8 with a 5.91 Mb single copy gain involving 8p11.21-8q11.1 as well as a 196.67 kb two copy gain within that segment (band 8p11.21). Derivative X chromosome with (a) a single copy gain of a 1.28 Mb segment involving Xp11.22 (HSD17B10, HUWE1, and PHF8 genes all associated with intellectual disability) and (b) a single copy gain of a 5.07Mb segment involving the pericentromeric region of the X chromosome Seven year old referred for developmental delay, ADHD, and speech apraxia 42.95kb deletion identified that knocks out one gene- SHANK3 This defines the Phelan- McDermid (22q13.3 microdeletion) syndrome; the deletion explains the child s phenotype Identification provides accurate genetic counseling and accurate risk assessment in future pregnancies CMA identifies a chromosome 3p triplication in a child with developmental delay CMA has Identified Novel Microdeletion/Microduplication Syndromes Ballif BC et al (2007): Discovery of a previously unrecognized microdeletion syndrome of 16p11.2-p12.1. Nat Genet 39(9), 1071-1073 Shaffer LG et al (2007): The discovery of microdeletion syndromes in the post-genomic era: review of the methodology and characterization of a new 1q41q42 microdeletion syndrome. Genet Med 9(9), 607-616. 4
CMA studies on children with developmental delay, mental retardation, and multiple congenital anomalies Beaumont Constitutional CMA Experience Reported diagnostic yield of CMA is 8-17% depending on methodology and selected patient population; most of these patients had a normal karyotype. Number of studies performed: Number of cases with one or more clinically significant abnormalities: 330 (October 2009-present) 74 (22.4% yield) CMA detects de novo CNVs in 10% of patient s with autism (16p11.2 autism locus). CMA identifies more microdeletions than microduplications (may be technical or patient selection bias due to milder phenotype in duplications) Determination of the clinical significance of CNVs remains an issue Smallest imbalance: 42.95 kb (SHANK3 gene - Phelan McDermid Syndrome) Largest Imbalance: 32.8 Mb (Cri-du-Chat syndrome) 16p deletion/duplication: 6/74 cases 8p inverted duplication/deletion syndrome: 4/74 cases Rare genetic syndromes identified: Pitt-Hopkins syndrome (535kb) Opitz/GBBB syndrome (140.58kb) Number of cases with abnormality(s) of uncertain clinical significance: 54 (16.4%) Prenatal Cytogenetic Analysis Conventional karyotyping detects chromosome abnormalities in: - 35% of pregnancies with fetal ultrasound abnormalities (All) - 2-3% of pregnancies presumed to be at-risk - 40% of prenatally ascertained heart defects (22q11.2 FISH) - 50% of pregnancies with cystic hygroma - 4.5-35% of pregnancies with omphalocele - >50% with nasal bone absence - 40% of pregnancies with nuchal fold (2 nd trimester) The clinical sensitivity of prenatal interphase FISH (for rapid aneuploidy detection) in the AMA population approaches 80%, but falls to 65-70% for all prenatal patients Classical cytogenetics (karyotyping) is not good at detecting small (<5Mb) deletions or duplications that may be clinically significant. FISH is great at detecting such imbalance, but only if there are clinical cues to guide appropriate choice of DNA probe used But what about a fetus with U/S anomalies and a normal karyotype? Will Prenatal Array Replace the Karyotype? NICHD study of 4,450 unselected prenatal cases from 29 centers (4 reference labs) - Normal karyotype with U/S anomalies: addl. 6.0% of cases - AMA: 1.7% - Screen positive: 1.6% - Variants of uncertain clinical significance (VOUS): 1.5% Emory University (ISCA Consortium) Experience - Normal karyotype with U/S anomalies (n=755): addl. 6.0% of cases - AMA (n=1,966): 1.7% - Screen positive (n=729): 1.7% - VOUS: 1.6% LabCorp Experience (>1,600 specimens) - 6.1% pathogenetic CNVs identified - 3.0% consanguinity, 3.6% UPD - VOUS: 1.3% Signature Genomic Laboratories - Normal karyotype with U/S anomalies: addl. 6.6% of cases - AMA: 0.3% - Screen positive: 5.2% - VOUS: 1.0% when accounting for parental inheritance Normal Karyotype With Abnormal CMA Result Utility of Prenatal Array in Pregnancies With Ultrasound Anomalies Amniocentesis for ventriculomegaly and agenesis of the corpus callosum 46,XX CMA done on 9 week old child for multiple congenital anomalies, ACC, and lack of normal development 2.88Mb chromosome 1q44 deletion which includes the 1qter critical region for corpus callosum agenesis/hypogenesis LabCorp (n=>1600) Signature Genomics (n=3,876) U/S Finding Positive Findings U/S Finding Positive Findings Single abn. 5.6% Single abn. 5.6% Multiple abns. 15.3% Multiple abns. 9.8% Brain 9.6% Abn. Of growth 2.6% Kidney 10.9% > One soft sign 2.6% Cardiac 6.8% Holoprosencephaly 11% Posterior fossa defect 14.6% Skeletal anomalies 10.8% VSD 11.4% Hypoplastic left heart 16.2% Based on these studies, ACOG will likely recommend that array should be used as a first-tier test in certain prenatal clinical situations 5
Considerations in Prenatal CMA Analysis Applications of SNP Array Analysis in Oncology The parental anxiety created by identifying CNVs of unknown significance The need to do parental CMA studies to further characterize uncertain results Familial variants are not always benign due to reduced penetrance and variable expressivity The need for pretest and posttest genetic counseling The identification of a genomic imbalance that is not related to the reason for referral nor to the observed phenotype, but rather to late-onset disorders, infertility, neurological, or cancer predisposition loci Will (and should) CMA replace standard karyotyping and FISH, and when? Chronic lymphocytic leukemia Myelodysplastic syndrome Plasma cell myeloma Acute lymphoblastic leukemia Acute myeloid leukemia Renal cell carcinoma Meduloblastoma Glial tumors Neuroblastoma SNP Oligonucleotide Arrays SNP arrays identify a set of genetic variants (single nucleotide polymorphisms) in an individual; Estimated one SNP/100-300 bp Probes on a SNP array consist of oligonucleotides that contain SNPs; SNP arrays can distinguish parental alleles at heterozygous loci 9.2 million SNPs have been reported (HapMap); 300,000-600,000 SNPs contain most of the information about the patterns of genetic variation SNP genotyping permits genome-wide high-resolution detection of copycopy number changes (Affymetrix CytoHD SNP array with 2.697 million DNA markers) SNP Arrays Detect Copy Number Neutral Loss of Heterozygosity (aupd) A B A A SNP C T C C Mitotic Recombination Error Resulting in Segmental UPD UPD has been described in CLL, PCV leading to JAK2 homozygosity, MDS/MPD, AML, and follicular lymphoma Principles of Copy Number Detection and Genotyping Using SNP Arrays The Power of SNP Array Analysis A Case of Mantle Cell Lymphoma Sato-Otsubo A et al (2012): Sem Oncol 39(1). The problem with current SNP arrays is that they can t detect balanced rearrangements! Normal karyotype Positive for IGH/CCND1 fusion and for monosomy 13 by FISH 6
The Power of SNP Array Analysis A Case of Mantle Cell Lymphoma The Power of SNP Array Analysis A Case of Mantle Cell Lymphoma UPD (CNNLOH) 4 or more copies Normal disomy Deletion * RB1 gene LAMP1 gene * FISH misidentified this as monosomy 13. In fact, it s far more complicated! SNP Array Analysis of Chronic Lymphocytic Leukemia CLL is a highly variable disease with life expectancies from a few months to many decades CLL FISH panel detects abnormalities that may not be found by karyotype CLL SNP array useful as the disease is characterized by genomic gain or loss CLL SNP array has high concordance with other studies; however, SNP array can identify 30-76% more abnormalities Elevated CLL genomic complexity is an independent and powerful marker to identify patients with aggressive disease and a shorter survival! SNP Array Analysis in AML Copy number alterations are identified in 15-40% of AML patients with a normal karyotype and predict an adverse clinical outcome Yi et al, 2011 The implementation of SNP array analysis in AML can increase the detection of clinically significant anomalies (including aupd) from ~50% by karyotyping to ~87% Copy number alterations were identified in 40-90% of MDS/AML patients with abnormal clones and nearly 100% of cases with complex karyotypes Hagenkord JM (2010) SNP Array Analysis in Renal Cell Carcinoma Clear cell RCC Papillary RCC Mucinous, tubular, and spindle cell carcinoma Chromophobe RCC Oncocytoma The Significance of Acquired UPD as Detected by SNP Array Analysis Copy neutral LOH is known to constitute between 50-80% of LOH in all human cancers aupds are found in: - 20% of AML - 30% of MDS - 80% of lymphomas O Shea et al (2009): Regions of aupd at diagnosis of follicular lymphoma are associated with both overall survival and risk of transformation. Blood 113(10). UPDs are associated with homozygous mutations of a number of tumor suppressor genes including TET2, CDKN2A/B, TP53, NF1, Rb, CEBPa, and RUNX1 UPDs are also associated with gain-of-function alleles of oncogenes including JAK2, NRAS, c-cbl, and FLT3 7
SNP Molecular Karyotyping in Neuroblastoma SNP Molecular Karyotyping in Neuroblastoma Neuroblastoma (NB), the most common extracranial tumor of childhood, can vary in clinical behavior from highly aggressive to spontaneous regression. Children <1 year present with localized tumor and have an excellent outcome while older children have more aggressive disease and a poorer prognosis. The prognosis is highly dependent on cytogenetic characterization: Good Prognosis Poor Prognosis Near-triploidy Near-diploidy or near-tetraploidy Numerical only with segmental rearrangements MYCN gene amplification Genome-wide microarray analysis is an excellent tool to assess cytogenetic status and is important to fully evaluate risk, aid diagnosis, and guide treatment del(11q) NBs constitute a unique group of unfavorable tumors with distinct features that differ from the other high-risk NB group (MYCN-amplified) Caren et al (2010): PNAS 107(9) Caren et al (2010): PNAS 107(9) Other Molecular Technologies May Supplant CMA The $1000 Genome Will Transform Medicine (and Make it a Lot More Complicated!) Ion Torrent Proton Sequencer Thank You 8