Clinical Utility of a Large Exome-Based Panel Greg Fischer and Moumita … · 2018-04-20 · Greg Fischer and Moumita Chaki CO-PRESENTING AUTHORS Diane Allingham-Hawkins, Ben Dorshorst,

Clinical Utility of a Large Exome-Based Panel in the Genetic Diagnosis of Autism Spectrum Disorders and Intellectual Disabilities

ABSTRACT

1. ASD-ID test reports include candidate genes identified through phenotypically-driven analyses and similar panels. In trio analyses, de novo variants within genes that are part of PreventionGenetics’s PGxome® (but not included in the ASD-ID panel) are also analyzed and reported depending on clinical fit.

2. The ASD-ID comprehensive test utilize the PGxome® workflow.

3. While only recently implemented, CNV Detection by NGS analysis has identified several causative CNVs likely responsible for the proband’s phenotype(s).

4. In case of trio analyses, a definitive molecular diagnoses is identified at a two-fold higher rate than singletons, even though such tests are ordered in only ~29% of cases.

5. Gene list updates will continue to occur twice per year, accounting for new information in public databases.

CONCLUSIONS

3800 South Business Park Avenue • Marshfield, Wisconsin 54449 • Phone: (715) 387-0484 • www.PreventionGenetics.com PUT US TO THE TEST

Greg Fischer and Moumita Chaki CO-PRESENTING AUTHORS

Diane Allingham-Hawkins, Ben Dorshorst, Gina Londre, Rachel Vanneste,

Jerry Machado, Robert Steiner, James WeberPREVENTIONGENETICS, LLC, MARSHFIELD, WI

INTRODUCTION: Autism Spectrum Disorders (ASDs) and Intellectual Disability (ID) are a heterogeneous group of highly comorbid neurodevelopmental disorders inherited in a multifactorial fashion. ASDs are characterized by variable social impairment, impaired communication ability, propensity for repetitive behavior(s), and restricted interests whereas ID refers to significant impairment of cognitive and adaptive development (IQ ≤70) due to abnormalities of brain structure and/or function. Chromosomal microarray (CMA) and FMR1 CGG-repeat expansion analysis have a combined diagnostic yield of ~11-15% and are recommended as the first tier of testing. Genetic variants (including SNVs, CNVs, and Fragile X testing) have been reported in at least 25-50% of ID cases and ~20% of ASD cases. Moreover, trio-based studies are reported to increase the molecular diagnostic rate to ~30-40% for developmental phenotypes. To date, over 800 genes have been implicated in ASD and ID-related phenotypes, with 97% of pathogenic variants found in coding regions. In early 2017, PreventionGenetics released an exome-based ‘large panel’ now including 1,913 genes with reported associations to syndromic and non-syndromic forms of ASD and/or ID (Autism Spectrum Disorders/Intellectual Disabilities (ASD-ID) Comprehensive Panel) as part of a comprehensive approach to genetic testing for patients with ASD and ID.

METHODS: Patient’s DNA is captured using Agilent Clinical Research Exome v2 hybridization probes and sequenced using Illumina’s NovaSeq 6000 using 2x150bp paired-end reads. Variants are filtered computationally to the 1,913 gene list before annotation and analysis via VarSeq (www.goldenhelix.com). For trio-based studies, this panel also considers de novo variants in any gene that is currently included in PreventionGenetics’ whole exome (PGxome®) sequencing analysis. Variant interpretations are distributed among our team of PhD geneticists with particular expertise for a given gene, leveraging the collective knowledge of our entire geneticist team. Furthermore, patient reports are limited to variants considered relevant to the reported phenotype, which are selected collectively by a team including several geneticists and genetic counselors. If desired, patients can reflex to our PGxome® whole exome sequencing analysis for a significantly reduced cost.

RESULTS: In the first 12 months of testing, a definitive molecular diagnosis was reported in 8% of cases, while 88% of cases had a variant(s) that may explain the patient’s clinical features. Our data suggests improved diagnostic rates (i.e. positive cases) with trio-based studies (~15% in trios vs. ~6% in singletons), which were ordered by providers in ~29% of cases. Due to the limited number of duo analyses requested, we are unable to appreciate any advantage of duo testing over singleton analysis.

CONCLUSION: In contrast to smaller NGS-based panels, the ASD-ID comprehensive panel leverages the advantages of our PGxome® workflow and has increased clinical sensitivity. This test covers multiple genes of interest in a single comprehensive sequencing test. Furthermore, this test also provides a reasonable alternative to whole-exome sequencing for patients with neurodevelopmental disorders.

Variant interpretation for each case is carried out by the entire PreventionGenetics Geneticist team, each focusing on a particular disease portfolio of which he or she is particularly familiar with. Clinical features are extracted from the requisition documents by the clinical team, which includes a Genetic Counselor and a Clinical Geneticist. Using this information, the assigned geneticist then reviews variants and CNV calls by NGS for a particular case and identifies candidates suspected to contribute to the proband’s phenotype. Selected variants (SNV and CNV) are presented to a team of geneticists (including a Clinical Geneticist) who have significant experience with exome analyses. Based on the approval of the exome team, a list of variants are then confirmed (by Sanger and array-CGH, as applicable) and included in the clinical report. The assigned geneticist prepares the report which is then second reviewed by an accredited geneticist. The second reviewer performs an independent assessment of the observed variants and is unbiased by previous discussion and analysis between the first reviewer and the exome team. Approval by the second reviewer is achieved within 30-35 days of sample arrival for clients ordering the ASD-ID comprehensive panel.

GTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCA

DNA capture using Agilent Clinical Research Exome v2 hybridization probes

+Sequencing using Illumina’s NovaSeq

6000 (2x150bp paired-end reads)

Variant Interpretation

Clinical Features

Exome Team *Clinical Geneticist

*

PhD GeneticistFIRST REVIEW GTCCGATTCCAGG

AGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCA

Report Writing andVariant Confirmation

Variant and CNV Candidates

Accredited GeneticistSECOND REVIEW

Report SentTAT: ~30-35 days

GTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCAGTCCGATTCCAGGAGGGCTGGGGTTTTCTGTGATGGACCTTTGATCCGTGCA

Pathogenic?Likely Pathogenic?

Uncertain?Likely Benign?

Benign?

UTILIZATION OF THE PreventionGenetics PGxome® WORKFLOW FOR ASD-ID TESTS

Fitzgibbon et al. 2009. PubMed ID: 19183217GeneReviews, Finucane et al. 2016. PubMed ID: 27308687Linhares et al. 2016. PubMed ID: 27561113Schaefer and Mendelsohn. 2013. PubMed ID: 23519317Sun et al. 2015. PubMed ID: 26158183

Suggested Guidelines for Testing Strategies

Phenotype-Driven Gene Panels for ASD-ID

Detection of Single Nucleotide Variant – Positive Finding

Detection of Copy Number Variations

Testing Metrics and Clinical Validity of the Panel in Year 1

FIRST TIER: Detection of clinically-relevant CNVs by genomic microarrays (array-CGH) and karyotyping (whenever applicable)

SECOND TIER: Screening for Fragile X syndrome by PCR

THIRD TIER: Application of Next-Generation sequencing using extensive gene panels (10 to >100 genes) (WES, WGS)

(Schaefer and Mendelsohn. 2013 and Sun et al. 2015)

Clinically-Relevant (ASD-ID) OMIM Genes1,500+ GENESAutosomal Recessive

300+ Genes

TRIO-ANALYSIS: Heterozygous De Novo Likely Pathogenic variant in KMT2ACLINICAL FEATURES: Global developmental delay, generalized hypotonia, Cheyne-Strokes respiration, hypertelorism, short palpebral fissures, tented upper lip vermillion, high-narrow palate, low-set ears, overfolded helix, short stature, paternal history of febrile seizures as a child and supraventricular tachycardia.

• Pathogenic variants in KMT2A have been documented in autosomal dominant Wiedemann-Steiner syndrome.

• Additionally, heterozygous for paternally-inherited variants of uncertain significance in TRAPPC9 (c.3176C>T, p.Ala1059Val and c.3469G>A, p.Ala1157Thr). Pathogenic TRAPPC9 variants are associated with autosomal dominant mental retardation 13.

c.2318dup (p.Ser774Valfs*12)in KMT2A

CASE ONE (SINGLETON): Heterozygous Large Likely Pathogenic Deletion in Chromosome 1CLINICAL FEATURES: Short stature, obesity, speech delay, mild intellectual disability, learning disability, developmental delay, dysmorphic features, upslanting palpebral fissures, round facies, retrognathia.

• 1p13.2 deletions involving NRAS have been reported in individuals with Noonan-like features, including short stature, developmental delay, intellectual disability and dysmorphic features (Fitzgibbon et al. 2009, Linhares et al. 2016).

• Due to the relative lack of probes in the centromeric region, it was not clear if this represents a single event or multiple events.

~6.29 Mb via CMA at 1p13.2-p11.2

AMPD1, NRAS, NGF, PHGDH, NOTCH2

~1.56 Mb could not confirm due to segmental

duplications

~29.167 Mb via NGS

CASE TWO (SINGLETON): Heterozygous for Pathogenic ~4.86 Mb Duplication on Chromosome 15 (15q Duplication Syndrome)CLINICAL FEATURES: Global developmental delay, intellectual disability, autism spectrum disorder, delayed speech and language development, behavioral abnormality, facial anomalies.

• Associated with 15q Duplication Syndrome (Dup15q; OMIM #608636)

• While several genes of interest, no single gene has been identified that, when duplicated, results in 15q Duplication Syndrome (GeneReviews, Finucane et al. 2016).

• Inheritance predominantly occurs de novo (85% of probands), while the in the remaining 15% of cases the duplication is inherited maternally (GeneReviews, Finucane et al. 2016).

~4.841 Mb via NGS

~4.86 Mb via CMA at 15q11.2-q13.1

Several genes of interest (ATP10A, CYFIP1, MAGEL2, NECDIN, SNRPN, UBE3A, GABAA receptor subunits)

68%3%

29%

Singleton Duo Trio

3%

88%

8%

Negative Indeterminate Positive

6%

0%

15%

8%

0%

5%

10%

15%

20%

Posi

tive

%

(A) (B)

(C)

REFE

REN

CES

(B) Considering all family structures, a definitive molecular diagnosis was determined in ~8% of cases overall, while in ~88% of cases at least a single variant was identified that may contribute to the patient’s phenotype (indeterminate).

(C) When considering the type of test ordered, a definitive molecular diagnosis is identified at an almost two-fold higher rate in trios than singletons or all family structures overall.

(A) Family structure breakdown of ASD-ID tests. Trio analyses account for approximately a third of the ordered tests.

X-Linked100+ Genes

X-Linked ID Panel - 123 GenesUPDATE IN PROGRESS - available through Phenotype Cluster - Custom Panel

Non-Syndromic ID Panel - 110 GenesASD-Specific Panel - 108 Genes

ASD-ID Comprehensive Panel or Syndromic ID Panel)1,913 GENES RECENTLY UPDATED

Autosomal DominantDe Novo