← Product Code [PFX](/submissions/IM/subpart-f%E2%80%94immunological-test-systems/PFX) · K163367

# GenetiSure Dx Postnatal Assay (K163367)

_Agilent Technologies, Inc. · PFX · Aug 11, 2017 · Immunology · SESE_

**Canonical URL:** https://fda.innolitics.com/submissions/MG/subpart-f%E2%80%94immunological-test-systems/PFX/K163367

## Device Facts

- **Applicant:** Agilent Technologies, Inc.
- **Product Code:** [PFX](/submissions/IM/subpart-f%E2%80%94immunological-test-systems/PFX.md)
- **Decision Date:** Aug 11, 2017
- **Decision:** SESE
- **Submission Type:** Traditional
- **Regulation:** 21 CFR 866.5920
- **Device Class:** Class 2
- **Review Panel:** Immunology
- **Attributes:** Pediatric

## Indications for Use

GenetiSure Dx Postnatal Assay is a qualitative assay intended for the postnatal detection of copy number variations (CNV) and copy-neutral loss of heterozygosity (cnLOH) in genomic DNA obtained from peripheral whole blood in patients referred for chromosomal testing based on clinical presentation. GenetiSure Dx Postnatal Assay is intended for the detection of CNVs and cnLOH associated with developmental delay, intellectual disability, congenital anomalies, or dysmorphic features. Assay results are intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, including confirmation by alternative methods, parental evaluation, clinical genetic evaluation, and counseling, as appropriate. Interpretation of assay results is intended to be performed only by healthcare professionals, board certified in clinical cytogenetics or molecular genetics. The assay is intended to be used on the SureScan Dx Microarray Scanner System and analyzed by CytoDx Software. This device is not intended to be used for standalone diagnostic purposes, preimplantation or prenatal testing or screening, population screening, or for the detection of, or screening for, acquired or somatic genetic aberrations.

## Device Story

In vitro diagnostic assay for molecular karyotyping using array comparative genomic hybridization (aCGH) and SNP analysis; inputs are gDNA from peripheral whole blood; patient sample and sex-matched reference sample are restriction-digested, fluorescently labeled (Cy5/Cy3), and co-hybridized to 4x180K aCGH+SNP microarray slides; slides are washed and scanned via SureScan Dx Microarray Scanner; CytoDx Software performs feature extraction, computes relative abundance of target sequences, and identifies CNVs and cnLOH; results are interpreted by board-certified cytogeneticists or molecular geneticists; output is a report of chromosomal aberrations; used in clinical laboratories to assist in diagnosing patients with developmental delay, intellectual disability, or congenital anomalies; benefits include identification of clinically relevant genomic abnormalities to guide clinical management and counseling.

## Clinical Evidence

Retrospective clinical study of 800 patient samples and 100 normal samples. Diagnostic yield for CNVs was 15%, increasing to 20% with cnLOH. PPA was 76.8% (CNV only) and 76.7% (CNV+cnLOH); NPA was 95.5% (CNV only) and 89.8% (CNV+cnLOH). Accuracy assessed against two independent microarray platforms; confirmation rates varied by aberration size and probe count. Bench testing confirmed stability, precision, and limit of detection (375ng).

## Technological Characteristics

Array comparative genomic hybridization (aCGH) + SNP microarray. 4x180K slides (approx. 107k CNV probes, 59k SNP probes). In-situ ink-jet probe synthesis. Fluorescence-based detection (Cy3/Cy5). SureScan Dx Microarray Scanner. CytoDx 1.0 software for analysis. Requires 500ng gDNA input. Sterilization/materials: standard laboratory reagents/glass slides.

## Regulatory Identification

A postnatal chromosomal copy number variation detection system is a qualitative assay intended for the detection of copy number variations (CNVs) in genomic DNA obtained from whole blood in patients referred for chromosomal testing based on clinical presentation. It is intended for the detection of CNVs associated with developmental delay, intellectual disability, congenital anomalies, or dysmorphic features. Assay results are intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, including confirmation by alternative methods, parental evaluation, clinical genetic evaluation, and counseling, as appropriate. Interpretation of assay results is intended to be performed by qualified healthcare professionals such as clinical cytogeneticists or molecular geneticists. This device is not intended to be used for standalone diagnostic purposes, pre-implantation or prenatal testing or screening, population screening, or for the detection of, or screening for, acquired or somatic genetic aberrations.

## Special Controls

*Classification.* Class II (special controls). The special controls for this device are:(1) Design verification and validation must include the following information:
(i) A detailed description of all components in the test system that includes:
(A) A description of the assay components, array composition and layout, all required reagents, instrumentation, and equipment, including illustrations or photographs of non-standard equipment or methods;
(B) A description of the design of the array in terms of chromosomal coverage and probe density for different regions;
(C) An identification of the number of probes and size of the CNVs reported at the lower range of the assay;
(D) Detailed documentation of the device software, including standalone software applications and hardware-based devices that incorporate software;
(E) Methodology and protocols for detecting copy number and visualizing results;
(F) A description of the result outputs along with sample reports, and a description of any links to external databases provided by the device to the user or accessed by the device;
(G) Specifications for the methods to be used in specimen collection, extraction (including DNA criteria for DNA quality and quantity to perform the assay), and storage; and
(H) A description of appropriate internal and external controls that are recommended or provided. The description must identify those control elements that are incorporated into the testing procedure.
(ii) Information that demonstrates the performance characteristics of the system, including:
(A) Device reproducibility data generated, at a minimum, using three sites, with two operators at each site, for three non-consecutive days using at least three instruments. A well-characterized panel of samples that provide a wide range of CNVs (
*i.e.,* gains, losses, adequate size coverage across the range of sizes claimed by the device, adequate chromosomal coverage, challenging regions in the genome, CNVs reported at the lower range of the assay, interstitial, subtelomeric, and pericentromeric rearrangements, aneuploidy, unbalanced translocations, mosaicism, and known syndromic regions) must be used. The results must be itemized for all CNVs detected in each sample across all replicates and summarized in a tabular format stratified by size range and range of probe numbers for gains and losses separately and calculated for overall. The results must be analyzed using pairwise replicate agreement, and summarized as overall pairwise replicate agreement as well as pairwise replicate agreement conditional on replicates having a positive copy number state call (gains or losses), call rate, CNV size variation, and endpoint agreement;(B) Device accuracy data using cell lines and clinical samples representing a variety of CNVs and syndromes. In this analytical study, accuracy must be determined for every CNV detected in a particular sample. The accuracy data provided must include the copy number state determination and endpoint accuracy. The accuracy samples must cover different genomic variations across the genome (
*i.e.,* gains, losses, adequate CNV size coverage across the range of sizes claimed by the device, adequate chromosomal coverage, challenging regions in the genome, CNVs reported at the lower range of the assay, interstitial, subtelomeric, and pericentromeric rearrangements, aneuploidy, unbalanced translocations, mosaicism, and known syndromic regions). CNVs identified by the device must be compared to comparator method(s). Agreement between the CNVs detected by the array and the comparator must be summarized in a tabular format that includes the positive percent agreement and false positive rate stratified by size range and range of probe numbers for gains and losses separately and calculated for overall;(C) Assay performance data for CNVs reported at the lower range of the assay for both gains and losses;
(D) Device analytical sensitivity data, including DNA input and limit of detection for mosaicism, if applicable;
(E) Device analytical specificity data, including interference, carryover, and cross-contamination data;
(F) Device stability data, including real-time stability under various storage times, temperatures, and freeze-thaw conditions;
(G) Specimen matrix comparison data if more than one specimen type or anticoagulant can be tested with the device;
(H) Data that demonstrates the clinical validity, including diagnostic yield, of the device using a minimum of 800 retrospective clinical samples that were collected prospectively and obtained from three or more clinical laboratories, with results interpretation equally divided between two or more qualified healthcare professionals (
*e.g.,* cytogeneticists). Patients must be representative of the intended use population and not limited to common syndromes. Diagnostic yield data must be summarized in tabular format and stratified by the comparison methodologies. Data must also be summarized comparing interpretation of results, with description of reasons for variability in calls between the device and the standard of care methods. Data to support the accuracy of calls for known syndromes must be included; and(I) Data that demonstrates device results when a minimum of 100 apparently healthy, phenotypically normal individuals are tested and interpreted by one or more cytogeneticists blinded to the patient status.
(iii) Identification of risk mitigation elements used by the device, including a description of all additional procedures, methods, and practices incorporated into the directions for use that mitigate risks associated with testing.
(2) The labeling required under § 809.10 of this chapter must include:
(i) A warning statement that the device is not intended to be used for standalone diagnostic purposes, pre-implantation or prenatal testing or screening, population screening, or for the detection of, or screening for, acquired or somatic genetic aberrations;
(ii) Limitations regarding the assay's performance with respect to validated CNVs reported at the lower range of the assay, stratified by size range and range of probe numbers for gains and losses separately; and limitations regarding problematic (hypervariable) regions, loss of heterozygosity, mosaicism, and inability to detect balanced translocations, as appropriate;
(iii) A warning statement that interpretation of assay results is intended to be performed by qualified healthcare professionals such as clinical cytogeneticists or molecular geneticists; and,
(iv) A description of the performance studies performed in accordance with paragraph (b)(1)(ii) of this section and a summary of the results.

## Predicate Devices

- Affymetrix CytoScan Dx Assay ([K130313](/device/K130313.md))

## Submission Summary (Full Text)

> This content was OCRed from public FDA records by [Innolitics](https://innolitics.com). If you use, quote, summarize, crawl, or train on this content, cite Innolitics at https://innolitics.com.
>
> Innolitics is a medical-device software consultancy. We help companies design, build, and clear FDA-regulated software and AI/ML devices, including [a 510(k)](https://innolitics.com/services/510ks/), [a De Novo](https://innolitics.com/services/regulatory/), [a SaMD](https://innolitics.com/services/end-to-end-samd/), [an AI/ML medical device](https://innolitics.com/services/medical-imaging-ai-development/), or [an FDA regulatory strategy](https://innolitics.com/services/regulatory/).

{0}

1

# 510(k) SUBSTANTIAL EQUIVALENCE DETERMINATION DECISION MEMORANDUM

A. 510(k) Number:
k163367

B. Purpose for Submission:
New device

C. Measurand:
Genome-wide chromosomal copy number variations

D. Type of Test:
Chromosomal Microarray

E. Applicant:
Agilent Technologies, Inc.

F. Proprietary and Established Names:
Agilent GenetiSure Dx Postnatal Assay

G. Regulatory Information:

1. Regulation section:
21 CFR 866.5920

2. Classification:
Class II (special controls)

3. Product code:
PFX -- System, Microarray-based, genome-wide, postnatal chromosomal abnormality detection

4. Panel:
Immunology

H. Intended Use:

{1}

1. Intended use(s):

GenetiSure Dx Postnatal Assay is a qualitative assay intended for the postnatal detection of copy number variations (CNV) and copy-neutral loss of heterozygosity (cnLOH) in genomic DNA obtained from peripheral whole blood in patients referred for chromosomal testing based on clinical presentation. GenetiSure Dx Postnatal Assay is intended for the detection of CNVs and cnLOH associated with developmental delay, intellectual disability, congenital anomalies, or dysmorphic features. Assay results are intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, including confirmation by alternative methods, parental evaluation, clinical genetic evaluation, and counseling, as appropriate. Interpretation of assay results is intended to be performed only by healthcare professionals, board certified in clinical cytogenetics or molecular genetics. The assay is intended to be used on the SureScan Dx Microarray Scanner System and analyzed by CytoDx Software.

This device is not intended to be used for standalone diagnostic purposes, preimplantation or prenatal testing or screening, population screening, or for the detection of, or screening for, acquired or somatic genetic aberrations.

2. Indication(s) for use:

Same as Intended use above.

3. Special conditions for use statement(s):

For prescription use.

4. Special instrument requirements:

SureScan Dx Microarray Scanner with CytoDx 1.0 software.

I. Device Description:

The GenetiSure Dx Postnatal Assay consists of the following components:

1. Six 4 x 180K Microarray slides capable of running 4 assays per slide and Gasket slides which hold the samples during hybridization to the microarrays

2. Reagents and Columns (materials supplied separately)

a) GenetiSure Dx DNA Labeling Kit contains sufficient two-color labeling reaction contains Random Primers and the Exo(-) Klenow fragment to differentially label DNA, enzymes, nucleotides, columns, and human reference DNA for the six 4x180K microarray slides included with the GenetiSure Dx Postnatal Assay.

b) GenetiSure Dx Hybridization Kit contains hybridization buffer and blocking agent used for hybridization to the microarrau

c) GenetiSure Dx Wash Buffer Set- The kit contains wash buffers for washing unhybridized labeled DNA from the microarrays.

{2}

d) GenetiSure Dx Cot-1 Human DNA - The kit contains a solution of fractionated human DNA that has been enriched for repetitive sequences.

3. CytoDx 1.0 Software- CytoDx Software performs feature extraction, CNV and cnLOH identification and reporting on the microarray TIF images generated by the SureScan Dx Microarray Scanner.

# J. Substantial Equivalence Information:

1. Predicate device name:

Affymetrix CytoScan Dx Assay

2. Predicate device 510(k) number:

DEN130018

4. Comparison with predicate:

|  SIMILARITIES  |   |   |
| --- | --- | --- |
|  Item | Device GenetiSure Dx Postnatal Assay | Predicate Affymetrix CytoScan Dx Assay  |
|  Indications for Use | GenetiSure Dx Postnatal Assay is a qualitative assay intended for the postnatal detection of copy number variations (CNV) and copy-neutral loss of heterozygosity (cnLOH) in genomic DNA obtained from peripheral whole blood in patients referred for chromosomal testing based on clinical presentation. GenetiSure Dx Postnatal Assay is intended for the detection of CNVs and cnLOH associated with developmental delay, intellectual disability, congenital anomalies, dysmorphic features. Assay results are intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, including confirmation by alternative methods, parental evaluation, clinical genetic evaluation, and counseling, as appropriate. | CytoScan® Dx Assay is a qualitative assay intended for the postnatal detection of copy number variations (CNV) in genomic DNA obtained from peripheral whole blood in patients referred for chromosomal testing based on clinical presentation. CytoScan® Dx Assay is intended for the detection of CNVs associated with developmental delay, intellectual disability, congenital anomalies, or dysmorphic features. Assay results are intended to be used in conjunction with other clinical and diagnostic findings, consistent with professional standards of practice, including confirmation by alternative methods, parental evaluation, clinical genetic evaluation, and counseling, as appropriate.  |

{3}

|  DIFFERENCES  |   |   |
| --- | --- | --- |
|  Item | Device GenetiSure Dx Postnatal Assay | Predicate Affymetrix CytoScan Dx Assay  |
|  Array Format | 60-mer probes Four microarrays on a single 1”x 3” glass slide | 25-mer oligos Individual microarrays housed in a GeneChip cartridge  |
|  Method of Array Manufacture | On slide (in-situ) synthesis of probes using ink-jet printing | On-wafer synthesis of probes using photolithography  |

{4}

|  DNA Fragmentation/ Labeling | Fragmented DNA is directly labelled with fluorescent dye (Cy3 and Cy5) before hybridization. Data is produced in two intensity channels which are then compared to generate a LogRatio | Fragmented DNA is PCR amplified and then labelled with biotin before hybridization. Single channel data is produced which is later compared to an in silico reference to produce a LogRatio.  |
| --- | --- | --- |
|  Hybridization | Cohybridization of labeled sample and reference for direct on-array comparison | Hybridization of single labeled sample which is compared to an in silico reference.  |
|  Washing / Staining of Microarrays | Manual washing process in accordance with instruction for the validated diagnostic assay | Automated processing with FS450Dx fluidics station for both washing and staining steps  |
|  Instrument | SureScan Dx Microarray Scanner | GeneChip System 3000Dx Scanner  |

# K. Standard/Guidance Document Referenced:

- CLSI EP07-A2 Interference Testing in Clinical Chemistry; Approved Guideline - Second Edition.
- CLSI EP12-A2 User Protocol for Evaluation of Qualitative Test Performance.
- EP25-A Evaluation of Stability of In Vitro Diagnostic Reagents; Approved Guideline.
- CLSI MM13-A Collection, Transport, Preparation, and Storage of Specimens for Molecular Methods; Approved Guideline.
- CLSI MM21 Genomic Copy Number Microarrays for Constitutional Genetic and Oncology Applications -First Edition.

# L. Test Principle:

The GenetiSure Dx Postnatal Assay uses array comparative genomic hybridization (aCGH) and single nucleotide polymorphism (SNP) technology (referred to as CGH +SNP microarray) to detect chromosomal imbalances in genomic DNA (gDNA) isolated from  $200\mu \mathrm{L}$  of EDTA-anticoagulated peripheral blood using the QIAamp DSP DNA Blood Mini kit. The gDNA is quantified and  $0.5\mu \mathrm{g}$  (500ng) is processed in parallel with  $0.5\mu \mathrm{g}$  of sex-matched reference DNA included in the GenetiSure Dx DNA Labeling Kit. gDNA is digested with restriction enzymes and labeled with fluorescent dyes in the Labeling Kit. The subject samples is labeled with cyanine 5 (Cy5) dye and the sex-matched reference samples is labeled with cyanine 3 (Cy3) dye. Sex-matched reference DNAs are used for data normalization and aberration detection on a per test sample basis. Labeled samples are co-hybridized onto a single GenetiSure Dx microarray using the reagents in the GenetiSure Dx Hybridization Kit, and Hybridization Chamber Kit. The in situ hybridization technique allows detection and mapping of DNA sequence copy differences between the two differentially labeled genomic DNAs (subject/test sample and a reference sample).

{5}

The microarray contains complementary nucleic acid sequences synthesized in situ on a microarray slide: approximately 107,000 probes for CNV analysis, and approximately 59,000 bi-allelic SNP probes. The CNV probes are distributed across the entire genome with a higher density of probes in regions designated by the International Standards for Cytogenomic Arrays (ISCA) consortium to be of clinical interest.

The relative amount of the hybridized target sequences is computed by the Analytics module, based on the relative intensities of the fluorophores in the patient and reference samples hybridized to each of the probe sequences. After hybridization and subsequent washing with the GenetiSure Dx Wash Buffer Set, the microarrays are scanned by SureScan Microarray Scanner. The scanner generated image is processed using the Agilent CytoDx Software, and the image data are converted to numeric data using the Feature Extraction module of the software. Internal control probes on each array are used to calculate array quality metrics and assess the quality of the data. Additional array QC metrics are employed to detect for potential anomalies. Locations of copy number variation (CNVs) and copy-neutral loss of heterozygosity (cnLOH) in the DNA segments of the subject sample genome are identified. The aberrations identified in a patient sample by the CytoDx algorithms can be viewed from the Triage View screen of the CytoDx software. The reported CNVs and cnLOH are interpreted by a Board Certified Cytogeneticist, Molecular Geneticist, Molecular Pathologist, or similarly qualified clinician who has been trained to identify the clinically relevant CNVs, determine clinical significance, and report out these findings.

## M. Performance Characteristics (if/when applicable):

### 1. Analytical performance:

#### a. Precision/Reproducibility:

Two reproducibility studies were conducted. The first was a site-to-site reproducibility study, and the second was a between-lot reproducibility study.

#### Study 1. Site-to-site reproducibility:

A reproducibility study was performed with 48 genomic DNA samples purified from cell lines representing a range of aberrations detected (gains, losses and copy neutral loss of heterozygosity (cnLOH)). Each sample was tested at 3 sites by 2 operators in triplicate across 3 non-consecutive weeks for a total of 18 replicates (3×2 × 3). The entire study was performed using one lot of reagents including each of the following: Qiagen DSP extraction kits, GenetiSure Dx Complete DNA Labeling kit, microarrays, gaskets, Oligo aCGH/ChIP-on-chip Hybridization Kit, Cot 1 DNA, Oligo aCGH/ChIP-on-chip Wash Buffer 1, Oligo aCGH/ChIP-on-chip Wash Buffer 2, and SureScan Dx Microarray Scanner.

The aberrations met the following criteria: common syndromes (i.e., known syndromic regions), analytically challenging regions, claimed minimal resolution, varying aberration size ranges, and genomic coverage of aberrations. Multiple samples had multiple aberrations spanning multiple criteria. Test sample selection criteria encompassed aberrations expected to be found in normal whole blood

6

{6}

7

samples.

For a given test sample (n=48), each aberration identified, regardless of expected pathogenicity, in each of the 18 replicates was reported. A total of 325 unique aberrations (87 gains, 193 losses, and 45 cnLOH) were detected from the 48 gDNA samples across their replicates. These aberrations were distributed across all 24 chromosomes in the human genome, with a collective coverage of 41.8% (16.7% for gains, 11.3% for losses, 22.4% for gains or losses, and 27.7% for cnLOH).

All individual aberrations reported were compared to their respective replicates (18 replicates for each aberration) by pairwise replicate analysis (PRA), requiring at least 50% overlap of chromosomal coordinates for confirmation. The aberrations reported across the replicate arrays, for a given sample, were first divided into specific "unique" aberration groups, each representing a specified genomic range with a specified aberration type (gain, loss, or cnLOH). Individual replicates for a unique aberration were then evaluated using pairwise analyses:

- For replicates i and j, where i ≠ j, an aberration found in replicate i was considered pairwise confirmed if it had ≥ 50% overlap with an aberration found in replicate j.
- If either or both replicates i or j had a gap, the gap region was considered as not being in agreement for the overlap calculation; no gap filling or segment joining was performed.
- The replicates must have the same copy number state (i.e., gain or loss) to be confirmed.
- An alternative, more stringent 80% overlap criteria was also assessed.

Percent Pairwise Replicate Agreement was calculated across all replicates and also individually by variable type (site, operator, test sample, processing week, probe size, length (kb) and aberration type such as gain, loss, or LOH). Additional breakpoint/endpoint agreement was assessed

The following statistics are reported: positive percent agreement (PPV), call rate, percent overlap, length, percent coefficient of variance (% CV), median percent absolute endpoint deviation, and standard deviation (SD) for both left and right endpoints.

The results demonstrated that when comparing all replicates of all test samples across all days, sites, and operators, using a 50% aberration overlap criterion, the overall pairwise replicate agreement across all CNVs was 85.0% (Table 1). Pairwise replicate agreement across the various kb (length) bins ranged from 75.9% to 100%. For copy number gains, the overall agreement was 85.7%; for losses, the overall agreement was 84.6%. For cnLOH, the overall pairwise replicate agreement was 89.1%. Applying a more stringent 80% overlap criterion produced overall agreements of 82.3% for CNVs, 84.4% for gain, 81.3% for losses, and 87.9% for LOH. When assessing specifically the agreement between replicates for positive aberration calls by PPA analysis, the agreement was 89.3% for all copy number calls and 92.7% for

{7}

LOH using the 50% overlap criterion. Call rate averaged 78.1% for CNVs, and 74.9% for cnLOH. The results are shown in Table 1.

Table 1. Reproducibility of Aberrations Categorized by Size (in kb) and Type Based on Call Rate, Pairwise Agreement between Replicates and Positive Percent Agreement (PPA) for Two Criteria (50% and 80% Overlap) in All Regions in the Site-to-Site Study.

|  Aberration Type | Aberration Range (kb) | # Aberrations | Call Rate (%) | Pairwise Replicate Agreement (%) |   | PPA (%)  |   |
| --- | --- | --- | --- | --- | --- | --- | --- |
|   |   |   |   |  50% Overlap | 80% Overlap | 50% Overlap | 80% Overlap  |
|  GAIN | 10 - 50 | 5 | 51.2 | 82.5 | 82.5 | 82.9 | 82.9  |
|   |  50 - 100 | 3 | 68.7 | 96.3 | 96.3 | 97.3 | 97.3  |
|   |  100 - 200 | 13 | 50.5 | 79.9 | 79.8 | 80.1 | 80.0  |
|   |  200 - 500 | 26 | 82.7 | 86.3 | 84.6 | 91.5 | 89.4  |
|   |  500 - 1000 | 9 | 79.7 | 80.2 | 78.9 | 87.6 | 86.0  |
|   |  1000 - 2000 | 7 | 72.1 | 90.7 | 82.8 | 92.4 | 81.6  |
|   |  2000 - 5000 | 11 | 65.1 | 75.9 | 75.9 | 79.7 | 79.7  |
|   |  5000 + | 13 | 93.2 | 98.4 | 98.4 | 99.1 | 99.1  |
|   |  Total | 87 | 73.8 | 85.7 | 84.4 | 89.9 | 88.2  |
|  LOSS | 10 - 50 | 14 | 51.6 | 76.8 | 76.1 | 77.6 | 76.2  |
|   |  50 - 100 | 2 | 100.0 | 100.0 | 89.5 | 100.0 | 89.5  |
|   |  100 - 200 | 23 | 81.4 | 82.3 | 78.1 | 88.1 | 82.9  |
|   |  200 - 500 | 31 | 82.6 | 81.8 | 75.5 | 86.3 | 78.7  |
|   |  500 - 1000 | 55 | 72.8 | 81.2 | 76.2 | 85.2 | 78.5  |
|   |  1000 - 2000 | 30 | 83.3 | 86.4 | 85.9 | 91.5 | 91.0  |
|   |  2000 - 5000 | 18 | 88.9 | 87.4 | 85.1 | 89.9 | 87.3  |
|   |  5000 + | 20 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0  |
|   |  Total | 193 | 80.1 | 84.6 | 81.3 | 89.0 | 84.8  |
|  ALL CNVs | Total | 280 | 78.1 | 85.0 | 82.3 | 89.3 | 85.8  |
|  cnLOH | 5000 - 10000 | 21 | 50.6 | 77.1 | 76.8 | 77.4 | 76.8  |
|   |  10000 - 20000 | 11 | 91.5 | 99.0 | 96.4 | 99.4 | 96.7  |
|   |  20000 + | 13 | 100.0 | 100.0 | 98.4 | 100.0 | 98.4  |
|   |  Total | 45 | 74.9 | 89.1 | 87.9 | 92.7 | 91.1  |

When results were grouped into various bins by the number of probes in an aberration, rather than length in kb, using the 50% overlap criterion, the overall pairwise replicate agreement was similar to the above: 86.2% for CNVs (ranging from 70.6% to 100%), 86.1% for gains alone, 86.3% for losses alone, and 89.1% for LOH. Using the 80% overlap criterion, overall agreements were 84.6% for CNVs, 85.3% for gains, 84.2% for losses, and 88.4% for LOH. PPA for the 50% overlap

{8}

criterion was 90.9% and 92.7% for CNVs and cnLOH, respectively. Call rate averaged 78.1% for CNVs, and 74.9% for cnLOH. The results are shown below in 2.

Table 2. Reproducibility of Aberrations Categorized by Probe Number and Type Based on Call Rate, Pairwise Agreement between Replicates and PPA for Two Criteria (50% and 80% Overlap) in All Regions in the Site-to-Site Study.

|  Aberration Type | Aberration Range (# Probes) | # Aberrations | Call Rate (%) | Pairwise Replicate Agreement (%) |   | PPA (%)  |   |
| --- | --- | --- | --- | --- | --- | --- | --- |
|   |   |   |   |  50% Overlap | 80% Overlap | 50% Overlap | 80% Overlap  |
|  GAIN | 5 - 7 | 11 | 38.0 | 76.5 | 76.5 | 69.0 | 69.0  |
|   |  7 - 10 | 15 | 54.1 | 70.6 | 69.4 | 72.8 | 70.6  |
|   |  10 - 15 | 23 | 87.9 | 89.4 | 88.3 | 94.0 | 92.7  |
|   |  15 - 20 | 11 | 66.5 | 82.4 | 79.9 | 86.8 | 83.1  |
|   |  20 - 30 | 9 | 89.6 | 97.5 | 97.5 | 97.9 | 97.9  |
|   |  30 - 100 | 3 | 70.3 | 93.0 | 93.0 | 95.0 | 95.0  |
|   |  100 - 500 | 3 | 72.3 | 90.2 | 90.2 | 93.2 | 93.2  |
|   |  500 + | 12 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0  |
|   |  Total | 87 | 73.8 | 86.1 | 85.3 | 90.5 | 89.4  |
|  LOSS | 5 - 7 | 36 | 61.1 | 76.6 | 75.6 | 80.9 | 79.1  |
|   |  7 - 10 | 39 | 65.5 | 77.6 | 75.4 | 82.8 | 79.6  |
|   |  10 - 15 | 42 | 81.9 | 85.5 | 81.0 | 90.5 | 85.0  |
|   |  15 - 20 | 18 | 96.9 | 95.9 | 94.0 | 97.6 | 95.7  |
|   |  20 - 30 | 16 | 87.5 | 89.1 | 85.7 | 91.6 | 87.9  |
|   |  30 - 100 | 10 | 92.2 | 93.2 | 92.9 | 96.3 | 96.0  |
|   |  100 - 500 | 17 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0  |
|   |  500 + | 15 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0  |
|   |  Total | 193 | 80.1 | 86.3 | 84.2 | 91.1 | 88.5  |
|  ALL CNVs | Total | 280 | 78.1 | 86.2 | 84.6 | 90.9 | 88.8  |
|  cnLOH | 100 to 200 | 25 | 54.8 | 80.3 | 80.3 | 82.0 | 82.0  |
|   |  200 to 500 | 13 | 100.0 | 100.0 | 97.7 | 100.0 | 97.7  |
|   |  > 500 | 7 | 100.0 | 100.0 | 100.0 | 100.0 | 100.0  |
|   |  Total | 45 | 74.9 | 89.1 | 88.4 | 92.7 | 91.8  |

For endpoint analysis, only those CNVs detected and with the same copy number state (gain/loss) were assessed for endpoint agreement (i.e., 'no calls' in replicates could not be included). Endpoint agreement is assessed by median % absolute endpoint deviation, standard deviation of left endpoint, and standard deviation of right endpoint. For copy number aberrations (CNVs), the Median % Absolute Endpoint Deviation was 3% (mean) when analyzed by both probe number and size (in kb) (Table 3 and 4). For cnLOH calls, the Median % Absolute Endpoint Deviation was 1% (mean) for both analyses. In addition, the average % overlap for all pairwise confirmed aberration was 82.1% (by probe number) and 80.8% (by size) for CNVs, and 84.7% (by probe number) and 84.5% (by size) for cnLOH calls.

{9}

Table 3. Reproducibility of Aberration Breakpoints by Probe Number

|  Aberration Type | Aberration Range (# Probes) | N | % CV Aberration Length | Average % Overlap | % Median Absolute Endpoint Deviation (x100) | SD Left Endpoint (# Probes) | SD Right Endpoint (# Probes)  |
| --- | --- | --- | --- | --- | --- | --- | --- |
|   |   |   |  Mean (Min, Median, Max) |   | Mean (Min, Median, Max)  |   |   |
|  GAIN | 5 - 7 | 11 | 3.0 (0.0, 0.0, 11.7) | 51.9 | 0.01 (0.00, 0.00, 0.13) | 0.1 (0.0, 0.0, 0.6) | 0.1 (0.0, 0.0, 0.5)  |
|   |  7 - 10 | 15 | 9.0 (0.0, 6.5, 35.4) | 55.1 | 0.05 (0.00, 0.00, 0.31) | 0.5 (0.0, 0.4, 2.0) | 0.5 (0.0, 0.0, 2.9)  |
|   |  10 - 15 | 23 | 5.1 (0.0, 4.5, 19.0) | 86.6 | 0.01 (0.00, 0.00, 0.08) | 0.4 (0.0, 0.2, 2.7) | 0.2 (0.0, 0.0, 1.5)  |
|   |  15 - 20 | 11 | 5.0 (0.0, 3.0, 17.0) | 74.8 | 0.02 (0.00, 0.00, 0.11) | 0.8 (0.0, 0.4, 2.8) | 0.1 (0.0, 0.0, 0.5)  |
|   |  20 - 30 | 9 | 3.5 (0.0, 1.5, 19.5) | 96.5 | 0.01 (0.00, 0.00, 0.05) | 0.8 (0.0, 0.0, 5.1) | 0.4 (0.0, 0.3, 1.7)  |
|   |  30 - 100 | 3 | 0.9 (0.0, 1.3, 1.4) | 90.2 | 0.00 (0.00, 0.00, 0.01) | 0.1 (0.0, 0.0, 0.3) | 0.3 (0.0, 0.5, 0.5)  |
|   |  100 - 500 | 3 | 0.1 (0.0, 0.2, 0.2) | 87.3 | 0.00 (0.00, 0.00, 0.00) | 0.2 (0.0, 0.0, 0.5) | 0.0 (0.0, 0.0, 0.0)  |
|   |  500 + | 12 | 0.1 (0.0, 0.1, 0.3) | 99.9 | 0.00 (0.00, 0.00, 0.00) | 1.0 (0.0, 0.5, 3.6) | 0.8 (0.0, 0.2, 3.2)  |
|   |  Total | 87 | 4.3 (0.0, 2.4, 35.4) | 81.4 | 0.02 (0.00, 0.00, 0.31) | 0.5 (0.0, 0.3, 5.1) | 0.3 (0.0, 0.0, 3.2)  |
|  LOSS | 5 - 7 | 36 | 6.9 (0.0, 7.4, 21.1) | 65.2 | 0.03 (0.00, 0.00, 0.20) | 0.1 (0.0, 0.0, 1.1) | 0.4 (0.0, 0.3, 1.5)  |
|   |  7 - 10 | 39 | 6.6 (0.0, 4.5, 28.6) | 68.2 | 0.02 (0.00, 0.00, 0.22) | 0.1 (0.0, 0.0, 0.9) | 0.5 (0.0, 0.3, 2.2)  |
|   |  10 - 15 | 42 | 9.0 (0.0, 3.6, 66.1) | 81.0 | 0.05 (0.00, 0.00, 0.60) | 0.8 (0.0, 0.0, 6.8) | 0.5 (0.0, 0.3, 3.0)  |
|   |  15 - 20 | 18 | 7.4 (0.0, 1.7, 53.6) | 93.2 | 0.10 (0.00, 0.00, 1.42) | 1.2 (0.0, 0.1, 9.4) | 0.7 (0.0, 0.2, 9.8)  |
|   |  20 - 30 | 16 | 8.5 (0.0, 0.4, 47.3) | 87.7 | 0.01 (0.00, 0.00, 0.18) | 1.4 (0.0, 0.1, 6.9) | 0.7 (0.0, 0.0, 5.1)  |
|   |  30 - 100 | 10 | 3.9 (0.0, 0.3, 24.0) | 92.1 | 0.01 (0.00, 0.00, 0.04) | 0.1 (0.0, 0.0, 1.0) | 1.2 (0.0, 0.0, 8.9)  |
|   |  100 - 500 | 17 | 0.3 (0.0, 0.1, 1.3) | 99.9 | 0.00 (0.00, 0.00, 0.00) | 0.3 (0.0, 0.0, 2.6) | 0.2 (0.0, 0.2, 0.4)  |
|   |  500 + | 15 | 0.1 (0.0, 0.1, 0.2) | 100.0 | 0.00 (0.00, 0.00, 0.00) | 0.3 (0.0, 0.0, 3.7) | 0.3 (0.0, 0.0, 1.1)  |
|   |  Total | 193 | 6.2 (0.0, 1.3, 66.1) | 82.4 | 0.03 (0.00, 0.00, 1.42) | 0.5 (0.0, 0.0, 9.4) | 0.5 (0.0, 0.0, 9.8)  |
|  ALL CNVs | Total | 280 | 5.6 (0.0, 1.9, 66.1) | 82.1 | 0.03 (0.00, 0.00, 1.42) | 0.5 (0.0, 0.0, 9.4) | 0.4 (0.0, 0.0, 9.8)  |
|  cnLOH | 100 - 200 | 25 | 4.8 (0.0, 4.8, 13.4) | 67.8 | 0.01 (0.00, 0.00, 0.11) | 4.7 (0.0, 0.3, 25.6) | 3.5 (0.0, 0.5, 17.2)  |
|   |  200 - 500 | 13 | 7.0 (4.1, 6.1, 12.5) | 98.2 | 0.00 (0.00, 0.00, 0.02) | 9.7 (0.0, 5.7, 32.3) | 8.0 (0.0, 2.9, 28.4)  |
|   |  > 500 | 7 | 6.0 (4.1, 5.4, 9.6) | 98.6 | 0.01 (0.00, 0.00, 0.04) | 32.2 (1.1, 36.4, 74.8) | 9.3 (0.0, 2.6, 42.9)  |
|   |  Total | 45 | 5.6 (0.0, 5.4, 13.4) | 84.7 | 0.01 (0.00, 0.00, 0.11) | 10.4 (0.0, 3.0, 74.8) | 5.7 (0.0, 0.9, 42.9)  |

{10}

Table 4. Reproducibility of Aberration Breakpoints by Size (kb)

|   |   |   | % CV Aberration Length | Average % Overlap | % Median Absolute Endpoint Deviation (x100) | SD Left Endpoint (# Probes) | SD Right Endpoint (# Probes)  |
| --- | --- | --- | --- | --- | --- | --- | --- |
|  Aberration Type | Aberration Range (kb) | N | Mean (Min, Median, Max) |   | Mean (Min, Median, Max)  |   |   |
|  GAIN | 20 - 50 | 6 | 4.2 (0.0, 0.0, 14.7) | 65.5 | 0.00 (0.00, 0.00, 0.00) | 1.0 (0.0, 0.0, 3.5) | 0.0 (0.0, 0.0, 0.0)  |
|   |  50 - 100 | 6 | 8.1 (0.0, 5.0, 27.3) | 80.5 | 0.01 (0.00, 0.00, 0.03) | 5.2 (0.0, 0.7, 21.9) | 1.5 (0.0, 0.0, 5.4)  |
|   |  100 - 200 | 16 | 4.0 (0.0, 0.0, 26.5) | 67.8 | 0.00 (0.00, 0.00, 0.00) | 2.5 (0.0, 0.0, 36.2) | 3.6 (0.0, 0.0, 25.1)  |
|   |  200 - 500 | 30 | 5.5 (0.0, 0.0, 73.9) | 72.3 | 0.02 (0.00, 0.00, 0.36) | 11.2 (0.0, 0.0, 264.8) | 11.6 (0.0, 0.0, 120.1)  |
|   |  500 - 1000 | 6 | 6.3 (0.0, 0.6, 23.6) | 69.0 | 0.00 (0.00, 0.00, 0.01) | 40.6 (0.0, 4.1, 136.6) | 1.3 (0.0, 0.0, 7.9)  |
|   |  1000 - 2000 | 4 | 3.2 (0.0, 1.7, 9.5) | 74.4 | 0.01 (0.00, 0.00, 0.02) | 9.9 (0.0, 0.0, 39.7) | 48.7 (0.0, 7.9, 178.8)  |
|   |  2000 - 5000 | 18 | 3.0 (0.0, 0.6, 37.3) | 88.1 | 0.00 (0.00, 0.00, 0.06) | 54.7 (0.0, 4.7, 751.3) | 18.8 (0.0, 0.0, 145.9)  |
|   |  5000 + | 17 | 0.1 (0.0, 0.1, 0.5) | 99.9 | 0.00 (0.00, 0.00, 0.00) | 28.5 (0.0, 11.1, 162.5) | 6.0 (0.0, 0.0, 48.2)  |
|   |  Total | 103 | 4.0 (0.0, 0.1, 73.9) | 80.7 | 0.01 (0.00, 0.00, 0.36) | 21.0 (0.0, 0.0, 751.3) | 10.3 (0.0, 0.0, 178.8)  |
|  LOSS | 10 - 50 | 20 | 4.0 (0.0, 0.0, 35.0) | 74.6 | 0.00 (0.00, 0.00, 0.00) | 0.0 (0.0, 0.0, 0.3) | 0.9 (0.0, 0.0, 7.9)  |
|   |  50 - 100 | 4 | 5.4 (0.0, 4.7, 12.1) | 88.1 | 0.00 (0.00, 0.00, 0.00) | 4.9 (0.0, 4.6, 10.2) | 0.0 (0.0, 0.0, 0.0)  |
|   |  100 - 200 | 33 | 9.1 (0.0, 0.0, 187.7) | 89.9 | 0.00 (0.00, 0.00, 0.11) | 11.9 (0.0, 0.0, 257.0) | 2.2 (0.0, 0.0, 27.6)  |
|   |  200 - 500 | 24 | 3.6 (0.0, 0.0, 31.5) | 83.8 | 0.02 (0.00, 0.00, 0.35) | 4.2 (0.0, 0.0, 45.1) | 9.5 (0.0, 0.0, 111.8)  |
|   |  500 - 1000 | 31 | 9.8 (0.0, 2.1, 45.6) | 86.5 | 0.00 (0.00, 0.00, 0.07) | 30.0 (0.0, 0.0, 259.4) | 44.5 (0.0, 0.0, 223.5)  |
|   |  1000 - 2000 | 15 | 2.1 (0.0, 0.0, 22.1) | 93.7 | 0.00 (0.00, 0.00, 0.02) | 15.7 (0.0, 0.0, 235.5) | 10.4 (0.0, 0.0, 64.6)  |
|   |  2000 - 5000 | 6 | 3.5 (0.0, 0.1, 17.9) | 98.2 | 0.00 (0.00, 0.00, 0.00) | 11.0 (0.0, 0.0, 66.1) | 10.9 (0.0, 0.0, 60.1)  |
|   |  5000 + | 33 | 1.1 (0.0, 0.0, 15.8) | 98.8 | 0.00 (0.00, 0.00, 0.04) | 20.3 (0.0, 0.0, 216.8) | 60.0 (0.0, 0.0, 870.1)  |
|   |  Total | 166 | 5.3 (0.0, 0.0, 187.7) | 89.3 | 0.00 (0.00, 0.00, 0.35) | 14.5 (0.0, 0.0, 259.4) | 23.5 (0.0, 0.0, 870.1)  |
|  ALL CNVs | Total | 269 | 4.8 (0.0, 0.0, 187.7) | 86.3 | 0.01 (0.00, 0.00, 0.36) | 17.0 (0.0, 0.0, 751.3) | 18.4 (0.0, 0.0, 870.1)  |
|  cnLOH | 5000 - 10000 | 23 | 3.5 (0.0, 1.7, 17.7) | 63.6 | 0.01 (0.00, 0.00, 0.11) | 132.6 (0.0, 4.2, 1637.7) | 206.1 (0.0, 92.4, 938.9)  |
|   |  10000 - 20000 | 18 | 9.8 (0.0, 8.4, 27.5) | 80.1 | 0.02 (0.00, 0.00, 0.17) | 791.8 (0.0, 890.2, 2038.5) | 536.4 (0.0, 33.9, 2487.3)  |
|   |  20000 + | 5 | 1.7 (0.4, 1.7, 3.0) | 91.3 | 0.00 (0.00, 0.00, 0.01) | 554.6 (14.3, 262.4, 1819) | 172.3 (15.4, 86.9, 631.1)  |
|   |  Total | 46 | 5.7 (0.0, 2.3, 27.5) | 73.7 | 0.01 (0.00, 0.00, 0.17) | 436.4 (0.0, 107.3, 2038.5) | 331.7 (0.0, 76.6, 2487.3)  |

{11}

A second supplemental study was conducted to increase the number of samples and representative CNV regions to align with the predicate evaluation using a panel of 48 genomic DNA encompassing a wide variety of chromosomal aberrations of interest. A total of 315 unique aberrations (103 CN gains, 166 CN losses, and 46 cnLOH) were detected from the 48 gDNA samples. These aberrations were distributed across all 24 chromosomes in the human genome, with collective genome coverage of 42.0% (18.6% for CN gains, 11.6% for CN losses, 27.1% for all CNVs, and 18.4% for cnLOH).

A total of nine (9) replicates for each sample in the diverse 48 aberrant gDNA panel were processed by multiple operators using combinations of three (3) reagent lots and three (3) scanner instruments, across three (3) processing weeks at a single site, for a total of 432 data points. The study spanned across 3 weeks (runs, or sample processing batches).

Individual aberrations called within each processed test sample were compared to their respective replicates (9 replicates for each aberration representing 3x3 reagent lot/scanner combinations) by pairwise replicate analysis (PRA), requiring at least 50% overlap of chromosomal coordinates for confirmation (agreement). Agreement was assessed separately for small copy number variants (CNVs, 5-20 probes contained within the aberration), larger CNVs (>20 probes), or cnLOH regions. The results demonstrate that the pre-defined acceptance criteria were met for each category with a pairwise replicate agreement of 83.33%, 98.39%, and 80.80%, respectively. Results were similar when using a more stringent 80% overlap criterion for pairwise replicate agreement. In addition, no substantial differences were observed when the pairwise replicate agreement was assessed separately for inter-lot vs. intra-lot replicate pairs, or for inter-scanner vs. intra-scanner replicate pairs.

## Study 2. Lot-to-Lot reproducibility:

A lot-to-lot reproducibility study was conducted using first a set of 6 samples and then a second set of 48 samples containing a range of chromosomal aberrations (gains, losses, and cnLOH). Samples were processed by one (1) operator, using combinations of three (3) reagent lots analyzed on three (3) scanner instruments across five (5) processing weeks at a single site. Aberrations were distributed across most of the 24 chromosomes in the human genome, with the exception of chromosomes 7, 17, 19, and 20. Data analysis methods were the same as those used in the site-to-site reproducibility study.

Individual aberrations called within each processed test sample were compared to their respective replicates (9 replicates for each aberration, representing 3x3 reagent-lot/scanner combinations) by pairwise replicate analysis (PRA), requiring at least 50% overlap of chromosomal coordinates for confirmation. Agreement was assessed separately for small copy number variants (CNVs, 5-20 probes contained within the aberration), larger CNVs (>20 probes), or cnLOH regions. The results demonstrate that the pre-defined acceptance criteria were met for each category with a pairwise

9

{12}

replicate agreement of 83.33%, 98.39%, and 80.80%, respectively. Results were similar when using a more stringent 80% overlap criteria for pairwise replicate agreement. In addition, no substantial differences were observed when the pairwise replicate agreement was assessed separately for inter-lot vs. intra-lot replicate pairs, or for inter-scanner vs. intra-scanner replicate pairs.

Data were further refined by size, probe number, aberration type, and study variable (e.g. reagent lot, scanner, processing week). Alternative metrics of positive percent agreement, call rate, and breakpoint accuracy/endpoint deviation were conducted and consistent with data observed in the site-to-site reproducibility.

The results demonstrated that when comparing all replicates of all test samples across all lots, scanners, and weeks, using a 50% aberration overlap criteria, the overall pairwise replicate agreement across all sizes of copy number gains and losses was 89.0%. Pairwise replicate agreement across the various kb bins ranged from 76.9% to 100%. For copy number gains, the overall agreement was 85.4%; for copy number losses, the overall agreement was 91.3%. For cnLOH, the overall pairwise replicate agreement was 80.8%. Applying a more stringent 80% overlap criteria produced overall agreements of 87.2% for all CNVs (gains or losses) combined, 84.0% for CN gains, 89.2% for CN losses, and 76.0% for cnLOH.

When assessing specifically the agreement between replicates for positive aberration calls by PPA analysis, the agreement was 93.0% for all copy number calls and 87.3% for cnLOH using the 50% overlap criteria. Call rate averaged 83.0% for CNVs, and 75.6% for cnLOH.

## DNA extraction precision:

A panel of twenty-four (24) samples was tested by each operator. The gDNA from the samples was extracted using the same lot of the QIAamp DSP DNA Blood Mini kit, at one site, in duplicate, by 3 operators, on 3 separate days, for a total of 432 extractions (3 operators x 3 days x 2 duplicates x 24 samples). Each of the three operators labelled a set of 48 extracted samples (24 different samples x 2 extractions) per week for 3 weeks for a total of 144 test results run per operator over the course of the study. The extracted gDNA replicates were tested in the GenetiSure Dx Postnatal Assay in 3 weeks, each corresponding to a specific day of extraction.

A total of 160 unique aberrations (65 gains, 79 losses, and 16 cnLOH) were detected from the 24 gDNA samples across their replicates. These aberrations were distributed across all chromosomes in the human genome with the exception of Chr17 and Chr20, with a collective coverage of 25.1% (13.8% for gains, 8.8% for losses, 16.6% for gains or losses, and 8.9% for cnLOH). Primary analysis was performed using pairwise comparison of aberration results on each of the 18 replicates (3 operators x 3 days x 2 duplicates) for each sample. An aberration was considered confirmed if at least 50% of the region of aberration overlapped between the replicates being compared.

The individual Pairwise Replicate Agreement % values stratified by week or operator

10

{13}

were similar to each other and the overall agreement averages 82.17% for CNVs called by 5-20 probes, 98.47% for CNVs called by >20 probes, and 81.15% for cnLOH, which further supports that similar assay performance can be expected from different extractions, personnel, days, and samples.

b. Linearity/assay reportable range:

N/A

c. Traceability, Stability, Expected values (controls, calibrators, or methods):

## Array and Reagent Stability:

Stability studies were performed to demonstrate the stability of the GenetiSure Dx Postnatal Assay and reagents (controls were included in the evaluation). Stability testing consisted of four different arms of the study:

- Shelf Life: testing after storage of assay components under recommended conditions at defined intervals after manufacturing.
- Multi-use testing: testing after components of the assay that are stored at -15°C to -25°C are subject to different numbers of freeze/thaw cycles.
- Transport: Similar to Shelf life testing, but assay components were subject to a transportation simulation prior to storage at recommended conditions.
- In-use testing: testing after different components of the assay are stored for various times after preparation or testing where reaction intermediates are stored for various times.

A leukocyte-depleted whole blood sample spiked in with human B-lymphocyte cell with known genotype, a whole blood from an anonymous male donor, and the Agilent male reference DNA combined with Agilent female reference DNA were used with a minimum of 2 replicates at a minimum of 4 time points in the study.

The final stability was determined based on the time point prior to that at which the 95% confidence interval of any of the QC metrics first overlaps with the acceptance criteria. Interim stability is defined based on the maximum duration for which the 95% confidence intervals for all QC metrics do not overlap with the acceptance criteria.

Results of the study to date indicate that all tested conditions pass all metrics under all conditions tested, supporting a shelf life of the assay components of 12 months. In addition, multi-use and in-use testing support the tested conditions of the assay:

All reagents should be stored in the dark. The stability claims are as follows:

- 8x freeze/thaw for Labeling Kit components;
- 16x freeze/thaw or 120 days' storage at -15°C to -25°C for reconstituted 10x Blocking Agent;

11

{14}

- 16x freeze/thaw for Cot-1 DNA;
- 60 days' storage of an open microarray package under specified conditions;
- 30 days' storage of the digested gDNA or labeled DNA reaction intermediates at -15°C to -25°C.

## Whole Blood Stability

To determine the stability of whole blood specimens prior to gDNA isolation, 24 whole blood specimens, 12 male and 12 female, were obtained from a blood bank and gDNA was isolated from the specimens at 1, 3, 7, and 10 days after initial collection.

A list of aberrations for each sample extracted on Days 3, 7 and 10 were reported and compared with the 'Day 1' list for the same sample. An aberration from the 'Day 1' sample was considered confirmed in the stored samples if the test result identifies a region of aberration that overlaps the 'Day 1' region by at least 50%. Blood was considered stable when stored for a given time when 75% of small gain/loss aberrations, called by 5-20 copy number probes, and 90% of the larger gain/loss aberrations, called by >20 probes, were confirmed.

An additional analysis was also performed using an 80% overlap criterion. The results of both the 50% and 80% overlap analysis methods demonstrated that whole blood specimens may be stored for up to 10 days at 2–8°C prior to gDNA isolation. Samples stored for this period of time and processed with the GenetiSure Dx Postnatal Assay produced acceptable results.

### d. Detection limit:

#### DNA input:

The amount of genomic DNA recommended for testing per samples with the GenetiSure Dx Postnatal Assay is 500 ng. To determine the performance of the assay across a range of genomic DNA input concentrations, 24 gDNA samples with known chromosomal aberrations were tested. These DNA samples were tested in the assay using 2 lots of reagents across a range of varied DNA input levels from 0.125 μg (125 ng) to 1 μg (1000 ng), with 0.5 μg (500 ng) as the recommended input quantity (standard).

The study assessed the impact of various gDNA input on aberration calling and determined the upper and lower limits of detection (ULOD and LLOD) of the assay by comparing the percentage of aberrations confirmed at each non-standard DNA input level against pre-defined acceptance criteria. The study data, and supplemental data generated under similar study conditions, support a conservative LLOD at 375 ng and a common ULOD at 1000 ng for both copy number and cnLOH aberrations. For copy number aberrations only, the LLOD could be further reduced to 250 ng. Data from this study support the use of 500 ng as the recommended input amount.

12

{15}

13

Mosaicism:

To determine the level of mosaicism reliably detected by the GenetiSure Dx Postnatal Assay, 24 aberrant cell line DNAs containing known copy number changes were mixed with a reference background DNA in different percentages to mimic various levels of mosaicism. 12 male and 12 female DNA samples, each with at least 4 previously identified aberrations, were each mixed with sex-matched reference DNA in the following ratios: 1:0, 9:1, 3:1, 1:1; 1:3, 1:9, and 0:1.

Large copy number aberrations could be reliably detected when present in a 50% or greater admixture. Some aberrations were correctly identified at lower than a 50% level, but the sensitivity of detection was reduced. Results were similar for both gains and losses. Smaller aberrations could not be reliability detected in any of the admixtures.

e. Analytical specificity:

Interfering Substances:

To assess the impact of interfering substances on the GenetiSure Dx Postnatal Assay, a study evaluating the impact of hemoglobin, conjugated bilirubin, unconjugated bilirubin and triglycerides (triolein) spiked into whole blood prior to gDNA isolation was conducted. Blood drawn from 12 phenotypically normal males and 12 phenotypically normal females was used in the testing. The list of aberrations for each sample containing a given interferent was reported and compared with the ‘non-adulterated control’ list for the same sample. An aberration was considered confirmed if the test result identified a region of aberration that overlaps between the sample and control by at least 50%. The test was considered robust to a given interferent when 75% of the gains/losses in the 5-20 probe category and 90% of the gains/losses in the >20 probe category in the ‘non-adulterated control’ were confirmed. No interference was observed with any of the tested conditions. The results of the study met the acceptance criteria.

Cross-Contamination:

To determine the effect of potential gDNA carry-over from 1 array to the next when processing multiple arrays, 2 male and 2 female gDNA samples from cell lines, each with distinctive sets of known chromosomal aberrations, were tested across multiple microarray slides under conditions that would either allow or prevent detection of cross contamination between the adjacent arrays on the slides. Four (4) microarray slides served as the “non-contaminated condition” with four replicates of the same sample placed on each of the four arrays of the slides. Six (6) slides served as a test for “potential cross-contamination” that could occur between adjacent arrays within a single slide during the hybridization set-up or overnight incubation. For these slides, the sample replicates were alternated on the slide with sample replicates from a different sample. The CNV and cnLOH aberration results from the “potential cross-contamination” microarray slides were compared to the aberration results from the “non-contaminated condition” microarray slides, using a 50% overlap criterion, to determine if detectable cross contamination had occurred on the test slides.

{16}

Additionally, gasket related cross-contamination was evaluated by use of three (3) different lots of gasket slides. No suspected cross contamination was detected.

f. Assay cut-off:

N/A

2. Comparison studies:

a. Accuracy — comparison to orthogonal methods:

Accuracy of the GenetiSure Dx Postnatal Assay results was assessed by comparing the CNVs identified by GenetiSure Dx Postnatal Assay to the results obtained using comparator microarray methods. A total of 556 out of 626 samples were eligible for testing based. Samples were excluded based on pre-specified exclusion criteria (e.g., lack of informed consent, lack of sufficient gDNA). The sample panel consisted of 451 aberrant genomic DNA (gDNA) samples derived from established commercial cell lines, 76 archived clinical gDNA samples isolated from whole blood specimens of anonymized patients, 5 globally recognized syndrome reference panel gDNA samples, and 24 fresh blood-derived gDNA samples extracted from whole blood of phenotypically normal subjects. The samples were selected to maximize the variation across the genome with consideration for gain and loss segments of various sizes/number of probes, chromosomal representation, CNV regions in genic and non-genic regions, and in telomeric and centromeric regions. A total of 2187 CNV regions covered 91% of the genome and were more prevalent in non-telomeric/non-centromeric regions than in telomeric/centromeric (1337 regions vs 1130 regions). A total of 21% (534 out of 2187 regions) of the CNVs had high (>45%) GC content.

Due to the lack of an applicable universal comparator, two independent (non-Agilent) commercially available microarray based assays analytically validated for copy number detection were employed to assist accuracy assessment of CNV aberrations and resolve discrepancies. The samples were tested with the GenetiSure Dx Postnatal Assay using standard procedures in a designated Agilent laboratory.

Three methods were used to analyze the data (Figure 1). Method 1 and Method 2 compared Agilent results with the two platform results each separately. Method 3 generated a composite dataset by consolidating results from these two platforms. A description of the differences between the methods is shown in Figure 1 and Table 5. The main differences were in the definition of comparator and overlap.

A target Agilent aberration was deemed "confirmed" if a minimum of percent overlap was found with comparator aberration call(s) of the same type (gain, loss, or cnLOH). Method 1 and 2 used 50% overlap, and method 3 used a 65% overlap criterion. All eligible Agilent aberrations were assessed. If an Agilent CNV aberration could not be "confirmed" by a microarray-based comparator, another analytically validated method was employed to adjudicate the results. To avoid bias in the assessment, an additional

14

{17}

5% randomly selected "confirmed" CNV aberrations (separately selected for CNVs>20 probes and CNVs with 5-20 probes) were also included in the discrepancy resolution testing. Other CNVs directly subject to a third method confirmation included those selected from normal whole blood samples near the limit of resolution.

![img-0.jpeg](img-0.jpeg)
Figure1. Methods to assess the accuracy

![img-1.jpeg](img-1.jpeg)

![img-2.jpeg](img-2.jpeg)

{18}

Table 5. Comparison of Assessment Criteria

|   | Comparator applicable to |   | Calculation of Overlap % |   |   | Minimum Overlap % considered as “confirmed”  |
| --- | --- | --- | --- | --- | --- | --- |
|   | Cell line Samples | Clinical lab Samples | Numerator | Denominator | How to treat Multiple comparator aberrations overlapping a single Agilent aberration  |   |
|  Method 1 | Comparator platform 1 and 2 compared separately | Comparator platform 1 only (same for all criteria) | Length of overlap summed | Union (merged length) of the Agilent aberration and overlapping comparator aberrations(s) | All overlaps considered together | 50%  |
|  Method 2 |   |   | Length of overlap itemized | Union (merged length) of the Agilent aberration and one overlapping comparator aberrations | Each overlap considered separately, one at a time. The maximum % overlap was chosen to assess confirmation  |   |
|  Method 3 | Composite Data set |   | Length of overlap summed | Length of the Agilent aberration | All overlaps considered together | 65%  |

The results are analyzed as confirmation rate/ false positive rate and are summarized for each method either including indeterminate CNVs as unconfirmed (referred to as scheme a) or excluding indeterminate CNVs (referred to as scheme b). The data in each table is stratified by copy number state, size (kb) or probe number, and genomic region (Tables 6-17). The results were generally consistent across all methods but differed dependent on probe vs size and whether indeterminate CNVs were included or excluded: gains ranged

In addition, refined aberration size binning either by probe number or length in kilobases (kb) was carried out in the accuracy evaluation. 25/26, or  $96.2\%$  confirmation rate near the resolution limit for CNV detection was confirmed using the aforementioned qPCR technology on selected small CNVs detected in normal whole blood-derived samples. Alternative assessment criteria for aberration confirmation (agreement) were also performed and produced comparable results. Breakpoint accuracy was evaluated on confirmed aberrations. Breakpoint agreement with comparators were  $91.0\%$  for CNVs and  $91.4\%$  for cnLOH.

{19}

Table 6. Accuracy based on number of probes
(Method 1, Scheme a: Including Indeterminate CNVs as "not confirmed")

|  TYPE | Aberration Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  Gain | 5-7 | 48 | 38 | 79.2% (65.7%, 88.3%) | 20.8% (11.7%, 34.3%)  |
|   |  7-10 | 101 | 85 | 84.2% (75.8%, 90.0%) | 15.8% (10.0%, 24.2%)  |
|   |  10-15 | 197 | 152 | 77.2% (70.8%, 82.5%) | 22.8% (17.5%, 29.2%)  |
|   |  15-20 | 101 | 83 | 82.2% (73.6%, 88.4%) | 17.8% (11.6%, 26.4%)  |
|   |  20-50 | 148 | 104 | 70.3% (62.5%, 77.0%) | 29.7% (23.0%, 37.5%)  |
|   |  50-500 | 82 | 64 | 78.0% (67.9%, 85.6%) | 22.0% (14.4%, 32.1%)  |
|   |  500 + | 169 | 166 | 98.2% (94.9%, 99.4%) | 1.8% (0.6%, 5.1%)  |
|   |  Total | 846 | 692 | 81.8% (79.1%, 84.3%) | 18.2% (15.7%, 20.9%)  |
|  Loss | 5-7 | 216 | 196 | 90.7% (86.1%, 93.9%) | 9.3% (6.1%, 13.9%)  |
|   |  7-10 | 202 | 165 | 81.7% (75.8%, 86.4%) | 18.3% (13.6%, 24.2%)  |
|   |  10-15 | 257 | 216 | 84.0% (79.1%, 88.0%) | 16.0% (12.0%, 20.9%)  |
|   |  15-20 | 125 | 90 | 72.0% (63.6%, 79.1%) | 28.0% (20.9%, 36.4%)  |
|   |  20-50 | 130 | 95 | 73.1% (64.9%, 80.0%) | 26.9% (20.0%, 35.1%)  |
|   |  50-500 | 225 | 217 | 96.4% (93.1%, 98.2%) | 3.6% (1.8%, 6.9%)  |
|   |  500 + | 186 | 180 | 96.8% (93.1%, 98.5%) | 3.2% (1.5%, 6.9%)  |
|   |  Total | 1341 | 1159 | 86.4% (84.5%, 88.2%) | 13.6% (11.8%, 15.5%)  |
|  All CNVs |   | 2187 | 1851 | 84.6% (83.1%, 86.1%) | 15.4% (13.9%, 16.9%)  |
|  cnLOH | 100-200 | 132 | 94 | 71.2% (63.0%, 78.2%) | 28.8% (21.8%, 37.0%)  |
|   |  200-500 | 102 | 96 | 94.1% (87.8%, 97.3%) | 5.9% (2.7%, 12.2%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 292 | 248 | 84.9% (80.4%, 88.6%) | 15.1% (11.4%, 19.6%)  |

{20}

Table 7. Accuracy based on number of probes
(Method 1, Scheme b: Excluding Indeterminate CNVs)

|  TYPE | Aberration Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  Gain | 5-7 | 43 | 38 | 88.4% (75.5%, 94.9%) | 11.6% (5.1%, 24.5%)  |
|   |  7-10 | 91 | 85 | 93.4% (86.4%, 96.9%) | 6.6% (3.1%, 13.6%)  |
|   |  10-15 | 175 | 152 | 86.9% (81.1%, 91.1%) | 13.1% (8.9%, 18.9%)  |
|   |  15-20 | 91 | 83 | 91.2% (83.6%, 95.5%) | 8.8% (4.5%, 16.4%)  |
|   |  20-50 | 124 | 104 | 83.9% (76.4%, 89.3%) | 16.1% (10.7%, 23.6%)  |
|   |  50-500 | 72 | 64 | 88.9% (79.6%, 94.3%) | 11.1% (5.7%, 20.4%)  |
|   |  500 + | 169 | 166 | 98.2% (94.9%, 99.4%) | 1.8% (0.6%, 5.1%)  |
|   |  Total | 765 | 692 | 90.5% (88.2%, 92.3%) | 9.5% (7.7%, 11.8%)  |
|  Loss | 5-7 | 197 | 196 | 99.5% (97.2%, 99.9%) | 0.5% (0.1%, 2.8%)  |
|   |  7-10 | 183 | 165 | 90.2% (85.0%, 93.7%) | 9.8% (6.3%, 15.0%)  |
|   |  10-15 | 231 | 216 | 93.5% (89.6%, 96.0%) | 6.5% (4.0%, 10.4%)  |
|   |  15-20 | 102 | 90 | 88.2% (80.6%, 93.1%) | 11.8% (6.9%, 19.4%)  |
|   |  20-50 | 114 | 95 | 83.3% (75.4%, 89.1%) | 16.7% (10.9%, 24.6%)  |
|   |  50-500 | 222 | 217 | 97.7% (94.8%, 99.0%) | 2.3% (1.0%, 5.2%)  |
|   |  500 + | 184 | 180 | 97.8% (94.5%, 99.2%) | 2.2% (0.8%, 5.5%)  |
|   |  Total | 1233 | 1159 | 94.0% (92.5%, 95.2%) | 6.0% (4.8%, 7.5%)  |
|  All CNVs |   | 1998 | 1851 | 92.6% (91.4%, 93.7%) | 7.4% (6.3%, 8.6%)  |
|  cnLOH | 100-200 | 132 | 94 | 71.2% (63.0%, 78.2%) | 28.8% (21.8%, 37.0%)  |
|   |  200-500 | 99 | 96 | 97.0% (91.5%, 99.0%) | 3.0% (1.0%, 8.5%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 289 | 248 | 85.8% (81.3%, 89.4%) | 14.2% (10.6%, 18.7%)  |

{21}

Table 8. Accuracy based on CNV size (kb)
(schemes a and b) (Method 1, Including Indeterminate CNVs as "Not Confirmed")

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  Gain | 20-100 | 69 | 37 | 53.6% (42.0%, 64.9%) | 46.4% (35.1%, 58.0%)  |
|   |  100-200 | 94 | 75 | 79.8% (70.6%, 86.7%) | 20.2% (13.3%, 29.4%)  |
|   |  200-300 | 136 | 120 | 88.2% (81.7%, 92.6%) | 11.8% (7.4%, 18.3%)  |
|   |  300-500 | 123 | 92 | 74.8% (66.5%, 81.6%) | 25.2% (18.4%, 33.5%)  |
|   |  500-1,000 | 90 | 70 | 77.8% (68.2%, 85.1%) | 22.2% (14.9%, 31.8%)  |
|   |  1000-10,000 | 168 | 136 | 81.0% (74.3%, 86.2%) | 19.0% (13.8%, 25.7%)  |
|   |  10,000 + | 166 | 162 | 97.6% (94.0%, 99.1%) | 2.4% (0.9%, 6.0%)  |
|   |  Total | 846 | 692 | 81.8% (79.1%, 84.3%) | 18.2% (15.7%, 20.9%)  |
|  Loss | 10-100 | 88 | 59 | 67.0% (56.7%, 76.0%) | 33.0% (24.0%, 43.3%)  |
|   |  100-200 | 207 | 180 | 87.0% (81.7%, 90.9%) | 13.0% (9.1%, 18.3%)  |
|   |  200-300 | 129 | 114 | 88.4% (81.7%, 92.8%) | 11.6% (7.2%, 18.3%)  |
|   |  300-500 | 116 | 103 | 88.8% (81.8%, 93.3%) | 11.2% (6.7%, 18.2%)  |
|   |  500-1,000 | 209 | 164 | 78.5% (72.4%, 83.5%) | 21.5% (16.5%, 27.6%)  |
|   |  1000-10,000 | 398 | 351 | 88.2% (84.6%, 91.0%) | 11.8% (9.0%, 15.4%)  |
|   |  10,000 + | 194 | 188 | 96.9% (93.4%, 98.6%) | 3.1% (1.4%, 6.6%)  |
|   |  Total | 1341 | 1159 | 86.4% (84.5%, 88.2%) | 13.6% (11.8%, 15.5%)  |
|  All CNVs |   | 2187 | 1851 | 84.6% (83.1%, 86.1%) | 15.4% (13.9%, 16.9%)  |
|  cnLOH | 5,000-10,000 | 93 | 61 | 65.6% (55.5%, 74.5%) | 34.4% (25.5%, 44.5%)  |
|   |  10,000-20,000 | 94 | 84 | 89.4% (81.5%, 94.1%) | 10.6% (5.9%, 18.5%)  |
|   |  20,000 + | 105 | 103 | 98.1% (93.3%, 99.5%) | 1.9% (0.5%, 6.7%)  |
|   |  Total | 292 | 248 | 84.9% (80.4%, 88.6%) | 15.1% (11.4%, 19.6%)  |

{22}

Table 9. Accuracy based on CNV size (kb)
(Method 1, Scheme b: including Excluding CNVs)

|  TYPE | Aberration Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  Gain | 5-7 | 43 | 38 | 88.4% (75.5%, 94.9%) | 11.6% (5.1%, 24.5%)  |
|   |  7-10 | 91 | 85 | 93.4% (86.4%, 96.9%) | 6.6% (3.1%, 13.6%)  |
|   |  10-15 | 175 | 152 | 86.9% (81.1%, 91.1%) | 13.1% (8.9%, 18.9%)  |
|   |  15-20 | 91 | 83 | 91.2% (83.6%, 95.5%) | 8.8% (4.5%, 16.4%)  |
|   |  20-50 | 124 | 104 | 83.9% (76.4%, 89.3%) | 16.1% (10.7%, 23.6%)  |
|   |  50-500 | 72 | 64 | 88.9% (79.6%, 94.3%) | 11.1% (5.7%, 20.4%)  |
|   |  500 + | 169 | 166 | 98.2% (94.9%, 99.4%) | 1.8% (0.6%, 5.1%)  |
|   |  Total | 765 | 692 | 90.5% (88.2%, 92.3%) | 9.5% (7.7%, 11.8%)  |
|  Loss | 5-7 | 197 | 196 | 99.5% (97.2%, 99.9%) | 0.5% (0.1%, 2.8%)  |
|   |  7-10 | 183 | 165 | 90.2% (85.0%, 93.7%) | 9.8% (6.3%, 15.0%)  |
|   |  10-15 | 231 | 216 | 93.5% (89.6%, 96.0%) | 6.5% (4.0%, 10.4%)  |
|   |  15-20 | 102 | 90 | 88.2% (80.6%, 93.1%) | 11.8% (6.9%, 19.4%)  |
|   |  20-50 | 114 | 95 | 83.3% (75.4%, 89.1%) | 16.7% (10.9%, 24.6%)  |
|   |  50-500 | 222 | 217 | 97.7% (94.8%, 99.0%) | 2.3% (1.0%, 5.2%)  |
|   |  500 + | 184 | 180 | 97.8% (94.5%, 99.2%) | 2.2% (0.8%, 5.5%)  |
|   |  Total | 1233 | 1159 | 94.0% (92.5%, 95.2%) | 6.0% (4.8%, 7.5%)  |
|  All CNVs |   | 1998 | 1851 | 92.6% (91.4%, 93.7%) | 7.4% (6.3%, 8.6%)  |
|  cnLOH | 100-200 | 132 | 94 | 71.2% (63.0%, 78.2%) | 28.8% (21.8%, 37.0%)  |
|   |  200-500 | 99 | 96 | 97.0% (91.5%, 99.0%) | 3.0% (1.0%, 8.5%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 289 | 248 | 85.8% (81.3%, 89.4%) | 14.2% (10.6%, 18.7%)  |

{23}

Table 10. Accuracy based on number of probes
(Method 2, Scheme a: Excluding Indeterminate CNVs)

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 20-100 | 69 | 37 | 53.6% (42.0%, 64.9%) | 46.4% (35.1%, 58.0%)  |
|   |  100-200 | 94 | 75 | 79.8% (70.6%, 86.7%) | 20.2% (13.3%, 29.4%)  |
|   |  200-300 | 136 | 120 | 88.2% (81.7%, 92.6%) | 11.8% (7.4%, 18.3%)  |
|   |  300-500 | 123 | 92 | 74.8% (66.5%, 81.6%) | 25.2% (18.4%, 33.5%)  |
|   |  500-1,000 | 90 | 70 | 77.8% (68.2%, 85.1%) | 22.2% (14.9%, 31.8%)  |
|   |  1000-10,000 | 168 | 136 | 81.0% (74.3%, 86.2%) | 19.0% (13.8%, 25.7%)  |
|   |  10,000 + | 166 | 162 | 97.6% (94.0%, 99.1%) | 2.4% (0.9%, 6.0%)  |
|   |  Total | 846 | 692 | 81.8% (79.1%, 84.3%) | 18.2% (15.7%, 20.9%)  |
|  LOSS | 10-100 | 88 | 59 | 67.0% (56.7%, 76.0%) | 33.0% (24.0%, 43.3%)  |
|   |  100-200 | 207 | 180 | 87.0% (81.7%, 90.9%) | 13.0% (9.1%, 18.3%)  |
|   |  200-300 | 129 | 114 | 88.4% (81.7%, 92.8%) | 11.6% (7.2%, 18.3%)  |
|   |  300-500 | 116 | 103 | 88.8% (81.8%, 93.3%) | 11.2% (6.7%, 18.2%)  |
|   |  500-1,000 | 209 | 164 | 78.5% (72.4%, 83.5%) | 21.5% (16.5%, 27.6%)  |
|   |  1000-10,000 | 398 | 351 | 88.2% (84.6%, 91.0%) | 11.8% (9.0%, 15.4%)  |
|   |  10,000 + | 194 | 188 | 96.9% (93.4%, 98.6%) | 3.1% (1.4%, 6.6%)  |
|   |  Total | 1341 | 1159 | 86.4% (84.5%, 88.2%) | 13.6% (11.8%, 15.5%)  |
|  ALL CNVs |   | 2187 | 1851 | 84.6% (83.1%, 86.1%) | 15.4% (13.9%, 16.9%)  |
|  cnLOH | 5,000-10,000 | 93 | 61 | 65.6% (55.5%, 74.5%) | 34.4% (25.5%, 44.5%)  |
|   |  10,000-20,000 | 94 | 84 | 89.4% (81.5%, 94.1%) | 10.6% (5.9%, 18.5%)  |
|   |  20,000 + | 105 | 103 | 98.1% (93.3%, 99.5%) | 1.9% (0.5%, 6.7%)  |
|   |  Total | 292 | 248 | 84.9% (80.4%, 88.6%) | 15.1% (11.4%, 19.6%)  |

* The number of aberrations analyzed in each range bin, including indeterminate CNVs as "not confirmed".

{24}

Table 11. Accuracy based on number of probes
(Method 2, Scheme b: including Indeterminate CNVs as "Not Confirmed")

|  TYPE | Aberration Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 5-7 | 48 | 38 | 79.2% (65.7%, 88.3%) | 20.8% (11.7%, 34.3%)  |
|   |  7-10 | 101 | 85 | 84.2% (75.8%, 90.0%) | 15.8% (10.0%, 24.2%)  |
|   |  10-15 | 197 | 151 | 76.6% (70.3%, 82.0%) | 23.4% (18.0%, 29.7%)  |
|   |  15-20 | 101 | 83 | 82.2% (73.6%, 88.4%) | 17.8% (11.6%, 26.4%)  |
|   |  20-50 | 148 | 104 | 70.3% (62.5%, 77.0%) | 29.7% (23.0%, 37.5%)  |
|   |  50-500 | 82 | 57 | 69.5% (58.9%, 78.4%) | 30.5% (21.6%, 41.1%)  |
|   |  500 + | 169 | 119 | 70.4% (63.1%, 76.8%) | 29.6% (23.2%, 36.9%)  |
|   |  Total | 846 | 637 | 75.3% (72.3%, 78.1%) | 24.7% (21.9%, 27.7%)  |
|  LOSS | 5-7 | 216 | 200 | 92.6% (88.3%, 95.4%) | 7.4% (4.6%, 11.7%)  |
|   |  7-10 | 202 | 165 | 81.7% (75.8%, 86.4%) | 18.3% (13.6%, 24.2%)  |
|   |  10-15 | 257 | 215 | 83.7% (78.6%, 87.7%) | 16.3% (12.3%, 21.4%)  |
|   |  15-20 | 125 | 89 | 71.2% (62.7%, 78.4%) | 28.8% (21.6%, 37.3%)  |
|   |  20-50 | 130 | 95 | 73.1% (64.9%, 80.0%) | 26.9% (20.0%, 35.1%)  |
|   |  50-500 | 225 | 212 | 94.2% (90.4%, 96.6%) | 5.8% (3.4%, 9.6%)  |
|   |  500 + | 186 | 176 | 94.6% (90.4%, 97.1%) | 5.4% (2.9%, 9.6%)  |
|   |  Total | 1341 | 1152 | 85.9% (83.9%, 87.7%) | 14.1% (12.3%, 16.1%)  |
|  ALL CNVs |   | 2187 | 1789 | 81.8% (80.1%, 83.4%) | 18.2% (16.6%, 19.9%)  |
|  cnLOH | 100-200 | 132 | 94 | 71.2% (63.0%, 78.2%) | 28.8% (21.8%, 37.0%)  |
|   |  200-500 | 102 | 95 | 93.1% (86.5%, 96.6%) | 6.9% (3.4%, 13.5%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 292 | 247 | 84.6% (80.0%, 88.3%) | 15.4% (11.7%, 20.0%)  |

* The number of aberrations analyzed in each range bin, including indeterminate CNVs as "not confirmed".

{25}

Table 12. Accuracy based on CNV size (kb)
(Method 2, Scheme a: Excluding Indeterminate CNVs)

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 20-100 | 53 | 37 | 69.8% (56.5%, 80.5%) | 30.2% (19.5%, 43.5%)  |
|   |  100-200 | 87 | 75 | 86.2% (77.4%, 91.9%) | 13.8% (8.1%, 22.6%)  |
|   |  200-300 | 123 | 120 | 97.6% (93.1%, 99.2%) | 2.4% (0.8%, 6.9%)  |
|   |  300-500 | 107 | 92 | 86.0% (78.2%, 91.3%) | 14.0% (8.7%, 21.8%)  |
|   |  500-1,000 | 75 | 68 | 90.7% (82.0%, 95.4%) | 9.3% (4.6%, 18.0%)  |
|   |  1000-10,000 | 145 | 128 | 88.3% (82.0%, 92.5%) | 11.7% (7.5%, 18.0%)  |
|   |  10,000 + | 120 | 117 | 97.5% (92.9%, 99.1%) | 2.5% (0.9%, 7.1%)  |
|   |  Total | 710 | 637 | 89.7% (87.3%, 91.7%) | 10.3% (8.3%, 12.7%)  |
|  LOSS | 10-100 | 76 | 59 | 77.6% (67.1%, 85.5%) | 22.4% (14.5%, 32.9%)  |
|   |  100-200 | 191 | 184 | 96.3% (92.6%, 98.2%) | 3.7% (1.8%, 7.4%)  |
|   |  200-300 | 115 | 114 | 99.1% (95.2%, 99.8%) | 0.9% (0.2%, 4.8%)  |
|   |  300-500 | 103 | 101 | 98.1% (93.2%, 99.5%) | 1.9% (0.5%, 6.8%)  |
|   |  500-1,000 | 179 | 164 | 91.6% (86.6%, 94.9%) | 8.4% (5.1%, 13.4%)  |
|   |  1000-10,000 | 378 | 350 | 92.6% (89.5%, 94.8%) | 7.4% (5.2%, 10.5%)  |
|   |  10,000 + | 184 | 180 | 97.8% (94.5%, 99.2%) | 2.2% (0.8%, 5.5%)  |
|   |  Total | 1226 | 1152 | 94.0% (92.5%, 95.2%) | 6.0% (4.8%, 7.5%)  |
|  ALL CNVs |   | 1936 | 1789 | 92.4% (91.1%, 93.5%) | 7.6% (6.5%, 8.9%)  |
|  cnLOH | 5,000-10,000 | 93 | 61 | 65.6% (55.5%, 74.5%) | 34.4% (25.5%, 44.5%)  |
|   |  10,000-20,000 | 92 | 83 | 90.2% (82.4%, 94.8%) | 9.8% (5.2%, 17.6%)  |
|   |  20,000 + | 104 | 103 | 99.0% (94.8%, 99.8%) | 1.0% (0.2%, 5.2%)  |
|   |  Total | 289 | 247 | 85.5% (80.9%, 89.1%) | 14.5% (10.9%, 19.1%)  |

* The number of aberrations analyzed in each range bin, excluding indeterminate CNVs

{26}

Table 13. Accuracy based on CNV size (kb)
(Method 2, Scheme b: including Indeterminate CNVs as "Not Confirmed")

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 20-100 | 69 | 37 | 53.6% (42.0%, 64.9%) | 46.4% (35.1%, 58.0%)  |
|   |  100-200 | 94 | 75 | 79.8% (70.6%, 86.7%) | 20.2% (13.3%, 29.4%)  |
|   |  200-300 | 136 | 120 | 88.2% (81.7%, 92.6%) | 11.8% (7.4%, 18.3%)  |
|   |  300-500 | 123 | 92 | 74.8% (66.5%, 81.6%) | 25.2% (18.4%, 33.5%)  |
|   |  500-1,000 | 90 | 68 | 75.6% (65.8%, 83.3%) | 24.4% (16.7%, 34.2%)  |
|   |  1000-10,000 | 168 | 128 | 76.2% (69.2%, 82.0%) | 23.8% (18.0%, 30.8%)  |
|   |  10,000 + | 166 | 117 | 70.5% (63.1%, 76.9%) | 29.5% (23.1%, 36.9%)  |
|   |  Total | 846 | 637 | 75.3% (72.3%, 78.1%) | 24.7% (21.9%, 27.7%)  |
|  LOSS | 10-100 | 88 | 59 | 67.0% (56.7%, 76.0%) | 33.0% (24.0%, 43.3%)  |
|   |  100-200 | 207 | 184 | 88.9% (83.9%, 92.5%) | 11.1% (7.5%, 16.1%)  |
|   |  200-300 | 129 | 114 | 88.4% (81.7%, 92.8%) | 11.6% (7.2%, 18.3%)  |
|   |  300-500 | 116 | 101 | 87.1% (79.8%, 92.0%) | 12.9% (8.0%, 20.2%)  |
|   |  500-1,000 | 209 | 164 | 78.5% (72.4%, 83.5%) | 21.5% (16.5%, 27.6%)  |
|   |  1000-10,000 | 398 | 350 | 87.9% (84.4%, 90.8%) | 12.1% (9.2%, 15.6%)  |
|   |  10,000 + | 194 | 180 | 92.8% (88.3%, 95.7%) | 7.2% (4.3%, 11.7%)  |
|   |  Total | 1341 | 1152 | 85.9% (83.9%, 87.7%) | 14.1% (12.3%, 16.1%)  |
|  ALL CNVs |   | 2187 | 1789 | 81.8% (80.1%, 83.4%) | 18.2% (16.6%, 19.9%)  |
|  cnLOH | 5,000-10,000 | 93 | 61 | 65.6% (55.5%, 74.5%) | 34.4% (25.5%, 44.5%)  |
|   |  10,000-20,000 | 94 | 83 | 88.3% (80.2%, 93.3%) | 11.7% (6.7%, 19.8%)  |
|   |  20,000 + | 105 | 103 | 98.1% (93.3%, 99.5%) | 1.9% (0.5%, 6.7%)  |
|   |  Total | 292 | 247 | 84.6% (80.0%, 88.3%) | 15.4% (11.7%, 20.0%)  |

* The number of aberrations analyzed in each range bin, including indeterminate CNVs as "not confirmed".

{27}

Table 14. Accuracy based on number of probes (Method 3, Scheme a: Excluding Indeterminate CNVs)

|  TYPE | Aberration Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI)** | FPR*** (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 5-7 | 44 | 39 | 88.6% (76.0%, 95.0%) | 11.4% (5.0%, 24.0%)  |
|   |  7-10 | 97 | 91 | 93.8% (87.2%, 97.1%) | 6.2% (2.9%, 12.8%)  |
|   |  10-15 | 188 | 165 | 87.8% (82.3%, 91.7%) | 12.2% (8.3%, 17.7%)  |
|   |  15-20 | 91 | 83 | 91.2% (83.6%, 95.5%) | 8.8% (4.5%, 16.4%)  |
|   |  20-50 | 127 | 107 | 84.3% (76.9%, 89.6%) | 15.7% (10.4%, 23.1%)  |
|   |  50-500 | 78 | 70 | 89.7% (81.0%, 94.7%) | 10.3% (5.3%, 19.0%)  |
|   |  500 + | 169 | 166 | 98.2% (94.9%, 99.4%) | 1.8% (0.6%, 5.1%)  |
|   |  Total | 794 | 721 | 90.8% (88.6%, 92.6%) | 9.2% (7.4%, 11.4%)  |
|  LOSS | 5-7 | 212 | 211 | 99.5% (97.4%, 99.9%) | 0.5% (0.1%, 2.6%)  |
|   |  7-10 | 187 | 169 | 90.4% (85.3%, 93.8%) | 9.6% (6.2%, 14.7%)  |
|   |  10-15 | 244 | 229 | 93.9% (90.1%, 96.2%) | 6.1% (3.8%, 9.9%)  |
|   |  15-20 | 107 | 95 | 88.8% (81.4%, 93.5%) | 11.2% (6.5%, 18.6%)  |
|   |  20-50 | 122 | 103 | 84.4% (77.0%, 89.8%) | 15.6% (10.2%, 23.0%)  |
|   |  50-500 | 225 | 220 | 97.8% (94.9%, 99.0%) | 2.2% (1.0%, 5.1%)  |
|   |  500 + | 184 | 180 | 97.8% (94.5%, 99.2%) | 2.2% (0.8%, 5.5%)  |
|   |  Total | 1281 | 1207 | 94.2% (92.8%, 95.4%) | 5.8% (4.6%, 7.2%)  |
|  All CNVs |   | 2075 | 1928 | 92.9% (91.7%, 93.9%) | 7.1% (6.1%, 8.3%)  |
|  cnLOH | 100-200 | 132 | 106 | 80.3% (72.7%, 86.2%) | 19.7% (13.8%, 27.3%)  |
|   |  200-500 | 102 | 99 | 97.1% (91.7%, 99.0%) | 2.9% (1.0%, 8.3%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 292 | 263 | 90.1% (86.1%, 93.0%) | 9.9% (7.0%, 13.9%)  |

* The number of aberrations analyzed in each range bin, excluding indeterminate CNVs
** Confirmation Rate = TP/(TP+FP), equivalent to "# Confirmed / Sample Size (N)". It can also be referred to as "% Agreement" or "Positive Predictive Value (PPV)". 95% CI calculated using the Wilson score method. Applicable to other tables in this report.
*** FPR (False Positive Rate) = Pr (Aberration "Not Confirmed" | Aberration detected by the GenetiSure Dx Postnatal Assay) in this context is "1-Agreement (Confirmation Rate)" rather than the conventional concept of "1-specificity". Applicable to other tables in this report.

{28}

Table 15. Accuracy based on number of probes w (Method 3, Scheme b: including Indeterminate CNVs)

|  TYPE | Aberratio n Range (# of Probes) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 5-7 | 48 | 39 | 81.3% (68.1%, 89.8%) | 18.7% (10.2%, 31.9%)  |
|   |  7-10 | 101 | 91 | 90.1% (82.7%, 94.5%) | 9.9% (5.5%, 17.3%)  |
|   |  10-15 | 197 | 165 | 83.8% (78.0%, 88.3%) | 16.2% (11.7%, 22.0%)  |
|   |  15-20 | 101 | 83 | 82.2% (73.6%, 88.4%) | 17.8% (11.6%, 26.4%)  |
|   |  20-50 | 148 | 107 | 72.3% (64.6%, 78.9%) | 27.7% (21.1%, 35.4%)  |
|   |  50-500 | 82 | 70 | 85.4% (76.1%, 91.4%) | 14.6% (8.6%, 23.9%)  |
|   |  500 + | 169 | 166 | 98.2% (94.9%, 99.4%) | 1.8% (0.6%, 5.1%)  |
|   |  Total | 846 | 721 | 85.2% (82.7%, 87.5%) | 14.8% (12.5%, 17.3%)  |
|  LOSS | 5-7 | 216 | 211 | 97.7% (94.7%, 99.0%) | 2.3% (1.0%, 5.3%)  |
|   |  7-10 | 202 | 169 | 83.7% (77.9%, 88.1%) | 16.3% (11.9%, 22.1%)  |
|   |  10-15 | 257 | 229 | 89.1% (84.7%, 92.4%) | 10.9% (7.6%, 15.3%)  |
|   |  15-20 | 125 | 95 | 76.0% (67.8%, 82.6%) | 24.0% (17.4%, 32.2%)  |
|   |  20-50 | 130 | 103 | 79.2% (71.5%, 85.3%) | 20.8% (14.7%, 28.5%)  |
|   |  50-500 | 225 | 220 | 97.8% (94.9%, 99.0%) | 2.2% (1.0%, 5.1%)  |
|   |  500 + | 186 | 180 | 96.8% (93.1%, 98.5%) | 3.2% (1.5%, 6.9%)  |
|   |  Total | 1341 | 1207 | 90.0% (88.3%, 91.5%) | 10.0% (8.5%, 11.7%)  |
|  All CNVs |   | 2187 | 1928 | 88.2% (86.7%, 89.4%) | 11.8% (10.6%, 13.3%)  |
|  cnLOH | 100-200 | 132 | 106 | 80.3% (72.7%, 86.2%) | 19.7% (13.8%, 27.3%)  |
|   |  200-500 | 102 | 99 | 97.1% (91.7%, 99.0%) | 2.9% (1.0%, 8.3%)  |
|   |  500 + | 58 | 58 | 100.0% (93.8%, 100.0%) | 0.0% (0.0%, 6.2%)  |
|   |  Total | 292 | 263 | 90.1% (86.1%, 93.0%) | 9.9% (7.0%, 13.9%)  |

* The number of aberrations analyzed in each range bin, including indeterminate CNVs as "not confirmed".

{29}

Table 16. Accuracy based on CNV size (kb)
(Method 3, Scheme a: Excluding Indeterminate CNVs)

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 20-100 | 57 | 41 | 71.9% (59.2%, 81.9%) | 28.1% (18.1%, 40.8%)  |
|   |  100-200 | 89 | 77 | 86.5% (77.9%, 92.1%) | 13.5% (7.9%, 22.1%)  |
|   |  200-300 | 131 | 128 | 97.7% (93.5%, 99.2%) | 2.3% (0.8%, 6.5%)  |
|   |  300-500 | 112 | 97 | 86.6% (79.1%, 91.7%) | 13.4% (8.3%, 20.9%)  |
|   |  500-1,000 | 84 | 77 | 91.7% (83.8%, 95.9%) | 8.3% (4.1%, 16.2%)  |
|   |  1000-10,000 | 155 | 138 | 89.0% (83.1%, 93.0%) | 11.0% (7.0%, 16.9%)  |
|   |  10,000 + | 166 | 163 | 98.2% (94.8%, 99.4%) | 1.8% (0.6%, 5.2%)  |
|   |  Total | 794 | 721 | 90.8% (88.6%, 92.6%) | 9.2% (7.4%, 11.4%)  |
|  LOSS | 10-100 | 82 | 65 | 79.3% (69.3%, 86.6%) | 20.7% (13.4%, 30.7%)  |
|   |  100-200 | 204 | 197 | 96.6% (93.1%, 98.3%) | 3.4% (1.7%, 6.9%)  |
|   |  200-300 | 121 | 120 | 99.2% (95.5%, 99.9%) | 0.8% (0.1%, 4.5%)  |
|   |  300-500 | 112 | 110 | 98.2% (93.7%, 99.5%) | 1.8% (0.5%, 6.3%)  |
|   |  500-1,000 | 185 | 170 | 91.9% (87.1%, 95.0%) | 8.1% (5.0%, 12.9%)  |
|   |  1000-10,000 | 385 | 357 | 92.7% (89.7%, 94.9%) | 7.3% (5.1%, 10.3%)  |
|   |  10,000 + | 192 | 188 | 97.9% (94.8%, 99.2%) | 2.1% (0.8%, 5.2%)  |
|   |  Total | 1281 | 1207 | 94.2% (92.8%, 95.4%) | 5.8% (4.6%, 7.2%)  |
|  All CNVs |   | 2075 | 1928 | 92.9% (91.7%, 93.9%) | 7.1% (6.1%, 8.3%)  |
|  cnLOH | 5,000-10,000 | 93 | 70 | 75.3% (65.6%, 82.9%) | 24.7% (17.1%, 34.4%)  |
|   |  10,000-20,000 | 94 | 89 | 94.7% (88.1%, 97.7%) | 5.3% (2.3%, 11.9%)  |
|   |  20,000 + | 105 | 104 | 99.0% (94.8%, 99.8%) | 1.0% (0.2%, 5.2%)  |
|   |  Total | 292 | 263 | 90.1% (86.1%, 93.0%) | 9.9% (7.0%, 13.9%)  |

* The number of aberrations analyzed in each range bin, excluding indeterminate CNVs

{30}

Table 17. Accuracy base don CNV size (kb)
(Method 3, Scheme b: including Indeterminate CNVs as "Not Confirmed")

|  TYPE | Aberration Range (kb) | Sample Size (N)* | # Confirmed | Confirmation Rate (95% CI) | FPR (95% CI)  |
| --- | --- | --- | --- | --- | --- |
|  GAIN | 20-100 | 69 | 41 | 59.4% (47.6%, 70.2%) | 40.6% (29.8%, 52.4%)  |
|   |  100-200 | 94 | 77 | 81.9% (72.9%, 88.4%) | 18.1% (11.6%, 27.1%)  |
|   |  200-300 | 136 | 128 | 94.1% (88.8%, 97.0%) | 5.9% (3.0%, 11.2%)  |
|   |  300-500 | 123 | 97 | 78.9% (70.8%, 85.1%) | 21.1% (14.9%, 29.2%)  |
|   |  500-1,000 | 90 | 77 | 85.6% (76.8%, 91.4%) | 14.4% (8.6%, 23.2%)  |
|   |  1000-10,000 | 168 | 138 | 82.1% (75.7%, 87.2%) | 17.9% (12.8%, 24.3%)  |
|   |  10,000 + | 166 | 163 | 98.2% (94.8%, 99.4%) | 1.8% (0.6%, 5.2%)  |
|   |  Total | 846 | 721 | 85.2% (82.7%, 87.5%) | 14.8% (12.5%, 17.3%)  |
|  LOSS | 10-100 | 88 | 65 | 73.9% (63.8%, 81.9%) | 26.1% (18.1%, 36.2%)  |
|   |  100-200 | 207 | 197 | 95.2% (91.3%, 97.4%) | 4.8% (2.6%, 8.7%)  |
|   |  200-300 | 129 | 120 | 93.0% (87.3%, 96.3%) | 7.0% (3.7%, 12.7%)  |
|   |  300-500 | 116 | 110 | 94.8% (89.2%, 97.6%) | 5.2% (2.4%, 10.8%)  |
|   |  500-1,000 | 209 | 170 | 81.3% (75.5%, 86.0%) | 18.7% (14.0%, 24.5%)  |
|   |  1000-10,000 | 398 | 357 | 89.7% (86.3%, 92.3%) | 10.3% (7.7%, 13.7%)  |
|   |  10,000 + | 194 | 188 | 96.9% (93.4%, 98.6%) | 3.1% (1.4%, 6.6%)  |
|   |  Total | 1341 | 1207 | 90.0% (88.3%, 91.5%) | 10.0% (8.5%, 11.7%)  |
|  All CNVs |   | 2187 | 1928 | 88.2% (86.7%, 89.4%) | 11.8% (10.6%, 13.3%)  |
|  cnLOH | 5,000-10,000 | 93 | 70 | 75.3% (65.6%, 82.9%) | 24.7% (17.1%, 34.4%)  |
|   |  10,000-20,000 | 94 | 89 | 94.7% (88.1%, 97.7%) | 5.3% (2.3%, 11.9%)  |
|   |  20,000 + | 105 | 104 | 99.0% (94.8%, 99.8%) | 1.0% (0.2%, 5.2%)  |
|   |  Total | 292 | 263 | 90.1% (86.1%, 93.0%) | 9.9% (7.0%, 13.9%)  |

* The number of aberrations analyzed in each range bin, including indeterminate CNVs as "not confirmed".

{31}

The endpoint agreements were also analyzed. Agreements using Method 1 are shown in Table 18-.19. All three methods were found to have similar results (data not shown).

Table 18. Endpoint Agreement for Method 1 Based: comparator platform 1 Only; Binning by Number of Probes; Start/Stop Breakpoints Combined

|  TYPE | Aberration Range (# of Probes) | Breakpoints N | Breakpoint Agreement N | Breakpoint Agreement % (95% CI)  |
| --- | --- | --- | --- | --- |
|  GAIN | 5-7 | 48 | 47 | 97.9% (89.1%, 99.6%)  |
|   |  7-10 | 126 | 116 | 92.1% (86.0%, 95.6%)  |
|   |  10-15 | 258 | 223 | 86.4% (81.7%, 90.1%)  |
|   |  15-20 | 110 | 88 | 80.0% (71.6%, 86.4%)  |
|   |  20-50 | 128 | 115 | 89.8% (83.4%, 94.0%)  |
|   |  50-500 | 124 | 104 | 83.9% (76.4%, 89.3%)  |
|   |  500 + | 296 | 258 | 87.2% (82.9%, 90.5%)  |
|   |  Total | 1090 | 951 | 87.2% (85.1%, 89.1%)  |
|  LOSS | 5-7 | 236 | 234 | 99.2% (97.0%, 99.8%)  |
|   |  7-10 | 124 | 117 | 94.4% (88.8%, 97.2%)  |
|   |  10-15 | 116 | 89 | 76.7% (68.3%, 83.5%)  |
|   |  15-20 | 84 | 69 | 82.1% (72.6%, 88.9%)  |
|   |  20-50 | 138 | 110 | 79.7% (72.2%, 85.6%)  |
|   |  50-500 | 420 | 382 | 91.0% (87.8%, 93.3%)  |
|   |  500 + | 350 | 283 | 80.9% (76.4%, 84.6%)  |
|   |  Total | 1468 | 1284 | 87.5% (85.7%, 89.1%)  |
|  All CNVs |   | 2558 | 2235 | 87.4% (86.0%, 88.6%)  |
|  cnLOH | 100-200 | 188 | 176 | 93.6% (89.2%, 96.3%)  |
|   |  200-500 | 194 | 176 | 90.7% (85.8%, 94.1%)  |
|   |  500 + | 116 | 104 | 89.7% (82.8%, 94.0%)  |
|   |  Total | 498 | 456 | 91.6% (88.8%, 93.7%)  |

{32}

Table 19. Endpoint Agreement for Method 1 Based: comparator platform 2 Only; Binning by Number of Probes; Start/Stop Breakpoints Combined

|  TYPE | Aberration Range (# of Probes) | Breakpoints N | Breakpoint Agreement N | Breakpoint Agreement % (95% CI)  |
| --- | --- | --- | --- | --- |
|  GAIN | 5-7 | 38 | 37 | 97.4% (86.5%, 99.5%)  |
|   |  7-10 | 124 | 120 | 96.8% (92.0%, 98.7%)  |
|   |  10-15 | 242 | 204 | 84.3% (79.2%, 88.3%)  |
|   |  15-20 | 58 | 49 | 84.5% (73.1%, 91.6%)  |
|   |  20-50 | 156 | 140 | 89.7% (84.0%, 93.6%)  |
|   |  50-500 | 140 | 119 | 85.0% (78.2%, 90.0%)  |
|   |  500 + | 306 | 261 | 85.3% (80.9%, 88.8%)  |
|   |  Total | 1064 | 930 | 87.4% (85.3%, 89.3%)  |
|  LOSS | 5-7 | 254 | 247 | 97.2% (94.4%, 98.7%)  |
|   |  7-10 | 124 | 121 | 97.6% (93.1%, 99.2%)  |
|   |  10-15 | 138 | 117 | 84.8% (77.9%, 89.8%)  |
|   |  15-20 | 72 | 63 | 87.5% (77.9%, 93.3%)  |
|   |  20-50 | 160 | 153 | 95.6% (91.2%, 97.9%)  |
|   |  50-500 | 426 | 398 | 93.4% (90.7%, 95.4%)  |
|   |  500 + | 352 | 324 | 92.0% (88.7%, 94.4%)  |
|   |  Total | 1526 | 1423 | 93.3% (91.9%, 94.4%)  |
|  All CNVs |   | 2590 | 2353 | 90.8% (89.7%, 91.9%)  |
|  cnLOH | 100-200 | 188 | 174 | 92.6% (87.9%, 95.5%)  |
|   |  200-500 | 196 | 176 | 89.8% (84.8%, 93.3%)  |
|   |  500 + | 116 | 104 | 89.7% (82.8%, 94.0%)  |
|   |  Total | 500 | 454 | 90.8% (87.9%, 93.0%)  |

b. Matrix comparison:

Not applicable. The device is for use with EDTA anticoagulated peripheral blood only.

# 3. Clinical Studies:

a. Clinical Sensitivity:

A retrospective clinical study was performed to characterize the clinical performance characteristics of GenetiSure Dx Postnatal Assay for the purpose of reporting the pathogenic detection rate (potential diagnostic yield) of the assay. A total of 800 gDNA samples from patients suspected of having pathogenic aberrations (SPA

{33}

samples) were collected from 3 regionally distinct clinical institutions that offer postnatal array testing for the detection of chromosomal abnormalities. One hundred (100) samples from phenotypically normal individuals were also processed using the GenetiSure Dx Postnatal Assay and were used to assess the aberrations that might be expected to be found in a normal (non-patient) population. The aberrations detected in each sample, for all nine hundred (900) samples, were interpreted by one of four cytogeneticists as Benign, Likely Benign, Variant Of Unknown Significance (VOUS), Likely Pathogenic, or Pathogenic.

The interpretations of the calls were in agreement if the cytogeneticists at both the CNC Test processing site and at the collection site determined an aberration to be of the same pathogenicity.

The test results, per sample, were compared to historical array data from the respective collection site, which were generated using the methods established at each laboratory. All reported Pathogenic and Likely Pathogenic copy number variants (CNVs), gains and losses, were subject to confirmation by alternative methods. Confirmation methods used at the collection sites were selected by the sites based on the assays availability as well as the nature of the aberration being confirmed. The methods used included one or more of the following:

a. G-banded karyotyping
b. Fluorescence in situ hybridization (FISH)
c. Multiplex Ligation-dependent Probe Amplification (MLPA)
d. Quantitative Polymerase Chain Reaction (qPCR)
e. Non-Agilent Comparative Genomic Hybridization oligonucleotide microarrays (molecular karyotyping)

The diagnostic yield based on the subset evaluated with the GenetiSure Dx Postnatal Assay, when considering only copy number aberrations, was 15%. This increased to 20% when cnLOH aberrations were also considered.

Table 20. Diagnostic Yield by Collection Site (95% CI)

|  Collection Site | Number of Samples | Collection Site: Number of Pathogenic Calls | Collection Site: Diagnostic Yield | GenetiSure Dx Postnatal Assay: Number of Pathogenic Calls | GenetiSure Dx Postnatal Assay: Diagnostic Yield  |
| --- | --- | --- | --- | --- | --- |
|  Copy Number Aberrations Only  |   |   |   |   |   |
|  Site 1 | 257 | 29 | 11% (8.0%, 15.7%) | 39 | 15% (11.3%, 20.1%)  |
|  Site 2 | 313 | 35 | 11% (8.2%, 15.2%) | 33 | 11% (7.6%, 14.4%)  |
|  Site 3 | 230 | 48 | 21% (16.1%, 26.6%) | 45 | 20% (15.0%, 25.2%)  |
|  TOTAL | 800 | 112 | 14% (11.8%, 16.6%) | 117 | 15% (12.3%, 17.2%)  |

{34}

|  All Aberrations (Copy Number and cnLOH)  |   |   |   |   |   |
| --- | --- | --- | --- | --- | --- |
|  Site 1 | 257 | 29 | 11% (8.0%, 15.7%) | 48 | 19% (14.4%, 23.9%)  |
|  Site 2 | 313 | 39 | 12% (9.2%, 16.6%) | 60 | 19% (15.2%, 23.9%)  |
|  Site 3 | 230 | 48 | 21% (16.1%, 26.6%) | 51 | 22% (17.3%, 28.0%)  |
|  TOTAL | 800 | 116 | 15% (12.2%, 17.1%) | 159 | 20% (17.3, 22.8%)  |

Results of the PPA and NPA analysis are presented considering only copy number aberrations (Table 21) or considering both copy number and cnLOH aberrations (Table 22). For the copy number aberration only analysis, samples with pathogenic cnLOH aberrations were considered as non-pathogenic, unless they also included a pathogenic copy number aberration.

Table 11. Comparison of Sample Classification between the Collection Site and the GenetiSure Dx Postnatal Assay, Considering only Copy Number Aberrations

|   | Collection Site Aberration Interpretation |   |   |   |   |   |   |
| --- | --- | --- | --- | --- | --- | --- | --- |
|   |   |  Pathogenic Interpretation |   | Non-Pathogenic Interpretation  |   |   |   |
|  GenetiSure Dx Postnatal Assay Interpretation |   | Pathogenic | Likely Pathogenic | VOUS | Likely Benign1 | Normal2 | Total  |
|  Pathogenic Interpretation | Pathogenic | 56 | 14 | 9 | 0 | 3 | 82  |
|   |  Likely Pathogenic | 12 | 4 | 11 | 0 | 8 | 35  |
|  Non-Pathogenic Interpretation | VOUS | 5 | 8 | 35 | 0 | 32 | 80  |
|   |  Normal2 | 6 | 7 | 80 | 1 | 509 | 603  |
|  Total |   | 79 | 33 | 135 | 1 | 552 | 800  |
|  PPA3 |   | 86/112 = 76.8% (95%CI=68.2% - 83.6%)  |   |   |   |   |   |
|  NPA4 |   | 657/688 = 95.5% (95%CI=93.7% - 96.8%)  |   |   |   |   |   |

One Site 2 sample was presented with the interpretation on Likely Benign.
Samples from the GenetiSure Dx Postnatal Assay or Site 1 with either only Benign or Likely Benign aberrations, or samples without aberrations are classified as "Normal". Site 3 and Site 2 provided sample classification of "Normal".
Positive Percent Agreement (PPA): Percent [(Agilent GenetiSure Dx Postnatal Assay = Pathogenic & Collection site classification = Pathogenic)/(Collection site classification = Pathogenic)]
Negative Percent Agreement (NPA): Percent [(Agilent GenetiSure Dx Postnatal Assay = Non-pathogenic & Collection site classification = Non-Pathogenic)/(Collection site classification = Non-Pathogenic)]

When considering only copy number aberrations in the sample classification, PPA was  $76.8\%$  and NPA was  $95.5\%$ . In total, 26 samples which were determined to have Pathogenic or Likely Pathogenic copy number aberrations by the collection sites were reported as non-pathogenic by the GenetiSure Dx Postnatal Assay. Most of these

{35}

aberrations were either detected by GenetiSure Dx Postnatal Assay, but interpreted differently by the cytogeneticist, or below the detection limit of the Assay.

Table 22. Comparison of Sample Classification between the Collection Site and the GenetiSure Dx Postnatal Assay, Considering Copy Number and cnLOH Aberrations

|   | Collection Site Aberration Interpretation |   |   |   |   |   |   |
| --- | --- | --- | --- | --- | --- | --- | --- |
|   |   |  Pathogenic Interpretation |   | Non-Pathogenic Interpretation  |   |   |   |
|  GenetiSure Dx Postnatal Assay Interpretation |   | Pathogenic | Likely Pathogenic | VOUS | Likely Benign1 | Normal2 | Total  |
|  Pathogenic Interpretation | Pathogenic | 56 | 14 | 9 | 0 | 3 | 82  |
|   |  Likely Pathogenic | 14 | 5 | 23 | 0 | 35 | 77  |
|  Non-Pathogenic Interpretation | VOUS | 5 | 10 | 59 | 0 | 46 | 120  |
|   |  Normal2 | 7 | 5 | 74 | 1 | 434 | 521  |
|  Total |   | 82 | 34 | 165 | 1 | 518 | 800  |
|  PPA3 |   | 89/116 = 76.7% (95%CI=68.3%-83.5%)  |   |   |   |   |   |
|  NPA4 |   | 614/684 = 89.8% (95%CI=87.3%-91.8%)  |   |   |   |   |   |

One Site 2 sample was presented with the interpretation on Likely Benign.
Samples from the GenetiSure Dx Postnatal Assay or Site 1 with either only Benign or Likely Benign aberrations, or samples without aberrations are classified as "Normal". Site 3 and Site 2 provided sample classification of "Normal".
Positive Percent Agreement (PPA): Percent [(Agilent GenetiSure Dx Postnatal Assay = Pathogenic & Collection site classification = Pathogenic)/(Collection site classification = Pathogenic)]
Negative Percent Agreement (NPA): Percent [(Agilent GenetiSure Dx Postnatal Assay = Non-pathogenic & Collection site classification = Non-Pathogenic)/(Collection site classification = Non-Pathogenic)]

When consid…

---

**Source:** [https://fda.innolitics.com/submissions/MG/subpart-f%E2%80%94immunological-test-systems/PFX/K163367](https://fda.innolitics.com/submissions/MG/subpart-f%E2%80%94immunological-test-systems/PFX/K163367)

**Published by [Innolitics](https://innolitics.com)** — a medical-device software consultancy. We help companies design, build, and clear FDA-regulated software and AI/ML devices. If you're preparing [a 510(k)](https://innolitics.com/services/510ks/), [a De Novo](https://innolitics.com/services/regulatory/), [a SaMD](https://innolitics.com/services/end-to-end-samd/), [an AI/ML medical device](https://innolitics.com/services/medical-imaging-ai-development/), or [an FDA regulatory strategy](https://innolitics.com/services/regulatory/), [get in touch](https://innolitics.com/contact).

**Cite:** Innolitics at https://innolitics.com