oncomine-dx-target-test-technical-validation-information-270

Established performance

The Ion Torrent Oncomine Dx Target Test is the first targeted next-generation sequencing (NGS) in vitro diagnostic test simultaneously delivering multiple biomarker results to aid selection of targeted therapies for NSCLC and CC patients. Concordance with FDA approved or validated reference methods based on FISH, PCR, Sanger sequencing, or NGS was established for EGFR, BRAF, ROS1, and RET inNSCLC and for IDH1 inCC.The variants for KRAS, MET, and PIK3CA were analytically validated in NSCLC. The safety and effectiveness of these three genes have not been established and they are not intended to be used to direct therapy. The performance of all other variants identified by the test, other than clinically validated therapeutic variants and analytically validated variants, was validated based on a representative method.

Request more info

Oncomine Dx Target Test content

Gene targets included for NSCLC

Gene targets for therapeutic use
BRAF: mutationEGFR: L858R, exon 19 deletions, and exon 20 insertionsROS1: fusionsRET: fusions
Analytically validated targets
KRASMET*PIK3CA
Additional targets**
AKT1ERBB2HRASMTORRET
ALK*ERBB3KITNRASROS1
CDK4FGFR2MAP2K1PDGFRA 
DDR2FGFR3MAP2K2RAF1 

Figure 1. Complete gene list. *The test reports fusion/translocation variants for ROS1 and RET only. The test only reports ALK and MET mutations. ** Performance for the additional gene target variants has been validated based on a representative method. Only IDH1 is reported for CC.


Clinical concordance for companion diagnostics markers for targeted therapies selection - NSCLC

Method comparison studies evaluated the accuracy of the Oncomine Dx Target Test for the detection of BRAF V600E, EGFR exon 19 deletions, L858R, and exon 20 insertions, ROS1 fusions, and RET fusions, using a BRAF V600E PCR assay, therascreen™ EGFR PCR Kit, ROS1 FISH assay, and validated NGS assay respectively. A summary of the concordance studies results are included in Table 1. For details see the User Manual.

 

Variants for therapy selection

 

 

Validated comparator methods

 

Excluding no calls or unknowns*Including no calls or unknowns*

Positive percent agreement

Negative percent agreement

Overall percent agreement

Positive percent agreement

Negative percent agreement

Overall percent agreement

 

BRAF V600E

Validated BRAF
V600E qPCR
test

100% (67/67)

100% (114/114)

100% (181/181)

91.8% (67/73)

97.4% (114/117)

95.3% (181/190)

EGFR

Therascreen™ EGFR PCR Kit

98.6% (71/72)

99.2% (120/121)

99.0% (191/193)

81.6% (71/87)

96.8% (120/124)

90.5% (191/211)

EGFR exon 19 deletions

97.6% (41/42)

99.3% (147/148)

99.0% (188/190)

74.6% (41/55)

94.2% (147/156)

89.1% (188/211)

EGFR exon 21 L858R

100% (30/30)

100% (167/167)

100% (197/197)

93.8% (30/32)

93.3% (167/179)

93.4% (197/211)

EGFR exon 20 insertionsValidated NGS assay100% (54/54)100% (95/95)100% (149/149)98.2% (54/55)90.5% (95/105)93.1% (149/160)

ROS1 fusions

Validated ROS1 FISH test

100% (9/9)

100%
(62/62)

100%
(71/71)

90.0% (9/10)

88.6% (62/70)

88.8% (71/80)

RET fusionsValidated NGS90.9% (40/44)91.8% (101/110)91.6% (141/154)90.9% (40/44)91.8% (101/110)91.6% (141/154)

* No-calls are for DNA variants and unknowns are for RNA fusions

Table 1. Method comparison between Oncomine Dx Target Test and reference methods for three companion diagnostic biomarkers in NSCLC.


Analytical validation performance - NSCLC

The Oncomine Dx Target Test also detects DNA sequence variations in an additional 19 genes in NSCLC. The variants for KRAS, MET, and PIK3CA have been analytically validated. Safety and effectiveness of these three genes have not been established and they are not intended to be used to direct therapy. The performance of all other variants identified by the test, other than clinically validated therapeutic variants and analytically validated variants, has not been directly demonstrated and was validated based on a representative method.

Limit of detection

Four LoD studies were performed to evaluate DNA variants, ROS1 fusions, and EGFR exon 20 insertions.

Study I: The limit of detection (LoD) was evaluated for 14 representative DNA variants representing 3 variant categories detected by the Oncomine Dx Target Test. The LoD is the lowest allele frequency of SNV, multi-nucleotide polymorphism (MNP), or deletion variants, that can be detected at least 95% of the time. The study demonstrated that the Oncomine Dx Target Test can detect DNA variants with allele frequencies between 6 and 8%.

Study II: The LoD was calculated for two clinical ROS1 RNA fusion variants using the updated RNA library preparation workflow, and determined at 516 fusion reads.

Study III: The LoD was calculated for two clinical RET fusion variants using the updated RNA library preparation workflow, and determined at 405 fusion reads.

Study IV: The LoD was calculated for two clinical EGFR exon 20 insertion-positive samples and determined to be 4.8–5.2% allele frequencies.

Accuracy

To evaluate the ability of the Oncomine Dx Target Test DNA and RNA panels to identify somatic variants in human specimens, 290 FFPE tumor samples were analyzed using the Oncomine Dx Target Test to demonstrate positive percent agreement (PPA) and negative percent agreement (NPA) concordance with validated reference detection methods.

The following reference detection methods were used:

  • Validated NGS method, to detect SNV and deletion hotspot variants
  • Validated ROS1 FISH test, to detect ROS1 fusions

The study demonstrated variant level PPA of 98.5%, NPA of 100%, and OPA of 100%, excluding invalids no-calls; and PPA level of 98.5%, NPA of 96.8%, and OPA of 96.8% including no-calls. A summary of the data are included in Table 2. For details see the User Manual.

Variant level measure of agreement

Percent agreement (N)
excluding no-calls

Percent agreement (N)
including no-calls

Positive percent agreement

98.5% (195/198)

98.5% (195/198)

Negative percent agreement

100.0% (118,155/118,159)

96.8% (118,155/122,012)

Overall percent agreement

100.0% (118,350/118,357)

96.8% (118,350/122,210)

Table 2. Variant level accuracy study results

Reproducibility

Four reproducibility studies were performed to evaluate DNA variants, ROS1 fusions, EGFR fusions, and EGFR exon 20 insertions.

Study I:

The reproducibility and repeatability of the Oncomine Dx Target Test was evaluated for 30 representative variants from 18 DNA samples. The study was designed to evaluate within-run precision performance (repeatability) and variability across sites, operators, and instruments (reproducibility). Due to the large number of variants detected by the test and the rarity of some of the variants, a representative variant approach was used. Variants were selected in the following categories:

  • Simple SNVs
  • Complex SNVs and MNPs, including SNVs in di- or tri-nucleotide repeat regions and SNVs in high-GC (>60%) or low-GC (<40%) content regions
  • Deletions (including deletions of 6, 9, 15, and 18 bp)

Excluding no calls, the percent of correct calls is >96%. The estimate of repeatability at each DNA variant location across all the samples was ≥98.8% (95% CI lower limit of ≥97.5%). A summary of results of the assay reproducibility study are included in Table 3. For details see the User Manual.

 

Description

No. of variants

Call rate excluding no-calls

Call rate including no-calls

Mean

Median

Mean

Median

DNA positive variants (positive calls)

46

96.60%

97.10%

94.50%

95.80%

WT DNA variant locations (negative calls)

872

96.10%

95.00%

96.10%

95.00%

Table 3. Assay reproducibility study I.

Study II: 

An additional study was performed to evaluate the reproducibility and repeatability of the Oncomine Dx Target Test for 6 representative variants from 11 DNA samples and 4 RNA samples. 1 WT DNA sample and 4 WT RNA samples were included in the study.

The study was designed to evaluate within-run precision performance (repeatability) and variability across sites, operators, and instrument platforms (reproducibility). The updated RNA library preparation workflow was used. Due to the large number of variants detected by the test and the rarity of some variants, a representative variant approach was used.

 Variants were selected in the following categories:

  • 15 bp deletion
  • Simple SNVs
  • Complex SNVs and MNPs
  • Fusions

Excluding no calls, the estimate of repeatability at each DNA variant location across all the samples was ≥94.4% (95% CI lower limit of ≥72.7%). The estimate of repeatability at each RNA clinical variant location was 100%. A summary of the results of Study II is included in Table 4 and 5.

 

Description

No. of variants

Call rate excluding no-calls 

Call rate including no-calls

Mean

Median

Mean

Median

DNA positive variants (positive calls) 

11

99%

100%

98%

99%

WT DNA variant locations (negative calls)

367

100%

100%

99%

100%

Table 4. Study II—assay reproducibility study results (DNA variants)

 

Description

No. of variants

Call rate including or excluding unknowns 

Mean

Median

ROS1 positive variants (positive calls)

4

100%

100%

WT RNA variant locations (negative calls)

4

99%

100%

Table 5. Study II—assay reproducibility study results (ROS1 fusions)

Study III:

An additional study was performed to evaluate the reproducibility and repeatability of the Oncomine Dx Target Test for 4 RET fusion positive samples and 2 RET fusion-negative samples. The study was designed to evaluate within-run precision performance (repeatability) and variability across sites, operators, and instrument platforms (reproducibility). The updated RNA library preparation workflow was used. Excluding unknowns, estimates of the repeatability ranged from 98.1% to 100% for two RET variants. A summary of the results of Study III is included in Table 6.

 

Description

No. of variants

Call rate including or excluding unknowns 

Mean

Median

RET positive variants (positive calls)

4

99%

100%

WT RNA variant locations (negative calls)

2

100%

100%

Table 6. Study III—assay reproducibility study results (RET fusions)

Study IV:

A study was performed to evaluate the reproducibility and repeatability of the Oncomine Dx Target Test for detection of EGFR exon 20 insertion variants using FFPE DNA from two EGFR variant-positive samples (blended with WT clinical samples) and two EGFR variant-negative (WT) samples. The study was designed to evaluate within-run precision performance (repeatability) and variability across sites, operators, and instrument platforms (reproducibility). Excluding no-calls, estimates of the repeatability is 100% for both EGFR exon 20 insertion variants. A summary of the results of Study IV is included in Table 7.

 

Description

No. of variants

Call rate excluding no-calls

Call rate including no-calls

Mean

Median

Mean

Median

EGFR insertion positive variants (positive calls)

2

100%

100%

100%

100%

WT DNA variants (negative calls)

2

100%

100%

100%

100%

Table 7. Study IV - assay reproducibility results (EGFR exon 20 insertions)


Clinical concordance for companion diagnostics markers for targeted therapies selection - CC

A clinical concordance study was conducted to evaluate the ability of the Oncomine Dx Target Test to identify five IDH1 biomarkers in FFPE cholangiocarcinoma tumor specimen compared to a validated Sanger assay. The study demonstrated OPA of 97.9%, excluding invalids and no calls. A summary of the data is included in Table 8.

 

Variants for therapy selection

 

 

Validated comparator methods

 

Excluding invalid and no callsIncluding invalid results and no-calls

Positive percent agreement

Negative percent agreement

Overall percent agreement

Positive percent agreement

Negative percent agreement

Overall percent agreement

 

BIDH1 R132

Validated Sanger assay

99.4% (163/164)

96.5% (164/170)

97.9% (327/334)

97.0% (163/168)

90.6% (164/181)

93.7% (327/349)

Table 8. Concordance between the Oncomine Dx Target Test and reference method for IDH1 R132


Analytical validation performance - CC

Limit of detection

The limit of detection (LoD) was evaluated for fove IDH1 R132 variants detected by the Oncomine Dx Target Test. The LoD is the lowest allele frequency of SNV that can be detected at least 95% of the time. The study demonstrated LoD of the five IDH1 R132 variants ranged from 4.5–5.7% allele frequencies, including 4.5% for R132C, 5.7% for R132G, 4.9% for R132H, 5.1% for R132L, and 5.3% for R132S.

Assay reproducibility

The reproducibility and repeatability of IDH1 R132 variant detection using Oncomine Dx Target Test were assessed with one IDH1 WT sample and three IDH1 R132 variant-positive samples at two allele frequency (AF) levels. Testing was performed at four testing sites, each site had two PGM Dx instrument systems, two operators, and using four lots of reagents. The overall positive call rate for IDH1 R132 variants was 92.6% when including no calls and 97.1% when excluding no calls. The negative call rate for IDH1 WT sample were 100% at all IDH1 R132 variant locations. See Table 9.

 

Sample COSMIC ID, variant

No. of valid sample results

Call rate  (95% CI)

Including no calls

Excluding no calls

D1
COSM28747, R132C

36

100%
(90.3%, 100%)
100%
(90.3%, 100%)
D2
COSM28747, R132C

36

97.2%
(85.5%, 99.9%)
100%
(90.0%, 100%)
D3
COSM28749, R132G
36100%
(90.3%, 100%)
100%
(90.3%, 100%)
D4
COSM28749, R132G
36100%
(90.3%, 100%)
100%
(90.3%, 100%)
D5
COSM28750, R132L
36100%
(90.3%, 100%)
100%
(90.3%, 100%)
D6
COSM28750, R132L 
3557.1%
(39.4%, 73.7%)
76.9%
(56.4%, 91.0%)
D1-D6
All Variants, R132 
21592.6%
(88.2%, 95.7%)
97.1%
(93.7%, 98.9%)
D7
Wild-Type 
36100%
(90.3%, 100%)
100%
(90.3%, 100%)

Table 9. Reproducibility results


The Oncomine Dx Target Test workflow—all results in 4 days

The Oncomine Dx Target Test workflow is a fully validated IVD workflow from beginning to end and includes all the reagents, consumables, instruments, and software to perform the test. It is possible to run 1–6 samples per run, plus 2 controls within 4 days (Figure 2).

dx-target-test-workflow
Figure 2. The Oncomine Dx Target Test utilizes a single streamlined NGS workflow for detecting cancer-associated biomarkers, incorporating reagents, instrument systems, and bioinformatics. The turnaround time, from FFPE sample to report, is 4 days.

Request more information

Abbreviated Intended Use: The Oncomine Dx Target Test is a qualitative in vitro diagnostic test that uses targeted high-throughput, parallel-sequencing technology to detect single nucleotide variants (SNVs) and deletions in 23 genes from DNA and fusions in ROS1 and RET from RNA isolated from formalin-fixed, paraffin-embedded (FFPE) tumor samples from patients with non-small cell lung cancer (NSCLC), and IDH1 R132 mutations from FFPE tumor tissue samples from patients with cholangiocarcinoma (CC) using the Ion PGM Dx System.

Test limitations and warnings

  • Use of this product must be limited to personnel trained in the techniques of PCR, NGS, and the use of the Oncomine Dx Target Test and the Ion PGM Dx System.
  • The Oncomine Dx Target Test has only been validated for use with NSCLC and CC FFPE tumor slide specimens.
  • The Oncomine Dx Target Test has been validated to detect the following somatic mutations: RNA fusions, single-nucleotide variations (SNVs), multi-nucleotide variations (MNVs), and deletions of 3, 6, 9, 12, 15, and 18 base pairs (bps), and insertions of 3, 6, 9, and 12 base pairs (bps) from DNA.
  • The Oncomine Dx Target Test is only validated for use with the Ion PGM Dx System and the Veriti Dx 96-Well Thermal Cycler.
  • The Oncomine Dx Target Test is only validated for use with 10 ng each of DNA and RNA per sample. Input amounts lower or higher than 10 ng are not recommended.
  • Both the DNA and RNA from a single sample extraction must meet the concentration requirements specified in the procedure. Do not use DNA from one extraction with RNA from a different extraction.
  • The effects of potential variations in FFPE specimen fixation have not been evaluated.
  • Extraction from FFPE sample curls has not been evaluated.
  • A potential source of contamination in the procedure is nucleic acid from previous sample processing steps. Follow good laboratory practices and all precautions and guidelines in these user guides to avoid cross-contamination between samples.
  • The Oncomine Dx Target Test is a qualitative test. The test is not for quantitative measurements of percent mutation.
  • The Ion OneTouch Rack Kit has only been designed to work with GeneMate SnapStrip 8-Strip 0.2 mL PCR Tubes. Tubes from other manufacturers may not fit properly in the rack, resulting in a higher risk of user error.

For NSCLC, the Oncomine Dx Target Test assay definition file includes prevalent but not all rare or newly identified RET isoforms, ROS1 isoforms, and EGFR exon 20 insertions. The Oncomine Dx Target Test may miss rare or newly identified:

  • RET isoforms carried by a subset of patients who may derive benefit from GAVRETO™ (pralsetinib)
  • ROS1 isoforms carried by a subset of patients who may derive benefit from XALKORI® (crizotinib)
  • EGFR exon 20 insertions carried by a subset of patients who may derive benefit from EXKIVITY™ (mobocertinib)