1) CTC Identified
Samples are processed by standard procedures and CTCs are identified (see standard enumeration process for details) for recovery guided by morphology and marker expression.
The Epic Sciences' platform analyzes all nucleated cells within a blood sample at single cell resolution. Cells from a patient’s blood sample are deposited on replicate glass slides and compared for morphological features, expression of biomarkers and nuclear integrity, using immunofluorescent staining. Guided by cell phenotype and marker expression, CTCs are individually recovered from the slide surface. Their genomes are amplified (WGA) and analyzed by next generation sequencing (NGS) for the presence of point mutations, copy number alterations, genome wide chromosomal instability, ploidy or genome wide scarring.
Samples are processed by standard procedures and CTCs are identified (see standard enumeration process for details) for recovery guided by morphology and marker expression.
Using the unique cell ID number and the associated XY coordinates, CTCs are relocated and recovered from the slide surface.
Cells are lysed. Their genomes are amplified and are then constructed into libraries.
Libraries are sequenced using a HiSeq 2500. Reads are trimmed and aligned.
Aligned reads are binned into genomic regions and normalized for variance, read number and WBC control.
CNV profiles are clustered to measure # of clonal populations per patient and associated with cell morphology and biomarker expression to characterize CTC clonality.
Example data showing the clonal evolution: CTCs from a single patient were identified and characterized through our CNV assay. The distinct clonal species and their relationship to each other are shown below.
Comparison of called variant function (left) and impact (right) across all patient CTCs: Moderate and high impact clonal and subclonal alterations detected across all patient CTCs (below).
| Gene | Mutatioin | %CTCs | Cosmic ID | Impact | Comment |
|---|---|---|---|---|---|
|
MLL3 |
p.Glyu838Ser/c.2512G>A |
71.9% |
Moderate |
~20% in crc associated with tumor progression and microsatellite instability |
|
|
SSX2 |
p.Awe174Gly/c.520A>G |
31.3% |
Moderate |
||
|
MLL3 |
p.Tyr816fs/c.2447dupA |
25.0% |
289942 |
High |
|
|
ATM |
p.Val1534Leu/c.4600G>C |
18.8% |
Moderate |
Homologous recombination deficiency ~ PARP inhibitors |
|
|
EIF4A2 |
p.Gln118*/c.352C>T |
9.4% |
High |
Co-occurance with MLL3 clonal alteration |
|
|
NOTCH2 |
pAsn46Ser/c.137A>G |
9.4% |
Moderate |
||
|
KRAS |
p.Gly12Asp/c.35G.A |
6.3% |
521 |
Moderate |
Associated with resistance to EGFR TKIs |
|
EML4 |
p.Glu130fs/c.387delA |
6.3% |
1408084 |
High |
|
|
ABL2 |
p.His544fs/c.1630delC |
3.1% |
High |
||
|
PIK3CA |
p.Pro27fs/c.80delC |
3.1% |
High |
Resistance to EFFR-TKIs |
|
|
TP53 |
p.Cys229*/c.687T>A |
3.1% |
45394 |
High |
Tumor suppressor inactivation |
|
BRCA1 |
pLys654fs/c.1961delA |
3.1% |
1383519 |
High |
Homologous recombination deficiency ~ PARP inhibitors |
Every time Epic’s AR-V7 test is administered the healthcare system saves about $15,000 per year of life extended. In other words approximately $7,000 to $9,000 per patient tested.
Of the roughly 30 million cells in a typical blood sample, about 5 are cancer cells on their way to form tumors elsewhere in the body. Modern cancer diagnostics technology can't see them.
We spend about $100 billion a year on cancer drugs. But because we’re not matching the right drug to the right patient and their disease in a personalized way, about 3/4ths of that drug spend is wasted.
The Epic approach answers a critical question for a doctor: When is a patient resistant? By answering that question accurately and precisely we increase patient life and save the system money at the same time.
Today, Epic Sciences is embedded in clinical trials with 48 leading producers of oncology drugs and 35 of the top academic cancer hospitals.
It takes only 7 minutes to acquire an entire image of 6 million stained cells on a 1 x 3 slide. Then that image, which is about 20GB of data, gets sent up to the cloud where the algorithms run and find the rare events in about 2 more minutes.