Biotechnology of High Sensitivity PCR for Oncology Biomarkers
1. Two novel, low frequency variant detection methods (castPCR and dPCR) show
promise as quantitative cancer biomarkers in clinical feasibility studies.
Kirsten Copren1 (PhD), Cassandra Adams1 (PhD), Robert Warren2 (MD), James Rubenstein3 (MD, PhD)
1Genome Analysis Core 2Department of Surgery 3Department of Medicine
The early detection of possible oncogenic driver mutations in clinically
relevant DNA samples has treatment and prognostic value in cancer.
Case Study 1: Colon Cancer
Gene: TSC1, Tuberous sclerosis 1 (TSC1), tumor growth suppressor. Two
potential oncogenic mutations not previously found in colon cancer
identified by WGS . Single assay could be used to screen patients.
Digital PCR (dPCR) Competitive allele specific
Taqman® PCR
(castPCR)
Case Study 2: Primary Central Nervous
System (CNS) lymphoma
Gene: Myd88, Myeloid differentiation primary response gene 88
Sample type: Cerebral Spinal Fluid (CSF). Synthetic oligo sensitive and
specific to 0.05ng input.
dPCR Application Results castPCR Application Results
• Using DNA extracted from cerebral spinal fluid (CSF) of primary
central nervous system (CNS) lymphoma patients to study
disease progression.
• Common mutation identifying primary CNS lymphoma cells is
Myd88 L265P.
• Used for validation of putative mutation L696* in Tsc1 by next-
generation sequencing.
Genome Analysis Core ● genomecore.ucsf.edu ● genome2@ucsf.edu
QuantStudio 3D. Images from Thermofisher.com
AB7900HT. Images from Thermofisher.com
• Using synthetic template DNA the assay can detect the
mutant allele to a dilution of 0.1% in the presence of the WT
background. Equal to 0.2pg of mutant DNA in a total of 0.2ng
DNA.
• Reproducible results to 3% in cell lines. Needs further
replication on cell lines prior to analysis on CSF samples.
12.5% 5% 1% 0.5%0.1%
NTC
• NGS detected the mutant allele (L969*) at a percentage of 7.2%.
• Using dPCR on the sample we were unable to detect the
mutation, despite having demonstrated sensitivity using template
DNA to 5%, therefore suggesting this mutation may be a
sequencing artifact.
dPCR data summary for sample run with L969* Concentrations (copies/µL) and precision values were calculated by the
QuantStudio 3D AnalysisSuite Cloud Software. The precision indicates the probability of a false positive, and values <10% are
desirable . Allele frequencies were calculated using the following formula: frequency = (allele conc./(Wt conc. + Mut conc.))*100.
All values represent the average of replicate chips.
Image from Thermofisher.com
Technical Challenges:
• The mutant allele is present at very low frequencies (<5%) among a
background of wild type.
• The starting DNA template is often fragmented and low input in clinically
relevant samples.
• Assay must be highly sensitive, accurate, and reproducible.
Case Study 1: PDX* Models of metastatic colon
cancer and spheroid (3-D) cell cultures
Goal: Treat both with small molecule drugs, investigate coincident response.
*PDX = Patient Derived Xenograft
Case Study 2: Are tumor cells
circulating in cerebral spinal fluid?
Goal: Prognostic value if the measured amount of the
mutant allele shows a decrease in the patient sample after
therapeutic treatment.
Sample type: FFPE tumor. Allele present at 7.2% of total sample
via WGS
Validation and calibration of assay using double stranded synthetic oligo
templates (gBlocks from IDTDNA) shows no off-target effects.
90% wildtype and 10% mutant mixture of synthetic templates requires
additional replicates for better reproducibility
Mut DNA template
amplification with Mut
assay
WT DNA template
amplification with WT
assay
WT DNA Mut DNA
WT assay Mut assay
Mut assay
WT DNAMut DNA
WT or Mut DNA template amplification using Mut assay