Performance characteristics of a polymerase chain reaction-based assay for the detection of EGFR mutations in plasma cell-free DNA from patients with non-small cell lung cancer using cell-free DNA collection tubes.

Survival rates in non-small cell lung cancer (NSCLC) are low. Detection of circulating tumor DNA in liquid biopsy (plasma) is increasingly used to identify targeted therapies for clinically actionable mutations, including EGFR mutations in NSCLC. The cobas® EGFR Mutation Test v2 (cobas EGFR test) is FDA-approved for EGFR mutation detection in tissue or liquid biopsy from NSCLC. Standard K2EDTA tubes require plasma separation from blood within 4 to 8 hours; however, Roche Cell-Free DNA (cfDNA) Collection Tubes (Roche cfDNA tube) enable whole blood stability for up to 7 days prior to plasma separation. This analysis assessed performance of Roche cfDNA tubes with the cobas EGFR test for the detection of EGFR mutations in plasma from healthy donors or patients with NSCLC. Overall, test performance was equally robust with either blood collection tube, eg, regarding limit of detection, linearity, and reproducibility, making Roche cfDNA tubes suitable for routine clinical laboratory use in this setting. Importantly, the Roche cfDNA tubes provided more flexibility for specimen handling versus K2EDTA tubes, eg, in terms of tube mixing, plasma separation, and sample stability, and do not require processing of blood within 8 hours thereby increasing the reach of plasma biopsies in NSCLC.

Biomarker-Directed Phase II Platform Study in Patients With EGFR Sensitizing Mutation-Positive Advanced/Metastatic Non-Small Cell Lung Cancer Whose Disease Has Progressed on First-Line Osimertinib Therapy (ORCHARD).

INTRODUCTION: Osimertinib, a third-generation, irreversible, epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI), selectively inhibits both EGFR-TKI sensitizing (EGFRm) and EGFR T790M resistance mutations and has demonstrated efficacy in non-small cell lung cancer (NSCLC) CNS metastases. Most patients with EGFRm NSCLC treated with osimertinib will eventually develop resistance. ORCHARD (NCT03944772) is a phase II study aiming to characterize first-line osimertinib resistance and identify post-progression treatments. METHODS: Adults aged ≥ 18 years (Japan ≥ 20 years), with EGFRm locally advanced/metastatic NSCLC will be allocated to one of three groups after first-line osimertinib progression, based on molecular profiling from a post-progression tumor biopsy. Group A will evaluate patients with protocol-determined biomarkers of resistance treated with novel osimertinib combination therapies, Group B will evaluate patients without a detectable protocol-determined biomarker treated with non-biomarker selected therapies that are chemotherapy- or EGFR-TKI-based, and Group C (observational) includes patients with histologically transformed disease, and/or a biomarker with an available therapy not investigated in ORCHARD. Group C patients will be treated as per local practice and followed to assess overall survival. The study's platform design allows for adaptability to include emerging treatments related to novel resistance mechanisms. The primary endpoint is confirmed objective response rate (investigator assessed). Other endpoints are progression-free survival, duration of response, overall survival, pharmacokinetics and safety. CONCLUSIONS: ORCHARD aims to characterize mechanisms of resistance to first-line osimertinib and explore treatments to overcome acquired resistance. The modular design allows for additional biomarker-directed cohorts and treatment options as understanding of osimertinib resistance mechanisms evolves.

Tissue and Plasma EGFR Mutation Analysis in the FLAURA Trial: Osimertinib versus Comparator EGFR Tyrosine Kinase Inhibitor as First-Line Treatment in Patients with EGFR-Mutated Advanced Non-Small Cell Lung Cancer.

PURPOSE: To assess the utility of the cobas EGFR Mutation Test, with tissue and plasma, for first-line osimertinib therapy for patients with EGFR-mutated (EGFRm; Ex19del and/or L858R) advanced or metastatic non-small cell lung cancer (NSCLC) from the FLAURA study (NCT02296125). EXPERIMENTAL DESIGN: Tumor tissue EGFRm status was determined at screening using the central cobas tissue test or a local tissue test. Baseline circulating tumor (ct)DNA EGFRm status was retrospectively determined with the central cobas plasma test. RESULTS: Of 994 patients screened, 556 were randomized (289 and 267 with central and local EGFR test results, respectively) and 438 failed screening. Of those randomized from local EGFR test results, 217 patients had available central test results; 211/217 (97%) were retrospectively confirmed EGFRm positive by central cobas tissue test. Using reference central cobas tissue test results, positive percent agreements with cobas plasma test results for Ex19del and L858R detection were 79% [95% confidence interval (CI), 74-84] and 68% (95% CI, 61-75), respectively. Progression-free survival (PFS) superiority with osimertinib over comparator EGFR-TKI remained consistent irrespective of randomization route (central/local EGFRm-positive tissue test). In both treatment arms, PFS was prolonged in plasma ctDNA EGFRm-negative (23.5 and 15.0 months) versus -positive patients (15.2 and 9.7 months). CONCLUSIONS: Our results support utility of cobas tissue and plasma testing to aid selection of patients with EGFRm advanced NSCLC for first-line osimertinib treatment. Lack of EGFRm detection in plasma was associated with prolonged PFS versus patients plasma EGFRm positive, potentially due to patients having lower tumor burden.

Fetal-juvenile origins of point mutations in the adult human tracheal-bronchial epithelium: absence of detectable effects of age, gender or smoking status.

Allele-specific mismatch amplification mutation assays (MAMA) of anatomically distinct sectors of the upper bronchial tracts of nine nonsmokers revealed many numerically dispersed clusters of the point mutations C742T, G746T, G747T of the TP53 gene, G35T of the KRAS gene and G508A of the HPRT1 gene. Assays of these five mutations in six smokers have yielded quantitatively similar results. One hundred and eighty four micro-anatomical sectors of 0.5-6x10(6) tracheal-bronchial epithelial cells represented en toto the equivalent of approximately 1.7 human smokers' bronchial trees to the fifth bifurcation. Statistically significant mutant copy numbers above the 95% upper confidence limits of historical background controls were found in 198 of 425 sector assays. No significant differences (P=0.1) for negative sector fractions, mutant fractions, distributions of mutant cluster size or anatomical positions were observed for smoking status, gender or age (38-76 year). Based on the modal cluster size of mitochondrial point mutants, the size of the adult bronchial epithelial maintenance turnover unit was estimated to be about 32 cells. When data from all 15 lungs were combined the log2 of nuclear mutant cluster size plotted against log2 of the number of clusters of a given cluster size displayed a slope of approximately 1.1 over a range of cluster sizes from approximately 2(6) to 2(15) mutant copies. A parsimonious interpretation of these nuclear and previously reported data for lung epithelial mitochondrial point mutant clusters is that they arose from mutations in stem cells at a high but constant rate per stem cell doubling during at least ten stem cell doublings of the later fetal-juvenile period. The upper and lower decile range of summed point mutant fractions among lungs was about 7.5-fold, suggesting an important source of stratification in the population with regard to risk of tumor initiation.

Analysis of mutational spectra by denaturing capillary electrophoresis.

The point mutational spectrum over nearly any 75- to 250-bp DNA sequence isolated from cells, tissues or large populations may be discovered using denaturing capillary electrophoresis (DCE). A modification of the standard DCE method that uses cycling temperature (e.g., +/-5 degrees C), CyDCE, permits optimal resolution of mutant sequences using computer-defined target sequences without preliminary optimization experiments. The protocol consists of three steps: computer design of target sequence including polymerase chain reaction (PCR) primers, high-fidelity DNA amplification by PCR and mutant sequence separation by CyDCE and takes about 6 h. DCE and CyDCE have been used to define quantitative point mutational spectra relating to errors of DNA polymerases, human cells in development and carcinogenesis, common gene-disease associations and microbial populations. Detection limits are about 5 x 10(-3) (mutants copies/total copies) but can be as low as 10(-6) (mutants copies/total copies) when DCE is used in combination with fraction collection for mutant enrichment. No other technological approach for unknown mutant detection and enumeration offers the sensitivity, generality and efficiency of the approach described herein.

Distributions of five common point mutants in the human tracheal-bronchial epithelium.

The mutations C742T, G746T, G747T in the TP53 gene and G35T in the KRAS gene have been repeatedly found in sectors of human tumors by direct DNA sequencing. The mutation G508A in the HPRT1 gene has been repeatedly found among peripheral T lymphocytes by clonal expansion under selective conditions. To discover if these mutations also occur frequently in normal tissues from which tumors arise, we have developed and validated allele-specific mismatch amplification mutation assays (MAMA) for each mutation. Reconstruction experiments demonstrated linearity in the range of 9-3000 mutant alleles among 3 x 10(6) wild-type alleles. The cumulative distributions of all negative controls established robust detection limits (P<0.05) of 34-125 mutants per 10(6) copies assayed depending on the mutation. One hundred and seventy-seven micro-anatomical samples of approximately (0.5-6)x10(6) tracheal-bronchial epithelial cells from nine non-smokers were assayed representing en toto the equivalent of approximately 1.6 human bronchial trees to the fifth bifurcation. Statistically significant mutant copy numbers were found in 257 of 463 assays. Clusters of mutant copies ranged from 10 to 1000 in 239/257 positive samples. As all five point mutations were detected at mutant fractions of >10(-5) in two or more lungs, we infer that they are mutational hotspots generated in lung epithelial stem cells. As the cancer-associated mutations did not differ in cluster size distribution from the HPRT1 mutation, we infer that none of the mutations conferred a growth advantage to somatic heterozygous clusters or maintenance turnover units. Specific mutants appeared in very large copy numbers, 1000-35,000, in 18/257 positive assays. Various hypotheses to account for the observed cluster size distributions are offered.

Design of an automated multicapillary instrument with fraction collection for DNA mutation discovery by constant denaturant capillary electrophoresis (CDCE).

A fundamental goal ingenomics is the discovery of genetic variation that contributes to disease states or to differential drug responses. Single nucleotide polymorphism (SNP) detection has been the focus of much attention in the study of genetic variation over the last decade. These SNPs typically occur at a frequency greater than 1% in the human genome. Recently, low-frequency alleles are also being increasingly recognized as critical to obtain an improved understanding of the correlation between genetic variation and disease. Although many methods have been reported for the discovery and scoringof SNPs, sensitive, automated, and cost-effective methods and platforms for the discovery of low-frequency alleles are not yet readily available. We describe here an automated multicapillary instrument for high-throughput detection of low-frequency alleles from pooled samples using constant denaturant capillary electrophoresis. The instrument features high optical sensitivity (1 x 10(-12) M fluorescein detection limit), precise and stable temperature control (+/- 0.01degrees C), and automation for sample delivery, injection, matrix replacement, and fraction collection. The capillary array is divided into six groups of four capillaries, each of which can be independently set at any temperature ranging from room temperature to 90 degrees C. The key performance characteristics of the instrument are reported.

Clustering of mutant mitochondrial DNA copies suggests stem cells are common in human bronchial epithelium.

Tissue maintenance stem cells, as opposed to transition and/or terminal cells in the epithelium, are possible progenitor cells for human tumors, but little is known about their frequency in human tissues. It occurred to us that the colonies of mutants that should be created when a stem cell mutates and transmits the rare mutation to its descendent transition and terminal cells should, given a quantitative mutation assay, define the average number of cells in a maintenance turnover unit and permit calculation of stem cell number. To test this concept we used a combination of high fidelity PCR and constant denaturant capillary electrophoresis to enumerate mitochondrial point mutations and define their number and distribution among multiple small samples of approximately one million cells containing about 400 million copies of mitochondrial DNA. The bulk of the data were best explained by a model in which most stem cells, defined here as long-lived cells, give rise to colonies of approximately 8-128 cells. In addition, we found that about 1.5% of colonies contained hundreds or even thousands of homoplasmic mutant cells. These expanded turnover units suggest the bronchial epithelium may contain large clusters of cells with mutations, and possibly phenotypic alterations as well.

Scanning the beta-globin gene for mutations in large populations by denaturing capillary and gel electrophoresis.

Separation of mutant from nonmutant DNA sequences of 100 bp may be accomplished by using defined denaturing conditions of chemical denaturant and/or elevated temperature during electrophoresis on either polyacrylamide slab gels (denaturing gradient gel electrophoresis, DGGE) or capillary gels (constant denaturant capillary electrophoresis, CDCE). In analysis of mutant directly from a polymerase chain reaction (PCR) product mixture, both have detection sensitivities of approximately 1%. CDCE that facilitates an intermediate mutant enrichment step permits detection of mutants at fractions as low as 2 x 10(-6). Here we report the successful application of both approaches to scan for mutations of the human beta-globin gene (HBB) in two human population samples of approximately 5000 persons in the HBB. Using DGGE, the coding region and flanking intronic splice sites of HBB were scanned in a population of 4949 Han Chinese individuals in pool sizes of 48 individual DNA samples. Four point mutations ranging in mutant frequency from 0.5 to 0.0002 were identified. Using CDCE with a mutant enrichment step, these same sequences were scanned in a population of 5028, predominantly African-American juveniles (<9 years) as a single pooled DNA sample. Three point mutations were identified ranging in mutant frequency from 0.13 to 0.0005. This study shows that both the DGGE/small pool and the CDCE/large pool approaches offer the means to define the fine structure map of genetic variation in large population samples, and with appropriately engineered facilities to provide high throughput, should be useful in pangenomic scans to discover genes carrying casual mutations for common diseases.

Detection and frequency estimation of rare variants in pools of genomic DNA from large populations using mutational spectrometry.

DNA variants underlying the inheritance of risk for common diseases are expected to have a wide range of population allele frequencies. The detection and scoring of the rare alleles (at frequencies of <0.01) presents significant practical problems, including the requirement for large sample sizes and the limitations inherent in current methodologies for allele discrimination. In the present report, we have applied mutational spectrometry based on constant denaturing capillary electrophoresis (CDCE) to DNA pools from large populations in order to improve the prospects of testing the role of rare variants in common diseases on a large scale. We conducted a pilot study of the cytotoxic T lymphocyte-associated antigen-4 gene (CTLA4) in type 1 diabetes (T1D). A total of 1228 bp, comprising 98% of the CTLA4 coding sequence, all adjacent intronic mRNA splice sites, and a 3' UTR sequence were scanned for unknown point mutations in pools of genomic DNA from a control population of 10,464 young American adults and two T1D populations, one American (1799 individuals) and one from the United Kingdom (2102 individuals). The data suggest that it is unlikely that rare variants in the scanned regions of CTLA4 represent a significant proportion of T1D risk and illustrate that CDCE-based mutational spectrometry of DNA pools offers a feasible and cost-effective means of testing the role of rare variants in susceptibility to common diseases.