Classifications

• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
  • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
  • C12Q1/6806—Preparing nucleic acids for analysis, e.g. for polymerase chain reaction [PCR] assay
• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
  • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
  • C12Q1/6876—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes
  • C12Q1/6883—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material
  • C12Q1/6886—Nucleic acid products used in the analysis of nucleic acids, e.g. primers or probes for diseases caused by alterations of genetic material for cancer
• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q1/00—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions
  • C12Q1/68—Measuring or testing processes involving enzymes, nucleic acids or microorganisms; Compositions therefor; Processes of preparing such compositions involving nucleic acids
  • C12Q1/6844—Nucleic acid amplification reactions
  • C12Q1/6858—Allele-specific amplification
• • C—CHEMISTRY; METALLURGY
  • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
  • C12Q—MEASURING OR TESTING PROCESSES INVOLVING ENZYMES, NUCLEIC ACIDS OR MICROORGANISMS; COMPOSITIONS OR TEST PAPERS THEREFOR; PROCESSES OF PREPARING SUCH COMPOSITIONS; CONDITION-RESPONSIVE CONTROL IN MICROBIOLOGICAL OR ENZYMOLOGICAL PROCESSES
  • C12Q2600/00—Oligonucleotides characterized by their use
  • C12Q2600/156—Polymorphic or mutational markers

Definitions

• the disclosed inventions relate generally to methods for detecting nucleic acid mutations and fusions using amplification methods such as the polymerase chain reaction (PCR).
• PCR polymerase chain reaction
• a method for determining the single nucleotide variants present in a lung squamous cell carcinoma includes generating a set of amplicons by performing a multiplex amplification reaction on nucleic acids isolated from a sample of blood or a fraction thereof from an individual suspected of having a lung squamous cell carcinoma, wherein each amplicon of the set of amplicons spans at least one single nucleotide variant loci of a set of single nucleotide variant loci known to be associated with lung cancer; and
• a method for supporting a lung cancer diagnosis for an individual suspected of having lung cancer from a sample of blood or a fraction thereof from the individual includes generating a set of amplicons by performing a multiplex amplification reaction on nucleic acids isolated from the sample, wherein each amplicon of the set of amplicons spans at least one single nucleotide variant loci of a set of single nucleotide variant loci known to be associated with lung cancer; and
• the presence of a single nucleotide variant supports a diagnosis of squamous cell carcinoma or a stage 2b or 3 a adenocarcinoma, and/or
• the presence of 5, 10, 15 or more single nucleotide variants supports a diagnosis of squamous cell carcinoma or a stage 2b or 3 adenocarcinoma.
• the presence of 5, 10, or 15 or more single nucleotide variants supports a diagnosis of squamous cell carcinoma or a stage 3 adenocarcinoma
• the amplification reaction is a PCR reaction
• the annealing temperature is between 1 and 15 °C greater than the melting temperature of at least 50, 60, 70, 85, 80, 90, 95, or 100% of the primers of the set of primers
• the length of the annealing step in the PCR reaction is between 15 and 60 minutes
• the primer concentration in the amplification reaction is between 1 and 10 nM
• the primers in the set of primers are designed to minimize primer dimer formation.
• an efficiency and an error rate per cycle can be determined for each amplification reaction of the multiplex amplification reaction of the set of single nucleotide variance loci, and the efficiency and the error rate can be used to determine whether a single nucleotide variant at the set of single variant loci is present in the sample.
• a confidence is determined and a SNV call is made if a cutoff confidence value is exceeded, such as 90%, 95%, or 98% confidence.
• inventive compositions and solid supports are provided herein.
• FIG. 1 is a workflow Diagram.
• FIG. 2 Top panel: the number of SNVs per sample; bottom panel: the working assays, sorted by driver category.
• Figure 3 Measured cfDNA concentration. Each data point refers to a plasma sample.
• FIG. 4 Samples showing good correlation between tissue VAF measurements determined previously (x axis) and here using mPCR-NGS (y axis). Each sample is shown in a separate box, and the VAF data points are colored by tissue subsection.
• FIG. 5 Samples showing poor correlation between tissue VAF measurements determined previously (x axis) and here using mPCR-NGS (y axis). Each sample is shown in a separate box, and the VAF data points are colored by tissue subsection.
• Figure 6A Depth of read histogram as a function of the resulting call. Top: the assay did not detect the expected plasma SNV. Bottom: the assay detected the expected plasma SNV.
• Figure 7 Number of SNVs detected in plasma by histological type.
• FIG. 1 SNV detection (left) and sample detection (right) in plasma by tumor stage.
• Figure 9 Plasma VAF as a function of tumor stage and SNV clonality.
• Figure 10 Number of SNVs detected in plasma from each sample as a function of the cfDNA input amount.
• FIG. 11 Plasma VAF as a function of average tumor VAF. Average tumor VAF was calculated across all the tumor sub-sections analyzed from each tumor.
• FIG. 12 shows the clonal ratios (red to blue) and mutant variant allele frequency (MutVAF) of each detected SNV.
• the total SNVs detected from each sample are placed in a single column and the samples are categorized by tumor stage (pTNMstage). Samples with no detected SNVs are included.
• the clonal ratio is defined as the ratio between the number of tumor subsections in which SNV was observed and the total number of subsections analyzed from that tumor.
• FIG. 13 shows the clonal status (blue for clonal and red for subclonal) and mutant variant allele frequency (MutVAF) of each detected SNV.
• the total SNVs detected from each sample are placed in a single column and the samples are categorized by tumor stage (pTNMstage). Samples with no detected SNVs are included.
• the clonal status was determined by PyCloneCluster using whole exome equencing data from the tumor tissue.
• FIG. 14 shows the clonal status (blue for clonal and red for subclonal) and mutant variant allele frequency (MutVAF) of each detected SNV where the top panel shows only the clonal SNVs and the bottom panel shows only the subclonal SNVs.
• the total SNVs detected from each sample are placed in a single column and the samples are categorized by tumor stage (pTNMstage). Samples with no detected SNVs are included.
• the clonal status was determined by PyCloneCluster using whole exome equencing data from the tumor tissue.
• FIG. 15 shows the number of SNVs detected in plasma as a function of histological type and tumor size.
• the histological type and tumor stage were determined by the pathology report.
• Each data point is colored by size, where red denotes the largest tumor size and blue denotes the smallest tumor size.
• FIG. 16 is a table of cfDNA analysis showing DNA concentration, genome copy equivalents into library prep, plasma hemolysis grade, and cDNA profile in all samples.
• FIG. 17 is a table of SNVs detected in the plasma for each sample.
• FIG. 18 is a table of additional SNVs detected in plasma.
• FIG. 19 is a table of assay count based for genes for the experiments in Example 1.
• FIG. 20 is a table of information regarding the samples analyzed in the study of Example 1 as well as data generated from the experiment provided in Example 1.
• FIG. 21 is an example of detected assays and their background allele fractions for a plasma sample at relapse time (LTX103).
• Methods and compositions provided herein improve the detection, diagnosis, staging, screening, treatment, and management of lung cancer.
• Methods provided herein in illustrative embodiments analyze single nucleotide variant mutations (SNVs) in circulating fluids, especially circulating tumor DNA.
• SNVs single nucleotide variant mutations
• the methods provide the advantage of identifying more of the mutations that are found in a tumor and clonal as well as subclonal mutations, in a single test, rather than multiple tests that would be required, if effective at all, that utilize tumor samples.
• the methods and compositions can be helpful on their own, or they can be helpful when used along with other methods for detection, diagnosis, staging, screening, treatment, and management of lung cancer, for example to help support the results of these other methods to provide more confidence and/or a definitive result.
• a method for determining the single nucleotide variants present in a lung squamous cell carcinoma by determining the single nucleotide variants present in a ctDNA sample from an individual, such as an individual having or suspected of having, squamous cell carcinoma, using a ctDNA SNV amplification/sequencing workflow provided herein.
• a method for detecting lung sqaumous cell carcinoma in a sample of blood or a fraction thereof from an individual, such as an individual suspected of having a cancer that includes determining the single nucleotide variants present in a sample by determining the single nucleotide variants present in a ctDNA sample using a ctDNA SNV amplification/sequencing workflow provided herein.
• the method includes performing a ctDNA SNV amplification/sequencing workflow as provided herein, and determining the variant allele frequency for each of the SNV loci based on the sequence of the plurality of copies of the series of amplicons.
• a higher relative allele frequency compared to the other single nucleotide variants of the plurality of single nucleotide variant loci is indicative of a clonal single nucleotide variant in the tumor.
• Variant allele frequencies are well known in the sequencing art. Support for this embodiment, is provided, for example in FIGs. 12-14.
• the method further includes determining a treatment plan, therapy and/or administering a compound to the individual that targets the one or more clonal single nucleotide variants.
• subclonal and/or other clonal SNVs are not targeted by therapy.
• Specific therapies and associated mutations are provided in other sections of this specification and are known in the art.
• the method further includes administering a compound to the individual, where the compound is known to be specifically effective in treating lung squamous cell carcinoma having one or more of the determined single nucleotide variants.
• a variant allele frequency of greater than 0.25%, 0.5%, 0.75%, 1.0%, 5% or 10% is indicative a clonal single nucleotide variant.
• the squamous cell carcinoma is a stage la, lb, or 2a squamous cell carcinoma. In certain examples of this embodiment, the squamous cell carcinoma is a stage la or lb squamous cell carcinoma. In certain examples of the embodiment, the individual is not subjected to surgery. In certain examples of the embodiment, the individual is not subjected to a biopsy.
• a clonal SNV is identified or further identified if other testing such as direct tumor testing suggest an on-test SNV is a clonal SNV, for any SNV on test that has a variable allele frequency greater than at least one quarter, one third, one half, or three quarters of the other single nucleotide variants that were determined.
• methods herein for detecting SNVs in ctDNA can be used instead of direct analysis of DNA from a tumor.
• Results provided herein demonstrate that SNVs that are much more likely to be clonal SNVs have higher VAFs (See e.g. FIGs. 12-14).
• data is provided on SNVs that are found in a tumor from the individual. Accordingly, in these embodiments, a SNV amplification/sequencing reaction is performed on one or more tumor samples from the individual.
• the ctDNA SNV amplification/sequencing reaction provided herein is still advantageous because it provides a liquid biopsy of clonal and subclonal mutations. Furthermore, as provided herein, clonal mutations can be more unambiguously identified in an individual that has lung cancer, if a high VAF percentage, for example, more than 1, 2, 3, 4, 5, 6, 7, 8, 9, 10% VAF in a ctDNA sample from the individual is determined for an SNV.
• method provided herein can be used to determine whether to isolate and analyze ctDNA from circulating free nucleic acids from an individual with lung cancer. First, it is determined whether the lung cancer is an adenocarcinoma or a squamous cell carcinoma. If the lung cancer is a squamous cell carcinoma circulating free nucleic acids are isolated from individual. The method in some examples, further includes determining the stage of the lung cancer, wherein if the lung cancer is squamous cell carcinoma or stage 3a adenocarcinoma, circulating free nucleic acids are isolated from the individual. Results provided in FIG. 15 and in tabular form in FIG.
• SNVs are prevalent in squamous cell carcinoma or stage 3a adenocarcinoma, However, SNVs are much less prevalent in earlier stage ADCs. Accordingly, important health care savings can be realized by saving from testing for SNVs in stage la, lb, and/or 2a ADC patients.
• the lung cancer is squamous cell carcinoma or stage 3a adenocarcinoma
• circulating free nucleic acids are isolated from the individual.
• the lung cancer is stage squamous cell carcinoma or stage 3a adenocarcinoma nucleic acids are not isolated from a lung tumor of the individual.
• inventive compositions and/or solid supports are inventive compositions and/or solid supports.
• Fl. A composition comprising circulating tumor nucleic acid fragments comprising a universal adapter, wherein the circulating tumor nucleic acids originated from a lung squamous cell carcinoma tumor.
• an inventive composition that includes circulating tumor nucleic acid fragments comprising a universal adapter, wherein the circulating tumor nucleic acids originated from a sample of blood or a fraction thereof, of an individual with lung squamous cell carcinoma.
• Results presented in Example 1 demonstrate the surprising advantage of ctDNA SNV amplification/sequencing test methods. These methods typically include formation of ctDNA fragment that include a universal adapter.
• such methods typically include the formation of a solid support especially a solid support for high throughput sequencing, that includes a plurality of clonal populations of nucleic acids, wherein the clonal populations comprise amplicons generated from a sample of circulating free nucleic acids, wherein the ctDNA.
• the ctDNA originated from a lung squamous cell carcinoma tumor.
• a solid support comprising a plurality of clonal populations of nucleic acids, wherein the clonal populations comprise nucleic acid fragments generated from a sample of circulating free nucleic acids from a sample of blood or a fraction thereof, from an individual with lung squamous cell carcinoma.
• the nucleic acid fragments in different clonal populations comprise the same universal adapter.
• Such a composition is typically formed during a high throughput sequencing reaction in methods of the present invention, as performed in Example 1.
• the clonal populations of nucleic acids can be derived from nucleic acid fragments from a set of samples from two or more individuals.
• the nucleic acid fragments comprise one of a series of molecular barcodes corresponding to a sample in the set of samples.
• the methods for determining whether a single nucleotide variant is present in the sample includes identifying a confidence value for each allele determination at each of the set of single nucleotide variance loci, which can be based at least in part on a depth of read for the loci.
• the confidence limit can be set at least 75%, 80%, 85%, 90%, 95%, 96%, 96%, 98%, or 99%.
• the confidence limit can be set at different levels for different types of mutations.
• the method can performed with a depth of read for the set of single nucleotide variance loci of at least 5, 10, 15, 20, 25, 50, 100, 150, 200, 250, 500, 1,000, 10,000, 25,000, 50,000, 100,000, 250,000, 500,000, or 1 million.
• FIG. 20 provides depth of read data for SNV loci successfully analyzed in Example 1.
• a method of any of the embodiments herein includes determining an efficiency and/or an error rate per cycle are determined for each amplification reaction of the multiplex amplification reaction of the single nucleotide variance loci. The efficiency and the error rate can then be used to determine whether a single nucleotide variant at the set of single variant loci is present in the sample. More detailed analytical steps provided in SNV Method 2 provided in the analytical method can be included as well, in certain embodiments.
• the set of single nucleotide variance loci includes all of the single nucleotide variance loci identified in the TCGA and COSMIC data sets for lung cancer, or for lung adenocarcinoma and/or especially lung squamous cell carcinoma.
• the set of single nucleotide variant loci include 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 40, 50, 75, 100, 250, 500, 1000, 2500, 5000, or 10,000 single nucleotide variance loci known to be associated with lung cancer, lung ADC, and/or especially lung SCC on the low end of the range, and , 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 40, 50, 75, 100, 250, 500, 1000, 2500, 5000, 10,000, 20,000 and 25,000 on the high end of the range.
• any of the methods for detecting SNVs herein that include a ctDNA SNV amplification/sequencing workflow improved amplification parameters for multiplex PCR can be employed.
• the amplification reaction is a PCR reaction and the annealing temperature is between 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10°C greater than the melting temperature on the low end of the range, and 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15° on the high end the range for at least 10, 20, 25, 30, 40, 50, 06, 70, 75, 80, 90, 95 or 100% the primers of the set of primers.
• the amplification reaction is a PCR reaction
• the length of the annealing step in the PCR reaction is between 10, 15, 20, 30, 45, and 60 minutes on the low end of the range, and 15, 20, 30, 45, 60, 120, 180, or 240 minutes on the high end of the range.
• the primer concentration in the amplification, such as the PCR reaction is between 1 and 10 nM.
• the primers in the set of primers are designed to minimize primer dimer formation.
• the amplification reaction is a PCR reaction
• the annealing temperature is between 1 and 10 °C greater than the melting temperature of at least 90% of the primers of the set of primers
• the length of the annealing step in the PCR reaction is between 15 and 60 minutes
• the primer concentration in the amplification reaction is between 1 and 10 nM
• the primers in the set of primers are designed to minimize primer dimer formation.
• the multiplex amplification reaction is performed under limiting primer conditions.
• a method for supporting a lung cancer diagnosis for an individual such as an individual suspected of having lung cancer, from a sample of blood or a fraction thereof from the individual, that includes performing a ctDNA SNV amplification/sequencing workflow as provided herein, to determine whether one or more single nucleotide variants are present in the plurality of single nucleotide variant loci.
• a ctDNA SNV amplification/sequencing workflow as provided herein, to determine whether one or more single nucleotide variants are present in the plurality of single nucleotide variant loci.
• the absence of a single nucleotide variant supports a diagnosis of stage la, lb, or 2a adenocarcinoma
• the presence of a single nucleotide variant supports a diagnosis of squamous cell carcinoma or a stage 2b or 3 a adenocarcinoma, and/or
• the presence of ten or more single nucleotide variants supports a diagnosis of squamous cell carcinoma or a stage 2b or 3 adenocarcinoma.
• Example 1 See e.g. the tabular data in FIG. 20). These results identify analysis using a ctDNA SNV amplification/sequencing workflow of lung ADC and SCC samples from an individual as a valuable method for identifying SNVs found in an ADC tumor, especially for stage 2b and 3a ADC tumors, and especially an SCC tumor at any stage (See e.g. FIG. 15 and FIG. 20).
• this embodiment further includes determining the stage of a lung cancer lesion by a non-invasive method, For example, the size of a tumor can be determined by non-invasive methods.
• methods herein for detecting SNVs can be used to direct a therapeutic regimen. Therapies are available and under development that target specific mutations associated with ADC and SCC (Nature Review Cancer. 14:535-551 (2014). For example, detection of an EGFR mutation at L858R or T790M can be informative for selecting a therapy. Erlotinib, gefitinib, afatinib, AZK9291, CO-1686, and HM61713 are current therapies approved in the U.S.
• a G12D, G12C, or G12V mutation in KRAS can be used to direct an individual to a therapy of a combination of Selumetinib plus docetaxel.
• a mutation of V600E in BRAF can be used to direct a subject to a treatment of Vemurafenib, dabrafenib, and trametinib.
• a sample analyzed in methods of the present invention in certain illustrative embodiments, is a blood sample, or a fraction thereof.
• Methods provided herein, in certain embodiments, are specially adapted for amplifying DNA fragments, especially tumor DNA fragments that are found in circulating tumor DNA (ctDNA). Such fragments are typically about 160 nucleotides in length.
• cell-free nucleic acid e.g cfDNA
• cfNA cell-free nucleic acid
• the cfDNA is fragmented and the size distribution of the fragments varies from 150- 350 bp to > 10000 bp.
• HCC hepatocellular carcinoma
• the circulating tumor DNA is isolated from blood using EDTA-2Na tube after removal of cellular debris and platelets by centrifugation.
• the plasma samples can be stored at -80oC until the DNA is extracted using, for example, QIAamp DNA Mini Kit (Qiagen, Hilden, Germany), (e.g. Hamakawa et al., Br J Cancer. 2015; 112:352-356).
• Hamakava et al. reported median concentration of extracted cell free DNA of all samples 43.1 ng per ml plasma (range 9.5-1338 ng ml/) and a mutant fraction range of 0.001-77.8%, with a median of 0.90%.
• the sample is a tumor.
• Methods are known in the art for isolating nucleic acid from a tumor and for creating a nucleic acid library from such a DNA sample given the teachings here.
• a skilled artisan will recognize how to create a nucleic acid library appropriate for the methods herein from other samples such as other liquid samples where the DNA is free floating in addition to ctDNA samples.
• Methods of the present invention typically include a step of generating and amplifying a nucleic acid library from the sample (i.e. library preparation).
• the nucleic acids from the sample during the library preparation step can have ligation adapters, often referred to as library tags or ligation adaptor tags (LTs), appended, where the ligation adapters contain a universal priming sequence, followed by a universal amplification. In an embodiment, this may be done using a standard protocol designed to create sequencing libraries after fragmentation.
• the DNA sample can be blunt ended, and then an A can be added at the 3' end.
• a Y-adaptor with a T-overhang can be added and ligated.
• other sticky ends can be used other than an A or T overhang.
• other adaptors can be added, for example looped ligation adaptors.
• the adaptors may have tag designed for PCR amplification.
• a number of the embodiments provided herein include detecting the SNVs in a ctDNA sample.
• Such methods include an amplification step and a sequencing step (Sometimes referred to herein as a "ctDNA SNV amplification/sequencing workflow).
• a ctDNA amplification/sequencing workflow can include generating a set of amplicons by performing a multiplex amplification reaction on nucleic acids isolated from a sample of blood or a fraction thereof from an individual, such as an individual suspected of having a lung cancer, for example a squamous cell carcinoma, wherein each amplicon of the set of amplicons spans at least one single nucleotide variant loci of a set of single nucleotide variant loci, such as an SNV loci known to be associated with lung cancer; and
• this exemplary method determines the single nucleotide variants present in the sample.
• Exemplary ctDNA SNV amplification/sequencing workflows in more detail can include forming an amplification reaction mixture by combining a polymerase, nucleotide triphosphates, nucleic acid fragments from a nucleic acid library generated from the sample, and a set of primers that each binds an effective distance from a single nucleotide variant loci, or a set of primer pairs that each span an effective region that includes a single nucleotide variant loci.
• the single nucleotide variant loci in exemplary embodiments, is one known to be associated with lung cancer, for example lung adenocarcinoma and/or in especially illustrative embodiemnts squamous cell carcinoma.
• the effective distance of binding of the primers can be within 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, 35, 40, 45, 50, 75, 100, 125, or 150 base pairs of a SNV loci.
• the effective range that a pair of primers spans typically includes an SNV and is typically 160 base pairs or less, and can be 150, 140, 130, 125, 100, 75, 50 or 25 base pairs or less.
• the effective range that a pair of primers spans is 20, 25, 30, 40, 50, 60, 70, 75, 100, 110, 120, 125, 130, 140, or 150 nucleotides from an SNV loci on the low end of the range, and 25, 30, 40, 50, 60, 70, 75, 100, 110, 120, 125, 130, 140, or 150, 160, 170, 175, or 200 on the high end of the range.
• nucleic acid sequencing data is generated for amplicons created by the tiled multiplex PCR.
• Algorithm design tools are available that can be used and/or adapted to analyze this data to determine within certain confidence limits, whether a mutation, such as a SNV is present in a target gene, as illustrated in Example 1 herein.
• Sequencing Reads can be demultiplexed using an in-house tool and mapped using the Burrows-Wheeler alignment software, Bwa mem function (BWA, Burrows-Wheeler Alignment Software (see Li H. and Durbin R. (2010) Fast and accurate long-read alignment with Burrows- Wheeler Transform. Bioinformatics, Epub. [PMID: 20080505]) on single end mode using pear merged reads to the hgl9 genome.
• Amplification statistics QC can be performed by analyzing total reads, number of mapped reads, number of mapped reads on target, and number of reads counted.
• any analytical method for detecting an SNV from nucleic acid sequencing data detection can be used with methods of the invention methods of the invention that include a step of detecting an SNV or determining whether an SNV is present.
• methods of the invention that utilize SNV METHOD 1 below are used.
• methods of the invention that include a step of detecting an SNV or determining whether an SNV is present at an SNV loci utilize SNV METHOD 2 below.
• SNV METHOD 1 For this embodiment, a background error model is constructed using normal plasma samples, which were sequenced on the same sequencing run to account for run- specific artifacts. In certain embodiments, 5, 10, 15, 20, 25, 30, 40, 50, 100, 150, 200, 250, or more than 250 normal plasma samples are analyzed on the same sequencing run. In certain illustrative embodiments, 20, 25, 40, or 50 normal plasma samples are analyzed on the same sequencing run. noisysy positions with normal median variant allele frequency greater than a cutoff are removed. For example this cutoff in certain embodiments is > 0.1%, 0.2%, 0.25%, 0.5%, 1%, 2%, 5%, or 10%.
• noisy positions with normal medial variant allele frequency greater than 0.5% are removed.
• Outlier samples were iteratively removed from the model to account for noise and contamination.
• samples with a Z score of greater than 5, 6, 7, 8, 9, or 10 are removed from the data analysis.
• the depth of read weighted mean and standard deviation of the error are calculated.
• Tumor or cell-free plasma samples' positions with at least 5 variant reads and a Z-score of 10 against the background error model for example, can be called as a candidate mutation.
• SNV METHOD 2 Single Nucleotide Variants (SNVs) are determined using plasma ctDNA data.
• the PCR process is modeled as a stochastic process, estimating the parameters using a training set and making the final SNV calls for a separate testing set.
• the propagation of the error across multiple PCR cycles is determined, and the mean and the variance of the background error are calculated, and in illustrative embodiments, background error is differentiated from real mutations.
• Xif number of generation i reads with mutation e generated in the PCR cycle j [0099] Moreover, in addition to normal molecules Xo, if there are additional f e Xo molecules with the mutation e at the beginning of the PCR process (hence fe/(l+fe) will be the fraction of mutated molecules in the initial mixture).
• SNV Method 2 is performed as follows:
• a confidence cutoff can be used to identify an SNV at an SNV loci. For example, a 90%, 95%, 96%, 97%, 98%, or 99% confidence cutoff can be used to call an SNV.
• the algorithm starts by estimating the efficiency and error rate per cycle using the training set. Let n denote the total number of PCR cycles. [0113] The number of reads Rb at each base b can be approximated by (1+pb) " Xo, where pb is the efficiency at base b. Then (Rb/ Xo) 1/n can be used to approximate 1+pb. Then, we can determine the mean and the standard variation of pb across all training samples, to estimate the parameters of the probability distribution (such as normal, beta, or similar distributions) for each base.
• the number of error e reads Rb e at each base b can be used to estimate p e .
• the mean and the standard deviation of the error rate across all training samples we approximate its probability distribution (such as normal, beta, or similar distributions) whose parameters are estimated using this mean and standard deviation values.
• Primer tails can improve the detection of fragmented DNA from universally tagged libraries. If the library tag and the primer-tails contain a homologous sequence, hybridization can be improved (for example, melting temperature (Tm) is lowered) and primers can be extended if only a portion of the primer target sequence is in the sample DNA fragment.
• Tm melting temperature
• 13 or more target specific base pairs may be used. In some embodiments, 10 to 12 target specific base pairs may be used. In some embodiments, 8 to 9 target specific base pairs may be used. In some embodiments, 6 to 7 target specific base pairs may be used.
• Libraries are generated from the samples above by ligating adaptors to the ends of DNA fragments in the samples, or to the ends of DNA fragments generated from DNA isolated from the samples.
• the fragments can then be amplified using PCR, for example, according to the following exemplary protocol:
• kits and methods are known in the art for generation of libraries of nucleic acids that include universal primer binding sites for subsequent amplification, for example clonal amplification, and for subsequence sequencing.
• library preparation and amplification can include end repair and adenylation (i.e. A-tailing).
• Kits especially adapted for preparing libraries from small nucleic acid fragments, especially circulating free DNA can be useful for practicing methods provided herein.
• the NEXTflex Cell Free kits available from Bioo Scientific () or the Natera Library Prep Kit (available from Natera, Inc. San Carlos, CA) .
• kits would typically be modified to include adaptors that are customized for the amplification and sequencing steps of the methods provided herein.
• Adaptor ligation can be performed using commercially available kits such as the ligation kit found in the AGILENT SURESELECT kit (Agilent, CA).
• Target regions of the nucleic acid library generated from DNA isolated from the sample, especially a circulating free DNA sample for the methods of the present invention are then amplified.
• a series of primers or primer pairs which can include between 5, 10, 15, 20, 25, 50, 100, 125, 150, 250, 500, 1000, 2500, 5000, 10,000, 20,000, 25,000, or 50,000 on the low end of the range and 15, 20, 25, 50, 100, 125, 150, 250, 500, 1000, 2500, 5000, 10,000, 20,000, 25,000, 50,000, 60,000, 75,000, or 100,000 primers on the upper end of the range, that each bind to one of a series of primer binding sites.
• Primer designs can be generated with Primer3 (Schgrasser A, Cutcutache I, Koressaar T, Ye J, Faircloth BC, Remm M, Rozen SG (2012) "Primer3 - new capabilities and interfaces.” Nucleic Acids Research 40(15):el l5 and Koressaar T, Remm M (2007) “Enhancements and modifications of primer design program Primer3.” Bioinformatics 23(10): 1289-91) source code available at primer3.sourceforge.net). Primer specificity can be evaluated by BLAST and added to existing primer design pipeline criteria:
• Primer specificities can be determined using the BLASTn program from the ncbi-blast- 2.2.29+ package.
• the task option “blastn-short” can be used to map the primers against hgl9 human genome.
• Primer designs can be determined as "specific” if the primer has less than 100 hits to the genome and the top hit is the target complementary primer binding region of the genome and is at least two scores higher than other hits (score is defined by BLASTn program). This can be done in order to have a unique hit to the genome and to not have many other hits throughout the genome.
• the final selected primers can be visualized in IGV (James T. Robinson, Helga Thorvaldsdottir, Wendy Winckler, Mitchell Guttman, Eric S. Lander, Gad Getz, Jill P. Mesirov. Integrative Genomics Viewer. Nature Biotechnology 29, 24-26 (2011)) and UCSC browser (Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, Haussler D. The human genome browser at UCSC. Genome Res. 2002 Jun;12(6):996-1006 ) using bed files and coverage maps for validation.
• Methods of the present invention include forming an amplification reaction mixture.
• the reaction mixture typically is formed by combining a polymerase, nucleotide triphosphates, nucleic acid fragments from a nucleic acid library generated from the sample, a set of forward and reverse primers specific for target regions that contain SNVs.
• An amplification reaction mixture useful for the present invention includes components known in the art for nucleic acid amplification, especially for PCR amplification.
• the reaction mixture typically includes nucleotide triphosphates, a polymerase, and magnesium.
• Polymerases that are useful for the present invention can include any polymerase that can be used in an amplification reaction especially those that are useful in PCR reactions. In certain embodiments, hot start Taq polymerases are especially useful.
• Amplification reaction mixtures useful for practicing the methods provided herein, such as AmpliTaq Gold master mix (Life Technologies, Carlsbad, CA), are available commercially.
• Amplification (e.g. temperature cycling) conditions for PCR are well known in the art.
• the methods provided herein can include any PCR cycling conditions that result in amplification of target nucleic acids such as target nucleic acids from a library.
• Non-limiting exemplary cycling conditions are provided in the Examples section herein.
• At least a portion and in illustrative examples the entire sequence of an amplicon, such as an outer primer target amplicon, is determined.
• Methods for determining the sequence of an amplicon are known in the art. Any of the sequencing methods known in the art, e.g. Sanger sequencing, can be used for such sequence determination.
• next-generation sequencing techniques also referred to herein as massively parallel sequencing techniques
• MYSEQ ILLUMINA
• HISEQ ILLUMINA
• ION TORRENT LIFE TECHNOLOGIES
• GENOME ANALYZER ILX ILLUMINA
• GS FLEX+ ROCHE 454
• High throughput genetic sequencers are amenable to the use of barcoding (i.e., sample tagging with distinctive nucleic acid sequences) so as to identify specific samples from individuals thereby permitting the simultaneous analysis of multiple samples in a single run of the DNA sequencer.
• barcoding i.e., sample tagging with distinctive nucleic acid sequences
• the number of times a given region of the genome in a library preparation (or other nucleic preparation of interest) is sequenced (number of reads) will be proportional to the number of copies of that sequence in the genome of interest (or expression level in the case of cDNA containing preparations). Biases in amplification efficiency can be taken into account in such quantitative determination.
• Target genes of the present invention are cancer-related genes, and in many illustrative embodiments, lung cancer-related genes.
• a cancer-related gene refers to a gene associated with an altered risk for a cancer (e.g. lung cancer or lung SCC or lung ADC, respectively) or an altered prognosis for a cancer.
• Exemplary cancer-related genes that promote cancer include oncogenes; genes that enhance cell proliferation, invasion, or metastasis; genes that inhibit apoptosis; and pro-angiogenesis genes.
• Cancer-related genes that inhibit cancer include, but are not limited to, tumor suppressor genes; genes that inhibit cell proliferation, invasion, or metastasis; genes that promote apoptosis; and anti-angiogenesis genes.
• An embodiment of the mutation detection method begins with the selection of the region of the gene that becomes the target. The region with known mutations is used to develop primers for mPCR-NGS to amplify and detect the mutation.
• Methods provided herein can be used to detect virtually any type of mutation, especially mutations known to be associated with cancer and most particularly the methods provided herein are directed to mutations, especially SNVs, associated with lung cancer, specifically adenocarcinoma and squamous cell carcinoma.
• Exemplary SNVs can be in one or more of the following genes: EGFR, FGFR1, FGFR2, ALK, MET, ROS 1, NTRK1, RET, HER2, DDR2, PDGFRA, KRAS, NF1, BRAF, PIK3CA, MEK1, NOTCH1, MLL2, EZH2, TET2, DNMT3A, SOX2, MYC, KEAP1, CDKN2A, NRG1, TP53, LKB 1, and PTEN, which have been identified in various lung cancer samples as being mutated, having increased copy numbers, or being fused to other genes and combinations thereof (Non-small-cell lung cancers: a heterogeneous set of diseases. Chen et al. Nat. Rev. Cancer.
• the list of genes are those listed above, where SNVs have been reported, such as in the cited Chen et al. reference.
• the SNVs can include SNVs found in one of the genes found in Table 19 herein. SNVs in the genes listed in Table 19 were analyzed in the experiment of Example 1. SNVs in these genes were detected in tumor samples matched to the ctDNA samples of Example 1.
• SNVs that are analyzed in methods provided herein can include any of the genes listed in this paragraph above or any of the genes in Table 19 that are not listed above. Provided herein, are methods that use the specific determination of a particular SNV in a particular gene to direct a targeted drug therapy.
• Amplification e.g. PCR
• Reaction Mixtures
• Methods of the present invention include forming an amplification reaction mixture.
• the reaction mixture typically is formed by combining a polymerase, nucleotide triphosphates, nucleic acid fragments from a nucleic acid library generated from the sample, a series of forward target- specific outer primers and a first strand reverse outer universal primer.
• Another illustrative embodiment is a reaction mixture that includes forward target- specific inner primers instead of the forward target- specific outer primers and amplicons from a first PCR reaction using the outer primers, instead of nucleic acid fragments from the nucleic acid library.
• the reaction mixtures are PCR reaction mixtures.
• PCR reaction mixtures typically include magnesium.
• the reaction mixture includes ethylenediaminetetraacetic acid (EDTA), magnesium, tetramethyl ammonium chloride (TMAC), or any combination thereof.
• EDTA ethylenediaminetetraacetic acid
• TMAC tetramethyl ammonium chloride
• the concentration of TMAC is between 20 and 70 mM, inclusive. While not meant to be bound to any particular theory, it is believed that TMAC binds to DNA, stabilizes duplexes, increases primer specificity, and/or equalizes the melting temperatures of different primers. In some embodiments, TMAC increases the uniformity in the amount of amplified products for the different targets.
• the concentration of magnesium (such as magnesium from magnesium chloride) is between 1 and 8 mM.
• the large number of primers used for multiplex PCR of a large number of targets may chelate a lot of the magnesium (2 phosphates in the primers chelate 1 magnesium). For example, if enough primers are used such that the concentration of phosphate from the primers is ⁇ 9 mM, then the primers may reduce the effective magnesium concentration by -4.5 mM.
• EDTA is used to decrease the amount of magnesium available as a cofactor for the polymerase since high concentrations of magnesium can result in PCR errors, such as amplification of non-target loci. In some embodiments, the concentration of EDTA reduces the amount of available magnesium to between 1 and 5 mM (such as between 3 and 5 mM).
• the pH is between 7.5 and 8.5, such as between 7.5 and 8, 8 and 8.3, or 8.3 and 8.5, inclusive.
• Tris is used at, for example, a concentration of between 10 and 100 mM, such as between 10 and 25 mM, 25 and 50 mM, 50 and 75 mM, or 25 and 75 mM, inclusive. In some embodiments, any of these concentrations of Tris are used at a pH between 7.5 and 8.5.
• a combination of KCl and (NH4)2SC1 ⁇ 4 is used, such as between 50 and 150 mM KCl and between 10 and 90 mM (NH 4 )2S04, inclusive.
• the concentration of KCl is between 0 and 30 mM, between 50 and 100 mM, or between 100 and 150 mM, inclusive.
• the concentration of (NH 4 )2S0 4 is between 10 and 50 mM, 50 and 90 mM, 10 and 20 mM, 20 and 40 mM, 40 and 60 mM, or 60 and 80 mM (NH 4 ) 2 S0 4 , inclusive.
• the ammonium [NH 4 + ] concentration is between 0 and 160 mM, such as between 0 to 50, 50 to 100, or 100 to 160 mM, inclusive.
• the sum of the potassium and ammonium concentration ([K + ] + [NH 4 + ]) is between 0 and 160 mM, such as between 0 to 25, 25 to 50, 50 to 150, 50 to 75, 75 to 100, 100 to 125, or 125 to 160 niM, inclusive.
• An exemplary buffer with [K + ] + [NH 4 + ] 120 mM is 20 mM KC1 and 50 mM (NH 4 ) 2 S0 4 .
• the buffer includes 25 to 75 mM Tris, pH 7.2 to 8, 0 to 50 mM KC1, 10 to 80 mM ammonium sulfate, and 3 to 6 mM magnesium, inclusive.
• the buffer includes 25 to 75 mM Tris pH 7 to 8.5, 3 to 6 mM MgCl 2 , 10 to 50 mM KC1, and 20 to 80 mM (NH 4 ) 2 S0 4 , inclusive. In some embodiments, 100 to 200 Units/mL of polymerase are used. In some embodiments, 100 mM KC1, 50 mM (NH 4 ) 2 S0 4 , 3 mM MgCl 2 , 7.5 nM of each primer in the library, 50 mM TMAC, and 7 ul DNA template in a 20 ul final volume at pH 8.1 is used.
• a crowding agent such as polyethylene glycol (PEG, such as PEG 8,000) or glycerol.
• PEG polyethylene glycol
• glycerol the amount of PEG (such as PEG 8,000) is between 0.1 to 20%, such as between 0.5 to 15%, 1 to 10%, 2 to 8%, or 4 to 8%, inclusive.
• the amount of glycerol is between 0.1 to 20%, such as between 0.5 to 15%, 1 to 10%, 2 to 8%, or 4 to 8%, inclusive.
• a crowding agent allows either a low polymerase concentration and/or a shorter annealing time to be used.
• a crowding agent improves the uniformity of the DOR and/or reduces dropouts (undetected alleles).
• Polymerases In some embodiments, a polymerase with proof-reading activity, a polymerase without (or with negligible) proof-reading activity, or a mixture of a polymerase with proof-reading activity and a polymerase without (or with negligible) proof-reading activity is used. In some embodiments, a hot start polymerase, a non-hot start polymerase, or a mixture of a hot start polymerase and a non-hot start polymerase is used. In some embodiments, a HotStarTaq DNA polymerase is used (see, for example, QIAGEN catalog No.
• AmpliTaq Gold® DNA Polymerase is used.
• a PrimeSTAR GXL DNA polymerase a high fidelity polymerase that provides efficient PCR amplification when there is excess template in the reaction mixture, and when amplifying long products, is used (Takara Clontech, Mountain View, CA).
• KAPA Taq DNA Polymerase or KAPA Taq HotStart DNA Polymerase is used; they are based on the single-subunit, wild-type Taq DNA polymerase of the thermophilic bacterium Thermus aquaticus.
• KAPA Taq and KAPA Taq HotStart DNA Polymerase have 5 '-3' polymerase and 5 '-3' exonuclease activities, but no 3' to 5' exonuclease (proofreading) activity (see, for example, KAPA BIOSYSTEMS catalog No. BK1000).
• Pfu DNA polymerase is used; it is a highly thermostable DNA polymerase from the hyperthermophilic archaeum Pyrococcus furiosus . The enzyme catalyzes the template-dependent polymerization of nucleotides into duplex DNA in the 5' ⁇ 3' direction.
• Pfu DNA Polymerase also exhibits 3' ⁇ 5' exonuclease (proofreading) activity that enables the polymerase to correct nucleotide incorporation errors. It has no 5' ⁇ 3' exonuclease activity (see, for example, Thermo Scientific catalog No. EP0501).
• Klentaql is used; it is a Klenow-fragment analog of Taq DNA polymerase, it has no exonuclease or endonuclease activity (see, for example, DNA POLYMERASE TECHNOLOGY, Inc, St. Louis, Missouri, catalog No. 100).
• the polymerase is a PHUSION DNA polymerase, such as PHUSION High Fidelity DNA polymerase (M0530S, New England BioLabs, Inc.) or PHUSION Hot Start Flex DNA polymerase (M0535S, New England BioLabs, Inc.).
• the polymerase is a Q5® DNA Polymerase, such as Q5® High-Fidelity DNA Polymerase (M0491S, New England BioLabs, Inc.) or Q5® Hot Start High-Fidelity DNA Polymerase (M0493S, New England BioLabs, Inc.).
• the polymerase is a T4 DNA polymerase (M0203S, New England BioLabs, Inc.).
• between 5 and 600 Units/mL (Units per 1 mL of reaction volume) of polymerase is used, such as between 5 to 100, 100 to 200, 200 to 300, 300 to 400, 400 to 500, or 500 to 600 Units/mL, inclusive.
• hot-start PCR is used to reduce or prevent polymerization prior to PCR thermocycling.
• Exemplary hot-start PCR methods include initial inhibition of the DNA polymerase, or physical separation of reaction components reaction until the reaction mixture reaches the higher temperatures.
• slow release of magnesium is used.
• DNA polymerase requires magnesium ions for activity, so the magnesium is chemically separated from the reaction by binding to a chemical compound, and is released into the solution only at high temperature.
• non-covalent binding of an inhibitor is used. In this method a peptide, antibody, or aptamer are non-covalently bound to the enzyme at low temperature and inhibit its activity. After incubation at elevated temperature, the inhibitor is released and the reaction starts.
• a cold-sensitive Taq polymerase such as a modified DNA polymerase with almost no activity at low temperature.
• chemical modification is used.
• a molecule is covalently bound to the side chain of an amino acid in the active site of the DNA polymerase. The molecule is released from the enzyme by incubation of the reaction mixture at elevated temperature. Once the molecule is released, the enzyme is activated.
• the amount to template nucleic acids (such as an RNA or DNA sample) is between 20 and 5,000 ng, such as between 20 to 200, 200 to 400, 400 to 600, 600 to 1,000; 1,000 to 1,500; or 2,000 to 3,000 ng, inclusive.
• a QIAGEN Multiplex PCR Kit is used (QIAGEN catalog No. 206143).
• the kit includes 2x QIAGEN Multiplex PCR Master Mix (providing a final concentration of 3 mM MgCl 2 , 3 x 0.85 ml), 5x Q-Solution (1 x 2.0 ml), and RNase-Free Water (2 x 1.7 ml).
• the QIAGEN Multiplex PCR Master Mix (MM) contains a combination of KC1 and (NH 4 ) 2 S04 as well as the PCR additive, Factor MP, which increases the local concentration of primers at the template.
• HotStarTaq DNA Polymerase is a modified form of Taq DNA polymerase and has no polymerase activity at ambient temperatures. In some embodiments, HotStarTaq DNA Polymerase is activated by a 15-minute incubation at 95 °C which can be incorporated into any existing thermal-cycler program.
• lx QIAGEN MM final concentration (the recommended concentration), 7.5 nM of each primer in the library, 50 mM TMAC, and 7 ul DNA template in a 20 ul final volume is used.
• the PCR thermocycling conditions include 95°C for 10 minutes (hot start); 20 cycles of 96°C for 30 seconds; 65°C for 15 minutes; and 72°C for 30 seconds; followed by 72°C for 2 minutes (final extension); and then a 4°C hold.
• 2x QIAGEN MM final concentration (twice the recommended concentration), 2 nM of each primer in the library, 70 mM TMAC, and 7 ul DNA template in a 20 ul total volume is used. In some embodiments, up to 4 mM EDTA is also included.
• the PCR thermocycling conditions include 95°C for 10 minutes (hot start); 25 cycles of 96°C for 30 seconds; 65°C for 20, 25, 30, 45, 60, 120, or 180 minutes; and optionally 72°C for 30 seconds); followed by 72°C for 2 minutes (final extension); and then a 4°C hold.
• Another exemplary set of conditions includes a semi-nested PCR approach.
• the first PCR reaction uses 20 ul a reaction volume with 2x QIAGEN MM final concentration, 1.875 nM of each primer in the library (outer forward and reverse primers), and DNA template.
• Thermocycling parameters include 95°C for 10 minutes; 25 cycles of 96°C for 30 seconds, 65°C for 1 minute, 58°C for 6 minutes, 60°C for 8 minutes, 65°C for 4 minutes, and 72°C for 30 seconds; and then 72°C for 2 minutes, and then a 4°C hold.
• 2 ul of the resulting product, diluted 1:200 is used as input in a second PCR reaction.
• This reaction uses a 10 ul reaction volume with lx QIAGEN MM final concentration, 20 nM of each inner forward primer, and 1 uM of reverse primer tag.
• Thermocycling parameters include 95°C for 10 minutes; 15 cycles of 95°C for 30 seconds, 65°C for 1 minute, 60°C for 5 minutes, 65°C for 5 minutes, and 72°C for 30 seconds; and then 72°C for 2 minutes, and then a 4°C hold.
• the annealing temperature can optionally be higher than the melting temperatures of some or all of the primers, as discussed herein (see U.S. Patent Application No. 14/918,544, filed Oct. 20, 2015, which is herein incorporated by reference in its entirety).
• the melting temperature (T m ) is the temperature at which one-half (50%) of a DNA duplex of an oligonucleotide (such as a primer) and its perfect complement dissociates and becomes single strand DNA.
• the annealing temperature (TA) is the temperature one runs the PCR protocol at. For prior methods, it is usually 5°C below the lowest T m of the primers used, thus close to all possible duplexes are formed (such that essentially all the primer molecules bind the template nucleic acid). While this is highly efficient, at lower temperatures there are more unspecific reactions bound to occur.
• the TA is higher than T m , where at a given moment only a small fraction of the targets have a primer annealed (such as only -1-5%). If these get extended, they are removed from the equilibrium of annealing and dissociating primers and target (as extension increases T m quickly to above 70°C), and a new -1-5% of targets has primers.
• T m the concentration of the targets in the range.
• the annealing temperature is between 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13 °C and 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, or 15 °C on the high end of the range, greater than the melting temperature (such as the empirically measured or calculated T m ) of at least 25, 50, 60, 70, 75, 80, 90, 95, or 100% of the non-identical primers.
• the annealing temperature is between 1 and 15 °C (such as between 1 to 10, 1 to 5, 1 to 3, 3 to 5, 5 to 10, 5 to 8, 8 to 10, 10 to 12, or 12 to 15 °C, inclusive) greater than the melting temperature (such as the empirically measured or calculated T m ) of at least 25; 50; 75; 100; 300; 500; 750; 1,000; 2,000; 5,000; 7,500; 10,000; 15,000; 19,000; 20,000; 25,000; 27,000; 28,000; 30,000; 40,000; 50,000; 75,000; 100,000; or all of the non-identical primers.
• the melting temperature such as the empirically measured or calculated T m
• the annealing temperature is between 1 and 15 °C (such as between 1 to 10, 1 to 5, 1 to 3, 3 to 5, 3 to 8, 5 to 10, 5 to 8, 8 to 10, 10 to 12, or 12 to 15 °C, inclusive) greater than the melting temperature (such as the empirically measured or calculated T m ) of at least 25%, 50%, 60%, 70%, 75%, 80%, 90%, 95%, or all of the non-identical primers, and the length of the annealing step (per PCR cycle) is between 5 and 180 minutes, such as 15 and 120 minutes, 15 and 60 minutes, 15 and 45 minutes, or 20 and 60 minutes, inclusive.
• long annealing times (as discussed herein and exemplified in Example 12) and/or low primer concentrations are used.
• limiting primer concentrations and/or conditions are used.
• the length of the annealing step is between 15, 20, 25, 30, 35, 40, 45, or 60 minutes on the low end of the range and 20, 25, 30, 35, 40, 45, 60, 120, or 180 minutes on the high end of the range.
• the length of the annealing step (per PCR cycle) is between 30 and 180 minutes.
• the annealing step can be between 30 and 60 minutes and the concentration of each primer can be less than 20, 15, 10, or 5 nM.
• the primer concentration is 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, or 25 nM on the low end of the range, and 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, and 50 on the high end of the range.
• the solution may become viscous due to the large amount of primers in solution. If the solution is too viscous, one can reduce the primer concentration to an amount that is still sufficient for the primers to bind the template DNA. In various embodiments, between 1,000 and 100,000 different primers are used and the concentration of each primer is less than 20 nM, such as less than 10 nM or between 1 and 10 nM, inclusive.
• EXAMPLE 1 Analysis of single nucleotide variants (SNVs) in circulating tumor DNA (ctDNA) from Lung Cancer Patients
• a prior pilot study demonstrated the successful detection of cancer-relevant point mutations in the plasma of cancer patients.
• the mutation profile of 4 lung cancer tumors was determined by whole exome sequencing (WES) or Ampliseq (Life Technologies, Carlsbad, CA), and a subset of those mutations were successfully detected in the corresponding plasma samples using a multiplex PCR-Next-Generation Sequencing (mPCR-NGS) method.
• WES whole exome sequencing
• Ampliseq Life Technologies, Carlsbad, CA
• mPCR-NGS multiplex PCR-Next-Generation Sequencing
• TRACERx the mPCR-NGS method was used to detect and track over time cancer-specific mutations in the plasma of cancer patients, and to evaluate the utility of the method in monitoring disease progression through treatment.
• the overall project design is shown in FIG. 1.
• the first phase of the project was the determination of the baseline mutation profile in the plasma of 50 treatment-naive lung cancer patients.
• Purified genomic DNA samples from several tumor regions (2-7 regions per tumor), purified germline DNA samples, and intact plasma samples from 50 patients were obtained.
• the mutation profile of all of the tumor regions was previously determined by WES and AmpliSeq, and a subset of mutations per patient was analyzed by mPCR- NGS. Those mutations included both driver and passenger mutations and both clonal and sub- clonal mutations.
• we designed multiplex PCR assays prepared primer pools (primers were obtained from IDT, Coralville, Iowa), QC'ed the primer pools, and optimized the mPCR protocol for each pool.
• Plasma cfDNA was purified, quantified, and converted into libraries. The libraries were then used as input into mPCR, and the products were sequenced and analyzed. A similar protocol was applied to the genomic DNA from tumor and matched normal samples.
• SNV information The mutation profile, including single nucleotide variants (SNVs) and copy number variants (CNVs), of each tumor subsection was determined by TRACERx using WES. The full mutation profile of each tumor was used to detect clonal structure and to reconstruct the phylogenetic tree of each tumor. PyClone (PyClone: statistical inference of clonal population structure in cancer. Roth et al, Nature Methods 11, 396-398 (2014)) was used to detect clonal structure. PyClone identifies a list of SNV subclones and calculates their cancer cell fraction. It also categorizes SNVs as either clonal or subclonal.
• the driver category of each SNV was determined and provided as the driver category (1-4, where 1 is most likely to be a driver mutation, and 4 is the least likely). For each patient, up to 108 SNVs, spanning all driver categories and including clonal and subclonal mutations, were analyzed. The detected allele fractions of each SNV in each tumor subsection, lymph node and matched normal DNA sample along with PyClone clonal/subclonal cluster information were compared.
• tumor size mm
• tumor location lung lobe
• tumor stage tumor pathological type
• number of lymph nodes affected vascular invasion status, as well as de-identified information on the collecting hospital.
• Assay design Natera's standard assay design pipeline was used to design Right and Left PCR primers for all given SNVs. A pair of Right and Left PCR primers targeting an SNV is defined as an assay for that particular SNV. Note that it is possible for one assay to cover more than 1 target SNV if they are in close proximity. For every pair of assays, the probability of forming a primer-dimer was calculated. The SNV allele fraction data in each tumor was used to reconstruct phylogenetic trees using Lichee (Fast and scalable inference of multi-sample cancer lineages. Popic et al. Genome Biol. 2015 May 6;16:91).
• the list of assays for each sample were filtered to remove primers that are predicted to form primer dimers while giving strong priority to assays covering driver 1 and 2 SNVs.
• the remaining assays were used to build 5 balanced pools. All assays pooled together were compatible meaning there were no primers predicted to form primer- dimers in a pool.
• the assays were chosen such that assays covering driver 1 and 2 SNVs have the highest priority and for each patient the number of selected SNVs per branch was proportional to the total number of SNVs of that branch from the reconstructed phylogenetic tree.
• cfDNA sometimes contains an intact DNA fraction that overlaps with the high size marker on the chip, which makes quantification of the mononucleosomal peak unreliable.
• a representative subset of the purified genomic DNA samples was quantified using Nanodrops (Wilmington, DE). All of the samples quantified were in the expected range (-10 ng/ ⁇ ).
• cfDNA library preparation The entire cfDNA amount from each plasma sample was used as input into Library Prep using the Natera library prep kit and following the kit instructions. For two samples with extremely high cfDNA amounts, the input amount into Library Prep was restricted to -50,000 genome equivalents (165 ng). The libraries were amplified to plateau and then purified using Ampure beads (Beckman Coulter, Brea, CA) following the manufacturer's protocol. The purified libraries were QCed on the LabChip.
• cfDNA multiplex PCR and Sequencing The library material from each plasma sample was used as input into multiplex PCR (mPCR) using the relevant assay pool and an optimized plasma mPCR protocol.
• the protocol utilized an annealing time of 15 minutes at a temperature of 60C or 62.5C, which was above the Tm of the primers.
• the Tms of the primers using theoretical calculations was 53 to 59C.
• a ⁇ primer concentration was used.
• the mPCR products were barcoded in a separate PCR step, and the barcoded PCR products were pooled according to the assay pooling information (see section above) into 5 pools.
• the pools were purified using Ampure beads following the manufacturer's protocol, QCed on a Bioanalyzer DNA1000 chip (Agilent, Santa Clara, CA), and quantified using the Qubit dsDNA Broad Range kit (Thermo Fisher Scientific, Waltham, MA).
• Each pool contained libraries prepared as disclosed above, from 10 cancer patient plasma samples and 20 negative controls (prepared from cfDNA extracted from presumed healthy volunteers). The negative control samples were obtained following the necessary regulatory procedures.
• Each pool was sequenced on a separate HiSeq 2500 Rapid run (Illumina, San Diego, CA) with 50 cycle paired end single index reads.
• gDNA multiplex PCR and sequencing The genomic DNA samples were used as input into a similar mPCR using the relevant assay pools and an optimized genomic mPCR protocol. The mPCR products were barcoded in a separate PCR step, and all the barcoded products were combined into one pool. The pool was purified using Ampure beads following the manufacturer's protocol, QCed on a Bioanalyzer DNA1000 chip, and quantified using the Qubit dsDNA Broad Range kit. The pool was sequenced on a single HiSeq2500 Rapid run with 50 cycle single end single index reads.
• FIG. 20 is a table showing detailed results of the analysis and detailed information regarding the samples that were analyzed in this study.
• cfDNA extraction and analysis The distribution of cfDNA concentrations for the 50 plasma samples ( Figure 3) followed the expected distribution based on 5 ml of plasma (median of 2,200 genome copy equivalents per ml of plasma). The cfDNA concentrations, the hemolysis grade (visually estimated) and the qualitative evaluation of the cfDNA size profile (visually estimated from the Bioanalyzer traces) are shown in tabular form in FIG. 16
• cfDNA analysis The purified cfDNA concentration, plasma hemolysis grade and cfDNA profile are shown in Figure 16.
• cfDNA concentration refers to the mononucleosomal peak only, and was determined from the mononucleosomal peak height using a calibration curve. Genome copy equivalents were calculated using a 3.3 pg/genome conversion factor; 40 ⁇ purified cfDNA is used as input into Library Prep; green highlights: for those samples, the input into library prep was restricted to 50,000 genome equivalents.
• cfDNA size profile 1: most of the cfDNA is in the mononucleosomal peak; 2: most of the cfDNA is in the mononucleosomal peak, but other sizes are seen; 3: a large peak of intact DNA (> 1,000 bp) is seen along with the mononucleosomal peak and some higher molecular weight peaks. Hemolysis was estimated visually based on the plasma color. 0: no hemolysis (yellow plasma); 1 : mild hemolysis (faint pink plasma); 2: severe hemolysis (bright pink or red plasma).
• VAF analysis in tumor subsections The sequence data from each of the tumor subsections was analyzed to determine the variant allele frequency (VAF) of each SNV in each tumor subsection, lymph node and matched normal sample. This data was compared with matched data provided separately from a different test site using different test methods, such as whole genome sequencing and exome sequencing. For most samples, the previously determined tissue VAF values from each tumor subsection closely matched the newly derived tissue VAF values ( Figure 4). However, there were a large number of samples in which significant discrepancies were seen ( Figure 5).
• Each run belonged to one assay pool and contained 10 cancer samples as well as 20 control samples.
• the set of SNVs covered by assays in a pool are considered as target SNVs for the associated run.
• To make an SNV call at a specific position of a cancer sample first a background error model for that position was built. The error model was constructed based on the 20 negative samples and the remaining cancer samples (8 or 9) that were not expected to contain an SNV at that position, based on the information provided. Positions with VAF > 20% were excluded from the background error model.
• a positive plasma SNV call was made if the confidence for that mutation in the corresponding plasma sample passed our confidence threshold of 95 to 98%.
• the overall SNV detection rate in plasma is 35.5% (310 out of 890), similar to a prior pilot study. While the algorithm made most confident true positive calls, the number of false positive calls are at an acceptable number ( ⁇ 0.25%).
• the average mutant allele frequency for the SNVs detected with high confidence is 0.875%, ranging from 0.011% to 13.93%.
• a sample was considered as 'detected in plasma' if at least one SNV expected to be present in that sample was confidently detected in plasma.
• the overall sample detection rate in plasma was 69% (34 out of 49 samples), and for those, and the average number of SNVs detected in plasma was 9.1 (ranging from 1 to 19).
• the number of SNVs detected in plasma for each sample is shown in tabular form in FIG. 17.
• the number of mutant reads is almost negligible. In fact, 36% of them have 0 mutant reads, 75% of them have more than 5 mutant reads, and the remaining 25% false negative calls have VAF ⁇ .1%.
• Tumor stage and size were some of the most important factors identified that influence the number of SNVs detected in the corresponding plasma sample ( Figure 8). Stage la tumors had the lowest chance of having at least one SNV detected, as well as the lowest success rate of detecting SNVs in the plasma. The VAF distribution for the SNVs that were detected from stage la tumors was also lower than for the rest of the tumors ( Figure 9). As tumor size and stage are correlated, a similar trend was seen with tumor size. As this was not due to assay failure or sensitivity limits (see below), the most likely explanation is that such tumors tend to not have cfDNA present in the plasma in quantities that are detectable in the plasma volumes used in this study.
• the ADC samples were more dependent on these factors with a general trend of far fewer SNVs detected in ctDNA than were detected in ctDNA of the SCC samples.
• SNVs were detected in the ctDNA of all of the SQCC samples regardless of stage: Three SNVs were detected in the ctDNA of one of the SCC samples and at least 5 SNVs detected in the ctDNA of the remainder of the SCC samples ( Figure 15). In fact, between 3 and 19 SNVs were detected in the ctDNA of SCC samples.
• stage la In 6 ADC samples that were stage la, an SNV was only detected in one of the ctDNA samples, and in that sample only a single SNV was detected. In none of the stage la ADC samples were more than 1 SNV detected in the ctDNA. In stage lb ADC samples, less than 5 SNVs were identified in all but two samples, with 7 SNVs identified in one of the stage lb ADC samples and 18 SNVs identified in one of the stage lb samples.
• Tumor VAF and clonality The clonality ratio was calculated for each mutation as (number of sub-sections of a tumor where the mutation is detected) / (total number of sub-sections of that tumor analyzed). Mutations that were observed in all analyzed tumor sections are considered 'clonal', all others are considered 'sub-clonal'.
• the VAF of SNVs detected in plasma correlates with the 'clonality' of the mutations, with more clonal mutations being responsible for the highest plasma VAF values ( Figures 9 and 12); similarly, SNVs present in multiple tumor sub-sections tend to be responsible for higher plasma VAFs in the corresponding plasma samples.
• the clonal status of each SNV was categorized by PyCloneCluster based on WES data from the tumor tissue. Clonal SNVs tended to have higher VAFs ( Figures 13 and 14).
• cfDNA input and tumor VAF There was no correlation between the amount of cfDNA and the number and proportion of SNVs detected in the plasma samples. The number of SNVs detected in plasma is not predicted by the cfDNA input amount; however, all samples with high input (>25,000 copies) had at least one SNV detected in plasma ( Figure 10). The plasma SNV VAF also correlates with the tumor SNV VAF( Figure 11).
• Multivariate analysis A regression analysis was performed to determine the variables that can be used to predict our detection of mutations. More specifically, a 0/1 response variable was used to annotate the mutations we called as present or not. The following independent variables were included in our model:
• EXAMPLE 2 A bespoke multiplex PCR protocol to track tumor mutations in plasma
• Natera's bespoke multiplex PCR (mPCR) protocol is designed to estimate plasma ctDNA level by tracking a set of patient- specific mutations identified from tumor tissue sample(s). Given a patient-specific mutation profile we designed custom mPCR panels that can be applied to time- series plasma samples of the corresponding patient.
• SNV targets The mutation profile, which included single nucleotide variants (SNVs) for each tumor subsection, was determined based on analyses of tumor sequencing. The full mutation profile of each tumor was used to reconstruct the phylogenetic tree of each tumor. PyClone (Roth, et al. (2014). PyClone: Statistical inference of clonal population structure in cancer. Nature Methods 11 : 396-398) was used to identify clusters of SNVs, and calculate their cancer-cell fraction. This was used to categorize SNVs as either clonal or subclonal. The driver category of each SNV was determined (1-4, where 1 was most likely to be a driver mutation, and 4 was the least likely).
• PCR conditions that yield the best percentage of on-target reads, depth-of-read uniformity, and the lowest number of drop-outs were determined. For all pools, the optimal conditions were 10 nM primers and 60°C or 62.5°C annealing temperature.
• cfDNA was extracted at Natera using the Qiagen NA kit following a protocol optimized for 5 ml of plasma. All cfDNA samples were QCed on Bioanalyzer High Sensitivity chips. The same Bioanalyzer High Sensitivity runs were also used to also quantify the cfDNA samples by interpolation of the mononucleosomal peak height on a calibration curve prepared from a pure cfDNA sample that was quantified previously. This is necessary because cfDNA sometimes contains an intact DNA fraction that overlaps with the high size marker on the chip, making quantification of the mononucleosomal peak unreliable.
• Genomic DNA samples (from tumor subsections, lymph nodes, and white blood cells) were quantified on the Nanodrop.
• cfDNA library preparation The entire cfDNA amount from each plasma sample was used as input into Library Prep using the Natera library prep kit and following the kit instructions. For two samples with extremely high cfDNA amounts, the input amount into Library Prep was restricted to -50,000 genome equivalents (165 ng). In brief, 40 ul of DNA extracted from plasma, which is present in fragments of mononucleosomal and polynucleosomal length, were end repaired and A-tailed, and Natera custom adapters ligated. The libraries were amplified for 15 cycles to plateau and then purified using Ampure beads following the manufacturer's protocol. The purified libraries were QCed on the LabChip.
• cfDNA multiplex PCR and Sequencing The library material from each plasma sample was used as input into multiplex PCR using the relevant assay pool and an optimized plasma mPCR protocol.
• the PCR composition was: lx in-house PCR master mix, 10 nM primers, 3 uL cfDNA library (corresponding to -600 ng DNA), in 10 uL total reaction volume.
• the thermocycling conditions were: 95°C, 10 minutes; 10 cycles of (95°C, 30 seconds; 60°C or 62.5°C, 15 minutes; 72°C, 30 seconds); 72°C, 2 minutes, 4°C hold.
• the mPCR products were barcoded in a separate PCR step, and the barcoded PCR products were pooled according to the assay pooling information.
• the pools were purified using Ampure beads following the manufacturer's protocol, QCed on a Bioanalyzer DNA1000 chip, and quantified using the Qubit dsDNA Broad Range kit.
• Each pool contained barcoded mPCR products of 10 cancer plasma libraries and 20 negative controls (prepared from cfDNA extracted from healthy volunteers). The negative control samples were obtained following the necessary regulatory procedures.
• Each pool was sequenced on a separate HiSeq2500 Rapid runs with 50 cycle paired end single index reads.
• Genomic DNA multiplex PCR and sequencing The genomic DNA samples (gDNA) were used as input into a similar mPCR using the relevant assay pools and an optimized genomic mPCR protocol; 50 ng gDNA was used as input.
• the mPCR products were barcoded in a separate PCR step, and all the barcoded products were combined into one pool.
• the pool was purified using Ampure beads following the manufacturer's protocol, QCed on a Bioanalyzer DNA1000 chip, and quantified using the Qubit dsDNA Broad Range kit.
• the pool was sequenced on a single HiSeq2500 Rapid run with 50 cycle single end single index reads.
• Targets with less than 1000 reads were considered failed and were filtered from further analyses.
• Quality control (QC) was performed using an in-house Java program checking for a wide list of statistics per sample that included total numbers of reads, mapped reads, on-target reads, number of failed targets, and average error rate. A sample with less than 90% mapped reads and more than 3 failed targets did not pass, and needed to be resequenced.
• Error rate (p e ) Error rate per cycle for mutation type e (e.g wildtype allele A to mutant allele G).
• the target-specific error propagation model was used to characterize the distribution of error molecules. As a molecule is replicated over the course of PCR process, more errors occur. If an error occurs in cycle i and there are Xj wildtype molecules in the system, that error molecule is duplicated in next cycle with probability p and new error molecules are produced from wildtype background molecules according to a binomial process B(Xj, p e ). Using a recursive relation, we computed the mean and variance of number of total molecules X n and number of error molecules E n after n PCR cycles as shown in FIG. 21.
• step d Use the mean and variance estimated in step d to compute the likelihood L(0) for each potential real mutant fraction. Select the value of ⁇ that maximizes this likelihood, (denoted by 6 M LE) an d compute the confidence score ( as ' ( ⁇ m ⁇ e) ).

Landscapes

• Chemical & Material Sciences (AREA)
• Life Sciences & Earth Sciences (AREA)
• Organic Chemistry (AREA)
• Health & Medical Sciences (AREA)
• Proteomics, Peptides & Aminoacids (AREA)
• Zoology (AREA)
• Wood Science & Technology (AREA)
• Engineering & Computer Science (AREA)
• Analytical Chemistry (AREA)
• Genetics & Genomics (AREA)
• Immunology (AREA)
• Molecular Biology (AREA)
• General Engineering & Computer Science (AREA)
• Biotechnology (AREA)
• General Health & Medical Sciences (AREA)
• Microbiology (AREA)
• Biochemistry (AREA)
• Bioinformatics & Cheminformatics (AREA)
• Physics & Mathematics (AREA)
• Biophysics (AREA)
• Pathology (AREA)
• Chemical Kinetics & Catalysis (AREA)
• Hospice & Palliative Care (AREA)
• Oncology (AREA)
• Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
• Investigating Or Analysing Biological Materials (AREA)

Priority Applications (10)

Applications Claiming Priority (2)

Publications (2)

Family

ID=59215913

Family Applications (1)

Country Status (10)

Cited By (47)

Families Citing this family (9)

Family Cites Families (622)