Methods for detecting and quantifying cancer nucleic acid

Abstract

Methods of detecting and optionally quantifying cancer nucleic acid in patient samples. The methods can include identifying one or more structural variants in cancer nucleic acid relative to non-cancer nucleic acid, designing primers that amplify at least one of the one or more structural variants in the cancer nucleic acid but lack a defined amplifiable target in the non-cancer nucleic acid, amplifying a test sample with the primers, and analyzing the amplification product for the presence of amplicons to thereby detect cancer nucleic acid in the test sample.

Description

[0001] METHODS FOR DETECTING AND QUANTIFYING CANCER NUCLEIC ACID

[0002] STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0003] This invention was made with government support under CA243078 and CA234306 awarded by the National Institutes of Health. The government has certain rights in the invention.

[0004] FIELD OF THE INVENTION

[0005] The invention is directed to detecting and optionally quantifying cancer nucleic acid in patient samples, such as detecting and optionally quantifying ctDNA from cancer patients with residual disease.

[0006] BACKGROUND

[0007] A small fraction of cancer patients treated with curative intent experience disease recurrence. There are few effective biomarkers currently available to detect minimal residual disease (MRD) after completion of therapy. Circulating tumor DNA (ctDNA) holds promise as a biomarker to detect tumor-derived DNA in blood from patients with MRD. Most MRD assays rely on prior identification of patient-specific single nucleotide variants (SNV) in DNA from the tumor biopsy, and their subsequent detection in plasma DNA. Due to polymerase errors during amplification and sequencing noise, confidence in ctDNA detection at and below 0.01 % is limited. Alternative methods for detecting and quantifying nucleic acid in patient samples, such as detecting and quantifying ctDNA from cancer patients with residual disease, are needed.

[0008] SUMMARY OF THE INVENTION

[0009] One aspect of the invention is directed to methods of detecting and optionally quantifying cancer nucleic acid in a test sample from a subject. In some versions, the methods comprise identifying one or more structural variants in the cancer nucleic acid from the subject relative to non-cancer nucleic acid. In some versions, the methods comprise providing an assay primer set comprising structural-variant primers, wherein the structural-variant primers comprise one or more structural-variant primer pairs, wherein each structural-variant primer pairs is configured to amplify at least one of the one or more structural variants in the cancer nucleic acid but lack a defined amplifiable target in the non-cancer nucleic acid. In some versions, the methods comprise amplify ing the test sample with the assay primer set to yield test amplification product. In some versions, the methods comprise analyzing the test amplification product for presence of testsample structural-variant amplicons from the structural-variant primers to thereby detect any cancer nucleic acid in the test sample. In some versions, the assay pnmer set compnses at least two structural-variant primer pairs. In some versions, the assay primer set comprises up to 80 structural-variant primer pairs. In some versions, the assay primer set comprises at least two structural-variant primer pairs configured to amplify different structural variants. In some versions, the assay primer set comprises at least 10 structural-variant primer pairs configured to amplify different structural variants.

[0010] In some versions, the identify ing one or more structural variants comprises sequencing the cancer nucleic acid from a cancer sample from the subject and / or sequencing the non-cancer nucleic acid from a non-cancer sample from the subject.

[0011] In some versions, the assay primer set further comprises control primers, wherein the control primers comprise one or more control primer pairs, wherein each control primer pair is configured to amplify a control target present in both the cancer nucleic acid and the non-cancer nucleic acid. In some versions, the assay primer set comprises at least two control primer pairs. In some versions, the assay primer set comprises up to 10 control primer pairs. In some versions, the assay primer set comprises at least two control primer pairs configured to amplify different control sequences. In some versions, the assay primer set comprises at least four control primer pairs configured to amplify' different control sequences.

[0012] In some versions, the methods comprise amplifying a cancer sample from the subject with the assay primer set to generate cancer-sample structural-variant amplicons from the structural- variant primers and cancer-sample control amplicons from the control primers. In some versions, the methods comprise determining a cancer-sample structural-variant amount corresponding to each of one or more of the structural-variant primer pairs, wherein each cancer-sample structural- variant amount is an amount of the cancer-sample structural -variant amplicons amplified from the cancer sample with a corresponding one of the one or more structural-variant primer pairs. In some versions, the methods comprise determining a cancer-sample normalization amount, wherein the cancer-sample normalization amount is an amount of the cancer-sample control amplicons amplified from the cancer sample with corresponding one or more, respectively, of the control primer pairs. In some versions, the methods comprise normalizing each cancer-sample structural-variant amount to the cancer-sample normalization amount to thereby obtain a normalization factor corresponding to each of the one or more structural -variant primer pairs.

[0013] In some versions, the methods comprise determining a test-sample structural-variant amount corresponding to each of the one or more of the structural-variant primer pairs, wherein each test-sample structural -variant amount is an amount of test-sample structural-variant amplicons amplified from the test sample w ith a corresponding one of the one or more structural- variant primer pairs. In some versions, the methods comprise normalizing each test-sample structural-variant amount to its corresponding normalization factor to thereby obtain a normalization-factor-adjusted test-sample structural -variant amount corresponding to each testsample structural-variant amount.

[0014] In some versions, the methods comprise determining a cancer-sample control amount corresponding to each of one or more of the control pnmer pairs, wherein each cancer-sample control amount is an amount of the cancer-sample control amplicons amplified from the cancer sample with a corresponding one of the one or more control primer pairs. In some versions, the methods comprise normalizing each cancer-sample control amount to the cancer-sample normalization amount to thereby obtain a normalization factor corresponding to each of the one or more control primer pairs. In some versions, the methods comprise determining a test-sample control amount corresponding to each of the one or more of the control primer pairs, wherein each test-sample control amount is an amount of test-sample control amplicons amplified from the test sample with a corresponding one of the one or more control primer pairs. In some versions, the methods comprise normalizing each test-sample control amount to its corresponding normalization factor to thereby obtain a normalization -factor-adjusted test-sample control amount corresponding to each test-sample control amount. In some versions, the methods comprise normalizing one or more of the normalization-factor-adjusted test-sample structural -variant amounts to the normalization-factor-adjusted test-sample control amount to thereby obtain a cancer-nucleic-acid fraction in the test sample.

[0015] In some versions, the amplifying the cancer sample with the assay primer set comprises amplifying the cancer sample with the entire assay primer set in a multiplex reaction.

[0016] In some versions, the amplifying the cancer sample with the assay primer set comprises amplify ing the cancer sample to exponential phase with the assay primer set.

[0017] In some versions, the amplifying the test sample with the assay primer set comprises amplifying the test sample with the entire assay primer set in a multiplex reaction. In some versions, the amplifying the test sample with the assay primer set comprises amplifying the test sample to exponential phase with the assay primer set.

[0018] In some versions, the test sample comprises nucleic acid obtained from the subject after being treated for cancer.

[0019] In some versions, the test sample comprises cell-free DNA from the subject.

[0020] In some versions, the analyzing the test amplification product comprises sequencing the test amplification product.

[0021] In some versions, the test sample comprises the cancer nucleic acid in a fraction relative to the non-cancer nucleic acid less than 0.01% and the cancer nucleic acid is detected.

[0022] In some versions, the test sample comprises total DNA in an amount less than 50 ng. The objects and advantages of the invention will appear more fully from the following detailed description of the preferred embodiment of the invention made in conjunction with the accompanying drawings.

[0023] BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIG. 1 . Schema outlining exemplary aspects of the invention.

[0025] FIG. 2. Ovendew of an exemplary process of the invention.

[0026] FIGS. 3A and 3B. Overview of the samples used for each step in exemplary laboratory' and computational processes of the invention.

[0027] FIG. 4. Exemplary types of structural variants. Structural genetic variants are a hallmark of many cancers and somatic structural variants are uncommon in normal tissues. Although there are many types of structural variants, they are all characterized by unique breakpoints that distinguish them from normal genomes. They have advantages over single nucleotide variants in that they are less vulnerable to polymerase errors introduced by PCR technologies.

[0028] FIG. 3. Exemplary structural -variant primer design strategies. Top row shows a primer design strategy for a deletion as a structural variant. Bottom row shows a primer design strategy’ for a translocation as a structural variant.

[0029] FIG. 6. Schema of matching sequenced fragments to expected structural -variant or housekeeping gene sequences and quantifying tumor content using housekeeping gene product counts as a reference.

[0030] FIG. 7. Schema of relative quantification of tumor fraction. Structural variant fragment counts are influenced by cfDNA tumor count and input. Housekeeping gene fragment counts are influenced only by input.

[0031] FIG. 8. Tumor DNA detection in contrived samples at multiple levels of input using an exemplary' method of the invention (SVEN). Each box represents a replicate reaction. Empty boxes indicate undetectable tumor structural variants. Filled shading indicate inferred- tumor fraction. AF, mutant allele fraction.

[0032] FIGS. 9 A and 9B. Comparison of observed tumor fractions (calculated using SVEN) and expected tumor fractions (in contrived dilution series of tumor in normal DNA from two breast cancer cell lines, HCC1954 (FIG. 9 A) and HCC1143 (FIG. 9B)). Each small dot represents an individual SVEN reaction at an input level of 5-40ng. Large dots represent average observed tumor fraction for each set of expected tumor fractions. DETAILED DESCRIPTION OF THE INVENTION

[0033] One aspect of the invention comprises methods of detecting and optionally quantifying cancer nucleic acid in a test sample from a subject. FIGS. 1-3B show schemas of various exemplary methods of the invention with exemplary steps.

[0034] "‘Nucleic acid” refers to DNA, RNA, or other nucleic acid forms. Exemplary types of DNA includes genomic DNA (gDNA), extra-chromosomal DNA, cell-free DNA (cfDNA), circulating-tumor DNA (ctDNA), etc. Exemplary types of RNA include messenger RNA (mRNA), ribosomal RNA (rRNA), transfer RNA (tRNA), microRNA (miRNA), etc.

[0035] “Cancer nucleic acid” refers to nucleic acid contained in, produced by, obtained from, or derived from (e.g., copied or amplified from) a cancer cell. The cancer nucleic acid can be in the form of genomic DNA from a cancer cell, mRNA from a cancer cell, cfDNA from a cancer call, ctDNA from a cancer cell, or other forms.

[0036] “Test sample” refers to any sample containing or suspected of containing cancer nucleic acid. The test sample can comprise a clinical sample or a sample containing material obtained from, processed from, or derived from a clinical sample. “Clinical sample” refers to a sample obtained directly from a subject. Clinical samples include solid biopsies, liquid biopsies (e ., blood samples and plasma samples), saliva samples, stool samples, nasal samples, etc. In some versions of the invention, the test sample is a cfDNA sample comprising cfDNA extracted from a plasma sample from a subject.

[0037] The subject can include any animal, such as a mammal or a human. The subject in various versions is an individual suspected of having, diagnosed with, being treated for, previously treated for, or in remission from cancer.

[0038] The methods of the invention can comprise identifying one or more structural variants in the cancer nucleic acid from the subject relative to non-cancer nucleic acid.

[0039] “Non-cancer nucleic acid” refers to nucleic acid contained in, produced by, obtained from, or derived from (e.g, copied or amplified from) a non-cancer (“normal”) cell. As with the cancer nucleic acid, the non-cancer nucleic acid can be in the form of genomic DNA from a non-cancer cell, mRNA from a non-cancer cell, cfDNA from a non-cancer call, ctDNA from a non-cancer cell, or other forms. In some versions, the non-cancer nucleic acid can comprise germline DNA.

[0040] As is used in the art, “structural variant” or “structural variation” refers to a large alteration in a first nucleic acid relative to a second, otherwise-corresponding nucleic acid. In the present case, “structural variant” refers to a large alteration in the cancer nucleic acid relative to the non-cancer nucleic acid. The structural variants of the invention in various versions can comprise a sequence difference in at least 50 contiguous bp, at least 100 contiguous bp, at least 200 contiguous bp, at least 300 contiguous bp, at least 400 contiguous bp, at least 500 contiguous bp, at least 600 contiguous bp, at least 700 contiguous bp, at least 800 contiguous bp, at least 900 contiguous bp, or at least 1000 contiguous bp in the cancer nucleic acid relative to the non-cancer nucleic acid. Exemplary structural variants include insertions, deletions, inversions (e.g, reciprocal inversions, fold-back inversions), translocations (e.g, balanced translocations, unbalanced translocations, reciprocal translocations), duplications (e.g, tandem duplications), and local-distant clusters, among others. See, e.g, FIG. 4, Yang et al. 2020 (Yang L. A Practical Guide for Structural Variation Detection in the Human Genome. Curr Protoc Hum Genet. 2020 Sep;107(l):el03) and Li et al. 2020 (Li Y, et al. Patterns of somatic structural variation in human cancer genomes. Nature. 2020 Feb; 578, 112-121).

[0041] In some versions, identifying the one or more structural variants can comprise sequencing the cancer nucleic acid from a cancer sample from the subject and / or sequencing the non-cancer nucleic acid from a non-cancer sample from the subject. In some versions, the non-cancer nucleic acid can be from a non-cancer sample from an unrelated individual (z.e., not the subject), or a pool of individuals. Any non-cancer nucleic acid not from the subject used for identifying the structural variants should be assumed to be representative of non- cancer nucleic acid from the subject. The sequences obtained thereby can then be compared to identify one or more structural variants in the cancer nucleic acid relative to the non-cancer nucleic acid. The cancer nucleic acid used for identifying the structural variants can be from a cancer sample from the subject. "Cancer sample'’ in this regard refers to a clinical sample from the subject containing cancer nucleic acid. In some versions, the cancer sample is a biopsy of cancer tissue or cancer cells. Similarly, the non-cancer nucleic acid used for identifying the structural variants can be from a non-cancer sample, whether from the subject an unrelated individual, or a pool of individuals. “Non-cancer sample” in this regard refers to a clinical sample containing non-cancer nucleic acid. In some versions, the non-cancer sample is a biopsy of non-cancer tissue or non-cancer cells. The sequencing can be performed with any sequencing methods known in the art. In some versions, only a portion of the cancer nucleic acid and / or non-cancer nucleic acid is sequenced. In some versions, most or all of the cancer nucleic acid and / or non-cancer nucleic acid is sequenced (e.g, whole-genome sequencing).

[0042] In some versions, identifying the structural variants is performed by comparing sequences from the cancer nucleic acid from the subject with a standard reference genome (e.g, GRCh38 / hg38, among others) not from the subject, wherein the standard reference genome is assumed to be representative of non-cancer nucleic acid from the subject. In some versions, identifying the structural variants comprises whole-genome sequencing of a patient tumor biopsy, aligning sequencing reads to a reference genome, analyzing the sequenced genome of the patient tumor biopsy with a structural variant caller (more than one (e.g, 2, 3, or more) can be used in parallel, and identifying consensus breakpoint calls, e.g. , via a network model.

[0043] The methods of the invention can comprise providing an assay primer set. The assay primer set can comprise structural-variant primers. The structural-variant primers can comprise one or more structural-variant primer pairs. The structural-variant primer pairs are preferably configured to amplify the one or more structural variants in the cancer nucleic acid but lack a defined amplifiable target in the non-cancer nucleic acid. See, e.g., FIG. 5.

[0044] ■‘Amplify the one or more structural variants’’ as used herein refers to amplifying at least a portion of the structural variant, w ith or without non-structural-variant sequence, and does not require amplifying the entire structural variant. To obtain primer pairs configured to amplify the one or more structural variants in the cancer nucleic acid, the structural-variant primer pairs are preferably configured bind to the cancer nucleic acid and extend toward each other across a breakpoint betw een the structural variant and non-structural-variant sequence at an amplifiable distance such that at least a portion of the structural variant, a portion of the non-structural-variant sequence, and the breakpoint are amplified.

[0045] “Lack of a defined amplifiable target” refers to a predicted and / or empirically determined inability of the primers to specifically amplify any defined portion of the noncancer nucleic acid. The lack of a defined amplifiable target in the non-cancer nucleic acid can result from the primers binding to the non-cancer nucleic acid but extending toward each other at anon-amplifiable distance, binding to the non-cancer nucleic acid but extending away each other, binding to different non-cancer nucleic acid molecules (e.g, different chromosomes), or other mechanisms. In some cases, the lack of a defined amplifiable target in the non-cancer nucleic acid will depend on the type of structural variant.

[0046] In some versions, the primer design can comprise identifying primers for all structural variant targets, filtering the primers by thermodynamic and sequence properties, running a simulated PCR, excluding primers that generate off-target products, selecting the best primer pair for each structural variant target for multiplex pool, and adding single-copy housekeeping gene control primer pairs.

[0047] In various versions of the invention, the assay primer set comprises at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, or at least 60 structural-variant primer pairs.

[0048] In various versions of the invention, the assay primer set comprises up to 10, up to 11, up to 12, up to 13, up to 14, up to 15, up to 16, up to 17, up to 18, up to 19, up to 20, up to 21, up to 22, up to 23, up to 24, up to 25, up to 26, up to 27, up to 28, up to 29, up to 30, up to 35, up to 40, up to 45, up to 50, up to 55, up to 60, up to 65, up to 70, up to 75, up to 80, up to 85, up to 90 or more structural-variant primer pairs.

[0049] In various versions of the invention, the assay primer set comprises at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, or at least 60 structural-variant primer pairs that are configured to amplify different structural variants.

[0050] In various versions of the invention, the assay primer set comprises up to 10, up to 11, up to 12, up to 13, up to 14, up to 15, up to 16, up to 17, up to 18, up to 19, up to 20, up to 21, up to 22, up to 23, up to 24, up to 25, up to 26, up to 27, up to 28, up to 29, up to 30, up to 35, up to 40, up to 45, up to 50, up to 55, up to 60, up to 65, up to 70, up to 75, up to 80, up to 85, up to 90 or more structural-variant primer pairs that are configured to amplify different structural variants.

[0051] Increasing the amount of different structural variants that are amplified increases the sensitivity of detecting low amounts of cancer nucleic acid in a sample and also improves quantification accuracy. Thus, higher numbers of structural-variant primer pairs that target different structural variants are preferred.

[0052] The assay primer set in some versions can comprise control primers. The control primers can comprise one or more control primer pairs. The control primer pairs are preferably configured to amplify a control target present in both the cancer nucleic acid and the noncancer nucleic acid. The control target can include any nucleic acid sequence present in both the cancer nucleic acid and the non-cancer nucleic acid that does not contain any portion of a structural variant. In some versions, the control target is a sequence from a “housekeeping gene.”

[0053] In various versions of the invention, the assay primer set comprises at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, or at least 60 control primer pairs.

[0054] In various versions of the invention, the assay primer set comprises up to 10, up to 11, up to 12, up to 13, up to 14, up to 15, up to 16, up to 17, up to 18, up to 19, up to 20, up to 21, up to 22, up to 23, up to 24, up to 25, up to 26, up to 27, up to 28, up to 29, up to 30, up to 35, up to 40, up to 45, up to 50, up to 55, up to 60, up to 65, up to 70, up to 75, up to 80, up to 85, up to 90 or more control primer pairs.

[0055] In various versions of the invention, the assay primer set comprises at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 11, at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21, at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 35, at least 40, at least 45, at least 50, at least 55, or at least 60 control primer pairs that are configured to amplify different control sequences.

[0056] In various versions of the invention, the assay primer set comprises up to 10, up to 11, up to 12, up to 13, up to 14, up to 15, up to 16, up to 17, up to 18, up to 19, up to 20, up to 21, up to 22, up to 23, up to 24, up to 25, up to 26, up to 27, up to 28, up to 29, up to 30, up to 35, up to 40, up to 45, up to 50, up to 55, up to 60, up to 65, up to 70, up to 75, up to 80, up to 85, up to 90 or more control primer pairs that are configured to amplify different control sequences.

[0057] The methods of the invention can comprise amplifying the test sample with the assay primer set to yield test amplification product. ‘‘Amplify the test sample with the assay primer sef’ refers to subjecting the test sample to conditions that would be effective to amplify any targets of the assay primer set in the test sample. Any method of nucleic acid amplification can be employed in this step. Preferred methods include PCR or other thermocycling methods.

[0058] In some versions of the invention, amplifying the test sample with the assay primer set comprises amplifying the test sample with the entire assay primer set in a multiplex reaction. In other words, the entire assay primer set is employed together and simultaneously in amplify ing at least one single test sample.

[0059] In some versions of the invention, amplifying the test sample with the assay primer set comprises amplifying the test sample to exponential phase with the assay primer set. Amplifying any given sample referred to herein “to exponential phase with the assay primer set” refers to generating any and all specific amplicons from the assay primer set to an exponential phase of amplification without entering a plateau phase of amplification. Exponential and plateau phases of amplification are well understood in the art. “Any and all specific amplicons from the assay primer set” refer to amplicons specifically targeted with the primer pairs of the assay primer set (e.g. , the structural-variant pnmer pairs and control primer pairs) and do not refer from amplicons from off-target amplification. “Any” in phrase “any and all specific amplicons” accommodates that fact that some of the primer pairs will not yield any specific amplifications in cases in which target is not present.

[0060] The methods of the invention can comprise analyzing the test amplification product for presence of test-sample structural -variant amplicons from the structural -variant primers to thereby detect any cancer nucleic acid in the test sample. This step can comprise sequencing the test amplification product for specific targets of the structural-variant primers to thereby detect the presence of the targeted structural variants in the test sample.

[0061] Some aspects of the invention are directed to quantifying cancer nucleic acid in a test sample.

[0062] The quantifying can include amplifying a cancer sample from the subject with the assay primer set to generate cancer-sample structural-variant amplicons and cancer-sample control amplicons. “Cancer-sample structural-variant amplicons” refer to amplicons produced in the cancer sample from the structural-variant primers. “Cancer-sample control amplicons” refer to amplicons produced in the cancer sample from the control primers. The cancer sample preferably contains a known amount of cancer nucleic acid contained therein, such as a known amount of cancer nucleic acid relative to non-cancer nucleic acid. In some versions, the cancer sample consists of or essentially consists of cancer nucleic acid, save for some minute amounts of contaminating non-cancer nucleic acid potentially present in an original clinical sample. In some versions, the cancer sample is the same cancer sample used to identify the structural variants in the cancer nucleic acid from the subj ect relative to non-cancer nucleic acid. In some versions, the cancer sample is different from the same cancer sample used to identify the structural variants in the cancer nucleic acid from the subject relative to non-cancer nucleic acid.

[0063] In some versions of the invention, amplifying the cancer sample with the assay primer set comprises amplifying the cancer sample with the entire assay primer set in a multiplex reaction. In other words, the entire assay primer set is employed together and simultaneously in amplifying at least one single cancer sample.

[0064] In some versions of the invention, amplifying the cancer sample with the assay primer set comprises amplifying the test sample to exponential phase with the with the assay primer set.

[0065] Once cancer-sample structural-variant amplicons and cancer-sample control amplicons are obtained, a normalization factor corresponding to each of one or more structural-variant primer pairs can be determined. The one or more structural-variant primer pairs can include any one or more of the structural-variant primer pairs, in any combination. In some versions, the one or more structural-variant primer pairs include each of the structural- variant primer pairs, such that a normalization factor is determined for each structural-variant primer pair.

[0066] The normalization factors can be determined by determining a cancer-sample structural-variant amount corresponding to each of one or more of the structural-variant primer pairs, determining a cancer-sample normalization amount, and normalizing each cancersample structural-variant amount to the cancer-sample normalization amount.

[0067] Each cancer-sample structural-variant amount is an amount of the cancer-sample structural-variant amplicons amplified from the cancer sample with a corresponding one of the one or more structural-variant primer pairs. “Corresponding” in the context of particular amplicons and primer pairs corresponding to each other refers to a relationship in which the particular amplicons are or were amplified with the particular primer pairs.

[0068] The cancer-sample normalization amount is the amount of the cancer-sample control amplicons amplified from the cancer sample with corresponding one or more, respectively, of the cancer-sample primers. “One or more” in this context refers to an amount of amplicons corresponding to any single cancer-sample control primer pair or a combination of more than one cancer-sample control primer pair. An amount of amplicons from a combination of more than one cancer-sample control primer pair is generally understood to constitute an average of the individual amounts of amplicons amplified from each of the cancer-sample control primer pairs in the combination. The amounts of the respective amplicons can be determined by sequencing the amplicons and analyzing the sequences. The amounts of the amplicons can be represented by amplicon count. In the exemplary version of the invention, the cancer-sample normalization amount is the amplicon amount corresponding to the cancer-sample primers producing the highest amplicon counts in the cancer sample.

[0069] Normalizing cancer-sample structural-variant amounts to the cancer-sample normalization amount to obtain the normalization factors can be performed by mathematically adjusting the cancer-sample structural-variant amounts to the cancer-sample normalization amount. Any suitable mathematical operation can be used. In the exemplary version of the invention, normalizing cancer-sample structural-variant amounts to the cancer-sample normalization amount is performed by dividing each of the cancer-sample structural-variant amounts by the cancer-sample normalization amount. Each normalization factor represents a value for the efficiency of a given structural -variant primer pair relative to one or more of the control primer pairs for a given amount of cancer nucleic acid in a sample. Once the normalization factors for the structural-variant primer pairs are determined, the normalization factors can be used to normalize amplification amounts from a test sample to adjust for differences in nucleic acid template input when amplifying the test sample. This process can comprise determining a test-sample structural-variant amount corresponding to each of the one or more of the structural-variant primer pairs and normalizing each test-sample structural-variant amount to its corresponding normalization factor. Each test-sample structural-variant amount is an amount of test-sample structural-variant amplicons amplified from the test sample with a corresponding one of the one or more structural-variant primer pairs. Normalizing test-sample structural-variant amounts to their corresponding normalization factors can be performed by mathematically adjusting the test-sample structural-variant amounts to the normalization factors. Any suitable mathematical operation can be used. In the exemplary version of the invention, normalizing test-sample structural- variant amounts to their corresponding normalization factors is performed by dividing each of the test-sample structural-variant amounts by its corresponding normalization factor. Normalizing each test-sample structural-variant amount to its corresponding normalization factor thereby obtains a normalization-factor-adjusted test-sample structural -variant amount corresponding to each test-sample structural-variant amount.

[0070] Similar operations relating to the control primer pairs can be used to determine a cancer-nucleic-acid fraction in a test sample. “Cancer-nucleic-acid fraction” refers to the amount of cancer nucleic acid as a fraction of total nucleic acid (cancer nucleic acid and noncancer nucleic acid) in a test sample. An example of a cancer-nucleic-acid fraction is “tumor fraction” as generally understood in the art. Any considerations outlined above with respect to the operations pertaining the structural-variant primer pairs, such as the considerations pertaining to the terms “amount,” “corresponding,” “one or more,” etc., apply to the operations performed with the control primer pairs, unless the context dictates otherwise.

[0071] Determining the cancer-nucleic-acid fraction can include determining a cancer-sample control amount corresponding to each of one or more of the control primer pairs. Each cancersample control amount is an amount of the cancer-sample control amplicons amplified from the cancer sample with a corresponding one of the one or more control primer pairs.

[0072] Another step can include normalizing each cancer-sample control amount to the cancer-sample normalization amount to thereby obtain a normalization factor corresponding to each of the one or more control primer pairs.

[0073] Another step can include determining a test-sample control amount corresponding to each of the one or more of the control primer pairs. Each test-sample control amount is an amount of test-sample control amplicons amplified from the test sample w ith a corresponding one of the one or more control primer pairs. In some versions, the test-sample control amount is the test-sample control amount corresponding to a single control primer pair. In some versions, the test-sample control amount is an average corresponding to multiple control primer pairs.

[0074] Another step can include normalizing each test-sample control amount to its corresponding normalization factor to thereby obtain a normalization-factor-adjusted testsample control amount corresponding to each test-sample control amount. In some versions, the normalization-factor-adjusted test-sample control amount is the normalization-factor- adjusted test-sample control amount corresponding to a single test-sample control amount. In some versions, the normalization-factor-adjusted test-sample control amount is an average corresponding to multiple test-sample control amounts.

[0075] Another step can include normalizing one or more of the normalization-factor-adjusted test-sample structural-variant amounts to the normalization-factor-adjusted test-sample control amount to thereby obtain the cancer-nucleic-acid fraction in the test sample. In some versions, the cancer-nucleic-acid fraction is the cancer-nucleic-acid fraction corresponding to a single normalization-factor-adjusted test-sample structural-variant amount. In some versions, the normalization-factor-adjusted test-sample control amount is an average corresponding to multiple normalization-factor-adjusted test-sample structural -variant amounts.

[0076] An example of the operations with the structural- variant primers and control primers using contrived cancer-sample (structural variant or control) amounts for a cancer sample containing a cancer-nucleic acid fraction of 1, contrived test-sample (structural variant or control) amounts for a test sample containing a cancer-nucleic acid fraction of 0.5, structural variant primer pairs S1-S3, and control primer pairs C1-C2 is shown in Table 1 below:

[0077] Table 1. Exemplary' determination of cancer-nuclei acid fraction.

[0078] Consistent with the discussion above, any one or more of structural primer pairs SI, S2, and S3 (e.g, SI; S2; S3; S I and S2; S2 and S3; SI and S3; or S I, S2, and S3) can be used for determining cancer-sample amounts and / or downstream operations. Any one or more of Cl and C2 (e.g, Cl, C2, or Cl and C2) can be used for determining cancer-sample amounts and / or downstream operations. The cancer-sample amount from any one or more of Cl and C2 (e.g., Cl, C2, or an average of Cl and C2) can be used as the cancer-sample normalization amount for determining the normalization factors. Any one or more of structural primer pairs SI, S2, and S3 (e.g., SI; S2; S3; SI and S2; S2 and S3; SI and S3; or SI, S2, and S3) can be used for determining test-sample amounts and / or downstream operations. The test-sample amount from any one or more of Cl and C2 (e.g, Cl, C2, or an average of Cl and C2) can be used as the cancer-sample normalization amount for determining the normalization factors. Within the structural-variant primers, the test sample amount for each primer pair can be individually adjusted with its corresponding normalization factor, or an average of test sample amounts for a given set of primer pairs can be adjusted with an average of the normalization factors corresponding thereto. Within control pnmers, the test sample amount for each primer pair can be individually adjusted with its corresponding normalization factor, or an average of test sample amounts for a given set of primer pairs can be adjusted with an average of the normalization factors corresponding thereto. The normalization-factor-adjusted test-sample amount from any one or more control primer pairs (averaged if using more than one) can be used to determine the cancer-nucleic acid fraction for the structural-variant primers. The cancer-nucleic-acid fraction from any one or more control primer pairs (averaged if using more than one) can be used to constitute the cancer-nucleic-acid fraction of the test sample. Similar considerations apply to any number of structural-variant primer pairs and control primer pairs. Additional modifications evident to a practitioner in the art are within the scope of the present invention.

[0079] The invention advantageously provides the ability7to detect cancer nucleic acid in samples having extremely low tumor fraction. Accordingly, in various versions of the invention, the test sample can comprise the cancer nucleic acid in a fraction relative to noncancer nucleic less than 1%, less than 0.75%, less than 0.5%, less than 0.25%, less than 0.1%, less than 0.075%, less than 0.05%, less than 0.01%, less than 0.0075%, less than 0.005%, less than 0.0025%, or less than 0.001%. In various versions of the invention, the test sample can comprise the cancer nucleic acid in a fraction relative to non-cancer nucleic acid of at least 0.75%, at least 0.5%, at least 0.25%, at least 0.1%, at least 0.075%, at least 0.05%, at least 0.01%, at least 0.0075%, at least 0.005%, at least 0.0025%, at least 0.001%, at least 0.00075%, at least 0.0005%, at least 0.00025%, or at least 0.0001%. In various versions of the invention, the methods of the invention can detect the cancer nucleic acid in a test sample comprising the cancer nucleic acid in a fraction relative to non-cancer nucleic less than 1%, less than 0.75%, less than 0.5%, less than 0.25%, less than 0.1 %, less than 0.075%, less than 0.05%, less than 0.01%, less than 0.0075%, less than 0.005%, less than 0.0025%, or less than 0.001%. In various versions of the invention, the methods of the invention can detect the cancer nucleic acid in a test sample comprising the cancer nucleic acid in a fraction relative to non-cancer nucleic acid of at least 0.75%, at least 0.5%, at least 0.25%, at least 0.1%, at least 0.075%, at least 0.05%, at least 0.01%, at least 0.0075%, at least 0.005%, at least 0.0025%, at least 0.001%, at least 0.00075%, at least 0.0005%, at least 0.00025%, or at least 0.0001%. Such fractions are determined as is performed in the art for determining tumor fraction DNA, such as mutant allele fraction, average mutant allele fraction, or tumor fraction inferred by analysis of copy number aberrations. See, e.g., McDonald et al. 2019 (McDonald BR, Contente-Cuomo T, Sammut SJ, Odenheimer-Bergman A, Ernst B, Perdigones N, Chin SF, Farooq M, Mejia R, Cronin PA, Anderson KS, KosiorekHE, Northfelt DW, McCullough AE, Patel BK, Weitzel JN, Slavin TP, Caldas C, Pockaj BA, Murtaza M. Personalized circulating tumor DNA analysis to detect residual disease after neoadjuvant therapy in breast cancer. Sci Transl Med. 2019 Aug 7;l l(504):eaax7392).

[0080] In various versions of the invention, the sample can comprise total nucleic acid, such as total DNA) in an amount less than 100 pg, less than 50 pg, less than 10 pg, less than 5 pg, less than 1 pg, less than 900 ng, less than 800 ng, less than 700 ng, less than 600 ng, less than 500 ng, less than 400 ng, less than 300 ng, less than 200 ng, less than 100 ng, less than 50 ng, less than 45 ng, less than 40 ng, less than 35 ng, less than 30 ng, less than 25 ng, less than 20 ng, less than 15 ng, less than 10 ng, less than 5 ng, or less than 1 ng. In various versions of the invention, the sample can comprise total nucleic acid, such as total DNA) in an amount at least 45 ng, at least 40 ng, at least 35 ng, at least 30 ng, at least 25 ng, at least 20 ng, at least 15 ng, at least 10 ng, at least 5 ng, at least 1 ng, or at least 0.5 ng. Amounts above and below these amounts are within the scope of the invention.

[0081] The elements and method steps described herein can be used in any combination whether explicitly described or not.

[0082] All combinations of method steps as used herein can be performed in any order, unless otherwise specified or clearly implied to the contrary7by the context in which the referenced combination is made. As used herein, the singular forms "a." “an,” and "the" include plural referents unless the content clearly dictates otherwise.

[0083] Numerical ranges as used herein are intended to include every number and subset of numbers contained within that range, whether specifically disclosed or not. Further, these numerical ranges should be construed as providing support for a claim directed to any number or subset of numbers in that range. For example, a disclosure of from 1 to 10 should be construed as supporting a range of from 2 to 8, from 3 to 7, from 5 to 6, from 1 to 9, from 3.6 to 4.6, from 3.5 to 9.9, and so forth.

[0084] All patents, patent publications, and peer- reviewed publications (z.e., "references") cited herein are expressly incorporated by reference to the same extent as if each individual reference were specifically and individually indicated as being incorporated by reference. In case of conflict between the present disclosure and the incorporated references, the present disclosure controls.

[0085] It is understood that the invention is not confined to the particular construction and arrangement of parts herein illustrated and described, but embraces such modified forms thereof as come within the scope of the claims.

[0086] EXAMPLES

[0087] STRUCTURAL VARIANT ENRICHMENT AND NORMALIZATION (SVEN): A TUMOR- GUIDED APPROACH FOR DETECTION AND QUANTIFICATION OF CTDNA IN CANCER PATIENTS WITH RESIDUAL DISEASE

[0088] Summary

[0089] A small fraction of cancer patients treated with curative intent experience disease recurrence. There are few7effective biomarkers currently available to detect minimal residual disease (MRD) after completion of therapy. Circulating tumor DNA (ctDNA) holds promise as a biomarker to detect tumor-derived DNA in blood from patients with MRD. Most MRD assays rely on prior identification of patient-specific single nucleotide variants (SNV) in DNA from the tumor biopsy, and their subsequent detection in plasma DNA. Due to polymerase errors during amplification and sequencing noise, confidence in ctDNA detection at and below 0.01% is limited. Here, we describe an alternative approach that relies on patient-specific structural variants (SVs), referred to as Structural Variant Enrichment and Normalization (SVEN). Overall (FIGS. 1-3B), SVEN is a process that includes identification of SV breakpoints from tumor and germline DNA whole genome sequencing, design of multiplexed primer panels for each patient, a wet-lab workflow to amplify and sequence patient-derived plasma DNA samples for patient-specific SV targets together with a prespecified set of control targets (in a single multiplexed reaction), and analysis of the resulting targeted sequencing data to establish tumor DNA detection and quantification of tumor fraction in the tested sample. We describe the process and demonstrate its performance in various sets of contrived samples generated using DNA from commercially available cancer cell lines.

[0090] Methods

[0091] Overall workflow (FIGS. 1-3B)

[0092] 1. Preparation and data generation for whole genome sequencing libraries from tumor DNA (from cell line, frozen tumor biopsy or FFPE tumor sections) and germline DNA (from peripheral blood cells).

[0093] 2. Computational analysis of whole genome sequencing data to identify tumor-derived (somatic) patient-specific SVs.

[0094] 3. Prioritization of SVs for inclusion in a patient-specific multiplexed primer panel.

[0095] 4. Design of patient-specific multiplexed primer panel for target SVs (along with 5 prespecified control targets).

[0096] 5. Testing of each patient-specific panel using control samples to ensure effective combined performance and minimize noise.

[0097] 6. Amplification and sequencing of test sample (plasma DNA sample or contrived sample) using patient-specific primers.

[0098] 7. Bioinformatic analysis of targeted sequencing data to detection of SVs.

[0099] 8. Statistical analysis to detect and quantify tumor-derived DNA in plasma sample.

[0100] 1. Preparation and data generation for whole genome sequencing of tumor and germline DNA (from cell lines or tissue samples)

[0101] We analyzed DNA samples from two sets of commercially available cancer (breast cancer) cell lines, each with matched tumor and normal cell line DNA, HCC1954 (ATCC) and HCC1143 (ATCC). HCC1954 is an epithelial cell line derived from primary tumor from a patient with Stage IIA ER- HER2+ invasive ductal carcinoma. HCC1134 is also an epithelial cell line derived from primary' tumor from a patient with Stage IIA ER- HER2- invasive ductal carcinoma. HCC1954-BL and HCC1143-BL are matched normal cell lines derived from peripheral blood of the same patients. Genomic DNA from HCC1954 and HCC1954 BL is commercially available from ATCC.

[0102] Intact genomic DNA from cell lines was sheared to an average size of 150 - 200 bp on a Bioruptor Pico shearing sonicator (Diagenode) to reflect the modal size of human cell free DNA. Tumor and match normal samples were sheared to be within 15 bp of each other. The cycling program was 30 s on, 30 s off for 24 - 30 cycles. DNA size was assayed on an Agilent TapeStation.

[0103] 2. Computational analysis for identification of tumor-derived patient-specific SVs

[0104] Illumina paired-end whole genome sequencing data for tumor and matched normal tissue is aligned to the human reference genome using bwa mem and samtools. Three published somatic structural variant callers, delly, mutect, and svaba, are each run in somatic mode to call putative structural variants. We consider the consensus of the three callers as follows: each breakpoint identified is considered confident if at least two callers identify the same breakpoint at base pair resolution with a coordinate that differs by less than lOObp.

[0105] 3 / 4. Prioritization of SVs for inclusion in panel and design of patient-specific primer panels

[0106] Using consensus structural variant calls, a PCR primer panel is designed for each patient to amplify their tumor-specific SVs. Potential primer sequences are generated for each consensus breakpoint where both sides are identified, and then filtered for melting temperature and sequence features that might result in poor primer performance such as poor GC content and repetitive sequences.

[0107] If there are significantly more SV targets than the desired panel size, SVs are further filtered by flanking region complexify, SV allele frequency in the tumor, SV type, amplicon size in non-tumor tissue, and the number of potential primers that passed primer filtering. So far, our goal has been to design primer panels that target -30-60 SVs in multiplex.

[0108] Once the best set of targets has been selected, all primers are analyzed using in-silico PCR to identify and exclude primers that may create off-target amplification. A set of known housekeeping gene primers are added to the panel at this stage to serve as controls during downstream quantitation. The best forward and reverse primers for each SV target are included in the final panel design.

[0109] 5. Testing of primer panels using control samples

[0110] Normal template samples and no template controls were assessed concomitantly for all multiplex primer panels to identify false positives. Eight replicates at multiple input amounts (5ng to 40ng) were performed for the normal template, and four replicates of no template controls were performed for each primer panel. Normal libraries were sequenced prior to and separately from tumor libraries. 6. Amplification and sequencing of test sample

[0111] Hot Start Taq DNA Polymerase (NEB) was used for targeted amplification, with optimization for multiplex reactions according to the manufacturer’s recommendations. Reactions were cycled in a BioRad C1000 thermal cycler using a touchdown PCR program with a drop of 1.5 °C in TA over each of the first 10 cycles from 70 °C to 55 °C followed by a limited number of amplification cycles (15) at 55 °C.

[0112] After thermal cycling, PCR reactions were cleaned by treatment with an exonuclease shrimp alkaline phosphatase cocktail, HT ExoSAP-IT (ThermoFisher) to degrade unincorporated primers, followed by size-selective bead-based cleanup with a 1: 1.8 ratio of SPRISelect paramagnetic beads (Beckman Coulter) to remove adapter dimers. There was no QC at this step as PCR products were typically below the level of detection by standard methods.

[0113] Sequencing Libraries were prepared directly from cleaned multiplex PCR products with the Takara ThruPlex DNA-Seq HV Library Preparation Kit and Takara Unique Dual Index Kits according to manufacturer’s instructions. Input was the entirety of the cleaned PCR products from the amplification step. Libraries underwent 11 cycles of amplification prior to cleanup. Prepared libraries were cleaned with a 1 : 1 ratio of SPRISelect beads then assayed for concentration and size on Agilent TapeStation D5000 tapes.

[0114] All sample libraries were pooled equimolarly and NTC libraries were added at 5 pl per library. The final pool was cleaned with a 1 :1 ratio of SPRISelect beads and eluted from in a volume estimated to give an appropriate concentration for sequencing. After QC on Agilent TapeStation, the final pool was prepared for loading at a concentration of 750 pM with 2% PhiX in RSBT. Samples were sequenced on a Pl or P2 flow cell for 300 cycles on a NextSeq 1000.

[0115] 7. Bioinformatic analysis of targeted sequencing data for detection and quantification

[0116] Paired end reads are merged into a single fragment sequence using fastp. These fragments are then aligned to the regions flanking the targeted SVs using LAST and matched to their original primer sequences by analyzing the fragment end sequences. The top LAST hit for each fragment is identified, requiring at least 40bp of alignment, overlap between the breakpoint and the fragment, and that the sequenced fragment be within 5bp of the expected length of amplicons targeting that breakpoint. Fragments passing these criteria are considered to be amplified copies of the target SV from the patient’s plasma cfDNA. See, e.g., FIG. 6.

[0117] We are currently evaluating an alternative approach whereby amplicons are only counted as matches to the expected sequence if both ends have a perfect match to the expected primer sequences for a given target. These are then aligned to the expected amplicon sequence, and only counted as true hits if they align with a certain score.

[0118] 8. Statistical analysis to detect and quantify tumor-derived DNA in plasma sample

[0119] One of the challenges in using SVs for ctDNA detection and quantification is the loss of quantitative signal in multiplexed bulk PCR and the differences between different primer efficiencies. We overcome this challenge by normalizing for the relative efficiency of each primer pair (see, e.g., FIG. 7):

[0120] 1. In 8 replicate samples of tumor DNA, the amplicon count for each SV and control target is divided by the amplicon count for the control target that is most abundant on average across all replicates. The median value provides a normalization factor for each target.

[0121] 2. To quantify tumor fraction in a given sample, the count of each amplicon is first divided by its normalization factor.

[0122] 3. The mean normalized count for control targets is calculated to use as the denominator for the tumor fraction.

[0123] 4. For each SV target, the allele fraction is calculated as the normalized count for that SV divided by the mean normalized control target count.

[0124] 5. Finally, the tumor fraction is given by the mean allele fraction of all SV targets.

[0125] Results

[0126] We obtained publicly available short-read Illumina whole genome sequencing data from all four cell lines, and analyzed fastq files using our in-house informatics pipeline for identification of SVs. For HCC1954, sequencing coverage was 86x and 60x for tumor and germline DNA respectively. For HCC1134, sequencing coverage was 88x and 74x for tumor and germline DNA respectively. For HCC1954, 1748 SV breakpoints were identified that passed our criteria for consensus between at least 2 of 3 SV callers. For HCC 1143, 674 SV breakpoints were identified that passed similar criteria. For HCC1954 and HCC1143, we prioritized and successfully design primer panels to target 28 and 29 variants, respectively. Each panel had an additional 5 universal control primers included as indicated earlier. Following control DNA and tumor DNA validation runs, we validated 24 (86%) and 23 (79%) targets for HCC1954 and HCC1143, respectively.

[0127] Intact genomic DNA from cell lines was sheared to an average size of 150 - 200 bp using a shearing sonicator to reflect the modal size of human cell free DNA. To evaluate quantitative performance, we performed the SVEN assay on dilution series of HCC1954 and HCC1143 samples at several different inputs and sequenced the resulting amplicon libraries on an Illumina NextSeq 2000 (Table 2). Altogether, including wildtype and tumor DNA dilution samples, we analyzed 352 SVEN reactions to establish quantitative accuracy in contrived samples using tw o different SVEN panels.

[0128] Table 2. Number in each cell indicates number of replicates analyzed at the corresponding tumor fraction and amount of input DNA. Experiments were performed using commercially available breast cancer cell lines HCC1954 and HCC1134, along with their corresponding peripheral blood normal cell lines.

[0129] We first evaluated if any targets amplified at least once in a replicate of normal DNA sample with no tumor fraction (32 replicates across 5ng-40ng for each cell line). No SVs were detected in wild-type samples for either cell line experiment, representing 100% specificity in these experiments. We detected tumor DNA represented by at least 1 positive re-arranged fragment across all tumor DNA samples from both cell lines between 1% and 0.03%, achieving a 100% sensitivity across 5ng to 40ng input (157 / 157 successfully analyzed replicates). For samples at 0.006% tumor fraction, we detected tumor DNA in 61 / 63 successful reactions, achieving 97% sensitivity. Only two samples, one each at 5 and lOng of input was missed for HCC 1143 dilutions where the tumor DNA was represented by 0.09 and 0.18 copies of the tumor genome respectively. For samples at 0.003% tumor fraction, we detected tumor DNA in 50 / 64 replicates, achieving 78% sensitivity (FIG. 8). To put these results into context, tumor is represented by 0.09-0.36 copies of the genome at 0.003% tumor fraction in 5-40 ng DNA samples. These sensitivity levels are achieved at 100% specificity in wild-type.

[0130] To quantify tumor fraction from a multiplex PCR reaction amplify ing SVs, we developed a novel approach for normalization of amplicon counts for SVs relative to control primers. Using this approach, we calculated SVEN-inferred tumor fractions for all samples from the dilution series of HCC 1954 and HCC1143 cell lines (FIGS. 9A and 9B). For HCC1954, the correlation between inferred tumor fractions and known tumor fractions across 144 data points (different inputs and tumor fraction levels) was 0.91. For HCC1134, the correlation between inferred tumor fractions and known tumor fractions across 141 data points (different inputs and tumor fraction levels) was 0.90. In HCC1954, the coefficient of variation (CV) between replicates were generally consistent across tumor fractions, with a decrease in CV at 0.3% tumor fraction as the number of SV molecules per reaction increased, reducing the impact of sampling noise (Table 3).

[0131] Table 3. Evaluation of quantitative performance of SVEN using contrived dilutions of HCC1954.

[0132] Discussion

[0133] Features of the exemplary approach

[0134] 1. Limited cycles of multiplexed amplification to minimize the contribution of differences in primer efficiency and capture the contribution of differences in starting template concentrations

[0135] 2. Single step quantification without the need to split the sample into multiple reactions, as needed for digital PCR, quantitative PCR, and recently described amplicon sequencing approaches for SV detection.

[0136] 3. The incorporation of pre-specified control primers in the molecular workflow as well as the statistical analysis for normalized of relative primer efficiency

[0137] 4. Standard workflow for every patient - no need for optimization of patient-specific primer panels. The same protocol is applied to all primer sets.

[0138] 5. The ability to quantify both tumor fraction and estimated tumor fragments per ml of plasma (or other body fluid) will be an added utility over SNVs. We have a multiplexed digital PCR approach that can quantify absolute number of cell-free DNA fragments in a plasma sample using the same pre-specified control targets used in the SVEN assay. Tumor fragments per ml can be calculated by tethering the two assays together.

[0139] Exemplary modifications

[0140] 1. Number of multiplexed targeted amplification cycles can be varied, depending on the balance desired between on-target rate and the ability to effectively calculate a normalization factor for each target. Additional cycles will add greater amplification bias between targets but will improve on-target sequencing rate.

[0141] 2. The number of SV targets tested can vary, we have found good performance with 25 targets, and are currently testing panels as large as 60 targets.

[0142] 3. The number and selection of pre-specified control targets can vary.

[0143] 4. The choice of master mix, enzyme, buffer, and temperature cycling conditions can vary for the multiplexed PCR. Similarly, the choice of library preparation downstream of multiplexed PCR can be variable.

[0144] 5. Plasma DNA is one potential template to analyze using SVEN, although SVEN can be used with any DNA template (or cDNA made from RNA for example). Since SVEN is a targeted amplification assay specific to SVs, it will do specially well when the tumor fraction is particularly low, and diluted by background DNA due to poor processing of sample or release of DNA from normal cells.

[0145] 6. Normalization factor can be calculated using any number of replicates of patient- derived tumor DNA or a dilution thereof, depending on the precision desired.

[0146] Implications of the current method and findings

[0147] In the current data so far, we have demonstrated that SVEN can amplify SVs identified in a tumor DNA sample and detect trace amounts of tumor DNA from mixtures of samples. The ability to detect 30 parts per million using just 25 targets from 5-20ng of DNA is a testament to potential implications of this approach. Our method allows us to directly quantify the tumor fraction using a set of pre-specific control targets that allow for internal calculation of PCR amplification biases (normalization factors) as well as for downstream calculation of tumor fraction.

[0148] Unlike SNV-assays that get affected by background DNA in samples that not well processed for ctDNA analysis (most bio-banked samples), SV assays only amplify tumor- derived fragments (ignoring all the background DNA from peripheral cells). We expect this feature to unlock sample banks from existing clinical trials of early-stage cancer patients and ease the stringent requirements for sample processing.

Claims

CLAIMSWhat is claimed is:

1. A method of detecting and optionally quantifying cancer nucleic acid in a test sample from a subject, the method comprising: identifying one or more structural variants in the cancer nucleic acid from the subject relative to non-cancer nucleic acid; providing an assay primer set comprising structural-variant primers, wherein the structural-variant primers comprise one or more structural-variant primer pairs, wherein each structural-variant primer pairs is configured to amplify at least one of the one or more structural variants in the cancer nucleic acid but lack a defined amplifiable target in the non-cancer nucleic acid; amplifying the test sample with the assay primer set to yield test amplification product; and analyzing the test amplification product for presence of test-sample structural-variant amplicons from the structural-variant primers to thereby detect any cancer nucleic acid in the test sample.

2. The method of claim 1, wherein the assay primer set comprises at least two structural- variant primer pairs.

3. The method of any prior claim, wherein the assay primer set comprises up to 80 structural- variant primer pairs.

4. The method of any prior claim, the assay primer set comprises at least two structural- variant primer pairs configured to amplify different structural variants.

5. The method of any prior claim, the assay primer set comprises at least 10 structural-variant primer pairs configured to amplify' different structural variants.

6. The method of any prior claim, wherein the identifying one or more structural variants comprises sequencing the cancer nucleic acid from a cancer sample from the subject and / or sequencing the non-cancer nucleic acid from a non-cancer sample from the subject.

7. The method of any prior claim, wherein the assay primer set further comprises control primers, wherein the control primers comprise one or more control primer pairs, wherein each control primer pair is configured to amplify a control target present in both the cancer nucleic acid and the non-cancer nucleic acid.

8. The method of claim 7, wherein the assay primer set comprises at least two control primer pairs.

9. The method of any one of claims 7-8, wherein the assay primer set comprises up to 10 control primer pairs.

10. The method of any one of claims 7-9, the assay primer set comprises at least two control primer pairs configured to amplify different control sequences.

11. The method of any one of claims 7-10, the assay primer set comprises at least four control primer pairs configured to amplify different control sequences.

12. The method of any one of claims 7-11, further comprising: amplifying a cancer sample from the subject with the assay primer set to generate cancersample structural-variant amplicons from the structural-variant primers and cancer-sample control amplicons from the control primers; determining a cancer-sample structural-variant amount corresponding to each of one or more of the structural-variant primer pairs, wherein each cancer-sample structural-variant amount is an amount of the cancer-sample structural-variant amplicons amplified from the cancer sample with a corresponding one of the one or more structural-variant primer pairs; determining a cancer-sample normalization amount, wherein the cancer-sample normalization amount is an amount of the cancer-sample control amplicons amplified from the cancer sample with corresponding one or more, respectively, of the control primer pairs; and normalizing each cancer-sample structural-variant amount to the cancer-sample normalization amount to thereby obtain a normalization factor corresponding to each of the one or more structural-variant primer pairs.

13. The method of claim 12, further comprising: determining a test-sample structural-variant amount corresponding to each of the one or more of the structural-variant primer pairs, wherein each test-sample structural-variant amount isan amount of test-sample structural-variant amplicons amplified from the test sample with a corresponding one of the one or more structural-variant primer pairs; and normalizing each test-sample structural -variant amount to its corresponding normalization factor to thereby obtain a normalization-factor-adjusted test-sample structural- variant amount corresponding to each test-sample structural-variant amount.

14. The method of claim 13, further comprising: determining a cancer-sample control amount corresponding to each of one or more of the control primer pairs, wherein each cancer-sample control amount is an amount of the cancersample control amplicons amplified from the cancer sample with a corresponding one of the one or more control primer pairs; normalizing each cancer-sample control amount to the cancer-sample normalization amount to thereby obtain a normalization factor corresponding to each of the one or more control primer pairs; determining a test-sample control amount corresponding to each of the one or more of the control primer pairs, wherein each test-sample control amount is an amount of test-sample control amplicons amplified from the test sample with a corresponding one of the one or more control primer pairs; normalizing each test-sample control amount to its corresponding normalization factor to thereby obtain a normalization-factor- adjusted test-sample control amount corresponding to each test-sample control amount; and normalizing one or more of the normalization -factor-adjusted test-sample structural- variant amounts to the normalization-factor-adjusted test-sample control amount to thereby obtain a cancer-nucleic-acid fraction in the test sample.

15. The method of any one of claims 12-14, wherein the amplifying the cancer sample with the assay primer set comprises amplifying the cancer sample with the entire assay primer set in a multiplex reaction.

16. The method of any one of claims 12-15, wherein the amplifying the cancer sample with the assay primer set comprises amplifying the cancer sample to exponential phase with the assay primer set.

17. The method of any prior claim, wherein the amplifying the test sample with the assay primer set comprises amplifying the test sample with the entire assay primer set in a multiplex reaction.

18. The method of any prior claim, wherein the amplifying the test sample with the assay primer set comprises amplifying the test sample to exponential phase wi th the assay primer set.

19. The method of any prior claim, wherein the test sample comprises nucleic acid obtained from the subject after being treated for cancer.

20. The method of any prior claim, wherein the test sample comprises cell-free DNA from the subj ect.

21. The method of any prior claim, wherein the analyzing the test amplification product comprises sequencing the test amplification product.

22. The method of any prior claim, wherein the test sample comprises the cancer nucleic acid in a fraction relative to the non-cancer nucleic acid less than 0.01% and the cancer nucleic acid is detected.

23. The method of any prior claim, wherein the test sample comprises total DNA in an amount less than 50 ng.

Images