Methods and materials for assessing homologous recombination deficiencies

Abstract

Methods and materials are provided for assessing samples (eg, cancer cells) for the presence of homologous recombination deficiency (HRD) or an HRD signature. The present invention provides an in vitro method for predicting patient response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor, the method comprising: (1) determining the number of indicator CA regions, including at least two of the indicator LOH regions, the indicator TAI regions, and / or the indicator LST regions, in at least one pair of human chromosomes in the cancer cells of a cancer patient in a sample containing cancer cells; and (2) diagnosing a patient whose sample contains a greater number of indicator LOH regions, TAI regions, or LST regions than the reference number as likely to respond to the cancer treatment regimen; The method includes:

Description

[Technical field]

[0001] Related Applications This application claims the benefit of priority to U.S. Provisional Application No. 62 / 037,764, filed August 15, 2014, and U.S. Utility Application No. 14 / 507,412, filed October 6, 2014 (the "'412 Application"), the entire contents of which are incorporated herein by reference. The '412 Application was originally filed in error as a U.S. utility application and was accorded the filing date set forth above. It is currently in the process of being converted to a U.S. provisional application, to which an application number has not yet been assigned. When such number is properly assigned by the U.S. Patent and Trademark Office, this international application will further claim, and will be deemed to have claimed, the benefit of priority of the provisional application. [Background technology]

[0002] background Cancer is a major public health problem, with 562,340 cancer deaths in the United States in 2009 alone. American Cancer Society, Cancer Facts & Figures 2009 (available on the American Cancer Society website). One of the most important challenges in cancer treatment is to discover the relevant and clinically useful features of a patient's own cancer and then, based on these features, administer the most appropriate treatment plan for the patient's cancer. Although this field of personalized medicine is progressing, there is still a great need for better molecular diagnostic tools to characterize a patient's cancer. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] American Cancer Society, Cancer Facts & Figures 2009 Summary of the Invention

[0004] overview This document relates to methods and materials relating to assessing a sample (e.g., cancer cells or nucleic acids derived therefrom) for homologous recombination deficiency (HRD) (e.g., HRD signature) based on the detection of specific chromosomal abnormalities ("CA"). For example, this document provides methods and materials for detecting CA regions to determine whether a cell (e.g., cancer cell) has HRD (e.g., exhibits an HRD signature). This document also provides materials and methods for identifying cancer patients likely to respond to a particular cancer treatment regimen based on the presence, absence, or severity of HRD. Throughout this document, unless otherwise indicated, HRD and homology-dependent repair (HDR) deficiency are used interchangeably.

[0005] In general, one aspect of the present invention features a method for evaluating HRD in cancer cells or DNA derived therefrom (e.g., genomic DNA). In some embodiments, the method comprises or consists essentially of (a) detecting CA regions (as defined herein) in at least one pair of human chromosomes (e.g., any human chromosome pair other than the human X / Y sex chromosome pair) in a sample or DNA derived therefrom, and (b) determining the number, size (e.g., length), and / or characteristics of the CA regions. In some embodiments, the CA regions are analyzed in several chromosome pairs representative of the entire genome (e.g., analyzing enough chromosomes such that the number and size of CA regions are expected to be representative of the number and size of CA regions across the entire genome).

[0006] Various aspects of the present invention involve the use of combined analysis of two or more CA regions to assess (e.g., detect) HRD in a sample. Three types of CA regions useful in such methods include (1) chromosomal regions that show loss of heterozygosity ("LOH regions" as defined herein), (2) chromosomal regions that show telomeric allele imbalance ("TAI regions" as defined herein), and (3) chromosomal regions that show large-scale transitions ("LST regions" as defined herein). CA regions with certain sizes, chromosomal locations, or characteristics (e.g., "indicator CA regions" as defined herein) are particularly useful in various aspects of the present invention described herein.

[0007] Thus, in one aspect, the present invention provides a method of assessing (e.g., detecting) HRD in a sample, comprising: (1) determining the total number of LOH regions in the sample having a certain size or characteristic (e.g., "indicator LOH regions" as defined herein); (2) determining the total number of TAI regions in the sample having a certain size or characteristic (e.g., "indicator TAI regions" as defined herein); and (3) assessing HRD in the sample based at least in part on the determinations made in (1) and (2). In another aspect, the present invention provides a method of assessing (e.g., detecting) HRD in a sample, comprising: (1) determining the total number of LOH regions in the sample having a certain size or characteristic (e.g., "indicator LOH regions" as defined herein); (2) determining the total number of LST regions in the sample having a certain size or characteristic (e.g., "indicator LST regions" as defined herein); and (3) assessing HRD in the sample based at least in part on the determinations made in (1) and (2). In another aspect, the present invention provides a method of assessing (e.g., detecting) HRD in a sample, comprising: (1) determining the total number of TAI regions in the sample having a particular size or characteristic (e.g., "indicator TAI regions" as defined herein); (2) determining the total number of LST regions in the sample having a particular size or characteristic (e.g., "indicator LST regions" as defined herein); and (3) assessing HRD in the sample based at least in part on the determinations made in (1) and (2).In another aspect, the present invention provides a method of assessing (e.g., detecting) HRD in a sample, the method comprising: (1) determining the total number of LOH regions in the sample having a certain size or characteristics (e.g., "indicator LOH regions" as defined herein); (2) determining the total number of TAI regions in the sample having a certain size or characteristics (e.g., "indicator TAI regions" as defined herein); (3) determining the total number of LST regions in the sample having a certain size or characteristics (e.g., "indicator LST regions" as defined herein); and (4) assessing (e.g., detecting) HRD in the sample based at least in part on the determinations made in (1), (2), and (3).

[0008] In one aspect, the present invention provides a method of diagnosing the presence or absence of HRD in a patient sample, the method comprising either: (1) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LOH regions of a certain size or characteristic in the sample (e.g., "indicator LOH regions" as defined herein); (2) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of TAI regions of a certain size or characteristic in the sample (e.g., "indicator TAI regions" as defined herein); and (3)(a) diagnosing the presence of HRD in the patient sample if the number from (1) and / or the number from (2) is above some reference number; or (3)(b) diagnosing the absence of HRD in the patient sample if neither the number from (1) nor the number from (2) is above some reference number. In another aspect, the present invention provides methods of diagnosing the presence or absence of HRD in a patient sample, the method comprising either: (1) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LOH regions of a certain size or characteristic in the sample (e.g., "indicator LOH regions" as defined herein); (2) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LST regions of a certain size or characteristic in the sample (e.g., "indicator LST regions" as defined herein); and (3)(a) diagnosing the presence of HRD in the patient sample if the number from (1) and / or the number from (2) is above some reference number; or (3)(b) diagnosing the absence of HRD in the patient sample if neither the number from (1) nor the number from (2) is above some reference number.In another aspect, the present invention provides methods for diagnosing the presence or absence of HRD in a patient sample, the method comprising either: (1) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of TAI regions in the samples having a certain size or characteristic (e.g., "indicator TAI regions" as defined herein); (2) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LST regions in the samples having a certain size or characteristic (e.g., "indicator LST regions" as defined herein); and (3)(a) diagnosing the presence of HRD in the patient sample if the number from (1) and / or the number from (2) is greater than some reference number; or (3)(b) diagnosing the absence of HRD in the patient sample if neither the number from (1) nor the number from (2) is greater than some reference number. In another aspect, the present invention provides a method of diagnosing the presence or absence of HRD in a patient sample, comprising: (1) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LOH regions having a certain size or characteristic in the sample (e.g., "indicator LOH regions" as defined herein); (2) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of TAI regions having a certain size or characteristic in the sample (e.g., "indicator TAI regions" as defined herein); and (3) analyzing (e.g., assaying) one or more patient samples to determine (e.g., detect) the total number of LST regions having a certain size or characteristic in the sample (e.g., "indicator LST regions" as defined herein); and either (3) (a) diagnosing the presence of HRD in the patient sample if the number from (1), the number from (2) and / or the number from (3) is above some reference number; or (3) (b) diagnosing the absence of HRD in the patient sample if none of the numbers from (1), (2) or (3) is above some reference number.

[0009] Various aspects of the present invention involve using the average (e.g., arithmetic mean) of three CA regions to assess (e.g., detect) HRD in a sample. Three CA regions useful in such methods include (1) chromosomal regions showing loss of heterozygosity ("LOH regions" as defined herein), (2) chromosomal regions showing telomeric allele imbalance ("TAI regions" as defined herein), and (3) chromosomal regions showing large-scale transitions ("LST regions" as defined herein). CA regions with certain sizes or characteristics (e.g., "indicator CA regions" as defined herein) may be particularly useful in various aspects of the present invention described herein. Thus, in one aspect, the present invention provides a method of assessing (e.g., detecting) HRD in a sample, comprising: (1) determining the total number of LOH regions in the sample having a certain size or characteristics (e.g., "indicator LOH regions" as defined herein); (2) determining the total number of TAI regions in the sample having a certain size or characteristics (e.g., "indicator TAI regions" as defined herein); (3) determining the total number of LST regions in the sample having a certain size or characteristics (e.g., "indicator LST regions" as defined herein); (4) calculating an average (e.g., arithmetic mean) of the determinations made in (1), (2), and (3); and (5) assessing HRD in the sample based at least in part on the calculated average (e.g., arithmetic mean) determined in (4).

[0010] In some embodiments, the assessment (e.g., detection) of HRD is based on a score ("CA region score" as defined herein) derived or calculated from (e.g., representative of or corresponding to) the detected CA region. Scores are described in more detail below. In some embodiments, HRD is detected when the CA region score for the sample is above some threshold (e.g., a reference score or index CA region score), and optionally HRD is not detected when the CA region score for the sample (e.g., a reference score or index CA region score, which in some embodiments may be the same threshold as for positive detection) is not above some threshold. Those skilled in the art will readily understand that scores in the opposite sense (e.g., HRD is detected when the CA region score is below a certain threshold and is not detected when the score is above a certain threshold) may be devised in this disclosure.

[0011] In some embodiments, the CA region score is a combination of scores derived or calculated from (e.g., representative of or corresponding to) two or more of: (1) the detected LOH regions ("LOH region score" as defined herein), (2) the detected TAI regions ("TAI region score" as defined herein), and / or (3) the detected LST regions ("LST region score" as defined herein). In some embodiments, the LOH region score and the TAI region score are combined to obtain the CA region score as follows: CA area score = A * (LOH region score)+B * (TAI domain score) In some embodiments, the LOH region score and the TAI region score are combined to obtain a CA region score as follows: CA area score = 0.32 * (LOH region score) +0.68 * (TAI domain score) In some embodiments, the LOH region score and the LST region score are combined to obtain a CA region score as follows: CA area score = A * (LOH region score)+B * (LST domain score) In some embodiments, the TAI and LST domain scores are combined to obtain a CA domain score as follows: CA area score = A * (TAI domain score)+B * (LST domain score) In some embodiments, the LOH region score, the TAI region score, and the LST region score are combined to obtain a CA region score as follows: CA area score = A * (LOH region score)+B * (TAI domain score)+C * (LST domain score) In some embodiments, the LOH region score, the TAI region score, and the LST region score are combined to obtain a CA region score as follows: CA area score = 0.21 * (LOH region score) +0.67 * (TAI domain score) +0.12 * (LST domain score)

[0012] In some embodiments, the CA region score is a combination of scores derived or calculated from (e.g., representative of or corresponding to) the average (e.g., arithmetic mean) of (1) the detected LOH regions ("LOH region score" as defined herein), (2) the detected TAI regions ("TAI region score" as defined herein), and / or (3) the detected LST regions ("LST region score" as defined herein) to obtain a CA region score: TIFF0007680995000001.tif9128

[0013] In another aspect, the present invention provides a method for predicting the status of BRCA1 and BRCA2 genes in a sample. Such a method is similar to the above-mentioned method, except that the determination of CA regions, LOH regions, TAI regions, LST regions, or scores incorporating these are used to evaluate (e.g., detect) BRCA1 and / or BRCA2 defects in a sample. In another aspect, the present invention provides a method for predicting the response of a cancer patient to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor. Such a method is similar to the above-mentioned method, except that the determination of CA regions, LOH regions, TAI regions, LST regions, or scores incorporating these are used to predict the likelihood that a cancer patient will respond to a cancer treatment regimen. In some embodiments, the patient is a treatment-naive patient. In another aspect, the present invention provides a method for treating cancer. Such methods are similar to the above methods, but differ in that a particular treatment regimen is administered (recommended, prescribed, etc.) based at least in part on the determination of CA regions, LOH regions, TAI regions, LST regions, or scores incorporating these. In another aspect, the present invention features the use of one or more drugs selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors in the manufacture of a medicament useful for treating cancer in a patient identified as having (or having had) cancer cells determined to have HRD (e.g., HRD signature) as described herein. In another aspect, this document features a method for evaluating a sample for the presence of a mutation in a gene from the HDR pathway. Such methods are similar to the above methods, but differ in that the determination of CA regions, LOH regions, TAI regions, LST regions, or scores incorporating these are used to detect the presence (or absence) of a mutation in a gene from the HDR pathway.

[0014] In another aspect, the present invention provides a method for evaluating a patient. The method includes or essentially consists of the steps of: (a) determining whether the patient has (or has had) cancer cells that exceed the reference number of CA regions (or, for example, CA region score that exceeds the reference CA region score); and (b)(1) diagnosing the patient has (or has had) cancer cells that exceed the reference number of CA regions (or, for example, CA region score that exceeds the reference CA region score) when the patient is determined to have (or has had) cancer cells that exceed the reference number of CA regions; or (b)(2) diagnosing the patient does not have (or has never had) cancer cells that exceed the reference number of CA regions (or, for example, the patient does not have (or has never had) cancer cells with CA region score that exceed the reference CA region score) when the patient is determined to not have (or has never had) cancer cells that exceed the reference number of CA regions).

[0015] In another aspect, the invention features the use of a plurality of oligonucleotides capable of hybridizing to a plurality of polymorphic regions of human genomic DNA in the manufacture of a diagnostic kit useful for determining the total number or combined length of CA regions in at least one pair of chromosomes (or DNA derived therefrom) in a sample obtained from a cancer patient, and also useful for detecting (a) HRD or likelihood of HRD in the sample (e.g., an HRD signature), (b) a defect (or likelihood of a defect) in the BRCA1 or BRCA2 gene in the sample, or (c) that a cancer patient is likely to respond to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor.

[0016] In another aspect, the invention features a system for detecting HRD (e.g., HRD signature) in a sample. The system includes or consists essentially of (a) a sample analyzer configured to generate a plurality of signals for genomic DNA of at least one pair of human chromosomes (or DNA derived therefrom) in the sample, and (b) a computer subsystem programmed to calculate the number or combined length of CA regions in at least one pair of human chromosomes based on the plurality of signals. The computer subsystem can be programmed to compare the number or combined length of CA regions to a reference number to detect (a) HRD or likelihood of HRD (e.g., HRD signature) in the sample, (b) loss (or likelihood of loss) of BRCA1 or BRCA2 genes in the sample, or (c) that a cancer patient is likely to respond to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor. The system may include an output module configured to display (a), (b), or (c). The system may include an output module configured to display a recommendation regarding the use of the cancer treatment regimen.

[0017] In another aspect, the present invention provides a computer program product embodied in a computer readable medium, which, when executed on a computer, provides instructions for detecting the presence or absence of any CA region (wherein the CA region is optionally an indicator CA region) present on one or more human chromosomes other than the X and Y sex chromosomes of humans; and for determining the total number or combined length of the CA regions in one or more pairs of chromosomes. The computer program product may also include other instructions.

[0018] In another aspect, the present invention provides a diagnostic kit. The kit comprises or consists essentially of at least 500 oligonucleotides that can hybridize with multiple polymorphic regions of human genomic DNA (or DNA derived therefrom); and a computer program product as provided herein. The computer program product may be embodied in a computer readable medium, and when executed on a computer, provides instructions for detecting the presence or absence of any CA region (wherein the CA region is optionally an indicator CA region) present on one or more human chromosomes other than human X and Y sex chromosomes; and for determining the total number or combined length of CA regions in one or more pairs of chromosomes. The computer program product may also include other instructions.

[0019] In some embodiments of any one or more of the aspects of the present invention described in the preceding paragraph, any one or more of the following may be applied as appropriate. CA regions can be determined in at least 2, 5, 10 or 21 pairs of human chromosomes. Cancer cells can be ovarian, breast, lung or esophageal cancer cells. Reference numbers can be 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 or 20 or more. At least one pair of human chromosomes can exclude human chromosome 17. The DNA damaging agent may be cisplatin, carboplatin, oxalaplatin, or picoplatin, the anthracycline may be epirubicin or doxorubicin, the topoisomerase I inhibitor may be campothecin, topotecan or irinotecan, or the PARP inhibitor may be iniparib, olaparib or velapirib. The patient may be a treatment-naïve patient.

[0020] Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention belongs.Methods and materials similar or equivalent to those described herein can be used in the practice of this invention, and suitable methods and materials are described below.All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety.In case of discrepancy, the present specification, including definitions, will prevail.In addition, materials, methods, and examples are merely illustrative and are not intended to be limiting.

[0021] The details of one or more embodiments of the invention are set forth in the following description and accompanying drawings. The materials, methods, and examples are illustrative only and are not intended to be limiting. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims. [Brief description of the drawings]

[0022] [Figure 1] Graphs showing allele dosages of breast cancer cells from fresh frozen samples from breast cancer patients plotted along the length of chromosomes, as determined using SNP arrays (top) and high-throughput sequencing (bottom). [Diagram 2] Graphs showing allele dosages of breast cancer cells from FFPE samples derived from breast cancer patients plotted along the length of chromosomes, as determined using SNP arrays (top) and high-throughput sequencing (bottom). [Diagram 3] 1 is a flow diagram of an exemplary process for evaluating the genome of a cell (e.g., a cancer cell) for an HRD signature. [Figure 4] 1 is a schematic diagram of an example of a computing device and a mobile computing device that can be used to implement the techniques described herein. [Figure 5A]The LOH region scores for each IHC subtype of breast cancer are shown. The top three panels are BRCA1 / 2-deficient samples. The bottom panel is a BRCA1 / 2-intact sample. [Figure 5B] The TAI region scores for different IHC subtypes of breast cancer are shown. The top three panels are BRCA1 / 2-deficient samples. The bottom panel is a BRCA1 / 2-intact sample. [Figure 6] Correlation between LOH region score and TAI region score is shown. Correlation coefficient=0.69. X-axis: LOH score; Y-axis: TAI score; Red dots: intact samples; Blue dots (with "X" overlaid): BRCA1 / 2 deletion samples. The area under the dots is proportional to the number of samples with the combination of LOH and TAI scores. p=10-39. [Figure 7A] Figure 1 shows the LOH region scores for the patients analyzed in Example 2 herein. The top three panels are BRCA1 / 2-deficient samples. The bottom panel is a BRCA1 / 2-intact sample. [Figure 7B] Figure 1 shows the TAI region scores for patients analyzed in Example 2 herein. The top three panels are BRCA1 / 2 deficient samples. The bottom panel is a BRCA1 / 2 intact sample. [Figure 7C] Figure 1 shows the LST region scores for the patients analyzed in Example 2 herein. The top three panels are BRCA1 / 2 deficient samples. The bottom panel is a BRCA1 / 2 intact sample. [Figure 7D] Shows LOH vs. TAI for patients analyzed in Example 2 herein. X-axis: LOH score; Y-axis: TAI score; Red dots: intact samples; Blue dots (with "X" overlaid): BRCA1 / 2 deletion samples. Area under the dots is proportional to the number of samples with the combined LOH and TAI scores. [Figure 7E]Shows LOH vs LST for patients analyzed in Example 2 herein. X-axis: LOH score; Y-axis: LST score; Red dots: intact samples; Blue dots (with "X" overlaid): BRCA1 / 2 deletion samples. Area under the dots is proportional to the number of samples with the combined LOH and LST scores. [Figure 7F] Shows TAI vs. LST for patients analyzed in Example 2 herein. X-axis: TAI score; Y-axis: LST score; Red dots: intact samples; Blue dots (with "X" overlaid): BRCA1 / 2-deficient samples. Area under the dots is proportional to the number of samples with the TAI and LST score combination. [Figure 8] 1 is a graph plotting the number of LOH regions longer than 15 Mb and shorter than the entire chromosome for ovarian cancer cell samples with somatic BRCA mutations, germline BRCA mutations, BRCA1 low expression, or intact BRCA (normal BRCA). The size of the circle is proportional to the number of samples with such a number of LOH regions. [Figure 9A] HRD-LOH scores are shown in BRCA 1 / 2 null (mutated or methylated) samples (upper panel) and intact samples (lower panel) in the whole-participant breast cohort. [Figure 9B] HRD-TAI scores are illustrated in BRCA1 / 2 null (mutated or methylated) samples (upper panel) and intact samples (lower panel) in the whole-participant breast cohort. [Figure 9C] HRD-LST scores are illustrated in BRCA1 / 2 null (mutated or methylated) samples (upper panel) and intact samples (lower panel) in the whole-participant breast cohort. [Figure 10] Illustrates mean (eg, arithmetic mean) HRD-composite scores (Y-axis) stratified by Miller-Payne score (X-axis) in the combined cisplatin-1 and cisplatin-2 cohorts. [Figure 11]Figure 1 shows the Spearman correlations of three different measures of HR deficit. The upper diagonal panel shows the correlations. The diagonal panel shows the density plots. [Figure 12] The associations between clinical variables and HRD-composite scores are illustrated. [Figure 13] Associations between clinical variables and BRCA1 / 2 defects are illustrated. The top two panels and the bottom left panel show the proportion of BRCA1 / 2 defective patients in each category of grade, stage, and breast cancer type. The width of each bar is proportional to the number of patients in each category. The bottom right panel shows conditional density estimates of age at diagnosis of BRCA1 / 2 defects. [Figure 14] Illustrates the determination of high HRD with a reference score of 42 or greater. [Figure 15] A histogram showing the distribution of HRD scores in the cisplatin cohort is shown, with the four columns on the left representing low HRD and the five columns on the right representing high HRD, with a reference score above 42. [Figure 16] The distribution of HRD scores in pCR, RCB-I, RCB-II and RCB-III class responses is illustrated. Boxes represent the interquartile range (IQR) of scores and horizontal lines represent medians. The dotted line at 42 represents the HRD threshold between low and high scores. [Figure 17] Response curves for quantitative HRD scores are illustrated. The curves are modeled by generalized logistic regression. The dark boxes indicate the probability of response in HR-deficient versus non-deficient samples. [Figure 18] HRD scores for individual HRD components (LOH, TAI, and LST) are shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] Detailed Description In general, one aspect of the present invention features a method for evaluating HRD in cancer cells or DNA (e.g., genomic DNA) derived therefrom. In some embodiments, the method comprises or consists essentially of: (a) determining CA regions in at least one pair of human chromosomes or DNA derived therefrom in a sample or DNA derived therefrom; and (b) determining the number, size (e.g., length), and / or characteristics of the CA regions.

[0024] As used herein, "chromosomal abnormality" or "CA" refers to a somatic change in the chromosomal DNA of a cell that falls into at least one of three overlapping categories: LOH, TAI, or LST. Polymorphic loci (e.g., single nucleotide polymorphisms (SNPs)) in the human genome are generally heterozygous in an individual's germline because an individual typically receives one copy from the biological father and one copy from the biological mother. However, in somatic cells, this heterozygosity can change (through mutation) to homozygosity. This change from heterozygosity to homozygosity is called loss of heterozygosity (LOH). LOH can occur by several mechanisms. For example, in some cases, a locus on one chromosome can be deleted in a somatic cell. Since only one copy (instead of two copies) of that locus is present in the genome of the affected cell, the locus that is still present on the other chromosome (the other non-sex chromosome in the case of males) will be LOH. This type of LOH event results in a reduction in copy number. In other cases, a locus on one chromosome in a somatic cell (e.g., one of the non-sex chromosomes in the case of males) may be replaced by a copy of the locus from the other chromosome, thereby eliminating any heterozygosity that may have existed in the replaced locus. In such cases, the locus still present on each chromosome becomes a LOH locus, which can be referred to as a copy neutral LOH locus. LOH and its use in determining HRD are described in detail in International Application No. PCT / US2011 / 040953 (published as WO / 2011 / 160063), the entire contents of which are incorporated herein by reference.

[0025] A broader class of chromosomal abnormalities that encompasses LOH is allelic imbalance. Allelic imbalance is when the relative copy number (i.e., the ratio of copies) at a particular locus in somatic cells is different from that in the germline. For example, if the germline has one copy of allele A and one copy of allele B at a particular locus, and the somatic cell has two copies of A and one copy of B, there is allelic imbalance at that locus because the ratio of copies in the somatic cell (2:1) is different from that in the germline (1:1). LOH is an example of allelic imbalance because somatic cells have a different ratio of copies (1:0 or 2:0) than that in the germline (1:1). However, allelic imbalance encompasses many more types of chromosomal abnormalities, such as 2:1 in the germline to 1:1 in the somatic cell; 1:0 in the germline to 1:1 in the somatic cell; 1:1 in the germline to 2:1 in the somatic cell, etc. The analysis of the region of allele imbalance in the telomere of chromosome is particularly useful in the present invention.Therefore, "telomere allele imbalance region" or "TAI region" is defined as a region with allele imbalance that (a) extends toward one of the subtelomeres, but (b) does not cross the centromere.TAI and its use in determining HRD are described in detail in U.S. Patent Application No. 13 / 818,425 (published as US20130281312A1) and No. 14 / 466,208 (published as US20150038340A1), the entire contents of each of which are incorporated herein by reference.

[0026] A broader class of chromosomal abnormalities, but still including LOH and TAI, is referred to herein as large-scale transition ("LST"). LST refers to any somatic copy number transition (i.e., discontinuity) along the length of chromosome, which exists between two regions that have at least some minimum length (e.g., at least 3, 4, 5, 6, 7, 8 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 megabases or more) after filtering out regions that are shorter than some maximum length (e.g., 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4 megabases or more). For example, if, after filtering out regions shorter than 3 megabases, a somatic cell has a copy number of 1:1 over, for example, at least 10 megabases, followed by a discontinuity transition to, for example, a region of at least 10 megabases where the copy number is 2:2, then this is an LST. An alternative way of defining the same phenomenon is as an LST region, which refers to a genomic region that has stable copy number over at least some minimum length (e.g., at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 megabases) and has a discontinuity (i.e., transition) bordering it where the copy number changes towards another region also having at least this minimum length. For example, if, after filtering out regions shorter than 3 megabases, a somatic cell has a region of at least 10 megabases with a copy number of 1:1, with a breakpoint transition at one boundary to a region of at least 10 megabases with a copy number of, for example, 2:2, and at the other boundary to a region of at least 10 megabases with a copy number of, for example, 1:2, then this is two LSTs. Note that this is more extensive than allelic imbalance, since such copy number changes would not be considered allelic imbalance (because the copy ratios of 1:1 and 2:2 are the same, i.e., there is no change in the copy ratio).LST and its use in determining HRD are described in detail in U.S. Patent Application No. 14 / 402,254 (published as US20150140122A1), the entire contents of which are incorporated herein by reference.

[0027] Various cutoffs for LST score can be used for "near-diploid" and "near-tetraploid" tumors to separate BRCA1 / 2 intact and BRCA1 / 2 defective samples. LST score sometimes increases with ploidy in both intact and defective samples. As an alternative to using ploidy-specific cutoffs, some embodiments can use a modified LST score adjusted by ploidy: LSTm=LST-kP, where P is ploidy and k is a constant. Based on multivariate logistic regression analysis with defective as outcome and LST and P as predictors, k=15.5, provided that it provides the best separation between intact and defective samples (however, other values ​​of k can be envisioned by those skilled in the art).

[0028] Chromosomal abnormalities can be spread across many loci to define a region of chromosomal abnormality, referred to herein as a "CA region". Such CA regions can be of any length (e.g., from less than about 1.5 Mb to the full length of a chromosome). The presence of many large CA regions ("indicator CA regions") indicates a defect in the cellular homology-dependent repair (HDR) mechanism. For each type of CA (e.g., LOH, TAI, LST), the definition of what constitutes a region of CA, and therefore an "indicator" region, depends on the unique characteristics of the CA. For example, a "LOH region" refers to at least some minimum number of consecutive loci that exhibit LOH, or some minimum stretch of genomic DNA with consecutive loci that exhibit LOH. "TAI region," on the other hand, refers to at least some minimum number of contiguous loci that exhibit allelic imbalance extending from the telomere into the remainder of the chromosome (or some minimum stretch of genomic DNA having contiguous loci that exhibit allelic imbalance extending from the telomere into the remainder of the chromosome). LSTs have been defined above in terms of a region of genomic DNA of at least some minimum size, so "LST" and "LST region" are used interchangeably in this document to refer to the minimum number of contiguous loci (or some minimum stretch of genomic DNA) with the same copy number that border a discontinuity or transition from that copy number to another.

[0029] In some embodiments, a CA region (either an LOH region, a TAI region, or an LST region) is an indicator CA region (either an indicator LOH region, an indicator TAI region, or an indicator LST region) if it is at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, or 100 megabases or more in length. In some embodiments, an indicator LOH region is an LOH region that is about 1.5, 5, 12, 13, 14, 15, 16, 17 megabases or more in length (preferably 14, 15, 16 megabases or more, more preferably 15 megabases or more in length), but less than the full length of the respective chromosome in which it is located. Alternatively or additionally, the total length of such indicator LOH regions may be determined. In some embodiments, the indicator TAI region is a TAI region with allelic imbalance that (a) spans one of the subtelomeres, (b) does not cross the centromere, and (c) is 1.5, 5, 12, 13, 14, 15, 16, 17 megabases or longer (preferably 10, 11, 12 megabases or longer, more preferably 11 megabases or longer). Alternatively or additionally, the total length of such indicator TAI regions may be determined. Since the concept of LST already includes a region with some minimum size (such a minimum size is determined based on its ability to distinguish HRD from HDR intact samples), the indicator LST region as used herein is the same as the LST region. Moreover, the LST region score can be derived from either the number of regions showing LST or the number of LST discontinuities as described above.In some embodiments, the minimum length of a region of stable copy number bordering an LST breakpoint is at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 megabases (preferably 8, 9, 10, 11 megabases or more, more preferably 10 megabases) and the maximum region remaining unscreened is 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4 megabases or less (preferably 2, 2.5, 3, 3.5 or 4 megabases or less, more preferably less than 3 megabases).

[0030] As used herein, a sample has an "HRD signature" if such sample has a number of indicator CA regions (as described herein) or a CA region score (as described herein) that exceeds a reference number as described herein, where a number or score exceeding such reference number is indicative of a homologous recombination deficiency.

[0031] Thus, the invention generally involves detecting and quantifying indicator CA regions in a sample to determine whether cells in the sample (or cells from which DNA in the sample is derived) have an HRD signature. Often this involves comparing the number of indicator CA regions (or a test value or score derived or calculated therefrom and corresponding to such a number) to a reference or index number (or score).

[0032] Various aspects of the invention include using a combined analysis of two or more CA regions (including two or more indicator CA regions) to assess (e.g., detect, diagnose) HRD in a sample. Thus, in one aspect, the invention provides a method of assessing (e.g., detect, diagnose) HRD in a sample, comprising: (1) determining the total number (or combined length) of indicator LOH regions in the sample; (2) determining the total number (or combined length) of indicator TAI regions in the sample; and (3) determining (e.g., detect, diagnose) the presence or absence of HRD in the sample based at least in part on the determinations made in (1) and (2). In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining the total number (or combined length) of indicator LOH regions in the sample; (2) determining the total number (or combined length) of indicator LST regions in the sample; and (3) determining (e.g., detecting, diagnosing) the presence or absence of HRD in the sample based at least in part on the determinations made in (1) and (2). In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining the total number (or combined length) of indicator TAI regions in the sample; (2) determining the total number (or combined length) of indicator LST regions in the sample; and (3) determining (e.g., detecting, diagnosing) the presence or absence of HRD in the sample based at least in part on the determinations made in (1) and (2).In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining the total number (or combined length) of indicator LOH regions in the sample; (2) determining the total number (or combined length) of indicator TAI regions in the sample; (3) determining the total number (or combined length) of indicator LST regions in the sample; and (4) determining (e.g., detecting, diagnosing) the presence or absence of HRD in the sample based at least in part on the determinations made in (1), (2), and (3).

[0033] Various aspects of the invention include using a combined analysis of the average of three different CA regions to assess (e.g., detect, diagnose) HRD in a sample. Thus, in one aspect, the invention provides a method of assessing (e.g., detect, diagnose) HRD in a sample, comprising: (1) determining the total number of LOH regions (e.g., "indicator LOH regions" as defined herein) in the sample having a certain size or characteristic; (2) determining the total number of TAI regions (e.g., "indicator TAI regions" as defined herein) in the sample having a certain size or characteristic; (3) determining the total number of LST regions (e.g., "indicator LST regions" as defined herein) in the sample having a certain size or characteristic; (4) calculating the average (e.g., arithmetic mean) of the determinations made in (1), (2), and (3); and (5) assessing HRD in the sample based at least in part on the calculated average (e.g., arithmetic mean) determined in (4).

[0034] As used herein, "CA region score" refers to a test value or score derived or calculated (e.g., representative of or corresponding to) the indicator CA region detected in a sample (e.g., a score or test value derived or calculated from the number of indicator CA regions detected in a sample). Similarly, as used herein, "LOH region score" refers to a test value or score that is a subset of the CA region score and derived or calculated (e.g., representative of or corresponding to) the indicator LOH region detected in a sample (e.g., a score or test value derived or calculated from the number of indicator LOH regions detected in a sample), as well as the TAI region score and the LST region score. In some embodiments, such a score may simply be the number of indicator CA regions detected in a sample. In some embodiments, the score is more complex and factors in the length of each indicator CA region or a subset of indicator CA regions detected.

[0035] As discussed above, the present invention generally involves combining the analysis of two or more CA region scores (which may include the number of such regions).Thus, in one aspect, the present invention provides a method for evaluating (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining the LOH region score for the sample; (2) determining the TAI region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample, at least in part, based on either the LOH region score exceeding the reference number or the TAI region score exceeding the reference number; or optionally (3)(b) detecting (or diagnosing) the absence of HRD in the sample, at least in part, based on both the LOH region score not exceeding the reference number and the TAI region score not exceeding the reference number. In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining an LOH region score for the sample; (2) determining an LST region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on either an LOH region score above a reference number or an LST region score above a reference number; or, optionally, (3)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on both an LOH region score not above the reference number and an LST region score not above the reference number. In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining a TAI region score for the sample; (2) determining an LST region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on either a TAI region score that exceeds a reference number or an LST region score that exceeds the reference number; or optionally (3)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on both a TAI region score that does not exceed the reference number and an LST region score that does not exceed the reference number.In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining an LOH region score for the sample; (2) determining a TAI region score for the sample; (3) determining an LST region score for the sample; and (4)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on either an LOH region score above a reference number, a TAI region score above a reference number, or an LST region score above a reference number; or optionally (4)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on an LOH region score not above the reference number, a TAI region score not above the reference number, and an LST region score not above the reference number.

[0036] In some embodiments, the CA region score is a combination of scores derived or calculated from (e.g., representative of or corresponding to) two or more of: (1) the detected LOH regions ("LOH region score" as defined herein), (2) the detected TAI regions ("TAI region score" as defined herein), and / or (3) the detected LST regions ("LST region score" as defined herein). In some embodiments, the LOH region score and the TAI region score are combined to obtain the CA region score as follows: CA area score = A * (LOH region score)+B * (TAI domain score) In some embodiments, the LOH region score and the TAI region score are combined to obtain a CA region score as follows: CA area score = 0.32 * (LOH region score) +0.68 * (TAI domain score) or CA area score = 0.34 * (LOH region score) +0.66 * (TAI domain score) In some embodiments, the LOH region score and the LST region score are combined to obtain a CA region score as follows: CA area score = A * (LOH region score)+B * (LST domain score) In some embodiments, the LOH region score for a sample and the LST region score for a sample are combined to obtain a CA region score as follows: CA area score = 0.85 * (LOH region score) +0.15 * (LST domain score) In some embodiments, the TAI and LST domain scores are combined to obtain a CA domain score as follows: CA area score = A * (TAI domain score)+B * (LST domain score) In some embodiments, the LOH region score, the TAI region score, and the LST region score are combined to obtain a CA region score as follows: CA area score = A * (LOH region score)+B * (TAI domain score)+C * (LST domain score) In some embodiments, the LOH region score, the TAI region score, and the LST region score are combined to obtain a CA region score as follows: CA area score = 0.21 * (LOH region score) +0.67 * (TAI domain score) +0.12 * (LST domain score) or CA area score = [0.24] * (LOH region score) + [0.65] * (TAI domain score)+[0.11] * (LST domain score) or CA area score = [0.11] * (LOH region score) + [0.25] * (TAI domain score)+[0.12] * (LST domain score)

[0037] In some embodiments, the CA region score is a combination of scores derived or calculated from (e.g., representative of or corresponding to) the average (e.g., arithmetic mean) of (1) the detected LOH regions ("LOH region score" as defined herein), (2) the detected TAI regions ("TAI region score" as defined herein), and / or (3) the detected LST regions ("LST region score" as defined herein) to obtain a CA region score from one of the following formulas: TIFF0007680995000002.tif54152 In some embodiments, including some specifically exemplified herein, one or more of these coefficients (i.e., A, B, or C, or any combination thereof) are 1, and in some embodiments, all three coefficients (i.e., A, B, and C) are 1. Thus, in some embodiments, CA region score = (LOH region score) + (TAI region score) + (LST region score), where LOH region score is the number of indicator LOH regions (or total length of LOH), TAI region score is the number of indicator TAI regions (or total length of TAI), and LST region score is the number of indicator LST regions (or total length of LST).

[0038] In some cases, a formula may not have all of the coefficients specified (and therefore may not incorporate corresponding variables). For example, the embodiment just mentioned may be applied to formula (2) where A in formula (2) is 0.95 and B in formula (2) is 0.61. C and D are considered inapplicable as these coefficients and their corresponding variables are not found in formula (2) (although the clinical variables are incorporated in the clinical score found in formula (2)). In some embodiments, A is 0.9 to 1, 0.9 to 0.99, 0.9 to 0.95, 0.85 to 0.95, 0.86 to 0.94, 0.87 to 0.93, 0.88 to 0.92, 0.89 to 0.91, 0.85 to 0.9, 0.8 to 0.95, 0.8 to 0.9, 0.8 to 0.85, 0.75 to 0.99, 0.75 to 0.95, 0.75 to 0.9, 0.75 to 0.85, or 0.75 to 0.8. In some embodiments, B is 0.40 to 1, 0.45 to 0.99, 0.45 to 0.95, 0.55 to 0.8, 0.55 to 0.7, 0.55 to 0.65, 0.59 to 0.63, or 0.6 to 0.62. In some embodiments, C is between 0.9 and 1, between 0.9 and 0.99, between 0.9 and 0.95, between 0.85 and 0.95, between 0.86 and 0.94, between 0.87 and 0.93, between 0.88 and 0.92, between 0.89 and 0.91, between 0.85 and 0.9, between 0.8 and 0.95, between 0.8 and 0.9, between 0.8 and 0.85, between 0.75 and 0.99, between 0.75 and 0.95, between 0.75 and 0.95, between 0.75 and 0.9, between 0.75 and 0.85, or between 0.75 and 0.8, as applicable. In some embodiments, D is, as applicable, 0.9 to 1, 0.9 to 0.99, 0.9 to 0.95, 0.85 to 0.95, 0.86 to 0.94, 0.87 to 0.93, 0.88 to 0.92, 0.89 to 0.91, 0.85 to 0.9, 0.8 to 0.95, 0.8 to 0.9, 0.8 to 0.85, 0.75 to 0.99, 0.75 to 0.95, 0.75 to 0.9, 0.75 to 0.85, or 0.75 to 0.8.

[0039] In some embodiments, A is between 0.1 and 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 20; or between 0.2 and 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 20; or between 0.3 and 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1 or between 0.4 and 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.5 and 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.6 and 0.7, 0 or between 0.7 and 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.8 and 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.8 and 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.9 and 1, 1.5, 2, 2.5, 3, 3.5, 4 , 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1 and 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1.5 and 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2 and 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2.5 and 3, 3.5, 4, 4.or between 3 and 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3.5 and 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4 and 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4.5 and 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 5 and 6, 7, 8 , 9, 10, 11, 12, 13, 14, 15 or 20; or between 6 and 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 7 and 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 8 and 9, 10, 11, 12, 13, 14, 15 or 20; or between 9 and 10, 11, 12, 13, 14, 15 or 20; or between 10 and 11, 12, 13, 14, 15 or 20; or between 11 and 12, 13, 14, 15 or 20; or between 12 and 13, 14, 15 or between 13 and 14, 15 or 20; or between 14 and 15 or 20; or between 15 and 20; B is between 0.1 and 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.2 and 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 1 or between 0.3 and 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.4 and 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.5 and 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.or between 0.6 and 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.7 and 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 , 15 or 20; or between 0.8 and 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.9 and 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1 and 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, or between 1.5 and 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2 and 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2.5 and 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2.5 and 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3 and 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3.5 and 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4 and 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or 4.between 5 and 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20;or between 5 and 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20;or between 6 and 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20;or between 7 and 8, 9, 10, 11, 12, 13, 14, 15 or 20;or between 8 and 9, 10, 11, 12, 13, 14, 15 or 20;or between 9 and 10, 11, 12, 13, 14, 15 or 20;or between 10 and 11, 12, 13, 14, 15 or 20; or between 11 and 12, 13, 14, 15 or 20; or between 12 and 13, 14, 15 or 20; or between 13 and 14, 15 or 20; or between 14 and 15 or 20; or between 15 and 20;C is, as applicable, between 0.1 and 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.2 and 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, or between 0.3 and 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.4 and 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.5 and 0.6, or between 0.6 and 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.7 and 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.7 and 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.8 and 0.9, 1, 1.5, 2, 2.or between 0.9 and 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1 and 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1.5 and 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2, or between 2.5 and 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3 and 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3.5 and 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4 and 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20 or between 4.5 and 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 5 and 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 6 and 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 7 and 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 8 and 9, 10, 11, 12, 13, 14, 15 or 20; or between 9 and 10, 11, 12, 13, 14, 15 or 20; or between 10 and 11, 12, 13, 14 , 15 or 20; or between 11 and 12, 13, 14, 15 or 20; or between 12 and 13, 14, 15 or 20; or between 13 and 14, 15 or 20; or between 14 and 15 or 20; or between 15 and 20; and D is, as applicable, between 0.1 and 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.2 and 0.3, 0.4, 0.5, 0.or between 0.3 and 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.4 and 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.5 and 0.6, 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.6 and 0.7, 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.7 and 0.8, 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20 or between 0.8 and 0.9, 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 0.9 and 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1 and 1.5, 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 1.5 and 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20 between 1, 12, 13, 14, 15 or 20; or between 2 and 2.5, 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 2.5 and 3, 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3 and 3.5, 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 3.5 and 4, 4.5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4 and 4.between 5, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 4.5 and 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 5 and 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 6 and 7, 8, 9, 10, 11, 12, 13, 14, 15 or 20; or between 7 and 8, 9, 10, 11, 12, 13, 14, 15 or 20; and is between 8 and 9, 10, 11, 12, 13, 14, 15 or 20; or between 9 and 10, 11, 12, 13, 14, 15 or 20; or between 10 and 11, 12, 13, 14, 15 or 20; or between 11 and 12, 13, 14, 15 or 20; or between 12 and 13, 14, 15 or 20; or between 13 and 14, 15 or 20; or between 14 and 15 or 20; or between 15 and 20. In some embodiments, A, B, and / or C are within the rounding range of any of these values ​​(e.g., A is between 0.45 and 0.54, etc.).

[0040] Thus, in one aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining an LOH region score for the sample; (2) determining a TAI region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on a combination of the LOH region score and the TAI region score that exceeds a reference number (e.g., a combined CA region score); or optionally (3)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on a combination of the LOH region score and the TAI region score that does not exceed a reference number (e.g., a combined CA region score). In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining an LOH region score for the sample; (2) determining an LST region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on a combination of the LOH region score and the LST region score (e.g., a combined CA region score) that exceeds a reference number; or, optionally, (3)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on a combination of the LOH region score and the LST region score (e.g., a combined CA region score) that does not exceed a reference number. In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining a TAI region score for the sample; (2) determining an LST region score for the sample; and (3)(a) detecting (or diagnosing) HRD in the sample based, at least in part, on a combination of the TAI region score and the LST region score (e.g., a composite CA region score) that exceeds a reference number; or optionally (3)(b) detecting (or diagnosing) the absence of HRD in the sample based, at least in part, on a combination of the TAI region score and the LST region score (e.g., a composite CA region score) that does not exceed a reference number.In another aspect, the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining an LOH region score for the sample; (2) determining a TAI region score for the sample; (3) determining an LST region score for the sample; and (4)(a) detecting (or diagnosing) HRD in the sample based at least in part on a combination of the LOH region score, the TAI region score and the LST region score (e.g., a combined CA region score) that exceeds a reference number; or optionally (4)(b) detecting (or diagnosing) the absence of HRD in the sample based at least in part on a LOH region score, the TAI region score and the LST region score (e.g., a combined CA region score) that does not exceed a reference number.

[0041] Accordingly, another aspect of the present invention provides a method of assessing (e.g., detecting, diagnosing) HRD in a sample, comprising: (1) determining the total number of LOH regions in the sample having a certain size or characteristics (e.g., "indicator LOH regions" as defined herein); (2) determining the total number of TAI regions in the sample having a certain size or characteristics (e.g., "indicator TAI regions" as defined herein); (3) determining the total number of LST regions in the sample having a certain size or characteristics (e.g., "indicator LST regions" as defined herein); (4) calculating an average (e.g., arithmetic mean) of the determinations made in (1), (2), and (3); and (5) assessing HRD in the sample based at least in part on the calculated average (e.g., arithmetic mean) determined in (4).

[0042] In some embodiments, the reference (or index) discussed above with respect to the CA region score (e.g., number of indicator CA regions) may be 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20 or more, preferably 5, preferably 8, more preferably 9 or 10, and most preferably 10. References to the total (e.g., combined) length of the indicator CA regions may be about 75, 90, 105, 120, 130, 135, 150, 175, 200, 225, 250, 275, 300, 325 350, 375, 400, 425, 450, 475, 500 megabases or more, preferably about 75 megabases or more, preferably about 90 or 105 megabases or more, more preferably about 120 or 130 megabases or more, more preferably about 135 megabases or more, and most preferably about 150 megabases or more. In some embodiments, the references discussed above regarding the combined CA region score (e.g., the combined number of indicator LOH regions, indicator, TAI regions and / or indicator LST regions) may be 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50 or more, preferably 5, preferably 10, preferably 15, preferably 20, preferably 25, preferably 30, preferably 35, preferably 40-44, most preferably 42 or more.References to the total (e.g., combined) length of the indicator LOH region, indicator TAI region and / or indicator LST region may be about 75, 90, 105, 120, 130, 135, 150, 175, 200, 225, 250, 275, 300, 325 350, 375, 400, 425, 450, 475, 500 megabases or more, preferably about 75 megabases or more, preferably about 90 or 105 megabases or more, more preferably about 120 or 130 megabases or more, more preferably about 135 megabases or more, and most preferably about 150 megabases or more.

[0043] In some embodiments, the present invention provides a method for detecting an HRD signature in a sample. Accordingly, another aspect of the present invention provides a method for detecting an HRD signature in a sample, comprising: (1) determining the total number of LOH regions (e.g., "indicator LOH regions" as defined herein) in the sample with a certain size or characteristic; (2) determining the total number of TAI regions (e.g., "indicator TAI regions" as defined herein) in the sample with a certain size or characteristic; (3) determining the total number of LST regions (e.g., "indicator LST regions" as defined herein) in the sample with a certain size or characteristic; (4) combining the determinations made in (1), (2) and (3) (e.g., calculating or deriving a composite CA region score); and (5) characterizing a sample with a composite CA region score greater than a reference value as having an HRD signature. In some embodiments, the reference value is 42. Thus, in some embodiments, a sample is characterized as having an HRD signature when the reference value is 42. In some embodiments, the reference number discussed above with respect to the composite CA region score (e.g., the combined number of indicator LOH regions, indicator, TAI regions and / or indicator LST regions) may be 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 18, 19, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50 or more, preferably 5, preferably 10, preferably 15, preferably 20, preferably 25, preferably 30, preferably 35, preferably 40-44, and most preferably 42 or more.

[0044] In some embodiments, the number of indicator CA regions (or combined length, CA region score, or composite CA region score) in a sample is determined to be "greater" than the reference if it is at least 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater than the reference, and in some embodiments, it is determined to be "greater" than the reference if it is at least 1-, 2-, 3-, 4-, 5-, 6-, 7-, 8-, 9-fold, or 10-fold greater than the reference. Conversely, in some embodiments, the number of indicator CA regions (or combined length, CA region score, or composite CA region score) in a sample is determined to be "not greater than" the reference if it is not 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater than the reference, and in some embodiments, it is determined to be "not greater than" if it is not 1-, 2-, 3-, 4-, 5-, 6-, 7-, 8-, 9-fold, or 10-fold greater than the reference.

[0045] In some embodiments, the reference number (or length, value or score) is derived from a valid reference population. Such reference population can include: (a) patients with the same cancer as the patient being tested, (b) patients with the same cancer subtype, (c) patients with cancer with similar genetic characteristics or other clinical or molecular characteristics, (d) patients who have responded to a certain treatment, (e) patients who have not responded to a certain treatment, (f) patients who are apparently healthy (e.g., patients who do not have any cancer, or at least those who do not have the cancer of the patient being tested), etc. The reference number (or length, value or score) may be (a) representative of the number (or length, value or score) found in the reference population as a whole, (b) the average (mean, median, etc.) of the numbers (or lengths, values ​​or scores) found in the reference population as a whole or in a particular subpopulation, (c) representative of the numbers (or lengths, values ​​or scores) found in the reference population as a tertile, quartile, quintile, etc., as ranked by (i) their respective numbers (or lengths, values ​​or scores) or (ii) the clinical features they are found to have (e.g., strength of response, prognosis (including time to cancer-specific death), etc.), or (d) selected to have high sensitivity for detecting HRD to predict response to a particular therapy (e.g., platinum, PARP inhibitors, etc.).

[0046] In some embodiments, if the reference or index is exceeded by the test value or score from the sample, it indicates that the HRD is the same as the reference, and if it is not exceeded by the test value or score from the sample, it indicates that the HRD (or functional HDR) is not present. In some embodiments, they are different.

[0047] In another aspect, the present invention provides a method for predicting the status of BRCA1 and BRCA2 genes in samples.Such method is similar to the above-mentioned method, but differs in that CA region, LOH region, TAI region, LST region, or the score that incorporates them is used to evaluate (e.g., detect) BRCA1 and / or BRCA2 defects in samples.

[0048] In another aspect, the present invention provides a method for predicting the response of a cancer patient to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor. Such a method is similar to the above-mentioned method, except that the determination of CA regions, LOH regions, TAI regions, LST regions, or scores incorporating these, including a high HRD score (e.g., HRD signature or high combined CA region score), is used to predict the likelihood that a cancer patient will respond to a cancer treatment regimen.

[0049] In some embodiments, the patient is an untreated patient.In another aspect, the present invention provides a method for treating cancer.Such method is similar to the above-mentioned method, but differs in that a certain treatment regimen is administered (recommended, prescribed, etc.) based at least in part on the determination of CA region, LOH region, TAI region, LST region, or the score that incorporates them.

[0050] In another aspect, the invention features the use of one or more drugs selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors in the manufacture of a medicament useful for treating cancer in a patient identified as having (or having had) cancer cells determined to have high levels of HRD (e.g., an HRD signature) as described herein.

[0051] In another aspect, this document features a method for evaluating a sample for the presence of mutations in genes from HDR pathway.Such a method is similar to the above-mentioned method, but differs in that CA region, LOH region, TAI region, LST region, or score determination incorporating these are used to detect the presence (or absence) of mutations in genes from HDR pathway.

[0052] In another aspect, this document features a method for evaluating a patient's cancer cells for the presence of an HRD signature. The method comprises or consists essentially of: (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes in a cancer patient's cancer cells that are greater than a reference number; and (b) identifying the patient as having cancer cells with an HRD signature. In another aspect, this document features a method for evaluating a patient's cancer cells for the presence of an HDR-deficient state. The method comprises or consists essentially of: (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes in a cancer patient's cancer cells that are greater than a reference number; and (b) identifying the patient as having cancer cells with an HDR-deficient state. In another aspect, this document features a method for evaluating a patient's cancer cells for the presence of an HRD signature. The method comprises or consists essentially of: (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of cancer cells of a cancer patient that are greater than a reference number; and (b) identifying the patient as having cancer cells with HRD signature. In another aspect, the document features a method for evaluating the presence of genetic mutations in genes from HDR pathways in cancer cells of a patient. The method comprises or consists essentially of: (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of cancer cells of a cancer patient that are greater than a reference number; and (b) identifying the patient as having cancer cells with genetic mutations.

[0053] In another aspect, this document features a method for determining whether a patient is likely to respond to a cancer treatment regimen comprising administering radiation or a drug selected from the group consisting of a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, and a PARP inhibitor. The method comprises or consists essentially of (a) detecting the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in cancer cells of a cancer patient, and (b) identifying the patient as likely to respond to the cancer treatment regimen. In another aspect, this document features a method for evaluating a patient. The method comprises or consists essentially of (a) determining that the patient contains cancer cells with an HRD signature, where the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in cancer cells of the cancer patient indicates that the cancer cells have an HRD signature, and (b) diagnosing the patient as having cancer cells with an HRD signature. In another aspect, this document features a method for evaluating a patient. The method includes or consists essentially of: (a) determining that the patient contains cancer cells with HDR-deficient status, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have HDR-deficient status; and (b) diagnosing the patient as having cancer cells with HDR-deficient status. In another aspect, the document features a method for evaluating a patient. The method includes or consists essentially of: (a) determining that the patient contains cancer cells with HDR-deficient status, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have high HDR; and (b) diagnosing the patient as having cancer cells with HDR-deficient status. In another aspect, the document features a method for evaluating a patient.The method includes or consists essentially of: (a) determining that the patient contains cancer cells having a genetic mutation in a gene from the HDR pathway, where the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in the cancer patient's cancer cells indicates that the cancer cells have the genetic mutation; and (b) diagnosing the patient as having cancer cells with the genetic mutation. In another aspect, this document features a method for assessing the likelihood that a patient will respond to a cancer treatment regimen that includes administering radiation or a drug selected from the group consisting of a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, and a PARP inhibitor. The method includes or consists essentially of: (a) determining that the patient contains cancer cells having an HRD signature, where the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in the cancer patient's cancer cells indicates that the cancer cells have an HRD signature, and (b) diagnosing the patient as likely to respond to a cancer treatment regimen based at least in part on the presence of the HRD signature. In another aspect, this document features a method for assessing the likelihood that a patient will respond to a cancer treatment regimen that includes administering radiation or a drug selected from the group consisting of a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, and a PARP inhibitor. The method includes, or consists essentially of, the steps of: (a) determining that the patient contains cancer cells having an HRD signature, wherein the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in the cancer patient's cancer cells indicates that the cancer cells have an HRD signature; and (b) diagnosing the patient as likely to respond to a cancer treatment regimen based, at least in part, on the presence of the HRD signature.

[0054] In another aspect, this document features a method for performing a diagnostic analysis of a patient's cancer cells. The method comprises or consists essentially of (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of a cancer cell above a reference number, and (b) identifying or classifying the patient as having cancer cells with an HRD signature. In another aspect, this document features a method for performing a diagnostic analysis of a patient's cancer cells. The method comprises or consists essentially of (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of a cancer cell above a reference number, and (b) identifying or classifying the patient as having cancer cells with an HDR-deficient state. In another aspect, this document features a method for performing a diagnostic analysis of a patient's cancer cells. The method comprises or consists essentially of (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of a cancer cell above a reference number, and (b) identifying or classifying the patient as having cancer cells with an HDR-deficient state. In another aspect, this document features a method for performing diagnostic analysis of cancer cells of a patient. The method includes or consists essentially of (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of the cancer cells that are relatively long and exceed the reference number, and (b) identifying or classifying the patient as having cancer cells with genetic mutations in genes from HDR pathway. In another aspect, this document features a method for performing diagnostic analysis of cancer cells of a patient to determine whether the cancer patient is likely to respond to a cancer treatment regimen that includes administering radiation or a drug selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors and PARP inhibitors. The method includes or consists essentially of (a) detecting the presence of indicator CA regions in at least one pair of human chromosomes of the cancer cells that exceed the reference number, and (b) identifying or classifying the patient as having a high likelihood of responding to the cancer treatment regimen.

[0055] In another aspect, this document features a method for diagnosing a patient as having cancer cells with HRD signature. The method comprises or consists essentially of: (a) determining that the patient comprises cells with HRD signature, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have HRD signature, and (b) diagnosing the patient as having cancer cells with HRD signature. In another aspect, this document features a method for diagnosing a patient as having cancer cells with HDR-deficient condition. The method comprises or consists essentially of: (a) determining that the patient comprises cancer cells with HDR-deficient condition, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have HDR-deficient condition, and (b) diagnosing the patient as having cancer cells with HDR-deficient condition. In another aspect, this document features a method for diagnosing a patient as having cancer cells with HDR-deficient conditions. The method comprises or consists essentially of: (a) determining that a patient comprises cancer cells with HDR-deficient conditions, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have HDR-deficient conditions; and (b) diagnosing the patient as having cancer cells with HDR-deficient conditions. In another aspect, this document features a method for diagnosing a patient as having cancer cells with genetic mutations in genes from HDR pathways. The method comprises or consists essentially of: (a) determining that a patient comprises cancer cells with genetic mutations, where the presence of indicator CA regions in at least one pair of human chromosomes of the cancer patient's cancer cells above a reference number indicates that the cancer cells have genetic mutations; and (b) diagnosing the patient as having cancer cells with genetic mutations.In another aspect, this document features a method for diagnosing a patient as a candidate for a cancer treatment regimen comprising administering radiation or a drug selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors. The method comprises or consists essentially of: (a) determining that the patient comprises cancer cells having an HRD signature, where the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in the cancer patient's cancer cells indicates that the cancer cells have an HRD signature; and (b) diagnosing the patient as likely to respond to the cancer treatment regimen based at least in part on the presence of the HRD signature. In another aspect, this document features a method for diagnosing a patient as a candidate for a cancer treatment regimen comprising administering radiation or a drug selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors. The method includes, or consists essentially of, the steps of: (a) determining that the patient contains cancer cells having a high HRD signature, wherein the presence of indicator CA regions above a reference number in at least one pair of human chromosomes in the cancer patient's cancer cells indicates that the cancer cells have an HRD signature; and (b) diagnosing the patient as likely to respond to a cancer treatment regimen based, at least in part, on the presence of the HRD signature.

[0056] In another aspect, the present invention provides a method for evaluating a patient. The method includes or consists essentially of the steps of: (a) determining whether the patient has (or has had) cancer cells that exceed the reference number of CA regions (or, for example, CA region score that exceeds the reference CA region score); and (b)(1) diagnosing the patient has (or has had) cancer cells that exceed the reference number of CA regions (or, for example, CA region score that exceeds the reference CA region score) when the patient is determined to have (or has had) cancer cells that exceed the reference number of CA regions; or (b)(2) diagnosing the patient does not have (or has never had) cancer cells that exceed the reference number of CA regions (or, for example, the patient does not have (or has never had) cancer cells with CA region score that exceed the reference CA region score) when the patient is determined to not have (or has never had) cancer cells that exceed the reference number of CA regions).

[0057] In another aspect, the invention features the use of a plurality of oligonucleotides capable of hybridizing to a plurality of polymorphic regions of human genomic DNA in the manufacture of a diagnostic kit useful for determining the total number or combined length of CA regions in at least one pair of chromosomes (or DNA derived therefrom) in a sample obtained from a cancer patient, and also useful for detecting (a) HRD, high HRD, or likelihood of HRD in the sample (e.g., an HRD signature, respectively); (b) a deletion (or likelihood of deletion) of the BRCA1 gene or the BRCA2 gene in the sample; or (c) that a cancer patient is likely to respond to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor.

[0058] In another aspect, the invention features a system for detecting HRD (e.g., HRD signature) in a sample. The system includes or consists essentially of (a) a sample analyzer configured to generate a plurality of signals for genomic DNA of at least one pair of human chromosomes (or DNA derived therefrom) in the sample, and (b) a computer subsystem programmed to calculate the number or combined length of CA regions in at least one pair of human chromosomes based on the plurality of signals. The computer subsystem can be programmed to compare the number or combined length of CA regions to a reference number to detect (a) HRD, high HRD, or likelihood of HRD in the sample (e.g., HRD signature, respectively), (b) loss (or likelihood of loss) of BRCA1 or BRCA2 genes in the sample, or (c) a high likelihood that a cancer patient will respond to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor. The system may include an output module configured to display (a), (b), or (c). The system may include an output module configured to display a recommendation regarding the use of the cancer treatment regimen.

[0059] In another aspect, the present invention provides a computer program product embodied in a computer readable medium, which, when executed on a computer, provides instructions for detecting the presence or absence of any CA region (wherein the CA region is optionally an indicator CA region) present on one or more human chromosomes other than the X and Y sex chromosomes of humans; and for determining the total number or combined length of the CA regions in one or more pairs of chromosomes. The computer program product may also include other instructions.

[0060] In another aspect, the present invention provides a diagnostic kit. The kit comprises or consists essentially of at least 500 oligonucleotides that can hybridize with multiple polymorphic regions of human genomic DNA (or DNA derived therefrom); and a computer program product as provided herein. The computer program product may be embodied in a computer readable medium, and when executed on a computer, provides instructions for detecting the presence or absence of any CA region (wherein the CA region is optionally an indicator CA region) present on one or more human chromosomes other than human X and Y sex chromosomes; and for determining the total number or combined length of CA regions in one or more pairs of chromosomes. The computer program product may also include other instructions.

[0061] In some embodiments of any one or more of the aspects of the present invention described in the preceding paragraph, any one or more of the following may be applied as appropriate. CA regions can be determined in at least 2, 5, 10 or 21 pairs of human chromosomes. Cancer cells can be ovarian, breast, lung or esophageal cancer cells. Reference numbers can be 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 or 20 or more. At least one pair of human chromosomes can exclude human chromosome 17. The DNA damaging agent may be cisplatin, carboplatin, oxalaplatin, or picoplatin, the anthracycline may be epirubicin or doxorubicin, the topoisomerase I inhibitor may be campothecin, topotecan or irinotecan, or the PARP inhibitor may be iniparib, olaparib or velapirib. The patient may be a treatment-naïve patient.

[0062] As described herein, a sample (e.g., a cancer cell sample, or a sample containing DNA from one or more cancer cells) can be identified as having an "HRD signature" (or alternatively referred to as an "HDR-deficient signature") if the genome of the cell being evaluated comprises (a) any of the LOH region score, TAI region score, or LST region score that exceeds the reference number, or (b) a composite CA region score that exceeds the reference number. Conversely, a sample (e.g., a cancer cell sample, or a sample containing DNA from one or more cancer cells) can be identified as lacking an "HRD signature" (or alternatively referred to as an "HDR-deficient signature") if the genome of the cell being evaluated comprises (a) any of the LOH region score, TAI region score, and LST region score that does not exceed the reference number, respectively, or (b) a composite CA region score that does not exceed the reference number.

[0063] Cells (e.g., cancer cells) identified as having HRD signature can be classified as likely to have HDR deficiency and / or likely to have a defective state of one or more genes in HDR pathway. For example, cancer cells identified as having HRD signature can be classified as likely to have HDR deficiency. In some cases, cancer cells identified as having HRD signature can also be classified as likely to have a defective state of one or more genes in HDR pathway. As used herein, defective state of a gene means that the sequence, structure, expression and / or activity of the gene or its product is defective compared to normal. Examples include, but are not limited to, low or no expression of mRNA or protein, harmful mutation, hypermethylation, reduced activity (e.g., enzyme activity, ability to bind with another biomolecule), and the like. As used herein, defective state of a pathway (e.g., HDR pathway) means that at least one gene (e.g., BRCA1) in the pathway is defective. Examples of highly harmful mutations include frameshift mutations, stop codon mutations, and mutations that lead to changes in RNA splicing.The defective state of genes in HDR pathway can lead to defective or reduced activity of homologous recombination repair in cancer cells.Examples of genes in HDR pathway include, but are not limited to, the genes listed in Table 1. Table 1. Selected HDR pathway genes TIFF0007680995000003.tif243145

[0064] As described herein, the identification of CA loci (as well as the size and number of CA regions) may include, first, determining the genotype of a sample at various genomic loci (e.g., SNP loci, individual bases in large-scale sequencing), and second, determining whether the loci exhibit LOH, TAI, or LST. Any suitable technique can be used to determine the genotype at loci of interest in the genome of a cell. For example, single nucleotide polymorphism (SNP) arrays (e.g., human genome-wide SNP arrays), targeted sequencing of loci of interest (e.g., sequencing of SNP loci and their surrounding sequences), and even large-scale sequencing (e.g., sequencing of the whole exome, transcriptome, or genome) can be used to identify loci as homozygous or heterozygous. Typically, the length of the CA region can be determined by analyzing the homozygosity or heterozygosity of loci over a certain length of a chromosome. For example, a series of SNP locations spaced apart along the length of a chromosome (e.g., spaced apart from about 25 kb to about 100 kb) can be evaluated using the results of an SNP array to determine not only the presence of homozygous regions (e.g., LOH) along the length of the chromosome, but also the length of the regions. Results from an SNP array can be used to generate a graph plotting allele dosage along the length of the chromosome. Allele dosage d for SNPi i is the adjusted signal intensity of the two alleles (A i and B. i ) can be calculated from: i =A i / (A i +B i). An example of such a graph is provided in Figures 1 and 2, which show the difference between fresh frozen and FFPE samples, and between SNP microarray and SNP sequencing analysis. Numerous variations on nucleic acid arrays useful in the present invention are known in the art. These include the arrays used in the various examples below (e.g., Affymetrix 500K GeneChip array in Example 3; Affymetrix OncoScan™ FFPE Express 2.0 Services (formerly MIP CN Services) in Example 4).

[0065] Once a sample has been genotyped for multiple loci (e.g., SNPs), common techniques can be used to identify loci and regions of LOH, TAI, and LST (including those described in International Application No. PCT / US2011 / 040953 (published as WO / 2011 / 160063); International Application No. PCT / US2011 / 048427 (published as WO / 2012 / 027224); Popova et al., Ploidy and large-scale genomic instability consistently identify basal-like breast carcinomas with BRCA1 / 2 inactivation, CANCER RES. (2012) 72:5454-5462). In some embodiments, determining chromosomal imbalances or large-scale transitions includes determining whether they are somatic or germline abnormalities. One technique for doing this is to compare somatic genotypes to germline. For example, the genotypes for multiple loci (e.g., SNPs) can be determined in both germline (e.g., blood) and somatic (e.g., tumor) samples. The genotypes for each sample can be compared (typically by computation) to determine where the genome of the germline cells is heterozygous and the genome of the somatic cells is homozygous. Such loci are LOH loci, and regions of such loci are LOH regions.

[0066] Computational methods can also be used to determine whether an abnormality is germline or somatic. Such techniques are particularly useful when germline samples are not available for analysis and comparison. For example, information from SNP arrays can be used to detect LOH regions using algorithms such as those described elsewhere (Nannya et al., Cancer Res. (2005) 65:6071-6079 (2005)). Typically, these algorithms do not explicitly take into account the contamination of tumor samples with benign tissue. See International Application No. PCT / US2011 / 026098 to Abkevich et al.; Goransson et al., PLoS One (2009) 4(6):e6057. This contamination is often high enough to make it difficult to detect LOH regions. Improved analytical methods according to the present invention for identifying LOH, TAI and LST despite contamination include those incorporated in computer software products such as:

[0067] The following is an example. When the observed ratio of the signals of the two alleles A and B is 2:1, there are two possibilities. The first possibility is that in a sample contaminated with 50% normal cells, the cancer cells have LOH with the deletion of allele B. The second possibility is that in a sample without contaminating normal cells, there is no LOH but allele A is duplicated. The algorithm can be implemented as a computer program described herein to reconstruct LOH regions based on genotype (e.g., SNP genotype) data. One of the key points of this algorithm is to first reconstruct the allele-specific copy number (ASCN) at each locus (e.g., SNP). The ASCN is the copy number of both the paternal allele and the maternal allele. The LOH region is then determined as a stretch of SNPs where one of the ASCNs (paternal or maternal) is 0. The algorithm can be based on maximizing a likelihood function and may be conceptually similar to previously described algorithms designed to reconstruct the total copy number (rather than the ASCN) at each locus (e.g., SNP). See International Application No. PCT / US2011 / 026098 to Abkevich et al. The likelihood function can be maximized over the ASCN of all loci, the benign tissue contamination level, the total copy number averaged over the whole genome, and the sample-specific noise level. The input data to the algorithm can include or consist of (1) sample-specific normalized signal intensities for both alleles of each locus, and (2) a set of assay-specific (specific for different SNP arrays and sequence-based approaches) parameters determined based on the analysis of a large number of samples with known ASCN profiles.

[0068] In some cases, the genotype of loci can be determined using nucleic acid sequencing techniques. For example, genomic DNA from cell samples (e.g., cancer cell samples) can be extracted and fragmented. Any suitable method can be used to extract and fragment genomic nucleic acid, including but not limited to commercially available kits such as QIAamp™ DNA Mini Kit (Qiagen™), MagNA™ Pure DNA Isolation Kit (Roche Applied Science™) and GenElute™ Mammalian Genomic DNA Miniprep Kit (Sigma-Aldrich™). Once extracted and fragmented, the genotype of the sample at loci can be determined by either targeted or untargeted sequencing. For example, whole genome, whole transcriptome, or whole exome sequencing can be performed to determine genotype at millions or even billions of base pairs (i.e., base pairs can be the "locus" to be evaluated).

[0069] In some cases, targeted sequencing of known polymorphic loci (e.g., SNPs and surrounding sequences) can be performed as an alternative to microarray analysis. For example, kits designed for this purpose (e.g., Agilent SureSelect™, Illumina TruSeq Capture™, and Nimblegen SeqCap EZ Choice™) can be used to enrich genomic DNA for fragments containing loci (e.g., SNP positions) to be analyzed. For example, genomic DNA containing loci to be analyzed can be hybridized with biotinylated capture RNA fragments to form biotinylated RNA / genomic DNA complexes. Alternatively, DNA capture probes can be utilized to form biotinylated DNA / genomic DNA hybrids. Streptavidin-coated magnetic beads and magnetic force can be used to separate biotinylated RNA / genomic DNA complexes from genomic DNA fragments that are not present in the biotinylated RNA / genomic DNA complexes. The resulting biotinylated RNA / genomic DNA complex can be processed to remove the captured RNA from the magnetic beads, thereby leaving intact genomic DNA fragments that contain the loci to be analyzed.These intact genomic DNA fragments that contain the loci to be analyzed can be amplified, for example, using PCR.The amplified genomic DNA fragments can be sequenced using high-throughput or next-generation sequencing technologies, such as Illumina HiSeq™, Illumina MiSeq™, Life Technologies SoLID™ or Ion Torrent™, or Roche 454™.

[0070] Similar to the microarray analysis described herein, sequencing results from genomic DNA fragments can be used to identify loci as exhibiting or not exhibiting CA. In some cases, the length of the CA region can be determined by analyzing the genotype of loci over a length of a chromosome. For example, a series of SNP positions spaced apart along the length of a chromosome (e.g., spaced apart from about 25 kb to about 100 kb) can be evaluated by sequencing, and the sequencing results can be used to determine not only the presence of CA regions along the length of the chromosome, but also the length of the CA regions. The resulting sequencing results can be used to generate a graph plotting allele dosages along the length of the chromosome. Allele dosage d for SNPi i is the adjustment number of the capture probe for the two alleles (A i and B. i ) can be calculated from: i =A i / (A i +B i ). An example of such a graph is provided in Figures 1 and 2. Determining whether an abnormality is germline or somatic can be performed as described herein.

[0071] In some cases, a selection process can be used to select loci (e.g., SNP loci) to be evaluated using assays (e.g., SNP array-based assays and sequencing-based assays) configured to genotype loci. For example, any human SNP location can be selected for inclusion in a SNP array-based assay or a sequencing-based assay configured to genotype loci. In some cases, 500,000, 1 million, 1.5 million, 2 million, 2.5 million or more SNP locations present in the human genome can be evaluated to identify SNPs that (a) are not on the Y chromosome, (b) are not mitochondrial SNPs, (c) have a minor allele frequency of at least about 5% in Caucasians, (d) have a minor allele frequency of at least about 1% in three non-Caucasian races (e.g., Chinese, Japanese, and Yoruba), and / or (e) do not significantly deviate from Hardy-Weinberg equilibrium in any of the four races. In some cases, more than 100,000, more than 150,000, or more than 200,000 human SNPs that meet the criteria (a) through (e) can be selected. Of the human SNPs that meet the criteria (a) through (e), a group of SNPs (e.g., the top 110,000 SNPs) can be selected such that the SNPs have a high allele frequency in Caucasians, cover the human genome at somewhat uniform intervals (e.g., at least one SNP every about 25 kb to about 500 kb), and are not in linkage disequilibrium with another selected SNP in any of the four races. In some cases, about 40,000, 50,000, 60,000, 70,000, 80,000, 90,000, 100,000, 110,000, 120,000, 130,000 or more SNPs can be selected as meeting these criteria and included in an assay configured to identify CA regions throughout the human genome. For example, about 70,000 to about 90,000 (e.g., about 80,000) SNPs can be selected for analysis using an SNP array-based assay, and about 45,000 to about 55,000 (e.g., about 54,000) SNPs can be selected for analysis using a sequencing-based assay.

[0072] As described herein, any suitable type of sample can be evaluated. For example, a sample containing cancer cells can be evaluated to determine whether the genome of the cancer cells contains HRD signature, lacks HRD signature, has an increased number of indicator CA regions, or has an increased CA region score. Examples of samples containing cancer cells that can be evaluated as described herein include, but are not limited to, tumor biopsy samples (e.g., breast tumor biopsy samples), formalin-fixed and paraffin-embedded tissue samples containing cancer cells, core needle biopsy samples, fine needle aspirates, and samples containing cancer cells sloughed off from tumors (e.g., blood, urine, or other bodily fluids). For formalin-fixed and paraffin-embedded tissue samples, samples can be prepared by DNA extraction using genomic DNA extraction kits optimized for FFPE tissues, including, but not limited to, the above-mentioned kits (e.g., QuickExtract™ FFPE DNA Extraction Kit (Epicentre™) and QIAamp™ DNA FFPE Tissue Kit (Qiagen™)).

[0073] In some cases, tissue samples can be laser dissected to minimize the number of non-cancerous cells in the cancer cell samples to be evaluated. In some cases, antibody-based purification methods can be used to enrich cancer cells and / or deplete non-cancerous cells. Examples of antibodies that can be used for cancer cell enrichment include, but are not limited to, anti-EpCAM, anti-TROP-2, anti-c-Met, anti-folate binding protein, anti-N-cadherin, anti-CD318, anti-anti-mesenchymal stem cell antigen, anti-Her2, anti-MUC1, anti-EGFR, anti-cytokeratin (e.g., cytokeratin 7, cytokeratin 20, etc.), anti-caveolin-1, anti-PSA, anti-CA125, and anti-surfactant protein antibodies.

[0074] Any kind of cancer cell can be evaluated using the methods and materials described herein.For example, breast cancer cell, ovarian cancer cell, liver cancer cell, esophageal cancer cell, lung cancer cell, head and neck cancer cell, prostate cancer cell, colon cancer cell, rectal cancer cell or colorectal cancer cell and pancreatic cancer cell can be evaluated to determine whether the genome of cancer cell contains HRD signature, lacks HRD signature, has an increased number of CA regions covering the entire chromosome, or has an increased CA region score.In some embodiments, cancer cell is the primary or metastatic cancer cell of ovarian cancer, breast cancer, lung cancer or esophageal cancer.

[0075] When the genome of a cancer cell is evaluated for the presence or absence of an HRD signature, one or more pairs of chromosomes (e.g., 1 pair, 2 pairs, 3 pairs, 4 pairs, 5 pairs, 6 pairs, 7 pairs, 8 pairs, 9 pairs, 10 pairs, 11 pairs, 12 pairs, 13 pairs, 14 pairs, 15 pairs, 16 pairs, 17 pairs, 18 pairs, 19 pairs, 20 pairs, 21 pairs, 22 pairs, or 23 pairs) can be evaluated. In some cases, the genome of a cancer cell is evaluated for the presence or absence of an HRD signature using one or more pairs of chromosomes (e.g., 1 pair, 2 pairs, 3 pairs, 4 pairs, 5 pairs, 6 pairs, 7 pairs, 8 pairs, 9 pairs, 10 pairs, 11 pairs, 12 pairs, 13 pairs, 14 pairs, 15 pairs, 16 pairs, 17 pairs, 18 pairs, 19 pairs, 20 pairs, 21 pairs, 22 pairs, or 23 pairs).

[0076] In some cases, it may be beneficial to exclude certain chromosomes from this analysis. For example, for females, the pair to be evaluated may include the X sex chromosome pair; whereas, for males, any pair of autosomes (i.e., any pair other than the X and Y sex chromosome pair) may be evaluated. As another example, in some cases, chromosome 17 pair may be excluded from analysis. It has been found that in certain cancers, certain chromosomes have higher than normal levels of CA, and therefore, when analyzing samples as described herein, it may be beneficial to exclude such chromosomes from patients with these cancers. In some cases, the sample is from a patient with ovarian cancer, and the chromosome to be excluded is chromosome 17.

[0077] Thus, a given number of chromosomes may be analyzed to determine the number of indicator CA regions (or CA region score or combined CA region score), preferably the total number of CA regions greater than 9 megabases, greater than 10 megabases, greater than 12 megabases, greater than 14 megabases, more preferably greater than 15 megabases in length. Alternatively or additionally, the sizes of all identified indicator CA regions may be summed to obtain the total length of the indicator CA region.

[0078] As described herein, the patient with cancer cell (or sample derived therefrom) identified as having HRD signature status can be classified as likely to respond to a certain cancer treatment regimen based at least in part on such HRD signature status.For example, the patient with cancer cell with genome with HRD signature can be classified as likely to respond to a cancer treatment regimen that includes the use of DNA damaging agent, synthetic lethal agent (e.g., PARP inhibitor), radiation, or combination thereof based at least in part on such HRD signature status.In some embodiments, the patient is an untreated patient. Examples of DNA damaging agents include, but are not limited to, platinum-based chemotherapeutic agents (e.g., cisplatin, carboplatin, oxaliplatin, and picoplatin), anthracyclines (e.g., epirubicin and doxorubicin), topoisomerase I inhibitors (e.g., campothecin, topotecan, and irinotecan), DNA crosslinking agents such as mitomycin C and triazene compounds (e.g., dacarbazine and temozolomide).Synthetic lethality therapeutic approaches typically involve administering an agent that inhibits at least one critical component of a biological pathway that is particularly important for the survival of a particular tumor cell. For example, if a tumor cell has a homologous repair pathway deficiency (e.g., as determined by the present invention), inhibitors of poly ADP-ribose polymerase (or platinum-based drugs, double-strand break repair inhibitors, etc.) may be particularly effective against such tumors, since two pathways critical for survival are blocked (one biologically, e.g., by BRCA1 mutations, and the other synthetically, e.g., by administration of a pathway drug).Synthetic lethal approaches to cancer therapy are described, for example, in O'Brien et al., Converting cancer mutations into therapeutic opportunities, EMBO MOL. MED. (2009) 1:297-299.Examples of synthetic lethal drugs include, but are not limited to, PARP inhibitors or double-strand break repair inhibitors in homologous repair-deficient tumor cells, PARP inhibitors in PTEN-deficient tumor cells, methotrexate in MSH2-deficient tumor cells, etc. Examples of PARP inhibitors include, but are not limited to, olaparib, iniparib, and veliparib. Examples of double-strand break repair inhibitors include, but are not limited to, KU55933 (ATM inhibitor) and NU7441 (DNA-PKcs inhibitor). Examples of information that can be used in addition to the presence of an HRD signature to form the basis of classification as likely to respond to a particular cancer treatment regimen include, but are not limited to, previous treatment results, germline or somatic DNA mutations, gene or protein expression profiling (e.g., ER / PR / HER2 status, PSA levels), tumor histology (e.g., adenocarcinoma, squamous cell carcinoma, papillary serous carcinoma, mucinous carcinoma, invasive ductal carcinoma, ductal carcinoma in situ (non-invasive), etc.), disease stage, tumor or cancer grade (e.g., well-differentiated, moderately differentiated, or poorly differentiated (e.g., Gleason, modified Bloom Richardson), etc.), number of previous courses of treatment, etc.

[0079] Once classified as likely to respond to a particular cancer treatment regimen (e.g., a cancer treatment regimen that includes the use of a DNA damaging agent, a PARP inhibitor, radiation, or a combination thereof), the cancer patient can be treated with such a cancer treatment regimen. In some embodiments, the patient is a treatment-naive patient. The present invention thus provides a method of treating a patient, comprising detecting an HRD signature as described herein, and administering (or recommending or prescribing) a treatment regimen that includes the use of a DNA damaging agent, a PARP inhibitor, radiation, or a combination thereof. The cancer patient identified as having cancer cells with an HRD signature status can be treated using any suitable method for treating the cancer in question. For example, a platinum-based chemotherapeutic agent or a combination of platinum-based chemotherapeutic agents can be used to treat the cancer as described elsewhere (see, for example, U.S. Patent Nos. 3,892,790, 3,904,663, 7,759,510, 7,759,488, and 7,754,684). In some cases, anthracycline or a combination of anthracyclines can be used to treat cancer as described elsewhere (see, for example, U.S. Patent Nos. 3,590,028, 4,138,480, 4,950,738, 6,087,340, 7,868,040 and 7,485,707). In some cases, topoisomerase I inhibitor or a combination of topoisomerase I inhibitors can be used to treat cancer as described elsewhere (see, for example, U.S. Patent Nos. 5,633,016 and 6,403,563). In some cases, PARP inhibitor or a combination of PARP inhibitors can be used to treat cancer as described elsewhere (see, for example, U.S. Patent Nos. 5,177,075, 7,915,280 and 7,351,701). In some cases, radiation can be used to treat cancer as described elsewhere (see, eg, US Pat. No. 5,295,944).In some cases, cancer can be treated with a combination that includes different drugs (e.g., a combination that includes any of platinum-based chemotherapy drugs, anthracyclines, topoisomerase I inhibitors, and / or PARP inhibitors) with or without radiation therapy.In some cases, the combination therapy can include any of the above drugs or treatments (e.g., DNA damaging drugs, PARP inhibitors, radiation, or combinations thereof) with another drug or treatment, such as a taxane drug (e.g., doxetaxel, paclitaxel, abraxane), a growth factor inhibitor or growth factor receptor inhibitor (e.g., erlotinib, gefitinib, lapatinib, sunitinib, bevacizumab, cetuximab, trastuzumab, panitumumab), and / or antimetabolite (e.g., 5-flourouracil, methotrexate).

[0080] In some cases, the patient identified as having cancer cells with genomes that lack HRD signatures based at least in part on the sample that lacks HRD signatures can be classified as unlikely to respond to a treatment regimen that includes DNA damaging agents, PARP inhibitors, radiation, or a combination thereof.Subsequently, such a patient can be classified as likely to respond to a cancer treatment regimen that includes the use of one or more cancer therapeutics that are not associated with HDR, such as taxanes (e.g., doxetaxel, paclitaxel, abraxane), growth factor inhibitors or growth factor receptor inhibitors (e.g., erlotinib, gefitinib, lapatinib, sunitinib, bevacizumab, cetuximab, trastuzumab, panitumumab), and / or antimetabolites (e.g., 5-flourouracil, methotrexate).In some embodiments, the patient is an untreated patient. Once classified as likely to respond to a particular cancer treatment regimen (e.g., a cancer treatment regimen that includes the use of a cancer therapeutic agent that is not associated with HDR), the cancer patient can be treated with such a cancer treatment regimen. The present invention thus provides a method of treating a patient, comprising detecting the absence of an HRD signature as described herein, and administering (or recommending or prescribing) a treatment regimen that does not include the use of a DNA damaging agent, a PARP inhibitor, radiation, or a combination thereof. In some embodiments, the treatment regimen comprises one or more of a taxane agent (e.g., doxetaxel, paclitaxel, abraxane), a growth factor inhibitor or growth factor receptor inhibitor (e.g., erlotinib, gefitinib, lapatinib, sunitinib, bevacizumab, cetuximab, trastuzumab, panitumumab), and / or antimetabolite (e.g., 5-flourouracil, methotrexate). Cancer patients identified as having cancer cells that lack the HRD signature can be treated using any method appropriate for the cancer being treated.Examples of information that can be used in addition to the absence of an HRD signature to form the basis of a classification as likely to respond to a particular cancer treatment regimen include, but are not limited to, previous treatment results, germline or somatic DNA mutations, gene or protein expression profiling (e.g., ER / PR / HER2 status, PSA levels), tumor histology (e.g., adenocarcinoma, squamous cell carcinoma, papillary serous carcinoma, mucinous carcinoma, invasive ductal carcinoma, ductal carcinoma in situ (non-invasive), etc.), disease stage, tumor or cancer grade (e.g., well-differentiated, moderately differentiated, or poorly differentiated (e.g., Gleason, modified Bloom Richardson), etc.), number of previous courses of treatment, etc.

[0081] Once the treatment has been administered for a certain period of time (e.g., 1 month to 6 months), the patient can be evaluated to determine whether the treatment regimen is effective. If a beneficial effect is detected, the patient can continue the same or similar cancer treatment regimen. If only a small beneficial effect or no beneficial effect is detected, adjustments can be made to the cancer treatment regimen. For example, the dose, frequency of administration, or duration of treatment can be increased. In some cases, additional anticancer drugs can be added to the treatment regimen, or a particular anticancer drug can be replaced with one or more different anticancer drugs. The treated patient can continue to be monitored as appropriate, and changes can be made to the cancer treatment regimen as appropriate.

[0082] In addition to predicting likely therapeutic response or selecting a desired therapeutic regimen, HRD signatures can also be used to determine a patient's prognosis.Thus, in one aspect, this document features a method for determining a patient's prognosis, based at least in part on detecting the presence or absence of HRD signatures in a sample from a patient.The method includes or essentially consists of: (a) determining whether a sample from a patient contains cancer cells (or whether the sample contains DNA derived from such cells) that have an HRD signature described herein (sometimes referred to herein as having high HRD) (e.g., the presence of indicator CA regions or higher CA region scores or composite CA region scores above a reference number); and (b)(1) determining that the patient's prognosis is relatively good, based at least in part on the presence of the HRD signature or having high HRD, or (b)(2) determining that the patient's prognosis is relatively poor, based at least in part on the absence of the HRD signature. Prognosis may include the patient's chances of survival (e.g., progression-free survival, overall survival), where a relatively good prognosis would include a higher chance of survival compared to some reference population (e.g., the average patient with this patient's cancer type / subtype, the average patient without the HRD signature, etc.). Conversely, a relatively poor prognosis for survival would include a lower chance of survival compared to some reference population (e.g., the average patient with this patient's cancer type / subtype, the average patient with the HRD signature, etc.).

[0083] As described herein, this document provides a method for evaluating a patient for cells (e.g., cancer cells) having a genome with an HRD signature. In some embodiments, one or more clinicians or medical practitioners can determine whether a sample from a patient contains cancer cells having an HRD signature (or whether the sample contains DNA derived from such cells). In some cases, one or more clinicians or medical practitioners can determine whether a patient contains cancer cells having an HRD signature by obtaining a cancer cell sample from a patient and evaluating the DNA of the cancer cells of the cancer cell sample to determine the presence or absence of an HRD signature as described herein.

[0084] In some cases, one or more clinicians or medical practitioners can obtain cancer cell samples from patients and provide the samples to a testing laboratory capable of evaluating the DNA of cancer cells in the cancer cell samples to indicate the presence or absence of HRD signature as described herein. In some embodiments, the patient is an untreated patient. In such cases, one or more clinicians or medical practitioners can determine whether the patient contains cancer cells with HRD signature (or whether the sample contains DNA derived from such cells) by receiving information about the presence or absence of HRD signature directly or indirectly from the testing laboratory. For example, after the testing laboratory evaluates the DNA of cancer cells for the presence or absence of HRD signature as described herein, the testing laboratory can provide or allow the clinician or medical practitioner to access a written, electronic, or verbal report or medical record that indicates the presence or absence of HRD signature for the particular patient (or patient sample) being evaluated. With such written, electronic, or verbal reports or medical records, one or more clinicians or medical personnel can determine whether a particular patient being evaluated contains cancer cells having an HRD signature.

[0085] Once a clinician or medical practitioner, or a group of clinicians or medical practitioners, determines that a particular patient being evaluated contains cancer cells with HRD signature, the clinician or medical practitioner (or group) can classify the patient as having cancer cells whose genome contains the presence of HRD signature. In some embodiments, the patient is an untreated patient. In some cases, the clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient whose genome is determined to have cancer cells whose genome contains the presence of HRD signature as having cancer cells that are HDR deficient (or likely to be HDR deficient). Such a diagnosis can be based solely on determining whether a sample from a patient contains cancer cells that have HRD signature (or whether the sample contains DNA derived from such cells) or can be based at least in part on determining whether a sample from a patient contains cancer cells that have HRD signature (or whether the sample contains DNA derived from such cells). For example, a patient determined to have cancer cells with an HRD signature can be diagnosed as likely to be HDR deficient based on a combination of the HRD signature and a deletion status of one or more tumor suppressor genes (e.g., BRCA1 / 2, RAD51C), a family history of cancer, or the presence of behavioral risk factors (e.g., smoking).

[0086] In some cases, a clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient whose genome is determined to have cancer cells that contain the presence of HRD signature as having cancer cells that are likely to contain genetic mutations in one or more genes in HDR pathway.In some embodiments, the patient is an untreated patient.Such a diagnosis can be based simply on the determination that the particular patient being evaluated has cancer cells that contain genomes that contain HRD signature, or can be based at least in part on the determination that the particular patient being evaluated has cancer cells that contain genomes that contain HRD signature.For example, a patient whose genome is determined to have cancer cells that contain the presence of HRD signature can be diagnosed as having cancer cells that are likely to contain genetic mutations in one or more genes in HDR pathway based on the combination of the presence of HRD signature and the presence of family history of cancer, or the presence of behavioral risk factors (e.g., smoking).

[0087] In some cases, a clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient who is determined to have cancer cells with HRD signature as having cancer cells that are likely to respond to a particular cancer treatment regimen. In some embodiments, the patient is an untreated patient. Such a diagnosis can be based solely on whether a sample from a patient contains cancer cells with HRD signature (or whether the sample contains DNA derived from such cells), or can be based at least in part on whether a sample from a patient contains cancer cells with HRD signature (or whether the sample contains DNA derived from such cells). For example, a patient who is determined to have cancer cells with HRD signature can be diagnosed as having a high probability of responding to a particular cancer treatment regimen based on a combination of HRD signature and the presence of one or more tumor suppressor genes (e.g., BRCA1 / 2, RAD51) defective status, family history of cancer, or behavioral risk factor (e.g., smoking). As described herein, a patient determined to have cancer cells with HRD signature can be diagnosed as likely to respond to a cancer treatment regimen that includes the use of platinum-based chemotherapy drugs, such as cisplatin, carboplatin, oxaliplatin or picoplatin, anthracyclines, such as epirubicin or doxorubicin, topoisomerase I inhibitors, such as campothecin, topotecan or irinotecan, PARP inhibitors, radiation, combinations thereof, or any of the above in combination with another anticancer drug.In some embodiments, the patient is a treatment-naive patient.

[0088] Once a clinician or medical practitioner, or a group of clinicians or medical practitioners, determines that a sample from a patient contains cancer cells with genomes that lack HRD signatures (or the sample contains DNA derived from such cells), the clinician or medical practitioner (or group) can classify the patient as having cancer cells whose genomes lack HRD signatures. In some embodiments, the patient is an untreated patient. In some cases, the clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient that is determined to have cancer cells that contain genomes that lack HRD signatures as having cancer cells that are likely to have functional HDR. In some cases, the clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient that is determined to have cancer cells that contain genomes that lack HRD signatures as having cancer cells that are unlikely to contain genetic mutations of one or more genes in HDR pathways. In some cases, a clinician or medical practitioner, or a group of clinicians or medical practitioners, can diagnose a patient who is determined to have cancer cells that contain genomes that lack HRD signatures or contain multiple CA regions that cover the entire chromosome as having cancer cells that are unlikely to respond to platinum-based chemotherapy drugs, such as cisplatin, carboplatin, oxaliplatin or picoplatin, anthracyclines, such as epirubicin or doxorubicin, topoisomerase I inhibitors, such as campothecin, topotecan or irinotecan, PARP inhibitors, or radiation, and / or are likely to respond to cancer treatment regimens that include the use of cancer therapeutic agents that are not associated with HDR, such as one or more taxanes, growth factor inhibitors or growth factor receptor inhibitors, antimetabolites, etc. In some embodiments, the patient is an untreated patient.

[0089] As described herein, this document also provides a method for performing diagnostic analysis of a nucleic acid sample (e.g., a genomic nucleic acid sample or a genomic nucleic acid sample amplified therefrom) of a cancer patient to determine whether the sample from the patient contains cancer cells that contain HRD signature and / or a large number of CA regions covering an entire chromosome (or whether the sample contains DNA derived from such cells). In some embodiments, the patient is an untreated patient. For example, one or more laboratory technicians or laboratory personnel can detect the presence or absence of HRD signature in the genome (or DNA derived therefrom) of the patient's cancer cells, or the presence or absence of a large number of CA regions covering an entire chromosome in the genome of the patient's cancer cells. In some cases, one or more laboratory technicians or laboratory workers (a) receive a cancer cell sample obtained from a patient, receive a genomic nucleic acid sample obtained from a cancer cell obtained from a patient, or receive such enriched and / or amplified genomic nucleic acid sample obtained from a cancer cell obtained from a patient, and (b) perform an analysis (e.g., SNP array-based assay or sequencing-based assay) using the received material to detect the presence or absence of an HRD signature as described herein, or the presence or absence of a number of CA regions covering an entire chromosome, thereby detecting the presence or absence of an HRD signature in the genome of the patient's cancer cells, or the presence or absence of a number of CA regions covering an entire chromosome. In some cases, one or more laboratory technicians or laboratory workers can receive the sample to be analyzed (e.g., a cancer cell sample obtained from a patient, a genomic nucleic acid sample obtained from a cancer cell obtained from a patient, or such enriched and / or amplified genomic nucleic acid sample obtained from a cancer cell obtained from a patient) directly or indirectly from a clinician or medical worker. In some embodiments, the patient is an untreated patient.

[0090] Once a laboratory technician or a laboratory worker, or a group of laboratory technicians or laboratory workers, detects the presence of an HRD signature as described herein, the laboratory technician or laboratory worker (or group) can associate the HRD signature or the result of the diagnostic analysis performed (or a plurality of results or a compilation of results) with the corresponding patient's name, medical record, symbol / number identifier, or a combination thereof. Such identification may be based solely on detecting the presence of the HRD signature, or may be based at least in part on detecting the presence of the HRD signature. For example, a laboratory technician or laboratory worker can identify a patient with cancer cells detected as having an HRD signature as having cancer cells that are potentially HDR deficient (or likely to respond to a particular treatment as described at length herein) based on a combination of the HRD signature and the results of other genetic and biochemical tests performed in the testing laboratory. In some embodiments, the patient is a treatment-naive patient.

[0091] The reverse of the above also applies. That is, once a laboratory technician or a laboratory worker, or a group of laboratory technicians or laboratory workers, detects the absence of HRD signature, the laboratory technician or laboratory worker (or a group) can associate the HRD signature or the result of the diagnostic analysis performed (or a plurality of results or a compilation of results) with the corresponding patient's name, medical record, symbol / number identifier, or a combination thereof. In some cases, the laboratory technician or laboratory worker, or a group of laboratory technicians or laboratory workers, can identify a patient with cancer cells detected as lacking HRD signature as having cancer cells with potentially intact HDR (or unlikely to respond to a particular treatment as described at length herein) based on the absence of HRD signature alone, or based on the presence of HRD signature combined with the results of other genetic and biochemical tests performed in the testing laboratory. In some embodiments, the patient is an untreated patient.

[0092] The results of any analysis according to the present invention are often communicated to a physician, genetic counselor and / or patient (or other parties such as researchers) in a communicable form that can be communicated or communicated to any of the above parties. Such forms may vary and may be tangible or intangible. The results may be embodied in a description, diagram, photograph, chart, image, or any other visual form. For example, a graph or diagram showing genotype or LOH (or HRD status) information may be used in explaining the results. The description and visual form may be recorded on a tangible medium, such as paper, a computer readable medium, such as a floppy disk, compact disk, flash memory, etc., or on an intangible medium, such as an electronic medium in the form of an email or a website on the Internet or an intranet. Additionally, the results may be recorded in audio form and communicated via any suitable medium, such as analog or digital cable lines, fiber optic cables, etc., telephone, facsimile, wireless mobile phone, Internet telephone, etc.

[0093] Thus, the information and data regarding the test results can be generated anywhere in the world and transmitted to different locations. As one illustrative example, if the assay is performed outside the United States, the information and data regarding the test results can be generated, cast in a transmittable form as described above, and then imported into the United States. Thus, the present invention also encompasses a method for generating a transmittable form of information regarding the HRD signature for at least one patient sample. The method includes the steps of (1) determining the HRD signature according to the method of the present invention; and (2) embodying the results of the determining step in a transmittable form. The transmittable form is the result of such a method.

[0094] Some embodiments of the invention described herein involve correlating an HRD signature according to the invention (e.g., a total number of indicator CA regions above a reference number or a CA region score or a composite CA region score) with a particular clinical feature (e.g., a high likelihood of BRCA1 or BRCA2 gene defects; a high likelihood of HDR deficiency; a high likelihood of response to a treatment regimen including a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, etc.), and optionally correlating the absence of an HRD signature with one or more other clinical features. Wherever such an embodiment is described throughout this document, other embodiments of the invention may involve one or both of the following steps in addition to or instead of the correlating step: (a) concluding that the patient has the clinical feature based at least in part on the presence or absence of the HRD signature; or (b) communicating that the patient has the clinical feature based at least in part on the presence or absence of the HRD signature.

[0095] By way of example, but not limitation, one embodiment described herein is a method of predicting a cancer patient's response to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising the steps of: (1) determining in a sample two or more of: (a) a LOH region score for the sample; (b) a TAI region score for the sample; or (c) a LST region score for the sample; and (2) determining in a sample two or more of: (a) a LOH region score, a TAI region score, and a LST region score that exceed a reference number. The method includes: (2) (b) correlating a combination of two or more (e.g., a composite CA region score) with a high likelihood of responding to the treatment regimen; or optionally (2) (b) correlating a combination of two or more of the LOH region score, TAI region score and LST region score that does not exceed a reference number (e.g., a composite CA region score) with a low likelihood of responding to the treatment regimen; or optionally (2) (c) correlating the average (e.g., arithmetic mean) of the LOH region score, TAI region score and LST region score. According to the previous paragraph, this description of this embodiment is interpreted as including the description of two related alternative embodiments.One such embodiment is a method of predicting a cancer patient's response to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising the steps of: (1) determining in a sample two or more of: (a) a LOH region score for the sample; (b) a TAI region score for the sample; or (c) an LST region score for the sample; or (d) an average (e.g., arithmetic mean) of the LOH region score, the TAI region score, and the LST region score; and (2) determining in a sample two or more of: (a) a LOH region score for the sample; (b) a TAI region score for the sample; or (c) an LST region score for the sample; or (d) an average (e.g., arithmetic mean) of the LOH region score, the TAI region score, and the LST region score that exceeds a reference number. or optionally (2)(b) concluding that the patient is unlikely to respond to the cancer treatment regimen based at least in part on a combination of two or more of the LOH region score, the TAI region score, and the LST region score (e.g., a combined CA region score), or an average (e.g., an arithmetic mean) of the LOH region score, the TAI region score, and the LST region score, not exceeding a reference number.Another such embodiment is a method of predicting a cancer patient's response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising the steps of: (1) determining in a sample two or more of: (a) an LOH region score for the sample; (b) a TAI region score for the sample; or (c) an LST region score for the sample; or (d) an average (e.g., arithmetic mean) of the LOH region score, the TAI region score, and the LST region score; and (2) determining in a sample a combination of two or more of the LOH region score, the TAI region score, and the LST region score that exceeds a reference number ( or (2)(b) a combination of two or more of the LOH region score, TAI region score, and LST region score (e.g., a combined CA region score); or an average (e.g., arithmetic mean) of the LOH region score, TAI region score, and LST region score, where the combination does not exceed a reference number; or (2)(b) a combination of two or more of the LOH region score, TAI region score, and LST region score (e.g., a combined CA region score); or an average (e.g., arithmetic mean) of the LOH region score, TAI region score, and LST region score, where the combination does not exceed a reference number;

[0096] In each embodiment described herein involving correlating the output of a particular assay or analysis (e.g., total number of indicator CA regions above a reference number, presence of an HRD signature, etc.) with some likelihood (e.g., increased, no increase, decreased, etc.) of some clinical feature (e.g., response to a particular treatment, cancer-specific mortality, etc.), or additionally or alternatively concluding or communicating such clinical feature based at least in part on the output of such a particular assay or analysis, such correlating, concluding, or communicating may include assigning a risk or likelihood of the clinical feature occurring based at least in part on the output of the particular assay or analysis. In some embodiments, such risk is a percentage probability of an event or outcome occurring. In some embodiments, patients are assigned to a risk group (e.g., low risk, medium risk, high risk, etc.). In some embodiments, "low risk" is any percentage probability of less than 5%, less than 10%, less than 15%, less than 20%, less than 25%, less than 30%, less than 35%, less than 40%, less than 45%, or less than 50%. In some embodiments, "medium risk" is any percent probability of more than 5%, more than 10%, more than 15%, more than 20%, more than 25%, more than 30%, more than 35%, more than 40%, more than 45%, or more than 50%, and less than 15%, less than 20%, less than 25%, less than 30%, less than 35%, less than 40%, less than 45%, less than 50%, less than 55%, less than 60%, less than 65%, less than 70%, or less than 75%. In some embodiments, "high risk" is any percent probability of more than 25%, more than 30%, more than 35%, more than 40%, more than 45%, more than 50%, more than 55%, more than 60%, more than 65%, more than 70%, more than 75%, more than 80%, more than 85%, more than 90%, more than 95%, or more than 99%.

[0097] As used herein, "communicating" a particular piece of information means making such information known to another person or transferring such information to a thing (e.g., a computer). In some methods of the present invention, a patient's prognosis or likelihood of response to a particular treatment is communicated. In some embodiments, information used to arrive at such a prognosis or response prediction (e.g., an HRD signature according to the present invention, etc.) is communicated. This communication may be audible (e.g., verbal), visual (e.g., written), electronic (e.g., data transferred from one computer system to another computer system), etc. In some embodiments, communicating the cancer classification (e.g., prognosis, likelihood of response, appropriate treatment, etc.) includes generating a report that communicates the cancer classification. In some embodiments, the report is a paper report, an audible report, or an electronic record. In some embodiments, the report is displayed and / or stored on a computing device (e.g., a handheld device, a desktop computer, a smart device, a website, etc.). In some embodiments, the cancer classification is communicated to a physician (e.g., a report that communicates the classification is provided to the physician). In some embodiments, the cancer classification is communicated to the patient (e.g., a report is provided to the patient communicating the classification). The cancer classification can also be communicated by transferring information (e.g., data) embodying the classification to a server computer and allowing an intermediate or end user to access such information (e.g., by viewing the information as displayed from the server, by downloading the information in the form of one or more files transferred from the server to the intermediate or end user's device, etc.).

[0098] Whenever an embodiment of the invention involves concluding some fact (e.g., the prognosis of a patient, or the likelihood that a patient will respond to a particular treatment regimen), this may in some embodiments involve a computer program that concludes such a fact, typically after running an algorithm that applies information about the CA region in accordance with the invention.

[0099] In each embodiment described herein involving several CA regions (e.g., indicator CA regions), or the combined calculated length of such CA regions, or the average (e.g., arithmetic mean) of the composite CAR region scores, the invention encompasses related embodiments involving test values ​​or scores (e.g., CA region scores, LOH region scores, etc.) derived from, incorporating, and / or reflecting at least to some extent such numbers or lengths. In other words, in various methods, systems, etc. of the invention, the number or length of bare CA regions need not be used, but test values ​​or scores derived from such numbers or lengths may be used.For example, one embodiment of the invention is a method of treating cancer in a patient, comprising: (1) determining in a sample from the patient two or more or an average (e.g., an arithmetic mean) of (a) the number of indicator LOH regions, (b) the number of indicator TAI regions, or (c) the number of indicator LST regions; (2) providing one or more test values ​​derived from the number of indicator LOH regions, indicator TAI regions, and / or indicator LST regions; (3) comparing the test values ​​to one or more reference values ​​(e.g., reference values ​​(e.g., mean, median, tertile, quartile, quintile, etc.) derived from the number of indicator LOH regions, indicator TAI regions, and / or indicator LST regions in a reference population); and (4)(a) administering an anti-cancer drug to said patient, or recommending, prescribing, or initiating a treatment regimen including chemotherapy and / or a synthetic lethal agent, based at least in part on said comparing step revealing that one or more of the test values ​​is greater than at least one of said reference values ​​(e.g., at least 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; at least 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater); or optionally (4)(b) one or more of the test values ​​is not greater than at least one of said reference values ​​(e.g., not 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; recommending, prescribing, or initiating a chemotherapy and / or synthetic lethal agent-free treatment regimen based at least in part on said comparing step revealing that the IL-16 expression level is not greater than 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater).The invention encompasses corresponding embodiments in which the test value or score is used mutatis mutandis to determine a patient's prognosis, the likelihood that a patient will respond to a particular treatment regimen, the likelihood that a patient or patient sample has BRCA1 deficiency, BRCA2 deficiency, RAD51C deficiency, HDR deficiency, etc.

[0100] FIG. 8 shows an exemplary process by which a computational system (or a computer program (e.g., software) including computer executable instructions) can identify LOH loci or regions from genotype data as described herein. This process can be adapted for use in determining TAI and LST, as will be apparent to those skilled in the art. When the observed ratio of signals of the two alleles A and B is 2:1, there are two possibilities. The first possibility is that the cancer cells have LOH with allele B deleted in a sample that is 50% contaminated by normal cells. The second possibility is that there is no LOH, but allele A is duplicated in a sample that is not contaminated by normal cells. The process starts at box 1500. In box 1500, the following data is collected by the computational system: (1) sample-specific normalized signal intensities of both alleles of each locus, and (2) an assay-specific (specific to different SNP arrays and sequence-based approaches) parameter set defined based on the analysis of a large number of samples with known ASCN profiles. As described herein, any suitable assay, such as an SNP array-based assay or a sequencing-based assay, can be used to assess the homozygosity or heterozygosity of loci along a chromosome. In some cases, a system including a signal detector and a computer can be used to collect data (e.g., fluorescent signals or sequencing results) on the homozygosity or heterozygosity of multiple loci (e.g., sample-specific normalized signal intensity of both alleles at each locus). In box 1510, an allele-specific copy number (ASCN) is reconstructed at each locus (e.g., each SNP). The ASCN is the copy number of both the paternal and maternal alleles. In box 1530, a likelihood function is used to determine whether a homozygous locus or region of a homozygous locus is due to LOH. This may be conceptually similar to the previously described algorithm designed to reconstruct the total copy number (not the ASCN) at each locus (e.g., SNP).See International Application No. PCT / US2011 / 026098 to Abkevich et al. The likelihood function can be maximized over the ASCN of all loci, the level of contamination by benign tissue, the total copy number averaged over the whole genome, and the sample-specific noise level. In box 1540, the LOH region is determined as a stretch of SNPs where one of the ASCNs (paternal or maternal) is 0. In some embodiments, the computer process further includes a step of querying whether the patient is untreated or a step of determining whether the patient is untreated.

[0101] FIG. 3 shows an exemplary process by which a computational system can determine the presence or absence of LOH signatures, which is included to illustrate how this process can also be applied to TAI and LST, as will be apparent to those skilled in the art. The process begins at box 300, where data on homozygosity or heterozygosity of multiple loci along a chromosome is collected by a computational system. As described herein, any suitable assay, such as an SNP array-based assay or a sequencing-based assay, can be used to evaluate the loci along a chromosome for homozygosity or heterozygosity. In some cases, a system comprising a signal detector and a computer can be used to collect data (e.g., fluorescent signals or sequencing results) on homozygosity or heterozygosity of multiple loci. At box 310, the data on homozygosity or heterozygosity of multiple loci and the position or spatial relationship of each locus are evaluated by a computational system to determine the length of any LOH regions present along the chromosome. In box 320, the data regarding the number of detected LOH regions and the length of each detected LOH region is evaluated by the computing system to determine the number of LOH regions that (a) are greater than or equal to a preset number of Mb (e.g., 15 Mb) and (b) have a length less than the full length of the chromosome containing the LOH region. Alternatively, the computing system can determine the full length of the LOH or determine the sum of the LOH lengths as described above. In box 330, the computing system formats an output indicating the presence or absence of the HRD signature. Once formatted, the computing system can present the output to a user (e.g., a laboratory technician, clinician, or medical professional). As described herein, the presence or absence of the HRD signature can be used to indicate the likely HDR status of the patient, to indicate the likely presence or absence of a genetic mutation of a gene in the HDR pathway, and / or to indicate a possible cancer treatment regimen.

[0102] 4 is a diagram of an example of a computing device 1400 and a mobile computing device 1450 that may be used with the techniques described herein. The computing device 1400 is intended to be various types of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The computing device 1450 is intended to be various types of mobile devices, such as personal digital assistants, mobile phones, smartphones, and other similar computing devices. The components shown, their connections and relationships, and their functions are intended to be merely exemplary and are not intended to limit the practice of the invention described and / or claimed herein.

[0103] Computing device 1400 includes processor 1402, memory 1404, storage device 1406, high-speed interface 1408 connecting memory 1404 and high-speed expansion port 1410, and low-speed interface 1415 connecting low-speed bus 1414 and storage device 1406. Each of components 1402, 1404, 1406, 1408, 1410, and 1415 may be interconnected using various buses and mounted on a common motherboard or in other manners as appropriate. Processor 1402 can process instructions for execution within computing device 1400, including instructions stored in memory 1404 or storage device 1406, to display external input / output devices, such as graphical information for a GUI on display 1416 in communication with high-speed interface 1408. In other implementations, multiple processors and / or multiple buses may be used, along with multiple memories and types of memories, as appropriate. Also, multiple computing devices 1400 may be connected, each providing a portion of the required operations (eg, as a server bank, a collection of blade servers, or a multi-processor system).

[0104] The memory 1404 stores information within the computing device 1400. In one implementation, the memory 1404 is a volatile memory unit. In another implementation, the memory 1404 is a non-volatile memory unit. The memory 1404 may also be another type of computer readable medium, such as a magnetic disk or an optical disk.

[0105] The storage device 1406 can provide mass storage for the computing device 1400. In one implementation, the storage device 1406 can be or comprise a number of different devices, including computer-readable media, such as a floppy disk drive, a hard disk drive, an optical disk drive, or a tape drive, a flash memory or other similar solid-state memory device, or a device in a storage area network or other configuration. The computer program product can be tangibly embodied in an information carrier. The computer program product can also include instructions that, when executed, perform one or more methods, such as the methods described herein. The information carrier is a computer-readable or machine-readable medium, such as the memory 1404, the storage device 1406, a memory installed in the processor 1402, or a propagating signal.

[0106] The high-speed controller 1408 manages bandwidth-intensive operations for the computing device 1400, while the low-speed controller 1415 manages less bandwidth-intensive operations. Such an allocation of functions is merely exemplary. In one implementation, the high-speed controller 1408 is coupled to the memory 1404, the display 1416 (e.g., via a graphics processor or accelerator), and a high-speed expansion port 1410, which may accept various expansion cards (not shown). In this implementation, the low-speed controller 1415 is coupled to the storage device 1406 and the low-speed expansion port 1414. The low-speed expansion port may comprise various communication ports (e.g., USB, Bluetooth, Ethernet, or wireless Ethernet) and may be coupled, for example, via a network adapter, to one or more input / output devices, such as a keyboard, a pointing device, a scanner, an optical reader, a fluorescent signal detector, or a networking device, such as a switch or a router.

[0107] As shown in the figure, the computing device 1400 can be implemented in many different ways. For example, the computing device 1400 can be implemented as a standard server 1420, or multiple times in a collection of such servers. The computing device 1400 can be implemented as part of a rack server system 1424. Additionally, the computing device 1400 can be implemented in a personal computer, such as a laptop computer 1422. Or, components from the computing device 1400 can be combined with other components, such as device 1450, in a mobile device (not shown). Each such device may comprise one or more of the computing devices 1400, 1450, and the entire system may consist of multiple computing devices 1400, 1450 in communication with each other.

[0108] Computing device 1450 includes a processor 1452, memory 1464, input / output devices such as a display 1454, a communication interface 1466, and a transceiver 1468, among other components (e.g., scanner, optical reader, fluorescent signal detector). Device 1450 may also be provided with a storage device such as a microdrive or other device to provide additional storage. Components 1450, 1452, 1464, 1454, 1466, and 1468 are each interconnected using various buses, and some of the components may be mounted on a common motherboard or in other manners as appropriate.

[0109] The processor 1452 can execute instructions within the computing device 1450, including instructions stored in the memory 1464. The processor may be implemented as a chipset of chips with separate and multiple analog and digital processors. The processor may provide, for example, control of a user interface, applications run by the device 1450, and wireless communication by the device 1450, for example, to coordinate other components of the device 1450.

[0110] The processor 1452 can communicate with a user via a control interface 1458 and a display interface 1456 coupled to a display 1454. The display 1454 can be, for example, a TFTLCD (thin film transistor liquid crystal display) or an OLED (organic light emitting diode) display or other suitable display technology. The display interface 1456 can include suitable circuitry for driving the display 1454 to present graphical and other information to a user. The control interface 1458 can receive commands from a user and translate the commands for transmission to the processor 1452. Additionally, an external interface 1462 can communicate with the processor 1452 to enable near area communication between the device 1450 and other devices. The external interface 1462 can provide, for example, wired communication in some implementations and wireless communication in other implementations, and multiple interfaces can be used.

[0111] The memory 1464 stores information within the computing device 1450. The memory 1464 may be implemented as one or more of a computer readable medium, a volatile memory unit, or a non-volatile memory unit. An expansion memory 1474 may be provided and connected to the device 1450 via an expansion interface 1472. The expansion interface 1472 may include, for example, a SIMM (single in-line memory module) card interface. Such expansion memory 1474 may provide extra storage space to the device 1450 and may store applications or other information for the device 1450. For example, the expansion memory 1474 may include instructions that perform or supplement the processes described herein and may also include secure information. Thus, for example, the expansion memory 1474 may be provided as a security module for the device 1450 and may be programmed with instructions that enable secure use of the device 1450. Additionally, secure applications may be provided via a SIMM card along with additional information, such as placing identifying information on the SIMM card so that it cannot be hacked.

[0112] The memory may include, for example, flash memory and / or NVRAM memory, as discussed below. In one implementation, the computer program product is tangibly embodied in an information carrier. The computer program product includes instructions that, when executed, perform one or more methods, such as the methods described herein. The information carrier is a computer-readable or machine-readable medium, such as memory 1464, expansion memory 1474, memory on-board the processor 1452, or a propagated signal that may be received, such as a propagated signal that may be received by the transceiver 1468 or the external interface 1462.

[0113] The device 1450 may communicate wirelessly via a communication interface 1466. The communication interface 1466 may include digital signal processing circuitry, if necessary. The communication interface 1466 may provide communications under various modes or protocols, such as GSM voice calls, SMS, EMS, or MMS messaging, CDMA, TDMA, PDC, WCDMA, CDMA2000, or GPRS, among others. Such communications may occur, for example, via a radio frequency transceiver 1468. Additionally, short-range communications may occur, for example, using Bluetooth, WiFi, or other such transceivers (not shown). Additionally, a GPS (Global Positioning System) receiver module 1470 may provide further navigation-related and location-related wireless data to the device 1450, which may be used by applications operating on the device 1450, as appropriate.

[0114] Device 1450 may also communicate audibly using audio codec 1460. Audio codec 1460 may receive voice information from a user and convert it into usable digital information. Additionally, audio codec 1460 may generate audible sounds to a user, for example at a handset of device 1450, for example via a speaker. Such sounds may include sounds from voice telephone calls, may include recorded sounds (e.g., voice mail, music files, etc.), and may also include sounds created by applications running on device 1450.

[0115] As shown in the figure, computing device 1450 may be implemented in a number of different ways. For example, computing device 1450 may be implemented as a mobile phone 1480. Computing device 1450 may also be implemented as part of a smartphone 1482, personal digital assistant, or other similar mobile device.

[0116] Various implementations of the systems and techniques described herein can be realized in digital electronic circuitry, integrated circuits, specially designed ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include execution in one or more computer programs executable and / or interpretable on a programmable system including at least one programmable processor, which may be special purpose or general purpose, coupled to receive data and instructions from a storage system, at least one input device, and at least one output device, and to transmit data and instructions to the storage system, at least one input device, and at least one output device.

[0117] These computer programs (also known as programs, software, software applications, or code) contain machine instructions for a programmable processor and may be executed in high-level procedural and / or object-oriented programming languages ​​and / or assembly / machine languages. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal (e.g., magnetic disks, optical disks, memory, and programmable logic devices (PLDs)). The term "machine-readable signal" refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0118] The systems and techniques described herein for providing user interaction can be executed on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user, as well as a keyboard and pointing device (e.g., a mouse or trackball) to allow the user to provide input to the computer. Other types of devices can also be used to provide user interaction. For example, feedback provided to the user can be any type of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback). Input from the user can be received in any form, including acoustic input, speech input, or tactile input.

[0119] The systems and techniques described herein can be implemented in a computing system that includes a back-end component (e.g., as a data server), or includes a middleware component (e.g., an application server), or includes a front-end component (e.g., a client computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communications network). Examples of communications networks include a local area network ("LAN"), a wide area network ("WAN"), and the Internet.

[0120] A computing system may include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.

[0121] In some cases, the computing system provided herein may be configured to include one or more sample analyzers. The sample analyzer may be configured to generate a plurality of signals for genomic DNA of at least one pair of human chromosomes from cancer cells. For example, the sample analyzer may generate a signal that can be interpreted in a manner that identifies homozygosity or heterozygosity of loci along the chromosome. In some cases, the sample analyzer may be configured to perform one or more steps of a SNP array-based assay or a sequencing-based assay, and may be configured to generate and / or capture signals from such assays. In some cases, the computing system provided herein may be configured to include a computing device. In such cases, the computing device may be configured to receive signals from the sample analyzer. The computing device may include computer executable instructions or a computer program (e.g., software) that includes computer executable instructions for performing one or more of the methods or steps described herein. In some cases, such computer executable instructions can instruct the computing device to analyze signals from a sample analyzer, another computing device, a SNP array-based assay, or a sequencing-based assay.Such signals can be analyzed to determine genotype, homozygosity or other chromosomal abnormalities at a particular locus, CA region, number of CA regions, determine the size of the CA region, determine the number of CA regions having a particular size or size range, determine whether the sample is positive for an HRD signature, determine the number of indicator CA regions in at least one pair of human chromosomes, determine the likelihood of loss of the BRCA1 and / or BRCA2 genes, determine the likelihood of loss of HDR, determine the likelihood that a cancer patient will respond to a particular cancer treatment regimen (e.g., a regimen including a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, a PARP inhibitor, or a combination thereof), or determine a combination of these items.

[0122] In some cases, the computational system provided herein may comprise a computer program (e.g., software) that includes computer executable instructions or computer executable instructions for formatting an output that provides an indication of the number of CA regions, the size of CA regions, the number of CA regions with a particular size or size range, whether a sample is positive for an HRD signature, the number of indicator CA regions in at least one pair of human chromosomes, the likelihood of loss of BRCA1 and / or BRCA2 genes, the likelihood of loss of HDR, the likelihood that a cancer patient will respond to a particular cancer treatment regimen (e.g., a regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, a PARP inhibitor, or a combination thereof), or a combination of these items. In some cases, the computational system provided herein may comprise a computer program (e.g., software) that includes computer executable instructions or computer executable instructions for determining a cancer treatment regimen that is desirable for a particular patient based at least in part on the presence or absence of an HRD signature or the number of indicator CA regions.

[0123] In some cases, the computational system provided herein may include a pre-processing device configured to process a sample (e.g., cancer cells) so that SNP array-based assays or sequencing-based assays can be performed. Examples of pre-processing devices include, but are not limited to, devices configured to enrich cell populations for cancer cells as opposed to non-cancerous cells, devices configured to lyse cells and / or extract genomic nucleic acid, and devices configured to enrich samples for specific genomic DNA fragments.

[0124] This document also provides a kit for evaluating a sample (e.g., cancer cells) as described herein. For example, this document provides a kit for evaluating cancer cells for the presence of HRD signatures or for determining the number of indicator CA regions in at least one pair of human chromosomes. The kit provided herein may include either SNP probes (e.g., an array of SNP probes for carrying out the SNP array-based assay described herein) or primers (e.g., primers designed to sequence SNP regions via sequencing-based assays) in combination with a computer program product that includes computer executable instructions for carrying out one or more of the methods or steps described herein (e.g., computer executable instructions for determining the number of CA regions with a particular size or size range). In some cases, the kit provided herein may include at least 500, 1000, 10,000, 25,000, or 50,000 SNP probes that can hybridize to polymorphic regions of human genomic DNA. In some cases, the kits provided herein may include at least 500, 1000, 10,000, 25,000, or 50,000 primers capable of sequencing polymorphic regions of human genomic DNA. In some cases, the kits provided herein may include one or more other components for performing SNP array-based assays or sequencing-based assays. Examples of such other components include, but are not limited to, buffers, sequencing nucleotides, enzymes (e.g., polymerases), and the like. This document also provides for the use of any suitable number of materials provided herein in the manufacture of a kit for carrying out one or more of the methods or steps described herein. For example, this document provides for the use of a collection of SNP probes (e.g., a collection of 10,000 to 100,000 SNP probes) and a computer program product provided herein in the manufacture of a kit for evaluating cancer cells for the presence of an HRD signature.As another example, this document provides the use of a collection of primers (e.g., a collection of 10,000-100,000 primers for sequencing SNP regions) and the computer program product provided herein in the manufacture of a kit for assessing cancer cells for the presence of an HRD signature.

[0125] Specific Aspects Below are illustrative, non-limiting details of particular aspects of the disclosure, namely, methods and systems in accordance with the more general description above.

[0126] In some embodiments, the sample used is a frozen tumor sample. In some embodiments, the sample is from a specific breast cancer subtype selected from triple negative, ER+ / HER2-, ER- / HER2+, or ER+ / HER2+. In some embodiments, the test assay portion of the method, system, etc. comprises assaying the sample to sequence BRCA1 and / or BRCA2 genes (as well as any other one or more genes in Table 1). In some embodiments, the test assay portion of the method, system, etc. comprises assaying the sample to determine allele dosage (e.g., genotype, copy number, etc.) for at least 10,000, 20,000, 30,000, 40,000, 50,000, 60,000, 70,000, 80,000, 90,000, 100,000 or more selected SNPs across the entire genome. In some embodiments, the SNP analysis is performed using oligonucleotide microarrays as discussed above. In some embodiments, BRCA sequence analysis, SNP analysis, or both are performed by using probe capture (e.g., a probe for each SNP to be analyzed and / or a probe to capture the entire coding region of BRCA1 and / or BRCA2) followed by a PCR enrichment technique (e.g., Agilent™ SureSelect XT). In some embodiments, BRCA sequence analysis, SNP analysis, or both are performed by processing the output from the enrichment technique using a "next-generation" sequencing platform (e.g., Illumina™ HiSeq2500). In some embodiments, samples are analyzed for somatic and / or germline mutations of BRCA1 / 2, which may include large-scale rearrangements. In some embodiments, samples are analyzed for BRCA1 promoter methylation (e.g., by qPCR assay (e.g., SA Biosciences)).In some embodiments, a sample is determined to have hypermethylation (or be "methylated") if it has greater than 10% (or 5%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%) methylation (e.g., percentage of methylation of promoter CpGs of BRCA1 or BRCA2). In some embodiments, DNA from matched normal (non-tumor) tissue of the patient may be analyzed, for example, to determine whether a BRCA1 or BRCA2 mutation is germline or somatic.

[0127] In some embodiments, the LOH region score can be calculated by calculating the number of LOH regions that are greater than 15 Mb in length but shorter than the length of the entire chromosome. In some embodiments, the TAI region score can be calculated by calculating the number of telomeric regions with allelic imbalance that are greater than 11 Mb in length and that extend into one of the subtelomeres but do not cross the centromere. In some embodiments, the LST region score can be calculated by calculating the number of discontinuities between regions longer than 10 megabases with stable copy number after filtering out regions shorter than 3 megabases. In some embodiments, the LST region score can be modified by adjusting for ploidy: LSTm=LST-kP, where P is ploidy and k is a constant (in some embodiments, k=15.5). In some embodiments, BRCA1 / 2 deficiency can be defined as loss of function due to BRCA1 or BRCA2 mutations or methylation of the promoter region of BRCA1 or BRCA2, with LOH in the affected gene. In some embodiments, the response to treatment may be a partial complete response ("pCR"), which in some embodiments may be defined as Miller-Payne 5 status after treatment (e.g., neoadjuvant).

[0128] In some embodiments, the claimed methods include ameliorating BRCA deficiency in at least 8 * 10 -12, 6 * 10 -6 , 0.0009, 0.01, 0.03, 2 * 10 -16 , 3 * 10 -6 , 10 -6 , 0.0009, 8 * 10 -12 , 2 * 10 -16 , 8 * 10 -8 , 6 * 10 -6 , 3 * 10 -6 or a p-value of 0.0002 (e.g., each CA region score is predefined and multiple scores are optionally combined to obtain these p-values). In some embodiments, the p-value is calculated by the Kolmogorov-Smirnov test. In some embodiments, the HRD score and age at diagnosis can be coded as numerical (e.g., integer) variables, breast cancer stage and subtype can be coded as categorical variables, and grade can be analyzed as either numerical or categorical variables or both.

[0129] In some embodiments, the p-value is two-sided. In some embodiments, logistic regression analysis can be used to predict BRCA1 / 2 defects based on the HRD score (including the HRD-composite score) as disclosed herein. In some embodiments, the various CA region scores are correlated according to the following correlation coefficient (e.g., defined for the purpose of achieving the following): LOH region score and TAI region score = 0.69 (p = 10 -39 ), between LOH and LST = 0.55 (p = 2 * 10 -19 ) and between TAI and LST = 0.39 (p = 10 -9 ).

[0130] In some embodiments, the method combines the LOH region score and the TAI region score to detect BRCA1 / 2 defects and / or to predict treatment response (e.g., response to platinum therapy (e.g., cisplatin)) as follows: combined CA region score = 0.32 * LOH area score +0.68 * TAI region score. In some embodiments, the method combines the LOH region score, the TAI region score and the LST region score to detect BRCA1 / 2 defects and / or predict treatment response (e.g., response to platinum therapy (e.g., cisplatin)) as follows: Combined CA region score = 0.21 * LOH area score +0.67 * TAI domain score +0.12 * LST Region Score. In some embodiments, the method combines the LOH region score, the TAI region score, and the LST region score to detect BRCA1 / 2 defects and / or to predict treatment response (e.g., response to platinum therapy (e.g., cisplatin)) as follows: Combined CA region score=0.1 1 * LOH area score +0.25 * TAI domain score +0.12 * LST region score. In some embodiments, the method combines the LOH region score, the TAI region score and the LST region score to detect BRCA1 / 2 defects and / or predict treatment response (e.g., response to platinum therapy (e.g., cisplatin)) as follows: Combined CA region score = arithmetic mean of the LOH region score, the TAI region score and the LST region score.

[0131] In some embodiments, the combination of BRCA-deficient status and HRD status can predict treatment response. For example, the present disclosure provides a method for predicting a patient's (e.g., a triple-negative breast cancer patient's) response to a cancer treatment regimen that includes a DNA damaging agent (e.g., a platinum agent, e.g., cisplatin), an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising: determining, in cancer cells from a patient sample, the number of indicator CA regions (e.g., indicator LOH regions, indicator TAI regions, indicator LST regions, or any combination thereof) in at least one pair of human chromosomes in the cancer cells of the cancer patient; Determining whether cancer cells from the patient sample have a BRCA1 or BRCA2 defect (e.g., deleterious mutation, hyperpromoter methylation); and diagnosing a patient whose sample is either (a) a number of said indicator CA regions greater than a reference number, or (b) a deletion in BRCA1 or BRCA2, or both (a) and (b), as likely to respond to said cancer treatment regimen; The method may include the steps of:

[0132] Further specific embodiments Embodiment 1. An in vitro method for predicting patient response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor, comprising: (1) determining the number of indicator CA regions, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in a cancer patient's cancer cells in a sample containing cancer cells; and (2) diagnosing a patient having a higher likelihood of responding to the cancer treatment regimen than the reference number in the sample, the patient having a higher likelihood of responding to the cancer treatment regimen ... A method comprising:

[0133] Embodiment 2. The method of embodiment 1, wherein at least one pair of human chromosomes represents the entire genome.

[0134] Embodiment 3. The method of embodiment 1 or embodiment 2, wherein the indicator CA regions are determined in at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 21 pairs of human chromosomes.

[0135] Embodiment 4. The method of any one of embodiments 1 to 3, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0136] Aspect 5. The number of references of the indicator LOH region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more, and the number of references of the indicator TAI region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 1 5. The method of any of embodiments 1-4, wherein the reference number of indicator LST regions is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more.

[0137] Aspect 6. An indicator LOH region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases in length or more, but shorter than either a complete chromosome or a complete chromosome arm, and an indicator TAI region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 6. The method of any of embodiments 1-5, wherein the TAI region is defined as a TAI region that is 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more but does not extend beyond the centromere, and the indicator LST region is defined as an LST region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more in length.

[0138] Embodiment 7. The method of any of embodiments 1-6, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0139] Embodiment 8. The method of any one of embodiments 1 to 7, further comprising administering a cancer treatment regimen to a patient diagnosed as likely to respond to the cancer treatment regimen.

[0140] 9. An in vitro method for predicting patient response to a cancer treatment regimen that includes a platinum-based anticancer agent, comprising: (1) determining, in a sample containing cancer cells, the number of indicator CA regions, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in cancer cells of a cancer patient; (2) determining whether the sample containing the cancer cells is defective in BRCA1 or BRCA2; and (3) diagnosing a patient in which the sample has either (a) a number of indicator LOH regions, indicator TAI regions, or indicator LST regions greater than a reference number, or (b) a deletion in BRCA1 or BRCA2, or both (a) and (b), as likely to respond to the cancer treatment regimen. A method comprising:

[0141] Embodiment 10. The method of embodiment 9, wherein at least one pair of human chromosomes represents the entire genome.

[0142] Embodiment 11. The method of embodiment 9 or embodiment 10, wherein the indicator CA regions are determined in at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 21 pairs of human chromosomes.

[0143] Embodiment 12. The method of any one of embodiments 9 to 11, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0144] Aspect 13. The number of references of the indicator LOH region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more, and the number of references of the indicator TAI region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 13. The method of any of embodiments 9-12, wherein the number of references of the indicator LST region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more.

[0145] Aspect 14. An indicator LOH region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases in length or more, but shorter than either a complete chromosome or a complete chromosome arm, and an indicator TAI region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 14. The method of any of embodiments 9-13, wherein the TAI region is defined as a TAI region that is 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more but does not extend beyond the centromere, and the indicator LST region is defined as an LST region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more in length.

[0146] Embodiment 15. The method of any of embodiments 9 to 14, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0147] Embodiment 16. A method according to any of embodiments 9 to 15, wherein a sample is deficient in BRCA1 or BRCA2 if a deleterious mutation, loss of heterozygosity or hypermethylation in either BRCA1 or BRCA2 is detected in the sample.

[0148] Embodiment 17. The method of embodiment 16, wherein hypermethylation is detected if methylation is detected in at least 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, or 50% or more of the analyzed BRCA1 or BRCA2 promoter CpGs.

[0149] 18. An in vitro method for predicting patient response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor, comprising: (1) determining, in a sample containing cancer cells, the number of indicator CA regions, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in cancer cells of a cancer patient; (2) providing a test value derived from the number of indicator CA regions; (3) comparing the test value to one or more reference values ​​derived from the number of the indicator CA regions in a reference population; and (4) diagnosing a patient in whose sample the test value is greater than the one or more reference numbers as likely to respond to the cancer treatment regimen. A method comprising:

[0150] Embodiment 19. The method of embodiment 18, wherein at least one pair of human chromosomes represents the entire genome.

[0151] Embodiment 20 The method of embodiment 18 or embodiment 19, wherein the indicator CA regions are determined in at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 21 pairs of human chromosomes.

[0152] Embodiment 21. The method according to any one of embodiments 18 to 20, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0153] 22. The number of references of the indicator LOH region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more, and the number of references of the indicator TAI region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 22, 23, 24, 25, 30, 35, 40, 45, 50 or more and the number of references of the indicator LST region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more.

[0154] Aspect 23. An indicator LOH region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases in length or more, but shorter than either a complete chromosome or a complete chromosome arm, and an indicator TAI region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 23. The method of any of embodiments 18-22, wherein the TAI region is defined as a TAI region that is 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more but does not extend beyond the centromere, and the indicator LST region is defined as an LST region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more in length.

[0155] Embodiment 24. A method according to any of embodiments 18 to 23, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0156] Aspect 25. The method of any of aspects 18-24, further comprising the steps of: diagnosing a patient whose sample has a test value not greater than the one or more reference numbers as not likely to respond to the cancer treatment regimen; and (5)(a) recommending, prescribing, initiating or continuing a treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor to a patient diagnosed as not likely to respond to the cancer treatment regimen; or (5)(b) recommending, prescribing, initiating or continuing a treatment regimen not comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor to a patient diagnosed as not likely to respond to the cancer treatment regimen.

[0157] Aspect 26. The test value is calculating the arithmetic mean of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions in the sample as follows: Calculating the arithmetic mean of the number of indicator LOH regions, indicator TAI regions and indicator LST regions in samples from the reference population derived by TIFF0007680995000004.tif9155 and one or more reference values ​​as follows: 26. The method according to any one of embodiments 18 to 25, derived from TIFF0007680995000005.tif9154.

[0158] Embodiment 27. The method of any of embodiments 18-26, comprising diagnosing as likely to respond to said cancer treatment regimen a patient whose sample has a test value that is at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 times greater, or at least 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 standard deviations greater, or at least 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% greater than one or more reference numbers.

[0159] 28. A method of treating a cancer patient, comprising: (1) determining, in a sample containing cancer cells, the number of indicator CA regions, including an indicator LOH region, an indicator TAI region, and an indicator LST region, in at least one pair of human chromosomes of the cancer cells of the cancer patient; (2) providing a test value derived from the number of indicator CA regions; (3) comparing the test value to one or more reference values ​​derived from the number of the indicator CA regions in a reference population; and (4)(a) recommending, prescribing, initiating, or continuing a treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor to a patient whose sample has a test value greater than at least one of the reference values; or (4)(b) recommending, prescribing, initiating, or continuing a treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor to a patient whose sample has a test value not greater than at least one of the reference values. A method comprising:

[0160] Embodiment 29. The method of embodiment 28, wherein at least one pair of human chromosomes represents the entire genome.

[0161] Embodiment 30. The method of embodiment 28 or embodiment 29, wherein the indicator CA regions are determined in at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 or 21 pairs of human chromosomes.

[0162] Embodiment 31. The method according to any one of embodiments 28 to 30, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0163] 32. The number of references of the indicator LOH region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more, and the number of references of the indicator TAI region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 32. The method of any of embodiments 28-31, wherein the number of references of the indicator LST region is 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 or more.

[0164] Aspect 33. An indicator LOH region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases in length or more, but shorter than either a complete chromosome or a complete chromosome arm, and an indicator TAI region is defined as an LOH region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 33. The method of any of embodiments 28-32, wherein the TAI region is defined as a TAI region that is 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more but does not extend beyond the centromere, and the indicator LST region is defined as an LST region that is at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 30, 35, 40, 45, 50 megabases or more in length.

[0165] Embodiment 34. A method according to any of embodiments 28 to 33, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0166] Aspect 35. The test value is calculating the arithmetic mean of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions in the sample as follows: Calculating the arithmetic mean of the number of indicator LOH regions, indicator TAI regions and indicator LST regions in samples from the reference population derived by TIFF0007680995000006.tif9155 and one or more reference values ​​as follows: A method according to any one of embodiments 28 to 34, derived from TIFF0007680995000007.tif9154.

[0167] Embodiment 36. The method of any of embodiments 28-35, comprising diagnosing as likely to respond to said cancer treatment regimen a patient whose sample has a test value that is at least 2, 3, 4, 5, 6, 7, 8, 9 or 10 times greater, or at least 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 standard deviations greater, or at least 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% greater than one or more reference numbers.

[0168] 37. A method for assessing HRD in a cancer cell or its genomic DNA, comprising: (a) detecting in a cancer cell or genomic DNA derived therefrom an indicator CA region of said cancer cell in at least one pair of human chromosomes, wherein said at least one pair of human chromosomes is not the human X / Y sex chromosome pair; and (b) determining the total number of indicator CA regions in said at least one pair of human chromosomes; The method includes:

[0169] 38. A method for predicting the status of BRCA1 and BRCA2 genes in cancer cells, comprising: determining in the cancer cell the total number of indicator CA regions in at least one pair of human chromosomes in said cancer cell; and diagnosing a patient in whom said total number in cancer cells is greater than the reference number as having a high likelihood of having a defect in the BRCA1 gene or the BRCA2 gene; The method includes:

[0170] 39. A method for predicting HDR status in a cancer cell, comprising: determining in the cancer cell the total number of indicator CA regions in at least one pair of human chromosomes in said cancer cell; and diagnosing a patient in which the total number in cancer cells is greater than the reference number as likely to have HDR deficiency; The method includes:

[0171] 40. A method for predicting a cancer patient's response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising: determining in cancer cells from said cancer patient the number of indicator CA regions in at least one pair of human chromosomes in said cancer cells from said cancer patient; and diagnosing a patient having a higher likelihood of responding to said cancer treatment regimen, said patient having a higher likelihood of responding to said cancer treatment regimen, said patient having a higher likelihood of responding to said cancer treatment regimen, The method includes:

[0172] 41. A method for predicting a cancer patient's response to a treatment regimen, comprising: determining in said cancer cells from said cancer patient the total number of indicator CA regions in at least one pair of human chromosomes in said cancer cells from said cancer patient; and diagnosing a patient having a cancer cell count greater than a reference count as likely to be unresponsive to a treatment regimen comprising paclitaxel or docetaxel; The method includes:

[0173] 42. A method of treating cancer, comprising: (a) determining in a cancer cell from a cancer patient or in genomic DNA obtained therefrom the total number of indicator CA regions in at least one pair of human chromosomes in the cancer cell; and (b) administering to the cancer patient, if the total number of indicator CA regions is greater than a reference number, a cancer treatment regimen comprising one or more drugs selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors; The method includes:

[0174] Embodiment 43. Use of one or more drugs selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors, and PARP inhibitors for the manufacture of a medicament useful for treating cancer in a patient identified as having cancer cells determined to have a total of five or more indicator CA regions.

[0175] 44. A system for determining the LOH status of cancer cells in a cancer patient, comprising: (a) a sample analysis device configured to generate a plurality of signals related to genomic DNA of at least one pair of human chromosomes of the cancer cell; and (b) a computer subsystem programmed to calculate a number of indicator CA regions in the at least one pair of human chromosomes based on the plurality of signals; A system including:

[0176] 45. The computer subsystem compares the number of indicator CA regions to a reference number, (a) the likelihood of a defect in the BRCA1 and / or BRCA2 genes in said cancer cells; (b) the likelihood of HDR deficiency in said cancer cells; or (c) the likelihood that the cancer patient will respond to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor; The system of embodiment 8, programmed to determine:

[0177] Aspect 46. When executed on a computer, detecting the presence or absence of any indicator CA region present on one or more of the human chromosomes; and determining the total number of said indicator CA regions in one or more pairs of chromosomes; 2. A computer program product embodied in a computer-readable medium, the computer program product comprising:

[0178] 47. At least 500 oligonucleotides capable of hybridizing to multiple polymorphic regions of human genomic DNA; and 11. The computer program product according to embodiment 10, A diagnostic kit comprising:

[0179] 48. For determining the total number of indicator CA regions in at least one chromosome pair of human cancer cells obtained from a cancer patient, and (a) there is a high likelihood that the cancer cells are defective in the BRCA1 or BRCA2 gene; (b) a high likelihood of HDR deficiency in the cancer cell; or (c) the cancer patient has a high likelihood of responding to a cancer treatment regimen that includes a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, or a PARP inhibitor; The use of a plurality of oligonucleotides capable of hybridizing to a plurality of polymorphic regions of human genomic DNA to prepare a diagnostic kit useful for detecting

[0180] Embodiment 49. The method according to any of embodiments 37 to 42, wherein the indicator CA regions are indicator LOH regions, indicator TAI regions and indicator LST regions, optionally determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0181] Embodiment 50. The method of any one of embodiments 36 to 42, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0182] Embodiment 51. The method of any of embodiments 36 to 42, wherein the total number of indicator LOH regions, indicator TAI regions, or indicator LST regions is 9, 15, 20 or more.

[0183] Embodiment 52. The method of any of embodiments 36 to 42, wherein the indicator LOH region, indicator TAI region or indicator LST region is defined as having a length of about 6, 12 or 15 megabases, or more.

[0184] Embodiment 53. The method according to any one of embodiments 36 to 42, wherein the reference number is 6, 7, 8, 9, 10, 11, 12 or 13 or more.

[0185] Embodiment 54. The use according to embodiment 43 or 48, wherein the indicator CA regions are indicator LOH regions, indicator TAI regions and indicator LST regions, optionally determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0186] Embodiment 55. The use according to embodiment 43 or 48, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0187] Embodiment 56. The use according to embodiment 43 or 48, wherein the total number of indicator LOH regions, indicator TAI regions, or indicator LST regions is 9, 15, 20 or more.

[0188] Embodiment 57. The use according to embodiment 43 or 48, wherein the indicator LOH region, the indicator TAI region or the indicator LST region is defined as having a length of about 6, 12 or 15 megabases or more.

[0189] Embodiment 58. The system of embodiment 44 or 45, wherein the indicator CA regions are indicator LOH regions, indicator TAI regions and indicator LST regions, optionally determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0190] Embodiment 59. The system according to embodiment 44 or 45, wherein the cancer cells are ovarian cancer cells, breast cancer cells, or esophageal cancer cells.

[0191] Embodiment 60. The system of embodiment 44 or 45, wherein the total number of indicator LOH regions, indicator TAI regions, or indicator LST regions is 9, 15, 20 or more.

[0192] Embodiment 61 The system of embodiment 44 or 45, wherein the indicator LOH region, the indicator TAI region or the indicator LST region is defined as having a length of about 6, 12 or 15 megabases, or more.

[0193] Embodiment 62. The computer program product of embodiment 46, wherein the indicator CA regions are indicator LOH regions, indicator TAI regions and indicator LST regions, optionally determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0194] Embodiment 63. The computer program product of embodiment 46, wherein the cancer cell is an ovarian cancer cell, a breast cancer cell, or an esophageal cancer cell.

[0195] Embodiment 64. The computer program product of embodiment 46, wherein the total number of indicator LOH regions, indicator TAI regions, or indicator LST regions is 9, 15, 20 or more.

[0196] Embodiment 65. The computer program product of embodiment 46, wherein the indicator LOH region, the indicator TAI region or the indicator LST region is defined as having a length of about 6, 12 or 15 megabases, or more.

[0197] Embodiment 66. The method according to any one of embodiments 36 to 42, wherein at least one pair of human chromosomes is not human chromosome 17.

[0198] Embodiment 67. The use according to embodiment 43 or 48, wherein the indicator CA region is not in human chromosome 17.

[0199] Embodiment 68. The system of embodiment 44 or 45, wherein the indicator CA region is not in human chromosome 17.

[0200] Embodiment 69. The computer program product of embodiment 46, wherein the indicator CA region is not in human chromosome 17.

[0201] Embodiment 70. The method of embodiment 40 or 42, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0202] Embodiment 71. The use according to embodiment 48, wherein the DNA damaging agent is a platinum-based chemotherapy drug, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0203] Embodiment 72. The system of embodiment 45, wherein the DNA damaging agent is a platinum-based chemotherapy drug, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0204] Embodiment 73. The computer program product of embodiment 46, wherein the DNA damaging agent is a platinum-based chemotherapy drug, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0205] 74. (a) detecting an indicator CA region in a cancer cell or genomic DNA derived therefrom, the indicator CA region including at least two of an indicator LOH region, an indicator TAI region, or an indicator LST region in a representative number of pairs of human chromosomes in the cancer cell; and (b) determining the number and size of the indicator CA regions; The method includes:

[0206] Embodiment 75. The method of embodiment 74, wherein a representative number of pairs of human chromosomes is representative of the entire genome.

[0207] Embodiment 76 The method of embodiment 74, further comprising correlating an increase in the number of indicator CA regions of a particular size with a high likelihood of HDR loss.

[0208] Embodiment 77. The method of embodiment 76, wherein the specified size is greater than about 1.5, 2, 2.5, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, 75 or 100 megabases and less than the length of an entire chromosome containing the indicator CA region.

[0209] Embodiment 78. The method of embodiment 76 or 77, wherein 6, 7, 8, 9, 10, 11, 12 or 13 or more indicator CA regions of a particular size are correlated with a high likelihood of HDR loss.

[0210] 79. A method for determining the prognosis of a cancer patient, comprising: (a) determining whether a sample comprising cancer cells has an HRD signature, wherein the presence of indicator CA regions above a reference number including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions in at least one pair of human chromosomes in a cancer patient's cancer cells indicates that the cancer cells have an HRD signature; and (b)(1) diagnosing a patient in whom an HRD signature is detected in the sample as having a relatively favorable prognosis; or (b)(2) diagnosing a patient in whom the HRD signature is not detected in the sample as having a relatively poor prognosis; The method includes:

[0211] Aspect 80. A composition comprising a therapeutic agent selected from the group consisting of a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor and a PARP inhibitor for use in treating a disease which is a cancer selected from the group consisting of breast cancer, ovarian cancer, liver cancer, esophageal cancer, lung cancer, head and neck cancer, prostate cancer, colon cancer, rectal cancer, colorectal cancer and pancreatic cancer in a patient in which the indicator CA region in at least one pair of human chromosomes in the patient's cancer cells is greater than a reference number.

[0212] Embodiment 81. The composition of embodiment 80, wherein the indicator CA regions are determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0213] Embodiment 82. The composition of embodiment 80, wherein the total number of indicator CA regions is 9, 15, 20 or more.

[0214] Embodiment 83. The composition of embodiment 80, wherein the initial length is about 6, 12 or 15 megabases or more.

[0215] Embodiment 84. The composition of embodiment 80, wherein the reference number is 6, 7, 8, 9, 10, 11, 12 or 13 or more.

[0216] 85. A method of treating cancer in a patient, comprising: determining in said patient sample the number of indicator CA regions, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in a cancer cell of the cancer patient, indicating that the cancer cell has an HRD signature; providing a test value derived from the number of said indicator CA regions; comparing the test value to one or more reference values ​​(e.g., mean, median, tertile, quartile, quintile, etc.) derived from the number of the indicator CA regions in a reference population; and administering an anti-cancer drug to said patient, or recommending, prescribing, or initiating a treatment regimen including chemotherapy and / or synthetic lethal agents, based at least in part on said comparing step revealing that the test value is greater than at least one of said reference values ​​(e.g., at least 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; at least 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater); or recommending, prescribing, or initiating a chemotherapy and / or synthetic lethal agent-free treatment regimen based at least in part on said comparing step revealing that the test value is not greater than at least one of said reference values ​​(e.g., not 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; not 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater); The method includes:

[0217] Embodiment 86. The method of embodiment 85, wherein the indicator CA regions are determined in at least 2, 5, 10 or 21 pairs of human chromosomes.

[0218] Embodiment 87. The method of embodiment 85, wherein the total number of indicator CA regions is 9, 15, 20 or more.

[0219] Embodiment 88. The method of embodiment 85, wherein the initial length is about 6, 12 or 15 megabases or more.

[0220] Embodiment 89. The method of embodiment 85, wherein the reference number is 6, 7, 8, 9, 10, 11, 12 or 13 or more.

[0221] Embodiment 90. The method of embodiment 85, wherein the chemotherapy is selected from the group consisting of DNA damaging agents, anthracyclines and topoisomerase I inhibitors, and / or the synthetic lethal agent is a PARP inhibitor.

[0222] Embodiment 91. The method of embodiment 85, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, and / or the PARP inhibitor is iniparib, olaparib or velapirib.

[0223] 92. A method for assessing HRD in a cancer cell or its genomic DNA, comprising: (a) detecting, in a cancer cell or genomic DNA derived therefrom, an indicator CA region comprising at least two selected from an indicator LOH region, an indicator TAI region, or an indicator LST region in at least one pair of human chromosomes of the cancer cell, wherein the at least one pair of human chromosomes is not a human X / Y sex chromosome pair; and (b) determining an average (e.g., arithmetic mean) over the total number of indicator CA regions by calculating the average number of each type of indicator CA region detected in the at least one pair of human chromosomes (e.g., if there are 16 indicator LOH regions and 18 indicator LST regions, the arithmetic mean is calculated to be 17); The method includes:

[0224] 93. A method for predicting the status of BRCA1 and BRCA2 genes in cancer cells, comprising: determining an average (e.g., arithmetic mean) across a total number of each type of indicator CA region, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in the cancer cell; and correlating said average (e.g., arithmetic mean) across total numbers greater than the reference number with the likelihood of a defect in the BRCA1 or BRCA2 gene; The method includes:

[0225] 94. A method for predicting HDR status in a cancer cell, comprising: determining an average (e.g., arithmetic mean) across a total number of each type of indicator CA region, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in the cancer cell; and correlating said average (e.g., arithmetic mean) over a total number greater than the reference number with a likelihood of loss of HDR; The method includes:

[0226] 95. A method for predicting a cancer patient's response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and / or a PARP inhibitor, comprising: Determining an average (e.g., arithmetic mean) across the total number of each type of indicator CA region, including at least two types selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in the sample comprising cancer cells (e.g., if there are 16 indicator LOH regions and 18 indicator LST regions, the arithmetic mean is determined to be 17); and determining that those patients in which the average (e.g., arithmetic mean) across all samples is greater than a reference number are likely to respond to the cancer treatment regimen; The method includes:

[0227] 96. A method for predicting a cancer patient's response to a treatment regimen, comprising: determining an average (e.g., arithmetic mean) over a total number of indicator CA regions, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in a patient sample comprising cancer cells, in at least one pair of human chromosomes in said patient sample; and diagnosing patients in which the average (e.g., arithmetic mean) across all samples is greater than a reference number as likely to not respond to a therapeutic regimen comprising paclitaxel or docetaxel; The method includes:

[0228] 97. A method for treating cancer, comprising: (a) determining the average (e.g., arithmetic mean) across the total number of each type of indicator CA region, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes of the cancer cells or in a patient sample containing genomic DNA obtained therefrom; and (b) administering to patients in whose samples said total number of indicator CA regions is greater than the reference number a cancer treatment regimen comprising one or more drugs selected from the group consisting of DNA damaging agents, anthracyclines, topoisomerase I inhibitors and PARP inhibitors; The method includes:

[0229] Embodiment 98. The method of embodiment 95 or 97, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, or the PARP inhibitor is iniparib, olaparib or velapirib.

[0230] Aspect 99. A composition comprising a therapeutic agent selected from the group consisting of a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor and a PARP inhibitor for use in treating a disease in a patient in which the average (e.g., arithmetic mean) across each type of indicator CA region, including at least two types selected from indicator LOH regions, indicator TAI regions, or indicator LST regions, in at least one pair of human chromosomes in the patient's cancer cells is greater than a reference number, the disease being a cancer selected from the group consisting of breast cancer, ovarian cancer, liver cancer, esophageal cancer, lung cancer, head and neck cancer, prostate cancer, colon cancer, rectal cancer, colorectal cancer and pancreatic cancer.

[0231] Embodiment 100. A method of treating cancer in a patient, comprising: determining an average (e.g., arithmetic mean) of the total number of indicator CA regions in at least one pair of human chromosomes of cancer cells of the cancer patient in a sample from the patient, indicating that the cancer cells have an HRD signature; providing a test value derived from an average (e.g., an arithmetic mean) over a number of each type of indicator CA region, including at least two selected from indicator LOH regions, indicator TAI regions, or indicator LST regions; comparing the test value to one or more reference values ​​(e.g., mean, median, tertile, quartile, quintile, etc.) derived from the average (e.g., arithmetic mean) counts across each species of indicator CA region in a reference population; and administering an anti-cancer drug to said patient, or recommending, prescribing, or initiating a treatment regimen including chemotherapy and / or synthetic lethal agents, based at least in part on said comparing step revealing that the test value is greater than at least one of said reference values ​​(e.g., at least 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; at least 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater); or recommending, prescribing, or initiating a chemotherapy and / or synthetic lethal agent-free treatment regimen based at least in part on said comparing step revealing that the test value is not greater than at least one of said reference values ​​(e.g., not 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or 10-fold greater; not 1 standard deviation, 2 standard deviations, 3 standard deviations, 4 standard deviations, 5 standard deviations, 6 standard deviations, 7 standard deviations, 8 standard deviations, 9 standard deviations, or 10 standard deviations greater); The method includes:

[0232] Embodiment 101. The method of embodiment 100, wherein the average (e.g., arithmetic mean) of the indicator CA regions across each species is determined in at least 2, 5, 10, or 21 pairs of human chromosomes.

[0233] Embodiment 102. The method according to embodiment 100, wherein the chemotherapy is selected from the group consisting of DNA damaging agents, anthracyclines and topoisomerase I inhibitors, and / or the synthetic lethal agent is a PARP inhibitor.

[0234] Embodiment 103. The method of embodiment 100, wherein the DNA damaging agent is cisplatin, carboplatin, oxalaplatin or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is campothecin, topotecan or irinotecan, and / or the PARP inhibitor is iniparib, olaparib or velapirib.

[0235] Embodiment 104. The method of embodiment 1, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0236] Embodiment 105. The method of embodiment 104, wherein the reference number is 42.

[0237] Embodiment 106. The method of embodiment 9, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0238] Embodiment 107. The method of embodiment 106, wherein the reference number is 42.

[0239] Embodiment 108. The method of embodiment 18, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0240] Embodiment 109. The method of embodiment 108, wherein the reference number is 42.

[0241] Embodiment 110. The method of embodiment 28, wherein the reference number is 42.

[0242] Embodiment 111. The method of embodiment 37, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0243] Embodiment 112. The method of embodiment 111, wherein the reference number is 42.

[0244] Embodiment 113. The method of embodiment 38, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0245] Embodiment 114. The method of embodiment 113, wherein the reference number is 42.

[0246] Embodiment 115. The method of embodiment 39, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0247] Embodiment 116. The method of embodiment 115, wherein the reference number is 42.

[0248] Embodiment 117. The method of embodiment 40, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0249] Embodiment 118. The method of embodiment 117, wherein the reference number is 42.

[0250] Embodiment 119. The method of embodiment 41, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0251] Embodiment 120. The method of embodiment 119, wherein the reference number is 42.

[0252] Embodiment 121. The method of embodiment 42, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0253] Embodiment 122. The method of embodiment 121, wherein the reference number is 42.

[0254] Embodiment 123. The method of embodiment 79, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0255] Embodiment 124. The method of embodiment 123, wherein the reference number is 42.

[0256] Embodiment 125. The method of embodiment 85, wherein the indicator CA region is a combination of an indicator LOH region, an indicator TAI region and an indicator LST region.

[0257] Embodiment 126. The method of embodiment 125, wherein the reference number is 42.

[0258] Embodiment 127. An in vitro method for predicting patient response to a cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor, comprising: (1) determining, in a sample comprising cancer cells, the number of indicator CA regions, including indicator LOH regions, indicator TAI regions, and indicator LST regions, in at least one pair of human chromosomes of the cancer cells of the cancer patient; (2) combining the indicator CA regions to obtain a test value as follows: Test Value = (number of indicator LOH regions) + (number of indicator TAI regions) + (number of indicator LST regions); and (3) providing a reference value for comparison to the test value; The method includes:

[0259] Embodiment 128. The method of embodiment 127, wherein the reference value corresponds to the 5th percentile value of the indicator CA region score in a training cohort of HDR-deficient patients.

[0260] Embodiment 129. The method of embodiment 127 or embodiment 128, wherein the reference value is 42.

[0261] Embodiment 130. The method of any of embodiments 127 to 129, further comprising comparing the test value to a reference value.

[0262] Embodiment 131. The method according to any of embodiments 127 to 130, further comprising diagnosing a patient whose test value in the sample is above the reference value as likely to respond to the cancer treatment regimen.

[0263] Embodiment 132. The determining step comprises determining at least 150, 200, 250, 300, 350, 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2100, 2200, 2300, 2400, 2500, 3000, 3500, 4000, 4000, 5000, 6000, 7000, 8000, 9000, 10000, 11000, 12000, 13000, 14000, 15000, 16000, 17000, 18000, 19000, 21000, 22000, 23000, 24000, 25000, 30000, 35000, 40000, 40000, 50000, 60000, 70000, 80000, 90000, 10000, 110000, 12000, 13000, 14000, 15000, 16000, 17000, 18000, 19000, 23000, 24000, 25000, 26000, 27000, 28000, 29000, 30000, 310 , 450, 500, 600, 700, 800, 900, 1,000, 1,500, 2,000, 2,500, 3,000, 3,500, 4,000, 4,500, 5,000, 6,000, 7,000, 8,000, 9,000, 10,000, 11,000, 12,000, 13,000, 14,000, 15,000, 16,000, 17,000, 18,000, 19,000, 20,000, 25,000, 30,000, 35,000, 40,000, 45,000, 50,000, 60,000, 70,000 132. The method of any of embodiments 127-131, comprising assaying the sample to determine the copy number of each allele for 80,000, 90,000, 100,000, 125,000, 150,000, 175,000, 200,000, 250,000, 300,000, 400,000, 500,000, 600,000, 700,000, 800,000, 900,000, 1,000,000 or more polymorphic genomic loci.

[0264] Embodiment 133. The method of embodiment 132, wherein the determining step comprises assaying polymorphic genomic loci in at least 10 pairs of autosomes.

[0265] Embodiment 134. The method of embodiment 133, wherein the polymorphic genomic loci are in 22 pairs of autosomes.

[0266] Embodiment 135. The method according to any of embodiments 132 to 134, wherein the determining step comprises assaying the sample to measure the copy number of each allele for at least 5,000 polymorphic genomic loci in the autosomal pair.

[0267] Embodiment 136. The method of embodiment 135, wherein the determining step comprises assaying the sample to measure the copy number of each allele for at least 10,000 polymorphic genomic loci in the autosomal pair.

[0268] Embodiment 137. The method of embodiment 136, wherein the determining step comprises assaying the sample to measure the copy number of each allele for at least 50,000 polymorphic genomic loci in the autosomal pair.

[0269] The invention is further described by the following examples, which do not limit the scope of the invention described in the claims. EXAMPLES

[0270] Example 1 - LOH and TAI region scores across breast cancer subtypes and association with BRCA1 / 2 loss An LOH signature based on genome-wide tumor LOH profiles has been developed that is highly correlated with loss of BRCA1 / 2 and other HDR pathway genes in ovarian cancer (Abkevich, et al., Patterns of Genomic Loss of Heterozygosity Predict Homologous Recombination Repair Defects, BR. J. CANCER (2012)) and predicts response to DNA damaging agents (e.g., platinum-based neoadjuvant) therapy in breast cancer (Telli et al., Homologous Recombination Deficiency (HRD) score predicts response following neoadjuvant platinum-based therapy in triple-negative and BRCA1 / 2 mutation-associated breast cancer (BC), CANCER RES. (2012)). A second score based on TAI also showed a strong correlation with BRCA1 / 2 loss and predicted response to platinum-based drug therapy in triple-negative breast cancer (Birkbak et al., Telomeric allelic imbalance indicates defective DNA repair and sensitivity to DNA-damaging agents, CANCER DISCOV. (2012)). In this study, we investigated the frequency of BRCA1 / 2 loss and increased LOH or TAI region scores across breast cancer subtypes defined by ER / PR / HER2 status.

[0271] Frozen tumors were purchased from three commercial tissue biobanks. Approximately 50 randomly ascertained tumors from each of four breast cancer subtypes (triple negative, ER+ / HER2-, ER- / HER2+, ER+ / HER2+) were selected for analysis. A targeted custom hybridization panel was developed targeting BRCA1, BRCA2 and 50,000 selected SNPs across the entire genome. This panel was used in combination with sequencing on an Illumina HiSeq2500 to analyze tumors for BRCA1 / 2 somatic and germline mutations, including large-scale rearrangements, and SNP allele burden. BRCA1 promoter methylation was determined by qPCR assay (SA Biosciences). When available, DNA from normal tissue was used to determine whether deleterious mutations were germline or somatic.

[0272] SNP data were analyzed using an algorithm that determined the most likely allele-specific copy number at each SNP position. LOH region scores were calculated by calculating the number of LOH regions >15 Mb in length but shorter than the length of the entire chromosome. TAI region scores were calculated by calculating the number of telomeric regions with allelic imbalance >11 Mb in length but not spanning the centromere. Samples with poor quality SNP data and / or high normal DNA contamination were excluded. 191 of 213 samples yielded robust scores. (Table 2) BRCA1 / 2 deficiency in breast cancer IHC subtypes TIFF0007680995000008.tif54142 (Table 3) Mutation screening was performed on matched normal tissue from 17 BRCA1 / 2 mutants. Thirteen of the 17 individuals (76.5%) had germline mutations. TIFF0007680995000009.tif103128 * Each individual had one germline and one somatic mutation in BRCA1. (Table 4) Association between LOH score or TAI score and BRCA1 / 2 deletion TIFF0007680995000010.tif66168

[0273] Figure 5 shows LOH and TAI region scores by IHC subtype of breast cancer. 5A: LOH score; 5B: TAI score. Blue bars: BRCA1 / 2 deletion samples. Red bars: BRCA1 / 2 intact samples. Figure 6 shows the correlation between LOH and TAI region scores (correlation coefficient = 0.69). X-axis: LOH score; Y-axis: TAI score; Red dots: intact samples; Blue dots: BRCA1 / 2 deletion samples. The area under the dots is proportional to the number of samples with the LOH and TAI score combination (p = 10 -39 ).

[0274] Logistic regression analysis was used to predict BRCA1 / 2 deletion based on LOH and TAI scores. Both scores were significant in multivariate analysis (chi-squared values ​​for LOH were 10.8 and for TAI were 44.7; p=0.001 and 2.3, respectively). * 10 -11 The best model for discriminating between BRCA1 / 2-deficient and intact samples was 0.32. * LOH area score +0.68 * TAI domain score (p=9 * 10 -18 ).

[0275] Conclusion: Increased LOH and TAI region scores are each highly associated with BRCA1 / 2 loss in all subtypes of breast cancer; LOH and TAI region scores are highly significantly correlated; the combined CA region score (i.e., a combination of LOH and TAI) shows the best correlation with BRCA1 / 2 loss in this dataset. Based on this disclosure, the combination of LOH-HRD and TAI-HRD scores can predict response to DNA damaging agents and other agents (e.g., platinum-based drug therapy) in triple-negative breast cancer, allowing the extension of the use of platinum-based agents to other breast cancer subtypes.

[0276] Example 2 - LOH, TAI and LST region scores across breast cancer subtypes and association with BRCA1 / 2 loss As described in Example 1, the ratio of SNP allele frequencies was calculated and used to calculate LOH region score, TAI region score and LST region score. LST score was defined as the number of discontinuities between regions longer than 10 megabases with stable copy number after filtering out regions shorter than 3 megabases. We observed that LST score increases with ploidy in both intact and missing samples. Therefore, in this Example 2, instead of using a ploidy-specific cutoff, we modified LST region score by adjusting it by ploidy: LSTm=LST-kP, where P is ploidy and k is a constant. According to multivariate logistic regression analysis with missing as outcome and LST and P as predictors, k=15.5.

[0277] Of the 214 samples, 191 had scores that passed the QC criteria used. Thirty-eight of these samples were BRCA1 / 2 defective. The corresponding p-value from the Kolmogorov-Smirnov test for the LOH region score was 8. * 10 -12 and for the TAI domain score, * 10 -16 and for the LST domain score, 8* 10 -8 Of the 191 samples, 53 were triple-negative breast cancers, including 22 with BRCA1 / 2 deletions. The corresponding p-values ​​were 6.0 and 6.1 for the LOH region score, TAI region score, and LST region score, respectively. * 10 -6 , 3 * 10 -6 and 0.0002. The same analysis was performed for each individual breast cancer subtype, with significant p-values ​​for at least one score for every subtype (Table 5). The distribution of scores is shown for BRCA1 / 2-deficient versus BRCA1 / 2-intact samples in Figures 7A-C.

[0278] The scores were then analyzed to determine whether they correlated (Figure 2D-F). The correlation coefficient between the LOH and TAI region scores was 0.69 (p=10 -39 ), and the correlation between LOH and LST was 0.55 (p=2 * 10 -19 ), and the correlation between TAI and LST was 0.39 (p=10 -9 ) was.

[0279] Logistic regression analysis was used to predict BRCA1 / 2 loss based on LOH, TAI and LST region scores. All three scores were significant in multivariate analysis (chi-squared values ​​for LOH were 5.1 (p=0.02) and for TAI were 44.7 (p=0.02). * 10 -11 ) and 5.4 for LST (p=0.02). The best model for discriminating between BRCA1 / 2-deficient and intact samples in this dataset was 0.21 * LOH+0.67 * TAI+0.12 * LST(p=10 -18This Example 2 extends the conclusions from Example 1 (i.e., a model combining LOH and TAI region scores) to a model combining LOH, TAI and LST region scores.

[0280] Other clinical data that were available for many of the samples included stage, grade, and age at diagnosis. Stage information was available for 64 / 191 samples. The correlation coefficients between stage and LOH region score (0.07) and TAI region score (0.1) were not significant. Grade information was available for 164 / 191 samples. The correlation coefficients between grade and LOH region score (0.33) and TAI region score (0.23) were significant (p=2 for each). * 10 -5 and 0.004). Age at diagnosis was available for 184 of 191 samples. The correlation coefficient between age and LOH region score (-0.13) was not significant. The correlation coefficient between age and TAI region score (-0.25) was significant (p=0.0009). (Table 5) TIFF0007680995000011.tif126157

[0281] Example 3 - Arithmetic Means of LOH, TAI and LST Region Scores Across Breast Cancer Subtypes and Association with BRCA1 / 2 Loss The following study shows how the HRD score as described herein can predict BRCA1 / 2 deficiency and the efficacy of drugs targeting HR deficiency in triple-negative breast cancer (TNBC). To examine the rate of BRCA1 / 2 deficiency across breast cancer subtypes, breast tumor samples were assayed for BRCA1 / 2 mutations and promoter methylation. The three HRD scores described in Example 2 were determined for the samples, and then examined for association with BRCA1 / 2 deficiency using the arithmetic mean of LOH / TAI / LST scores. An analysis of the neoadjuvant TNBC cohort treated with cisplatin was further performed for the relationship between all three HRD scores and response.

[0282] Invasive breast tumor samples and matched normal tissues were obtained from three commercial sources. Samples were selected to provide approximately equal numbers of all breast cancer subtypes defined by IHC analysis for ER, PR, and HER2. BRCA1 promoter methylation analysis was performed by qPCR. BRCA1 / 2 mutation screening and genome-wide SNP profiles were generated using custom Agilent SureSelect XT capture followed by sequencing on an Illumina HiSeq2500. These data were used to calculate HRD-LOH, HRD-TAI, and HRD-LST scores.

[0283] SNP microarray and clinical data were downloaded from public repositories for the cisplatin-1 and cisplatin-2 trial cohorts. BRCA1 / 2 mutation data were not available for one of these cohorts. All three HRD scores were calculated from the public data and analyzed for association with response to cisplatin. The two cohorts were combined to increase power.

[0284] To calculate the HRD score, the SNP data were analyzed using an algorithm that determines the most likely allele-specific copy number at each SNP position. HRD-LOH was calculated by counting the number of LOH regions longer than 15 Mb but shorter than the length of the entire chromosome. HRD-TAI score was calculated by counting the number of regions with allelic imbalance longer than 11 Mb in length, spanning one of the subtelomeres but not across the centromere. HRD-LST score was the number of discontinuities between regions longer than 10 Mb after filtering out regions shorter than 3 Mb.

[0285] The composite score was the arithmetic mean of the LOH / TAI / LST scores. All p-values ​​were from logistic regression models with BRCA deficiency or response to cisplatin as dependent variables.

[0286] Table 6 shows the BRCA1 / 2 mutation and BRCA1 promoter methylation frequencies across the four breast cancer subtypes. BRCA1 / 2 variant analysis was successful in 100% of the samples, whereas large-scale rearrangement analysis was less robust, with 198 of 214 samples providing data that passed the QC criteria. Deleterious mutations were observed in 24 of 214 individuals (one had a somatic mutation in BRCA1 and a germline mutation in BRCA2). Corresponding normal DNA was available for 23 of the 24 mutants and was used to determine whether the identified mutations were germline or somatic. BRCA1 promoter methylation analysis was successful in 100% of the samples. Figure 9 illustrates the HRD scores in BRCA1 / 2-deficient samples. (Table 6) TIFF0007680995000012.tif43170 * Includes one individual who still retains an intact, functional copy of BRCA1. † Includes one individual whose BRCA1 functional status could not be determined.

[0287] Table 7 shows the association between the three HRD scores and BRCA1 / 2 defects in the all-participant breast cohort. The composite score was the arithmetic mean of the three HRD scores. (Table 7) TIFF0007680995000013.tif148170

[0288] Table 8 shows the association between HRD score and pCR (Miller-Payne 5) in TNBC treated with cisplatin in the neoadjuvant setting. Data were available from samples from the Cisplatin-1 trial (Silver et al., Efficacy of neoadjuvant Cisplatin in triple-negative breast cancer. J. CLIN. ONCOL. 28:1145-53 (2010)) and the Cisplatin-2 trial (Birkbak et al, (2012)). pCR was defined as patients with Miller-Payne 5 status after neoadjuvant treatment. The HRD-composite was the arithmetic mean of the three HRD scores. (Table 8) TIFF0007680995000014.tif38170

[0289] Conclusions: BRCA1 / 2 loss and increased HRD scores were observed in all breast cancer subtypes, and HRD scores detected BRCA1 / 2 loss. All three HRD scores predicted / detected treatment response to cisplatin in TNBC. The average (arithmetic mean) of the three HRD scores detected BRCA1 / 2 status in the all-participant breast cohort and cisplatin response in a second independent TNBC cohort. The arithmetic mean HRD-composite was a stronger predictor / detector of BRCA1 / 2 loss or treatment response than the individual HRD scores.

[0290] Example 4 - Multivariate analysis of BRCA1 / 2 status and DNA-based assays for homologous recombination deficiency Previous examples have described DNA-based scores measuring homologous recombination deficiency (HRD), which demonstrated that each score was significantly associated with BRCA1 / 2 deficiency, as well as the HRD-composite score, defined as the arithmetic mean of the three HRD scores. This example extends the results of previous examples by examining (1) the association of each of the three scores with the HRD-composite score, (2) the association of clinical variables with the HRD-composite score, and (3) the association of clinical variables and the HRD-composite score with BRCA1 / 2 deficiency.

[0291] Methods: The analysis in this Example 4 includes the same 197 patient samples as described in the previous Examples. Briefly, 215 breast tumor samples were purchased as fresh frozen specimens from three commercial sources. Samples were selected to provide approximately equal representation of breast cancer subtypes by IHC analysis for ER, PR, and HER2. 198 samples provided reliable HRD scores by Kolmogorov-Smirnov quality criteria. One patient who passed the HRD score had an unusual breast cancer subtype (ER / PR+ HER2-) and was therefore excluded from the analysis. Details of the patient's tumor and clinical characteristics are shown in Table 9.

[0292] Patient clinical data were provided for 91 variables, but data for most variables were insufficient to include in the analysis. Breast cancer subtype (TNBC, ER+ / HER2-, ER- / HER2+, ER+ / HER2+) was available for all patients. Other variables considered were age at diagnosis (obtained in 196 / 197 patients), stage (obtained in 191 / 197 patients), and grade (obtained in 190 / 197 patients). (Table 9) TIFF0007680995000015.tif127170

[0293] BRCA1 / 2 mutation screening and genome-wide SNP profiles were generated using custom Agilent SureSelect XT capture followed by sequencing on an Illumina HiSeq2500. Methylation of the BRCA-1 promoter region was determined by qPCR. Samples with >10% methylation were classified as methylated.

[0294] The HRD score was calculated from three HRD scores that are combined in the "HRD-composite score" discussed in this Example 4: genome-wide tumor loss of heterozygosity (LOH) profile (HRD-LOH), telomeric allelic imbalance (HRD-TAI) and large-scale state transitions (HRD-LST).

[0295] BRCA1 / 2 deficiency was defined as loss of function due to mutations in BRCA1 or BRCA2 or methylation of the promoter region of BRCA1 accompanied by loss of heterozygosity (LOH) in the affected genes.

[0296] All statistical analyses were performed using R version 3.0.2. All reported p-values ​​are two-sided. Statistical tools used included Spearman rank correlation test, Kruskal-Wallis one-way analysis of variance, and logistic regression.

[0297] For logistic regression modeling, HRD score and age at diagnosis were coded as numerical variables. Breast cancer stage and subtype were coded as categorical variables. Grade was coded both numerically and categorically, except where noted otherwise. Coding grade as a numerical variable is inappropriate unless the increased odds of BRCA1 / 2 deletion are the same when comparing grade 2 patients with grade 1 patients and when comparing grade 3 patients with grade 2 patients.

[0298] P-values ​​reported for univariate logistic regression models are based on partial likelihood ratios. Multivariate p-values ​​are based on partial likelihood ratios for the change in deviation between the full model (including all relevant predictors) and the reduced model (including all predictors excluding the predictor being evaluated and any interaction terms involving the predictor being evaluated). Odds ratios for HRD scores are reported with interquartile ranges.

[0299] Results: Pairwise correlations between HRD-LOH, HRD-TAI and HRD-LST scores were examined graphically (Figure 1) and quantified by Spearman rank correlation. Spearman rank correlation is preferred over the more commonly used Pearson product-moment correlation because the HRD score distribution has a right tail and outliers were observed. All pairwise comparisons of scores showed positive correlations that were significantly different from 0 (p<10 -16 ).

[0300] The degree of independent BRCA1 / 2 deletion information captured by each of the HRD-LOH, HRD-TAI and HRD-LST scores was measured by examining a multivariate logistic regression model with all three scores included as predictors of BRCA1 / 2 deletion status (Table 10). The HRD-TAI score captured significant BRCA1 / 2 deletion information independent of that provided by the other two scores (p=0.00016), as did the HRD-LST score (p=0.00014). At the 5% significance level, the HRD-LOH score did not add significant independent BRCA1 / 2 deletion information (p=0.069). (Table 10) TIFF0007680995000016.tif27163

[0301] Table 10 illustrates the results from a three-term multivariate logistic regression model using HRD-LOH, HRD-TAI and HRD-LST as predictors of BRCA1 / 2 loss.

[0302] To evaluate whether the HRD-composite score adequately captures the BRCA1 / 2 deletion information of its three components, we tested three bivariate logistic regression models. Each model included the HRD-composite score and one of the HRD-LOH score, HRD-TAI score, or HRD-LST score. None of the component scores added significantly to the HRD-composite score at the 5% significance level (HRD-LOH p=0.89, HRD-TAI p=0.090, HRD-LST p=0.28). This suggests that the HRD-composite score adequately captures the BRCA1 / 2 deletion information of the HRD-LOH score, HRD TAI score, and HRD-LST score.

[0303] Finally, the HRD-composite score was also compared to a model-based composite score optimized to predict BRCA1 / 2 loss in this patient set. The HRD-composite score weights the HRD-LOH, HRD-TAI, and HRD-LST scores equally, whereas the model-based score weights the HRD-TAI score approximately twice as much as the HRD-LOH or HRD-LST scores. The formula for the model-based score is: HRD-model = 0.11 × (HRD-LOH) + 0.25 × (HRD-TAI) + 0.12 × (HRD-LST) is given by:

[0304] The results from the univariate analysis (Table 11) show that the HRD-Model score outperforms the HRD-Composite score by approximately one order of magnitude (HRD model p=2.5×10 -25 , HRD-composite p=1.1×10 -24 ). (Table 11) TIFF0007680995000017.tif102163

[0305] Table 11 shows the results from univariate logistic regression. Odds ratios for HRD score are reported per IQR of score. Odds ratios for age are reported per year. Odds ratios for grades are per unit.

[0306] In the bivariate logistic regression model, the HRD-model score did not add significant independent BRCA1 / 2 deletion information to the HRD-composite score (p=0.089), further suggesting that the HRD-composite score adequately captures the BRCA1 / 2 deletion information of the HRD-LOH score, HRD-TAI score and HRD-LST score.

[0307] Associations between clinical variables and the HRD-composite score are shown in Figure 12. The HRD-composite score correlated significantly with tumor grade (Spearman correlation 0.23, p=0.0017). Correlations with breast cancer stage and age at diagnosis were not significantly different from 0 at the 5% level. According to Kruskal-Wallis one-way analysis of variance, the mean HRD-composite score differed significantly by breast cancer subtype (p=1.6×10 -5 ).

[0308] The heterogeneity of HRD-composite scores between clinical subgroups was examined by examining the significance of interaction terms in multivariate logistic regression models. For each clinical variable, we added an interaction term with HRD-composite scores to the model that included all clinical variables and HRD-composite scores. None of the interaction terms reached significance at the 5% significance level. Therefore, there is no evidence that the probability of BRCA1 / 2 deletion conferred by HRD-composite scores differs between clinical subgroups.

[0309] Similar tests for HRD-LOH, HRD-TAI and HRD-LST scores, respectively, indicated significant interactions between HRD-TAI score and age (p=0.0072) and grade (p=0.015), as well as between HRD-LST score and breast cancer subtype (p=0.021). Adjusting for multiple comparisons, only the interaction between HRD-TAI score and age remained significant at the 5% level (p=0.029). The significance of this interaction suggests that the increase in the odds of BRCA1 / 2 deletion per unit increase in HRD-TAI score diminishes with increasing age.

[0310] Associations between clinical variables and BRCA1 / 2 defects are presented in Figure 13. Clinical variables and HRD-composite scores were evaluated in univariate (Table 11) and multivariate (Table 12) logistic regression models. Odds ratios for HRD scores are reported per IQR. Odds ratios for age at diagnosis are reported per year. (Table 12) TIFF0007680995000018.tif76163

[0311] Table 12 shows the results from multivariate logistic regression. Odds ratios for HRD score are reported per IQR of the score. Odds ratios for age are reported per year.

[0312] In univariate analysis, each of the HRD scores (HRD-LOH, HRD-TAI, HRD-LST, HRD-composite and HRD-model) was significantly associated with BRCA1 / 2 loss. Higher scores indicated a higher likelihood of loss. Older age at diagnosis was significantly associated with a lower risk of BRCA1 / 2 loss (p=0.0071). Univariate results for breast cancer subtype and tumor grade (both categorical and numerical) were also statistically significant. Cancer stage was not associated with BRCA1 / 2 status.

[0313] In multivariate analysis, a model based on the HRD-composite score and all available clinical variables was considered. The HRD-composite score captured significant BRCA1 / 2 deletion information that was not captured by the clinical variables (p=1.2×10 -16 ). Of the available clinical variables, only age at diagnosis remained significant in the multivariate environment (p = 0.027). When grade was coded as a categorical variable, it was not statistically significant (p = 0.40). Grade was also not significant when coded as a numerical variable (p = 0.28). Quadratic and cubic effects on HRD-composite score were examined in a multivariate model including all clinical variables and were not statistically significant.

[0314] Discussion. In this Example 4, the frequency of BRCA1 / 2 deletions across the four subtypes of breast cancer defined by IHC subtyping ranged from approximately 9% to approximately 16%. Sequencing of matched tumor and normal DNA samples suggested that approximately 75% of the observed mutations were of germline origin. Although the predominant method for loss of the second allele in breast cancer is via LOH, approximately 24% of tumors also harbored subsequent somatic deleterious mutations in the second allele. In addition, a likely sporadic breast tumor was observed in one individual harboring a somatic deleterious mutation in BRCA2.

[0315] All three HRD scores showed strong correlation with BRCA1 / 2 deficiency, regardless of subtype, and the frequency of increased scores suggested that a significant proportion of all breast tumor subtypes harbor defects in the homologous recombination DNA repair pathway. Considering these findings, especially in conjunction with Example 3 above, agents that target or exploit DNA damage repair (e.g., platinum-based anticancer agents) may be effective across a subset of tumors from all subtypes of breast cancer (those with homologous recombination deficiencies as detected by the present disclosure).

[0316] In a clinical setting, these HRD scores, either alone or in combination, are best implemented using assays that are compatible with formalin-fixed, paraffin-embedded core needle biopsy samples ("FFPE"). These types of samples provide very low amounts and poor quality DNA. DNA extracted from these FFPE-processed samples often underperforms in SNP microarray analysis.

[0317] Liquid hybridization-based target enrichment techniques have been developed to prepare libraries for next-generation sequencing. These methods allow targeted sequencing of regions of interest after genomic complexity reduction, resulting in reduced sequencing costs. Preliminary tests indicate that available assays are compatible with DNA derived from FFPE DNA. In this Example 4, we report the development of a capture panel that targets approximately 54,000 SNPs distributed throughout the genome. The allele counts from the sequencing information obtained by this panel can be used for copy number and LOH reconstruction, and for the calculation of all three HRD scores. In addition, as in this Example 4, BRCA1 and BRCA2 capture probes can be included in the panel, which allows high-quality mutation screening for harmful variants in these genes with the same assay.

[0318] All three scores were significantly correlated with each other, suggesting that they all measure the same core genomic phenomenon, but logistic regression analysis indicates that combining the scores may yield a stronger association with BRCA1 / 2 defects in this dataset.

[0319] The combination of a robust score capable of identifying tumors with defects in homologous recombination DNA repair, and assay compatibility with formalin-fixed, paraffin-embedded clinical pathology specimens, facilitates the diagnostic identification and classification of patients likely to respond to agents that target double-stranded DNA damage repair. In addition, such agents may have utility across all subtypes of breast cancer in which HRD is detected according to the present disclosure.

[0320] Example 5 - High HRD Threshold (e.g., an example of an HRD signature) In this example, the determination of high HRD is demonstrated. The threshold reference value was selected to have high sensitivity for detecting HRD in breast and ovarian tumors that were non-specific for both treatment response and outcome. The total number of LOH, TAI and LST regions was determined. To calculate the HRD score, the SNP data was analyzed using an algorithm that determines the most likely allele-specific copy number at each SNP position. HRD-LOH was calculated by calculating the number of LOH regions that were longer than 15 Mb but shorter than the length of the entire chromosome. The HRD-TAI score was calculated by calculating the number of regions with allelic imbalance that were longer than 11 Mb and extended into one of the subtelomeres but did not cross the centromere. The HRD-LST score was the number of discontinuities between regions longer than 10 Mb after filtering out regions shorter than 3 Mb. The composite score (HRD score) was the sum of the LOH / TAI / LST scores.

[0321] The training set was collected from four different cohorts (497 breast and 561 ovarian). The set consisted of 78 breast and 190 ovarian tumors lacking functional copies of BRCA1 or BRCA2, because the distribution of HRD scores in BRCA-deficient samples is comparable to the distribution of scores in HRD samples overall. The threshold was set at the 5th percentile of the HRD scores in the training set, which gives a sensitivity of over 95% for detecting HR deficiency. High HRD (or HRD signature) was defined as having a reference score of 42 or higher (Figure 14).

[0322] Example 6 - HRD predicts response to cisplatin in triple-negative breast cancer This example demonstrates how the HRD score described herein can predict the efficacy of drugs targeting HR defects in triple-negative breast cancer (TNBC) samples. Analysis of a neoadjuvant TNBC cohort treated with cisplatin was examined against the relationship between all three HRD scores and response. All p-values ​​are from a logistic regression model with response to cisplatin as the dependent variable.

[0323] HR-deficient status was determined for 62 of 70 samples (70 individual patients) from the cisplatin cohort (8 patients had insufficient tumors for analysis). Of these, 31 (50%) were HR-deficient, 22 (35%) were non-HR-deficient, and 9 (15%) were indeterminate. Figure 15 presents a histogram showing the distribution of HRD scores in the cohort. A score of 42 or higher was considered to have high HRD (see also Example 5). The bimodality depicted in Figure 15 indicates that the HRD score effectively distinguished between HR-deficient and non-deficient tumor status. Pathological complete response (pCR), which is associated with long-term survival, was defined as 0 residual cancer burden (RBC), which was observed in 11 of 59 (19%) samples. Pathological response (PR) was defined as 0 or 1 RBC, which was observed in 22 of 59 (37%) samples. These overall response rates correlated with those expected for monotherapy.

[0324] Statistical analyses were performed according to a predefined statistical analysis plan (SAP), which included primary, secondary, and BRCA wild-type subset analyses.

[0325] In the primary analysis, HR deficient status was used to predict response in 50 samples. As shown in Table 13, HR deficient samples provided better predictors of response for both PR and pCR. For example, 52% of HR deficient samples had a pathological response compared to 9.5% of non-deficient samples. Similarly, 28% of HR deficient samples had a pathological complete response compared to 0% of non-deficient samples. TIFF0007680995000019.tif52163 (Table 13) Primary analysis using HR defect to predict response

[0326] In a secondary analysis, the quantitative HRD score described in Example 5 was used to predict response in 48 samples. As shown in Table 14, HRD scores, whether defined as PR or pCR, were significantly higher in samples from responders than non-responders. TIFF0007680995000020.tif38163 (Table 14) Secondary analysis using quantitative HRD scores to predict response

[0327] The distribution of HRD scores within each response class defined by BRCA mutation status for the secondary analysis is illustrated in Figure 16, where the dotted line at 42 represents the HRD threshold between low and high scores. The response curves for the secondary analysis, or the probability of PR associated with each value of the quantitative HRD score, are illustrated in Figure 17. The curves shown in Figure 17 were modeled by generalized logistic regression, which estimates four parameters: the shape of the curve, the scale, and the lower and upper limits. The dark boxes show the probability of response in HR-deficient and non-deficient samples versus each other. Table 15 shows that HR status remained significantly associated with pathological response in the secondary analysis. TIFF0007680995000021.tif63163 (Table 15) Multivariate model for pathological response * Odds ratio per IQR

[0328] The individual HRD component scores versus pathological response are shown in Table 16 and illustrated in Figure 18. Table 16 shows that each component score, i.e., LOH, TAI, and LST, predicted response and their sum, i.e., the HRD score, was equally significant or more significant than either of the individual components (p-value for HRD=3.1×10 -4 ). Figure 18 illustrates the strong pairwise correlations between component scores. TIFF0007680995000022.tif48163 (Table 16) Quantitative HRD component scores versus PR

[0329] In secondary analyses, we further explored the association between BRCA1 / 2 mutation status and response. Table 17 confirms that BRCA mutation status is associated with response; however, the association was not significant in this cohort (n=51), and BRCA mutation status was not as predictive as HR deficiency. TIFF0007680995000023.tif43163 (Table 17) Secondary analysis using BRCA mutation status to predict response

[0330] To demonstrate that HR deficiency is predictive in samples without BRCA1 / 2 mutations, a subset analysis was further performed using HR deficiency status in 38 BRCA wild-type samples. As shown in Table 18, HR deficiency samples provided a better response predictor for both PR and pCR in BRCA wild-type samples. For example, 52.6% of HR deficiency samples had pathological response, compared with 10.5% of non-deficient samples with pathological response. Similarly, 26.3% of HR deficiency samples had pathological complete response, compared with 0% of non-deficient samples with pathological complete response. TIFF0007680995000024.tif60163 (Table 18) Subset analysis using HR deficiency to predict response in BRCA wild-type samples

[0331] Subset analysis using quantitative HRD scores was further performed in 38 BRCA wild-type samples. As shown in Table 19, samples with high HRD (score 42 or higher) provided a better response predictor for both PR and pCR in BRCA wild-type samples. TIFF0007680995000025.tif33163 (Table 19) Subset analysis using quantitative HRD scores to predict response in BRCA wild-type samples

[0332] In conclusion, this example demonstrates that the sum of all three HRD scores significantly predicted response to cisplatin treatment in TNBC.

[0333] Other Aspects Although the present invention has been described in conjunction with a detailed description thereof, it is to be understood that the foregoing description is intended to be illustrative of the invention and not limiting, which is defined by the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

1. 1. An in vitro method for determining the homologous recombination (HR) deficiency status of cancer cells in a patient, comprising: (1) determining the combined number of indicator LOH regions, indicator TAI regions, and indicator LST regions in at least one pair of human chromosomes of the cancer cells in a sample containing the patient's cancer cells, (a) the indicator LOH region is a LOH region that is greater than 15 megabases in length but less than the full length of the respective chromosome in which it is located; (b) the indicator TAI region is a TAI region with allelic imbalance that (i) spans one of the subtelomeres, (ii) does not cross the centromere, and (iii) is greater than 11 megabases in length; (c) the indicator LST region is a somatic copy number discontinuity along the length of a chromosome that is between two regions that are at least 10 megabases in length after filtering out regions shorter than 3 megabases in length; step; (2) calculating a test value, the test value being the sum of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions; and (3) identifying the cancer cell as HR deficient if the test value is at or above the reference number of 32. A method comprising:

2. 2. The method of claim 1, wherein the test value is the arithmetic mean of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions, and the reference number is 11 or greater.

3. 2. The method of claim 1, wherein the reference number is 40 or more, optionally 42 or more.

4. 4. The method of claim 3, wherein the test value is the arithmetic mean of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions, and the reference number is 14 or greater.

5. The test value is the arithmetic mean of the number of indicator LOH regions, indicator TAI regions, and indicator LST regions in the sample as follows: The method according to any one of claims 1 to 4, wherein

6. The method of any one of claims 1 to 5, wherein at least one pair of human chromosomes is an autosome.

7. The human chromosome is an autosome, and The combined number of indicator LOH regions, indicator TAI regions, and indicator LST regions is determined in at least 10 pairs or optionally at least 15 pairs of autosomes; The method according to any one of claims 1 to 5.

8. The method of any one of claims 6 to 7, further comprising assaying at least 150 polymorphic genomic loci in each autosomal pair.

9. further comprising the step of assaying at least 5,000 polymorphic loci in at least 20 pairs of human chromosomes; the chromosome is an autosome, The method according to any one of claims 1 to 5.

10. identifying the patient as likely to respond to a breast or ovarian cancer treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor based on identifying the cancer cells as likely to be HR deficient. Further comprising: Optionally, the DNA damaging agent is cisplatin, carboplatin, oxaliplatin, or picoplatin, the anthracycline is epirubicin or doxorubicin, the topoisomerase I inhibitor is camptothecin, topotecan, or irinotecan, or the PARP inhibitor is iniparib, olaparib, or veliparib. The method of any one of claims 1 to 9.

11. identifying the patient as likely to respond to a breast or ovarian cancer treatment regimen that does not include a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, or a PARP inhibitor based on identifying the cancer cells as likely to not be HR deficient. Further comprising: Optionally, identifying the patient as likely to respond to a treatment regimen comprising one or more taxane agents, growth factor inhibitors or growth factor receptor inhibitors, or antimetabolites based on identifying the cancer cells as likely to not be HR defective. Further comprising: Optionally, the taxane agent is doxetaxel, paclitaxel, abraxane, the growth factor inhibitor or growth factor receptor inhibitor is erlotinib, gefitinib, lapatinib, sunitinib, bevacizumab, cetuximab, trastuzumab, panitumumab, or the antimetabolite is 5-fluorouracil, or methotrexate. The method of any one of claims 1 to 9.

12. The method of any one of claims 1 to 11, wherein the cancer cells are breast cancer cells or ovarian cancer cells.

Images