Method of treating cancer
Inventors
Abkevich, Victor • Gutin, Alexander • Timms, Kirsten • Lanchbury, Jerry
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11174519-B2
Publication Date
2021-11-16
Expiration Date
Abstract
This document provides methods and materials involved in assessing samples (e.g., cancer cells) for the presence of a loss of heterozygosity (LOH) signature. For example, methods and materials for determining whether or not a cell (e.g., a cancer cell) contains an LOH signature are provided. Materials and methods for identifying cells (e.g., cancer cells) having a deficiency in homology directed repair (HDR) as well as materials and methods for identifying cancer patients likely to respond to a particular cancer treatment regimen also are provided.
Core Innovation
The document describes cancer diagnostics and treatment selection by quantifying loss of heterozygosity (LOH) regions in tumor genomes. An LOH signature is defined based on multiple LOH regions that are longer than a specified megabase threshold and shorter than the length of the whole chromosome containing the LOH region, excluding X/Y sex chromosomes and optionally excluding chromosome 17. The LOH regions used for the signature are referred to as “Indicator LOH Regions.”
The disclosure correlates a positive LOH signature status with an increased likelihood that a cancer patient will have likelihoods associated with homologous recombination deficiency, including HDR deficiency and/or BRCA1/BRCA2 deficiency. The LOH signature status is also linked to increased likelihood of responding to treatment regimens that include DNA damaging agents and/or regimens comprising anthracyclines, topoisomerase I inhibitors, PARP inhibitors, and/or radiation.
The document further provides computational concepts for reconstructing LOH regions from SNP array or sequencing data using allele-specific copy number reconstruction and likelihood-based inference. A diagnostic kit/system/computer program is described in connection with defining LOH signatures and using the LOH-region metrics to predict response and select regimens, including based on training and validation analyses using ovarian tumor data and chemosensitivity stratification using cell line assays.
The disclosed treatment-selection methods include predicting response to regimens by determining a total number of LOH regions in at least one pair of human chromosomes in a cancer cell with LOH regions bounded by a length threshold, where the chromosome pair is not a human X/Y sex chromosome pair. The predicted outcome is then used to administer a regimen that includes DNA damaging agents and specified classes of agents such as anthracyclines, topoisomerase I inhibitors, radiation, and PARP inhibitors.
Claims Coverage
The document provides two independent methods of treating cancer (predicting response and then administering therapy). Across the independent claims, the central inventive features relate to determining a total number of LOH regions within defined chromosome-pair and length boundaries, correlating that total with an increased likelihood of response or non-response using a reference number, and selecting administration of a regimen that includes specified classes of DNA damaging agents.
LOH-region counting for non-X/Y chromosome pairs
Determining, in the cancer cell, the total number of LOH regions in at least one pair of human chromosomes of said cancer cell that are longer than a first length but shorter than the length of the whole chromosome containing the LOH region, wherein said at least one pair of human chromosomes is not a human X/Y sex chromosome pair, wherein said first length is about 15 or more megabases, wherein the cancer cell is a breast cancer cell or an ovarian cancer cell.
LOH correlation to response likelihood with a threshold reference number
Correlating said total number of LOH regions that is greater than a reference number that is at least 6 with an increased likelihood that said cancer patient will respond to said cancer treatment regimen.
Regimen administration based on LOH correlation for response prediction
Based on correlating said total number of LOH regions with an increased likelihood that said cancer patient will respond to said cancer treatment regimen, administering to the cancer patient a treatment regimen comprising a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and/or a PARP inhibitor.
LOH correlation to non-response likelihood for paclitaxel/docetaxel regimens
Correlating said total number of LOH regions that is greater than a reference number that is at least 6 with an increased likelihood that said cancer patient will not respond to the treatment regimen including paclitaxel or docetaxel.
Regimen administration based on LOH correlation for non-response prediction
Based on correlating said total number of LOH regions with an increased likelihood that said cancer patient will not respond to said cancer treatment regimen, administering to the patient a treatment regimen including a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and/or a PARP inhibitor.
In both independent claims, the predictive core is counting a total number of LOH regions defined by length boundaries and by excluding X/Y sex chromosomes, then correlating that total (greater than a reference number) with increased likelihood of response or non-response. Administration of a regimen containing DNA damaging agents, anthracyclines, topoisomerase I inhibitors, radiation, and/or PARP inhibitors is performed based on the LOH-based prediction.
Stated Advantages
Increased likelihood that a cancer patient will respond to cancer treatment regimens including DNA damaging agents, anthracyclines, topoisomerase I inhibitors, radiation, and/or PARP inhibitors, based on correlating LOH-region metrics with a reference number.
Increased likelihood that a cancer patient will not respond to treatment regimens including paclitaxel or docetaxel, based on correlating LOH-region metrics with a reference number.
Documented Applications
Predicting a breast cancer or ovarian cancer patient’s response to cancer treatment regimens that include a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and/or a PARP inhibitor, using a LOH-based determination and correlation of a total number of LOH regions.
Predicting non-response of a cancer patient to a treatment regimen including paclitaxel or docetaxel, and administering a regimen including a DNA damaging agent, an anthracycline, a topoisomerase I inhibitor, radiation, and/or a PARP inhibitor based on the LOH-based prediction.
Computational reconstruction of LOH regions from SNP array or sequencing data using allele-specific copy number reconstruction and likelihood-based inference, in support of defining LOH signatures and diagnostic kit/system/computer program concepts.
Training/validation analyses using ovarian tumor training/validation datasets showing large LOH region counts correlate with HDR deficiency.
Cell line chemosensitivity assays stratifying response versus non-response using a threshold number of large LOH regions.
Interested in licensing this patent?