A new computational framework designed to predict cancer-associated T cell receptors (caTCRs) may enable the early detection of the disease with just a blood sample, according to a report in this week's Science Translational Medicine. A group of UT Southwestern Medical Center scientists trained a deep-learning approach — called DeepCAT — by applying a computational method for detecting tumor-infiltrating T cell CDR3 sequences from RNA-sequencing data from thousands of samples from The Cancer Genome Atlas. The researchers then tested DeepCAT on blood samples from 13 clinical studies that included cancer patients, people with viral infections, and healthy individuals, and found that it could identify patients with breast, pancreatic, ovarian, and colorectal cancers with near perfect accuracy. DeepCAT was also able to identify caTCRs in blood samples from patients with early-stage kidney, ovarian, pancreatic, or lung cancer. The authors state that the approach does have certain limitations including the inability to determine a cancer's tissue of origin, and they note that inflammatory conditions might affect DeepCAT's performance. Thus, "the cancer score is not intended to replace the current diagnostic methods at this time," they write. "Rather, future efforts should be made to explore whether the combined use of the cancer score with existing screening modalities … can improve diagnostic accuracy in patients." 360Dx has more on this, here.
A new study has identified the loss of the DNA demethylase TET2 as a method by which cancer cells resist PARP inhibitors, a class of drugs used to treat BRCA-deficient breast and ovarian cancers. In these cancers, the loss of BRCA proteins leads to defective homologous recombination-dependent repair of DNA double-strand breaks. This makes the cancer cells sensitive to PARP inhibitors, which work by blocking the repair and restart of stalled replication forks, yet treatment resistance is common. To better understand potential resistance mechanisms, a team led by investigators from the National Cancer Institute performed a genome-wide RNAi screen in PARP-sensitive mouse embryonic stem cells deficient in BRCA2. As reported in Science Signaling, they find that resistance to multiple PARP inhibitors results from reduced TET2 expression. The scientists show that a product of TET2, called 5hmc, is produced at stalled replication forks and serves to recruit the base excision repair-associated endonuclease APE1. Without TET2, however, stalled replication forks were stabilized instead of degraded, both reducing PARP inhibitor sensitivity and contributing to genomic instability