Development of a novel cell-free DNA fragmentomics assay for monitoring disease progression in real-time for stage IV cancer patients.

e14544 Background: Advances in oncology have improved treatment response rates and overall survival. Many treatments, even targeted therapies, can fail to deliver an initial response to stage IV patients. We aimed to evaluate cfDNA as a biomarker of non-response to therapy in metastatic cancer. Methods: Blood samples were prospectively collected in Streck tubes at baseline and 12 - 21 days after initiating treatment, along with detailed clinical data, including imaging assessments captured in a cohort of metastatic non-small-cell lung (50), small-cell lung (8), breast (21), and colorectal (51) cancer study participants. To determine the fragmentation pattern of the DNA in the patient plasma, cfDNA was extracted using QIAamp Circulating Nucleic ACID Kit (Qiagen). cfDNA fragment concentrations were quantitated using two multiplexed qPCR assays on a QuantStudio 5. Primers and probes for the qPCR assay were designed using the high copy number retrotransposons Alu and SVA, represented across all 23 chromosomes. The two assays were performed on each specimen: an >80bp, >265bp and internal positive control (IPC) assay and an >105bp, >265bp with IPC assay. Following analytic validation of the assays, specimens from all participants were analyzed. The results from both assays were combined and merged with the clinical results and statistically analyzed. Results: 130 patients were included in the analysis. The qPCR assays work across all tumor and therapy types analyzed. A logit regression model was developed which calculates the likelihood that a patient’s cancer is progressing. The area under the ROC is 0.93 with a p-value of <0.001. Results clustered above the logit value of 0.65 (n=15) are all progressors by imaging (PPV=100.0%). All logit values below 0.045 (n=69) are non-progressors by imaging (NPV=0.0%). Bootstrap resampling (1,000 iterations) suggests a 99.3% PPV at the 0.9 logit cut point. Healthy subjects (9) and patients with a variety of conditions (33) known to create elevated fragmented levels of cfDNA were also analyzed to determine which patients are not appropriate patients for the assay. Results indicate patients with following conditions within 10 days of a blood draw should be excluded: stroke, myocardial infarction, severe acute autoimmune event, and severe acute COPD exacerbation; and patients treated with a chemotherapy agent (i.e., methotrexate) within 10 days of a blood draw. Conclusions: These data support the value of a blood-based biopsy test for evaluating cancer cfDNA. cfDNA fragment assay is applicable to lung, colorectal, and breast cancer, and is agnostic to any mutation status. The cfDNA fragment analysis is a promising biomarker to monitor patient treatment and save time and the expense of clinically ineffective therapies, allowing the medical oncologist to evaluate the response to therapy in real-time and provide the best care possible for the patient.