Whole-genome error corrected sequencing for early stage melanoma detection without matched tumors
Description: Cancer monitoring tools are underdeveloped, but liquid biopsies aim to revolutionize this space by profiling for disease from circulating tumor DNA (ctDNA; DNA that is shed from tumor cells and enters the circulation). However, many tools rely on sequencing a patient’s tumor to create a list of mutations to mine for in the blood. Therefore, patients with scant or unavailable tumor material cannot be monitored with current methods.
I propose to develop error-corrected whole genome sequencing to monitor patients without matched tumor sequencing. The principle behind my research is that dramatically reducing the error rate of sequencing technologies (currently ~1:1000) will allow for the robust identification of cancer mutations from ctDNA alone. For example, we will be able to profile for characteristic mutations of ultraviolet light induced DNA mutations in the plasma of skin cancer patients, and differentiate those mutations from aging related mutations in the plasma.
To achieve this reduction in errors, I will develop a molecular barcoding strategy to tag double-stranded DNA molecules with a unique DNA identifier. Sequencing the identifier and the matching DNA strands will allow me to computationally link the two strands of a DNA molecule. I will develop bioinformatic tools to scan for variants in both DNA strands, and only variants found in both strands will be considered as a potential cancer signal. Importantly, pre-analytical errors (such as sequencing or sample preparation errors) are unlikely to occur on both strands of the DNA, as opposed to true mutations. To accomplish this project in a cost-effective manner, which is essential for clinical deployment, I will leverage the Ultima Genomics sequencer, a new sequencing machine capable of producing high-throughput data at an affordable price (1USD/GB).
I will apply my technology to the challenging context of early stage (stage II) melanoma. By definition, these cancers have not metastasized to the lymph nodes, which could have been used for tumor-sequencing. Treatment of these patients with adjuvant immunotherapy further decreases tumor burden, making ctDNA concentrations low in this patient population. However, there is a high risk of recurrence in patients, highlighting the need for continuous monitoring. I will apply my technology to a cohort of 60 patients who underwent resection to evaluate postoperative ctDNA tumor burden. Preoperative ctDNA burdens will be correlated with the extent of disease and postoperative burdens will be correlated to disease recurrence.
Finally, I will develop a complementary technology to simultaneously measure mutations and methylation of ctDNA. Methylation changes have been linked as an early biomarker of melanoma, yet current methods to probe these changes do so at the expense of mutation identification. I will develop an approach to simultaneously profile cytosine methylation markers and mutations. Taken together, this approach would provide an impactful noninvasive biomarker for melanoma, particularly in low burden disease scenarios.
Funder: American Cancer Society
Amount: $217,500
Funding type: award
Start date: 2024-01-01
End date: 2026-12-31
Investigator: Alexandre Pellan Cheng
Institution: Weill Medical College of Cornell University
Subfield: Cancer Research
Domain: Life Sciences
Funded outputs count: 0
Funder award ID: PF-23-1073957-01-CDP
OpenAlex ID: G4014615667