Cross-Cancer Analysis Clarifies Characteristics of Circulating Tumor DNA
February 19, 2014
NEW YORK (GenomeWeb News) – A study in Science Translational Medicine has started cataloguing the characteristics, constraints, and potential clinical applications of circulating tumor DNA.
Members of an international team led by investigators at Johns Hopkins University started by using tumor-guided PCR-based approaches to consider ctDNA levels in blood samples from hundreds of individuals with early or advanced cancers — a comparison that revealed both tumor stage and tumor type-related differences in the ability to detect ctDNA.
In addition to showing that ctDNA can be detected in instances where circulating cells cannot, the group went on to consider the sensitivities and specificities associated with using ctDNA to find treatment-related mutations and follow treatment response in individuals with metastatic colorectal cancer.
“This is really a survey of a variety of different [ctDNA] applications across multiple different tumor types,” co-corresponding author Luis Diaz, with the Ludwig Center for Cancer Genetics and Therapeutics at Johns Hopkins University’s Sidney Kimmel Comprehensive Cancer Center, told GenomeWeb Daily News.
Diaz and his colleagues started with samples from 136 individuals with metastatic forms of cancer from 14 tumor types. They also assessed samples from 41 individuals with primary glioma and medulloblastoma tumors — brain cancers that are often deadly, despite relatively rare metastasis — and from 10 ovarian or liver cancer cases that were advanced but not metastatic.
After using targeted-, exome-, or whole-genome sequencing and approaches developed by the Maryland-based company Personal Genome Diagnostics to find telltale point mutations or rearrangements in each individual’s tumor sample, the team turned to digital PCR in an effort to look for ctDNA, and assess its levels, in blood samples from the patients.
Results from that analysis indicated that almost all of the individuals with advanced cancers carried ctDNA in their blood. But there were exceptions: the researchers saw ctDNA in samples from fewer than 10 percent of those with glioma and less than half of samples from individuals with medulloblastoma.
The team detected variability in the proportion of patients with detectable ctDNA for individuals whose primary tumors were outside the brain, too.
Whereas ctDNA was almost always found in blood samples from individuals with pancreas, colon, breast, ovarian, and other cancer types, a minority of those with metastatic kidney, prostate, or thyroid cancer carried detectable amounts of tumor DNA in their blood.
“For the first time, we show that ctDNA is rare in patients with brain tumors and with certain types of tumors,” Diaz said, “while in other tumors such as ovarian cancer, colon cancer, and breast cancer it’s very abundant.”
Variability was also common amongst individuals with tumors that were not as advanced, the researchers reported. For instance, when they tested 223 individuals with localized cancers, the study’s authors detected ctDNA in more than half of the cases.
Again, though, cancer stage came into play, as did tumor type. While ctDNA turned up in more than 80 percent of samples from individuals with stage IV cancer, for example, that proportion dipped to 47 percent for samples from patients with stage I disease.
Even so, Diaz noted that the ability to find signs of cancer in blood samples from nearly half of those with stage I disease could prove useful for non-invasively diagnosing at least some cancer cases early, when relatively simple surgical interventions might still be possible.
The study also highlighted the potential of using ctDNA to find mutations for predicting patient outcomes, targeting cancer therapeutics, and monitoring treatment response and resistance.
In particular, the researchers demonstrated that they could uncover treatment-related KRAS gene mutations in blood samples from colorectal cancer patients with more than 87 percent sensitivity and specificity that exceeded 99 percent.
In another paper slated to appear in the same issue of STM, Italian researchers further illustrated the potential clinical applications of ctDNA in colorectal cancer. That study used ctDNA as a means of monitoring treatment response and resistance in metastatic colorectal cancer patients treated with cetuximab or panitumumab, treatments that target mutations in the epidermal growth factor pathway.
After identifying mutations to genes such as KRAS, NRAS, and BRAF that can reactivate pathways downstream of EGFR, members of the team took post-treatment blood samples from four individuals who initially responded to cetuximab or panitumumab but eventually became resistant to the drugs.
With the help of a PCR-based method known as BEAMing, the team used ctDNA to follow the development of various KRAS, NRAS, and BRAF mutations that coincided with drug resistance in the patients.
Using ctDNA as a proxy for finding resistance mechanisms present in metastatic tumors “turned out to be quite effective,” University of Torino oncology researcher Alberto Bardelli told GWDN.
“Many [colorectal cancer] patients undergo therapy relapse and that is detected in the blood,” explained Bardelli, senior author on the Italian study and a co-author on the ctDNA study led by Diaz and his colleagues.