Amos Douvdevani, director of the laboratory of nephrology at Soroka University Medical Center and Ben-Gurion University in Be'er Sheva, Israel, and his colleagues have developed a new assay to detect significant levels of cell-free DNA in patients with cancer that is simpler to perform than currently available assays. The researchers used their test to evaluate patients with primary colorectal cancer and demonstrated that the new method was capable of reporting CFD levels that were prognostic of disease and outcome. Genome Technology's Matthew Dublin recently caught up with Douvdevani to discuss his work, published in the February issue of the American Journal of Clinical Pathology.
Genome Technology: Given that elevated cell-free DNA levels in cancer patients' blood were first reported in 1977, what was your motivation for this study?
Amos Douvdevani: That discovery was a milestone in diagnostic medicine because it was the first time the clinical potential of circulating nucleic acids as a molecular marker in cancer had been explored. Many studies have confirmed this initial observation and found elevated CFD levels in patients with malignancies, including patients with CRC [colorectal cancer]. However, an unacceptable false-positive rate and prohibitive costs, coupled with the lack of specificity, have led to recommendations against CEA [carcinoembryonic antigen] testing for CRC screening. So we recently developed a convenient DNA assay, applied directly to biologic samples, that works by adding diluted fluorochrome to the samples that is capable of measuring fluorescence in an accurate, sensitive, and reproducible way. The aim of this study was to evaluate this new method ... in an animal model and in patients with CRC.
GT: What further developments are needed to implement this assay in the clinic?
AD: For implementation of this assay for clinical use, the assay has to be performed by an automated machine. Since this assay requires only mixture of serum samples with a fluorescent reagent and measurement of fluorescence, automation is relatively simple. In addition, normal and cutoff values have to be established in large populations.
GT: How do your findings demonstrate the promise of this new technique?
AD: Our assay reflects tumor progression in an animal model. In mice that were inoculated with cancer cells, tumor growth and CFD levels positively correlated and concurrently peaked at day 14 and declined with tumor rejection at day 18. Notable is the elevation of CFD levels at day six, which preceded the appearance of the tumor at day eight. Also important is the associated decline of CFD levels with tumor rejection, which suggests the possible use of this marker not only for detection and follow-up, but also for ... response to treatment. The CFD levels in our cohort of 38 patients with CRC were significantly higher than in the control group, with a median CFD value of more than double that of the control group. These levels were prognostic ... at one year from resection.
We also compared the CFD levels in our patients with CRC with CEA levels, the best available serologic marker for recurrence of CRC. CFD was advantageous in detecting almost 50 percent of the samples from patients with low-stage tumors.
Our data indicate that combining the two markers might be beneficial for the detection of primary tumors. When both markers were combined, 50 percent of patients were positive for one or two markers. Reduction of the CFD cutoff to a specificity of 82 percent increased the combined detection sensitivity to 80 percent.
GT: Going forward, how do you plan to further evaluate the efficacy of this new technique?
AD: A large prospective study will need to be performed to truly evaluate the efficacy of this method.