While much has been said about how advances in genetic testing technology have improved the diagnosis of Mendelian genetic disease, this so-called "next generation" genetic technology is finding other niches. The ability of next-generation sequencing methods to quantitatively measure all genomic material in a laboratory specimen has led several groups to search for modest genetic signals representing a minority of the overall genetic material within a sample.
One area garnering increasing attention is the application of high-throughput sequencing to measure the presence of low levels of tumor DNA in blood as a marker of early metastatic disease.
Except in the most advanced and widespread terminal cancer cases, human cancer cells represent a relatively small population of undesirable cells among a far larger number of normal, ostensibly healthy cells in a patient. Like an unwelcome guest in an otherwise harmonious community, a cancer and its subsequent metastatic offshoots are initially very difficult to detect, because of their being composed of small numbers of cells and being too small in volume to produce a visibly detectable tumor.
Presumably, cancer cells detected at this subclinical stage could be more amenable to treatment and even cure if they could be reliably detected and measured. A large field of biomarker driven research has arisen directly to develop novel strategies for locating proverbial "needle in a haystack" markers of early cancer or early metastatic disease.
Since the discovery of small amounts of circulating cell-free tumor DNA in cancer patients in the past decade, the race has been on to exploit circulating tumor DNA as a possible biomarker of recurrent disease.
A recent paper published in the New England Journal of Medicine illustrates how quickly this approach is becoming applicable in the clinical setting (2013;368:1199-209).
The study focused on women with metastatic breast cancer undergoing systemic chemotherapy. The researchers investigated whether circulating tumor DNA in the women was detectable and had promise as a marker of disease progression and response to therapy. In parallel, the cancer marker CA 15-3 and an assay to capture and count circulating tumor cells were also studied.
Over the course of the 1-year study, 52 women were recruited, 30 of whom had DNA alterations that were believed to be detectable in blood by a genetic approach. These mutations represented DNA differences in the tumor cells from the patient, effectively providing a calling card for each cancer. At approximately 3-week intervals, blood was collected and subjected to either a targeted tumor-gene analysis or broad whole-genome analysis searching for the presence of the tumor-cell DNA calling-card signatures from cell-free DNA.
In the case of the targeted approach, two genes commonly mutated in breast cancer (PIK3CA and TP53) were selectively analyzed and found to contain mutations in 25 of the 52 patients.
The investigators performed whole-genome analysis in 9 of the 52 patients, comparing tumor tissue to matched normal tissue from each patient. This yielded eight patients in whom significant genome differences were found between tumor and normal tissue – and five of those eight patients had no PIK3CA or TP53 mutation.
Thus, 30 of the 52 women had identified genomic alterations. Using either the targeted or whole-genome approach across the 141 serial samples in these 30 patients with mutations, cell-free DNA was detected in 29 (97%) of the women and 115 (82%) of the samples. The one patient with no detectable tumor DNA had a low tumor burden and did not progress during the study.
The levels of cell-free tumor DNA correlated well with the response to therapy during the study, and the number of copies of the tumor DNA found in blood correlated with patient prognosis.
Overall, the study is principally remarkable for its value as a proof-of-concept study. It showed that tumor DNA was detectable in about 60% of the patients with metastatic disease, and that tumor DNA metrics correlated with clinical progression measures.
The two-gene targeted approach was probably more cost effective. However, over the long haul, the need to develop custom PIK3CA and TP53 mutation assays for each patient – and the soon-to-arrive ability of whole-genome approaches to detect PIK3CA and TP53 and many additional mutations at a lower cost – will likely make the genomic approach more favorable.
New algorithms also will need to be developed to improve the sensitivity of variant detection in the remaining 40% of cases that did not have a genetic tumor signature in this study.
In spite of the early stage conclusions from this study, it is tempting to imagine how cell-free tumor DNA analysis might have a role in the initial staging of a cancer diagnosis and as a measure of patient response to therapy.