This release describes very preliminary work on a lung cancer screening test that examines blood plasma from lung cancer patients and healthy controls. The work focused on isolating specific methylated DNA (cells that influence gene expression) markers in the blood that existed in cancer patients but which were absent in controls. The research will be presented in a poster session at an upcoming meeting of the American Association of Cancer Research (ACCR).
The release gets credit for employing mostly cautious language when describing results but it comes up short in other areas; namely it omits costs, and pertinent descriptions of benefits, availability and quality of evidence. In addition, the release doesn’t mention that the study involved smokers — one has to go to the abstract to find that information — and that may lead some readers to think the screening test could be appropriate for anyone, including healthy non-smokers.
In addition, red flags go up whenever we see the words “simple blood test” in a news release or story. There is no such thing, as we’ve addressed numerous times.
The release points out that, “Lung cancer is the leading cause of cancer mortality, resulting in 1.7 million deaths globally and more than 150,000 deaths in the United States every year.” (The American Cancer Society cites about 156,000 lung cancer deaths annually.) And since a lung cancer diagnosis can often come as the disease has progressed considerably, the dream of an earlier warning is enticing.
There is a use for a “simple blood test” to guide further diagnostic testing in lung cancer, such as being able to differentiate benign lung nodules picked up incidentally during a chest scan from cancer. Such a test could theoretically reduce unnecessary repeat CT scans.
However, readers should bring a healthy amount of skepticism to early reports about such tests and reserve confidence until substantive studies show strong, comprehensive, positive results. Otherwise, readers will suffer disappointment at believing in advances that fail to actually hold up to the hype.
There is no mention of the possible costs of a diagnostic test suggested by this release. Such DNA analyses carry costs that are not trivial and since the sponsor of the research mentions it already offers a diagnostic test for colorectal cancer that uses similar technology, it should be easy to include at least a reasonable estimate of the cost of such a test.
The news release presents the potential benefits of such a test with the following — “Using two independent regression modeling approaches, a panel of four novel methylated DNA markers demonstrated a sensitivity of 91-96 percent at a specificity of 90-94 percent.” While the release gets credit for giving the actual numbers for sensitivity and specificity, this is very complex language for the average reader.
The news release addresses harms, but in relation to CT scanning for lung cancer and its potential for over- and under-diagnosis, not to the screening test under development In regard to CT scans it states:
“This approach has a sensitivity for lung cancer above 90 percent, but its specificity may fall below 75 percent (a false positive rate of more than 25 percent) because indeterminate lung nodules are so common. Evaluation of these false-positives leads to unnecessary, costly and potentially harmful procedures.”
The suggestion here is that the blood test eliminates these problems with CT scans, but is that accurate? Are there other potential harms for a blood test that might be different from an imaging test? The release doesn’t say.
The news release mentions the size of the study, but should have offered more caveats to clarify that it was a very small, preliminary study.
The study involved a small selection of patients already known to have lung cancer or not. A more challenging test is whether it can maintain the same accuracy level in a broader population of people whose cancer status is unknown. Second, the release didn’t caution that there is no proof that using the test will improve lung cancer survival. Use of other blood-based screening tests — notably for ovarian cancer — have been shown not to improve survival.
It also notes that the research will be presented at a poster session at the annual meeting of the American Association of Cancer Research (AACR). Presenting a study abstract is a far cry from being published in a peer-reviewed journal.
There’s no obvious disease mongering in the release. The prevalence of lung cancer, and the problem of pulmonary nodules picked up incidentally, is accurately portrayed here.
However, the release never states that the study involved smokers. They do later talk about the test being used as an adjunct to screening, which is recommended for smokers. However, could a casual reader of the release be misled into thinking that the test is appropriate for anyone, and that healthy non-smokers should be considering it?
While the news release never states it specifically, it’s pretty clear that the company, Exact Sciences Corp., supported the research. However, it never really identifies any of the researchers involved, and only quotes two individuals — the CEO of the company and a researcher from the Mayo Clinic. The release does provide a link to the abstract of the research and it reveals that 17 of the 18 researchers listed work for Exact Sciences. Only the quoted Mayo researcher is not from the company and while the abstract does include disclosure information, it is not detailed enough to determine if there are any conflicts of interest.
The news release mentions that currently, “Lung cancer screening is approved for smokers using chest CT scanning,” and points out the shortcomings of that approach.
The release says nothing about when such a test might be available, if ever, and de-emphasizes that this research is in very early stages. As mentioned above, results presented at a poster session at a scientific meeting may be interesting but they may be years from actual use, if any.
We often see claims that simple blood tests will reveal a disease or condition but often these claims turn out to be premature and require much more research. While this release rightly pointed out that, “More studies are needed to corroborate accuracy” we think this study is too premature to demonstrate a novel advancement in lung cancer screening.
The use of the phrase “simple blood draw” raises a red flag, as noted in the review summary, above.