One of the more pressing concerns in medicine these days is the need for an effective screening test for lung cancer, the leading cause of cancer deaths in this country among both men and women. As with many cancers, the five-year survival rate is good if lung cancer is detected in its early stages. However, because this is seldom the case, the overall cure rate for lung cancer is just 5%. To present the problem in an even more stark perspective, lung cancer takes more lives each year than cancers of the breast, colon, cervix, and prostate combined.
Part of the problem is that most people do not experience symptoms from lung cancer until it has advanced to the point at which treatment is unlikely to offer a cure. Some studies have shown annual low-dose CT screening can detect lung cancer at an early stage, but this method is controversial, because no randomized controlled trials of the approach have yet been completed. There is also concern about the cost of this type of screening, the risks associated with radiation, and the adverse consequences of false-positive results. Consequently, guidelines do not recommend this type of screening.
Although researchers continue to evaluate the effectiveness of annual CT scans as a screening method for lung cancer, investigators are making progress with other potential screening methods that could be available to physicians and their patients within five years.
Breath test shows promise
One promising development involves the use of a highly sensitive device that investigators report can pick up signs of early-stage lung cancer through a simple, no-risk breath test. "We have developed what is, essentially, the world's most sensitive [breath analyzer]," says Michael Phillips, MD, CEO of Fort Lee, NJ-based Menssana Research, Inc. "It is about one billion times more sensitive than the kinds of [breath analyzers] that people use for alcohol measurement in the breath." The Menssana device picks up roughly 200 different volatile organic compounds in a person's breath, most of which are present in very low concentrations but reproducible and present in everyone's breath, says Phillips.
Although most of these compounds are probably just products of normal metabolism, investigators have found that several are biomarkers of disease. The FDA already has approved the company's breath test for heart transplant rejection, for example. And work is under way on breath tests for pulmonary tuberculosis, breast cancer, and ischemic heart disease.
However, with respect to lung cancer, over several years, investigators have used sophisticated pattern recognition analysis to pinpoint a pattern of volatile, organic compounds in the breath that is distinctive to the disease. And the company has just published a National Institutes of Health (NIH)-funded study showing that the test predicted lung cancer with nearly the same degree of accuracy as CT screening in a group of 404 smokers and nonsmokers over the age of 60.1 "Essentially, what we found was a confirmation of what we have found [in previous studies], only we did it rather more accurately this time," says Phillips, noting that it was a multicenter study that included a group of patients with confirmed lung cancer and a control group that was cancer-free, according to CT scans.
Phillips emphasizes that he does not see the breath test as a replacement or substitute for CT scanning in the diagnosis of lung cancer. However, he points out that it does offer certain advantages over CT as an initial screening option. "You can't get a chest CT for much less than $2,000. We can do [a breath test] for far less than that. And it is completely safe; there is no radiation involved," says Phillips. "It is very simple and very quick, and it only takes about two minutes of a patient's time."
Further, the process of collecting a breath sample is simple enough to be done in a doctor's office with the assistance of an inexpensive breath-collecting device, according to Phillips. The instruments needed for analyzing the breath sample are sophisticated and expensive, but Phillips notes that you only need to have one lab equipped to conduct these analyses, and it can analyze hundreds of samples. For every test, the lab uses a mathematical algorithm to determine the probability that a person has lung cancer. How these results ultimately will be reported has not been formalized as of yet, but Phillips suggests that the risk level could be reported as low, intermediate, or high, and there may be some numbers indicating the predictive value of the test.
Also yet to be determined is how the test would ultimately be applied in clinical practice. "That is something that needs to be worked on and negotiated in the future," says Phillips, noting that the next step for Menssana is a three-year NIH study designed to generate the kind of validation data that the FDA will require in order to approve the breath test for large-scale use. "All we have really done thus far is demonstrated that the test works."
Sputum test offers potential advantages
Researchers at the University of Maryland School of Medicine in Baltimore also are working on a screening test for lung cancer, but their approach works in a completely different way from the breath test. Investigators there are working to perfect an inexpensive and noninvasive gene probe that can be used to detect genetic signals for lung cancer in a person's sputum. This screening test works by checking to see whether two genes, which are believed to be tumor suppressors, are deleted in cells found in the sputum. In a just-published study funded by the National Cancer Institute researchers report that the test identified 76% of stage I lung cancer patients whose tumors exhibited the same genetic characteristics.2
Although the approach shows promise, the study was very small, involving fewer than 100 participants. And it is still not accurate enough for large-scale screening. "We know the sensitivity of the test is 76%, but even though that sensitivity is higher than current genetic tests, it is not high enough," says Feng Jiang, MD, PhD, assistant professor of pathology at the University of Maryland. "We need at least 90% sensitivity . . . for a diagnostic test." Consequently, in addition to working toward validating the early results from the approach investigators are in the process of enhancing the probe so that it can look for up to eight genes in the sputum. However, Jiang notes that the work is challenging, because investigators must isolate altered genes that are only predictive of cancer and not just cellular damage from smoking.
If researchers succeed in producing a test of sufficient sensitivity to detect lung cancer without excessive false-positive results, the approach could potentially offer several advantages over existing methodologies. For example, it is easy for patients to cough a sample into a cup in the doctor's office. And the test itself is easy enough to perform in a clinic lab, according to Jiang. Further, he notes that the test could also potentially be used to monitor lung cancer patients for their response to treatment or for early evidence of relapse.
CT screening: A controversial option
Until a newer screening method becomes available, CT is the only technique available that can reliably detect lung cancer at an early stage. However, because of concerns relating to cost and false-positive results, there is no consensus about when or whether annual CT screening makes sense. One organization strongly in favor of the use of annual CT screening in high-risk groups is the International Early Lung Cancer Action Program (I-ELCAP), an international collaboration of 38 medical institutions in seven countries that has been studying the value of annual CT screening since 1993.
Reporting on its research in October 2006, investigators said that lung cancer can be detected at its earliest stage in 85% of patients by using annual low-dose CT screening.3 Further, when followed by prompt surgical removal, the 10-year survival rate for these individuals is 92%. Of the more than 30,000 participants in the study, CT screening found 484 cases of lung cancer, including 412 cases at stage I. The participants included people who were considered at higher risk for the disease because of a history of smoking or exposure to known carcinogens such as asbestos, radon, beryllium, uranium, or secondhand smoke. However, contrary to these results, researchers involved with a longitudinal analysis of 3,246 asymptomatic current or former smokers who underwent annual CT screening for three to four years beginning in 1998 concluded that although CT screening may increase the rate of diagnosis of the disease, it may not meaningfully reduce the risk of advanced lung cancer or death from the disease.
As a result of these findings, investigators recommended against annual CT screening for asymptomatic patients until more conclusive data are available. The principal investigator involved with the I-ELCAP study, Claudia Henschke, PhD, MD, suggests that these data are flawed, because the study did not continue long enough to provide meaningful results. And she has the same problem with a randomized controlled trial looking at the value of annual CT screening that is now under way. "The traditional paradigm is that you do a randomized screening trial where you randomize people into screening or no screening, and in order to do that correctly you really have to do screening for some 10 years-and then it takes another two or three years to write up all of that," says Henschke, pointing out that the trial now underway only includes three rounds of screening. "That is a long-term process by which you hold back the new methodology for everyone because you have to wait for the results."
Henschke acknowledges that a randomized controlled trial would provide the answers that people are looking for if it continued for 10 years-and if there were enough participants involved-but she suggests that no country in the world can afford to do that kind of trial. Despite the controversy that surrounds the issue, participating I-ELCAP centers continue to offer CT screening to patients in high-risk groups-although insurance rarely picks up the tab. Further, Henschke points out that the group is continually refining its protocol to reduce the rate of false-positive results and enhance accuracy. "There are many findings that accumulate over [a person's] lifetime that may be present on a CT scan, so you need someone who is familiar [with those findings] and has experience and training with the protocol," she says.
References 1. Phillips M, Altork N, Austin J, et al. Prediction of lung cancer using volatile biomarkers in breath. Cancer Biomarkers, 2006; 22: 1-15. 2. Li R, Nevins T, Qiu Q, et al.N. Genetic Deletions in Sputum as Diagnostic Markers for Early Detection of Stage I Non-Small Cell Lung Cancer. Clinical Cancer Research, 2007; 13:482-487. 3. Henschke C, Yankelevitz D, Libby D, et al. Survival of Patients with Stage I Lung Cancer Detected on CT Screening. New England Journal of Medicine, 2006; 355:1763-1771. 4. Bach P, Jett J, Pastorino U, et al. Computed Tomography Screening and Lung Cancer Outcomes. Journal of the American Medical Association; 297:953-961.