Novel test detects lung cancer in smokers’ nasal passages
Researchers have identified a 30-gene biomarker in the nasal passages of current and former smokers that could help distinguish between benign and malignant lung lesions.
“About one of three surgical biopsies that we perform in this country for suspicion of lung cancer turn out to benign disease. This means that we are subjecting a lot of people to a lot of unnecessary procedures,” Avrum Spira, MD, MSc, professor of medicine, pathology and bioinformatics at Boston University School of Medicine, told HemOnc Today. “Our 30-gene biomarker ... improves the ability of the physician to detect cancer, and it allows physicians to be relatively confident that there is no cancer present when the biomarker is negative.”
Spira and colleagues are investigating new approaches to improve early detection of lung cancer among those with indeterminate lesions identified on CT scan.
They previously identified and validated a bronchial epithelial gene expression biomarker that detects lung cancer in current and former smokers.
Bronchial and nasal epithelial gene expression are similarly altered by cigarette smoke exposure; therefore, the researchers aimed to determine if cancer-associated gene expression could potentially be detected in the nasal epithelium of 505 current and former smokers who underwent diagnostic testing for pulmonary lesions suspicious for lung cancer.
The results, published in Journal of the National Cancer Institute, showed that the activity of a set of 535 genes was altered in the nasal passages of smokers who subsequently received lung cancer diagnoses compared with those who had benign lung disease.
HemOnc Today spoke with Spira about the study results and what additional research must be conducted before the test is ready for widespread use.
Question: What prompted you and your colleagues to conduct this study?
Answer: There is a critical unmet need for early detection of lung cancer. Lung cancer is the leading cause of cancer death among both men and women, primarily because it is often diagnosed at a late stage. During the past 5 to 6 years, we have begun annually screening high-risk smokers for lung cancer with CT scans to the chest, with the idea that perhaps we can detect lung cancer earlier and reduce mortality. The problem we are beginning to face is that we cannot determine whether many of the nodules or lesions we find on CT are cancerous. As a result, we are subjecting many of these people to unnecessary invasive biopsy of their lungs, often for benign disease.
Q: How did you conduct the study?
A: We leveraged a large cohort of more than 500 patients, all of whom were underwent a diagnostic workup for suspicion of lung cancer because of a lesion found on CT scan to the chest. We performed nasal swabs and measured genomic activity of a large number of genes inside of the nasal cells. In this prospective study, we sampled the nose at the time they underwent diagnostic work-up, and we followed them for up to 1 year to see whether the nodules we found on CT scan turned out to be benign or malignant. This helped show whether our genomic biomarker accurately predicted whether the patient would be diagnosed with lung cancer.
Q: What did you find?
A: We found alterations in the activity of a set of 535 genes in the nasal passages of smokers who went on to be diagnosed with lung cancer. Intriguingly, the genes overlapped with genes we previously had shown changed in the windpipe or lower airway of smokers who had lung cancer. About 2 years ago, we published a paper in The New England Journal of Medicine that showed one could brush the airway cells that line the windpipe in smokers who have lesions on CT scans of the chest, and the genomic activity in those cells can tell whether the lesion is malignant. This test, Percepta (Veracyte), now is commercially available. However, the test can only be applied to patients who undergo bronchoscopy as part of their diagnostic workup in order to collect the airway cells. Not all patients that have lesions on their scan undergo this diagnostic procedure. This most recent paper moves our approach to the nasal passages, which are easier to sample and do not require that the patient undergoes a diagnostic bronchoscopy.
The genes that change in the nose when someone has cancer change in a very similar fashion as those in the windpipe. This gives us the confidence that what we are measuring in the nose is reflective of what is happening more deeply in the lungs.
Q: What did you do next?
A: We took those 535 genes and found the optimal combination that could work together as a biomarker. This turned out to be 30 genes. We developed this biomarker on 375 patients within a training set, then we validated them in an independent set of nose samples from an additional 130 patients. We demonstrated that this biomarker can detect the presence of lung cancer. More importantly, it adds information on top of clinical risk models that physicians use to decide whether a lung nodule is likely to be cancer.
Q: Can you describe the potential clinical impact?
A: We are seeing an epidemic of lung nodules in our clinics that are detected incidentally or upon CT chest screening. The vast majority of these nodules are benign. However, physicians struggle to identify which of these patients they should send for invasive lung biopsy — which carries many risks to patients and also increases health care costs — and which ones can be monitored with repeat imaging to see if the lesion grows over time. The test we developed gives physicians and patients a better sense of whether the nodule on the CT scan is benign or malignant. It provides a noninvasive way to inform the physician and patient about how aggressive to be in the diagnostic workup.
Q: What are the next steps in research?
A: We have plans for additional research based on the biological premise for this study, the so-called ‘field of injury.’ The concept is that — when one is exposed to a cancer-causing agent in the environment — all of the cells that line the respiratory tract are exposed to everything that is inhaled, and there are many genomic alterations as a result. The 15% to 20% of lifetime smokers that develop lung cancer have a different pattern of genomic damage in these cells that line the airway, and that is what allows us to use these cells as the ‘canary in a coalmine.’ Although the initial application of the test is for diagnosing the presence of cancer, one area for additional research is to see whether these changes in the airway genes occur years before someone gets lung cancer. If they do, we can imagine using this as a screening tool as opposed to a diagnostic test for people with nodules. This could provide us with a more accurate molecular tool to select high-risk individuals who are more likely to benefit from annual CT chest screening.
Q: How long will it be before this test is widely used?
A: We are several years out from the nasal diagnostic test going prime time. We are in the midst of additional clinical studies — which include hundreds of patients with indeterminate pulmonary nodules — to confirm that this test has high sensitivity for detecting lung cancer. These studies include individuals who did not undergo bronchoscopy as part of their diagnostic workup. Upon completion of these studies within the next 2 to 3 years, this test should move into the clinic with a commercial partner. As for the screening version, we are further away from primetime use, as this is a much more difficult set of studies to complete.
Q: Is there anything else that you would like to mention?
A: The paradigm that underlies our test, in many respects, could be applied to other cancer types beyond lung cancer. The idea that we can sample cells that are relatively accessible and pick up the presence of cancer deep in an organ that is hard to biopsy is one that I think we will see moving outside of the lung and into other cancers, such as esophageal cancer. – by Jennifer Southall
Silvestri GA, et al. N Engl J Med. 2015;doi:10.1056/NEJMoa1504601.
Spira A, et al. J Natl Cancer Inst. 2017;doi:10.1093/jnci/djw327.
For more information:
Avrum Spira, MD, MSc, can be reached at Boston University School of Medicine, 72 E. Concord St., Boston, MA 02118; email: firstname.lastname@example.org.
Disclosure: Spira reports financial fees from Allegro Diagnostics Inc. and Veracyte Inc.; grants from the Department of Defense, NCI and NIH; and patents licensed to Allegro Diagnostics Inc. Grants from the NIH Early Detection Research Network supported this study.