Blood Test Distinguishes Early Lung Cancer From Benign Nodules

Jenni Laidman

October 18, 2013

A blood test for early-stage lung cancer that can distinguish malignant from benign lung nodules with 90% negative predictive value could reduce the number of unnecessary lung biopsies dramatically, according to a study published online October 16 in Science Translational Medicine.

However, "I don't think it's a game changer," pulmonologist Norman H. Edelman, MD, chief medical officer at the American Lung Association, who was not involved in the study, told Medscape Medical News.

The screening tool tests for levels of 13 proteins that differentiate benign from malignant lung nodules identified on CT scans. Xiao-jun Li, director of Bioinformatics for Integrated Diagnostics (known as Indi) in Seattle, and colleagues developed the 13-protein panel by first testing 371 candidate proteins with a technique called multiple reaction monitoring mass spectroscopy.

An assay was created for each protein and tested on plasma samples from 143 patients with benign or stage IA lung cancer.

Those assays yielded the 13-protein test, which was then validated on 104 independent plasma samples. The effectiveness of the test was independent of nodule size, patient smoking history, or patient age, the researchers report.

The 13-protein panel identified benign lesions with 94% accuracy. Pathway analysis showed that the proteins are probably modulated by transcription factors associated with lung cancer, lung inflammation, and oxidative stress.

Such a test could go some way toward reducing unnecessary biopsies. "The classifier provided insightful assessment on the disease status of lung nodules beyond the clinical risk factors currently used by clinicians," the researchers write.

Only 20% of patients biopsied for lung nodules have a malignancy. "By measuring protein abundance in blood samples of patients, the classifier can be used to prevent patients with benign lung nodules from undergoing unnecessary invasive procedures," they explain.

"We believe this technology, when applied to a commercial protein expression test, will be of tremendous interest to pulmonologists," said Albert A. Luderer, PhD, chief executive officer at Indi, the test developer. "Clinicians encounter patients with lung nodules in their practices on a regular basis, which often leads to invasive procedures. We expect that a noninvasive test based on Indi's technology might be able to assist physicians in determining which nodules do not require risky and costly clinical interventions. Eliminating those invasive procedures in patients with benign nodules should lead to significant savings in the healthcare system."

However, Dr. Edelman said that test results stating that a given nodule is benign would not induce him to change his practice pattern. If I had a patient with a nodule that had a 90% chance of being benign, "I'm not going to send the patient home and say don't worry. I'm going to follow it with periodic CT scans."

Further, he said, because of the test's high negative predictive value, there is a natural corresponding tradeoff in specificity. Test specificity of about 50% "means there are whole lot of false positives.... If there is a 90% chance they're right that the tumor is benign, that means the test is going to say 'yes' to a lot of tumors that are not cancer," he explained. Such a scenario could mean not a reduction in biopsies, but a potential increase, he argued.

This diagnostic test provides objective quantitative information, explained Ken Fang, MD, chief medical officer for Indi. "Right now, even after considering age, smoking history, and nodule size, this test will tell you which nodules are benign so they can be followed with CT scanning. If you're on the fence, this provides objective information. It's also quantitative. We'll be providing the probability that the nodule is benign," he told Medscape Medical News.

In the validation study of 104 patient samples, at an assumed cancer prevalence of 15%, test sensitivity was 71%, specificity was 44%, and negative predictive value was 90%.

Another expert approached for comment, Lance A. Liotta, MD, PhD, from George Mason University in Manassas, Virginia, and codirector of its Center for Applied Proteomics and Molecular Medicine, who was not involved with the research, said this blood test is an important advance in the development of biomarker diagnostic testing.

"This study fulfills 2 prophecies that have been discussed in the cancer biomarker research community for many years," Dr. Liotta said. First, it demonstrates that marker combinations are superior to a single molecule biomarker for cancer. "Panels make more sense, because cancer is a heterogeneous disease and diagnostic biomarkers will shed from both the tumor cells and the host microenvironment," he said.

The second prophecy, he said, is that "blood biomarkers for cancer will not be standalone screening tests. Instead, they will be used in conjunction with other diagnostic modalities, such as imaging, to increase the specificity of both the biomarker panel and the imaging diagnosis."

"This is just the beginning," said study coauthor Leroy Hood, MD, PhD, in a press release. Dr. Hood is cofounder of the Institute for Systems Biology in Seattle, which collaborated with Indi on the study. "The principles used to develop this classifier should be applicable to a range of unmet diagnostic medical needs," he said.

Sci Transl Med. 2013;5:207ra142. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.