Roxanne Nelson

April 18, 2011

April 18, 2011 (Orlando, Florida) — A panel of biomarkers appears to be able to identify the presence of lung cancer in the blood samples of people who have never smoked, according to new data presented here at the American Association for Cancer Research 102nd Annual Meeting.

"This test has good potential for use in patients after they have undergone a spiral CT [computed tomography] scan. The blood test can help patients be appropriately triaged, and help determine which patients should go on for further assessment," David Sidransky, MD, director of head and neck cancer research at Johns Hopkins Medical Institutions, in Baltimore, Maryland, told Medscape Medical News.

The 6-biomarker panel that can detect lung cancer in never smokers was developed from 4 independent smoking studies. Overall, they showed a strong performance, with a sensitivity of 85% and a specificity of 83%, in identifying lung cancer. All stages of lung cancer and histologic cell types were distinguished.

"In recent years, more attention has focused on this never-smoker population of lung cancer cases," said lead author Charlie Birse, PhD, associate director of product development at Celera Corporation, during a press briefing. "The proportion of lung cancer cases among never smokers is projected to increase in the coming years as smoking prevention and cessation programs are introduced."

A number of clinical, epidemiologic, and molecular variations suggest that lung cancers that arise in smokers and nonsmokers are significantly different. For example, women are disproportionately affected, and lung cancers in never smokers are more commonly adenocarcinomas and often bronchioloalveolar carcinomas, he pointed out.

Importance of Imaging

CT technology has greater sensitivity in detecting pulmonary lesions, compared with regular chest x-rays. Preliminary data from the National Lung Screening Test demonstrate that CT screening confers a ~20% decrease in lung cancer mortality, compared with chest x-ray. However, there continue to be concerns about the low specificity (~75%) and the associated morbidity, Dr. Birse noted.

Imaging plays a central role in the detection of lung cancer, but in addition to intentional CT scans for lung cancer, a large number of individuals undergo scans for heart disease prevention or other conditions, and incidental nodules appear in the lungs that may or may not be benign, he explained. "This panel of biomarkers would allow these imaging tests to be further evaluated and provide a degree of certainty in diagnosis."

High Sensitivity and Specificity

Dr. Birse and colleagues previously employed a novel mass-spectrometry-based approach to identify serum biomarkers that were able to detect nonsmall-cell lung cancer (NSCLC) in a smoking population representing all 4 stages of disease. In the current study, the authors applied these findings to a cohort of lung cancer patients who had never smoked.

A total of 9 biomarkers were initially assayed in serum samples collected from smokers with NSCLC and appropriate controls. For this study, more than 600 specimens were collected from 4 independent sites. The samples were randomly divided into a training set (128 NSCLC patients and 191 control subjects) and a testing set (141 NSCLC patients and 175 control subjects). These data were used to create a regression-based algorithm for lung cancer detection.

Subsequently, the authors conducted an independent validation study among lung cancer patients (n = 40) who had never smoked. All stages of cancer (8 stage  I, 4 stage II, 17 stage III, and 11 stage IV) and all major histologic cell types (21 adenocarcinoma, 7 squamous, 8 bronchioloalveolar, and 4 other) were represented in the cohort. The cohort was then matched with appropriate controls (n = 40).

The authors subsequently identified a global 6-marker regression model that identified smoking-associated cancer cases with good individual performance (training set area under a curve [AUC] = 0.877; testing set AUC = 0.868).

"We were encouraged that all 6 markers showed a strong performance in the never-smoker study," said Dr. Birse. When the model was fitted to the never-smoker cohort, the algorithm again showed the malignant cases with strong performance (AUC = 0.906).

"This translates into a sensitivity of 85% and a specificity of 83%," he said.

The performance of this test suggests that it is useful as a complementary tool to imaging for detecting NSCLC in both smokers and never smokers, Dr. Birse concluded. "We are aware that we need to confirm the performance in the never-smoker population using a larger sample size. In particular, we need to confirm performance in distinguishing malignant from benign lesions in the never-smoker population."

Diagnostics Meeting Clinical Needs

Scientists, clinicians and the lay public have come to understand the importance of detecting cancer early, said Dr. Sidransky. "Efforts to detect tumors early can really give the greatest benefit of all the research approaches and efforts that we have made in our ongoing fight against cancer."

The researchers who are developing these technologies have a better understanding of what the clinical needs are for physicians, explained Dr. Sidransky.

American Association for Cancer Research (AACR) 102nd Annual Meeting: Abstract 2813. Presented April 4, 2011.


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.