STANFORD, Calif. -- Lung cancer is one of the deadliest and most common types of cancer, and accurately determining its stage -- how far it has progressed in a patient -- is crucial in determining the most appropriate treatment. The most commonly used tool for staging lung cancer is a CT scan, often followed by a biopsy.

A study by researchers at the VA Palo Alto Health Care System and Stanford University's Center for Health Policy/Center for Primary Care and Outcomes Research, however, finds that an emerging technology known as PET-FDG is considerably more accurate than CT in indicating whether a patient's cancer has spread outside the lungs, a key aspect of staging.

By using PET, the results suggest, physicians could avoid missed opportunities to operate on patients whose cancer is still at an early stage -- patients whose lives could be prolonged for five or more years with surgery. PET could also help physicians avoid the opposite mistake: unnecessarily operating on patients whose cancer is too advanced for surgery. The study focused on non-small-cell lung cancer, which accounts for up to 80 percent of all lung-cancer cases.

"The use of PET can improve doctors' ability to stage non-small-cell lung cancer, and that has important implications for treatment planning," said Michael Gould, MD, assistant professor of pulmonary and critical care medicine at the VA Palo Alto Health Care System and lead author of the paper, to be published in the Dec. 2 issue of the Annals of Internal Medicine.

The Stanford/VA study -- a comprehensive analysis of all previous research on the subject -- is also the first to show that the accuracy of PET in lung cancer staging is influenced by CT results, and that the two tests must be interpreted together to yield the most accurate diagnostic information. A series of graphs produced by the researchers can help physicians interpret their CT and PET results to make the best treatment decisions, explained Gould, a fellow at Stanford's Center for Health Policy.

Lung cancer is one of the most common cancers in the United States, accounting for about 15 percent of all cancer cases, or 170,000 new cases annually. It has one of the highest death rates of all cancers; 85 to 90 percent of those with the disease die within five years.

Computed tomography or CT, a non-invasive imaging test that yields detailed cross-sectional images of the body, has long been used for staging lung cancer. Because CT often yields false-positive results, however, many patients require a biopsy to confirm positive CT findings.

In the past five to 10 years, PET-FDG, or positron emission tomography with 18-fluorodeoxyglucose, has emerged as an alternative. PET-FDG detects cancerous tumors on the basis of their increased metabolic rate. Cancer cells metabolize (or use) glucose at a much higher rate than normal tissues.

To compare the accuracy of PET versus CT in detecting whether non-small-cell lung cancer has spread to the lymph nodes in the chest, Gould's research team systematically analyzed and synthesized all previous studies on the subject that met their criteria for scientific rigor (39 studies). For each study, researchers plotted the sensitivity and specificity of PET and CT, two key measures of diagnostic accuracy. Sensitivity refers to the likelihood that a diagnostic test will accurately detect the presence of disease (in this case, the presence of cancer in the lymph nodes); specificity refers to the likelihood that the test will correctly identify when no disease is present.

The researchers' analyses showed that PET was superior to CT in sensitivity and specificity. The average sensitivity and specificity of PET were 85 percent and 90 percent, respectively, compared with 61 percent and 79 percent for CT.

The researchers also found that the accuracy of PET depends on whether the CT scan shows that lymph nodes are enlarged. Specifically, the researchers found that when CT reveals enlarged lymph nodes, PET is better at identifying the presence of cancer; when CT shows that lymph nodes are not enlarged, PET is better at excluding the presence of cancer. "Doctors need to consider that the performance of PET differs depending on the CT findings," Gould explained. "If they don't, they'll make errors in staging the disease."

Gould cautioned that despite PET's strengths, the test is not perfect, and positive findings on PET should be confirmed with a biopsy before surgery is ruled out. Furthermore, he said, PET should not necessarily be used for all lung cancer patients, particularly given its high cost (about $2,000 per scan). Gould's ongoing research is exploring exactly when PET should be used and on which patients.

The study was funded by the U.S. Department of Veterans Affairs. Collaborators include Douglas K. Owens, MD, an investigator at the VA Palo Alto Health Care System and professor of medicine at Stanford; Ware Kuschner, MD, an investigator at the VA Palo Alto and assistant professor of medicine at Stanford; and CHP/PCOR research associate Jo Kay Chan.