In Review - News from the 2004 Meeting of the RSNA

Lung screening researcher claims survival benefit

Henschke touts benefits of early detection while observers await published results

By: Jane Lowers

CT screening of high-risk patients can identify lung cancers early. It can also keep patients alive longer than if they did not undergo screening. That's the claim made by the investigators in a long-running and controversial study; but not everyone is embracing this conclusion.

For the first time, Dr. Claudia Henschke, chief of chest imaging and healthcare policy and technology assessment at New York Weill Cornell Medical Center, made a correlation between early detection, the resulting treatment, and long-term survival for lung cancer. Data from the International Early Lung Cancer Action Project (I-ELCAP) indicate that as many as 95% to 98% of lung cancers caught at stage IA and treated can be cured. Screening, therefore, could give patients years of added survival beyond what would be expected from historical data on patients whose lung cancer is found through normal means, often at later stages.

Equally promising, the percentages of lung nodules that need to be treated after baseline screening (15%) and each subsequent year of screening (6%) are low enough that screening need not unduly burden the healthcare system. Critics voiced fears of such a consequence during the study's early years.

Patients can be triaged based on their existing risk factors, such as age and smoking history, to determine if yearly screening is appropriate.

"We've always evaluated diagnosis and treatment separately in this study, and we're just now getting the treatment results," said investigator Dr. Daniel Yankelevitz, director of inpatient radiology at New York Weill. "We're getting to be able to stratify whom to screen and, after one screening, to determine the benefit of a second screen."

The I-ELCAP has collected lung screening information for 27,701 patients in the U.S. and at international collaboration sites. Demographic data for the population have enabled the investigators to begin categorizing and predicting patient risk. When a stage I cancer is found at baseline, patients have an average 76% chance of being cured if standard treatment is initiated, Henschke said.

The number rises to 78% for stage IA cancers found on subsequent yearly screenings. Without screening, only 5% to 10% of patients are likely to be cured. More than 80% of cancers found on screening were stage I. Among ELCAP patients who had lung cancers resected, the eight-year lung cancer fatality rate was 4%.

"These are really astoundingly good data," she said.

Henschke presented a calculation of age, smoking history, cessation of smoking, and other factors that can help determine whether a patient is an appropriate candidate for such screening. Comorbidities and overall expected survival for at least 10 years can affect whether the patient will receive the full benefit of the screening.

Such claims could bolster the case for ELCAP, which has been scrutinized since its inception because it is an observational, rather than randomized, trial. Other researchers are still waiting for published data so they can review the claims.

"If you have 10 years of data, why not publish it? We don't know how they make their conclusions unless we see it in a peer-reviewed journal," said Dr. Edward Patz Jr., a professor of radiology and of pharmacology/cancer biology at Duke University Medical Center and lead author of a modeled lung screening study published in the Journal of Clinical Oncology in June. "From the abstract presented at RSNA, these aren't the data we need to know: Does it reduce mortality?"

With a growing list of centers participating in the trial worldwide, the study is adding more data annually, and it is becoming more complex. Scans 10 years ago might have been acquired with 10-mm slices, whereas today's CT scanners with thinner slices might find more potential nodules with 1-mm resolution. The ELCAP and other studies are beginning to sort out the thorny question of which lung masses need follow-up, and how fast. A subset of ELCAP patients had mediastinal nodules, of which nearly 90% stayed the same size year to year or shrank on their own. Setting up a watch-and-wait plan for appropriate masses can decrease unnecessary procedures.

For suspicious lung nodules, however, Henschke and colleagues are calling for early action.

"Deaths from stage I lung cancer were surprisingly low after surgery, but only if treatment is pursued," she said. "Delaying treatment by more than six months resulted in increased tumor disease and often a higher stage of the disease."