A Breath of Fresh Air: Lung Cancer Screening Post-NLST
The decision of the National Cancer Institute (NCI) to end the National Lung Cancer Screening Trial (NLST) promises to have far-reaching impacts. The data could spur development of a CT lung cancer screening program in the U.S. This next step requires a revised model characterized by clinical collaboration pre-and post-screening and thoughtful and systematic patient selection. Experts paint a picture of the new model.

The NCI prematurely ceased the NLST after initial results demonstrated that CT screening with at least three annual screens in a high-risk population of current and former smokers aged 55 to 74 years conferred a 20 percent drop in lung cancer mortality compared with x-ray screening. The announcement has spurred clinical conversations about screening and patient management.

“[Before the NLST study results] if a patient asked me about screening CT, I would have actively discouraged it and told him or her screening CT was not ready for prime time. Now my answer would be different,” says Greg Otterson, MD, interim co-director of the division of hematology and oncology at the Ohio State University Comprehensive Cancer Center in Columbus. He is a member of the National Comprehensive Cancer Network (NCCN) Lung Cancer Screening Guidelines Panel.

“The NLST results will have a significant impact on public health policy. Exactly who would benefit from CT screening, where that screening should be performed and the frequency and duration of screening are all questions that remain to the determined and will require more careful analysis of the considerable data collected from the NLST over the coming weeks,” offers Denise R. Aberle, MD, professor of radiology and bioengineering at University of California, Los Angeles and national principal investigator for NLST.

As researchers analyze the data by gender, race and smoking history, stakeholders are considering its implications. The American Cancer Society and the NCCN are working on screening recommendations, which are likely to be based on NLST criteria: current and former smokers aged 55 to 74 years with a smoking history of at least 30 pack years. NCCN recommendations should be available in a few months, says Otterson.

CT screening in evolution

NLST was implemented primarily at major medical institutions, with an infrastructure to address positive screens, Aberle pointed out. The trial reported an overall 24 percent rate of positive screen results with the first CT screen, which decreased with subsequent screens. The majority of positive screens were false positives.

Small nodules detected by screening CT require additional CT studies and possibly PET scans and more invasive biopsies or surgical lung resections. Recommendations for followup of indeterminate nodules, such as the Fleischner Society guidelines, already exist and provide a good starting point for protocols for assessing individuals with positive screens until analysis of new data revises best practices, Aberle says.

“Ideally, CT screening should be implemented in practice settings in which current best practice guidelines exist for the performance of low-dose CT screening and judicious followup is practiced,” recommends Aberle.

University of Michigan is evaluating the potential for a high-risk lung cancer clinic, shares Ella Kazerooni, MD, director of cardiothoracic radiology at University of Michigan in Ann Arbor, and site principal investigator for NLST. Such a program could include screening CT and allow for more systematic management of the high-risk population.

Otterson envisions the target population fitting a bell-shaped distribution curve. Some patients outside of the NSLT criteria, such as the healthy 49-year-old smoker with a 45 pack year history, probably should be screened, he says. Anxious non-smokers as well as patients who will not tolerate diagnostic followup will be more challenging. A patient with advanced COPD would not tolerate follow-up surgery or radiation; the knowledge of lung cancer may not be helpful, shares Otterson. The oncologist anticipates limits on screening for low- and high-risk patients.

Targeted screening?

A 72-year-old man presented for evaluation of progressive dyspnea and cough. He reported smoking one to two packs of cigarettes a day since age 15. Standard chest radiography showed a suspicious lesion in the right thoracic cavity. CT of his chest revealed bullous emphysema (bottom arrow), a tumor involving the middle lobe of the right lung (top arrow), and a pack of cigarettes in his shirt-pocket (asterisk). Biopsy of the lesion confirmed the presence of non-small-cell lung cancer. Source: N Engl J Med 2006; 354:397, David Michael McMullan, MD, Gordon Alan Cohen, MD, PhD, University of Washington, Seattle, WA 98195
Screening models based on age and pack year exposure may be problematic. Yearly detection rates for CT screening range from 0.5 to 1 percent, and false positive rates can stretch as high as 25 percent, which results in downstream testing and costs.

However, it may be possible to stratify high-risk candidates and target CT exams. “[T]here’s a pressing need to better target screening resources on those individuals who are most at risk from this disease. It may not be feasible to screen 50 million Americans,” explains Robert Young, MD, associate professor of medicine and molecular genetics at the University of Auckland in New Zealand.

Young recently shared the results of a prospective study of a gene-based screening tool at the American Association of Cancer Research (AACR) annual conference. Researchers followed 1,212 confirmed lung cancer cases and 1,200 controls gathered from smokers in New Zealand, Spain and the United Kingdom for four to 10 years.

Researchers administered a gene-based test that incorporates 20 markers associated with smoking-related lung damage and propensity to lung cancer along with clinical factors to derive a risk score on a 1 to 12 scale with higher scores correlating with higher risk. Current or former smokers with scores of 6 or higher, comprising 20 percent of the at-risk population, accounted for more than 50 percent of those with lung cancer. In the study, the test demonstrated specificity of 80 percent.

“At scores of 6 or more … only 25 percent of otherwise eligible smokers would be screened, but over half of lung cancers would potentially be detected, many in a treatable stage,” explains Young, who suggests that increasing the detection rate of lung cancer per number of patients screened could improve the cost-effectiveness of CT screening.

Impact on radiology

If screening CT is accepted as a clinical tool, radiology will need to adapt. “They’ll need to make sure the right patients are coming to them, which requires communication with referring providers,” explains Kazerooni.

Radiologists can expect to participate in both downstream conversations with colleagues in pulmonary medicine and oncology as well as upstream conversations with primary care physicians as the entire community grapples with the data.

“It will be up [to radiologists] to ensure standardization of followup, both for indeterminate nodules as well as other findings—like lesions of the thyroid, liver and kidney as well as cardiovascular and other abnormalities—that may be detected,” offers Aberle.

The drive for standardization and uniformity coupled with increased utilization could fuel technology innovation. “It could drive the development and potential reimbursement for chest CT CAD,” predicts Kazerooni.

Cost issues

As researchers and clinicians wrangle with the clinical implications of NLST, healthcare stakeholders also are considering economic ramifications. A major secondary aim of the trial is a cost-effectiveness analysis, shares Aberle. Researchers obtained quality of life measures for participants, detailed data on medical resource utilization for positive screens and screens that detected other findings and information about the management and outcomes of individuals with lung cancers.

“All of this information will be tremendously valuable to the NLST investigators and others as public health policy is crafted and we secure Medicare and insurance reimbursement for CT screening,” explains Aberle, who says the cost-effectiveness results should be available in 2011.

The continuing data analysis will bring new findings, but a few results are clear. “We know that the data demonstrated a benefit to 55 to 74 year old current and former smokers, and the positive finding of the 20 percent decrease in mortality cannot be overemphasized,” concludes Kazerooni.