James L. Mulshine, MD
Head
Intervention Section
Cell and Cancer Biology Branch
Co-chair
Upper Aerodigestive Chemoprevention Faculty
Center for Cancer Research
National Cancer Institute
National Institutes of Health
Bethesda, Maryland
Elizabeth E. Warner, MD
Department of Surgery
Georgetown University Hospital
Washington, DC
Visiting Researcher
Intervention Section
Cell and Cancer Biology Branch
Upper Aerodigestive Chemoprevention Faculty
Center for Cancer Research
National Cancer Institute
National Institutes of Health
Bethesda, Maryland
Last year, there were approximately 171,900 new cases of lung cancer and 157,200 deaths due to this disease in the US alone. The year 2003 also had the unfortunate distinction of being the first year that the number of lung cancer deaths outstripped the number of coronary artery disease-related deaths directly attributable to tobacco use in this country. There are over 90 million current and former smokers in the US at risk for developing lung cancer. Should they develop lung cancer, they face a 5-year survival rate of 15%. Clearly, lung cancer remains a public health crisis in this country. While it is important to continue efforts aimed at smoking prevention and cessation, focusing energy on these goals alone neglects the 50 million people who have already quit smoking but remain at an elevated risk for developing lung cancer. Meanwhile, lung cancer deaths among former and never-smokers in this country continue to outnumber cancer deaths from breast and colon cancer combined. Developing new methods of identifying lung cancer at an early stage could lead to a reduction in cancer deaths.
Lung cancer's lethality is due in large part to the fact that, at the time symptoms become apparent, the disease has usually spread to other parts of the body and, therefore, is not curable. Efforts have been made in the past to detect lung cancer at an earlier stage based on the belief that earlier diagnosis might lead to improved cure rates. Studies, such as the Mayo Lung Project in the early 1980s, examined the use of serial chest x-ray (repeated
s over time) as a potential screening tool for the early detection of lung cancer. However, it was concluded that, while some cancers were indeed found in advance of symptoms, only 30% were still stage I at the time of detection, and an overall lung cancer-related mortality benefit was not shown.
Rapidly evolving improvements in imaging technology, however, have raised this issue yet again. CT scanning has evolved remarkably over the last decade due to faster image acquisition capabilities coupled with massive improvements in microprocessing power. This includes the use of spiral CT scanning for the detection of lung cancer. Spiral CT provides detailed images of areas inside the body. The images are created by a computer linked to a machine that scans the body in a spiral path.
Currently available spiral CT scanners are standard instruments in most hospitals and clinics in the developed world. Rapid advances in resolution and reconstruction technologies have made it possible to frequently detect lung nodules in the subcentimeter range, which greatly exceeds the capabilities of conventional chest x-ray. The evolving field of computer-assisted diagnosis (CAD), already widely used in assisting with mammography reading and the evaluation of Pap tests for cervical cancer screening, is showing promise as a means to enhance the detection of small, suspicious nodules within the lungs while easing the potentially crushing burden that a huge number of new screening scans would present to the radiology community.
Some members of the medical community are concerned that detection of smaller primary lung cancers will not ultimately lead to a mortality benefit. However, studies have shown that some other types of cancer behave in a relatively predictable way---they begin as small, localized lesions with the likelihood of spreading to other areas of the body increasing over time and with the size of the original lesion. Numerous studies have confirmed this same pattern for lung cancer. There is growing evidence indicating that more favorable outcomes are associated with detecting smaller primary cancers. Meanwhile, groups such as the Early Lung Cancer Action Project (ELCAP) at the Weill Cornell Medical Center in New York and the Anti-Lung Cancer Association in Japan have demonstrated the ability to detect stage Ia lung cancer with spiral CT in better than 70% of patients with lung cancer.
While several groups have demonstrated the ability of spiral CT to detect early stage lung cancer, there is no evidence yet to show that this improvement results in an improved outcome in terms of reducing deaths due to lung cancer. However, the National Cancer Institute (NCI) has launched the National Lung Screening Trial (NLST), a major randomized trial of conventional chest x-ray versus spiral CT scan designed to evaluate their impact on lung cancer-related mortality (NLST-
http://www.cancer.gov/nlst 
). This trial has recently completed accrual of 50,000 patients. Each patient receives either a chest X-ray or a spiral CT once a year for three years and then is followed for clinical outcome. Trial results will not be available until definitive follow-up information is available. The Dutch also have recently launched a large national randomized screening trial.
In the meantime, a great deal of research effort is required to optimize the downstream management of lung cancer to decrease morbidity as well as cost.
The surgical management of lung cancer, for example, typically involves a fairly extensive operation--thoracotomy incision with an anatomic resection in the form of a lobectomy (removal of part of a lung) along with mediastinal dissection. Investigators in the past concluded that doing a less extensive surgery for stage la disease exposed patients to an unacceptably high risk of recurrence, but this issue may warrant reevaluation, in part, because the average size of CT-detected lung cancer is so much smaller than with typical chest x-ray-detected lung cancer. There is also a significant risk of patients eventually developing second or third primary lung cancers which mandates greater efforts at lung parenchyma preservation.
Ongoing technological advances in the fields of diagnostic imaging and CAD diagnosis have provided more sensitive tools to detect early lung cancer. While lung cancer remains the most lethal cancer in the world, the hope for major improvement in this disease has led to new definitive lung cancer screening trials. Complementary research into strategies for downstream management of CT-detected early lung cancer may also allow for improved screening outcomes.
