Searching for Lung Cancer
With no standard screening test, doctors and patients await research findings.

By Elizabeth Whittington

Every year near his birthday, Lewis Pomper gets a physical. His doctor checks his blood pressure, heart rate and weight. He does the appropriate blood work and gets a chest X-ray. In April of 1999, everything looked good.

Pomper, a 65-year-old from West Palm Beach, Florida, smoked before he could vote but quit in 1998 after 45 years because of increasing publicity about health risks associated with smoking. So when he learned Weill Medical College of Cornell University was conducting a lung cancer screening program, he signed up.

Three months after his chest X-ray came back clean, Pomper was screened for lung cancer with a low-radiation computed tomography (CT) scan, which revealed lung cancer. Fortunately, the cancer was stage 1A and Pomper had the tumor removed—no chemotherapy, no radiation.

“It surprised me. I felt that the doctor or radiologist who read my X-ray blew it,” Pomper says. “I got the X-rays from my doctor, brought them to Cornell and asked them to show me where the cancer was.”

The 2-centimeter tumor wasn’t on the chest X-ray, though. With the tumor tucked behind a rib, the radiologist could not see it. “They told me it would have been nine to 18 months before that cancer would have shown up on an X-ray,” Pomper says. “By that time, it would have been too late.”

Early detection appears to be the best way to beat lung cancer, with five-year survival rates of stage 1 cancers reaching 70 percent or higher. But with symptoms appearing only at advanced stages and the debate continuing over whether people should even be screened for lung cancer, only 15 percent of diagnoses come early. Experts say the findings from ongoing studies, such as the one involving Pomper, could revolutionize lung cancer screening.

Lung Cancer Screening: Past and Future

The link between smoking and lung cancer, discovered more than 50 years ago, led to the first screening test for lung cancer. Called sputum (saliva) cytology, the test examines mucus coughed up from the lungs for abnormal cells. The test wasn’t successful in identifying early cancers, so researchers have since developed several other options, including imaging scans, bronchoscopy, advanced sputum cytology and breath analysis.

CT imaging has generated the most interest. First introduced in the 1970s, a CT scan took several hours because each X-ray required a few minutes. In the 1990s, the development of spiral CTs meant a series of X-rays could create a three-dimensional image of the chest in only a few seconds. CTs have gone from four-slice, taking four images in one rotation, in 1998 to a 64-slice scan just in the past year. The 64-slice scan takes a higher resolution image in less time.

With advances in CT technology over the past decade, the average detectable size of a lung tumor has decreased considerably, often to less than 10 millimeters. In addition, a recent study reported that combining spiral CT with PET (positron emission tomography) improved identification of lung cancer. However, suspicious areas on imaging scans may not be lung cancer, so doctors recommend biopsy or surgery to determine if lesions are cancerous.

Bronchoscopy uses a small camera inserted through the mouth and into the lungs to look for tumors in the main lung pathways. A modified type of bronchoscopy called autofluorescence bronchoscopy employs a laser scope to examine the main lung branches but like standard bronchoscopy that uses light, it can miss precancerous lesions in the smaller airways. However, a new technique that allows doctors to travel deep into the lungs received Food and Drug Administration approval late last year. The SuperDimension Bronchus system uses three-dimensional images to see farther inside the lungs without invasive surgery.

Newer and better sputum screening tests have emerged, including LungAlert™, a method currently in clinical trials that detects a cancer-associated sugar in phlegm. Also in development are screening tests that examine exhaled breath for chemicals found in lung tumors. But CT imaging is gaining the most excitement.

“This is a field that is changing at the speed of light,” says James Mulshine, MD, former head of the intervention section in the Cell and Cancer Biology Branch at the National Cancer Institute. The average lung cancer now found at follow-up in the screened population at Cornell is 8 mm, but Dr. Mulshine says no standard teaching practice exists for lung tumors that small. And with newer technology comes the possibility that tumors could be detected at even earlier stages.

High Survival for Early-Stage Lung Cancer

Screening programs, such as the International Early Lung Cancer Action Program (I-ELCAP) at Weill Cornell and institutions around the world, find 80 percent or more lung cancers at an early stage.

Claudia Henschke, MD, PhD, professor of radiology at Weill Cornell, has been conducting its lung cancer screening trial since 1993. She says the trial, which has produced the longest-term information in the field to date, shows lung cancers detected early enough and without lymph node involvement have a greater than 90 percent long-term survival rate. Most lung cancers detected at advanced stages have a long-term survival rate of 10 to 15 percent. Results of the ongoing trial will be available in 2007.

In an earlier study, the Weill Cornell group screened 1,000 participants with an initial low-dose CT scan. Of the 27 cancers found, 83 percent were stage 1. Annual follow-up of all participants found additional cancers, with 85 percent at stage 1. Generally, 85 percent of lung cancers are diagnosed at stage 3 or 4.

Another trial, the National Lung Screening Trial (NLST) conducted by the National Cancer Institute, will show if screening with spiral CT compared with chest X-ray reduces lung cancer deaths. The trial, which began in 2002, will be completed in 2009.

Who Should Be Screened

Genetics affects lung cancer risk as do environmental carcinogen exposure and smoking history, including how long a person smoked, how much, the type of tobacco and if the person currently smokes. What constitutes a high-risk person remains debatable, but typically, smokers and former smokers who quit less than 10 to 20 years ago, smoked a pack a day for 10 years and are 50 or older are at high risk for lung cancer.

Identifying a protein or genetic mutation indicative of lung cancer provides a more scientific method of who could benefit from screening. Sputum cytology, blood or exhaled breath tests could identify genetic mutations frequently found in lung cancers. About half of all non—small-cell lung cancers and 75 percent of small-cell lung cancers express the p53 gene, but currently no known biomarker exists in all lung tumors. Many institutions involved in the NLST study are collecting blood, saliva and urine samples to search for a common biomarker.

Jury Still Out

A useful screening test decreases mortality, increases survival by finding cancers early, has a low rate of false-positives and is safe. Based on these criteria, lung cancer screening options have yet to prove useful.

No screening study performed to date has shown a reduction in lung cancer deaths. Three large lung cancer screening trials in the 1970s found that despite catching many lung cancers early, mortality did not decrease. Because of unimpressive study results, professional organizations, such as the American Cancer Society and the American Lung Association, say they cannot endorse lung cancer screening until data show a reduction in mortality rates.

Norman Edelman, MD, executive vice president and chief medical officer of the American Lung Association, says spiral CT shows potential, but studies proving its effectiveness are needed before the organization will endorse lung cancer screening. “It’s promising and we hope that eventually, it will reach the point where absolutely, high-risk patients should be screened by spiral CT,” Dr. Edelman says. “We’re pleased that a very large federal study is taking place to determine that.”

Opponents of lung cancer screening say early-stage patients may experience needless worry and invasive surgery if they have tumors that grow so slowly that death results from another cause later in life. Dr. Henschke says I-ELCAP works around this problem by checking the growth rate of each tumor one to three months after diagnosis. If the tumor is slow growing, they wait and watch. If the tumor grows at a high rate, doctors surgically remove it.

False positives mark another negative for lung cancer screening. Particularly with CT scans, the difficulty becomes distinguishing a potential tumor from scar tissue, a lung disease such as emphysema or another non-cancerous cause. A biopsy or surgery can confirm cancer, but these procedures carry the risk of a punctured lung or infection. “If you have to do nine operations to take one cancer out, is that going to be better or worse in the long run?” asks Dr. Edelman. “If you start operating on a lot of benign lesions, is it going to do more harm than good? That’s why people are cautious, because a definitive answer isn’t in yet.”

Because organizations and most doctors don’t currently endorse lung cancer screening, insurance doesn’t often cover screening tests. But that’s not keeping some high-risk patients from footing the bill. Dr. Henschke advises individuals to find a facility adept in screenings and follow-up, including facilities participating in the I-ELCAP or NLST trials. With 1 to 3 percent of lung cancer survivors developing new lung cancers each year, annual follow-up is critical.

While the medical community waits for the evidence, Pomper tells everyone he can about his lung cancer screening test. “I owe my life to that test.”