WASHINGTON-When used to diagnose early lung cancer, low-dose spiral CT scanning has a high specificity-as well as high sensitivity-and it is cost-effective, according to as-yet-unpublished data from the New York City-based Early Lung Cancer Action Program (ELCAP). Moreover, the technique can diagnose emphysema at an earlier stage than existing tests, which has helped smokers in the study to stop.
WASHINGTONWhen used to diagnose early lung cancer, low-dose spiral CT scanning has a high specificityas well as high sensitivityand it is cost-effective, according to as-yet-unpublished data from the New York City-based Early Lung Cancer Action Program (ELCAP). Moreover, the technique can diagnose emphysema at an earlier stage than existing tests, which has helped smokers in the study to stop.
We think that screening saves lives of former smokers and current smokers, said Claudia I. Henschke, MD, PhD, professor of radiology, Cornell Universitys Weill Medical College, and director of chest imaging, New York Presbyterian Hospital. It is cost-effective, and it encourages smoking cessation and should be combined with smoking cessation. That combination will be even more cost-effective.
ELCAP enrolled 1,000 men and women age 60 or older who had smoked at least a pack of cigarettes a day for 10 years or two packs a day for 5 years. Dr. Henschke and her colleagues reported findings from the first year of the study in 1999.
Their report set off what some have called call the spiral CT rush, as many radiology centers rushed to offer the test to smokers and former smokers as a way of diagnosing early lung cancer.
Dr. Henschke described some data from the studys second year at a conference on reducing lung cancer mortality sponsored by the Cancer Research Foundation of America and the Roy Castle International Center for Lung Cancer Research, Liverpool, England.
The ELCAP team initially reported that 233 of the people scanned had suspicious lesions, defined as one to six noncalcified nodules (see Figures 1-3). Of these 233 patients, 27 were diagnosed with lung cancer, and of those who were staged, 85% were found to have stage I disease.
Thats a marked reversal to the 8% that is found in the United States currently, Dr. Henschke said. The study clearly demonstrated that spiral CT can pick up lung tumors when they are small and treatment offers the greatest chance of survival.
She and her colleagues are now beginning a much larger New York City and state ELCAP study that will enroll 10,000 high-risk subjects at up to 12 centers. In addition, more than 10,000 subjects at 10 centers will be enrolled in the International Collaboration to Screen for Lung Cancer. The international study will have somewhat more flexible entry criteria than ELCAP in terms of age, she noted.
ELCAP participants were rescanned 1 year after their initial spiral CT screening. The key thing in the screening program is what happens on annual repeat, not so much on baseline, Dr. Henschke said. On annual repeat, this really becomes a terrific test.
Team members found that 1.5% of those rescreened had new noncalcified nodules, many of them being small focal infections, which were treated with antibiotics and resolved. As expected, a larger percentage of nodules found on repeat examination41%were cancerous, and 83% of those were stage I.
So now it becomes a highly specific test, about 98% specific, as well as being highly sensitive, Dr. Henschke said.
Evidence supports lung cancer screening as being cost-effective, she added, citing a study reported in the May issue of the Canadian Medical Journal. That study showed annual screening costs with low-dose spiral CT of between $5,000 and $10,000 per life-year saved.
Weve shown that baseline screening costs only about $1,000 per life-year saved because were finding many more cancers at baseline than on annual repeat, she said.
Benefits of Resection
The ELCAP team addressed the issue of whether early tumors picked up by spiral CT scanning are actually deadly. To do that, they examined data from the National Cancer Institutes Surveillance, Epidemiology, and End Results (SEER) program on the survival rates of stage I lung cancer patients who had undergone full resection, partial resection, or no resection, a category that included people who underwent radiotherapy or chemotherapy without surgery.
Theres a dramatic difference between no resection and full resection, Dr. Henschke said. These early lung cancers are ones that can kill or that do kill.
Earlier NCI-sponsored studies at Memorial Sloan-Kettering, Johns Hopkins University, and the Mayo Clinic showed that resected stage I lung cancer patients had a survival rate better than 70%, while those who refused or could not undergo surgery had less than 10% survival. Studies in Japan have also found a survival rate of 10% among stage I lung cancer patients who did not have resection.
To answer those who might complain that patients who were not resected might have been incorrectly staged, the researchers looked at SEER survival data for lung cancer patients with other stages of the disease, including those with node involvement. They found similar gaps in survival between resected and unresected patients.
This really should show everybody convincingly that theres a dramatic difference between resection and nonresec-tion, and even those small cancers kill, Dr. Henschke said.
Bringing Down Lung Cancer Mortality Rates
A projected 164,100 new lung cancers will be diagnosed in the United States this year, and 156,900 lung cancer patients will die. Although the survival rate for stage I lung cancer patients is about 70% for those whose tumors are resected, the overall survival rate for the disease is between 10% and 15%.
We have about as many lung cancers as breast and prostate cancers, and yet the annual death rates of prostate, breast, colon, and cervical cancer are less than lung cancer, Dr. Claudia Henschke said. We think that with screening, theres no reason that lung cancer cant be right down along with these other cancers in terms of cure.
She noted that currently spiral CT scanning can pick up tumors at about 1 mm or 2 mm and predicted that as the technology evolves, were going to be marching down this pathway and detecting lung cancer earlier and earlier.
Although it is too early in ELCAP to determine cure rates, the team has done some statistical projections for the relationship between tumor size and survival. Using SEER data, the researchers plotted the survival rates of lung cancer patients out 10 years.
We show that theres a statistically significant difference between the survival of individuals who started out with tumors less than 15 mm, 16 mm to 25 mm, and so on, Dr. Henschke said. And we can show by logistic regression that theres a decrease in the cure rate for every millimeter increase in tumor size.
In addition, the team looked at SEER data on a subset of lung tumors that were 15 mm or smaller, in which the large majority of cancers were between 10 and 15 mm. Even for these tiny lung cancers, theres a dramatic difference between full resection and nonresection, Dr. Henschke said.
Again using SEER data, ELCAP researchers have examined the relationship between tumor size and metastasis. We see an exponential increase of progression of disease or just the metastasisincurable diseaseby size, Dr. Henschke said. It really points out how much we should focus on this interval, looking for lung cancers below 1 cm.
The ELCAP screenings returned an unanticipated benefitevidence of very early emphysema on CT screening helps patients stop smoking.
We would go through the CT scans of the individuals, and they were all very shocked about their emphysema, Dr. Henschke said. So we would talk about it, and then they would call up several months later and say, You know, that was really very helpful. Every time I lit a cigarette, I had the image of my CT scan in front of me, and it helped me to quit.
After hearing a number of such anecdotes, the team commissioned a survey of 300 of the 1,000 ELCAP participants by researchers outside the group.
We found that among those who were still smoking at the time of enrollment, 23% quit, another 23% decreased their smoking, and the rest stayed the same. Almost all of them attributed the change to enrollment in the ELCAP program, Dr. Henschke said.
This finding has led the team to suggest combining annual screening with efforts to get smokers to stop. You can really personalize smoking with this information, and encourage smokers to quit, Dr. Henschke said.