A small CT lung cancer screening study of U.S. veterans found a higher rate of disease compared to the landmark National Lung Screening Trial (NLST), according to results presented at this week's American Thoracic Society (ATS) meeting in Philadelphia. The higher prevalence could fuel the debate surrounding the criteria for patient selection.
The study, conducted at the West Roxbury Division of the Veterans Affairs (VA) Boston Healthcare System, suggests that CT screening could be effective in finding cancer in high-risk individuals with fewer than 30 pack-years of smoking history, the threshold for patients in NLST.
The researchers found a 6% prevalence of the disease in individuals with at least 20 pack-years of smoking history and at least one additional risk factor, per the 2012 National Comprehensive Cancer Network (NCCN) lung cancer screening guidelines. In comparison, NLST reported a 3.8% prevalence of lung cancer among more than 50,000 individuals with a history of 30 or more pack-years.
The landmark NLST results, published in 2011, showed that in a population of heavy smokers ages 55 to 74 years, those screened with CT had a 20% drop in lung cancer mortality compared to those screened with chest x-ray. Since then, debate has centered on whether screening should include a broader range of at-risk individuals. The stakes are particularly high as the U.S. government, concerned with questions of cost and benefit, considers when, how, and if to implement widespread screening of current and former smokers.
Shorter smoking histories
In the present study, one-fifth of the 84 individuals had shorter smoking histories compared to NLST, and most of the cohort had multiple risk factors that the larger trial didn't encounter in its population, said Dr. Ronald Goldstein, chief of pulmonary medicine at the West Roxbury Division.
"We were shocked to see all these cancers that were popping up in a relatively small group of people," Goldstein said in an interview with AuntMinnie.com. "These are veterans who are really at super-high risk because they're coming to pulmonary clinics, so they often have underlying pulmonary diseases to begin with. Some of them got exposed to Agent Orange in Vietnam, for example, and a lot of them had low-level exposure to asbestos, so this is a different group of people we're looking at than a primary clinic."
Adding to the mix, researchers in the ongoing VA study are now also receiving patients from the primary care clinic who are not as ill, Goldstein added.
All told, the 84 CT screening subjects (74 from the pulmonary clinic, 10 from primary care providers) had a median age of 65 years, with a smoking history of 30 pack-years (80%) or 20 pack-years (20%). They also had one additional risk factor, such as chronic obstructive pulmonary disease, a family history of lung cancer, or work exposure to asbestos or other carcinogens.
Seven patients had a history of cancer, including one with a history of lung cancer. A total of 54% of the VA subjects had a nodule 4 mm or larger at baseline screening (versus 27% in NLST). Eleven percent of the subjects had an additional minor abnormality (versus 7.5% in NLST).
Among the 45 patients with a nodule 4 mm or larger, seven were deemed suspicious for malignancy. Positive CT cases included the following:
- 4 patients with pathologically proven lung cancer (1 stage 1A adenocarcinoma, 1 stage 1B adenocarcinoma, 1 stage 3A adenocarcinoma, and 1 stage 3B squamous cell carcinoma with neuroendocrine differentiation)
- 1 patient with mild PET uptake and bronchial obstruction, but bronchoscopy was negative
- 1 patient with mild PET uptake who declined biopsy; unable to determine malignancy status
- 1 patient with highly suspicious PET who refused biopsy; presumed stage 3A or M1 lung cancer
Regarding the last case above, the patient had a very large nodule and mediastinal involvement at CT, said co-investigator Sue Yoon, a nurse practitioner at the center.
"By the imaging study, he probably had stage 3 or even stage 4 cancer because he had extensive disease, but he made a decision not to pursue any biopsy," she said.
Guidelines and gray areas
Like the screening criteria in NLST and subsequent recommendations by the American College of Chest Physicians and American Society of Clinical Oncology, the 2012 NCCN guidelines call for CT screening of individuals ages 55 to 74 with 30 pack-years or more of smoking history. If the subjects have quit smoking, then it must have occurred within the previous 15 years.
Unlike these guidelines, however, the NCCN recommendations also include a weaker (category 2B) recommendation for screening individuals age 50 or older with a smoking history of 20 or more pack-years and one additional risk factor not including secondhand smoke.
As for screening patients with fewer than 30 pack-years, "I'm in favor of doing that with our high-risk patients because we see so much cancer," Goldstein said. One patient with pathology-proven lung cancer fell into the 20 pack-year group. But baseline screening criteria are just the tip of the iceberg.
"Following all these patients is a challenge, and we have to follow them for a minimum of three years," he said.
But even that is questionable considering that the NCCN guidelines are ambiguous on the issue of screening limits, calling for annual screening "for three years and until the age of 74," Goldstein noted. "We don't really know if they mean 20 scans for someone starting at the age of 55. It doesn't make sense to stop at three years if you have a young person who is still smoking. On the other hand, do they really mean you have to scan someone for 20 years? It may mean you should scan every other year; it's just ambiguous to me what you should do."
Stopping at age 74 is also problematic in the older VA population, added Yoon. "We already have patients about that age," one of whom "we just diagnosed" with lung cancer, she said.
It's going to take a lot of time and effort to determine the parameters, and following the Fleischner Society guidelines for follow-up of suspicious nodules is also a tricky business, Goldstein said.
"If you have a malignancy, and it's 4 mm and you're going to wait a year -- that might not be so good," he said.
Screening won't be cheap either, he added. "But we're hoping, when the primary care people start doing it, to have some sort of electronic system to ensure that they can follow-up their patients, and refer to us only patients that need urgent evaluation within the next three months."
"The nodules that look really threatening, those are the ones we want to see soon," Goldstein said. "But the critical thing for us is what size nodule needs urgent follow-up, because we see so many. Is it 6 mm? Is it 8 mm? There's that gray zone that is problematic."
