Clinicians need to learn more about the benefits of lung cancer screening with CT, biopsy their patients less, and not worry too much about the survival prospects of patients with minimally invasive adenocarcinomas, according to three studies presented on Thursday morning at the Multidisciplinary Symposium in Thoracic Oncology in Chicago.
One study found that biopsy costs are by far the most expensive part of lung cancer diagnosis workup, and clinicians need to find better ways to minimize their use, while a second one found extremely high five-year survival rates in minimally invasive adenocarcinoma -- statistically the same as adenocarcinoma in situ.
A third study surveyed primary care doctors and found astonishingly low levels of awareness for low-dose CT lung cancer screening, along with realistic fears that the costs of CT screening were a significant barrier at one institution.
"Most providers do not know the current guidelines for lung cancer screening, and less than half of providers perceived low-dose CT as effective in reducing cancer-specific mortality," said Dr. Jennifer Lewis, lead study author and assistant chief of medicine at Wake Forest Baptist Medical Center. "Twelve thousand lives could be saved each year in the U.S. if low-dose screening were fully implemented," she said. However, "before this can happen, our providers need education."
Costly negative biopsies
Researchers from data analysis firm Xcenda in Palm Harbor, FL, looked at all of the costs of workup following a typical lung cancer screening exam that showed nodules suspicious for lung cancer, using the Medicare claims database from 2009 through 2011. They found that biopsy costs dwarfed all other costs associated with a lung cancer diagnosis, and recommended the development of better tools to reduce the number of negative biopsies.
The retrospective study used a random 5% sample of Medicare beneficiaries from 65 to 74 years old (mean age, 69.3 years) who had a scan showing swelling, a mass, or a nodule on the lung following lung cancer screening. Diagnostic tests used to diagnose the disease in the sample included x-ray (54.4%, n = 4,885), CT (32.9%, n = 2,954), and PET scans (0.4%, n = 36).
Biopsies were performed for 19.4% of all patients (n = 1,742); however, 43.7% (n = 761) were negative, and these patients were not diagnosed with lung cancer, said lead author Tasneem Lokhandwala, PhD, from Xcenda. Moreover, the costs of biopsy were high -- by far the costliest part of workup.
For patients eventually diagnosed with lung cancer, the average total cost of the workup was $7,567, versus $3,558 in patients not diagnosed with lung cancer. The median cost of a biopsy was $3,784, but the cost was about four times higher in patients who had an adverse event -- about 20% of all patients undergoing biopsy.
The average cost of a lung biopsy with complications was $37,745, compared with $8,869 for a complication-free biopsy. All incidental costs for services required for a biopsy were included in the cost, she said. On the other hand, the costs for CT and x-ray included only the specific procedure.
"The biopsies were the costliest tool in diagnostic testing ... and patients who had a biopsy incurred significantly greater costs than patients who did not have a biopsy," she said.
"The total lung cancer diagnostic costs in this population were about $38 million, of which 43% was accounted for by negative biopsies," she said. "We need to develop more precise risk stratification tools to better identify patients for biopsy -- this has the potential to reduce costs and improve patient outcomes."
Results questioned
In an email to AuntMinnie.com, Dr. Ella Kazerooni, from the University of Michigan and the American College of Radiology, said the costs reported were not part of a lung cancer screening evaluation, but rather workups "for patients who have clinical signs of lung cancer, which leads to diagnostic tests like CT and ultimately some form of tissue sampling to figure out if they do or do not have lung cancer."
"Some of these patients will have infection or inflammatory conditions that need to be diagnosed and treated too, so the fact that they undergo biopsy is entirely appropriate in many cases," continued Kazerooni, who was not involved in the study. Moreover, "761 negatives for cancer biopsy patients does not mean 761 negative biopsy results, as some have specific noncancer diagnoses made, and this is not discussed at all."
Finally, the total cost of $16.5 million for 761 negative workups doesn't add up if the average cost of workup in patients without lung cancer was $3,558 apiece, as reported, she said.
"It would be much more useful if they reported what percentage of patients did and did not undergo a workup in accordance with the [National Comprehensive Cancer Network (NCCN)] guidelines, which they allude to but do not state" in the abstract, Kazerooni said.
No survival penalty for minimally invasive disease
In another study, patients with minimally invasive adenocarcinoma (MIA) survive as long as those with adenocarcinoma in situ (AIS), with five-year survival rates of 96% to 97%, reported researchers from Emory University. The results are surprising because MIAs are considered to be more challenging treatment targets than adenocarcinoma in situ, and some have advocated for separate categorizations.
Adenocarcinomas account for 60% of all lung cancers. Bronchoalveolar carcinomas, which affect cells lining the alveoli, are considered invasive when tumors extend beyond the alveolar lining. Tumors 3 cm or smaller in size are reclassified as adenocarcinoma in situ, no invasion or minimally invasive adenocarcinoma if the invasion is less than 0.5 cm, according to 2011 guidelines from the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society.
"We conducted this study, using data from the published literature, to understand if there are really any differences between the two categories as far as patient outcomes," said Dr. Madhusmita Behera, PhD, associate director of research in hematology and medical oncology at Emory.
The pooled analysis of 863 patients (mean age, 67.5 years) encompassed 18 studies with lung cancers categorized as AIS or MIA. Only 43% of the subjects were smokers, and most had undergone resection.
The five-year disease-free survival rate for the two groups pooled together was 97.7%.
"When comparing the two groups, no differences were found between the five-year disease-free survival rates between AIS and MIA," Behera said. Five-year survival for AIS was 97%, versus 96.7% for MIA. The five-year overall survival rate was also equivalent between the two groups.
"We found that patients with these two tumor types have excellent five-year survival outcomes of nearly 100%," she said in her presentation. "Our findings raise questions about the necessity of classifying these groups into two separate entities."
Primary care docs unaware of screening benefits
In the last study, patients at high risk of lung cancer are more likely to get low-dose CT screening when their primary care providers are aware of the current guidelines recommending screening, according to a physician survey from Wake Forest Baptist Medical Center.
In the wake of the National Lung Screening Trial (NLST) showing a 20% morality benefit in high-risk smokers, several organizations have recommended the CT test, including the American Lung Association, the American College of Chest Physicians, the National Comprehensive Cancer Network, and the U.S. Preventive Services Task Force, which published guidelines for screening.
But after the screening guidelines were published, "we noticed that very few if any providers at our own institution were providing screening; therefore, we became interested in determining the lung cancer screening practices at our own institution, as well as assessing providers' knowledge of the guidelines and attitudes about whether they feel this test is effective at saving lives from lung cancer," said Dr. Jennifer Lewis, assistant chief of medicine at Wake Forest.
The online survey was sent to 488 primary care providers at the institution, including physicians, physician assistants, and nurse practitioners, of whom 60% (n = 293) responded.
Twenty-one percent of respondents had ordered chest radiography, "a nonrecommended lung cancer screening test," Lewis said, while just 12% had ordered low-dose CT. An additional 3% ordered sputum cytology.
Less than half (48%, n = 102) of the respondents knew three or more of the six guideline components, including whom to screen, the recommended screening intervals, and how many screening rounds should occur. In addition, 24% of the respondents didn't know any of the guidelines.
Providers who knew three or more of the guidelines were likely to order low-dose CT screening for individuals at high risk for lung cancer, with an odds ratio of 5.0.
Only 42% of respondents said that CT was either moderately or very effective at saving lives, and many providers were unaware of the effectiveness of CT, Lewis added.
"But if you look at the numbers of patients needing to be screened to avoid a single death, lung cancer screening is more effective than mammography or even flexible sigmoidoscopy," she said.
Providers cited financial costs to patients as the single greatest barrier to CT screening. Other potential barriers included false positives, lack of patient awareness of screening, incidental findings, and lack of awareness of coverage.
More than 12,000 lives could be saved every year by lung cancer screening, but more education is needed before it can come to pass, she said. Since the survey about a year ago, the facility has published a brochure on CT lung cancer screening and distributed it to all providers.
Kazerooni told AuntMinnie.com that the results of this study confirm the need for more education.
"The fact that lung cancer CT screening is more likely to appropriately occur when primary care physicians are familiar with the guidelines reinforces that we need to continue our efforts to educate the primary care community, so that patients can benefit from this lifesaving test," she said.
