American Lung Association recommends CT screening for lung cancer

The American Lung Association has become the largest advisory body to recommend lung cancer screening for high-risk people, advising nearly all people aged 55-74 with a 30+ pack-year smoking history (the entry criteria for the National Lung Screening Trial, or NLST) to undergo low-dose CT scanning to detect early lung cancer.

The National Comprehensive Cancer Network, whose recommendations on off-label chemotherapy use often drive insurance companies' policies on payment, endorsed low-dose CT screening last year. The U.S. Preventive Services Task Force (USPSTF), whose recommendations are considered the most authoritative (and with its advice on mammogram breast cancer screening, recently the most controversial) has not endorsed CT scanning for lung cancer screening, but may update its 2004 guidelines in the near future. The American Cancer Society also does not endorse lung cancer screening, and emphasizes that physicians must explain the uncertainties of the risks and benefits if talking with patients about getting screened with CT scans.

The ALA issued its results in an "interim report" targeted at the public, and plans a later report targeted at policymakers, hospitals, etc.

Benefits of Low-Dose CT Scanning in the NLST

Over 53,000 people were enrolled in the National Lung Screening Trial, 26,700 of whom received annual low-dose CT scans, 26,700 undergoing annual chest films, with additional scans and some invasive procedures for radiographic abnormalities. The NLST results were published in the New England Journal of Medicine in August 2011, and showed:

• Low-dose CT scanning reduced death from lung cancer by a relative 20% compared to screening with chest X-rays (247 vs. 309 deaths per 100,000 person-years).

• One lung cancer death was prevented for every 320 people screened with low-dose CT.

• 39% of patients had at least one abnormal CT scan among the 3 annual scans.

• Among those with abnormal CT scans, 96% were determined not to be lung cancer (false positives).

After multiple failures of previous lung cancer screening trials over decades, the NLST saved lives and was a success by almost any measure. However, serious questions have remained about how to bring lung cancer screening to the U.S. population in a rational, effective, and safe way.

 Uncertainties About Community-Based Lung Cancer Screening

Given the clear mortality reduction seen in the NLST, why haven't the American Cancer Society, the USPSTF, and academic societies already endorsed low-dose CT scanning for the general public? Many believe that replicating the success of the well-controlled NLST in the patchwork and profit-driven American medical care system will be problematic. The balance of benefits and risks of screening in the much larger population who don't fit the strict inclusion criteria of the NLST are unknown. Annual CT scanning of millions of adults would also carry significant costs. The American Lung Association alluded to these issues in its own report:

The Committee acknowledges that cancer screening is associated with both benefits and risks and unfortunately, the NLST could not answer a number of questions on the advantages and safety of screening in the general population. In spite of this, the Committee ... [etc., etc., and recommends lung cancer screening for patients meeting the NLST entry criteria].

Concerns voiced by the ALA in its report included:

Unknown risks and benefits among the broader population; questionable ethics of advertising.  7 million smokers fit the NLST entry criteria, but there are 94 million current and former smokers in the U.S. who may be influenced by the ALA's recommendations. It's reasonably certain that many lower-risk people will seek screening, for whom the risks of screening may well outweigh any benefits, resulting in harm. Today, anyone with $100 can get a CT scan, by responding to widespread advertising by medical centers. The ALA passes the buck on this questionable practice, saying only "a call to action should be issued to hospitals ... to establish ethical policies for advertising and promoting lung cancer CT screening services."

Community-based management of pulmonary nodules may result in higher complication rates from invasive procedures and surgeries than seen in the NLST. For example, a recent observational, database analysis of 15,865 transthoracic needle biopsies mostly performed in the community revealed a ~7% rate of pneumothorax requiring a chest tube and a ~0.5% chance of dying in the hospital. Another 8% of patients had pneumothorax not requiring a chest tube, often requiring hospital admission. Multiplied by millions of people, these figures represent a potentially large burden on the medical system and on otherwise healthy patients (this is a screening test, remember). Also, how many healthy people with harmless nodules will also be subjected to needless bronchoscopy by "aggressive" pulmonologists with a monthly overhead to cover?

Quality control. The NLST was carried out at medical centers with specialized chest radiologists working from a standardized protocol, referring patients to treatment centers with strong interdisciplinary teams and abundant experience in diagnosing and treating lung cancer. Most community centers will not likely replicate that quality, and possibly not the NLST results.

We don't know the distribution of comorbid pulmonary and cardiac disease in the NLST participants. Lung function and cardiac disease are key predictors of complications after needle biopsies and of outcome after lung cancer resection, but we don't know how these illnesses were spread among those with and without complications after biopsies from the NLST. Most people enrolling in clinical trials are healthier than average, so those in the community getting screened might be more ill and at elevated risk from procedures.

Cost. Otis Brawley, then-chief medical officer of the American Cancer Society, estimated the costs of each abnormal CT scan in the NLST at about $45,000. With a ~40% false-positive rate among millions of people, that's a lot of Benjamins. Maybe it's a good time to be a pulmonologist. (It was already a good time to be a radiologist).

Lastly, it's worth considering that although they are undoubtedly well-meaning and staffed by smart, caring people, disease advocacy groups may have a natural incentive to promote screening and broader definitions of disease and "pre-disease" in general, for two reasons:

• They have ordinary people constantly beating down their door asking questions about screening -- many of whom believe in screening, want screening, demand screening, and are frustrated that it's not being recommended. As the USPSTF found out, it's incredibly hard to say no to people with this mindset. If you're the American Lung Association, it's even harder, because you rely on those people and their families for current and future donations.

• Screening creates illness, which creates political and financial strength for disease advocacy organizations. The more people who have a condition, or a near-condition, or believe they are at risk for a condition, the greater the awareness (and sometimes fear) among the general public. The greater the public awareness, the more the attention and credibility that accrues to the advocacy organization, translating directly into political clout (ability to mobilize more people for political action) and to financial health (more donations).

These processes and their influence on the medical care system have been well-described by people like H. Gilbert Welch in his book Overdiagnosis, by Carl Elliott in White Coat, Black Hat, and (though I haven't read this one) by Otis Brawley in How We Do Harm.

Despite its concerns, in its advice to patients, the ALA doesn't mention that based on NLST data, patients have a ~25-40% chance of having an abnormal CT scan, requiring further follow-up, and a ~95% chance that anything abnormal will turn out not to be cancer. They obliquely refer to the possibility of "more tests and procedures," but don't say that screening may harm some people without lung cancer. The American Cancer Society, by contrast, omits lung cancer screening with CT scanning entirely from its list of recommended screening tests, and publishes articles with titles like "Lung Biopsies Carry Risks."

Wellpoint is the only large insurer (to my knowledge) that has stated its plans to pay for low-dose screening CT scans. Most academic centers are offering self-pay CT scans around cost, in the $300+ range, but community medical centers are advertising CTs for as low as $99. All are presumably hoping to capture the associated revenues that follow an abnormal CT scan (and make sure St. Elsewhere across town doesn't get all of it).

Samet JM et al. Providing Guidance for Lung Cancer Screening: The American Lung Association Interim Report on Lung Cancer Screening. (American Lung Association website)