Lung Cancer Alliance Issues Statement on Recently Released Federal Study on Lung Cancer Screening
Washington, DC [April 28, 2010] --
Lung Cancer Alliance acknowledges the on-going work by the Office of
Disease Prevention (ODP) at the National Institutes of Health (NIH) on
lung cancer screening research and the need for people who may be at
risk for lung cancer to be well informed.
However, the paper by Jennifer M. Croswell, MD and co-authors published last week in the Annals of Internal Medicine,
claiming that CT scans yield twice as many false positives as chest
x-rays and lead to a high percentage of invasive follow-up procedures,
is based on a study (the Lung Screening Trial -LSS) done in 2000-2001
that used outdated equipment and failed to follow a uniform diagnostic
This repackaged data from a flawed ten year old study
ignores the exponential advances in CT imaging that have been made
since then, and fails to acknowledge the great progress that dedicated
screening researchers in the United States and abroad have made and are
continuing to make in perfecting the CT screening protocol.
It is not surprising that CT scans picked up more
suspicious nodules than the chest x-rays in the LSS study. CT scanners
are designed to be more sensitive and can pick up smaller nodules which
chest x-rays will frequently miss.
But screening is not a one step procedure. The protocol
for how those nodules are evaluated, followed-up and diagnosed is
critical to the screening process.
The LSS did not have a defined, uniform,
well-considered procedure in place and, from a patient advocate
perspective, showed the pitfalls of screening without a good diagnostic
protocol. Indeed, every screening research trial since then has
Recently, the first published papers on the
Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) trial, a
randomized controlled trial of 16,000 people in the Netherlands and
Belgium that uses a defined diagnostic work-up protocol, reported that
over 70% of cases were being diagnosed at Stage 1, with unprecedented
levels of 95% sensitivity and 99% specificity in diagnosing lung cancer
The International Early Lung Cancer Action Program
(I-ELCAP), the longest and largest ongoing CT research program
involving over 50,000 people at 50 sites in nine countries, has
developed a diagnostic and management protocol that incorporates
technology advances as they come on line and which is continuously
update to improve detection and reduce risk. Under this optimized
protocol, 80% of lung cancers are being diagnosed at Stage 1 and those
treated promptly have 10 year survival rates of nearly 90%.
Both these trials show a marked distinction from 16% rate of Stage 1 diagnoses for lung cancer outside of screening.
The first actuarial analysis ever done of over 300,000
lung cancer patients in the NCI's SEER registry was published two
months ago. Carried out by Milliman Inc, an internationally renowned
insurance actuarial firm, the analysis found that diagnosing a greater
frequency of newly detected lung cancer cases at Stage I rather than
Stage III or VI could be associated with a reduction of lung cancer
deaths by 70,000 lives a year.
If there were a more balanced presentation of the
current state-of-the-art in the promising areas of CT-based lung cancer
screening, the headline should be: “How far we have come in 10 years
to better detect, diagnose and treat lung cancer – the most lethal of