The authors begin their article by presenting the current state of affairs regarding lung cancer and the rationale as to why it would seem to be an obvious candidate to benefit from a program of early diagnosis followed by early treatment. Briefly summarized, the disease is the number one cancer killer, it is nearly uniformly fatal when diagnosis is symptom-prompted, and it is highly curable when found in its early stage.
Currently there are two main initiatives in progress in the United States to evaluate the benefit of computed tomographic (CT) screening. The first is the Early Lung Cancer Action Project (ELCAP), and the second is the National Lung Screening Trial (NLST).
The goal of ELCAP is to study the component issues in screening. This includes learning the frequency with which a regimen of screening leads to early diagnosis as well as learning how often an earlydiagnosed cancer can be cured. It is the view of the ELCAP investigators that knowledge of these two separate components is critical in terms of understanding screening, and that with knowledge of these components, all other aspects related to screening can be fully understood (including various biases and so forth). The alternative, traditional view in regard to screening is that a randomized controlled trial with a mortality end point must be performed, as has been initiated with the NLST. Proponents view this approach as the only way to prove a mortality benefit and control for biases.
While there has been much debate as to the merits of each approach, the simple truth is that both of these initiatives are ongoing. The ELCAP has accrued over 27,000 subjects that have undergone baseline CT screening, and the NLST has enrolled 25,000 subjects into a CT screening arm and another 25,000 into a chest x-ray screening arm. It is important to consider what can be learned by each approach and review where we are with CT screening today.
Early Lung Cancer Action Project
The ELCAP started its pilot project some 11 years ago. Between then and now, CT technology has undergone dramatic changes, progressively going from 10-mm to 1-mm slice thickness while maintaining the same amount of coverage. Such changes in technology-including the rapidly developing area of computer-aided diagnosis-are expected to continue over the next 10 years.
The ELCAP approach to this evolution is to continually update the protocol so as to stay current with the latest technology. Because ELCAP separates the diagnostic and interventional components of screening, each can be updated separately and stay current with the latest changes. During this time, the ELCAP investigators have also regularly updated their management protocol. For example, they have learned what to do with the ever-increasing number of nodules found on the initial baseline study. Thus, while there was concern that too many nodules were being identified that would require additional diagnostic work-up, the ELCAP protocol has now been updated so that only 12% of baseline studies need an additional study prior to the annual repeat study. These efficiencies can be expected to improve each year as new data continue to accumulate. ELCAP researchers have also learned how to incorporate new diagnostic technologies (such as computer-aided growth assessment and positron-emission tomography [PET]-CT) into their protocols and will continue to do so along the way.
On the intervention side, ELCAP is providing for learning about the various types of cancers that are being diagnosed. It has provided information about a new class of early lesions — the nonsolid nodules (so-called ground glass opacities, or GGOs) — and has learned about their prognostic implications relative to solid cancers. ELCAP investigators are also developing trials to learn about new treatment approaches to these lesions. This includes performing randomized treatment trials for small cancers, comparing limited resection to standard resection (lobectomy).
Randomized treatment trials are fully consistent and feasible within the ELCAP study design and allow for advances in treatments to be integrated as further knowledge is accumulated.[ 9] This includes the addition of new chemoprevention techniques as well. In addition to staying current with advances in technology, ELCAP also provides for ongoing assessment of cost-effectiveness. For example, the project has defined various risk categories as to who should be screened, has incorporated new ideas about risk stratification with advancing knowledge of molecular approaches, and can fully participate in these types of studies.
National Lung Screening Trial
The design of the NLST is based on the traditional screening randomized controlled trial approach. Its management protocol was substantially derived from an old ELCAP protocol. It is locked into its current paradigm until its completion sometime in 2008 or 2009, when it will announce results. Assuming that it can produce results given the many concerns about its design and that it has not been too severely contaminated by patients in the chest x-ray arm also getting CT scans, the NLST will announce whether there is a mortality benefit of CT compared to chest x-ray. It will also need to rely on the results of its predecessor study-the Prostate, Lung, Colon and Ovary (PLCO) trial-the lung component of which set out to learn the benefit of chest x-ray screening compared to no screening. That study began in 1993 and is projected to produce results in 2014. Notably, it is evaluating the chest x-ray with film, a technology that is obsolete in this new digital world.
Nevertheless, since both the ELCAP and the NLST are ongoing, it represents an extraordinary time to actually learn about the differences in these approaches. We can learn to understand what one vs the other provides, and possibly how they might complement each other. With the changes in technology (including understanding of genetics) occurring so rapidly, and with screening (particularly for cancer) having such enormous health-care implications, developing a new paradigm to efficiently evaluate the respective benefits of these approaches would be a watershed event.