The issue of CT screening for lung cancer is a big one, and to handle it properly I’m going to write about it in a few installments. It’s also quite controversial, so today I’ll start with the reasons in favor of CT screening. Just by means of background, I’ll start by saying that chest x-rays have been studied for screening, but they don’t provide enough detail, requiring tumors to be larger before they are reliably detectable, and location of the tumor can be a problem. For instance, you can see here a chest x-ray (CXR) and a CT scan from the same patient, in whom there is really no way to find anything wrong with the CXR on the left, but the CT shows a small, well-hidden nodule (circled in red) behind the mediastinal blood vessels and other structures:
Other relevant background information is that only approximately 25% of lung cancers are detected as stage I or II NSCLC (or about 30-35% as LD-SCLC), so most patients present with at least locally advanced NSCLC, and at least 40-50% of patients with lung cancer first have it diagnosed as advanced disease, when it cannot be cured. There is no government or medical authority that recommends routine screening for lung cancer at this time, so the current policy is essentially that you don’t start a work-up unless or until someone develops symptoms, at which time, they often have very advanced disease that has little or no chance of being cured. Here’s an unfortunate CXR appearance we see far too often at the start of a work-up:
It’s a system that leaves a lot of room for improvement.
There is really no question that CT scans can detect smaller cancers. There are multiple studies that have shown that an initial CT scan followed by a follow-up yearly CT scan in patients with a high enough level of risk (often defined as age of 40 or greater and a smoking history of at least 10 pack-years, the product of number of years smoked and average number of packs smoked per day) can detect far more cancers than a similar group of patients who undergo repeated CXRs. Among the most well-known trials is the International Early Lung Cancer Action Project, or I-ELCAP, which had results published in the New England Journal of Medicine in October, 2006 (abstract here). This was an international collaboration of academic and some community-based research sites that participated in a protocol in which over 31,000 patients were enrolled who were 40 and over and with a smoking history, or in some locations with occupational exposure to asbestos, radon, or other potential toxins, or possibly exposure to significant second-hand smoke. They had a baseline CT scan followed by yearly annual screening, unless they had a solid nodule at least 5 mm in size or a larger (8 mm or greater) non-solid nodule, in which case they were recommended to undergo additional work-up. If the nodules were still fairly small, the recommended follow-up was a repeat CT 3 months later to look for interval growth, and a biopsy if that was the case. Another options was to have patients undergo a PET scan to see if the nodule was metabolically active, and to biopsy if it did appear abnormal on PET. Finally, patients with a nodule of 15 mm or larger could undergo a biopsy immediately as part of the protocol. For cases in which infection was suspected based on appearance on CT or the clinical picture, antibiotics were given for a couple of weeks. In addition to the large number who initially participated, over 27,000 repeat screenings were performed.
The reason this was published in the New England Journal of Medicine was that lung cancer was detected in 484 patients, including 405 at the initial CT and 79 on follow-up scans. Of these, 412 (or 85%) had a stage I cancer, which is of course remarkably better than what you’d expect to see in the general population (10-15% stage I). The estimated 10-year survival of these patients was 88%, which is amazing for lung cancer in general and even very impressive for stage I lung cancer. The argument that these were not just clinically insignificant was bolstered by the fact that the 8 patients who had an apparent stage I lung cancer detected and declined treatment died within 5 years after diagnosis.
Importantly, several other smaller screening studies have shown a similar markedly higher likelihood of finding a cancer early and seeing better survival among those patients than you would ever expect for lung cancer. For instrance, one of these trials comes from Italy (abstract here), and another from the Mayo Clinic (abstract here). The results are consistent, with the I-ELCAP trial dominating because of the size and the fact that it included patients from multiple countries and treatment settings.
There are also a few other advantages. First, many of the patients who participate on lung cancer screening studies are currently smoking. Here’s the CT of one gentleman with both an early lung cancer and his cigarettes and lighter in his front pocket:
So having people there and thinking about the health risks of lung cancer provides a “teachable moment”. In fact, one investigator found that 14% of patients who were current smokers at the start of a screening studies had quit at their one-year follow-up, and it was all the way up to 24% of the people returning for their 3-year follow up (abstract here). However, at the same time, there are some people who receive good news from a screening CT and feel that this “clears” them to continue smoking…Ughh.
Finally, although there aren’t detailed studies of it, there is also the potential for screening CT scans to detect other cancers or life-threatening diseases. We can often see evidence of atherosclerosis, or hardening of the arteries, around the heart, and that may be another piece of evidence to convince someone to stop smoking, get a cardiac workup, etc.
So there is really no question that there is a good side to screening. However, there are downsides to consider, and I’ll cover those issues next.