I love to see the new words that are added each year to the Oxford English Dictionary. For example, this year 'selfie' made the cut. For Stedman's, 'overdiagnosis' is surely about to be added, since as Rick and I discuss on PodMed this week, we're seeing that term applied to more and more cancer screening modalities. Interesting also that the timeline for applying this term has become truncated; raising the specter of overdiagnosis with regard to prostate cancer took some several years. This week, JAMA Internal Medicine has published an analysis of the National Lung Screening Trial, hailed just a couple of short years ago as a major advance in detecting early lung cancers among those at risk for the disease by employing spiral CT, but now reinterpreting the data to demonstrate the likelihood of falsely fingering lung cancers that may not have ever caused a problem. Huh?
As a reminder, the National Lung Screening Trial employed either low-dose CT scanning or routine X-ray to annually screen almost 54,000 people with a 30-pack year history of cigarette smoking over a period of three years, with up to five years of additional follow-up. The upshot of the first foray through the data revealed that low dose spiral CT was superior to routine X-ray in detecting lung tumors and on that basis, was inserted as the screening modality of choice for those at high risk of lung cancer.
This data spin calculates several additional numbers: excess cancers detected, both from a clinical and a public health perspective using different denominators, for different types of lung cancers, and the number of overdiagnoses relative to the number needed to screen to prevent one lung cancer death. The researchers determined that there's an 18.5% probability that any lung cancer detected by low dose CT represented an overdiagnosis, and that with regard to certain types of lung tumors that probability was almost 23% for non-small cell lung cancer and a whopping 78.9% when the cancer was bronchoalveolar! If we accept the calculation that 320 people would need to be screened to prevent one lung cancer death in this population, then 1.38 cases of overdiagnosis would occur in this group as well.
So what does this mean for people who are at high risk for developing lung cancer? Rick opines that these folks should still undergo screening, but if they are told they have a suspicious lesion or frank tumor, they should seek an opinion and treatment from someone with abundant experience in this area. Lung biopsies of course are not without risk and overtreatment can carry considerable risks. The authors as well as Rick and me are hopeful that better imaging techniques, biomarkers and a host of other initiatives will bear fruit and provide us with a better method of figuring out who can be safely left alone. And of course I wouldn't be me if I didn't call for simply abolishing cigarettes as the most effective strategy.
Other topics this week include a lack of benefit seen with genetic analysis in assisting with Coumadin dosing in NEJM, diabetes and hypoglycemia risk in JAMA Internal Medicine, and lipid management in chronic kidney disease in Annals of Internal Medicine. Until next week, y'all live well.