Dr. Tim writes that the New England Journal of Medicine study on lung cancer screening to be released Thursday cannot be put in perspective without understanding the following points:
1. Screening tests (in general) may find cancers earlier but may not ultimately make a difference in death rates.
In other words, they may lead to longer survival statistics simply because they were found earlier or because they may find cancers that would not have ultimately been lethal even if left untreated. (Part of the ongoing debate about PSA screening for prostate cancer is precisely about these issues.)
2. The only definitive way to find out if a treatment or test makes a real difference in outcome is to do a randomized, blinded trial that directly compares two similar groups as opposed to an observational trial like this one that uses past data for comparison (so-called historical controls).
We painfully learned that lesson with HRT (hormone replacement therapy) in women; the observational trials suggested marked benefits but the gold standard randomized trial by the WHI (Womens Health Initiative) showed the opposite -- not only no benefit but some danger. The concluding sentence in the editorial in the NEJM about this study states, "The study ... is a provocative, welcome salvo in the long struggle to reduce the tremendous burden of lung cancer on society."
That is right on: The study is provocative. It is a welcome salvo, but it is not final or definitive.
3. That's why most experts/organizations (including the American Cancer Society -- no shrinking violet when it comes to cancer screening in general) are not ready to suggest a change in official guidelines. They are waiting for the results of the randomized National Lung Screening Trial now under way.
(The Mayo Clinic is leading another randomized trial.)
4. Unfortunately, those results won't be available until probably late 2009, unless the study is stopped early, like the hormone study was. So what should individual doctors and patients at high risk (smokers, ex-smokers, etc.) do now? I can fully understand why individuals at high risk might not want to wait until 2009. And if they can get good scans with competent follow-up at reasonable cost (or covered by insurance) why not?
The only reason I would offer is that they might end up having unnecessary further testing -- including lung biopsy -- to prove that a suspicious spot on a scan is not cancer after all.
Many people will say the peace of mind from a clean scan outweighs any risk of further testing -- and again, I can understand that feeling. But one of the things we hope to learn from the randomized trial is whether the benefits of screening outweight these risks.
In summary, this study adds to the debate but it does not settle the ultimate question: Will CT screening actually prevent deaths that would otherwise occur, or will they simply prolong survival time, and that's survival time with unnecessary procedures or prolonged treatment that ultimately make no difference.