A study, published online January 6 in the Journal of the National Cancer Institute, claims to show that screening men 55 years old or older for prostate cancer does not significantly decrease mortality from the disease.
This study made headlines, in large part because it runs contrary to conventional wisdom -- that cancer screening tests, by detecting the disease at an early stage, save lives. A number of things struck me about this study and the way it was reported.
This is one of several relatively recent studies that has called into question the premise behind cancer screening tests. Two years ago, I had a post about a revision of the US Preventive Services Task Force's position on Pap smears and mammograms. In that post, I talked more about how screening tests might be applied to particular patient populations. But cervical cancer and breast cancer are very different from prostate cancer. Women rarely die WITH breast cancer... they die OF breast cancer. In contrast, many men die WITH prostate cancer, but not because they had prostate cancer. This is a key difference, and strongly influences thoughts about screening tests. KevinMD did a great job addressing this issue, highlighting the idea that screening tests are not without their costs.
This episode also demonstrates that it is important to read the whole article, not just the headline. (Reading the original research is best, but only if you know how to interpret it appropriately). The Washington Post's headline, for example, reads, "Routine prostate cancer testing does not save lives." Only if you read deep into the article does it become clear that one group of men was screened by the study doctors with PSA blood tests and digital rectal exams, while the other group of men were cared for by their regular doctors. Of course, the regular doctors were allowed to do PSA tests and digital rectal exams, too. But those data were not recorded, making it challenging to know how different the screening of the two groups really was. This is important, because the "screened" group had a higher rate of cancer than the "unscreened" group, but no difference in the rate of death from prostate cancer.
How can that be?
The answer may come from pediatric oncology. Neuroblastoma, the most common solid tumor (other than brain cancer) in children, can be detected by a simple urine test. Since neuroblastoma primarily strikes kids less than 3 years old, a population of children who see their pediatricians quite frequently, it seems sensible to do a urine screening test for neuroblastoma as part of routine well child care. In fact, that was the routine in Japan for many years. In the beginning of this century, though, two large studies showed that screening for neuroblastoma increased the rate of diagnosis, but did not change the death rate. On the basis of those two studies, Japan no longer screens children for neuroblastoma.
How can a screening program increase detection but not decrease the death rate, especially if there are effective treatments available?
In the case of neuroblastoma, we believe the answer lies in the biology of the tumor. Some tumors are very aggressive, grow rapidly, and kill a high proportion of patients, while others are more slow growing and can be cured pretty easily, often with surgery alone. If screening detects more of the low grade tumors, but the high grade tumors end up being detected because they cause symptoms, then screening programs won't change the rate of detection of the aggressive tumors... the ones that cause most of the deaths.
If this same biological principle applies to prostate cancer, the findings of the screening trial make sense. Which brings me to my last point: at the end of the article in the Post, Dr. Jonathan W. Simons was quoted as saying, "We need a better test than PSA." That may be true, but it needs to be better not because it can detect prostate cancer even earlier, but because it can detect aggressive prostate cancer earlier. Screening for that may change death rates. Maybe we don't need a "better" test... we need a different test.
More What You'd Call "Guidelines" than Actual Rules
Is the Medical Community Complicit?
HPV, STI's, and Teenaged Girls: What does 1 in 4 mean, and what can be done?