Breast Cancer Early Detection vs. Overdiagnosis
The NEJM has found, after a long-term analysis, that mammography for women who have no signs of breast cancer (“screening mammography”) leads to more overdiagnosis than it does early breast cancer detection.
U.S. investigators analyzed population-based cancer registry data in women aged 40 and older. Data from 1975 to 1979, an era prior to widespread screening, were compared to data from 2000 to 2002, the most recent period for which 10-year follow-up data were available.
When widespread breast cancer screening began, incidence of small in situ tumors—meaning groups of abnormal cells that are not yet spreading—increased by 162 cases per 100,000 women, from 38% to 68%, or a 30% increase. Incidence of larger tumors decreased by 30 cases per 100,000, from 64% to 32%, a 32% decrease.
The logic in favor of early screening has always been to detect small malignant tumors before they grow large enough to cause symptoms. Effective screening should therefore lead to the detection of a greater number of small tumors, followed by fewer large tumors over time. “Small tumors” are defined here as less than 2 cm in diameter.
But a careful look at the data suggests that the downward trend of large-tumor detection was guided by an increase in detection of small tumors, not by a significant decrease in large-tumor incidence. In other words, after screening mammography became widely practiced, many more cases of cancer were detected on screening than would have ever led to clinical symptoms of cancer. The rate of large-tumor detection fell not because we got better at spotting cancer early, but because the percentage of large-tumor incidence in relation to overall cancer detections went way down—the result of there now being too many cancer detections, or false positives.
Based on the decline in the tumor size-specific case fatality rate, the study estimates that screening was responsible for no more than a third of the reduction in breast cancer mortality, the other 66% accounted for by improved treatment. And the risk of overdiagnosis appears to outweigh the benefit of modestly reduced mortality due to screening mammography.
The authors write, “Assuming that the underlying disease burden was stable, only 30 of the 162 additional small tumors per 100,000 women that were diagnosed were expected to progress to become large,” which implied 132 cases of overdiagnosis per 100,000 women. Overdiagnosis, then, is a significant and still under-recognized concern. Average-risk patients should ask their doctors to follow U.S. Preventive Services Task Force guidelines: screenings every two years starting at age 50.
Using probability to determine true risk
In the larger picture, this important study also illustrates how we as a culture, clinicians not excepted, too often struggle to understand statistics and probability, a topic broached in “Understand Probability To Make Smarter Health Choices,” in the Nov/Dec 2014 issue of Running & FitNews®.
We ought to dive a little deeper in the context of breast cancer screening to see how math plays tricks on us that can have real-world implications, such as in the form of alarming rates of overdiagnosis and false positive-reporting to patients.
In 2014 we looked at studies in Germany and the U.S. in which researchers asked physicians to estimate the probability that an asymptomatic woman aged 40 to 50 who has a cancer-positive mammogram actually has breast cancer if 7% of mammograms show cancer when there is none. While the correct answer was that a cancer-positive mammogram was due to cancer in only about 9% of the cases, 95% of American physicians estimated the probability to be approximately 75%.
Here is a slightly easier exercise that illustrates how to correctly use probabilistic thinking. Note that the following is a mathematical word problem—not actual cancer data.
Let’s say the facts are these: 100 out of 10,000 women at age 40 who participate in routine screening have breast cancer. 80 of every 100 women with breast cancer will get a positive mammography. 950 out of 9,900 women without breast cancer will also get a positive mammography. If 10,000 women in this age group undergo a routine screening, about what fraction of these women with positive mammographies will actually have breast cancer?
The answer is just 7.8%.
To arrive at that correct answer, begin by understanding that in this case, we want to know what percentage of the women with positive mammographies actually have breast cancer. So how many positive mammographies are there? That number becomes the denominator in a simple division problem.
Since 950 of the 9,900 women that do not have breast cancer will have a positive mammography, and 80 out of the 100 women who do have breast cancer will get a positive test result, 1,030 women will have a positive test result.
How many of those 1,030 women with a positive test result actually have breast cancer? Our hypothetical data tell us that 80 of the 100 women with breast cancer will get a positive test result, so 80 becomes the numerator of the division problem.
In this example, the fraction of women with positive test results who actually have breast cancer is 80/1,030, or .0777, or 7.8% probability.
Needless to say, conveying this information to a 40-year-old patient who just tested positive after a routine mammography is a lot less stressful than assessing her cancer risk at 10 times that. And that in itself is good for your health.