Last week blew by me in a blur. Because I was in full grant writing frenzy to get an R01 in the can by Friday, pretty much anything that wasn’t totally urgent got shoved aside, at least after Wednesday. Of course, it was last Wednesday that yet another mammography study was being touted as a “landmark” study. I had just enough time to look it over briefly and decide that I really should blog about it, particularly given that it came hot on the heels of a Norwegian study less than a week before that found the benefits of mammography to be less than previously believed and even more particularly because this study apparently showed that mammography was much more beneficial to women between the ages of 40-49 than previously thought. Given the kerfuffle over the USPSTF update of its guidelines for mammography last fall, where the USPSTF recommended starting regular mammography at the age of 50 rather than 40 and ran into a buzzsaw of criticism, both honest, dishonest, and from those who were protecting their turf. Naturally, with that background, the question is: Does this study help to clarify the question of assessing the benefits versus the risks of screening mammography? Or does it complicate things even more?
Perhaps a little of both.
As I’ve written about over the last couple of years, evidence has been accumulating that is muddying the picture regarding the benefits of screening mammography, So, before I move on to the study, let’s be absolutely clear on what it is that we are discussing here. Screening mammography is different from diagnostic mammography in that it is performed at regular intervals in asymptomatic women in order to detect cancer at an earlier stage and thereby allow earlier intervention, resulting in the saving of more lives than if we waited until breast cancer produces symptoms (such as a lump) that lead to diagnosis. If a woman feels a lump or some change in her breast and undergoes mammography, that is not screening. In that case, mammography is being done for diagnostic purposes. We are not discussing diagnostic mammography. We are discussing screening mammography. I can’t emphasize that distinction enough.
The study that was reported last week was performed in Sweden and reported under such headlines as New mammogram study stirs debate for women in 40s; Mammogram Benefit Seen for Women in Their 40s; and Swedish mammography study sows more confusion about screening for breast cancer. I found radiologists I know circulating the study as though it were vindication, and Dr. Len of the American Cancer Society touted it as being very important, while Dr. Daniel B. Kopans (whom we’ve met before crudely and blatantly defending his turf over the USPSTF guidelines) declared that the study “should end any debate and end the use of age 50 as a threshold for screening.”
Not so fast there, pardner.
None of this is to say that mammographic screening is not potentially beneficial to women in their 40s. I also happen to have come to the conclusion based on more recent studies that that benefit is not as great as previously believed. Certainly, the potential benefit is much smaller than it is for women over 50. In fact, the introduction of the article, published by Hellquist et al and entitled Effectiveness of population-based service screening with mammography for women ages 40 to 49 years: Evaluation of the Swedish Mammography Screening in Young Women (SCRY) cohort:
Consensus has been reached that mammography screening is efficient for women ages 50 to 69 years; however, the effectiveness of such screening for women ages 40 to 49 years still is questioned. Randomized controlled trials (RCTs) have revealed a significant effect for women aged 40 years. Recommendations to invite women from age 40 years to screening based on these RCTs later were contested when meta-analyses and overviews that focused on women ages 40 to 49 years revealed no statistically significant effect (throughout this report, results are considered statistically significant at the 5% level). However, both the Gothenburg trial and the MalmÃ¶ trial reported significant mortality reductions among women aged <50 years at randomization. A few studies have focused on screening for the group ages 40 to 49. years. The Canadian National Breast Screening Study randomized women ages 40 to 49 years and invited them to 4 or 5 annual screens, but that study demonstrated no significant effect on breast cancer mortality. In the Age trial, which is the only RCT that was designed to study this age group, women were randomized at ages 39 to 41 years, and the results indicated a statistically nonsignificant 17% reduction in mortality. Few studies have investigated the effectiveness of service screening for the group ages 40 to 49 years. A study comparing breast cancer mortality in Swedish counties in which women ages 40 to 49 years were invited to screening versus breast cancer mortality among women in counties in which the same age group was not invited to screening indicated a statistically nonsignificant 14% reduction in mortality for the women who were invited to screening and were followed for 10 years. A study in northern Sweden indicated a statistically significant 36% reduction in mortality for this age group.
See what I mean? Clear as mud. The situation regarding whether regular mammographic screening for women 40-49 years of age is controversial, and the data are conflicting. Unfortunately, I don’t think this study will change that one way or the other, in favor of regular mammographic screening in this age group or against it. My personal interpretation of the situation is that there is a benefit to beginning screening at age 40, but that it is not nearly as compelling as beginning it at age 50, and there is the potential for harm through overdiagnosis and overtreatment. Remember again that I am referring to the screening of asymptomatic women who are at an average risk for breast cancer. This analysis does not apply for women with symptoms or who are at a high risk of breast cancer. I know I keep harping on that, but it’s important.
Back to the Swedish study. What the investigators was similar to what the investigators who published th Norwegian study did in that they took advantage of a quirk in how mammographic screening was rolled out in their country a couple of decades ago. In the case of the Norwegian study, some counties offered mammography before others did. The Norwegian investigators compared improvements in breast cancer survival in counties that introduced mammographic screening early with those in counties that introduced it later in order to produce an estimate of how much of the improvement in breast cancer survival was due to screening mammography and how much was due to other factors, part of which is likely to be better treatments. In this study, Swedish investigators took advantage of how different counties invited women for screening mammography after 1988. Basically, the national government had recommended inviting women above age 40 for screening mammography in 1986, it backtracked. Because of a lack of resources, the government recommended that the counties concentrate on women aged 50 and above. As a result, about half the counties continued to recommend mammography for women over 40 and the other half only invited women over 50 years old.
The primary objective of the study was to compare breast cancer mortality between the counties that did and did not invite women ages 40-49 for screening. Investigators chose as the study group women from areas where women aged 40-49 underwent screening mammography for at least 6 years between 1986 and 2005. The control group included women from areas where screening didn’t begin until age 50. These two groups represent the Mammography Screening of Young Women (SCRY) cohort, and in 1990 there were 620,620 women in this cohort. The investigators then compared the risk of dying of breast cancer between women in the control group and the experimental group. The results reported were that women who were invited for screening beginning at age 40 had a 26% lower risk of dying of breast cancer than women who did not undergo regular screening between ages 40-49, and women who actually underwent regular screening had a 29% lower risk of dying of breast cancer. The estimated number needed to screen (NNS) during a ten year period was 1,252, which means that 1,252 women would have to be screened to save one life from breast cancer. This is within the range of estimates for mammographic screening of women over 50.
Unfortunately, there are a lot of problems with this study. These go beyond the usual problems with epidemiological studies looking at retrospective, population-based data that can bias a study one way or the other. If you think about it, I bet you can pick one of these problems out right away. Think about it. This was not randomized data. Some counties chose to start screening women at age 40, and some did not. Why might that be? Governments don’t make such choices randomly. Perhaps the county governments that did choose to begin screening at age 40 were better off than the counties that did not. Perhaps there were other differences between the two sets of counties that might account for a decreased death rate from breast cancer. Breast cancer treatment improved enormously between 1986 and 2006. It’s quite reasonable to speculate that counties that were willing to invest in mammographic screening at a younger age might also have been willing to invest more in other areas of health care related to breast cancer. Another problem is the fact that various counties changed their screening strategy to begin either at 40 or at 50. One even decided to split the difference and start at age 45. These required various statistical adjustments. Whether these were done correctly or in a fashion that would tend to introduce bias is impossible to tell from the manuscript.
Finally, do you remember the term lead time bias? As I explained so long ago, lead time bias is a phenomenon that leads to the appearance that cancer survival is longer, even though it is not. The phenomenon exists because for an individual patient cancer survival is defined as the time from diagnosis to the time to death. If a cancer is diagnosed earlier, survival will appear to be longer even if treatment has no effect. I’ll reuse a diagram that I’ve used to teach the concept of lead time bias in order to help:
As I explained before, unless the rate of progression from the point of a screen-detected abnormality to a clinically detected abnormality is known, it is very difficult to figure out whether a treatment of a screen-detected tumor is actually improving survival when compared to tumors detected later. For a genuine benefit in terms of survival to be shown, the lead time needs to be known and subtracted from the group with the test-based diagnoses. The problem is that the use of the more sensitive detection tests usually precede such knowledge of the true lead time by several years. The adjustment for lead time assumes that the screening test-detected tumors will progress at the same rate as those detected later clinically.
The authors of this study concentrated on cancers detected when women were in their 40s, whether they died in their 40s or many years later of their tumors. This is appropriate given that the entire hypothesis behind screening is that early detection will result in better outcomes. However, there is a problem comparing these women to women who began screening in their 50s. In this latter group of women, it’s unknown which of them would have been diagnosed earlier if the woman had been screened in her 40s, how fast they would have progressed, and whether treatment would have had an effect. There’s no way to know which of these women should be counted and which should not. So the authors made another statistical adjustment whose rationale, truth be told, I can’t for the life of me figure out:
The person-years that were added or subtracted corresponded to the number in the continuous age interval (50 LT, 50) and in the time interval (S, S Ã¾ LT), respectively, where LT is the estimated lead time, and S the start point. However, only the lead time for women who actually died from breast cancer can cause a bias. Therefore, the lead time was estimated for this group. The lead time for women ages 40 to 49 years who died from breast cancer reportedly was much shorter than the usual lead time for all women who are targeted or who have disease detected in a screening program. In the current study, the lead time for women who died of breast cancer, which we estimated as the difference in the mean time from diagnosis to breast cancer death between the study group and the control group, was approximately 1 month. An alternative estimate based on 49 women from the RCT WE4 was approximately 1 year.
I have a hard time thinking that these guys just made it up as they went along. Lead time in breast cancer is generally much longer than one month. It’s usually at least several months and may be as long as several years. One year is not entirely unreasonable, but it’s at the low end of most estimates of lead time. Other studies use estimates of three or four years.
Deciding what diseases to screen for, what screening tests to use, what age at which to begin screening, and over what intervals are all incredibly complex questions, particularly in cancer. Although it “makes sense” that earlier detection should save lives, such is not always the case but it is some of the time. The problem is that we can’t differentiate between the cases where early detection will save lives and when it won’t. We don’t always know what the lead time bias is, nor can we identify the subset of patients whose screen-diagnosed cancers would never progress to endanger their lives. Moreover, screening always has the potential to cause harm through overdiagnosis and overtreatment, anbd that har is not always as minor as having to undergo biopsies that, in retrospect, were unnecessary. Sometimes that overdiagnosis leads to overtreatment that includes surgery, chemotherapy, and radiation. This study did not even consider false positives or other adverse outcomes.
In the end, whether to screen or not is a decision that involves scientific, cultural, and economic considerations. There is no doubt that screening mammography saves lives. What we are undergoing is a recalibration of our understanding of just how beneficial it is and at what cost. There’s little doubt that the benefits outweigh the risks for women over 50. What we are still not sure of is whether the same is true for women between 40 and 49, which is why screening between these ages still remains somewhat controversial.
The Swedish study doesn’t change that.
Hellquist, B., Duffy, S., Abdsaleh, S., BjÃ¶rneld, L., BordÃ¡s, P., TabÃ¡r, L., VitÃ¡k, B., Zackrisson, S., NystrÃ¶m, L., & Jonsson, H. (2010). Effectiveness of population-based service screening with mammography for women ages 40 to 49 years Cancer DOI: 10.1002/cncr.25650