Wednesday, December 3, 2014

The United States Preventative Services Task Force (USPSTF) Mammography recommendations REVISITED

A couple of months ago, I received a form letter from a large local primary care group here in Denver.  In the letter, I was asked to put the USPSTF mammography screening recommendations into the results letters we send to their patients.  As you all probably know, the Task Force recommendations are in contradiction to those of the majority of scientific and advocacy groups.  While most organizations continue to recommend annual screening mammograms for all women 40 years old and older, the Task Force recommendations, which came out in November, 2009, do not routinely recommend screening at all in women in their 40s and only biennially for women aged 50-74.  Below is my response to this request.  Although it is a little lengthy, I hope that it expresses my position and the position of thousands of physicians and dozens of medical organizations in a logical and evidenced-based way.  Thanks for reading!




I am in receipt of your letter dated 10 Oct 2014 regarding the verbiage placed on normal screening mammogram reports generated from our comprehensive Breast Care Center at Exempla Saint Joseph Hospital.  You are requesting that we place mammography screening recommendations by both the ACS (annually from age 40) and the USPSTF (not routinely recommended prior to age 50 and every other year from 50-74) on the reports generated for your patients.  I regret that we will be unable to do so.  I will here try to explain why.

I am sensitive to the fact that the American Academy of Family Physicians and the American College of Physicians, the governing bodies of Family Physicians and Internists, have embraced the 2009 USPSTF recommendations for mammography screening.  Of note, however, is that 7 of the 16 members of the Task Force were either Family Practice or Internal Medicine physicians, while none were in specialties which screen, diagnose or treat women with breast cancer.  And the Task Force themselves state: "Recommendations made by the USPSTF are independent of the US government.  They should not be construed as an official position of the Agency for Healthcare Research and Quality or the US Department of Health and Human Services."

A closer look at these recommendations, furthermore, may make clinicans think twice about continuing to follow them. 

Two of the most glaring errors of the Task Force recommendations were 1) the estimate of the value of screening women in their 40s compared to age 50 and over and 2) the overemphasis of the theoretical 'harms' of screening (as opposed to the true harm of dying of breast cancer).

In their own estimation discussed on the Task Force website, the RR of breast cancer death for 39-49 y/o women undergoing screening was nearly identical to the RR for women aged 50-59 (0.85 and 0.86 respectively).

Although they admit that 'the risk for breast cancer increases steeply with age starting at age 40,' they made a 'C' recommendation for screening women in their 40s, recommending against routine screening.  This was largely due to their estimate of the 'number needed to be invited to be screened' in the original RCTs of 1904 for women aged 40-49 and 1339 for women aged 50-59, making the judgment that the former number was too large to justify screening in the younger age group.  This despite their own website recognition that the death rate from breast cancer has fallen 3.3% annually in women in their 40s since widespread mammography screening programs were introduced in 1990.

The main mistake they made was using the number needed to be invited to be screened, rather than the actual number needed to be screened to save one life. In assessing the utility of a screening exam, the benefit can only be judged by women actually receiving the screen.  Peer-reviewed estimates of service screening programs of how many women need to be screened annually to save a life are 746 (age 40-49), 351 (age 50-59) and 232 (60-69).  Taking the entire age group 40-84, the estimated number of women who need to be screened annually to save one life is only 84. (AJR:198, March, 2012, pp. 723-728). 

Despite ones opinion about the utility of screening certain age groups at certain intervals, some basic scientific truths are irrefutable: 

Firstly, the main predictors of surviving breast cancer are the size of the tumor (higher survival below 1.5 cm at the time of diagnosis), the presence or absence of regional lymph node metastases and the histologic grade at the time of diagnosis. 

Secondly, the average size of a clinically diagnosed breast cancer is 3-4cm. 

Thirdly, the sojourn time (tumor growth rate) is the time it takes a breast cancer to go from a screen-detectable size to being felt on exam. 

And, fourthly, dedifferentiation of histologic grade occurs when a lower grade tumor becomes higher grade the longer it goes undiagnosed in the breast. 

Since screening mammography can easily detect smaller, non-metastatic cancers and since small (<1.5cm) cancers of high histologic grade have a similar survival to lower grade cancers, mammography is an effective screening test for breast cancer and is the only test proven to decrease the death rate from the disease (Int J Cancer; 66, 413-419, 1996; Cancer August 1, 1999  volm 86  #3; 449-462).  This is true for both the 40-49 group and the 50 and older group (Cancer, Feb 15, 2011, volm 117,  #4;714-722). 

Furthermore, since premenopausal women have cancers more likely to grow faster, metastasize sooner and dedifferentiate to higher grades, screening women in their 40s should be self-evident.  And, since the sojourn time for breast cancers in women in their 40s is right at 2 years, screening less frequently than yearly is much less effective.  Even the Task Force's own words for women in their 40s that there is 'moderate evidence that the net benefit is small' (emphasis mine) is not a resounding or confident condemnation of screening in this age group.    

Finally, my own clinical experience with 'screen drift' indicates that recommending women to have annual screens frequently results in screens every 12-18 months while recommending biennial screening frequently results in screens greater than 24 months apart. 

In terms of harms, the Task Force, on their website, sites 'anxiety, distress and other psychological effects' from abnormal mammography results and the 'inconvenience due to false-positive screening results.'  They also mention the harms of the 'early treatment of cancer' that 'would not have shortened a woman's life.'  However, they, like other critics of screening mammography, make no effort to identify exactly which cancers would have or would not have killed the patient had they gone undiagnosed by screening mammography. 

In fact, the 'harms' of abnormal mammograms and the additional imaging and biopsies they cause, including increased radiation exposure, are overblown and have been calculated as follows for women aged 40-49:
:
·         Risk of recall for additional diagnostic work up:  once every 12 years
·         Risk of undergoing a negative biopsy:  once every 149 years
·         Risk of a missed breast cancer: once every 1000 years
·         Risk of suffering a fatal radiation-induced breast cancer: once every 76,000-97,000 years

Missing, too, from the Task Force's discussion is the difference in lives saved between following their recommendations as opposed to annual screening from age 40-84 (23.2% vs. 39.6% mortality reduction), as mortality is the end measurement of screening programs.  It is estimated that at a 64% screening rate, nearly 65 000 more lives would be saved in the lifetimes of women turning age 40 for US women currently aged 30-39 by following annual recommendations. Likewise, for all US women 40 and over, nearly 6500 more lives are saved annually by following the annual recommendations of the ACS (AJR:196, February 2011;W112-116). 

The Task Force even ignored their own models comparing various annual and biennial screening regimens, published concurrently with their 2009 recommendations.  They all showed higher mortality reductions with annual screening, with the most lives saved with screens from age 40-84 (Annals of Internal Medicine, Volume 151,Number 10, 17 Nov 2009. p. 738-747. W243-W247). 

Finally, I can only draw on my own 17 years of practice experience to say that the vast majority of women have no lasting harmful effects of a 'false positive' screen or a negative biopsy.  On the contrary, they are almost universally appreciative of the thoughtful and thorough care that they receive in our clinic. 

I am all for shared decision making.  I do it every day in my clinic when I can show a patient her actual finding on an actual mammogram and discuss with her the pros and cons of ignoring it, following it or biopsying it.  And although the thought of a thorough discussion with a newly-turned 40 year-old patient on the pros and cons of starting mammographic screening is laudable, the evidence shows that clinicians undergo shared decision making with their patients less than 40% of the time

As stated above, the Task Force admits that their recommendations are not US government policy and, in fact, an amendment was placed into the Affordable Care Law which ignores their recommendations and mandates no-cost annual screening for all women from age 40 on.


To me, the evidence is overwhelming for recommending annual screening mammography for all average risk women from age 40 on.  The actual screen itself is not the harmful event. It's what happens after the screen which determines whether the patient benefits from that screen. This is a responsibility which falls on my shoulders as a fellowship trained breast radiologist.  It's a responsibility that I take very seriously and I happily engage every patient with an abnormal mammogram in the shared decision making of what is best for her.  

1 comment:

  1. This comment has been removed by the author.

    ReplyDelete