Recommendation Statement From USPSTF: Screening for Breast Cancer

US Preventive Services Task Force (USPSTF)


December 17, 2009

In This Article


Burden of Disease

Breast cancer is the most frequently diagnosed cancer in women in the United States, not including skin cancer, and is second only to lung cancer as a cause of cancer deaths. In 2008, an estimated 182,460 cases of invasive cancer and 67,770 cases of in situ breast cancer were diagnosed, and 40,480 breast cancer deaths occurred.[9] The National Cancer Institute, on the basis of Surveillance Epidemiology and End Result data, estimates the lifetime risk for a woman to develop breast cancer at 12%.[10] The risk for breast cancer increases with age. The 10-year risk for breast cancer is 1 in 69 for a woman at age 40 years, 1 in 42 at age 50 years, and 1 in 29 at age 60 years.[11] The incidence rate of breast cancer has increased since the 1970s; however, recent data show that this rate seems to be decreasing, both overall and on an age-adjusted basis. The incidence rate in 2003 was 124.2 per 100,000 women, a 6.7% decrease from the previous year.[12] Discontinuation of hormone replacement therapy may be largely responsible for this observed decrease,[12,13] although slowed growth or even a decline in screening mammography also may have contributed.[14] Breast cancer mortality has been decreasing since 1990 by 2.3% per year overall and by 3.3% for women aged 40-50 years. This decrease is largely attributed to the combination of mammography screening with improved treatment.[15]

Scope of Review

The systematic evidence review undertaken in support of this recommendation[7] addressed the efficacy of 5 breast cancer screening methods for reducing breast cancer mortality -- film mammography, CBE, BSE, digital mammography, and MRI -- by using published reports of randomized, controlled screening trials and specifically updated information from mammography trials among women in the age groups of 40-49 years and 70 years or older. Information on harms of breast cancer screening, such as false-positive test results, pain, anxiety, and biopsy rates, was sought from multiple sources, including systematic reviews, meta-analyses, and recently published literature. To assess the follow-up testing and other outcomes of a mammography screening program, the reviewers included data from the Breast Cancer Surveillance Consortium from 2000 to 2005.

In addition to the systematic review of screening tests, the USPSTF requested a report from the CISNET Breast Cancer Modeling Group to provide data from comparative decision models on optimal starting and stopping ages and intervals for screening mammography.[8]

Accuracy of Screening Tests

Mammography, CBE, and BSE are recognized approaches for breast cancer screening. Since the 2002 USPSTF recommendation statement, digital (as opposed to film-based) mammography has been increasingly used, and MRI is being used with greater frequency for screening women at increased risk for breast cancer. The sensitivity of mammography screening is 77%-95%, whereas specificity is 94%-97%.[16] Multiple factors, including age, time since last examination, breast tissue density, equipment, and the skill of the interpreting radiologist, can affect sensitivity and specificity.[17] A single, large comparison study of film and digital mammography[18] demonstrated similar diagnostic accuracy for the 2 methods, although digital mammography was better at detecting lesions in women who were younger than 50 years or premenopausal or had radiographically dense breasts. Studies of MRI in high-risk women without cancer[19] showed that MRI has a sensitivity of 71%-100% and a specificity of 81%-97%; MRI is, therefore, recommended by the American Cancer Society for women at high risk for breast cancer. However, no studies have been done on using MRI to screen women at average risk.

CBE has a sensitivity of 40%-69% and a specificity in the range of 88%-99%. For BSE, sensitivity ranges from 12% to 41%, lower than that of CBE and mammography, and is age dependent.[16]

Effectiveness of Early Detection

The newly updated meta-analysis by Nelson and colleagues[7] confirms an earlier finding[16] that screening mammography reduces mortality. Improvements in the RR for death due to breast cancer for women aged 39-49 years and 50-59 years are similar at 0.85 (95% confidence interval [CI], 0.75-0.96) and 0.86 (CI, 0.75-0.99), respectively. An even greater improvement was found for women aged 60-69 years (RR, 0.68 [CI, 0.54-0.87]). Results were uncertain for older women, with a new report from a previously included trial with longer follow-up data showing an RR of 1.12 (CI, 0.73-1.72) for breast cancer death associated with screening women aged 70-74 years (based on a small number of participants). At the time of the previous meta-analysis, data from 2 trials showed an RR of 0.78 (CI, 0.62-0.99) for breast cancer death in screened women aged 65-74 years.[16] Mortality reduction and life-years gained[8] were both considered important outcomes of screening in forming this recommendation; in the end, mortality reduction as observed in trials was the metric chosen to express the benefits of screening in each age group. Digital mammography has been shown to perform similarly to film mammography, although it has never been studied in relation to health outcomes. MRI has not been evaluated for its potential benefit in screening average-risk women. CBE is inexpensive and easy to perform; however, few studies have compared the effectiveness of CBE with that of no screening, and none has examined CBE and mammography vs mammography alone. Two large trials of teaching BSE outside the United States[7] demonstrated no mortality benefit in the intervention groups.

Potential Harms of Screening

Mammography. False-positive results are common with mammography and can cause anxiety and lead to additional imaging studies and invasive procedures (such as biopsy or fine-needle aspiration). False-positive results and accompanying additional imaging studies are more common in younger women. Biopsies may occur as a consequence of false-positive mammography results; biopsy rates are more common among older women. Anxiety, distress, and other psychosocial effects can exist with abnormal mammography results but fortunately are usually transient, and some research suggests that these effects are not a barrier to screening. False-negative results occur at a relatively low rate for all ages but are slightly higher in women older than 70 years. Other potential harms, such as pain caused by the procedure, exist but are thought to have little effect on mammography use.

Overdiagnosis can occur when screening detects early-stage invasive breast cancer or DCIS in a woman, typically older, who is likely to die of another cause before the breast cancer would be clinically detected. Overdiagnosis can also occur in younger women if a detected DCIS or other early-stage lesion never progresses to invasive cancer. Methods for estimating overdiagnosis at a population level are not well established, and, thus, the proportion of all detected DCIS lesions that constitute overdiagnosis is uncertain.[7] Similarly, unnecessary earlier treatment can occur at any age when screening detects a slower-growing cancer that would have eventually become clinically apparent but would never have caused death.

Radiation exposure may increase the risk for breast cancer but usually only at much higher doses than those used in mammography, although regular mammography could contribute to cumulative radiation doses from additional imaging for other reasons.[7]

Digital mammography and magnetic resonance imaging. Digital mammography can be expected to have harms similar to those of film mammography. No studies have evaluated MRI screening of average-risk women; in theory, because MRI could be considerably less specific than mammography screening, it could potentially be associated with higher biopsy rates and greater degrees of overdiagnosis if used in low-risk populations.

Clinical breast examination. A theoretical harm of CBE is that its lower specificity than that of mammography may result in more women undergoing biopsy. Few data are available to evaluate harms associated with CBE.

Breast self-examination. One study indicated that anxiety was not a concern with BSE. The 2 available trials[20,21] indicated that more additional imaging procedures and biopsies were done for women who performed BSE than for control participants.

Estimate of Magnitude of Net Benefit

In 2002, the USPSTF concluded that there was fair evidence that mammography screening every 12-33 months could significantly reduce breast cancer mortality. The evidence was strongest for women aged 50-69 years, with weaker evidence supporting mammography screening for women aged 40-49 years. Since that recommendation, 1 new trial and updated data from an older study have been published that specifically address screening in women in the younger age group. These findings were combined in an updated meta-analysis, which resulted in an RR for breast cancer death of 0.85 (CI, 0.75-0.96; 8 trials) and a number needed to invite for screening of 1904 (CI, 929-6378) to prevent 1 breast cancer death in women aged 39-49 years. A meta-analysis of 6 trials among women aged 50-59 years and 2 trials among women aged 60-69 years provided pooled RRs for breast cancer death in the screened group of 0.86 (CI, 0.75-0.99; number needed to invite, 1339 [CI, 322-7455]) and 0.68 (CI, 0.54-0.87; number needed to invite, 377 [CI, 230-1050]), respectively. Only 1 study provided data on women older than 70 years, yielding an RR of 1.12 (CI, 0.73-1.72), although an RR of 0.78 (CI, 0.62-0.99) was found for women aged 65-74 years by combining the results of 2 studies used in the 2002 review.

A decision analysis performed for the USPSTF projected that biennial screening produced 70%-99% of the benefit of annual screening, with a significant reduction in the number of mammograms required and, therefore, a decreased risk for harms. Screening between the ages of 50 and 69 years produced a projected 17% (range, 15%-23%) reduction in mortality (compared with no screening), whereas extending the age range produced only minor improvements (additional 3% reduction from starting at age 40 years and 7% from extending to age 79 years).[8]

The USPSTF noted with moderate certainty that the net benefits of screening mammography in women aged 50-74 years were at least moderate and that the greatest benefits were seen in women aged 60-69 years. For women aged 40-49 years, the USPSTF had moderate certainty that the net benefits were small. Because of the uncertainties related to harms of screening, particularly overdiagnosis, and the near total lack of trial data for older women, the USPSTF had low certainty about the net benefits of screening mammography for women 75 years or older.

How Does Evidence Fit With Biological Understanding?

Current knowledge about the development of breast cancer is limited. The effectiveness of screening mammography seen in trials presumably results from the early detection of smaller, earlier-stage tumors, which are more responsive to available treatments. Although the most common breast cancer occurs in the epithelial cells that line the duct system of the gland (ductal carcinoma), the sequence of development of invasive cancer is not entirely known. For example, DCIS does not always represent a precursor to invasive ductal cancer. Studies of women with untreated DCIS showed progression to invasive disease in half or fewer of the cases. Because DCIS is often found only by mammography, its incidence has increased steadily since the advent of widespread screening mammography. In 1983, 4900 cases of DCIS were diagnosed; by 2008, that number was expected to be 67,770.[22] Because the likelihood that DCIS will progress to invasive cancer is unknown, surgical removal -- with or without adjuvant treatment -- may represent overdiagnosis or overtreatment. Lobular carcinoma in situ, in contrast, is not considered a true precursor lesion but connotes a higher risk for subsequent invasive lobular or ductal cancer in either breast. Lobular carcinoma in situ is often multifocal, appearing in several distinct locations. Knowledge of what determines the rapidity with which invasive cancer can spread (tumor characteristics, host factors, hormonal triggers) is limited.[23] Because of these elements of biological uncertainty, it is clear that lesion sensitivity alone is not a sufficient metric for assessing effectiveness of new screening methods.[24]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: