A multiparous, 45-year-old woman attends your clinic for an annual physical examination. She does not have a family history of breast cancer and is not currently using estrogen therapy.
Should we offer this patient routine breast cancer screening with mammography?
While there is good evidence that mammography starting at age 50 reduces breast cancer mortality, this meta-analysis1does not provide evidence of a mortality benefit for mammography screening in women aged 40 to 49. Controversy exists over whether to include in the analysis trials with methodologic flaws that would strengthen the evidence of a benefit.2Nevertheless, there is a growing consensus that physicians should encourage the use of screening mammography in women aged 40 and older. The review did not address screening intervals or clinical breast examination.
Background. Mammographic screening for breast cancer is controversial, as reflected in greatly varying national policies.
Objectives. To assess the effect of screening for breast cancer with mammography on mortality and morbidity.
Search Strategy. The sources of data included MEDLINE (May 16, 2000), the Cochrane Breast Cancer Group's trial register (January 24, 2000) and reference lists, letters, abstracts, and unpublished trials. Authors were contacted.
Selection Criteria. Randomized trials comparing mammographic screening with no mammo-graphic screening.
Data Collection and Analysis. Data were extracted independently by both authors.
Primary Results. Seven completed, eligible trials involving 500,000 women were identified. The two best trials provided medium-quality data and, when combined, yield a relative risk for overall mortality of 1.00 (95 percent confidence interval [CI], 0.96 to 1.05) after 13 years. However, the trials are underpowered for all-cause mortality, and CIs include a possible worthwhile effect as well as a possible detrimental effect. If data from all eligible trials (excluding flawed studies) are considered, the relative risk (RR) for overall mortality after 13 years is 1.01 (95 percent CI, 0.99 to 1.03).
The best trials failed to show a significant reduction in breast cancer mortality (RR, 0.97; 95 percent CI, 0.82 to 1.14). If data from all eligible trials (excluding flawed studies) are considered, the RR for breast cancer mortality after 13 years is 0.80 (95 percent CI, 0.71 to 0.89). However, breast cancer mortality is considered an unreliable outcome and biased in favor of screening. Flaws are the result of differential exclusion of women with breast cancer from analysis and differential misclassification of cause of death.
Reviewers' Conclusions. The currently available reliable evidence does not show a survival benefit of mass screening for breast cancer (and the evidence is inconclusive for breast cancer mortality). Women, clinicians, and policy makers should consider these findings when they decide whether or not to attend or support screening programs.
These summaries have been derived from Cochrane reviews published in the Cochrane Database of Systematic Reviews in the Cochrane Library. Their content has, as far as possible, been checked with the authors of the original reviews, but the summaries should not be regarded as an official product of the Cochrane Collaboration; minor editing changes have been made to the text (http://www.cochrane.org)
Did the authors address a focused clinical question? Yes.
Were the criteria used to select articles for inclusion appropriate? Yes.
Is it likely that important relevant articles were missed? No.
Was the validity of the individual articles appraised? Yes.
Were the assessments of studies reproducible? Yes.
Were the results similar from study to study? No. There was statistical heterogeneity across studies. Evidence for reduction in breast cancer mortality was seen only when poor-quality studies were included. The analysis was under-powered to detect benefits in all-cause mortality, and results suggested a possible benefit or harm from mammography on all-cause mortality.
How precise were the results? Breast cancer mortality results were reported as “unreliable.”
Can the results be applied to patient care? Yes.
Do the conclusions make biologic and clinical sense? Yes. Breast cancer and death from breast cancer are less common in younger women, and mammography is less sensitive in women younger than age 50.3Furthermore, other studies have shown that benefits of earlier screening typically do not become apparent until after age 50.4,5
Are the benefits worth the harms and cost? Yes. While there are psychologic harms from false-positive results, the overall benefit in mortality and cost-effectiveness supports screening in women over age 50. There is less benefit in screening women under age 50, and the decision on whether to screen should be based on the patient's level of risk, concern about breast cancer, and concern about false-positive results.
When considering whether to implement a screening test, several factors must be considered: (1) Does early diagnosis lead to improved survival or quality of life, or both? (2) Are early-diagnosed patients willing partners in the treatment strategy? (3) Are the time and energy it takes to confirm the diagnosis and provide lifelong care well spent? (4) Do the frequency and severity of the target disorder warrant this degree of effort and expenditure?8
|Misclassification refers to errors in the categorization of an intervention or a disease state. Misclassification errors are present in most studies, but the consequences of this type of bias depend on whether the misclassification is random.6When misclassification occurs equally in two groups (e.g., equal proportions of women with breast cancer are excluded from both groups in the analysis), misclassification is “random” or “nondifferential.” However, if more cases of breast cancer were excluded from the control group than from the mammography group, the misclassification is not independent of the disease status and is termed “nonrandom” or “differential.” According to the Cochrane review authors, differential misclassification occurred in the New York trial.7The authors wrote: “We calculated that 853 women were excluded from the screened group because of previous breast cancer, compared with only 336 in the control group. If only 10 percent of these excluded breast cancer cases are added as breast cancer deaths after 18 years of follow-up, the breast cancer mortality becomes higher in the screened group than in the control group, since the difference in breast cancer mortality at that time was 44 deaths.”1|
The analysis primarily addresses the first question. There appears to be a benefit in reducing breast cancer mortality in the 50- to 69-year range, but this meta-analysis did not detect a significant mortality benefit in women aged 40 to 49. However, in a less restrictive meta-analysis, the U.S. Preventive Services Task Force (USPSTF) determined a summary RR for breast cancer mortality of 0.85 (95 percent CI, 0.73 to 0.99). The key issue is whether the recognized flaws of several of the mammography trials are serious enough to disqualify them from inclusion in a meta-analysis. The USPSTF determined that observed mortality reductions in the “flawed” trials were not likely to be explained by the biases potentially introduced by the flaws. The USPSTF included these trials, whereas the Cochrane review authors thought that the trials were “fatally” flawed and should be excluded. The Cochrane review did not include summary data on flawed trials, but other meta-analyses have demonstrated a weaker but significant benefit with screening initiated in the 40- to 49-year age group and evidence of increasing benefit and cost-effectiveness with age. Most women diagnosed with early-stage breast cancer are willing to undergo treatment9and report a favorable quality of life.10 Mammography in selected women is cost-effective and comparable to screening for cervical cancer.
When considering one of the included studies,11screening with mammography every two years beginning at age 50 prevented one breast cancer death for every 4,000 women screened in a given year, per 1,460 mam-mographic examinations, per 13.5 biopsies, and per 7.4 breast cancers detected.12 In the United States in 2001, approximately 203,500 new cases of breast cancer were diagnosed, and 39,600 women died of the disease.13 Given the frequency and severity of breast cancer, which is the second leading cause of cancer death in women,13 the expense and potential harm of mammography seem to be warranted.
Epidemiologic data from the United States and the United Kingdom demonstrate a dramatic decline in breast cancer mortality beginning in the late 1980s.14,15 It is not known, however, to what extent this benefit is the result of widespread use of mammography or improved treatment regimens with adjuvant chemotherapy and radiation therapy. Women who are diagnosed early and undergo treatment have lower mortality rates than women who do not. Identifying and screening women at high risk,16 improving the proportion of eligible women screened,12 and eliminating disparities in access to mammography and treatment are evidence-based ways to reduce breast cancer mortality.17