Breast Cancer Screening: At What Age to Stop?

ABSTRACT: Some women 75 and older who are in good health and have excellent functional status may benefit from mammography screening, while others who are in poor health and have short life expectancies probably do not. The most significant risk of screening the oldest-old women is the detection of tumors that may not become clinically important during a patient's lifetime. The American Geriatrics Society encourages screening mammography for women younger than 85 who have at least 5 years' life expectancy and for healthy women 85 and older who have excellent functional status or who feel strongly about the benefits of screening. Several tools are available to help you target screening to older women with longer life expectancies. When you discuss stopping screening with older women who have shorter life expectancies, it is important to address the issues that make women want to continue screening, including habit and the need for reassurance.
________________________________________________________________________________________________________________________________________________

As more and more adults are leading healthy and independent lives well into their 80s and 90s, clinicians are increasingly faced with the difficult decision of whether to screen these patients for cancer. Guidelines from the US Preventive Services Task Force (USPSTF) now provide age cut-offs at which to stop screening older adults for prostate and colon cancers.^1,2 The USPSTF does not provide an age cut-off at which to stop screening older women with mammography; however, the USPSTF recently concluded that the evidence is insufficient to recommend for or against breast cancer screening for women 75 years and older and further recommends that these women understand the uncertainty of the balance of benefits and harms of mammography before undergoing screening.³

Although experts recommend stopping mammography screening among older women with less than 5 to 10 years' life expectancy,^4,5 I found that 39% of women 80 and older with probable life expectancies of less than 5 years were screened with mammography.⁶ This may be partly attributable to clinicians' discomfort about discussing with older women the decision to stop screening, difficulty in determining patient life expectancy, and uncertainty about guidelines.⁷ 
_____________________________________________________________________________________________________________________________________________
Related Content
The Case for Continued Breast Cancer Screening in Older Women
Challenges in the Diagnosis and Treatment of Breast Cancer in the Elderly
____________________________________________________________________________________________________________________________________________

One of the reasons that guidelines for breast cancer screening among older women are vague is that none of the 8 randomized controlled trials that evaluated mammography included women 75 and older. Observational studies demonstrate a possible decrease in breast cancer mortality among healthy women 80 and older who are regularly screened with mammography^8-10; however, these studies are limited by lead time, length time, and selection bias.

While some experts generally encourage screening elderly women, others warn that screening may result in the detection of tumors that would not have become clinically important if they had not been found with mammography (overdiagnosis).⁴ In these cases, oldest-old women are exposed to the burdens of treatment and the anxiety of living with cancer without the benefit of improved quality or quantity of life.⁴

In this article, I discuss the benefits and risks of screening older women with mammography, and I review current policy and guidelines. I also provide tools to help you decide whether or not to screen an individual patient.

EVIDENCE SUGGESTING BENEFIT OF MAMMOGRAPHY SCREENING FOR THE OLDEST-OLD

While breast cancer–specific mortality rates have declined among women younger than 70 years, they have risen for women 80 years and older; these trends may be a result of underscreening and/or undertreatment of the oldest-old.^11-13 The oldest-old contribute the greatest burden to the number of total deaths from breast cancer each year (approximately 27% of breast cancer deaths in 2001 were among women 80 years and older, even though they had only 13% of incident breast cancers).¹²

Three observational studies have examined the effect of screening mammography on breast cancer mortality^8-10:

• McCarthy and associates9  found that regular mammography was associated with lower breast cancer mortality in women older than 75 years, but the mortality reduction was not statistically significant for women 85 and older after adjusting for comorbidity and considering lead time bias. The adjusted hazard ratio was 1.5 (95% confidence interval [CI], 0.6-3.3) for non-users of mammography 85 and older compared with 2.5 (95% CI, 1.7-3.6) for those aged 75 to 84 years and 2.1 (95% CI, 1.5-3.0) for those aged 67 to 74 years.

• McPherson and colleagues10  found a significant decrease in all-cause mortality for women aged 75 to 84 with no or moderate comorbidity whose tumors were detected mammographically, but they found no mortality benefit for women 75 to 84 years with severe comorbidity (ie, Charlson Comorbidity Index of 2 or greater).

• Badgwell and associates8  found a breast cancer survival benefit for women 80 and older who were regularly screened with mammography; however, they also found that mammography was associated with a survival benefit for causes of death other than breast cancer, which suggests selection bias.

Although it remains unknown whether earlier detection with mammography leads to a survival benefit in older women, several studies demonstrate that screening older women is associated with detection of breast cancer at earlier stages, which may result in reduced morbidity from advanced disease.^8-10 In addition, since the sensitivity of mammography increases as women age, older women are less likely to experience a false-positive test than younger women.¹⁴ The most common benefit of mammography screening may occur every time the results are negative and the oldest-old are reassured of their health.¹⁵

EVIDENCE SUGGESTING RISK OF MAMMOGRAPHY SCREENING FOR THE OLDEST-OLD

The most significant risk of screening the oldest-old may be finding tumors that may not have become clinically important during a patient's lifetime.⁴ Oldest-old women may be at greater risk for overdiagnosis than younger women for the following reasons:

•Changes in breast cancer biology:  Evidence suggests that the interval between when a cancer is detected by mammography and when it becomes clinically evident (sojourn time) increases with age, from 2 years for women in their 50s to as long as 9 years for women aged 80 to 84.¹⁶ Some data suggest that older women are more likely to have less aggressive tumors and more favorable biological characteristics (fewer positive lymph nodes, greater percentage of estrogen- and progesterone-receptor–positive tumors) than younger women; moreover, the metastatic spread of breast cancer is slower in older women.¹⁷ In general, the growth rate of a tumor is related to its aggressiveness. Because less aggressive tumors usually have longer sojourn times, they are also more likely to manifest later in life and to be more prevalent among older women.¹⁷

•Decreased life expectancy:  Women with life expectancies of less than 5 to 10 years are unlikely to derive any survival benefit from screening mammography.^4,5  Meanwhile, the average life expectancy for a woman aged 80 is 9.8 years and for a woman aged 85 is 7.2 years.¹⁸

•Ductal carcinoma in situ (DCIS): Only 28% to 39% of cases of DCIS, a noninvasive form of breast cancer, are thought to develop into invasive breast cancer after a 10- to 15-year follow-up.¹⁹  Because of the long period before DCIS may progress to invasive breast cancer, identifying DCIS in the oldest-old probably represents overdiagnosis.

•Competing risks:  Fewer than 2% of women 80 and older die of breast cancer.²⁰  Even among women with a history of breast cancer, the oldestold are much less likely to die of this disease than younger women (73% of women aged 50 to 54 with breast cancer die of this disease compared with 29% of those 85 and older).²¹ One study estimated that women with breast cancer and more than 3 comorbid conditions were more than 20 times more likely to die of a cause other than breast cancer within 3 years.²¹ Given this high risk of noncancer death, treatment of breast cancer in patients with significant comorbidity could be construed as overtreatment (treatment of breast cancers that would not have become clinically important).

•Uncertainty about benefits of treatment of breast cancer in elderly women: A key prerequisite for screening a population for a disease is the availability of treatments that when given early in the course of the disease result in greater longevity or improved quality of life than when given when the disease becomes clinically detectable.²²  However, the benefits of treating breast cancer in the oldest-old are uncertain because few of these women participate in clinical trials and also because treatments vary.

Many studies have shown that older women are less likely to receive definitive or recommended care for a new diagnosis of breast cancer than younger women.^23,24 In addition, treatment of breast cancer in the oldest- old women is associated with greater morbidity. Complications from surgery increase as women age (operative mortality rates between 1% and 2% have been reported for older women undergoing breast cancer surgery).²⁵ Short-term decreases in cognitive function may follow the administration of general anesthesia,²⁶ and older women have greater arm morbidity after axillary lymph node dissections.²⁷

Although few oldest-old women receive chemotherapy for breast cancer (1% of oldest-old women compared with 30% of patients aged 55 to 64 years), increased toxicity and mortality are associated with chemotherapy among older women (1.5% chemotherapy- related mortality for breast cancer patients 65 years and older compared with 0.2% for patients 50 years and younger).²⁸ In addition, some older women may experience cognitive decline within 6 months after adjuvant chemotherapy.²⁹ Nearly all women experience fatigue during treatment with radiotherapy, and radiotherapy can also cause breast pain and edema even months after treatment.^{30, 31}  The side effects of tamoxifen (increased risk of thrombophlebitis, sleepiness) and aromatase inhibitors (arthritis, bone pain, fatigue), which are more commonly given to older women, can significantly impair their quality of life.³²

•Other:  Although false-positive mammograms are less frequent among older women, some patients may experience anxiety as a result and some may have complications from followup diagnostic procedures.⁴ 

OUTCOMES OF MAMMOGRAPHY SCREENING

To better understand outcomes of mammography screening among women 80 and older, I recently performed a cohort study in which I examined receipt of mammography screening among 2011 community-dwelling women 80 and older without a history of breast cancer who received their primary care at one large academic medical center or two community affiliates between 1994 and 2004.³³ After a median of 5 years of follow-up, I found that 51% (n = 1034) of the women were screened at least once after age 80 and 11% (110/1034) experienced a false-positive mammogram (19 of whom underwent breast biopsy).

Of the 143 women who had an abnormal screening mammogram, 17% experienced a true-positive test (DCIS or breast cancer was diagnosed). Almost all women who were told they had a normal screening mammogram had a true-negative mammogram; fewer than 1% experienced a false-negative. About 2% received a diagnosis of invasive breast cancer whether or not they were screened, and there was no difference in stage at diagnosis between the 2 groups. One woman who was screened died of breast cancer, while 2 women who were not screened died of the disease. One additional woman who was not screened had a breast cancer recurrence but did not die of breast cancer.

Overall, I found that fewer than 2% of women 80 and older potentially benefited from mammography screening (ie, early-stage breast cancer was diagnosed, and the patient received treatment and lived 2 years after diagnosis). However, since similar proportions of women who were not screened were also found to have early-stage disease and lived 2 years after diagnosis, it was unclear whether mammography screening was beneficial among women 80 and older. Nearly 13% of the women who were screened experienced burdens from screening (false-positive mammogram, detection of DCIS), and only 0.3% were potentially harmed from not being screened (diagnosed with late-stage disease); 2 of these women had probably never been screened.

POLICY AND GUIDELINES FOR SCREENING OLDER WOMEN

Mammography screening. Despite the uncertainty of benefit for the oldest-old and the potential costs of screening all women for a lifetime, the Medicare program instituted biennial coverage for screening mammography in 1991 and broadened coverage in 1998 to cover annual mammograms for all women over age 65 with no upper age limit.³⁴ Screening mammography has increased since Medicare first began covering the test. In 2000, 51% of the estimated 3.8 million community-dwelling women 80 and older in the United States had received screening mammography in the previous 2 years compared with about 25% in the early 1990s.^6,35

The current figure likely represents both overscreening and underscreening of older women. Among the 1.3 million oldest-old women with probable life expectancies of less than 5 years, 39% were screened even though they had little chance to benefit. On the other hand, among the 1.1 million oldest-old women with probable life expectancies of 10 or more years, 38% were not screened.⁶ Wealthy older women are more likely to be overscreened, while poor older women are at risk for not being screened.³⁶

Table 1 shows different mammography screening guidelines for older women. Most guidelines do not give an absolute age to stop screening older women but do encourage physicians to consider patient life expectancy.^3,37-40 The American Geriatrics Society (AGS) recommends that clinicians offer screening to women 85 and younger with at least 5 years' life expectancy after an individualized review of the potential benefits and risks of screening and a discussion of patient preferences.³⁷ The AGS also recommends that mammography be reserved for healthy women older than 85 who have excellent functional status or those who strongly believe that they will benefit through peace of mind or improved quality of life.

Several studies have examined the cost-effectiveness of screening older women for breast cancer with mammography; most have found that the positive effects of screening balance with the negative effects at about age 80.^41,42 However, Mandelblatt and colleagues⁴¹ concluded that screening women 80 and older in the top quartile of health for their age-group may be cost-effective under certain circumstances.

Clinical breast examination and breast self-examination. Two large randomized controlled trials that included women of all ages found no benefit of breast self-examination (BSE) when compared with no breast cancer screening. No trials have compared clinical breast examination (CBE) alone with no screening. The AGS recommends that CBEs be performed periodically among older women and neither endorses nor discourages BSE. The USPSTF finds insufficient evidence to recommend for or against performing CBE and recommends against teaching BSE.³ 

TOOLS FOR DISCUSSING SCREENING OF OLDER WOMEN

Tools for estimating life expectancy. Several tools are available to help you target mammography screening to older women by life expectancy. Lee and associates⁴³ created a prognostic index that includes 12 questions (eg, history of heart failure, difficulty in walking several blocks) which patients can answer during an office visit to help predict their risk of 4-year mortality (Box I and Table 2). I created a 9-question index that patients can answer to help predict their 5-year mortality (Box II and Table 3).⁴⁴ Women who score 14 or more points on this index are highly unlikely to benefit from mammography screening because they have approximately a 50% or greater chance of mortality in 5 years.

Walter and Covinsky⁴ also described a framework to help you with estimating patients' life expectancies. First you must estimate which quartile of health your patient is in for her age. After you decide on your patient's relative health, you can use the Figure to estimate her life expectancy.

Tips on talking to patients. Discussing the decision to stop screening with women who have been screened for years can be uncomfortable.⁷ However, just as it is important to learn how to encourage patients to be screened for cancer, it is also important to know how to discuss stopping screening with women who have shorter life expectancies.

Several factors influence the decision to continue screening, including habit and the need for reassurance.¹⁵ Therefore, when you talk with an older woman in poor health, you may want to acknowledge that you understand how hard it must be for her to stop receiving mammograms when she has been doing so for many years. You may also want to explain that just as there was a time to start screening, there is also a time to stop. Include in your discussion the risks of screening older women with comorbidities. To address patients' need for reassurance about their health, discuss the many preventive measures that you will now focus on that are likely to benefit your patient in a short time period (eg, exercise, screening for incontinence or depression).

You may need to initiate and reinitiate discussions about stopping screening during several visits until a patient feels informed enough to make a decision. A decision aid may also help. In Australia, Mathieu and associates⁴⁵ developed a decision aid to help women aged 70 decide whether to continue screening. It was found to increase older women's knowledge about mammography screening.