Factors Associated with Surgical Decision Making in Women with Early-Stage Breast Cancer: A Literature Review

Influence of age

Reports differ about the relationship between surgical treatment choices for early breast cancer treatment and the patient's age. Several factors—including differing age cutoffs among studies, year in which the studies were performed, and area (rural vs. urban) from which data were collected—are likely to contribute to these disparate findings.

Studies showing an association between younger age and increased mastectomy rate include a report by Adkisson et al., ⁷ in which 65% of women younger than 50 years opted for mastectomy versus 40% of women older than 70 years who chose mastectomy. An analysis of data from 1973 through 1995 from Surveillance Epidemiology and End Results (SEER) of 307,115 patients, including those from the San Antonio database, found that older patients underwent BCT more frequently (10% at age 55–74 years, 13% at age 75–84 years, and 24% at age ≥85 years) than younger patients. ⁸ When age was subdivided into less than 40 years, 40–60 years, and greater than 60 years, SEER data evaluation among 29,666 breast cancer patients showed that women aged 40 to 60 years were significantly (p<0.0001) more likely to undergo BCT than those younger than 40 years and those older than 60 years. ⁹

However, some studies have showed an increased mastectomy rate in older women. ^{10 –13} A study by Pappo et al. ¹⁰ involving 829 patients found that patients aged 70 years and older were significantly less likely to undergo breast-conserving surgery than patients younger than 70 years (36.1% vs. 59.4%; p=0.0001). Patients older than 60 years and those older than 80 years also showed a significant difference compared with younger patients. Staradub et al., ¹² in a 2002 report, reviewed their institution's database from 1995 through 1998 of all ductal cancer in situ and stage I and II breast cancer. They found that use of mastectomy was associated with older age; younger women were more likely to have had BCT (mean age, 59 vs. 54 years; p<0.001).

International experience mirrors these mixed results of the influence of age on mastectomy rate in U.S. populations. Several studies have showed that older age was associated with increased mastectomy rates. ^11,13,14 A study from the Western Australia Cancer Registry evaluating 2713 patients compared three age groups of women: younger than 40 years, 40–59 years, and 60 years and older. They found no differences in the two younger age groups; both chose BCT over mastectomy. The investigators found that women aged 60 years and older were more likely to choose mastectomy. ¹¹ Ridao-López et al. ¹⁴ studied more than 81,000 women in nearly 200 geographical regions of Spain in whom screening mammography detected breast cancer. The investigators found that age greater than 70 years was a factor associated with mastectomy. Wyld et al., ¹³ in a study from the United Kingdom, demonstrated that older women were more likely to have chosen a mastectomy than a different surgical treatment of breast cancer, despite their eligibility for conservative surgery.

Studies that did not show age as a factor in the choice of breast cancer surgery include those by Caldon et al. ¹⁵ and White et al. ¹⁶ In a study of more than 5000 women in Trent, United Kingdom, age was not found to be associated with increased mastectomy rates. ¹⁵ Similarly, White et al. ¹⁶ reported that in Australia, subsequent to the publication of breast cancer surgical treatment guidelines, more women were likely to undergo conservative therapy regardless of other factors, including age.

On the basis of all the interactions, including distance from radiation therapy sites, testing for BRCA gene, breast imaging, and decision aid use, age alone does not appear to be a strong predictor for choice of surgery. Because the various studies used different age groups, it is difficult to draw firm conclusions about the role age has in determining the surgical treatment used for breast cancer. However, the United States–based studies show that in younger women (age <50 years), mastectomy was favored over BCT.

Influence of ethnicity and race

African American women have been reported as being more likely to undergo BCT than white women (odds ratio [OR], 1.62; 95% confidence interval [CI], 1.11–2.37). ¹⁷ In contrast, Hispanic women were 38% less likely to receive BCT than white non-Hispanic women (OR, 0.62; 95% CI, 0.55–0.71). ¹⁸ Among both highly and less highly acculturated Hispanic women, African American women, and white women, patient involvement in decision making was associated with an increased use of mastectomy. Patients identified fear of breast cancer recurrence and adverse effects of or limited access to radiation therapy, as well as body image, as factors in their decision making. Input from family and friends, particularly among the less acculturated Hispanic women, was an important part of the decision-making process. ¹⁹ Similarly, the use of BCT among Asian/Pacific Islander women has been shown to be less than in white women (47% vs. 59%). The differences were not related to health status, demographic characteristics, or socioeconomic status in this ethnic group and were thought to be related, in part, to poor acculturation, limited access to medical care, communication difficulties, and cultural preferences or differences. ^{20 –22}

Influence of socioeconomic, education, and marital status

Socioeconomic status is thought to influence the decision making about receiving breast cancer surgery through its influence on both the patient's emotional responses to the cancer diagnosis and the physician–patient relationship. In a Nebraska study, low-income women were believed to base their decisions on the same criteria as higher-income women. ²³ Decision making was found to be based, in part, on a “paternalistic” pattern of interaction on the part of the surgeon or on poor comprehension of the treatment options that were available. Patients described feeling overwhelmed and pressured to make a decision in the face of what was felt to be a “life-threatening” illness. The investigators raised concern that higher-income women felt more empowered to question their physician regarding treatment choices than the other women.

Private insurance has been shown to be associated with BCT use. Patients with breast cancer who have private insurance have been reported to more likely undergo BCT than mastectomy than those with no insurance (OR, 3.90; 95% CI, 1.20–12.67) ²⁴ or with government insurance (OR, 1.49; 95% CI, 1.20–1.86). ¹⁸

In one study, marital status was noted to be associated with surgical decision making because married women were 23% more likely to receive BCT than single women (OR, 1.23; 95% CI, 1.09–1.40). ¹⁸

Socioeconomic status and education level are found to influence the decision regarding mastectomy and BCT in other countries as well. A large, descriptive study in Spain between 2002 and 2006 found that women from regions of high economic and education levels were more likely to have conservative surgery than women from regions with lower economic and education levels. ¹⁴

Influence of geographic location and access to radiation therapy facility

Several factors can influence the association between the patient's geographic location and the choice of breast surgery, but access to radiation therapy facilities is probably the most important factor influencing this association.

Voti et al. ¹⁸ examined the effect of distance to a radiation therapy facility on the likelihood of receiving BCT in 18,903 breast cancer patients in Florida between 1997 and 2000. The investigators showed that the distance to the closest radiation therapy facility was negatively associated with BCT, with the OR decreasing by 3% per 5-mile increase in distance.

Similar results were seen in a study of patients with early-stage breast cancer diagnosed from 1996 to 2000, identified through the Virginia state registry, which evaluated a subgroup of 11,597 patients. ²⁵ Mastectomy rate varied by distance to a radiotherapy center (31% at ≤10 miles, 36% at >10 to 25 miles, 41% at >25 to 50 miles, and 49% at >50 miles; p<0.001). Nattinger et al. ²⁶ evaluated 21,135 women who received a diagnosis of breast cancer in 1991 and 1992 who were registered in SEER. They reported a significant decrease in BCT for those patients who lived more than 15 miles from a radiation treatment center (OR, 0.52; 95% CI, 0.46–0.58) and in postlumpectomy radiation therapy for the women who lived more than 40 miles from the nearest radiation therapy center (OR, 0.55; 95% CI, 0.37–0.82).

A study analyzing all cases of localized breast cancer diagnosed in New Mexico in 1994 and 1995 showed that of 1122 women with localized breast cancer, 533 (48%) received BCT and 409 (77%) of these patients had radiotherapy after BCT. ²⁷ The authors reported that travel distance to the nearest radiation center was an important variable as to whether or not the women received radiation treatment after BCT. Travel distances were noted as 75 miles or more, 50 to 74.9 miles, and less than 50 miles, with radiation rates of 51%, 69%, and 82%, respectively (p<0.0001). However, there was no difference in the mastectomy vs. BCT rate by travel distance.

SEER data from Hawaii, which used the Hawaii Tumor Registry, identified trends in decision making regarding BCT. ²⁸ The investigators concluded that the primary determinant in decision making was location; the women living on Oahu were 70% more likely to undergo BCT than the women living on the other islands. The only other islands with radiation therapy facilities were Hawaii and Maui, requiring patients from the other islands to either fly in daily for treatments or move temporarily to another island.

Similar findings have been reported from studies in Canada ²⁹ and Western Australia. ³⁰ In contrast, a large study across 180 regions in Spain found a large geographic variation in mastectomy rates, with no identifiable impact of availability to radiation therapy. ¹⁴

Celaya and colleagues ³¹ showed that the time of year in which the radiation therapy following BCT was to take place also influenced women's choice of treatment. Although women chose BCT more often than mastectomy for early-stage breast cancer (66% vs. 34%), those who did choose BCT were less likely to undergo radiation treatment in winter (December through February) compared with the nonwinter months (March through November).

It appears that geographic location, which in reality means proximity to an irradiation center, does have a large role in the choice of surgery. Women take into account their age, closeness to an irradiation center, and possible winter weather conditions before making the choice between BCT and mastectomy. As irradiation techniques change and with the advent of shorter regimens, we may see future changes in surgical choices.

Influence of BRCA gene mutation status

BRCA testing, introduced in 1996, has given women the opportunity to evaluate their genetic breast cancer risk and has had an impact on surgical decision making. ³² The decision regarding what is optimal local therapy for women with BRCA mutations has been controversial, and a study by Robson et al. ³³ suggested that women with BRCA-associated breast carcinoma who undergo BCT appear to have risks of metachronous ipsilateral breast carcinoma that are similar to those reported for young women. BRCA mutation carriers have been generally advised to undergo mastectomy because of an increased risk of ipsilateral breast cancer recurrences. ^33,34 However, Pierce and colleagues ³⁵ reviewed 655 women with BRCA1/2 mutation who had received a diagnosis of breast cancer and found no differences in survival, regional recurrence, or systemic recurrence among the women who had BCT and irradiation versus those who had mastectomy. They did find, however, a local recurrence rate of 23.5% vs. 5.5% (p<0.0001) at 15 years (BCT vs. mastectomy), which they believed to be a second breast cancer rather than a recurrence of the original cancer. They also showed that chemotherapy reduced the risk of an ipsilateral breast cancer in the BCT group.

Contralateral prophylactic mastectomy (CPM) has been discussed with patients who have early-stage breast cancer and BRCA mutation as another treatment option because of increased risk of contralateral breast cancer. ³⁴ Mislowsky et al. ³⁶ showed a significant decrease in unilateral mastectomy and an increase in bilateral mastectomy after the period of 1996 to 2000 in women who received a diagnosis of unilateral breast cancer and were carriers of BRCA1 or BRCA2. The investigators also showed that CPM was the choice favored by most women whose initial decision had been lumpectomy before the BRCA test result was known (86% chose CPM vs. 14% who did not). In a review of 2504 women who had a diagnosis of unilateral stage 0–III breast cancer, Yi et al. ³⁷ showed that testing for BRCA1/2 genetic mutation, even when the result was ultimately negative, was associated with CPM.

Role of reconstruction

Reconstruction following mastectomy has been shown to improve quality of life. ³⁸ Arrington et al. ³⁹ showed that the immediate reconstruction rate was 62.4% in women who underwent a CPM but was only 33.9% in women who had a unilateral mastectomy. In studies in which women were contemplating bilateral prophylactic mastectomy because of their increased risk of breast cancer, the availability of breast reconstruction had a large part in the decision making to go forward with that plan. Nelson and colleagues ⁴⁰ showed that 93% of the high-risk group of women said the reconstruction options made them more willing to consider bilateral prophylactic mastectomy. These patients also reported that their surgeon's opinions were also a strong influence. Although bilateral prophylactic mastectomy is undertaken to prevent BC, the principles are similar for women undergoing bilateral mastectomy for a unilateral BC. Increasingly, women view CPM as an option if immediate reconstruction is available.

Influence of MRI

Performing a staging MRI has been shown to increase the rate of mastectomy over BCT. An observational study done between 1997 and 2006 showed that 52% of women who had a staging MRI before surgery had a mastectomy compared with 38% of those who had not had such imaging. ⁴¹ All of the women had early-stage breast cancer. Another study had similar results in comparing the periods of 1998 through 2000 and 2003 through 2005. ⁴² Over these time spans, the MRI use, mastectomy rate, and CPM rate increased for stage I and II breast cancers.

A 2011 review of 710 women with breast cancer who had a preoperative MRI reported the detection of an additional abnormal finding in 48% of the women. ⁷ When the result of a preoperative biopsy of the abnormal MRI finding was benign, the mastectomy rate was similar to that of the women who had an MRI with no additional findings (41% vs. 37%). Among the women, 46% declined evaluation of the abnormal MRI finding, resulting in a much higher mastectomy rate (83%) in this group. Younger women were more likely to decline biopsy of the MRI findings than older women.

Thus, it is important to discuss with women the role of a staging MRI before undergoing this test. Advising her, in advance, of the possible need for further biopsies may reduce her anxiety and disappointment. It may allow her to take the investigational steps needed, rather than proceed with an immediate bilateral mastectomy.

Role of the surgeon in decision making

Multiple studies have showed the influence of the surgeon's recommendation on a patient's decision regarding surgical intervention for breast cancer. A survey of women with early-stage breast cancer (stage 0–II) showed that of 1984 women, 23% underwent mastectomy, with 13% reporting that their decision was based on the surgeon's recommendation. ⁴³

In a 2005 study reported from China, only 49% of 443 women indicated that they were offered a choice of surgery. ⁴⁴ Of 43 women whose surgeon advised BCT, 79% chose BCT. Of 96 women whose surgeons did not make a specific recommendation, 37% chose BCT. Women who chose mastectomy did so on the basis of the risk of recurrence and need for further treatment. However, women who chose mastectomy and reconstruction and those who preferred BCT placed emphasis on appearance and body image.

Surgeons, their practice type, and hospital factors can all influence the surgical decision. A survey of 1000 breast cancer surgeons in the United States conducted during 1997 and 1998 showed that surgeons who practiced in areas of high Medicare BCT fees and those with the strongest beliefs in patient participation in treatment decisions were more likely to perform BCT. ⁴⁵ Studies have showed mixed results for how the sex of the treating surgeons influenced decisions. Although one study showed that male surgeons were more likely to provide BCT than their female colleagues, ⁴⁶ another study contradicted this result, showing that male surgeons were more likely than female surgeons to do mastectomy. ⁴⁵ Arrington et al. ³⁹ showed that CPM rates were higher for female surgeons than for younger male surgeons (age <50 years). In a study of analysis of variance in BCT use in the United States between 1988 and 1994, Jerome-D'Emilia and Begun ⁴⁷ reported that the rate of BCT use was highest in academic teaching hospitals and lowest in community hospitals. Hawley et al. ⁴⁸ evaluated SEER data for BC patients from December 2001 through January 2003 and found that high breast surgery volume was associated with a higher BCT rate. They found that almost 10% of the variation in the surgical procedure was due to surgeon preferences.

Role of decision aids and shared decision making

Decision aids have been incorporated by physicians into the discussion of early breast cancer treatment options, to both help educate women with breast cancer and help clarify the options available to them. One would surmise that the knowledge of equivalent survival between BCT and mastectomy as two surgical options would lead to a higher rate of BCT than of mastectomy. However, studies have showed mixed results.

Molenaar et al. ⁴⁹ evaluated the introduction of an interactive CD-ROM as a decision aid for patients with early-stage breast cancer, with patients with breast cancer who received usual care as the control group. No effect was found on patient treatment decisions when patients used the CD-ROM. In contrast, Whelan et al., ⁵⁰ in their study evaluating the introduction of a decision aid for the different surgical treatment options in Ontario, found that women using the decision aid had greater knowledge about the treatment options, had less decisional conflict, were more satisfied with the resulting treatment, and were more likely to choose BCT than mastectomy (94% vs. 76% for the latter; p=0.03).

Patients' perception of their involvement in the surgical decision, beyond their knowledge of equivalent surgical outcomes, appears to affect both their surgical decision and their long-term quality of life. ^{51 –55} Investigators have found that matching the desired degree of decision making with the actual degree of patient involvement in decision making is important to the decision made. Having less involvement than desired and more involvement than desired both increased the risk of low patient satisfaction. ^{48,55 –57} In another report, Katz et al. ⁵⁸ identified 183 women, through 1998 SEER data from Detroit, who had surgery for their breast cancer. Knowledge of equivalence of survival between BCT and mastectomy was associated with a higher lumpectomy rate (OR, 2.1; 95% CI, 1.1–4.01). The investigators reported that satisfaction with the decision making was less in the women who did not perceive they had a choice, suggesting that the decision-making process was as important as the choice of surgery.

In addition, Katz et al. ⁵⁹ demonstrated that greater patient involvement was associated with greater mastectomy rates, on the basis of their analysis of 2002 SEER data from Detroit and Los Angeles of women with early-stage breast cancer. Among white women who reported that the surgeon had made the surgical decision, 5.3% received a mastectomy. Among white women, those who reported the decision was shared had a 16.8% mastectomy rate, whereas those who said they had made the decision by themselves had a 27.0% mastectomy rate (p<0.001). In contrast, a study of 175 women in Western Australia who were treated for 6 months between 1996 and 1997 found that the women who preferred to take a more active role in decision making, among other factors, were more likely to choose BCT. ³⁰

A survey from 2005 through 2007 showed no difference among the knowledge scores of the women who chose mastectomy and those who chose BCT. ⁶⁰ Despite their knowledge of the similar outcomes of BCT versus mastectomy, 35% of the women chose to have a mastectomy. Those who opted for mastectomy rated “peace of mind” and “avoiding radiation” higher than those who chose BCT. Hence, although patient knowledge of outcomes of the two surgical options is important in facilitating the shared decision-making process, a woman's own values and preferences clearly contribute to her ultimate choice for type of breast surgery.