Etiology and Risk Factors




Clinical features of breast masses that favor benign etiology: firm rather than hard texture; well-circumscribed borders, mobility

Fibroadenomata--benign lesions; risk nor related to age of menarche or menopause or use of HRT; risk decreased in obese women, muliparous women, and smokers. Most common in the upper-outer quadrant. May regress over time.

I. Epidemiology of breast Ca

  1. Lifetime risk 1:8
  2. Annual incidence 2:1000
  3. 200,000 dx's, 46,000 deaths/yrs in U.S.
  4. Upper outer quadrant--last part to be replaced by fat, so most frequent site for Ca in older women

II. Etiology & risk factors for breast Ca

  1. Incidence increases with age after 40yo; this is by far the most sig. risk factor
  2. Family hx (most significant is for premenopausal breast Ca)
  3. Genetic factors
    1. Mutations of BRCA1 and BRCA2-See also Ovarian Neoplasms
      1. Tumor-suppression genes
      2. Mutations that inactivate these genes are ass'd with elevated risks for malignancies of breast, ovaries, pancreas; colon, cervix and uterus (BRCA1 only); and stomach, biliary tract/gallbladder, and melanoma (BRCA2 only).  Mutations are most prevalent among Ashkenazi Jews
      3. Harmful mutations in BRCA1 and BRCA2 are associated with a 35-84% risk of breast Ca by age 70y in women.
      4. See below re: screening regimen in women who are known to have BRCA1 or BRCA2 mutations
      5. Recommendations for BRCA mutation testing in women with a known history of breast Ca (per National Comprehensive Cancer Network, 2010):
        1. Diagnosis at <45 yo
        2. Diagnosis <50 yo with >1 1st-, 2nd- or 3rd-degree relative with breast cancer at <50 years of age
        3. Two breast primary tumors with first diagnosis at <50 years of age
        4. >21st-, 2nd- or 3rd-degree relatives with breast and/or epithelial ovarian/fallopian tube/primary peritoneal cancer
        5. 1st-, 2nd- or 3rd-degree male relative with breast cancer
        6. Personal history of epithelial ovarian/fallopian tube/primary peritoneal cancer
        7. Ethnicity associated with higher mutation frequency (e.g., Ashkenazi Jewish)
  4. Hormonal stimuli of breast cells: estrogens, progestins, prolactin, insulin, thyroid hormones
    1. In a prospective study of 97 postmenopausal women w/breast Ca & 244 controls, the women with breast Ca had sig. higher levels of all estrogens & androgens tested, including estradiol, etrone, androstenedione, dehydroepiandrosterone, and testosterone (Ann. Int. Med. 130:270, 1999--JW)
  1. Hormone Replacement Therapy--FAR FROM CLEAR that there is an association (see the HRT section for details)
  2. Previous breast Ca (annual risk 1.4%)
  3. Parity: RR is 1.6 if nulliparous or older than 30 for first term pregnancy; RR is 0.8 if 21yo or younger at first full term pregnancy
  4. Menarche: RR is 1.2-1.5 for menarche before age 12
  5. Menopause: RR is 2.0 for menopause after age 55
  6. Lactation: small protective effect if breastfeed for >4 mos at young age
  7. Oral Contraceptives
  8. Alcohol use
    1. RR of breast Ca 1.41 among women who drink 2-5 drinks/d c/w nondrinkers (JAMA 279:535, 1998--JW)
    2. The excess breast Ca risk due to high EtOH intake may be mitigated to some degree by dietary folate content (JAMA 281:1632, 1999--JW)
  1. Certain benign breast disease
  1. Slight increased risk with moderate to severe hyperplasia, papilloma, gross cysts with fam. h/o breast Ca, complex fibroadenomas, sclerosing adenosis
  2. Atypical hyperplasia (RR = 4.0-5.0) (NEJM 353:229, 2005--JW)
  1. Nutrition
  1. Dietary fat--unclear; in a 14y Nurse's Health Study analysis of 88,795 women there was no association between breast Ca incidence and intake of total fat, animal fat, vegetable fat, polyunsaturated fat, saturated fat, etc. etc. (JAMA 281:914, 1999--JW)
  2. Decreased vit. A associated with increased risk
  1. Body size
  1. Overweight: increased risk of postmenopausal Ca
  2. Lean: increased risk of pre-menopausal Ca
  3. Tall: increased risk
  1. Aspirin and other NSAIDs-May be protective; click on link for details

III. Screening for breast cancer

  1. Mammography
  1. Purported to find Ca at earlier stage, when it's less likely to be invasive (60% invasive by mammography vs. 91% by exam alone in Mass Gen'l study going from '78-'91; with better survival in mammography group)
  2. False-positive mammo's are more common in younger women, because their breast tissue tends to be more radiodense (more glands, less fat)
  3. Evidence of clinical benefit
    1. In individual randomized trials, reduces mortality form breast Ca by around 30% in some randomized studies
    2. See the section on Mammography in women 40-49yo for a list of the individual randomized trials
    3. A very controversial meta-analysis of the 8 randomized trials of mammography screening to date called benefit of screening mammography into question for women of any age:
Gotzsche P, Olsen O. Is screening for breast cancer with mammography justifiable? Lancet 2000; 355: 129-134.

This study was performed by the Scandinavian Cochrane group, though it was NOT an "official" Cochrane review. They reviewed the issue of whether the control vs. intervention (screened) groups had similar characteristics in terms of known risk factors for breast cancer. They concluded that 6 of the 8 RCTs of breast cancer screening had randomization processes that failed to produce equivalence in the control vs. intervention groups. Differences included characteristics like age and history of prior breast lumps that might affect risk of breast cancer mortality.  They also note that "Breast cancer mortality was unreliable and biased in favour of screening."  The two trials that were judged to have used adequate randomization were the Canadian and Malmo trials. These studies individually showed no effect of screening on breast cancer mortality.

They also performed a meta-analysis of the eight RCTs. For mortality from breast-cancer, they calculated a RR of 1.04 using only the two "unbiased" trials (95% CI 0.84-1.27) but 0.75 (95% CI 0.67-0.83) among the six "biased" trials. For all-cause mortality they calculated a RR of 1.06 (95% CI 1.04-1.08) among screened patients, BUT after adjusting for age (which was higher in the screened groups), RR of 1.00. In an analysis of all-cause mortality in the two "unbiased" trials, the RR was 0.99 (95% CI 0.93-1.05) among screened patients. This study did NOT include any subgroup analysis of women 40-49yo.  This paper's exclusion of the six "biased" trials was heavily criticized in letters-to-the-editor by various other reseachers.

Note--The same authors published results of this analysis as a formal Cochrane Review in 2001, which gave very similar #'s for the same comparisons.  It also included more information on age-specific subgroups; for women < 50yo at randomization, breast Ca mortality RR 1.03 (95% CI 0.77-1.38) in the two good trials and 0.89 (95% CI 0.72-1.10) in all trials; form women > 50yo at randomization, breast Ca mortality RR 0.94 (95% CI 0.77-1.15) in the two good trials and 0.76 (95% CI 0.66-0.86) in all trials  (Olsen O, G°tzsche PC. Screening for breast cancer with mammography (Cochrane Review). In: The Cochrane Library, Issue 4, 2001. Oxford: Update Software)

  1. In a study comparing incidence of breast Ca in counties in Norway where mammography screening was introduced vs. others where it was not introduced during the same time period, risk reduction attributable to screening was 2.4 deaths per 100,000 population over a ten-year period.  
  2. In a population study in Sweden, mammography screening was associated with a sig. reduction in breast Ca mortality (RR 0.71) (Cancer 9/29/2010 e-pub ahead of printing at:
  3. Mammography in women 40-49yo
    1. Overall risk reduction achieved with mammography may be small in 40-50yo women; many unnecessary biopsies will be performed
    2. Click HERE for an extensive summary of the data
  4. Digital mammography
    1. Digital mammography--Digital (i.e. filmless) capture of the image allows 
      manipulation of contrast, etc. May increase sensitivity c/w conventional 
      film-based mammography, but the latter may still find some tumors not found on the former (NEJM 353:1773, 2005--JW)
  1. Breast self-exam
    1. In a 10-year randomized trial on about 260,000 Chinese women (to whom mammography and clinical exam were mostly unavailable), detailed breast self-exam instruction was not associated with any significant difference in breast cancer mortality (J. Nat. Ca. Inst. 94:1445, 2002--JW)
  2. Ductal lavage
    1. A minimally invasive technique for obtaining breast epithelial cells for cytologic analysis and examination for molecular markers associated with cancer
    2. Suction is applied then individual ducts which yield fluid with suction are cannulated and lavaged with saline to obtain epithelial cells for cytologic analysis. More likely than nipple aspiration alone to yield adequate cellular material for cytologic analysis; may also be a target for examination for molecular markers associated with Ca. 
    3. Shown to demonstrate cytologic atypia in approximately 20 percent of women at high risk for breast cancer who have normal physical exams of the breast and benign-appearing mammograms.  Cytologic atypia in breast tissue obtained by other methods (e.g. nipple aspirate or fine needle aspiration) has been shown to be associated with an increased risk for eventual development of breast cancer. It has been proposed that a finding of cytologic atypia on a ductal lavage specimen could assist in the decision of whether to use tamoxifen for primary prevention of breast cancer (Cancer Control 2002;9:473-9).  However, as of 2004, the accuracy of ductal lavage in diagnosing breast cancer or predicting subsequent development of breast cancer remains unknown. It is also unknown whether using ductal lavage findings to guide further diagnostic or therapeutic interventions will lead to improved clinical outcomes (Breast Cencer Res 2002;4:51-53.).
    4. Sensitivity of ductal lavage for Ca was only 13% in one study of 32 of women about to undergo mastectomy for biopsy-proven Ca (J. Nat. Ca. Inst. 96:1510, 2004--AFP)
  3. Ultrasound-A poor screening test but may be a useful adjunct to mammography in women with dense breast tissue
  4. Contrast-enhanced MRI as an adjunct to mammography for screening for breast Ca:
    1. In a cohort study of 1,909 women 25-70yo with (strong family h/o breast Ca or positive screening for genes associated with breast Ca risk), MRI and mammography each identified cancers missed by the other.  Overall, MRI was more sensitive but less specific than mammography (NEJM 351:427, 2004--JW)
    2. In a study of 821 women referred for breast bx for mammographic abnormalities or clinically suspicious breast exam findings, all of whom underwent MRI before biopsy, MRI had sensitivity/specificity of 88%/67% (JAMA 292:2735, 2004--JW)
    3. In a study in 969 women with recently diagnosed breast Ca but no mammographic or clinical abnormalities in the controlateral breast, 3% had contralateral breast Ca detected on MRI. Positive predictive value was 21%.  Sensitivity for dx of contralateral breast Ca over the next year was 91% (NEJM 356:1295, 2007--JW)
    4. American Cancer Society in 2007 recommended that women meeting the following criteria get it annually along with mammography from age 30yo (CA Cancer J. Clin. 57:75, 2007--JW):
      • BRCA mutation or first-degree relative with BRCA mutation (based on non-randomized screening studies)
      • Lifetime breast Ca risk > 19% according to predictive models (based on non-randomized screening studies)
      • History of receiving chest radiation between ages 10-30yo (based on expert opinion)
    1. Screening and other management in women with harmful BRCA1 or BRCA2 mutations-Recommendations per National Comprehensive Care Network 2010:
      1. Breast self-exam monthly from 18yo
      2. Clinical breast exam Q6mos from 25yo
      3. Mammogram AND MRI Q1y from 25yo
      4. Discuss option of prophylactic mastectomy
      5. Recommend prophylactic salpingo-oophorectomy, ideally from 35-40yo and upon completion of childbearing-If declines, ttransvaginal ultrasound + CA-125 Q6mos from 35yo or 5-10 years before than the earliest age of first diagnosis of ovarian cancer in the family
      6. Consider chemoprevention options (tamoxifen, raloxifene, etc.) for breast and ovarian cancer

IV. Prevention of Breast Cancer

  1. Tamoxifen for primary prevention of breast Ca
  1. National Surgical Adjuvant Breast and Bowel Project (NSABP-P1)
    1. 4y study showing decrease in incidence of breast Ca in women at increased risk
  1. The Breast Cancer Prevention Trial (J. NCI 90:1371, 1998)
    1. 13,388 women either >60yo or 35-60yo but with risk factors for breast Ca or 35-60yo with h/o lobular CIS were randomized to tamoxifen 20mg/d vs. placebo
      1. Risk assessment based on algorithm that took into account age, # of 1st-degree relatives with breast Ca; nulliparity or age at first live birth, # of breast biopsies, dx of atypical hyperplasia, and age at menarche (based on J. NCI. 81:1879, 1989 which gave RR's based on these factors and also on absolute risk data from the 1984-1988 Surveillance, Epidemiology, and End Results program). Criteria was est'd 5y risk of > 1.66% for invasive breast Ca
    2. Study stopped early (avg 69mos f/u) b/c of sig. reduction in invasive breast Ca in tamoxifen group (RR 0.51)
      1. Similar risk reduction in all age groups
      2. Also reduced risk of noninvasive breast Ca by 50% (sig.)
      3. All the reduction in risk was for estrogen receptor-positive Ca; no reduction in risk of estrogen receptor-negative Ca
    3. Sig. increase in risk of endometrial Ca (RR 2.53; all in Tamoxifen group were stage I) & pulmonary embolism (RR 3.01) with tamoxifen; nonsig. increase in risk of CVA w/tamoxifen
    4. Nonsignificant reduction in death in Tamoxifen group (RR 0.81, 95% CI 0.56-1.16)
    5. In a follow-up report on the entire cohort (who, after the trial was halted, were all offered the opportunity to complete a 5y course of tamoxifen), after mean total f/u of 6y, the group originally randomized to tamoxifen still had sig. lower incidence of invasive breast Ca (.36%/yr vs. .63%/yr) and fractures (.20%/yr vs. .29%/yr), but had sig.higher incidence of uterine Ca (.22%/yr vs. .07%/yr) and pulmonary embolus (.07%/yr vs. .03%/yr) (J. Nat. Ca. Inst. 97;1652, 2005--JW)
  1. Royal Marsden Hospital Tamoxifen Randomised Chemoprevention trial (Lancet 352:98, 1998--AFP/JW)
  1. 2471 women 30-70yo (62% < 50yo) with family h/o breast Ca (summary didn't provide specifics) randomized to tamoxifen 20mg QD vs. placebo and followed median 70mos
  2. No sig. diff. in incidence of breast Ca in 2 groups (34 w/tamox vs. 36 w/placebo); 4 cases of endometrial Ca w/tamox vs. 1 w/placebo
  3. In a follow-up report on the same cohort, after mean 13y f/u, there was no sig. diff. in all-cause mortality or breast Ca incidence. (J. Nat. Ca. Inst. 99:283, 2007--JW)
  1. Lancet 352:93, 1998--JW
    1. 5,408 women (40% < 50yo) s/p hysterectomy randomized to tamoxifen 20mg/d vs. placebo. Over avg. 4y f/u, no sig. diff. in incidence of breast Ca
  1. International Breast Cancer Intervention Study ("IBIS-I")
    1. 7,145 women 35-70 at high risk for breast Ca randomized to Tamoxifen 20mg QD vs. placebo x 5y.  Over median 4.2y f/u, incidence of breast Ca was sig. lower in tamoxifen group (RR 0.68) but incidence of thromboembolic events and all-cause mortality were both sig. higher (RR 2.5 and 2.3, respectively); incidence of endometrial Ca was nonsig. higher in tamoxifen group (Lancet 360:817, 2002--JW)
    2. In a follow-up report on the IBIS-I participants, after mean 3y s/p completion of treatment, the tamoxifen group still had sig. reduction in breast Ca incidence (4.0% vs. 5.5%); incidence of thromboembolic events was also still elevated in the tamoxifen group (3.3% vs. 1.9%); no sig. diff. in all-cause mortality (J. Nat. Ca. Inst. 99:272, 2007--JW)
  1. Raloxifene for primary prevention of breast Ca
    1. 10,575 healthy postmenopausal women randomized to Raloxifene vs. placebo; over median 3.3y f/u, RR of breast Ca was 0.45 for Raloxifene group (results presented at an unspecified meeting--UW Pharm. Letter)
    2. 7705 postemenopausal women < 81yo (mean 66.5yo) with osteoporosis (defined as vertebral fx or femoral neck or spine T-score < 2.5SD below mean for young healthy women) and no current use of estrogen and no h/o breast Ca randomized to Raloxifene 120mg/d, Raloxifene 60mg/d, or placebo. RR of breast Ca over median 40mo f/u was 0.22 for the 60mg/d and 0.26 for the 120mg/d groups (both sig.)--the risk reduction was mostly for estrogen-receptor-positive breast Ca's. RR of thromboembolic disease was about 3.0 in Raloxifene group; no diff. in incidence of endometrial Ca. There was no sig. diff. in overall mortality rate. ("MORE" Study, JAMA 281:2189, 1999--note to myself; the discussion section of this paper has a good review of studies on hormonal methods for primary prevention of breast Ca which, however, I didn't read in its entirety)
    3. An analysis of the "MORE" Study data looking at subgroups by baseline estradiol level showed that women with baseline estradiol levels > 10pmol/L had RR of breast Ca of 0.24% with Raloxifene vs. placebo; women with undetectable estradiol levels had no sig. diff in breast Ca incidence between Raloxifene and placebo.  No data presented in this paper on overall mortality (JAMA 287:216, 2002--JW)
    4. In an extension of the MORE study in which about 5,000 pts continued raloxifene vs. placebo for mean 4y past the original end date of the study, the incidence of invasive breast Ca was sig. lower in raloxifene recipients (2 vs. 5 cases per 1,000 woman-years); raloxifene recipients had sig. higher incidence of pulmonary embolus (0.62% vs. 0.16%) ; no sig. diff. in mortality ("CORE" Study, J. Nat. Ca. Inst. 96:1751, 2004--JW)
  1. Tamoxifen vs. Ralofixene for prevention of breast cancer
    1. In a study in 19, 747 postmenopausal women with high breast Ca risk randomized to tamoxifen 20mg/d vs. raloxifene 60mg/d x 5y; at 5y, there was no sig. diff. in incidence of invasive breast Ca, noninvasive breast Ca, uterine Ca, other invasive Ca, coronary events, CVA, osteoporotic fractures, or overall mortality, but raloxifene recipients had sig. lower incidence of thromboembolic events (RR 0.70) and cataracts (RR 0.79). (NSABP Study of Tamoxifen and Raloxifene ("STAR-P2") Trial, JAMA 295:2727, 2006--abst)
  2. Aromatase Inhibitors for prevention of breast cancer
    1. In a study in 4,560 women with at least one of (age > 60yo, estimated 5y breast Ca risk > 1.66%, or prior intraepithelial neoplasia) randomized to exemestane vs. placebo, over median 3y f/u, the incidence of invasive breast Ca was sig. lower in the exemestane group (0.19% vs. 0.55%); exemestane recipients had more joint pain and hot flashes; no sig. diff. in incidence of fracture or adverse cardiovascular events (NEJM 364, 2381, 2011-JW)


  3. Prophylactic bilateral mastectomy--reduces risk > 90% from predicted in one retrospective study of 1,065 pts with median f/u 14y (NEJM 340:77, 1999)
  1. Criteria used to define "high-risk" in this study: family h/o breast Ca meeting one of the following patterns
    1. 2 or more 1st-degree relatives
    2. 1 first-degree and 2 or more second- or third-degree relatives
    3. 1 first-degree with onset < 45yo and one other relative
    4. 1 first-degree and 1 or more relatives with ovarian Ca
    5. 2 second- or third-degree relatives and one or more with ovarian Ca
    6. 1 second- or third-degree relative and 2 or more with ovarian Ca
    7. 3 or more second- or thrid-degree relatives with breast Cas
    8. One first-degree relative with bilateral breast Ca
  1. Aspirin and other NSAIDs for prevention of breast Ca--Click HERE for details
  2. Statins for prevention of breast Ca--Click HERE for details
  3. Dietary soy intake for prevention of breast Ca
    1. Soy is high in isoflavones which have both estrogen-like and antiestrogenic actions
    2. For secondary prevention: In a study using data from the Shanghai Breast Cancer Survival Study, a longitudinal cohort study, in 5,042 women 20-75yo with a h/o breast Ca (usually within 6mos of study enrollment), over median 3.9y f/u, women in highest quartile of sy intake had sig. lower recurrence rates (8.0% vs. 11.2%) and mortality (7.4% vs. 10.3%) compared with women in the lowest quartile; a dose-response relationship was sobserved up to a soy protein intake of 11g/d; there was no differerence in the degree of benefit based on estrogen receptor status or menopausal status (JAMA 302:2437, 2009-AFP)

V. Treatment of Breast Cancer

  1. Treatment determined by staging
  1. Ductal Carcinoma in Situ (DCIS)
    1. = Malignant transformation of epithelial cells in ductolobular system w/o invasion through the basement membrane.
    2. As many as 60% will fail to become invasive without treatment. 8y mortality rate was 1.4% in one series (BMJ 317:734, 1998--AFP)
    3. Modified radical mastectomy is often performed (because usually multicentric), though more recently has been treated with lumpectomy
    4. Radiation
      1. Reduces the recurrence of CIS though in studies as of 1998, no proven survival advantage to adding radiation to lumpectomy for CIS (J. Clin. Oncol. 16:441, 1998--JW)
      2. Benefit of radiation in reducing recurrence may be limited to cases where the surgical margins were < 10mm (NEJM 340:1455, 1999--JW)
  1. Tumor characteristics predicting local recurrence: estrogen receptor studies (better prognosis if pos.), ploidy (worse if aneuploid), doubling time/mitotic rate (faster is worse), catheptin-D receptor (bad if pos.)
  1. Surgery
  1. Halstead's Radical Mastectomy: complete removal of breast, axillary LN, and pec maj & minor mm. Rare today
  2. Modified Radical Mastectomy: removal of breast, axillary LN, poss. pec. minor. Similar outcomes as with Halstead's and better cosmetically
  3. Segmental Resection (aka "Wide Local Excision," "Partial Mastectomy")
    1. Local excision with healthy margin
    2. Often performed with axillary dissection & post-op radiotherapy
  4. In a trial in women with breast Ca & clinically negative axillary nodes randomized to radical mastectomy, total mastectomy w/XRT, or total mastectomy alone, 25y outcomes  were similar in all groups; ditto for women with clinically positive axillary nodes randomized to radical mastectomy vs. total mastectomy + radiation (NEJM 347:567, 2002--JW)
  5. Axillary lymph node dissection
    1. Surgical removal of the axillary nodes on the side of a breast cancer.
    2. Goal is to remove cancer that has spread to the axillary nodes before it metastases further.
    3. May be associated with complications including wound infection, seroma, paresthesias , lymphedema, pain, and limitation in shoulder range of motion. 
    4. In a prospective study in about 900 pts with localized breast Ca, no palpable adenopathy, no evident metastases, and a sentinel node POSITIVE for cancer, all of whom underwent lumpectomy and were randomized to axillary dissection vs. no dissection (most pts also received radiation and chemo or endocrine therapy), after median 6.3y f/u, there was no sig. diff. in overall survival, disease-free survival, or incidence of local recurrence (JAMA 305:569, 2011-JW)
  6. Testing "Sentinel Node" before doing axillary dissection
    1. The Sentinel Node is the first lymph node receiving drainage from a tumor and can be identified by intraoperative injection of radiocolloid tracer.  Presence of cancer in the sentinel node is associated with a higher risk of presence in other axillary lymph nodes.  When the sentinal node is cancer-free, incidence of cancer in other axillary nodes is low; thus, one alternative to routine axillary-node dissection in breast Ca surgery is testing of the sentinal node and no axillary dissection if it is cancer-free.
    2. In one series of 443 pts who had sentinel node identification and bx and then axillary node dissection, 11% of pts with a negative sentinel node had 1 or more positive nodes on axillary dissection (NEJM 339:941, 1998--JW)
    3. In a trial in 516 women with breast Ca, all of whom underwent lumpectomy and sentinel-node bx (followed afterward by radiation therapy), then randomized to axillary dissection no matter what vs. axillary dissection only if sentinel node was positive for cancer, over mean 4y f/u, there was no diff. between the groups in incidence of recurrent or metastatic breast Ca (NEJM 349:546, 2003--JW)
    4. In a study in 1,031 pts with clinically node-negatie invasive breast Ca randomized to axillary dissection vs. sentinal-node biopsy, the incidence of lymphedema, sensory deficits, and quality-of-life scores at 12mos were sig. lower in the sentinal node-biopsy group (J. Nat. Ca. Inst. 98:599, 2006--JW)
  1. Adjuvant chemotherapy
  1. Decision to give chemo is generally based on: Invasion of lymphatics, nerves, or small blood vessels; high nuclear grade; high histological grade; involvement of greater than 4 axillary lymph nodes
  2. Adjuvant (i.e. postsurgical) chemotherapy with cyclophosphamide, methotrexate, and fluorouracil (CMF) in women with early-stage breast cancer was associated with significant improvements in mortality in a systematic review of randomized and observational studies (RR for relapse 0.71, RR for death 0.79 compared with surgery alone, over median 28.5y f/u) (BMJ 330:220, 2005--JW)
  3. 1,577 post-surgical women with node-negative, estrogen-receptor-positive breast cancer randomized to cyclophosphamide, methotrexate, fluorouracil, and tamoxifen (CMFT) vs. tamoxifen alone. After 12y f/u, CMFT recipients had a sig. higher incidence of recurrence-free survival (89% vs. 79%). The diff. was greatest in women younger than 50, in premenopausal women, and in women whose tumors exhibited low concentrations of estrogen receptors. CMFT offered no significant advantage over tamoxifen alone in women who were older than 60 (Lancet 364:858, 2004--JW)
  4. "Neoadjuvant" chemo (preoperative)
    1. 1495 pts with stage I or II breast Ca randomized to receive chemo (doxorubicin + cyclophosphamide) pre-op vs. postop. No diff. in 5y disease-free survival or overall survival; about 9% of the women in the pre-op group were able to undergo lumpectomy + radiation instead of mastectomy (J. Clin. Onc. 16:2672, 1998--JW)
  5. High-dose chemotherapy with autologous stem-cell transplantation
    1. In a study of 885 pts with breast Ca and 4 or more positive axillary nodes but no known distant metastases randomized to conventional adjuvant chemotherapy or high-dose chemo plus stem-cell rescue, after median 4.8y f/u, no sig. diff. in 5y disease-free or overall survival; for subgroup of pts with 10 or more positive nodes, relapse-free survival was sig. greater in high-dose chemo group (61% vs. 51%) (NEJM 349:7, 2003--JW)
    2. In a study of 540 pts with breast Ca and 10 or more positive axillary nodes but no known distant metastases randomized to conventional adjuvant chemotherapy or high-dose chemo plus stem-cell rescue, after median 6.1y f/u, no sig. diff. in disease-free or overall survival  (NEJM 349:17, 2003--JW)
    3. High-dose chemo + autologous stem-cell support, vs. standard-dose chemo, over median 3.8y f/u, was ass'd with no sig. diff. in 4y event-free survival rate in a randomized trial of 302 women with breast Ca and > 9 positive nodes, but no identifiable distance metastases 
  1. Radiotherapy
    1. Generally used as an adjunctive treatment to surgery for women with localized breast cancer
    2. Can cause cardiac disease, brachial plexus injury, UE edema, rib fracture, penumonitis, and secondary malignangies (soft tissue sarcomata, lung Ca, and esophageal Ca)
    3. 1385 potsmenopausal women with stage II or III breast Ca s/p mastectomy and axillary dissection randomized to postoperative radiotherapy to the chest wall & regional lymph nodes vs. no tx; all received tamoxifen 30mg/d x 1y. Over median 10y f/u, risk of local recurrence was 8% in radiotherapy group vs. 35% in non-radiotherapy group; disease-free survival was 36% in radiotherapy group and 24% in non-radiotherapy group; overall survival was 45% in radiotherapy group vs. 36% in non-radiotherapy group (all differences were sig.) (Lancet 353:1641, 1999--JW)
    4. Axillary radiotherapy as an alternative to axillary node dissection in women with small (<2cm) tumors and no obvious clinical axillary involvement: 105 such pts, all postmenopausal, were tx'd with wide local excission then radiotherapy 5-6wks post-op to breast, axillary, and supraclavicular areas; mopst also got Tmoxifen 20mg/d; at average 41mo f/u, overall survival was 83%, disease-free survival was 82%; arm edema and impaired shoulder motion was present in 11% and 17%, respectively. (Cancer 88:1633, 2000--AFP)
    5. Comparisons of radiotherapy and tamoxifen in women with localized breast Ca s/p lumpectomy
      1. Radiotherapy was ass'd with sig. lower risk of ipsilateral recurrence c/w tamoxifen in an 8y randomized trial of 1009 women with breast Ca 1cm in diameter or smaller and negative axillary lymph nodes, all of whom underwent lumpectomy; no sig. diffs in incidence of death (J. Clin. Oncol. 20;4141, 2002--JW)
    6. Comparisons of radiotherapy vs. no radiotherapy in conjunction with tamoxifen in women with localized breast Ca s/p lumpectomy
      1. 769 women > 50yo with localized breast Ca (tumor size < 5cm) s/p lumpectomy randomized to tamoxifen 20mg PO QD x 5d vs. tamoxifen + radiation; 5y incidence of local recurrence was sig. lower in the combined group (0.6% vs. 7.7%); no sig. diff. in incidence of distant recurrence or total mortality (NEJM 351:963, 2004--JW)
      2. 636 women > 70yo with localized breast Ca (tumor size < 2cm) s/p lumpectomy randomized to tamoxifen 20mg PO QD x 5d vs. tamoxifen + radiation; 5y incidence of local recurrence was sig. lower in the combined group (1% vs. 4%); no sig. diff. in incidence of distant recurrence or total mortality (NEJM 351:971, 2004--JW)
  1. Estrogen suppression
    1. Tamoxifen (Novaldex)
  1. A nonsteroidal "selective estrogen receptor modulator" (SERM)
  2. Acts as agonist on some tissues, e.g. vaginal epithelium, endometrium, and bone; and an antagonist on others, e.g. breast
  3. Decreases LDL-cholesterol and increases bone density
  4. Clinical effects are mediated by its active metabolite endoxifen, to which it is metabolized by cytochrome P450 2D6 (CYP2D6), which is inhibited by various drugs including paroxetine and fluoxetine, which can thus diminish tamoxifen's effects (sertraline is a moderate inhibitor of CYP2D6; citalopram and escitalopram have mild effects, and venlafaxine and desvenlafaxine have no CYP2D6 inhibitory effects)
  5. Side effects include irregular menses, hot flashes, CNS effects, cataracts thromboembolic events, and increased risk of endometrial cancer (see above under "Prevention")
  6. Surveillance for endometrial Ca or pre-cancerous changes in women on Tamoxifen
    1. One proposed protocol is described in Ann. int. Med. 131:127, 1999
    2. Transvaginal ultrasound with biopsy of pts with increased endometrial thickness failed to identify a substantial # of women with serious abnormalities including cancer (J. Clin. Oncol. 18:3464, 2000--JW)
  7. Estrogen replacement therapy may add to tamoxifen effect
  8. Studies of use in Ductal Carcinoma in Situ
    1. Tamoxifen 10mg BID x 5y after lumpectomy + radiation was associated with sig. less recurrence of breast Ca (8.2% vs. 13.4%) over avg. 6y f/u in a randomized trial of1804 women with ductal carcinoma in situ (Lancet 353:1993, 1999--JW)
  9. Studies of use in Breast Cancer for prevention of recurrence
    1. Associated with decreased recurrence and increased survival in pts with ER+ but not in pts with ER- breast Ca (overview of 55 randomized trials with 37,000 pts total and avg 1y f/u--Lancet 351:1451, 1998--JW)
    2. Tamoxifen was + radiotherapy was ass'd with sig. lower risk of ipsilateral recurrence c/w either alone, in an 8y randomized trial of 1009 women with breast Ca 1cm in diameter or smaller and negative axillary lymph nodes, all of whom underwent lumpectomy; no sig. diffs in incidence of death (J. Clin. Oncol. 20;4141, 2002--JW)
    3. 2,892 post-surgical women with node-negative, estrogen receptor-positive breast cancer randomized to tamoxifen vs. placebo; after 15 years of follow-up, tamoxifen recipients had sig. higher incidence of recurrence-free survival (78% vs. 65%) and overall survival (71% vs. 65%) placebo recipients (Lancet 364:858, 2004--JW)
  1. Aromatase inhibitors
  1. Block synthesis of estrogens from androgens, causing circulating estrogen levels to fall by about 99%
  2. Side f/x include osteopenia, hot flashes, arthralgias, tenosynovitis, and myalgias
  3. Specific medications
    1. Anastrozole (Arimidex)
    2. Letrozole (Femara)
    3. Exemestane (Aromasin)
  4. Comparisons with Tamoxifen-Overall more effective at reducing breast Ca recurrence
    1. Anastrozole 1mg QD ass'd with similar response rates as and longer time to progression (11.1 vs. 5.6mos) and fewer thromboembolic events (RR 0.5) than Tamoxifen 20mg QD in a randomized trial in 1000 pts with locally advanced or metastatic breast Ca (J. Clin. Ontol. 18:3758, 2000--JW)
    2. In a study in 448 postmenopausal women s/p excision of estrogen-receptor-positive breast cancer, positive axillary nodes, and no known distant metastases, s/p 2-3y of tamoxifen, randomized to continue tamoxifen vs. start anastrozole to complete 5y of adjuvant hormonal therapy. Over median 36mo f/u, anastrozole group had sig. lower incidence of disease recurrence (5% vs. 14%) (J. Clin. Oncol. 23:5138, 2005--JW)
    3. In a trial in 9,366 postmenopausal women s/p primary therapy for localized breast Ca randomized to tamoxifen, anastrozole, or both; trial halted at 47 months when it became evident that anastrozole was associated with sig. better disease-free survival and lower incidence of contralateral Ca than tamoxifen; after median 68mo f/u, there was no sig. diff. in breast Ca mortality or total mortality, but anastrazole group had sig. greater disease-free survival (HR for qualifying event 0.87) and time to recurrence, and sig. lower incidence of contralateral breast Ca (HR 0.58), endometrial Ca, thromboembolic events, ischemic cerebrovascular events, and hot flashes, but sig. higher incidence of fractures (HR 1.45) ("ATAC trial", Lancet 365:60, 2005--JW)
    4. 4,742 postmenopausal women s/p 2-3y of tamoxifen for nonmetastatic breast Ca randomized to additional 2-3y of tamoxifen vs. 2-3y of exemestane.  After median 31mo f/u, 3y disease-free survival sig. higher in exemestane group (91.5% vs. 86.8%); no sig. diff. in overall mortality.  Tamoxifen group had sig. higher incidence of thromboembolic events and new non-breast cancers (NEJM 350:1081, 2004--JW)
    5. In a study in 8,010 postmenopausal women with hormone receptor-positive breast cancer randomized to letrozole vs. tamoxifen, over median 26mo f/u, incidence of (recurrence, new invasive contralateral breast Ca, any non-breast Ca, or non-cancer death) was sig. lower in letrozole group (HR 0.81) though the letrozole group had sig. higher incidence of fractures and cardiac events (NEJM 353:2747, 2005--JW)
  5. Studies of aromatase inhibitors after initial treatment with tamoxifen
    1. Letrozole 2.5mg/d vs. placebo was associated with sig. greater disease-free survival (projected 4y disease-free survival 93% vs. 87%) and nonsig. lower all-cause mortality over median 2.4y f/u in a randomized trial in 5,157 postmenopausal women with history of hormone-receptor-positive breast Ca who had received 5y of tamoxifen w/no evidence of recurrence.  Letrozole group had sig. greater incidence of hot flashes, arthralgias, and myalgia. (NEJM 349:1793, 2003--JW)
    2. In a study in 4,742 postmenopausal women with h/o breast Ca (mostly estrogen-receptor-positive) who had already received tamoxifen x 2-3y after surgery, randomized to continued tamoxifen x exemestane x 2-3y; over median 31mo f/u, the 3y disease-free survival rate was sig. higher in the exemestane group (91.5% vs. 86.8%), though no sig. overall mortality; tamoxifen group had sig. higher incidence of thromboembolic events and new non-breast cancers (NEJM 350:1081, 2004--JW)
      1. In a follow-up report showing data 5y after randomization, incidence of (recurrent breast Ca, contralateral breast Ca, or presumed breast Ca death) was sig. lower in exemestane group (15% vs. 19%); among women with estrogen-receptor-positive breast Ca; the mortality rate was also sig. lower in exemestane recipients.  Incidence of osteoporosis and fractures were sig. higher in exemestane recipients (Lancet 369:559, 2007--JW)
  1. Goserelin (Zoladex)--An LHRH analog that suppresses ovarian estrogen production; may have additive effect with tamoxifen
  1. Monoclonal antibodies
    1. Trastuzumab (Hrceptin)
      1. A humanized version of a mouse monoclonal antibody against human epidermal growth factor receptor 2 (HER2)
        1. The gene for HER2 is overexpressed in about 1/3 of breast cancers
        2. Amplification or overexpression of HER2 is associated with poor outcomes 
      2. Shown to improve survival as an adjunct to standard chemotherapy in randomized studies in women with tumors which over-express HER2 (e.g. NEJM 34:783, 2001--AFP)
      3. Can cause reversible cardiac dysfunction
      4. In a series of studies in about 7,000 women with HER2-positive breast Ca and no evidence of distant metastases (but some with positive axillary nodes), all of whom underwent surgery (and most radiation therapy):
        1. Adjuvant chemotherapy followed by trastuzumab IV Q3wks x 1y vs. adjuvant chemotherapy alone was ass'd with sig. greater 2y disease-free survival (86% vs. 77%)
        2. Adjuvant chemotherapy + overlapping trastuzumab IV Qwk x 1y vs. adjujant chemotherapy alone was ass'd with sig. greater 3y disease-free survival (87% vs. 75%)
          (NEJM 353:1659 and 1673, 2005--JW)
        3. In a follow-up paper describing results of the 2nd year of the study, after median 2y f/u, trastuzumab group still had sig. fewer recurrences and sig. fewer deaths (3.5% vs. 5.3%). Trastuzumab had sig. higher incidence of heart failure (5.6% vs. 0.6% of controls) (Lancet 369:29, 2007--JW)
          Can cause cardiotoxicity--Associated with CHF (yearly incidence of 0.5-4% in randomized trials in pts with no h/o heart disease)
  2. Bisphosphonates (see under "Osteoporosis")
    1. May reduce skeletal complications from bone metastases from breast Ca and may even reduce risk of new bone mets (NEJM 339:357, 1998--JW)
    2. In a study in 1,803 premenopausal women s/p surgery for stage I or II breast Ca positive for estrogen or progesterone receptors, all receiving radiation therapy and ovarian-suppression therapy with GNRH, randomized to zoledronic acid 4mg IV Q6mos vs. no zoledronic acid, over median 48mo f/u, zoledronic acid recipients had sig. higher disease-free survival (94.0% vs. 90.8%) though higher incidence of bone pain (35% vs. 25%), arthralgias (24% vs. 18%), and fever (9% vs. 2%) (NEJM 360:679, 2009-JW)
  3. Dietary soy intake-See above under "Prevention"

(Sources include Core Content Review of Family Medicine, 2012)