Breast cancer is a malignant proliferation of epithelial cells lining the ducts or lobules of the breast. In the year 2007, about 180,510 cases of invasive breast cancer and 40,910 deaths occurred in the United States. Epithelial malignancies of the breast are the most common cause of cancer in women (excluding skin cancer), accounting for about one-third of all cancer in women. As a result of improved treatment and earlier detection, mortality from breast cancer has begun to decrease substantially in the United States. This chapter will not consider rare malignancies presenting in the breast, such as sarcomas and lymphomas, but will focus on the epithelial cancers. Human breast cancer is a clonal disease; a single transformed cell—the product of a series of somatic (acquired) or germline mutations—is eventually able to express full malignant potential. Thus, breast cancer may exist for a long period as either a noninvasive disease or an invasive but nonmetastatic disease. These facts have significant clinical ramifications.
Not more than 10% of human breast cancers can be linked directly to germline mutations. Several genes have been implicated in familial cases. The Li-Fraumeni syndrome is characterized by inherited mutations in the p53 tumor-suppressor gene, which lead to an increased incidence of breast cancer, osteogenic sarcomas, and other malignancies. Inherited mutations in PTEN have also been reported in breast cancer.
Another tumor-suppressor gene, BRCA-1, has been identified at the chromosomal locus 17q21; this gene encodes a zinc finger protein, and the product therefore may function as a transcription factor. The gene appears to be involved in gene repair. Women who inherit a mutated allele of this gene from either parent have at least a 60–80% lifetime chance of developing breast cancer and about a 33% chance of developing ovarian cancer. The risk is higher among women born after 1940, presumably due to promotional effects of hormonal factors. Men who carry a mutant allele of the gene have an increased incidence of prostate cancer and breast cancer. A fourth gene, termed BRCA-2, which has been localized to chromosome 13q12, is also associated with an increased incidence of breast cancer in men and women.
Germline mutations in BRCA-1 and BRCA-2 can be readily detected; patients with these mutations can be counseled appropriately. All women with strong family histories for breast cancer should be referred to genetic screening programs, particularly women of Ashkenazi Jewish descent who have a high likelihood of a specific BRCA-1 mutation (deletion of adenine and guanine at position 185).
Even more important than the role these genes play in inherited forms of breast cancer may be their role in sporadic breast cancer. The p53 mutation is present in nearly 40% of human breast cancers as an acquired defect. Acquired mutations in PTEN occur in about 10% of the cases. BRCA-1 mutation in sporadic primary breast cancer has not been reported. However, decreased expression of BRCA-1 mRNA (possibly via gene methylation) and abnormal cellular location of the BRCA-1 protein have been found in some breast cancers. Loss of heterozygosity of BRCA-1 and BRCA-2 suggests that tumor-suppressor activity may be inactivated in sporadic cases of human breast cancer. Finally, increased expression of a dominant oncogene plays a role in about a quarter of human breast cancer cases. The product of this gene, a member of the epidermal growth factor receptor superfamily, is called erbB2 (HER-2, neu) and is overexpressed in these breast cancers due to gene amplification; this overexpression can contribute to transformation of human breast epithelium and is the target of effective systemic therapy in adjuvant and metastatic disease settings.
Breast cancer is a hormone-dependent disease. Women without functioning ovaries who never receive estrogen-replacement therapy do not develop breast cancer. The female:male ratio is about 150:1. For most epithelial malignancies, a log-log plot of incidence versus age shows a single-component straight-line increase with every year of life. A similar plot for breast cancer shows two components: a straight-line increase with age but with a decrease in slope beginning at the age of menopause. The three dates in a woman's life that have a major impact on breast cancer incidence are age at menarche, age at first full-term pregnancy, and age at menopause. Women who experience menarche at age 16 have only 50–60% of the breast cancer risk of a woman having menarche at age 12; the lower risk persists throughout life. Similarly, menopause occurring 10 years before the median age of menopause (52 years), whether natural or surgically induced, reduces lifetime breast cancer risk by about 35%. Women who have a first full-term pregnancy by age 18 have a 30–40% lower risk of breast cancer compared with nulliparous women. Thus, length of menstrual life—particularly the fraction occurring before first full-term pregnancy—is a substantial component of the total risk of breast cancer. These three factors (menarche, age of first full-term pregnancy, and menopause) can account for 70–80% of the variation in breast cancer frequency in different countries. A meta-analysis has shown that duration of maternal nursing correlates with substantial risk reduction independent of either parity or age at first full-term pregnancy.
International variation in incidence has provided some of the most important clues on hormonal carcinogenesis. A woman living to age 80 in North America has one chance in nine of developing invasive breast cancer. Asian women have one-fifth to one-tenth the risk of breast cancer of women in North America or Western Europe. Asian women have substantially lower concentrations of estrogens and progesterone. These differences cannot be explained on a genetic basis because Asian women living in a western environment have sex steroid hormone concentrations and risks identical to those of their western counterparts. These migrant women and more notably their daughters also differ markedly in height and weight from Asian women in Asia; height and weight are critical regulators of age of menarche and have substantial effects on plasma concentrations of estrogens.
The role of diet in breast cancer etiology is controversial. While there are associative links between total caloric and fat intake and breast cancer risk, the exact role of fat in the diet is unproven. Increased caloric intake contributes to breast cancer risk in multiple ways: earlier menarche, later age at menopause, and increased postmenopausal estrogen concentrations reflecting enhanced aromatase activities in fatty tissues. Moderate alcohol intake also increases the risk by an unknown mechanism. Folic acid supplementation appears to modify risk in women who use alcohol but is not additionally protective in abstainers. Recommendations favoring abstinence from alcohol must be weighed against other social pressures and the possible cardioprotective effect of moderate alcohol intake.
Understanding the potential role of exogenous hormones in breast cancer is of extraordinary importance because millions of American women regularly use oral contraceptives and postmenopausal hormone replacement therapy. The most credible meta-analyses of oral contraceptive use suggest that these agents cause little if any increased risk of breast cancer. By contrast, oral contraceptives offer a substantial protective effect against ovarian epithelial tumors and endometrial cancers. Far more controversial are the data surrounding HRT in postmenopausal women. Data from the Women's Health Initiative (WHI) trial showed in a prospectively randomized design that conjugated equine estrogens plus progestins increased the risk of breast cancer and adverse cardiovascular events but with decreases in bone fractures and colorectal cancer. On balance there were more negative events with HRT. A parallel WHI trial with >12,000 women enrolled testing conjugated estrogens alone (in women who have had hysterectomies) showed no significant increase in breast cancer incidence. A meta-analysis of nonrandomized HRT studies suggests that most of the previously attributed benefit of HRT can be accounted for by higher socioeconomic status among users, which is presumably associated with better access to health care and healthier behaviors. Certain potential benefits of HRT, such as a putative protective effect on cognition with age, were not assessed in WHI. HRT is an area of rapid reevaluation, but it would appear (at least from breast cancer and cardiovascular disease vantage points) that there are serious concerns about long-term HRT use. HRT in women previously diagnosed with breast cancer increases recurrence rates.
In addition to the other factors, radiation is a risk factor in younger women. Women who have been exposed before age 30 to radiation in the form of multiple fluoroscopies (200–300 cGy) or treatment for Hodgkin's disease ( more than 3600 cGy) have a substantial increase in risk of breast cancer, whereas radiation exposure after age 30 appears to have a minimal carcinogenic effect on the breast.
Evaluation of Breast Masses in Men and Women
Because the breasts are a common site of potentially fatal malignancy in women and because they frequently provide clues to underlying systemic diseases in both men and women, examination of the breast is an essential part of the physical examination. Unfortunately, internists frequently do not examine breasts in men, and, in women, they are apt to defer this evaluation to gynecologists. Because of the plausible association between early detection and improved outcome, it is the duty of every physician to distinguish breast abnormalities at the earliest possible stage and to institute a diagnostic workup. Women should be trained in breast self-examination (BSE). Although breast cancer in men is unusual, unilateral lesions should be evaluated in the same manner as in women, with the recognition that gynecomastia in men can sometimes begin unilaterally and is often asymmetric.
Virtually all breast cancer is diagnosed by biopsy of a nodule detected either on a mammogram or by palpation. Algorithms have been developed to enhance the likelihood of diagnosing breast cancer and reduce the frequency of unnecessary biopsy (Fig. 86-1).
The Palpable Breast Mass
Women should be strongly encouraged to examine their breasts monthly. A potentially flawed study from China has suggested that BSE does not alter survival, but given its safety, the procedure should still be encouraged. At worst, this practice increases the likelihood of detecting a mass at a smaller size when it can be treated with more limited surgery. Breast examination by the physician should be performed in good light so as to see retractions and other skin changes. The nipple and areolae should be inspected, and an attempt should be made to elicit nipple discharge. All regional lymph node groups should be examined, and any lesions should be measured. Physical examination alone cannot exclude malignancy. Lesions with certain features are more likely to be cancerous (hard, irregular, tethered or fixed, or painless lesions). A negative mammogram in the presence of a persistent lump in the breast does not exclude malignancy. Palpable lesions require additional diagnostic procedures including biopsy.
In premenopausal women, lesions that are either equivocal or nonsuspicious on physical examination should be reexamined in 2–4 weeks, during the follicular phase of the menstrual cycle. Days 5–7 of the cycle are the best time for breast examination. A dominant mass in a postmenopausal woman or a dominant mass that persists through a menstrual cycle in a premenopausal woman should be aspirated by fine-needle biopsy or referred to a surgeon. If nonbloody fluid is aspirated, the diagnosis (cyst) and therapy have been accomplished together. Solid lesions that are persistent, recurrent, complex, or bloody cysts require mammography and biopsy, although in selected patients the so-called triple diagnostic techniques (palpation, mammography, aspiration) can be used to avoid biopsy (Figs. 86-1, 86-2, and 86-3). Ultrasound can be used in place of fine-needle aspiration to distinguish cysts from solid lesions. Not all solid masses are detected by ultrasound; thus, a palpable mass that is not visualized on ultrasound must be presumed to be solid.
Several points are essential in pursuing these management decision trees. First, risk-factor analysis is not part of the decision structure. No constellation of risk factors, by their presence or absence, can be used to exclude biopsy. Second, fine-needle aspiration should be used only in centers that have proven skill in obtaining such specimens and analyzing them. The likelihood of cancer is low in the setting of a "triple negative" (benign-feeling lump, negative mammogram, and negative fine-needle aspiration), but it is not zero. The patient and physician must be aware of a 1% risk of false negatives. Third, additional technologies such as MRI, ultrasound, and sestamibi imaging cannot be used to exclude the need for biopsy, although in unusual circumstances they may provoke a biopsy.
The Abnormal Mammogram
Diagnostic mammography should not be confused with screening mammography, which is performed after a palpable abnormality has been detected. Diagnostic mammography is aimed at evaluating the rest of the breast before biopsy is performed or occasionally is part of the triple-test strategy to exclude immediate biopsy.
Subtle abnormalities that are first detected by screening mammography should be evaluated carefully by compression or magnified views. These abnormalities include clustered microcalcifications, densities (especially if spiculated), and new or enlarging architectural distortion. For some nonpalpable lesions, ultrasound may be helpful either to identify cysts or to guide biopsy. If there is no palpable lesion and detailed mammographic studies are unequivocally benign, the patient should have routine follow-up appropriate to the patient's age.
If a nonpalpable mammographic lesion has a low index of suspicion, mammographic follow-up in 3–6 months is reasonable. Workup of indeterminate and suspicious lesions has been rendered more complex by the advent of stereotactic biopsies. Morrow and colleagues have suggested that these procedures are indicated for lesions that require biopsy but are likely to be benign—that is, for cases in which the procedure probably will eliminate additional surgery. When a lesion is more probably malignant, open biopsy should be performed with a needle localization technique. Others have proposed more widespread use of stereotactic core biopsies for nonpalpable lesions on economic grounds and because diagnosis leads to earlier treatment planning. However, stereotactic diagnosis of a malignant lesion does not eliminate the need for definitive surgical procedures, particularly if breast conservation is attempted. For example, after a breast biopsy with needle localization (i.e., local excision) of a stereotactically diagnosed malignancy, reexcision may still be necessary to achieve negative margins. To some extent, these issues are decided on the basis of referral pattern and the availability of the resources for stereotactic core biopsies. A reasonable approach is shown in Fig. 86-4
During pregnancy, the breast grows under the influence of estrogen, progesterone, prolactin, and human placental lactogen. Lactation is suppressed by progesterone, which blocks the effects of prolactin. After delivery, lactation is promoted by the fall in progesterone levels, which leaves the effects of prolactin unopposed. The development of a dominant mass during pregnancy or lactation should never be attributed to hormonal changes. A dominant mass must be treated with the same concern in a pregnant woman as any other. Breast cancer develops in 1 in every 3000–4000 pregnancies. Stage for stage, breast cancer in pregnant patients is no different from premenopausal breast cancer in nonpregnant patients. However, pregnant women often have more advanced disease because the significance of a breast mass was not fully considered and/or because of endogenous hormone stimulation. Persistent lumps in the breast of pregnant or lactating women cannot be attributed to benign changes based on physical findings; such patients should be promptly referred for diagnostic evaluation.
Only about 1 in every 5–10 breast biopsies leads to a diagnosis of cancer, although the rate of positive biopsies varies in different countries and clinical settings. (These differences may be related to interpretation, medicolegal considerations, and availability of mammograms.) The vast majority of benign breast masses are due to "fibrocystic" disease, a descriptive term for small fluid-filled cysts and modest epithelial cell and fibrous tissue hyperplasia. However, fibrocystic disease is a histologic, not a clinical, diagnosis, and women who have had a biopsy with benign findings are at greater risk of developing breast cancer than those who have not had a biopsy. The subset of women with ductal or lobular cell proliferation (about 30% of patients), particularly the small fraction (3%) with atypical hyperplasia, have a fourfold greater risk of developing breast cancer than unbiopsied women, and the increase in the risk is about ninefold for women in this category who also have an affected first-degree relative. Thus, careful follow-up of these patients is required. By contrast, patients with a benign biopsy without atypical hyperplasia are at little risk and may be followed routinely.
Breast cancer is virtually unique among the epithelial tumors in adults in that screening (in the form of annual mammography) improves survival. Meta-analysis examining outcomes from every randomized trial of mammography conclusively shows a 25–30% reduction in the chance of dying from breast cancer with annual screening after age 50; the data for women between ages 40 and 50 are almost as positive. While controversy continues to surround the assessment of screening mammography, the preponderance of data strongly supports the benefits of screening mammography. New analyses of older randomized studies have suggested that screening may not work. While the design defects in some older studies cannot be retrospectively corrected, most experts, including panels of the American Society of Clinical Oncology and the American Cancer Society, continue to believe that screening conveys substantial benefit. Furthermore, the profound drop in breast cancer mortality seen over the past decade is unlikely to be solely attributable to improvements in therapy. It seems prudent to recommend annual mammography for women past the age of 40.
Although no randomized study of BSE has ever shown any improvement in survival, its major benefit is identification of tumors appropriate for conservative local therapy. Better mammographic technology, including digitized mammography, routine use of magnified views, and greater skill in mammographic interpretation, combined with newer diagnostic techniques (MRI, magnetic resonance spectroscopy, positron emission tomography, etc.) may make it possible to identify breast cancers even more reliably and earlier. Screening by any technique other than mammography is not indicated; however, younger women who are BRCA-1 or BRCA-2 carriers may benefit from MRI screening where the higher sensitivity may outweigh the loss of specificity.
Correct staging of breast cancer patients is of extraordinary importance. Not only does it permit an accurate prognosis, but in many cases therapeutic decision-making is based largely on the TNM (primary tumor, regional nodes, metastasis) classification (Table 86-1). Comparison with historic series should be undertaken with caution, as the staging has changed several times in the past 20 years. The current staging is complex and results in significant changes in outcome by stage as compared with prior staging systems.
aRT-PCR, reverse transcriptase/polymerase chain reaction.
bClinically apparent is defined as detected by imaging studies (excluding lymphoscintigraphy) or by clinical examination.
cT1 includes T1mic.
Breast Cancer: Treatment
Primary Breast Cancer
Breast-conserving treatments, consisting of the removal of the primary tumor by some form of lumpectomy with or without irradiating the breast, result in a survival that is as good as (or slightly superior to) that after extensive surgical procedures, such as mastectomy or modified radical mastectomy, with or without further irradiation. Postlumpectomy breast irradiation greatly reduces the risk of recurrence in the breast. While breast conservation is associated with a possibility of recurrence in the breast, 10-year survival is at least as good as that after more radical surgery. Postoperative radiation to regional nodes following mastectomy is also associated with an improvement in survival. Since radiation therapy can also reduce the rate of local or regional recurrence, it should be strongly considered following mastectomy for women with high-risk primary tumors (i.e., T2 in size, positive margins, positive nodes).
At present, nearly one-third of women in the United States are managed by lumpectomy. Breast-conserving surgery is not suitable for all patients: it is not generally suitable for tumors more than 5 cm (or for smaller tumors if the breast is small), for tumors involving the nipple areola complex, for tumors with extensive intraductal disease involving multiple quadrants of the breast, for women with a history of collagen-vascular disease, and for women who either do not have the motivation for breast conservation or do not have convenient access to radiation therapy. However, these groups probably do not account for more than one-third of patients who are treated with mastectomy. Thus, a great many women still undergo mastectomy who could safely avoid this procedure and probably would if appropriately counseled.
An extensive intraductal component is a predictor of recurrence in the breast, and so are several clinical variables. Both axillary lymph node involvement and involvement of vascular or lymphatic channels by metastatic tumor in the breast are associated with a higher risk of relapse in the breast but are not contraindications to breast-conserving treatment. When these patients are excluded, and when lumpectomy with negative tumor margins is achieved, breast conservation is associated with a recurrence rate in the breast of substantially more than 10%. The survival of patients who have recurrence in the breast is somewhat worse than that of women who do not. Thus, recurrence in the breast is a negative prognostic variable for long-term survival. However, recurrence in the breast is not the cause of distant metastasis. If recurrence in the breast caused metastatic disease, then women treated with lumpectomy, who have a higher rate of recurrence in the breast, should have poorer survival than women treated with mastectomy, and they do not. Most patients should consult with a radiation oncologist before making a final decision concerning local therapy. However, a multimodality clinic in which the surgeon, radiation oncologist, medical oncologist, and other caregivers cooperate to evaluate the patient and develop a treatment is usually considered a major advantage by patients.
The use of systemic therapy after local management of breast cancer substantially improves survival. More than one-third of the women who would otherwise die of metastatic breast cancer remain disease-free when treated with the appropriate systemic regimen.
The most important prognostic variables are provided by tumor staging. The size of the tumor and the status of the axillary lymph nodes provide reasonably accurate information on the likelihood of tumor relapse. The relation of pathologic stage to 5-year survival is shown in Table 86-2. For most women, the need for adjuvant therapy can be readily defined on this basis alone. In the absence of lymph node involvement, involvement of microvessels (either capillaries or lymphatic channels) in tumors is nearly equivalent to lymph node involvement. The greatest controversy concerns women with intermediate prognoses. There is rarely justification for adjuvant chemotherapy in most women with tumors more than 1 cm in size whose axillary lymph nodes are negative. Detection of breast cancer cells either in the circulation or bone marrow is associated with an increased relapse rate. The most exciting development in this area is the use of gene expression arrays to analyze patterns of tumor gene expression. Several groups have independently defined gene sets that reliably predict disease-free and overall survival far more accurately than any single prognostic variable. Their value is now being assessed in prospective randomized trials. In addition, gene sets capable of predicting responses to endocrine therapy and specific chemotherapeutic drugs have also been described.
Estrogen and progesterone receptor status are of prognostic significance. Tumors that lack either or both of these receptors are more likely to recur than tumors that have them.
Several measures of tumor growth rate correlate with early relapse. S-phase analysis using flow cytometry is the most accurate measure. Indirect S-phase assessments using antigens associated with the cell cycle, such as PCNA (Ki67), are also valuable. Tumors with a high proportion (more than the median) of cells in S phase pose a greater risk of relapse; chemotherapy offers the greatest survival benefit for these tumors. Assessment of DNA content in the form of ploidy is of modest value, with nondiploid tumors having a somewhat worse prognosis.
Histologic classification of the tumor has also been used as a prognostic factor. Tumors with a poor nuclear grade have a higher risk of recurrence than tumors with a good nuclear grade. Semiquantitative measures such as the Elston score improve the reproducibility of this measurement.
Molecular changes in the tumor are also useful. Tumors that overexpress erbB2 (HER-2/neu) or have a mutated p53 gene have a worse prognosis. Particular interest has centered on erbB2 overexpression as measured by histochemistry or by fluorescence in situ hybridization. Tumors that overexpress erbB2 are more likely to respond to higher doses of doxorubicin-containing regimens and predict those tumors that will respond to HER-2/neu antibodies (trastuzumab) (herceptin) and a Her-2/neu kinase inhibitor.
To grow, tumors must generate a neovasculature. The presence of more microvessels in a tumor, particularly when localized in so-called "hot spots," is associated with a worse prognosis. This may assume even greater significance in light of blood vessel–targeting therapies such as bevacizumab (avastin).
Other variables that have also been used to evaluate prognosis include proteins associated with invasiveness, such as type IV collagenase, cathepsin D, plasminogen activator, plasminogen activator receptor, and the metastasis-suppressor gene nm23. None of these has been widely accepted as a prognostic variable for therapeutic decision-making. One problem in interpreting these prognostic variables is that most of them have not been examined in a study using a large cohort of patients.
Adjuvant therapy is the use of systemic therapies in patients whose known disease has received local therapy but who are at risk of relapse. Selection of appropriate adjuvant chemotherapy or hormone therapy is highly controversial in some situations. Meta-analyses have helped to define broad limits for therapy but do not help in choosing optimal regimens or in choosing a regimen for certain subgroups of patients. A summary of recommendations is shown in Table 86-3. In general, premenopausal women for whom any form of adjuvant systemic therapy is indicated should receive multidrug chemotherapy. The antiestrogen tamoxifen improves survival in premenopausal patients with positive estrogen receptors and should be added following completion of chemotherapy. Prophylactic castration may also be associated with a substantial survival benefit (primarily in estrogen receptor–positive patients) but is not widely used in this country.
aAs determined by pathologic examination.
Data on postmenopausal women are also controversial. The impact of adjuvant chemotherapy is quantitatively less clear-cut than in premenopausal patients, although survival advantages have been shown. The first decision is whether chemotherapy or endocrine therapy should be used. While adjuvant tamoxifen improves survival regardless of axillary lymph node status, the improvement in survival is modest for patients in whom multiple lymph nodes are involved. For this reason, it has been usual to give chemotherapy to postmenopausal patients who have no medical contraindications and who have more than one positive lymph node; tamoxifen is commonly given simultaneously or subsequently. For postmenopausal women for whom systemic therapy is warranted but who have a more favorable prognosis, tamoxifen may be used as a single agent. Large clinical trials have shown superiority for aromatase inhibitors over tamoxifen alone in the adjuvant setting. Unfortunately the optimal plan is unclear. Tamoxifen for 5 years followed by an aromatase inhibitor, the reverse strategy, or even switching to an aromatase inhibitor after 2–3 years of tamoxifen has been shown to be better than tamoxifen alone. No valid information currently permits selection among the three clinically approved aromatase inhibitors. Large clinical trials currently underway will help address these questions.
Most comparisons of adjuvant chemotherapy regimens show little difference among them, although small advantages for doxorubicin-containing regimens are usually seen.
One approach—so-called neoadjuvant chemotherapy—involves the administration of adjuvant therapy before definitive surgery and radiation therapy. Because the objective response rates of patients with breast cancer to systemic therapy in this setting exceed 75%, many patients will be "downstaged" and may become candidates for breast-conserving therapy. However, overall survival has not been improved using this approach.
Other adjuvant treatments under investigation include the use of taxanes, such as paclitaxel and docetaxel, and therapy based on alternative kinetic and biologic models. In such approaches, high doses of single agents are used separately in relatively dose-intensive cycling regimens. Node-positive patients treated with doxorubicin-cyclophosphamide for four cycles followed by four cycles of a taxane have a substantial improvement in survival as compared with women receiving doxorubicin-cyclophosphamide alone, particularly in women with estrogen receptor–negative tumors. In addition, administration of the same drug combinations at the same dose but at more frequent intervals (q2 weeks with cytokine support as compared with the standard q3 weeks) is even more effective. Among the 25% of women whose tumors overexpress HER-2/neu, addition of trastuzumab given concurrently with a taxane and then for a year after chemotherapy produces significant improvement in survival. Though longer follow-up will be important, this is now the standard care for most women with HER-2/neu positive breast cancers. Cardiotoxicity, immediate and long-term, remains a concern, and further efforts to exploit nonanthracycline-containing regimens are being pursued. Very-high-dose therapy with stem cell transplantation in the adjuvant setting has not proved superior to standard dose therapy and should not be routinely used.
Systemic Therapy of Metastatic Disease
Nearly half of patients treated for apparently localized breast cancer develop metastatic disease. Although a small number of these patients enjoy long remissions when treated with combinations of systemic and local therapy, most eventually succumb to metastatic disease. Soft tissue, bony, and visceral (lung and liver) metastases each account for approximately one-third of sites of initial relapses. However, by the time of death, most patients will have bony involvement. Recurrences can appear at any time after primary therapy. Half of all initial cancer recurrences occur more than 5 years after initial therapy.
Because the diagnosis of metastatic disease alters the outlook for the patient so drastically, it should rarely be made without biopsy. Every oncologist has seen patients with tuberculosis, gallstones, sarcoidosis, or other nonmalignant diseases misdiagnosed and treated as though they had metastatic breast cancer or even second malignancies such as multiple myeloma thought to be recurrent breast cancer. This is a catastrophic mistake and justifies biopsy for virtually every patient at the time of initial suspicion of metastatic disease.
The choice of therapy requires consideration of local therapy needs, the overall medical condition of the patient, and the hormone receptor status of the tumor, as well as clinical judgment. Because therapy of systemic disease is palliative, the potential toxicities of therapies should be balanced against the response rates. Several variables influence the response to systemic therapy. For example, the presence of estrogen and progesterone receptors is a strong indication for endocrine therapy. On the other hand, patients with short disease-free intervals, rapidly progressive visceral disease, lymphangitic pulmonary disease, or intracranial disease are unlikely to respond to endocrine therapy.
In many cases, systemic therapy can be withheld while the patient is managed with appropriate local therapy. Radiation therapy and occasionally surgery are effective at relieving the symptoms of metastatic disease, particularly when bony sites are involved. Many patients with bone-only or bone-dominant disease have a relatively indolent course. Under such circumstances, systemic chemotherapy has a modest effect, whereas radiation therapy may be effective for long periods. Other systemic treatments, such as strontium 89 and/or bisphosphonates, may provide a palliative benefit without inducing objective responses. Most patients with metastatic disease and certainly all who have bone involvement should receive concurrent bisphosphonates. Since the goal of therapy is to maintain well-being for as long as possible, emphasis should be placed on avoiding the most hazardous complications of metastatic disease, including pathologic fracture of the axial skeleton and spinal cord compression. New back pain in patients with cancer should be explored aggressively on an emergent basis; to wait for neurologic symptoms is a potentially catastrophic error. Metastatic involvement of endocrine organs can cause profound dysfunction, including adrenal insufficiency and hypopituitarism. Similarly, obstruction of the biliary tree or other impaired organ function may be better managed with a local therapy than with a systemic approach.
Normal breast tissue is estrogen-dependent. Both primary and metastatic breast cancer may retain this phenotype. The best means of ascertaining whether a breast cancer is hormone-dependent is through analysis of estrogen and progesterone receptor levels on the tumor. Tumors that are positive for the estrogen receptor and negative for the progesterone receptor have a response rate of ~30%. Tumors that have both receptors have a response rate approaching 70%. If neither receptor is present, the objective response rates are more than 10%. Receptor analyses provide information as to the correct ordering of endocrine therapies as opposed to chemotherapy. Because of their lack of toxicity and because some patients whose receptor analyses are reported as negative respond to endocrine therapy, an endocrine treatment should be attempted in virtually every patient with metastatic breast cancer. Potential endocrine therapies are summarized in Table 86-4. The choice of endocrine therapy is usually determined by toxicity profile and availability.
In most patients, the initial endocrine therapy should be an aromatase inhibitor rather than tamoxifen. For the subset of women who are ER positive but also HER-2/neu positive, response rates to aromatase inhibitors are very substantially higher than to tamoxifen. Newer "pure" antiestrogens that are free of agonistic effects are also in clinical trial. Cases in which tumors shrink in response to tamoxifen withdrawal (as well as withdrawal of pharmacologic doses of estrogens) have been reported. Endogenous estrogen formation may be blocked by analogues of luteinizing hormone–releasing hormone in premenopausal women.
Additive endocrine therapies, including treatment with progestogens, estrogens, and androgens, may also be tried in patients who respond to initial endocrine therapy; the mechanism of action of these latter therapies is unknown. Patients who respond to one endocrine therapy have at least a 50% chance of responding to a second endocrine therapy. It is not uncommon for patients to respond to two or three sequential endocrine therapies; however, combination endocrine therapies do not appear to be superior to individual agents, and combinations of chemotherapy with endocrine therapy are not useful. The median survival of patients with metastatic disease is approximately 2 years, and many patients, particularly older persons and those with hormone-dependent disease, may respond to endocrine therapy for 3–5 years or longer.
Note: LHRH, luteinizing hormone–releasing hormone.
Unlike many other epithelial malignancies, breast cancer responds to multiple chemotherapeutic agents, including anthracyclines, alkylating agents, taxanes, and antimetabolites. Multiple combinations of these agents have been found to improve response rates somewhat, but they have had little effect on duration of response or survival. The choice among multidrug combinations frequently depends on whether adjuvant chemotherapy was administered and, if so, what type. While patients treated with adjuvant regimens such as cyclophosphamide, methotrexate, and fluorouracil (CMF regimens) may subsequently respond to the same combination in the metastatic disease setting, most oncologists use drugs to which the patients have not been previously exposed. Once patients have progressed after combination drug therapy, it is most common to treat them with single agents. Given the significant toxicity of most drugs, the use of a single effective agent will minimize toxicity by sparing the patient exposure to drugs that would be of little value. No method to select the drugs most efficacious for a given patient has been demonstrated to be useful.
Most oncologists use either an anthracycline or paclitaxel following failure with the initial regimen. However, the choice has to be balanced with individual needs. One randomized study has suggested docetaxel may be superior to paclitaxel. A nanoparticle formulation of paclitaxel (abraxane) has also shown promise.
The use of a humanized antibody to erbB2 [trastuzumab (Herceptin)] combined with paclitaxel can improve response rate and survival for women whose metastatic tumors overexpress erbB2. The magnitude of the survival extension is modest in patients with metastatic disease. Similarly, the use of bevacizumab (avastin) has improved the response rate and response duration to paclitaxel. Objective responses in previously treated patients may also be seen with gemcitabine, capecitabine, navelbine, and oral etoposide.
High-Dose Chemotherapy Including Autologous Bone Marrow Transplantation
Autologous bone marrow transplantation combined with high doses of single agents can produce objective responses even in heavily pretreated patients. However, such responses are rarely durable and do not alter the clinical course for most patients with advanced metastatic disease.
Stage III Breast Cancer
Between 10 and 25% of patients present with so-called locally advanced, or stage III, breast cancer at diagnosis. Many of these cancers are technically operable, whereas others, particularly cancers with chest wall involvement, inflammatory breast cancers, or cancers with large matted axillary lymph nodes, cannot be managed with surgery initially. Although no randomized trials have proved the efficacy of neoadjuvant chemotherapy, this approach has gained widespread use. More than 90% of patients with locally advanced breast cancer show a partial or better response to multidrug chemotherapy regimens that include an anthracycline. Early administration of this treatment reduces the bulk of the disease and frequently makes the patient a suitable candidate for salvage surgery and/or radiation therapy. These patients should be managed in multimodality clinics to coordinate surgery, radiation therapy, and systemic chemotherapy. Such approaches produce long-term disease-free survival in about 30–50% of patients.
Breast Cancer Prevention
Women who have one breast cancer are at risk of developing a contralateral breast cancer at a rate of approximately 0.5% per year. When adjuvant tamoxifen is administered to these patients, the rate of development of contralateral breast cancers is reduced. In other tissues of the body, tamoxifen has estrogen-like effects that are beneficial: preservation of bone mineral density and long-term lowering of cholesterol. However, tamoxifen has estrogen-like effects on the uterus, leading to an increased risk of uterine cancer (0.75% incidence after 5 years on tamoxifen). Tamoxifen also increases the risk of cataract formation. The Breast Cancer Prevention Trial (BCPT) revealed a more than 49% reduction in breast cancer among women with a risk of at least 1.66% taking the drug for 5 years. Raloxifene has shown similar breast cancer prevention potency but may have different effects on bone and heart. The two agents have been compared in a prospective randomized prevention trial (the STAR trial). The agents are approximately equivalent in preventing breast cancer with fewer thromboembolic events and endometrial cancers with raloxifene; however, raloxifene did not reduce noninvasive cancers as effectively as tamoxifen, so no clear winner has emerged.
Noninvasive Breast Cancer
Breast cancer develops as a series of molecular changes in the epithelial cells that lead to ever more malignant behavior. Increased use of mammography has led to more frequent diagnosis of noninvasive breast cancer. These lesions fall into two groups: ductal carcinoma in situ (DCIS) and lobular carcinoma in situ (lobular neoplasia). The management of both entities is controversial
Ductal Carcinoma In Situ (DCIS)
Proliferation of cytologically malignant breast epithelial cells within the ducts is termed DCIS. Atypical hyperplasia may be difficult to differentiate from DCIS. At least one-third of patients with untreated DCIS develop invasive breast cancer within 5 years. For many years, the standard treatment for this disease was mastectomy. However, treatment of this condition by lumpectomy and radiation therapy gives survival that is as good as the survival for invasive breast cancer treated by mastectomy. In one randomized trial, the combination of wide excision plus irradiation for DCIS caused a substantial reduction in the local recurrence rate as compared with wide excision alone with negative margins, though survival was identical in the two arms. No studies have compared either of these regimens to mastectomy. Addition of tamoxifen to any DCIS surgical/radiation therapy regimen further improves local control. Data for aromatase inhibitors in this setting are not available.
Several prognostic features may help to identify patients at high risk for local recurrence after either lumpectomy alone or lumpectomy with radiation therapy. These include extensive disease; age more than 40; and cytologic features such as necrosis, poor nuclear grade, and comedo subtype with overexpression of erbB2. Some data suggest that adequate excision with careful determination of pathologically clear margins is associated with a low recurrence rate. When surgery is combined with radiation therapy, recurrence (which is usually in the same quadrant) occurs with a frequency of 10%. Given the fact that half of these recurrences will be invasive, about 5% of the initial cohort will eventually develop invasive breast cancer. A reasonable expectation of mortality for these patients is about 1%, a figure that approximates the mortality rate for DCIS managed by mastectomy. Although this train of reasoning has not formally been proved valid, it is reasonable to recommend that patients who desire breast preservation, and in whom DCIS appears to be reasonably localized, be managed by adequate surgery with meticulous pathologic evaluation, followed by breast irradiation and tamoxifen. For patients with localized DCIS, axillary lymph node dissection is unnecessary. More controversial is the question of what management is optimal when there is any degree of invasion. Because of a significant likelihood (10–15%) of axillary lymph node involvement even when the primary lesion shows only microscopic invasion, it is prudent to do at least a level 1 and 2 axillary lymph node dissection for all patients with any degree of invasion; sentinel node biopsy may be substituted. Further management is dictated by the presence of nodal spread.
Proliferation of cytologically malignant cells within the lobules is termed lobular neoplasia. Nearly 30% of patients who have had adequate local excision of the lesion develop breast cancer (usually infiltrating ductal carcinoma) over the next 15–20 years. Ipsilateral and contralateral cancers are equally common. Therefore, lobular neoplasia may be a premalignant lesion that suggests an elevated risk of subsequent breast cancer, rather than a form of malignancy itself, and aggressive local management seems unreasonable. Most patients should be treated with tamoxifen for 5 years and followed with careful annual mammography and semiannual physical examinations. Additional molecular analysis of these lesions may make it possible to discriminate between patients who are at risk of further progression and require additional therapy and those in whom simple follow-up is adequate.
Male Breast Cancer
Breast cancer is about 1/150th as frequent in men as in women; 1720 men developed breast cancer in 2006. It usually presents as a unilateral lump in the breast and is frequently not diagnosed promptly. Given the small amount of soft tissue and the unexpected nature of the problem, locally advanced presentations are somewhat more common. When male breast cancer is matched to female breast cancer by age and stage, its overall prognosis is identical. Although gynecomastia may initially be unilateral or asymmetric, any unilateral mass in a man over the age of 40 should receive a careful workup including biopsy. On the other hand, bilateral symmetric breast development rarely represents breast cancer and is almost invariably due to endocrine disease or a drug effect. It should be kept in mind, nevertheless, that the risk of cancer is much greater in men with gynecomastia; in such men, gross asymmetry of the breasts should arouse suspicion of cancer. Male breast cancer is best managed by mastectomy and axillary lymph node dissection (modified radical mastectomy). Patients with locally advanced disease or positive nodes should also be treated with irradiation.
Approximately 90% of male breast cancers contain estrogen receptors, and approximately 60% of cases with metastatic disease respond to endocrine therapy. No randomized studies have evaluated adjuvant therapy for male breast cancer. Two historic experiences suggest that the disease responds well to adjuvant systemic therapy, and, if not medically contraindicated, the same criteria for the use of adjuvant therapy in women should be applied to men.
The sites of relapse and spectrum of response to chemotherapeutic drugs are virtually identical for breast cancers in either sex.
Follow-Up of Breast Cancer Patients
Despite the availability of sophisticated and expensive imaging techniques and a wide range of serum tumor marker tests, survival is not influenced by early diagnosis of relapse. Surveillance guidelines are given in Table 86-5.