Breast Cancer: Setting Priorities for Effectiveness Research (1990)

The research investment in breast cancer is quite large and continues to grow. Efficacy studies, when done properly, contribute immensely to the knowledge base, although they cannot be used across all the major clinical questions and population groups. One requirement in selecting key patient management topics for effectiveness studies is to understand what unanswered questions remain and whether they are suitable for projects outside the realm of RCTs.

Several clinical subjects are briefly reviewed here because they provide the context within which the committee’s discussions about those unanswered questions took place.³ They include detection and diagnosis, staging, and options for treatment of specific patient groups (e.g., women with different stages of primary disease or women with recurrent disease). These particular subjects raised issues that influenced or led to specific patient management and research recommendations that are discussed more fully later in this report.

Breast cancer can be detected by means of breast self-examination, physical examination by a physician or other health care professional, and mammography. Mammography has the greatest sensitivity as a diagnostic method, and it has been successful in identifying very early cancer, which has an excellent prognosis, even in women as old as 74.

Mammography is widely used as a screening test for asymptomatic women. Recent guidelines recommend annual mammograms for women age 50 and older, but there is controversy about recommendations for women age 40 to 49 and for women 75 and older. Other uncertainties concern the appropriate frequency—annually or less often—of screening mammography for older women.

Nevertheless, perhaps the most important problem in this matter is the paucity of older women who get breast cancer screening at all. Until the passage of the Medicare Catastrophic Coverage Act of 1988, Medicare reimbursed for mammography only when used as a diagnostic test (for a palpable breast mass or other symptom), not when used for screening.⁴ An unknown fraction of mammograms done among Medicare-eligible women is believed to be for screening purposes, even though they may be billed as diagnostic in order to obtain reimbursement.

Screening mammography is alleged to have high false-positive rates, although that rate decreases somewhat for older women. In the United States, false positives generally run from 65 to 80 percent (i.e., only 20 to 35 percent of biopsies for nonpalpable mammographic abnormalities turn up cancer). A high false-positive rate leads to unnecessary anxiety and concern on the part of women who at least temporarily believe they may have breast cancer. It may also lead to unnecessary biopsies being performed on such women. These problems indicate a need for better criteria for biopsy and more cost-effective ways to evaluate breast lesions (through, e.g., fine-needle aspiration biopsy or possibly magnetic resonance imaging).

Experts raise several other issues about mammography as a detection device, including its occasional unacceptability to women, laggardly follow-up of indeterminate mammography results, limitations in the specificity of mammography, problems with access to and reimbursement for mammography (especially in the context of Medicare), and questions about the technical quality of the test and its interpretation in average practice settings.

Staging, a common concept for neoplastic disease, refers to the classification of patients into clinically distinct prognostic categories. The major staging system for breast cancer and other solid tumors is an anatomically based “TNM” system. The classification of patients according to tumor size and local extent of tumor (T), degree of nodal involvement (N), and presence or absence of metastasis (M) can be based on clinical or pathol

ogic measures; the pathologic measures are considerably more accurate. Stages of breast cancer are defined in Table 1.

The anatomical staging system, by categorizing the extent of the tumor in the breast as well as the axillary lymph nodes, helps guide the appropriate use of surgery, radiotherapy, and systemic therapeutic approaches. The staging system divides patients into various prognostic subgroups, with the likelihood of cure and long-term survival decreasing as one moves from Stage I to Stage IV.

The majority of women with breast cancer are diagnosed with Stage I disease (e.g., small primary tumor and histologically negative nodes) or Stage II disease (e.g., operable primary tumor with histologically negative or positive nodes). These patients are potentially curable with localized treatment modalities (surgery or radiation therapy or both).

Some recent evidence, however, suggests that some Stage I and many Stage II women may benefit from adjuvant systemic therapy. For instance, recent work has shown that the failure rate for “local” therapeutic approaches even in patients with Stage I disease may be as high as 35 percent or even 50 percent for some subsets of patients. Thus, the need for effective systemic adjuvant therapy becomes even more compelling.

Women with locally advanced, inoperable disease (Stage III) and those with distant metastases (Stage IV) generally cannot be cured with currently available therapies. Combined modality therapy (chemoendocrine therapy, radiation therapy, surgery) appears to shrink tumors, relieve symptoms, and prolong survival for some women at these stages.

The staging system of breast cancer depends on tumor size and axillary node status as the main prognostic factors. Biologic factors that may identify patients with early breast cancer (i.e., Stages I and II) at high risk for recurrence are also sought. For instance, hormone (estrogen and progesterone) receptors in the primary tumor are important prognostic factors for three major reasons. First, they serve as a marker for the aggressiveness of

the tumor; when estrogen and progesterone receptors are positive, they imply a relatively good prognosis (and, when negative, a poor prognosis). Second, they serve as a marker for the sensitivity of the tumor cell to hormonal manipulation; for instance, tumors that are hormone-receptor positive are likely to respond to therapy with antiestrogen therapy such as tamoxifen. Third, they are important in understanding the biology of breast cancer and the growth of tumor cells.

Other advanced quantitative measures of physiologic and morphologic factors may further improve staging precision. Under investigation are cell kinetics (i.e., proliferative capacity, or percentage of cells in S phase), growth factor receptors, oncogenes (particularly the her/neu oncogene), and more refined ways of assessing morphologic features such as nuclear or histologic grade.

In short, the next decade may bring considerable improvements in predicting prognosis for individual patients. Such advances must, however, be based on large prospective studies and many years of follow-up to confirm the usefulness of new prognostic indicators in breast cancer. Such studies would not be contemplated as an early part of an effectiveness research agenda. For the Medicare population at least, effectiveness studies might contribute to this work (for instance, by hypothesis generation for randomized controlled trials or other prospective studies) to the degree permitted by exploitation of the large HCFA databases supplemented by medical record abstraction and patient follow-up.

The therapeutic approaches to breast cancer differ according to stage and whether the approach is local or systemic. Table 2 outlines possible treatment options, which are discussed in more detail in the remainder of this section.

Modified radical mastectomy and lumpectomy combined with axillary dissection and radiation therapy are the standard surgical treatment options for Stages I and II disease.⁵ Four prospective randomized clinical trials

have failed to document a difference between these two approaches in overall survival, disease-free survival (the amount of time between date of diagnosis and either recurrence or end of the observation period), or local or regional recurrence. In addition, these studies and one additional trial have failed to substantiate the theoretical concern of radiation carcinogenesis, having found no increased incidence of contralateral breast cancer or second non-breast-cancer malignancy in patients receiving radiation.

The essentials of the breast-conserving approach include a complete excision of the primary tumor⁶ and axillary node dissection to evaluate the status of the axillary nodes and to remove grossly positive nodes, a moderate dose of radiation to the entire breast with or without a booster dose of radiation to the primary tumor bed, and (when indicated) systemic therapy with either chemotherapy or hormonal agents. With this approach, three-quarters of all women are able to preserve the breast.

In elderly women undergoing conservative surgery and radiation, local treatment issues include whether a negative axilla should be dissected and whether radiotherapy should routinely follow lumpectomy. Axillary dissection causes postoperative morbidity, such as increased pain and less shoulder mobility, and it prolongs hospitalization. Nevertheless, it is considered a routine part of the surgical management of breast cancer because of its paramount importance in determining the status of the axillary lymph nodes. Clinical physical examination of the axilla can be erroneous in 30 percent of cases; that is, women believed to have negative axillary nodes will be proven to have metastases by histological analysis. Because one of the main indications for adjuvant chemotherapy or hormonal management is positive axillary nodes, accurate assessment of the axilla is essential.

If, however, a decision regarding adjuvant chemotherapy in the elderly patient is not going to be based on the histologic status of the axillary nodes, axillary dissection can be forgone. The National Surgical Adjuvant Breast and Bowel Project clinical trials have shown, after eight years of follow-up, that about 90 percent of women treated with lumpectomy and radiation were free of recurrence of cancer in the same breast, as compared with about 60 percent of those treated only with lumpectomy. Among women who had positive axillary nodes and received lumpectomy and chemotherapy, about 6 percent of those who received radiation had no recurrence as compared with 43 percent of those who did not. This latter finding suggests that chemotherapy alone may not be sufficient to prevent breast cancer recurrences. If the purpose of the conservative approach is to preserve the breast, the omission of radiation after an excisional biopsy results in an unacceptably high recurrence rate even with chemotherapy and may require subsequent mastectomy for salvage, thus defeating the purpose of conserving the breast.

In addition to these issues, data from the National Cancer Institute’s Surveillance, Epidemiology, and End Results Program (SEER)⁷ have dem-

onstrated that survival for elderly women (i.e., women 65 years of age and older) with localized or regional disease is comparable to that of younger women up to eight years after treatment. Despite this finding, elderly patients are less likely to be adequately staged and axillary node dissection is less likely to be performed. Approximately 60 percent of women over the age of 60 to 65 years are not referred for radiation after lumpectomy compared with only 20 percent of younger patients. Evidently the assumption is that the elderly patient may not live long enough to develop a recurrence, but the validity of this assumption can be challenged on the basis of life expectancy. In one study, 38 percent of the patients over the age of 70 developed a breast cancer recurrence when treated with lumpectomy and no radiation compared with 6 percent among those receiving radiation.

Although the treatment of elderly women with breast cancer may be complicated by the presence of other medical conditions, no evidence suggests that they tolerate radiation less well than younger women or have increased complications. Radiotherapy as part of breast conservation may present logistical difficulties to the elderly patient because it requires daily hospital visits for six weeks; such problems should be taken into account but not be a primary consideration in clinical recommendations about therapy. Nevertheless, radiation might be selectively forgone for some elderly women with limited expected life spans and significant coexisting conditions.

In short, for early breast cancer, optimal local-regional control is achieved with complete excision of the primary tumor and radiation. Conservative surgery with radiation produces results equivalent to mastectomy. Because the incidence of breast cancer increases with age beyond 45 years and because women are living longer, the number of elderly women with breast cancer is increasing. Such women should be offered appropriate curative therapies. The traditional notions that less aggressive treatment is warranted in elderly patients with breast cancer because their disease is less aggressive or because their life span is limited must be reexamined.

Treatment options for locally advanced breast cancer (LABC) vary greatly depending on the clinical characteristics of the tumor, hormone receptor status, and patient age and level of activity. Distant metastases represent a major problem in these patients. The emphasis of treatment has been on

systemic therapy, primarily chemotherapy. For patients whose breast tumors are operable, local treatment is mastectomy with or without radiation. When surgical resection cannot be performed, radiation alone may be considered. The treatment of LABC may be complex and frequently requires combined modalities. Accurate staging and other data are crucial in helping to direct patient management decisions and thus are important for valid effectiveness research in this group of patients.

Although a variety of chemotherapeutic and endocrine regimens can cause tumor regression and symptomatic relief in patients with advanced breast cancer, complete recovery is rarely achievable. Skillful use of the systemic therapies in conjunction with radiotherapy and surgery can improve the quality of life and perhaps prolong life. The selection of appropriate therapy for the patient with recurrent breast cancer should be based on a number of factors, including the general medical condition of the patient, the disease-free interval, sites of metastases, rates of progression of the tumor, menopausal status and age of the patient, estrogen and progesterone receptors, and history of prior systemic therapy.

In arriving at a treatment recommendation for a patient with recurrent disease, the physician must be aware of how patients may respond to major treatment modalities as outlined in Table 3. For example, an elderly woman with a disease-free interval of greater than two years, a strongly estrogen-receptor-positive tumor, and an asymptomatic pulmonary metastasis would be an excellent candidate for endocrine therapy with tamoxifen, an anti-estrogen agent. By contrast, a young woman with a disease-free interval of less than one year, a receptor-negative tumor, and multiple sites of visceral metastases that are symptomatic would be a more likely candidate for vigorous systemic cytotoxic chemotherapy. The challenges for the clinician

are the many gradations between these two extremes of clinical presentations of metastatic disease.

Intraductal carcinoma (ductal carcinoma in situ, or DCIS) is a neoplasm that is confined to the duct system of the breast; there is no microscopic evidence of penetration of the basement membrane.⁸ It represents breast cancer at its earliest stage and is characterized by an extremely low incidence of axillary nodal metastases or distant metastases (0 to 5 percent). It is currently detected chiefly by mammogram before the patient is symptomatic. Before routine use of high-quality mammography, DCISs detected by physical examination were usually large or widespread lesions. The growing use of screening mammography has increased the diagnosis of small DCIS, at least in nonelderly women. Because of this, treatment issues have become more visible and complex. In 1978, an American College of Surgeons survey found that DCIS accounted for only 3.2 percent of 10,000 breast cancer cases. However, the Malmö mammographic screening trial reported in 1988 that 16 percent of cancers in the screened group were intraductal.

Simple mastectomy without axillary dissection has been the therapy of choice for DCIS, with a 98 to 100 percent rate of survival. However, review of the small number of cases of microscopic DCIS treated by biopsy alone indicates that only 30 percent of women so treated went on to develop subsequent invasive cancer. The choice of surgical approach is further complicated by the fact that many of the older reports of DCIS are dealing with a gross (palpable) form of DCIS, which appears to behave in a more aggressive fashion than the microscopic form usually found by screening mammogram. Reports of approximately 180 cases of gross DCIS treated with excision and radiation indicate that breast cancer recurrence occurs in 4 to 10 percent of cases, and this appears to have a minimal impact on survival. Further information on treatment of microscopic DCIS will be available with the completion of a current randomized clinical trial that is comparing treatment by excision alone and excision plus radiation.

Thus, at present it is safe to say that the treatment of DCIS is in a state of evolution. When screening mammography becomes a covered Medicare service and more such cases are diagnosed among the elderly, questions relating to the effectiveness of alternative therapeutic approaches are certain to arise.

Questions of patient preferences for quantity versus quality of life are important in many of the management decisions regarding breast cancer. It is imperative that the patient, regardless of whether she has nonmetastatic or metastatic disease, be well informed regarding the potential benefits but also the limitations and the expected morbidity of therapy for her condition. Both medical and philosophic questions must be thoroughly discussed with the patient. An understanding of patient views and preferences about alternative therapeutic strategies and the attendant impacts on quality of life will be critical to an adequate evaluation of the effectiveness of those therapeutic strategies.