Inl J. Radmron Oncologic Biol Phys, Vol. Printed in the U.S.A. All rights resewed.

23. pp.

1071-1078

0360.30 Copyright 8 1992

t .oO

I6/92 $5.00 Pergamon Press Ltd.

0 Oncology Intelligence PRE-CANCEROUS BREAST LESIONS: IMPLICATIONS FOR BREAST CANCER PREVENTION TRIALS MONICA MORROW,

M.D.

Department of Surgery, University of Chicago The relationship between benign breast disease and breast cancer has become the focus of increased clinical attention as breast cancer prevention becomes a clinical reality. In this setting an understanding of the magnitude of the increase in risk conferred by the various types of benign breast disease assumes increasing importance. When benign breast disease is considered as an aggregate, estimates of the relative risk of breast cancer development range from 0.8 to 4.5. This article describes the risk associated with proliferative and non-proliferative benign breast disease, as well as the rationale for considering lobular carcinoma in situ and ductal carcinoma in situ (in some cases) as risk factors for breast carcinoma rather than actual malignant lesions. Mathematical models can provide a more precise estimate of breast cancer risk, but these efforts may be confounded by the lack of uniformity in the pathologic definition of borderline benign breast lesions. The breast cancer prevention trials offer a unique opportunity to improve our database on the natural history of high risk benign breast lesions, while attempting to reduce the 44,000 deaths occurring annually from breast cancer. Breast cancer risk, Benign breast disease, Atypical hyperplasia,

index breast in this report. More recently, relative risks ranging from less than one to 4.5 have been reported in women with benign breast disease (7, 9, 12, 28, 44, 47, 54,66, 69) (Table 1). Additional observations from more recent studies include the fact that the risk of cancer development varies with the degree of atypia of the benign lesion (36), that the increased risk persists for at least 30 years after biopsy (44), and the risk of breast carcinoma is bilateral, even with unilateral benign disease (11). The discrepancies in risk reported in these studies may be explained by the variability in the patient populations studied, a lack of uniform histopathologic definitions for benign lesions or the failure to review biopsy material, varying lengths of follow-up, the use of inappropriate control populations, and the failure to control for other known breast cancer risk factors. Considerable progress in resolving the confusion surrounding benign breast disease and breast cancer risk has been made in recent years, largely through the work of DuPont and Page (13, 14, 15, 49, 30). DuPont and Page reviewed 10,366 benign breast biopsies, and classified them according to strict pathologic definitions as nonproliferative lesions, proliferative lesions, and proliferative lesions with atypia ( 14). The incidence of subsequent invasive breast carcinoma was assessed at a median followup of 17 years and compared to reference data for breast

A number of factors which increase the risk of breast cancer development have been identified. These include both endogenous characteristics, such as family history and reproductive history, and exogenous influences, such as diet, hormone usage, and radiation exposure. Anatomic factors such as benign breast diseases have also been suggested to increase breast cancer risk. In this review, the data on breast cancer risk associated with both benign lesions and “borderline lesions” such as lobular carcinoma in situ and ductal carcinoma in situ, will be discussed, and the rationale for the current trials of breast cancer prevention reviewed. BENIGN

BREAST

Borderline breast lesions.

DISEASE

The importance of benign breast disease as a risk factor for breast carcinoma has long been a subject of controversy. In 1964 Davis et al. ( 10) reviewed the available literature on benign breast disease and cancer risk and noted a relative risk of 3 to 5 in women with biopsy proven benign disease. The authors reported a relative risk of 1.7 for 3 17 women with benign disease from their own practice. When the benign disease was further classified as hyperplastic or non-hyperplastic, the relative risks were 2.5 and 1.2, respectively. Of note, subsequent carcinomas occurred as frequently in the contralateral breast as the

Address for reprints: 584 I South Maryland Ave., MC 503 1, Chicago, IL 60637.

Accepted for publication

1071

27 March 1992.

1072

I. J. Radiation Oncology 0 Biology 0 Physics

Table 1. Previous

benign breast disease and breast cancer risk

Author

Relative risk

Brinton (7) Wynder (69) Veronesi (66) Paffenbarger (47) Hutchinson (28) Monson (44) Roberts (54) Coombs (9) Duffy ( 12)

0.8 1.0 1.6 1.6 2.1 2.5 2.7 3.0 4.5

cancer incidence in women in Atlanta in the Third National Cancer Survey as well as a reference population from the Connecticut Tumor Registry. The histologic diagnoses grouped under the headings of proliferative and nonproliferative are shown in Table 2. Women with proliferative disease were found to have a relative risk of breast cancer of 1.9, and the subcategory of women with atypical hyperplasia were found to have a relative risk of 4.4. Nonproliferative breast disease was associated with no excess risk of breast cancer. The risk of breast cancer development after a diagnosis of proliferative breast disease with or without atypia was found to be greatest in the first 10 years after biopsy (13), with the relative risk associated with atypical hyperplasia halved after this interval, and the relative risk for women with proliferative disease without atypia returning to the level of the index population. A marked interaction between atypia and a family history of a first degree relative with breast cancer was also noted by DuPont and Page (14, 49). This subgroup of women had a risk 11 times that of women with nonproliferative disease. The absolute risk of breast cancer development in women with a positive family history and atypical hyperplasia was 20% at 15 years compared to 8% in women with atypical hyperplasia and a negative family history of breast carcinoma. The presence of proliferative disease with or without atypia was a significant risk factor for breast cancer regardless of patient age at the time of biopsy. In addition to a family history of breast cancer in a first degree relative, the presence of calcifications in a

Table 2. Classification Non-proliferative Adenosis Cysts, macro or micro Duct ectasia Fibroadenoma Fibrosis Mastitis Metaplasia, apocrine or squamous Mild hyperplasia

of benign breast diseases Proliferative Moderate or florid hyperplasia Papilloma with fibrovascular core Atypical hyperplasia, ductal or lobular

Volume 23, Number 5, 1992

biopsy specimen showing atypical hyperplasia also significantly increased the relative risk of breast cancer from 4.0 to 6.5. No increased risk of breast cancer development was observed in women with a diagnosis of proliferative disease who used estrogens after their breast biopsy (15). In an additional report on atypical hyperplastic lesions of the breast (49), Page and coauthors noted that the interval from the diagnosis of atypical ductal hyperplasia to the development of carcinoma was 8.2 years, and only 56% of the carcinomas occurred in the index breast. In contrast, in women with atypical lobular hyperplasia the interval to cancer development was 11.9 years and 69% of the carcinomas occurred in the same breast as the index biopsy. The overall incidence of carcinoma did not differ between women with atypical ductal and atypical lobular hyperplasia, and the presence of a first degree relative with breast carcinoma doubled the relative risk of cancer development for either category of atypical hyperplasia. The classification of benign breast lesions as nonproliferative, proliferative, or proliferative with atypia for the purpose of assigning cancer risk was subsequently adopted by the American College of Pathologists at a consensus meeting in October 1985 (29). In the consensus statement nonproliferative lesions were assigned no increased risk of breast cancer development, proliferative lesions without atypia a relative risk of 1.5 to 2, and atypical hyperplasia a relative risk of 5. Since the development of this statement, several studies (30,4 1) have shown that sclerosing adenosis is associated with an increased risk of breast cancer and should be classified as a proliferative lesion without atypia rather than a nonproliferative lesion. In the original work of DuPont and Page (69), the study population consisted of women with clinically evident breast lesions. In this group, 69.7% of biopsies demonstrated nonproliferative disease, 26.7% proliferative disease without atypia, and only 3.6% atypical hyperplasia. A review of 2983 mammographically directed biopsies revealed nonproliferative disease in 68% of cases, proliferative disease in 25%, and atypical hyperplasia in 7% (46). Rubin et al. (59) specifically assessed the incidence of proliferative disease as defined by DuPont and Page in 80 mammographically directed benign breast biopsies and found 26% to have proliferative disease without atypia and 10% atypical hyperplasia. In a review of the last 100 benign, mammographically directed biopsies at the University of Chicago, an 18% incidence of atypia and a 35% incidence of proliferative disease was noted. The increased incidence of atypia observed in mammographically directed biopsies may be due to the fact that approximately 50% of these biopsies were done for the finding of microcalcifications. Benign microcalcifications are frequently seen with sclerosing adenosis which is known to be associated with atypical lobular hyperplasia (30). In addition, differences in the pathologic criteria used to define atypia may be responsible for the observed variability in the incidence of atypia (see below).

1073

Pre-cancerous breast lesions 0 M. MORROW

LOBULAR

CARCINOMA

IN SITU

In 1941 Foote and Stewart described a noninvasive form of mammary carcinoma arising from the lobules and terminal ducts which they called lobular carcinoma in situ (19) (LCIS). Their initial report identified three important features of LCIS: (a) The lesion is an incidental microscopic finding and cannot be identified clinically or by gross pathologic examination, (b) Invasive cancers which develop following LCIS may be either infiltrating ductal or infiltrating lobular neoplasms, and (c) The lesion is multifocal in the breast. These observations led Foote and Stewart to conclude that LCIS was a premalignant lesion best treated by simple mastectomy. Additional information accrued since the report of Foote and Stewart suggests that LCIS is a risk factor for breast cancer rather than a pre-malignant lesion. The true incidence of LCIS is difficult to determine since it lacks mammographic or clinical features. The histologic changes of LCIS are noted in from 0.8 to 8% of breast biopsies (22, 6 1, 68), with an incidence of 2.5% thought to reflect the frequency of the lesion when all breast biopsies are considered. Lobular carcinoma in situ is noted to be more frequent in premenopausal women, with a mean age of 44 to 46 reported in several series (37, 6 1, 68). The frequency with which LCIS is diagnosed appears to be increasing, with a 15% rise in the number of cases seen from 1973 to 1988 reported in one series (39). This apparent increase in the incidence of LCIS may simply be due to the increased number of breast biopsies done for mammographic abnormalities, since although LCIS lacks mammographic signs (50), it accounted for 14% of mammographically detected cancers in one large series (63) or may reflect increased awareness of the histologic features of LCIS. The major issue in the management of LCIS is whether the LCIS is a premalignant lesion or a marker of increased risk for the development of carcinoma. Mastectomy was the treatment Foote and Steward recommended for LCIS (19) based on the belief that the lesion was a precursor of invasive carcinoma. A significant body of data exists which supports the idea that LCIS is a risk factor rather than a precancerous lesion. The major finding which supports this assertion is the bilaterality of the breast cancer risk after a diagnosis of LCIS, even if the LCIS is present in only one breast (22, 26). McDivitt and coauthors (40) followed 40 women with LCIS and reported a 30% risk of ipsilateral breast carcinoma at 15 years and a 15% incidence of contralateral breast cancer in the same time period. Rosen er al. (57) reported 99 patients followed for a mean of 24 years and found an equal incidence of carcinoma in the index breast and the contralateral breast, while Haagensen reported 2 11 women with LCIS followed for a mean of 14 years with a 17% incidence of invasive carcinoma, equally distributed between the index and the contralateral breast (26). Andersen reviewed 228 published

cases of LCIS treated by biopsy alone and found that 15.5% developed invasive carcinoma of the ipsilateral breast and 9.3% had a contralateral carcinoma (2). The interval to the development of carcinoma after a diagnosis of LCIS can be quite protracted. Rosen (57) noted that the majority of subsequent carcinoma in his series occurred 15 or more years after the diagnosis of LCIS, and 38% presented 20 or more years after the initial biopsy. In contrast to the data presented for atypical hyperplasia, the increased risk of breast cancer seen with LCIS does not diminish with follow-up through 25 years. Estimates of relative risks of breast cancer in women with LCIS range from 6.9 to 12 (2,26,57). Efforts to identify features of LCIS associated with a higher likelihood of the development of malignancy have been largely unsuccessful. Haagensen noted that the relative risk of breast cancer increased from 5.7 for women with LCIS alone to 8.5 in women with both a positive family history and LCIS (26), an association similar to that observed for atypical hyperplasia and a family history of breast carcinoma (14). Histologic features, including the amount of LCIS present, have not been predictive of the subsequent development of invasive carcinoma (57). Based on the preceding information, a logical management option for the woman with LCIS is careful observation, as would be carried out for any woman known to be at increased risk for the development of breast cancer due to a personal history of breast cancer or a positive family history. For the majority of women, this will be an acceptable option. In women unwilling to accept the 20-30% risk of the development of breast cancer, surgical therapy must be directed toward both breasts. Bilateral total (simple) mastectomy, usually with immediate breast reconstruction, is the treatment of choice if prophylactic therapy is elected. Axillary dissection is not indicated. Radiotherapy has no role in the management of LCIS, and if observation is carried out it is unnecessary to obtain histologically negative margins since LCIS is known to be a diffuse lesion. Treatment strategies which address one breast, such as ipsilateral mastectomy with contralateral biopsy, would seem to be illogical, since the risk of LCIS is bilateral regardless of the findings of the contralateral biopsy. Women with LCIS are excellent candidates for breast cancer prevention trials due to their high risk status and persistence of this risk over time. DUCTAL CARCINOMA

IN SITU

Ductal carcinoma in situ (DCIS) has traditionally been considered to be a precursor of invasive breast carcinoma rather than a risk factor for the disease, and in the past, most patients with DCIS were treated with mastectomy. As a result, little information is available on the natural history of DCIS. In addition, most older series (35, 65, 67) consist of patients with palpable DCIS. Gross or palpable DCIS is a relatively uncommon lesion, accounting

1074

1. J. Radiation Oncology 0 Biology 0 Physics

for only 2% of 10,054 breast cancers in an American College of Surgeons Survey in 1987 (56). The incidence of DCIS appears to be increasing (27, 39), and much of the DCIS seen in 1992 is detected as a result of screening mammography and is clinically occult. Whether the malignant potential of microscopic DCIS is equivalent to that of gross DCIS is unknown. In spite of these uncertainties, the available data suggests that some, but not all, DCIS progresses to invasive carcinoma. Long term follow-up is available for several small series of women found to have DCIS on review of biopsies originally classified as benign. Page et al. (48) reported 25 cases of untreated DCIS, with a mean follow-up of 16 years. Seven women subsequently developed invasive carcinoma (28%). This represented a relative risk of 11 compared to age matched controls from the Third National Cancer Survey for White Women in Atlanta. Rosen (58) reported 30 women with untreated DCIS, with complete follow-up available for only 15. Seven invasive cancers occurred at a mean of 9.7 years after the diagnosis of DCIS, an incidence of 27% if all cases are included or 53% if only patients with complete follow-up are considered. In the reports of both Page and Rosen, all carcinomas were in the index breast, usually in the vicinity of the biopsy site. It is important to point out that only noncomedo type DCIS was included in these series, and the lesions studied were low grade, representing one end of the histologic spectrum of DCIS. In a similar report, Eusebi and co-authors (16) described 28 cases of DCIS with an 11% incidence of invasive carcinoma at a median follow-up of 16.7 years. One of the three subsequent carcinomas in this study was in the contralateral breast. Insight into the biologic behavior of more florid forms of DCIS can be gained by a review of the results of treatment of DCIS by excision alone. These studies differ from the preceding ones in that the lesion was recognized as DCIS at the time of treatment. However, patient selection criteria vary considerably and a uniform effort to achieve negative margins of resection was not made. Gallagher and co-authors (24) reported 13 women with DCIS treated by excision alone. The median tumor size was 1 cm, and median follow-up was 100 months. Margins were retrospectively assessed and felt to be negative in all cases. Five local recurrences (38%) were noted at a median of 47 months from excision. Four of the 8 women followed for greater than 9 years developed local recurrence. Two additional reports with long term follow-up after excision alone are available. Millis and Thynne (43) treated eight women with wide excision; at a mean follow-up of 10 years, two (25%) had recurred. A 63% incidence of breast recurrence was noted by Price et al. in 35 women followed a median of 108 months after excision alone (52). Several more recent studies with shorter follow-up offer conflicting data on the propensity of DCIS to progress to invasive carcinoma. A review of the pathologic materials of women entered into the National Surgical Adjuvant Breast Pro-

Volume 23, Number 5, 1992

Protocol B06 (18) identified 22 women with DCIS which was initially misdiagnosed as invasive carcinoma, who were treated with wide excision alone. Histologically negative margins, inked at the time of excision, were required for study entry. All of the tumors but one were clinically evident, and the mean tumor size was 2.2 ? 1.3 cm. At a mean follow-up of only 39 months, a 23% recurrence rate was observed in these patients. In contrast, Lagios and co-workers (38) reported 79 carefully selected women with DCIS followed a median of 44 months with a breast recurrence rate of only 10%. The tumors in this study were detected mammographically, and careful pathologic and radiographic correlations were done for measurements of tumor size (median 0.6 cm) and assessment of margins. Although the incidence of local recurrence varies in the studies discussed, the pattern of local failure is similar. Approximately 50% of the recurrences seen after treatment of DCIS by excision alone will be invasive carcinoma, and failure is most likely to occur at or adjacent to the biopsy site. Several authors have attempted to identify prognostic factors which might predict the malignant potential of DCIS. Histologic subtype of DCIS is one factor which has been suggested to be of prognostic significance. The comedo type of DCIS is more frequently associated with microinvasion (27), as well as an elevated thymidine labelling index (42), expression of the p2 1 ras gene (53), and amplification of the HER-2 neu oncogene (64), all of which are postulated as markers of tumor aggressiveness. Two clinical studies have suggested a higher rate of recurrence after excision of comedo type DCIS when compared to the non-comedo type (4, 38). Cellular DNA content (aneuploid vs diploid) has also been suggested as a method for predicting the biologic aggressiveness of DCIS (8). The “best” treatment for DCIS remains a matter of controversy, largely due to the lack of data on large numbers of patients with precisely classified lesions treated in the same fashion and followed for an extended period of time. Mature data from the NSABP and EORTC trials, when available, will help resolve this dilemma. While awaiting this information, it is reasonable to consider that some women with DCIS, particularly those with non-comedo, clinically occult, localized lesions have a risk for developing invasive breast cancer which is similar to that seen with other conditions considered to be high risk rather than actual malignancy. ject’s

PROBLEMS

WITH IDENTIFICATION RISK WOMEN

OF HIGH

Until very recently, the major problem with the identification of high risk women was that no intervention which might decrease breast cancer risk was available. Efforts were directed toward early detection of cancer in the absence of proven methods of prevention. The clas-

1075

Pre-cancerous breast lesions 0 M. MORROW

sification of a woman as high risk is frequently a cause of anxiety for both the patient and her physician, as evidenced by a British study in which high risk women underwent five times the number of breast biopsies as expected, with a predictive value of only 20% (54). In addition, there is a tendency to give increased weight to risk due to benign or borderline breast pathology in the belief that such pathology represents disease, when a similar degree of risk due to reproductive factors and family history may be ignored. Unfortunately, considerable variation in the criteria used to classify a lesion as atypical hyperplasia or in situ carcinoma exists. This was emphasized in a recent study by Rosai (55) in which five breast pathologists, acknowledged as having a special interest in the field, were asked to examine 17 proliferative breast lesions and render a diagnosis. In none of the cases reviewed did all five pathologists agree on the diagnosis, and in only three cases did 4 of the 5 concur. Perhaps even more alarming, in one-third of the cases the diagnosis ranged from hyperplasia without atypia to carcinoma in situ. This degree of interobserver variability is clearly an issue if women are being classified as high risk on the basis of a single risk factor such as atypical hyperplasia. The assessment of risk and the interaction of various risk factors is poorly understood by most clinicians. In addition, most reports evaluating risk factors have considered only the factor under study, providing no information on its interaction with other risk variables. Gail and co-authors (23) have developed a model which integrates the risk associated with a family history of breast carcinoma, age at menarche, age at first birth, and number of breast biopsies to provide an individualized risk estimate over a number of time intervals. For example, from this model it can be estimated that a 40 year old woman with menarche at age 13, a history of one previous breast biopsy, no children, and one relative with breast cancer has a 6% chance of developing breast cancer in the next 10 years and a 30 year risk of slightly less than 20%. This model does not subclassify benign breast disease as atypical hyperplasia, hyperplasia, or nonproliferative and is not applicable to women with LCIS or DCIS. The authors suggest that a correction in the model in women known to have atypical hyperplasia might further improve its sensitivity. An accurate estimation of an individual woman’s breast cancer risk is essential before considering any

Table 3. Incidence

of second primary

intervention, whether prophylactic surgery or entry into a breast cancer prevention trial. CURRENT TRIALS OF BREAST CANCER PREVENTION Epidemiologic evidence clearly indicates that hormonal factors are important in breast cancer risk, and hormonal manipulations are known to slow breast cancer growth in some women, suggesting that endocrine manipulation could be effective in breast cancer prevention. Experiments in mice infected with the mouse mammary tumor virus (33) or rats treated with the mammary carcinogens DMBA (32) and NMU (54) demonstrate that treatment with the anti-estrogen tamoxifen is effective in preventing or reducing the development of mammary carcinoma. This experimental evidence suggesting that tamoxifen may be effective as a chemopreventive agent is supported by data on the incidence of second primary breast cancers in women with Stage I and II breast cancer treated with adjuvant tamoxifen. In the Stockholm trial, 1846 postmenopausal women under age 7 1 were randomized to no endocrine therapy or 2 or 5 years of tamoxifen at a dose of 40 mg daily (60). At a median follow-up of 7 years, a 40% reduction in the incidence of contralateral breast cancers was observed in the tamoxifen treated women when compared to controls. No differences in contralateral cancer incidence were observed between women treated for 2 or 5 years with tamoxifen. Tamoxifen at a dose of 20 mg daily has also been shown to decrease breast cancer risk. In NSABP protocol B 14,2,892 pre- and postmenopausal women with node negative breast cancer were treated with tamoxifen or placebo for at least 5 years. At a mean follow-up of 59 months, a 50% reduction in contralateral breast cancer was seen in the tamoxifen treated women (17). Similar reductions were observed in the Cancer Research Campaign Trial (1) and the Scottish trial (6) and are summarized in Table 3. Based on these findings, two trials using tamoxifen as a chemopreventive agent have been initiated. Powles et al. (5 1) randomized a pilot group of 435 women, defined as high risk by virtue of a family history of breast cancer, with at least one first degree relative who had developed breast cancer under the age of 40, or bilateral breast cancer at any age, or at least two

breast cancers in adjuvant

tamoxifen

trials # Cancers

Trial Stockholm (55) NSABP (56) Scottish (58) trial Cancer (57) Research

Dose

Campaign

40mgqdX2orSyrs 20 mg qd X 5 years 20 mg qd X 5 years 20 mg qd X 2 years

Control 47 52 14 17

Tamoxifen 29 28 8 7

% Reduction 40 46 43 59

1076

1.J. Radiation Oncology 0 Biology 0 Physics

Table 4. Possible

side effects of tamoxifen

Estrogenic _ Thromboembolism Uterine carcinoma Hepatocellular carcinoma

Antiestrogenic Osteoporosis Atherosclerosis

Other Ophthalmic first degree relatives with breast cancer at any age. The participants ages ranged from 30 to 66 years. Initial compliance data revealed a greater than 80% compliance rate at 12 months and no significant difference in acute toxicity between tamoxifen and placebo other than an increased incidence of hot flashes in the tamoxifen treated women. In April 1992 the National Surgical Adjuvant Breast Project (NSABP) opened a double blind trial of tamoxifen versus placebo with an accrual goal of 16,000 high risk women over the next two years. Those eligible for entry into the study include any woman over the age of 60, or women between the ages of 35 and 59 whose 5-year risk of developing breast cancer, as predicted by the Gail model (23), equals that of a 60 year old woman. Any woman over age 35 with a diagnosis of LCIS is eligible for study entry, while women with DCIS are excluded due to controversy over the nature of this lesion. A number of concerns about the possible toxicity of tamoxifen have been raised. These can be classified as antiestrogenic effects, estrogenic effects, and effects apparently unrelated to the hormonal actions of tamoxifen and are summarized in Table 4. Of interest, available data suggest that tamoxifen does not accelerate osteoporosis (2 1,25,5 1) or increase serum lipid levels (3,5). Significant increases in the incidence of endometrial carcinoma have reported at tamoxifen doses of 20 mg daily (34), but were

Volume 23, Number 5, 1992 seen in the Swedish trial where 40 mg doses were employed (20). No increase in the incidence of hepatocellular carcinoma in humans has been observed to date (34). Careful monitoring of toxicity is a crucial aspect of the tamoxifen prevention trials to determine if the risk-benefit ratio is acceptable. Chemoprevention with a second group of compounds, the retinoids, is also under study. Experimental evidence in animals suggests that the retinoids may be effective in breast cancer prevention (45, 62). A trial using N-Chydroxyphenyl retanamide has been undertaken by the National Cancer Institute in Milan, Italy. Three thousand women with Tl or T2 node negative breast cancers not receiving chemohormonal therapy were selected as the high risk population. At present, no data are available from this study.

CONCLUSIONS

The accurate assessment of risk has assumed increased importance as chemoprevention strategies for breast cancer are studied. Considerable variation exists in the criteria used to define atypia and in situ carcinoma, and major uncertainties persist regarding the appropriate management of these lesions. The chemoprevention trials offer an ideal avenue for improving our data base on the natural history of high risk lesions while assessing potential interventions. However, it should be remembered that for the vast majority of “high risk” women, their lifetime risk of breast cancer development is considerably less than 50%. Thus, any breast cancer development is considerably less than 50%, and any chemopreventive agent must have a low incidence of toxic side effects. Until detailed toxicity data is available, it is essential that the use of tamoxifen for chemoprevention be limited to trial participants.

REFERENCES 1. Abram, W.; Baum, M.; Berstock, D.; et al. Cyclophosphamide and tamoxifen as adjuvant therapies in the management of breast cancer. Preliminary analysis by the CRC Adjuvant Breast Trial Working Party. Br. J. Cancer 57:604607; 1988. 2. Andersen, J. Lobular carcinoma in situ of the breast. An approach to rational treatment. Cancer 39:2597-2602; 1977. 3. Bagdate, J.; Wolter, J.; Subbaiah, P.; Ryan, W. Effects of tamoxifen treatment on plasma lipids and lipoprotein lipid composition. J. Clin. Endocrinology Metabolism 70: 1 l321135; 1990. 4. Baird, R.; Worth, A.; Hislop, G. Recurrence after lumpectomy for comedo-type intraductal carcinoma of the breast. Am. J. Surg. 159:479-48 1; 1990. 5. Bertelli, G.; Pronzato, P.; Amoroso, D.; et al. Adjuvant tamoxifen in primary breast cancer: Influence on plasma lipids and antithrombin III levels. Breast Cancer Res. Treat. 12: 307-310; 1988. 6. Breast Cancer Trials Committee. Adjuvant tamoxifen in the management of operable breast cancer: The Scottish Trial. Lancet 2:171-175; 1987.

7. Brinton, L.; William, R.; Hoover, R.; et al. Breast cancer risk factors among screening program participants. JNCI 62~37-44; 1979. 8. Carpenter, R.; Gibbs, N.; Matthews, J.; Cooke, T. Importance of cellular DNA content in pre-malignant breast disease and pre-invasive carcinoma of the female breast. Br. J. Surg. 74:905-906; 1987. 9. Coombs, L. J.; Lilienfeld, A.; Bross, I.; Burnett, W. A prospective study ofthe relationship between benign breast diseases and breast carcinoma. Prev. Med. 8:40-52; 1979. IO. Davis, H.; Simons, M.; Davis, J. Cystic disease of the breast: Relationship to carcinoma. Cancer 17:957-978; 1964. Il. Donnelly, P.; Baker, K.; Carney, J.; O’Fallon, W. Benign breast lesions and subsequent breast carcinoma in Rochester, Minnesota. Mayo Clinic Proc. 50:650-656; 1975. 12. Due, S.; Roberts, M.; Elton, R. Risk factors for breast cancer: relevance to screening. J. Epidemiol. Community Health 37:127-131; 1983. 13. DuPont, W.; Page, D. Relative risk of breast cancer varies with time since diagnosis of atypical hyperplasia. Hum. Pathol. 20:723-725; 1989.

Pre-cancerous breast lesions 0 M. MORROW 14. DuPont, W.; Page, D. Risk factors for breast cancer in women with proliferative disease. NEJM 3 12: 146- 15 1; 1985. 15. Dupont, W.; Page, D.; Rogers, L.; Parl, F. Influence of exogenous estrogens, proliferative breast disease, and other variables on breast cancer risk. Cancer 63:948-957; 1989. 16. Eusebi, V.; Foschini, M.; Cook, M.; et al. Long term followup of in situ carcinoma of the breast with special emphasis on clinging carcinoma. Sem. Diag. Pathol. 6:165-173; 1989. 17. Fisher, B.; Costanino, J.; Redmond, C.; et al. A randomized clinical trial evaluating tamoxifen in the treatment of patients with node-negative breast cancer who have estrogen receptor-positive tumors. NEJM 320:479-484; 1989. 18. Fisher, E.; Sass, R.; Fisher, B.; et al. Pathologic findings from the National Surgical Adjuvant Breast Project (Protocol 6)l. Intraductal carcinoma (DCIS). Cancer 57: 197208; 1986. 19. Foote, F.; Stewart, F. Lobular carcinoma in situ: A rare form of mammary carcinoma. Am. J. Pathol. 17:49 1-496; 1941. 20. Fomander, T.; Rutqvist, L.; Cedermark, B.; et d. Adjuvant tamoxifen in early breast cancer: occurrence of new primary cancers. Lancet 1: 117; 1989. 21. Fornander, T.; Rutqvist, L.; Sjoberg, H.; et al. Long term adjuvant tamoxifen in early breast cancer: effect on bone mineral density in postmenopausal women. JCO 8: 10191024; 1990. 22. Frykberg, E.; Santiago, F.; Betsill, W.; O’Brien, P. Lobular carcinoma in situ of the breast. SG&O 164:285-301; 1987. 23. Gail, M.; Brinton, L.; Byar, D.; et al. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. JNCI 8 1: 18791886; 1989. 24. Gallagher, W.; Koerner, F.; Wood, W. Treatment of intraductal carcinoma with limited surgery: Long term followup. J. Clin. Oncol. 7:376-380; 1989. 25. Gottfredsen, A.; Christiansen, C.; Palshof, T. The effect of tamoxifen on bone mineral content in pre-menopausal women with breast cancer. Cancer 53:853-857; 1984. 26. Haagensen, C.; Lane, N.; Lattes, R.; Bodian, C. Lobular neoplasia (so called lobular carcinoma in situ) of the breast. Cancer 42~737-769; 1978. 27. Han-is, J. Clinical management of ductal carcinoma in situ in Breast Diseases. Harris, Hellman, Henderson, Kinne, eds. J. B. Lippincott; 1991:233-239. 28. Hutchinson, W.; Thomas, D.; Hamlin, W.; et al. Risk of breast cancer in women with benign breast disease. JNCI 65: 13-20; 1980. 29. Hutter, R. Consensus meeting: Is “fibrocystic disease” of the breast precancerous? Arch. Pathol. Lab. Med. 110: 17 l3; 1986. 30. Jensen, R.; Page, D.; DuPont, W.; et al. Invasive breast cancer risk in women with sclerosing adenosis. Cancer 64: 19771983; 1989. 31. Jordan, V. C. Chemosuppression of breast cancer with tamoxifen-laboratory evidence and future clinical investigations. Cancer Invest. 6:589; 1988. 32. Jordan, V. C. Effect of tamoxifen (ICI46,474) on initiation and growth of DMBA induced rat mammary carcinomata. Eur. J. Cancer 12:419; 1976. 33. Jordan, V. C.; Labadidi, M.; Mine&i, D. The antiestrogenic and antitumor properties of prolonged tamoxifen therapy in C34/OUJ mice. Eur. J. Cancer 26:718; 1990. 34. Jordan, V. C. Tamoxifen for the prevention of breast cancer. Breast Cancer Prevention. August 1990: l- 12. 35. Kinne, D.; Petrek, J.; Osborne, M.; et al. Breast carcinoma in situ. Arch. Surg. 124:33-36; 1989.

1077

36. Kodlin, D.; Winger, E.; Morgenstem, N.; Chen, V. Chronic mastopathy and breast cancer. A follow-up study. Cancer 39:2603-2607; 1977. 37. Lagios, M. Human breast precancer: Current status. Cancer Surveys 2:383-402; 1983. 38. Lagios, M.; Margolin, F.; Westdahl, P.; et al. Mammographically detected duct carcinoma in situ: frequency of local recurrence following tylectomy and prognostic effect of nuclear grade on local recurrence. Cancer 63:6 18-624; 1989. 39. Lemanne, D.; Simon, M.; Martino, S.; et al. Breast carcinoma in situ: Greater rise in ductal carcinoma in situ vs lobular carcinoma in situ. Proceedings ASCO 10:45; 199 1. 40. McDivitt, R.; Hutter, R.; Foote, F.; Stewart, F. In situ lobular carcinoma. A prospective follow-up study indicating cumulative patient risks. JAMA 20 1:82-86; 1967. 41. McDivitt, R.; Rubin, G.; Stevens, J.; et a/. Benign breast disease histology and the risk of breast cancer. Lab. Invest. 50:62; 1988. 42. Meyer, J. Cell kinetics of histologic variants of in situ breast carcinoma. Breast Cancer Res. Treat. 7: 17 l- 180; 1986. 43. Millis, R.; Thynne, G. In situ intraduct carcinoma of the breast: A long term follow-up study. Br. J. Surg. 62:957962; 1975. 44. Monson, R.; Yen, S.; MacMahon, B.; Shields, W. Chronic mastitis and carcinoma of the breast. Lancet 2:224-226; 1976. 45. Moon, R.; Thompson, H.; Becci, P.; et al. N(4-hydroxyphenyl) retinamide, a new retinoid for prevention of breast cancer in the rat. Cancer Res. 39: 1339- 1346; 1979. 46. Morrow, M. Management of nonpalpable breast masses. PPO Updates 4: l- 11; 1990. 47. Paffenbarger, R.; Kampert, J.; Chang, H. Characteristics that predict risk of breast cancer before and after the menopause. Am. J. Epidemiol. 112:258-268; 1980. 48. Page, D.; DuPont, W.; Rogers, L.; Landenberger, M. Intraductal carcinoma ofthe breast: Follow-up after biopsy only. Cancer 49:75 1-758; 1982. 49. Page, D.; DuPont, W.; Rogers, L.; Rados, M. Atypical hyperplastic lesions of the female breast. A long-term followup study. Cancer 55:2698-2708; 1985. 50. Pope, T.; Fechner, R.; Wilhelm, M.; et al. Lobular carcinoma in situ of the breast: Mammographic features. Radiology 168:63-66; 1988. 51. Poweles, T.; Tillyer, C.; Jones, A.; et al. Prevention of breast cancer with tamoxifen-An update on the Royal Marsden Hospital pilot program. Eur. J. Cancer 26:680-684; 1990. 52. Price, P.; Sinnett, H.; Gusterson, B.; et al. Duct carcinoma in situ: predictors of local recurrence and progression in patients treated by surgery alone. Br. J. Cancer. 61:869872; 1990. 53. Querzoli, P.; Marchetti, E.; Bagni, A.; et al. Expression of p2 1 ras gene products in breast cancer relates to histological types and to receptor and nodal status. Breast Cancer Res. Treat. 12:23-30; 1988. 54. Roberts, M.; Jones, V.; Elton, R.; et al. Risk of breast cancer in women with a history of benign disease of the breast. Br. Med. J. 288:275-278; 1984. 55. Rosai, J. Borderline epithelial lesions of the breast. Am. J. Surg. Path. 15:209-22 1; 199 I. 56. Rosen, D.; Bedwani, R.; Vana, J.; et al. Noninvasive breast carcinoma. Results of a national survey by the American College of Surgeons. Ann. Surg. 192: 139- 147; 1980. 57. Rosen, P.; Lieberman, P.; Braun, D.; et al. Lobular carcinoma in situ of the breast. Am. J. Surg. Path. 2:225-25 1; 1978.

1078

I. J. Radiation Oncology 0 Biology0 Physics

58. Rosen, P.; Braun, D.; Kinne, D. The clinical significance of pre-invasive breast carcinoma. Cancer 46:919-925; 1980. 59. Rubin, E.; Visscher, D.; Alexander, R.; et al. Proliferative disease and atypia in biopsies performed for nonpalpable lesions directed mammogmphically. Cancer 6 1:2077-2082; 1988. 60. Rutqvist, L.; Cedermark, B.; Glas, V.; et al. Contralateral primary tumors in breast cancer patients in a randomized trial of adjuvant tamoxifen. JNCI 83: 1299- 1306; 199 1. 61. Schwartz, G.; Feig, S.; Rosenberg, A.; et al. Staging and treatment of clinically occult breast cancer. Cancer 53: 1379; 1984. 62. Silverman, J.; Katayama, S.; Radok, R.; et al. Effect of shortterm administration of N(4-hydroxyphenyl) all-trans-retinamide on chemically induced mammary tumors. Nutr. Cancer 4:186-191; 1983. 63. Silverstein, M.; Gamagami, P.; Colburn, W.; et al. Nonpalpable breast lesions: Diagnoses with slightly overpene-

Volume 23, Number 5, 1992

64.

65. 66. 67.

68.

69.

trated screen film mammography and hook wire-directed biopsy in 1014 cases. Radiology 171:632-638; 1989. Van de Vijver, M.; Peterse, J.; Mooi, W.; et al. Neu-protein overexpression in breast cancer. NEJM 3 19: 1239- 1245; 1988. Von Rueden, D.; Wilson, R. Intraductal carcinoma of the breast. SG & 0 158:105-l 11; 1984. Veronesi, U.; Pizzacaro, G. Breast cancer in women subsequent to cystic disease. SG & 0 126:529-532; 1968. Westbrook, K.; Gallagher, H. Intraductal carcinoma of the breast: A comparative study. Am. J. Surg. 130:667-670; 1975. Wheeler, T.; Enterline, H.; Roseman, J.; et al. Lobular carcinoma in situ of the breast. Long term follow-up. Cancer 34554-563; 1974. Wynder, E.; MacCormack, F.; Stellman, S. The epidemiology of breast cancer in 785 U.S. Caucasian women. Cancer 41:2341-2354; 1978.