the effects of hormonal and chemotherapy on tumoral and nonneoplastic breast tissue

The Effects of Hormonal and Chemotherapy on Tumoral and Nonneoplastic Breast Tissue

SUSAN KENNEDY, MBBCH, MARIA J. MERINO, MD, SANDRA M. SWAIN, MD, AND MARC E, LIPPMAN, MD

One hundred four sequential pretherapy and posttherapy breast

tissue specimens from 57 patients with locally advanced and met-

astatic breast cancer were evaluated in an attempt to define the

effects of systemic chemotherapeutic agents on the tumors and residual nonneoplastic breast tissue. The patients were treated

uniformly at the National Cancer Institute on an experimental

protocol combining systemic chemotherapy with attempted hor-

monal synchronization. Tumors were sampled prior to and fol-

lowing several cycles of chemotherapy to a maximum objective

clinical response (average number of cycles, 7). In 38 cases, the

posttreatment biopsy was positive for tumor. The most striking histologic change was extreme vacuolization of tumor cells that

often resembled histiocytes. Atrophy of the terminal duct lobular

unit (TDLU) and atypia of epithelial cells in TDLU and large

ducts were also seen. Severe degrees of epithelial atypia occa-

sionally proved to be difficult to distinguish from residual intra-

ductal carcinoma. Breast biopsies were stained with antibodies to

cytokeratin, epithelial membrane antigen (EMA), B72.3, lactalbu-

min, and SPl using immunoperoxidase techniques. The number

of cases showing immunoreactivity with antibodies to cytokera-

tin, EMA, and B72.3 remained approximately the same before

and after therapy, while SPl expression decreased and lactalbu-

min expression increased after therapy. Recognition of chemo-

therapeutic changes in breast tissue is important since systemic

chemotherapy plays an important role in the management of

breast cancer. HUM PATHOL 21:192-198. 0 1990 by W.B. Saun-

ders Company.

It has been recognized that resectable breast can- cer may be a systemic disease at initial presentation.’ Systemic chemotherapy has thus become an integral part of the primary treatment in patients with both advanced-stage and operable breast cancer.? Many clinical trials have been completed in the past several years using different combinations of chemotherapy and hormones. The morphologic changes that may occur in the neoplastic and nonneoplastic breast are not reported in detail in these trials, and most studies focus on the clinical response rather than the histopathology.2,J

In order to define histologic changes in both tu- mor and nonneoplastic breast tissue due to chemo- therapy, we studied the pathologic material of 57 pa- tients treated initially with chemotherapy and at- tempted hormonal synchronization. The patients

From the Department of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, MD: and Georgetown University, Washington, DC. Accepted for publication June 30, 1989.

Key words: chemotherapy, hormone therapy, breast cancer. Address correspondence and reprint requests to Maria J. Me-

rino, MD, Laboratory of Pathology, Bldg 10. Room 2N212, Na- tional Cancer Institute, National Institutes of Health, 9000 Rock- ville Pike, Bethesda, MD 20892.

0 1990 by W.B. Saunders Company. 0046.8177/90/2102-0010$.5.00/0

were treated on an identical protocol at the National Cancer Institute (NCI). The patients had an initial diagnostic biopsy and were rebiopsied after several cycles of systemic chemotherapy to a maximal objec- tive clinical response. The second biopsy specimen was obtained prior to the initiation of radiotherapy. We also evaluated the expression of commonly used tumor markers on the pretherapy and posttherapy biopsy specimens to determine whether they could be useful in recognizing morphologically altered tumor cells. The results of our study constitute the basis of this report.

MATERIALS AND METHODS

Our study is based on the pathologic material obtained from patients with stages III or IV locally advanced breast cancer initially diagnosed by incisional biopsy. The patients were clinically staged prior to biopsy using the American Joint Commission-International Umon Against Cancer’ staging system. Clinical information was obtained by re- viewing clinical charts. The patients received primary in- duction chemotherapy with attempted hormonal synchro- nization to a maximum objective clinical and pathologic re- sponse. Patients achieving a pathologic complete response with no evidence of tumor on tissue biopsy received radia- tion therapy, while those with biopsy-proven residual tu- mor underwent mastectomy prior to radiation therapv. Both groups were given consolidative chemotherapy for’6 months. The initial treatment regimen consisted of cyclo- phosphamide (500 mgimm” intravenously [IV]), doxorubi- tin (30 mg/mm” IV), tamoxifen (10 mg for 5 days), pre- marin (three doses of 0.625 mg), methotrexate (300 mgi mm2 IV), 5-fluorouracil (500 mgimm” IV), and leucovorin rescue (six doses of 10 mgimm” orally).

One hundred four prechemotherapy and postchemo- therapy biopsy specimens were available for 57 patients, and mastectomy specimens were available for the 38 pa- tients who had a positive postchemotherapy biopsy. All of the original pretreatment material consisted of tru-cut nee- dle or incisional biopsy specimens, with an average of three hematoxylin-eosin (H&E)-stained slides. The posttreat- ment biopsy specimens consisted of either needle or inci- sional biopsy alone or with subsequent mastectomy speci- mens (38 cases with an average of six H&E-stained slides). Histologic grade was assigned to each tumor using the cri- teria of Fisher et al5 and Black and Speer.” The histologic classification system of McDivitt et al’ was used.

In the cases positive for tumor following treatment (38 cases) selected blocks from both prechemotherapy and postchemotherapy biopsy specimens were assessed by im- munohistochemistry to determine whether there were any differences in antigen expression. Antibodies against epi- thelial membrane antigen (EMA) (Dako, Santa Barbara, CA), cytokeratin (Hybritech, San Diego, CA), B72.3 (Dr J. Schlom, NCI), Leu-Ml (Becton Dickinson, Cockesville. MD), lactalbumin (Dako), lysozyme (Dako), and pregnancy-

192

specific p-glycoprotein (SPl) (Dako) were used. The studies were performed on formalin-fixed, paraffin-embedded material using the avidin-biotin technique of Hsu et al.’

RESULTS

Clinical Data

The mean age of patients at presentation was 47 years (range, 22 to 77 years). Thirty of the patients were postmenopausal, 27 were premenopausal; 49 were white and eight were black. The initial com- plaint of 53 patients was a palpable breast lump; 12 complained of redness and swelling of the breast, eight of breast discomfort, and two of nipple dis- charge. Tlne length of time before diagnostic evalua- tion varied between 2 weeks and 5 years, with a me- dian of 3 months. The size of the tumor ranged from 4 to 13 cm, with a mean size of 6 cm. Palpable nodes were found in 27 patients, five patients had deep fix- ation of the tumor, and five had skin involvement. Prior to in.itiation of chemotherapy, all patients had a negative metastatic work-up to qualify for inclusion in the protocol. Biochemical studies revealed estrogen receptor positivic y in 12 cases.

Histologic Findings

Forty-four of the 57 original biopsy specimens showed mfiltrating ductal carcinoma. Ten were lob- ular carcinomas. one was mutinous, and two were metaplastic. Forty-three of the cases were grade II and 14 were grade II I. Several alterations were con- sistently noted in both the tumor and residual breast tissue after chemotherapy. The most important change was an imiocuous appearance of the residual tumor cells, which were easily mistaken for histiocytes (Fig 1). The c:ytoplasm of these cells was greatly in-

-__ - L , w..’ -’ _ & ; ‘i - ” y-p-_ _-,_f _ ,_.

. 1

:

- n. ;

1. _

__ t - -_ ‘.-__: *.

CHEMOTHERAPY IN THE BREAST [Kennedy et al)

creased in size and had prominent vacuolization con- sisting of multiple tiny or coalescing vacuoles. In some cases, the cells had a brightly eosinophilic gran- ular cytoplasm with well-marked but irregular cellu- lar outlines (Figs 2 and 3). The nuclei were eccentric and hyperchromatic, with an irregular nuclear mem- brane. Chromatin clumping and prominent nucleoli were often seen. Some cells were enucleated. These morphologic alterations were prominent features in 16 of the 38 positive posttherapy cases (47%) (Table 1).

The altered cells appeared for the most part as single cells (Fig 4); they occasionally formed small clumps that infiltrated the fibrous stroma in a subtle fashion, without significant desmoplasia. leading to difficulties in the interpretation, particularly on fro- zen sections or when they were found in association with chronic inflammatory cells, macrophages, or for- eign body giant cells near the previous biopsy site. Similar tumor cells were also seen in the in situ com- ponent in ducts and lobules.

Several histologic changes were consistently noted in the residual noncarcinomatous breast tissue following therapy. Alterations were noted in lobules, large ducts, and stromal compartments. Atrophic changes were seen in lobules, and epithelial atypia was noted both in lobules and ducts.

A spectrum of intralobular fibrosis \Tarying from basement membrane thickening to diffuse lobular at- rophy was seen in 39 of 57 cases (68.4%), with the changes equally frequent in premenopausal and post- menopausal women. The epithelium of the terminal duct lobular unit (TDLU) was inconspicuous or ab- sent where severe atrophy and fibrous obliteration of individual acini had occurred (Fig 5). Lobular atro- phy was diagnosed when the morphologic alterations discussed above had taken place, and estimation of the average number of lobules was two or less per low power field in a premenopausal woman and one or less per two low power fields in a postmenopausal woman. Using the criteria of Haagensen,<’ the aver- age number of lobules per low power field in a pre- menopausal woman varies from five to seven depend- ing on the phase of the menstrual cycle; the average in a postmenopausal woman is two lobules per low power field. In many cases in the premenopausal group, there were only two or three lobules per low power field following therapy. The average number of acini per lobule was decreased as well.

Both lobules and ducts exhibited focal cellular atypia. Atypical epithelial cells were demonstrated in the lobules in 18 of 57 cases (31.5%) (Fig 6). Atypia was rare when severe degrees of atrophy were present. The atypia was manifested by overall en- largement of the cells and loss of uniformity from cell to cell. The cytoplasm was pale and vacuolated or had a granular eosinophilic hue. Nuclei were prominent, with small basophilic nucleoli and thick nuclear mem- branes.

FIGURE 1. Residual tumor cells were found scattered throughoul the breast parenchyma with minimal distortion of the normal ar- chitecture. H&E stain; magnification x 100.)

Similar atypical cells were seen in the large ducts in 50.8%) of the cases. The myoepithelial layer was much less likely to show cytologic atypia than the

193

HUMAN PATHOLOGY Volume 21, No. 2 (February 1990)

FIGURE 2. (Left] The tumor cells resembled histiocytes and had abundant eosinophilic cytoplasm with well-demarcated cellular outlines. (H&E stain; magnification X 250.) (Right) Cells with the same histologic features were found within lymphatic spaces. [H&E stain; magnifi- cation X 250.)

overlying epithelium. Morphologic alterations seen in both myoepithelial and epithelial cells in approxi- mately 20% of ducts and 10% of lobules included clear cytoplasm and a “teardrop” appearance of the luminal lining cells (Fig 7).

Many large ducts were filled with clusters of pig- mented macrophages. Less common findings in- cluded intraductal necrosis and intraductal calcifica- tion.

The stromal alterations in the treated cases were as follows: fibrous involution, fatty involution, in- creased elastosis, inflammatory infiltrate, necrosis, and calcification. A uniform paucicellular stromal fi- brosis with scant atrophic lobules blending into the background was seen in roughly equal percentages of

FIGURE 3. Utilizing the immunoperoxidase technique, the tumor cells stained strongly positive for cytokeratin. (Magnification x 250.)

premenopausal(44.4%) and postmenopausal (43.3%) women. There was almost complete adipose replace- ment of the parenchyma in 43.8% of the cases. A lymphoid infiltrate, usually periductal, was noted in 28% of cases, while periductal and stromal elastosis was noted in 21% of cases.

Findings of immunohistochemical positivity with antibodies to cytokeratin (Fig 3), EMA (Fig 4, right), B72.3, lactalbumin, and SPI lend further support to the epithelial origin of the cells. Analysis of the im- munohistochemical staining of the tumor cells with the various antibodies before and after chemotherapy revealed that the majority of cases stained strongly and diffusely with antibodies to EMA (92%) and cy- tokeratin (85%). This positive staining was retained without any reduction in intensity in the posttreat- ment biopsy specimens. Lactalbumin was positive in more posttreatment (60%) than pretreatment (40%) biopsy specimens, usually staining approximately

TABLE 1. Histologic Alterations in the Nonneoplastic Breast

No. of Cases (75)

Lobular atrophy Prominent cytoplasmic vacuolization Lobular cellular atypia Ductal atypia Cytoplasmic vacuolization Pigmented macrophages Fibrous stromal involution

Premenopausal Postmenopausal

Adipose tissue present Premenopausal Postmenopausal

Elastosis Inflammatory infiltrate Stromal necrosis Stromal calcification

39 (68)

6 (10) 18 (31) 29 (51) 13 (23) 11 (19) 11 (19)

19 of 27 (70) 19 of 30 (63)

12 of 27 (44) 13 of 30 (43)

12 (21) 16 (28)

8 (14) 10 (17.5)

194

CHEMOTHERAPY IN THE BREAST (Kennedy et al)

FIGURE 4. (Left) Some biopsy specimens had dense hialinized fibrous stroma in which smaller single cells were noted. [H&E stain; mag- nification x :250.) (Right) These cells stained positive for EMA and other epithelial markers. [Magnification x 250.)

50% of the cells. 1372.3 stained only a small percent- age of cells when positive (less than 10%); it was pos- itive in 50% of the cases before treatment and in 35%, of the cases after treatment. SPl was expressed in 41 a8 of the cases before treatment and 42%’ of the cases after treatment (Table 2).

Follow-up

Follow-up information was available for 44 of 57 patients for a range of 1 to 5 years (mean, 3 years). Twelve patients died with clinical evidence of residual disease, 17 are alive with clinical evidence of disease, and 15 have no evidence of progressive disease. Dur- ing the course of treatment, most of the premeno-

FIGURE 5. Marked lobular atrophy was a common finding. (H&E stain; magnification x 100.)

pausal patients became amenorrheic. ‘I wenty-four patients exhibited a clinical complete response at the time of the second biopsy, but this was substantiated by histopathology in only 19 of these cases.

DISCUSSION

Until recently, chemotherapy played a minor role in the treatment of early stages of breast cancer. Many clinical trials, however, have been conducted using different combinations of chemotherapeutic drugs and hormones. Recent reports of various che- motherapeutic regimens focus on the clinical re- sponse to therapy, but the morphologic changes that may occur in the residual breast tissue and tumor cells following therapy have not been described in detail.

We studied these changes in a unique group of 57 patients who underwent incisional or needle biop- sies for diagnosis and were then treated with aggres- sive chemotherapy. The histologic changes were stud- ied by comparing initial biopsy specimens with post- therapy tissue biopsy specimens obtained to evaluate therapy efficacy.

The most striking finding was the histiocyte-like appearance of tumor cells following chemotherapy. The cells infiltrated the stroma as single cells in ap- proximately 40% of the cases. This bland appearance caused considerable diagnostic difficulties with both the frozen sections and the tru-cut needle biopsy specimens.

This type of cellular change was most common when an almost complete clinical response had oc- curred and only scattered microscopic foci of tumor remained in a sea of bland fibrous tissue. If the tumor exhibited little overall clinical response in terms of size reduction, the morphology of the tumor closely resembled the pretreatment biopsy. A.ltered tumor

195

HUMAN PATHOLOGY Volume 21, No. 2 (February 1990)

FIGURE 6. (Left] Abnormal epithelial cells with large, sometimes multilobulated, nuclei were noted throughout the ductal system. (H&E stain; magnification x 200.) (Right) In some cases, the atypical ductal changes were easily misdiagnosed on ductal carcinoma in situ. (H&E stain: magnification x 200.)

cells were seen in five biopsy specimens clinically lobules, and the distinction between intraductal car- staged as having a complete response but showing cinoma and epithelial atypia proved the most chal- histologic evidence of tumor on microscopic exami- lenging diagnostic problem. This was an important nation. In these cases, tumor deposits were very focal, distinction, because stratification of further treatment consisting of small collections of isolated single tumor was based on whether the biopsy specimen was diag- cells. nosed as benign or malignant.

Both architectural and cytologic features were useful in confirming the presence of cancer after therapy. Useful cytologic features were the irregular- ity and smudging of the nuclear outlines, hyperchro- masia, and coarse chromatin clumping. Similar cyto- logic abnormalities were seen in malignant ducts and

The morphologic alterations in the atypical epi- thelial cells included an increase in cell size with clear- ing or granularity of the cytoplasm, nucleomegaly, multinucleation, and small nucleoli. The changes in the epithelial cells were similar to those seen in tumor cells, but the nuclear membranes remained sharp; hy- perchromatism and nucleoli were less prominent. These changes were usually focal in both lobules and ducts. The presence of mixed atypical and normal cells, the lack of necrosis, and the retention of a con- tinuous layer of normal myoepithelium were useful criteria in diagnosing atypia rather than carcinoma. When severe cytologic atypia was present and the di- agnosis could not be made with certainty, further needle biopsies or open biopsy were requested to re- solve the issue.

FIGURE 7. Focally, the myoeprthelral cells showed abundant clear cytoplasm and the epithelial cells had a “teardrop” effect. These changes are probably due to hormone therapy. (H&E stain; mag- nification x 166.1

The teardrop appearance or vacuolization of the cytoplasm, which was seen in 22.8% of cases, is prob- ably an effect of the hormonal therapy, similar to that

TABLE 2. lmmunohistochemical Analysis”

Antibody Therapy

Initial Biopsy (%)

Biopsy Following Therapy (‘%)

Cytokeratin 92 92 EMA 85 77.5 B72.3 50 35 Lactalbumin 40 60 SPl 40 24

* Results expressed as percentage of positive cases.

196

described in the luteal or premenstrual phase of the menstrual cvcle “‘,11 These changes have also been I . reported to be present focally in normal breast and in patients receiving antihypertensive or antipsychotic drugs.”

The meffec.ts of chemotherapeutic drugs, particu- larly busulfan, on epithelial and other cell types are well-known. ‘The presence of giant nuclei has been reported in bronchioles, pulmonary alveoli, pancreas, thyroid, adrenal glands, liver, and lymph nodes fol- lowing 2 years of busulfan treatment for chronic granulocvtic leukemia. I3 Cytologic abnormalities, in- cluding those found in the breast epithelium. were reported in an autopsy study of a female patient treated with busulfan for 4 years.‘:‘,“’ Forni et al’” reported the occurrence of significant cancermimetic cellular abnormalities in the epithelial surface of the lower urinary tract following treatment with cyclo- phosphamide. No cellular abnormalities were noted in other organs. The changes described by these in- 1,estigato.rs were similar to those in the present series, namely, overall cellular enlargement; irregular, en- larged hyperchromatic nuclei and nucleoli; and cyto- plasmic \acuolization. Interestingly. in that study, the atypia regressed following cessation of treatment.

A recent study of ploidy and cytomorphologic c.hanges on sequential prechemotherapy and post- chemotherapy cytopunctures was reported by Brifford ct al.‘” In that series, tumor cells with enlarged nuclei and nucleoli, nuclear vacuolization, chromatin clear- ing, and foamy cytoplasm were noted following ther- apy in 20 of 35 ‘cases. In our series, we found that nuclear vacuolization and chromatin clearing were equally prevalent in pretherapy and posttherapy tu- mor cells.

A spectrum of forms of lobular atrophy was seen In our c,ases. varying from concentric thickening of the basement membrane to almost complete fibrous obliteration and sclerosis of the lobules. In the latter cases, residual acini were lined by scant flattened ep- ithelial cletls. The atrophic changes were quite marked in some of the premenopausal women who had nor- mal-appmearing lobules at the initial biopsy procedure. ‘The noted changes resemble those described during the normal process of involution in menopause. Sim- ilarly, the stromal involution and adipose replace- ment, seen to a significant degree in several premeno- pausal uomen, suggest the possibility that the therapy or resultant ovarian suppression have accelerated the natural involution of the breast. Almost all premeno- pausal patients developed either amenorrhea or oli- gomenorrhea during the course of treatment. Exper- imental evidence has suggested that the effects of ad- juvant chemotherapy in premenopausal women may be partlv mediated by suppression of ovarian function. i 7. Ix

Breast carcinoma is known to consist of a heter- ogenous population of cells, not all of which are af- fected ‘equally by chemotherapy and hormones. Therapeutic interventions attempt to alter the pre- dominant clone or destroy the most rapidly prolifer-

CHEMOTHERAPY IN THE BREAST [Kennedy et al)

ating subtypes. We’” have described alterations in tu- mor cell DNA content by image analysis. These changes included loss of aneuploidy and emergence of a new aneuploid population following therapy. The presence of tumor markers may prove to be a useful index of cellular differentiation.

Several studies have attempted to correlate pa- tient survival rate with tumor markers. For instance, it has been suggested that strong immunohistochem- ical expression in breast cancer of antigens related to EMA is indicative of a better prognosis than is pre- dicted for those cases without such expression. Wilkinson et a12” demonstrated that lack of reactivity with HMFGl, an antibody raised against human milk fat globule, was associated with a poor prognosis. Ellis et a121 have shown a strong positive correlation be- tween tumor cell expression of NCRL I 1, an EMA- type antigen, and improved survival rate; however, other investigators have not been able to confirm these results. Some evidence suggests that the pres- ence of SPl in breast carcinoma is associated with a significantly shorter survival rate”? and worse prog- nosis in patients stratified by histologic grade.23,‘-l

In the present study, the tumor markers cyto- keratin, EMA, B72.3, lactalbumin, and SPI were eval- uated before and after therapy. The results are out- lined in Table 3. Lactalbumin was expressed in a greater number of cases following therapy (60% com- pared with 400/c), and a decrease in SPl expression was shown following therapy (40% to 24%). These results could indicate a chemotherapv-induced in- crease in tumor differentiation. Expression of cyto- keratin did not change with treatment. There was a small decrease in the number of cases immunoreac- tive to anti-EMA antibodies. The morphologically al- tered tumor cells most often showed positivity for cy- tokeratin and EMA. B72.3 proved to be a disappoint- ing marker, as staining was faint and very focal in both pretreatment and posttreatment biopsy speci- mens. No reduction in staining intensity or marked change in the percentage of positive cells was seen with any of the markers following therapy.

The interlobular stroma had a bland homoge- nous appearance. The atypical flbroblasts and vascu- lar changes often seen following radiation therapy were absent. The iatrogenic effects of radiation ther- apy, outlined by Schnitt et al.“” have features in com- mon with the changes induced by systemic therapy, including lobular fibrosis, atrophy, and epithelial atypia. The most striking findings in the series of Schnitt et al involved the ‘T’DLL’. Atypical cells with

TABLE 3. Histologic Types of Carcinoma

Infiltrating ductal 44 L.OhLllar IO Mutinous 1 Metaplastic ‘,

Histologic grade 46 grade I I I4 grade III

Nuclear grade all grad? III Estrogen receptor status 12 positnbe: 45 negative

197

HUMAN PATHOLOGY Volume 21, No. 2 (February 1990)

features similar to those of the present series were described in the TDLLJ. A pseudoinfiltrative pattern was noted in the most severely diseased lobules in which the acini appeared to be squeezed by the fibro- collagenous connective tissue. In our series, the lob- ular fibrosis always appeared bland, causing the acini to disappear into the fibrous background rather than to resemble an infiltrative process.

As chemotherapy becomes an important tool in the treatment of early breast cancer, pathologists will be required to recognize these changes in the treated tissues. Immunohistochemistry for EMA and cyto- keratins is, in our experience, a useful tool to confirm the epithelial nature of cytologically altered tumor cells.

We conclude that changes in normal breast and tumor cells are common following systemic chemo- hormonal therapy and correlative with a decrease in tumor bulk and susceptibility to therapy.

REFERENCES

1. Fisher ER: The impact of pathology on the biologic, diag- nostic, prognostic and therapeutic considerations in breast cancer. Surg Clin North Am 64:1073-1093, 1984

2. Israel L, Breau J-L, Morene JF, et al: Two years of high dose cyclophosphamide and 5-fluorouracil followed by surgery af- ter 3 months for acute inflammatory breast carcinoma. Cancer 5724-28, 1986

3. Morrow M. Braverman A, Thelmo W. et al: Multimodal therapy for locally advanced breast cancer. Arch Surg 12 1: 129 l- 1296, 1986

4. Harmer MH (ed): TNM Classification of Malignant Tu- mours (ed 3). Geneva. UICC, 1978

5. Fisher ER, Redmond C, Fisher B: Histologic grading of breast cancer. Path01 Am 15:239-251, 1980

6. Black MD, Speer FS: Nuclear structure in cancer tissues. Surg Gynecol Obstet 105:97-102. 1957

7. McDivitt RW. Stewart FW, Berg JW: Tumors of the breast, in Atlas of Tumor Pathology, Series 2. Washington, DC, Armed Forces Institute of Pathology, 1969

8. Hsu SM, Raine J, Fanger H. et al: Use of avidin biotin peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (Pap) proce- dures, J Histochem Cytochem 29:577-580. 1981

9. Haagensen CD: Diseases of the Breast, Philadelphia, PA, Saunders, 1986, pp 47-52

10. Longacre TA, Bartow SA: A correlative study of human breast and endometrium in the menstrual cycle. Am J Surg Path01 10:382-393, 1986

11. Vogel PM, Geordiade NC. Fetter BW, et al: The correla- tion of histologic changes in the human breast with the menstrual cycle. Am J Path01 104:23-34, 1981

12. Tavassoli FA, Tienyeh I: Lactational and clear cell changes of the breast in nonlactating, nonpregnant women. Am J Clin Path01 87:23-29, 1987

13. Wailer U: Giant nuclei after myleran therapy and splenic irradiation. Path01 Microbial 23:283-290, 1960

14. Nelson BM, Andrews GA: Breast cancer and cytologic dysplasia in many organs after busulfan. Am J Clin Path01 43:37- 44, 1964

15. Forni AM, Koss LG. Geller W: Cytologic study of the ef- fects of cyclophosphamide on the epithelium of the urinary blad- der in man. Cancer 17:1348-1355, 1964

16. Brifford M, Spyratos F, Tubiana-Huhn M, et al: Sequen- tial cytopunctures durhg preoperative chemotherapy for primary breast cancer. Cancer 63:631-637. 1989

17. Rose DP, Davis TE: Effects of adjuvant chemohormonal therapy on the ovarian and adrenal function of breast cancer pa- tients Cancer Res 40:4043-4047, 1980

18. Samaan NA, DeAsi DN Jr, Buzdar AV, et al: Pituitary- ovarian function in breast cancer patients on adjuvant chemoim- munotherapy. Cancer 41:2084-2087, 1978

19. Kennedy SM, Merino MJ, Swain S. et al: Breast cancer kinetics and ploidy before and after chemotherapy. Lab Invest 60:47A, 1989 (abstr)

20. Wilkinson MJ, Howell A, Harris M, et al: The prognostic significance of two epithelial membrane antigens expressed by hu- man mammary carcinomas. Int J Cancer 33:299-304, 1984

21. Ellis IO, Hinton CP, MacNay J, et al: Immunocytochem- ical staining of breast carcinoma with the monoclonal antibody NCRLll: A new prognostic indicator. Br Med J 290:881-883, 1985

22. Berry N, Jones DB, Smallwood J, et al: The prognostic value of the monoclonal antibodies HMFGl and HMFG2 in breast cancer. Br J Cancer 51:179-186, 1985

23. Horne CH, Reid IN, Milne GD: Prognostic significance of inappropriate production of pregnancy proteins by breast cancers. Lancet 2:279-282, 1976

24. Wright C, Angus B, Napier J. et al: Prognostic factors in breast cancer. Immunohistochemical staining for SPl and NCR1 1 related to survival, tumor epidermal growth factor receptor and oestrogen receptor status. J Pathol 155:325-331, 1987

25. Schnitt SJ, Connolly JL. Harris JR, et al: Radiation in- duced changes in the breast. HUM PATHOL 15:545-550, 1984

198