Br. J. Surg. Vol. 66 (1979) 848-852

Circulating breast tumour antigen-sensitive T lymphocytes in early breast cancer and high risk benign breast disease 9

W I L L I A M G. RAMEY, G E O R G E A. H A S H I M A N D H U G H F. F I T Z P A T R I C K * SUMMARY Levels of circulating T lymphocytes sensitized to breast tumour associated antigens (BTA) were correlated with pathological tumour stage or benign histopathology in preoperative studies of 180 patients by the antigen- stimulated active rosette-forming T cell (AgARFC) assay. Incubation of lymphocytes with allogeneic BTA extracts produced increased AgARFC compared with incubation without BTA. Sign$cant levels of BTA-sensitive T cells were found in 78per cent of breast cancer patients compared with 23 per cent of patients with benign disease (P<0.0005, by x2). Over 93 per cent of stage I cancer patients responded to BTA, compared with 69 per cent of stage II patients (P< 0.025) and 59 per cent of stages III-IVpatients (P< 040.5). Twenty-nine per cent of 42 patients with fibro- cystic disease were positive to BTA in contrast to 8 per cent of 2.5 patients with fibroadenomas. This was a 3.6-fold higher incidence of BTA-sensitive T cells associated with fibrocystic disease than with fibro- adenomas, which was in agreement with the increased breast cancer risk rate associated with jibrocystic disease. These findings suggest that the AgARFC assay may detect early malignant change in fibrocystic disease. The AgARFC assay was found to reliably detect early invasive carcinoma.

CELL mediated immunity to breast tumour associated antigens has been investigated as a modality for improved early detection of breast cancer (Farrow, 1970; Herberman, 1976; Order, 1977; Waters, 1978). In vitro assays of peripheral blood leucocytes from breast cancer patients have demonstrated proliferation, cell mediated cytotoxicity, migration inhibition and adherence inhibition responses to tumour cells and to soluble tumour associated antigens (Harris and Sinkovics, 1976; Zimmerman et al., 1977). Many of these studies, however, reported either low sensitivity in the detection of malignancies or low specificity in the differentiation of malignant from benign lesions. Re- cently, we demonstrated T cell mediated immunity to soluble extracts of breast adenocarcinoma by the anti- gen-stimulated active rosette-forming Tcell (AgARFC) assay (Ramey et al., 1979a, b). Peripheral blood T cells from breast cancer patients formed significantly higher percentages of active rosettes with heterologous erythrocytes when incubated with soluble breast tumour associated antigens than when incubated in antigen-free media. We now report the incidence and relationship of peripheral blood T cell responses to breast tumour associated antigens detected by the AgARFC assay, with tumour stage and benign histo- pathology in patients with clinical breast disease.

Patients and methods Patients, histopathology and staging One hundred and eighty inpatients at St Luke's Hospital Center were assayed for circulating T lymphocyte reactivity to soluble

breast adenocarcinoma-associated antigens by the AgARFC assay. Heparinized venous blood samples drawn before biopsy or operative manipulation of breast lesions were obtained from 164 women with clinical breast disease. Following surgery the patients were classified according to clinical findings in the medical records and pathological diagnoses were obtained from the routine surgical reports of the Department of Pathology. Patients with mammary carcinomas were classified according to the post-surgical treatment-pathological staging modification of the TNM staging system (American Joint Committee for Cancer Staging and End-Results Reporting, 1977). Patients with benign mastopathy were grouped accord- ing to the dominant lesion of the biopsy specimen: (a) fibro- cystic disease, (b) fibroadenoma, (c) periductular or peri- vascular inflammation and ( d ) other benign lesions (including fibrosis, ductal ectasia, ductal or lobular hyperplasia and intra- ductal papilloma). Sixteen patients with non-mammary carcinomas of a variety of organs were also assayed pre- operatively for lymphocyte activity to breast tumour associ- ated antigens. Patients who had received radiation therapy or immunosuppressive drugs, or who had a past history of treated breast carcinoma were not included in this study.

Preparation of breast tumour associated and histocompatibility antigen extracts Soluble breast tumour associated antigens (BTA) were prepared from fresh tumour tissue (histologically well- differentiated adenocarcinoma) obtained from two radical mastectomy specimens by extraction techniques which have been described previously (Ramey et al., 1979b). In brief, soluble antigen preparations of each tumour (designated BTA, and BTA2) were made by hypertonic potassium chloride extraction of homogenates of minced tumour tissue (2 g wet tumour per 2 ml of 3 M KCl solution) at 4 "C for 16 h. Extracts were ultracentrifuged at 16 000 g for 60 min and the super- natants were dialysed three times against 200 volumes of 0.9 per cent NaCl. Dialysates were then centrifuged at 10 000 g and the supernatants of the tumour extracts were stored in aliquots at -20 "C. Aliquots were thawed once at room temperature and diluted 1 : 10 in fresh Hank's balanced salt solution (HBSS, Gibco, Long Island, New York) adjusted to pH 7.4, prior to use in the AgARFC assay.

Parallel extracts of peripheral blood lymphocyte histo- compatibility antigens (HA) were prepared from heparinized venous blood drawn before mastectomy from the tumour donors (designated HA, and HA,), corresponding to BTA, and BTA,). Lymphocytes were isolated by Ficoll-Hypaque density flotation, washed in HBSS and extracted (los lympho- cytes per 20 ml of 3 M KC1) for 16 h at 4 "C (Reisfeld and Kahan, 1970). Ultracentrifugation, dialysis and storage of aliquots were performed as described above. Histocom- patibility antigen extracts were diluted to a concentration of 5 x lo8 cell-equivalents per ml of HBSS at pH 7.4 before use in the AgARFC assay.

AgARFC, ARFC and TRFC assays The AgARFC assay for T cell sensitization to soluble antigens has been described in detail elsewhere (Hashim et al., 1977; Ramey et al., 1977; Hashim et al., 1978) and application of this assay to soluble buman HA (Burrows et al., 1978) and to

s

c

*

*Department of Surgery, S t Luke's Hospital Center and Columbia University, Amsterdam Avenue at 114th Street, New York 10025.

Breast tumour antigen-sensitive T lymphocytes

Table I: CIRCULATING BTA-SENSITIVE T LYMPHOCYTES DETECTED BY THE AgARFC ASSAY IN MALIGNANT AND BENIGN BREAST DISEASE

849

Incidence of BTA-sensitive T cells*

Breast pathology No. positive % Breast carcinoma 50 78.1 Benign mastouathv 23 23.0

BTA (Ramey et al., 1979a, b) has been previously reported. In brief, peripheral lymphocytes, isolated from heparinized venous blood samples by Ficoll-Hypaque density flotation, were washed in HBSS, incubated with latex microparticles to label monocytes and suspended in 10 per cent minimal essential medium. Aliquots containing 300 000 lymphocytes per 0.05 ml were incubated in the presence of equal volumes of BTA for 15 min at 37 "C. Incubation was continued for 60 min follow- ing addition of 0.05 ml of sheep erythrocyte-adsorbed, gamma globulin-free fetal calf serum. Washed sheep red blood cells (SRBC) (0.05 ml of 0.25 per cent suspension) were added and the mixture was centrifuged at 200 g for 5 min. The pellet was gently resuspended and a minimum of 200 lymphocytes were counted on a haemocytometer at 400 times magnification to determine the percentage of rosette-forming lymphocytes (those having three or more adherent SRBC). Non-stimulated active rosette-forming T cells (ARFC) were determined by the same techniques, with 0.05 ml of HBSS substituted for soluble antigen preparation. Total T cell (TRFC) levels were deter- mined with 300 000 lymphocytes in 0.05 ml of HBSS added to 0.05 ml of 2 per cent SRBC and 0.05 ml of adsorbed fetal calf serum, centrifuged at 200g for 5min and incubated for 60 min at 25 "C. All assays were performed in duplicate.

Nomenclature, calculations and statistical treatment The subpopulation of peripheral blood lymphocytes that responded to in uitro incubation with specific soluble antigen was calculated as the difference between the percentage of antigen-stimulated active rosette-forming T cells (AgARFC, incubated in the presence of antigens) and the non-stimulated active rosette-forming T cells (ARFC, incubated in the absence of antigens). This difference was expressed as the percentage (AgARFC-ARFC). Patients in whom the percent- age (AgARFC-ARFC) with either BTA, or BTA, exceeded 5.5 per cent (the 99.5 per cent upper confidence limit for replicate variation of percentage ARFC determinations for the entire group of patients) and in whom the percentage (AgARFC-ARFC) for corresponding HA was less than 5.5 per cent were designated as testing-positive for circulating BTA-sensitive T cells. All other patients, including those in whom the percentage (AgARFGARFC) was greater than 5.5 per cent with BTA and with HA, were considered to be negative for sensitization to BTA. All group results were expressed as the mean&the standard error (s.e.) and group means were analysed by the two-tailed Student's t test. Signifi- cant differences between the incidences of positive patient responses in each patient group were determined by x2 analysis.

Results Sixty-four of the 164 patients with breast lesions studied preoperatively were found to have breast carcinoma at operation: 31 with stage I, 16 with stage I1 and 17 with stages I11 or IV disease. The remaining 100 patients had benign histopathology on biopsy: (a) 42 had fibrocystic disease, (6) 25 had fibro- adenomas, (c) 10 had periductal or perivascular inflammation and ( d ) 23 had other benign lesions (fibrosis, ductal ectasia, ductular or lobular hyper- plasia, or intraductal papilloma). Cellular sensitization to BTA in malignant and benign breast disease Fifty of 64 breast cancer patients (78.1 per cent), all stages, were found by preoperative assay to have greater than 5.5 per cent (AgARFC-ARFC) differences upon incubation with BTA, or BTA,. This incidence of circulating T cells sensitized to BTA in breast cancer patients was significantly higher than the 23.0 per cent of 100 benign mastopathy patients shown to have cellular sensitization to BTA (P< 0.0005, Table I ) . These values were also higher than the 12.5 per cent of non-mammary carcinoma patients responding to BTA (P<OWO5, Table IZ) in the AgARFC assay.

P<0.0005. * Percentage (AgARFC-ARFC) greater than 5.5 when incubated with BTA.

Table 11: BTA-SENSITIVE T CELLS IN NON-MAMMARY CARCINOMA

Incidence of BTA-sensitive T cells*

Diagnosis No. positive % Non-mammary carcinoma 2 12.5 Breast carcinoma 50 78.1

P < 0.0005. * Percentage (AgARFGARFC) greater than 5.5 when incubated with BTA.

Table ILI: CORRELATION OF BREAST CARCINOMA STAGE TO INCIDENCE OF CIRCULATING BTA-SENSITIVE T CELLS

Patients with BTA* Pathological stage No. positive %

I 29 93.5" I1 11 68.Xb 111-IV 10 58@

Significance (x2): a u. b, P<O.O25; a u. c, Pt0.005. * Percentage (AgARFGARFC) greater than 5.5 when incubated with BTA.

Stage-response relationship in breast carcinoma patients An inverse relationship between the pathological stage of breast carcinoma patients and the incidence of BTA-sensitive T lymphocytes was observed (Table IZZ). Twenty-nine of 31 patients (93.5 per cent) with early invasive carcinoma, without axillary nodal metastases (stage I), were found to have BTA- sensitive T lymphocytes. This incidence was signifi- cantly higher than the 68.8 per cent of patients with tumour in axillary nodes (stage 11, P<0.025) and the 58.8 per cent of patients with locally advanced or widely metastatic Carcinoma (stages 111-IV, P< 0.005) positive for BTA-sensitive T cells.

Sensitization to BTA inpatients with benign mastopathy Twenty-three of 100 patients found at operation to have benign breast pathology had greater than 5.5 per cent BTA-sensitive peripheral blood T lymphocytes. Classified according to benign histopathology, 12 of 42 patients with fibrocystic disease (28.6 per cent), 2 of 25 patients with fibroadenomas (8.0 per cent), 5 of 10 patients with predominant periductular or peri- vascular lymphocytic infiltrates (50.0 per cent), and 5 of 23 patients with other benign conditions (21.7 per cent) were demonstrated to have peripheral blood T cells reactive with either BTA, or BTA,. There were significantly more patients in the fibrocystic disease group reactive to BTA than in the fibro- adenoma group (P< 0.05, Table ZV). Similarly, more of the group of patients with predominant perivascular or periductular infiltrates were shown to have circula- ting T cell sensitization to BTA extracts than the

850

Table IV: BTA-SENSITIVE T CELLS IN BENIGN BREAST CONDITIONS OF DIFFERING MALIGNANT POTENTIAL

W. G. Ramey et al.

Patients with BTA* Benign mastopathy No. positive % Fibrocystic disease 12 286 Fibroadenoma 2 8.0

Significance (xz) : P< 0.05. * Percentage (AgARFC-ARFC) greater than 5.5 when incubated with BTA.

Table V: BTA-SENSITIVE T CELLS IN OPERABLE BREAST CARCINOMA AND BENIGN BREAST DISEASE

Patients with BTA* Significance Breast pathology No. positive % (x2 test)

Stage I or I1 breast 40 85.1

Fibrocystic disease 12 286 P<O.0005 Fibroadenoma 2 8.0 P<O.0005

carcinoma

Periductal or perivascular 5 50.0 P < 0.025

Other benign conditions? 5 21.7 P<0.0005 * Percentage (AgARFC-ARFC) greater than 5.5 when incubated with BTA. t Fibrosis, ductal ectasia, hyperplasia, papillomatosis.

inflammation

Table M: PERIPHERAL LYMPHOCYTE COUNTS AND NON-SPECIFIC T CELL SUBPOPULATIONS IN STAGES OF BREAST CARCINOMA

Stage III- Stage I Stage I1 IV

Population (n= 31) (n= 16) (n= 17)

% TRFC 70*8f1*9* 76.7k2.0 73.0k2.1 % ARFC 41.6-122 44.0&3-9 46.2k2-2 TRFC/mm8 19112~145 1448-1206 1833-1352 ARFC/mms 904-1104 847*175 1147+218

* P<O.O5 compared with %TRFC for stage I1 patients.

Lymphocytes/mms 21 14k 186 1873 k 248 2452 f 5 10

patients with fibroadenomas (P< 0.01). However, all other differences among the groups of patients with benign mastopathy were not significant.

Differentiation between operable breast carcinoma and benign mastopathy The rate of positive assay results was highest in patients with the best prognosis by stage (93.5 per cent in stage I). BTA-sensitive T cell detection by the AgARFC assay was positive in 85.1 per cent of patients with resectable breast carcinoma (stages I and 11, combined). This overall incidence of positive BTA- sensitive T cell assays in patients with stages I and I1 breast carcinoma was significantly higher than in any of the benign mastopathy groups (P<0.025 to P< 0.0005, Table V).

Peripheral lymphocyte counts and non-specific T cell subpopulations in patients with breast carcinoma The mean values of peripheral blood lymphocyte counts, percentage or absolute number of circulating total T cells (TRFC) and percentage or absolute number of active T cells (ARFC), were not highly significantly different among stage I, stage I1 and stages 111-IV breast carcinoma patients (Table VI).

'

Discussion Positive responses to soluble breast tumour associated antigens by circulating T lymphocytes were detected in 93.5 per cent of patients with early invasive breast carcinoma by the antigen-stimulated active rosette- forming T cell (AgARFC) assay. Significantly lower percentages of patients with axillary nodal metastases, locally advanced disease and widespread breast cancer were positive. The relationship between patho- logical tumour stage and incidence of circulating BTA-sensitive T cells confirms previous findings reported from this laboratory (Ramey et al., 1979a, b). Furthermore, the relationship of decreasing cell mediated immune responses to BTA with advancing tumour stage shown in this study is in agreement with other reports which demonstrated depressed cell mediated immunity in advanced breast cancer. Disease stage and patient survival have been correlated with depressed delayed hypersensitivity reactions to autologous tumour transplants (Southam et al., 1962; Brunschwig et al., 1965) and to cryostat sections of autologous tumour tissue applied to skin windows (Black and Leis, 1971); however, the relationship of disease stage to skin test responses to soluble breast carcinoma tissue extracts has not been clearly defined (Hughes and Lytton, 1964; Stewart and Orizaga, 1971). Leucocyte migration inhibition (LMI) induced by BTA antigens has been shown to correlate with clinical stage and prognosis in several studies (Black et al., 1975; Roberts and Bathgate, 1975; Rieche et al., 1976). The highest degree of LMI reactivity to auto- logous and allogeneic breast tumour antigen was demonstrated in patients with in situ and early invasive breast carcinoma. Leucocyte adherence inhibition (LAI) induced by breast carcinoma antigen extracts was reported to be most frequent with leucocytes from patients with stage I or I1 breast carcinoma, while progressively lower rates of LA1 reactivity were assayed with patients with stage I11 (45 per cent), or stage IV (29 per cent) disease (Grosser and Thompson, 1975; Fujisawa et al., 1977; Sanner et al., 1979). These results are in agreement with the incidence of BTA-sensitive T cells detected by the AgARFC assay in different pathological stages of breast cancer.

The mechanism underlying the unresponsiveness of T lymphocytes from patients with advanced breast cancer to BTA in the AgARFC assay has not been established. General parameters of cellular immune competence measured in this study were not signifi- cantly depressed in patients with stages 11, I11 or IV breast cancer. In other studies, lymphoproliferative (LP) responses to alloantigens were depressed in patients with breast cancer, even when LP responses to mitogens were normal (Golub et al., 1974; Dean et al., 1977; Jerrells et al., 1978a, b). Mitogen- responsive T lymphocytes were correlated with the ARFC subpopulation whereas alloantigen-responsive T cells were correlated with the non-ARFC T cell sub- population (Jerrells et al., 1978a). Similarly, BTA- sensitive T cells were detected in the non-ARFC subpopulation by the AgARFC assay, as were allo- antigen-sensitized T cells in allotransplantation studies (Ramey et al., 1977; Burrows et al., 1978). Depressed LP reactivity of circulating alloantigen-sensitive T cells and depressed AgARFC reactivity of circulating BTA-sensitive T cells in patients with breast carcin- oma may be the result of either depletion of such cells

from the circulating lymphocyte pool, or excess of the sensitizing antigen which has been shown to block lymphocyte responses in the AgARFC assay (Hashim et al., 1979). Breast cancer serum has been shown to exert blocking activity on circulating leucocytes in a number of in vitro assay systems (Vanky et al., 1971 ; Whitehead et al., 1976; Jeejeebhoy, 1977; Steward, 1977), attributed to either antigen-antibody complexes (Sjogren et al., 1971; Matthews and Whitehead, 1976; Hoffken et al., 1977; Nordquist et al., 1977), or to excess circulating free tumour antigen (Grosser and Thomson, 1976; Fujisawa et al., 1977; Steward, 1977). Depression of BTA-sensitive T cells observed with AgARFC assays of peripheral blood lympho- cytes of patients with advanced breast cancer could, therefore, represent an in vivo effect of either excess tumour antigen, or of tumour antigen-antibody complexes.

Recognition of breast tumour associated antigens by circulating T cells was observed in 23.0 per cent of patients with benign mastopathy. These responses may reflect the development of immunity to early malignant change in a continuum of breast disease progressing from benign mastopathy through ‘pre- cancerous’ hyperplastic lesions to in situ and eventually invasive carcinoma (Kern and Brooks, 1969; Black, 1976; Black et al., 1976; Cardiff et al., 1977). Recogni- tion of BTA by patients with benign mastopathy has been reported in serologic assays (Humphrey et al., 1974), and LA1 assays (Sanner et al., 1979) to a similar degree to that observed in the AgARFC assay. Higher rates of response in benign mastopathy have been reported with LMI (Cannon et al., 1978) and cell mediated cytotoxicity (Fossati et al., 1972; Avis et al., 1974; Oldham et al., 1975; Avis et al., 1976; Canevari et al., 1976) assays.

The incidence of BTA-sensitive T cells detected was significantly higher in patients with fibrocystic disease than in patients with fibroadenomas. This 3.6-fold increase in the incidence of positive assays for BTA-sensitive T cells in fibrocystic disease over the incidence with fibroadenomas parallels the increased risk of eventual breast cancer development associated with fibrocystic disease (Warren, 1940; Haagensen, 1956; Davis et al., 1964; Veronesi and Pizzocara, 1968). The question of whether these patients with fibrocystic disease will eventually develop invasive breast carcinoma was not addressed in this study; however, the data suggest that the AgARFC assay may be detecting early malignant change in fibro- cystic disease. Such reactivity may have been responsible for the BTA-sensitive T cells noted in 5 of 10 patients with benign lesions consisting of peri- vascular or periductular lymphocyte infiltrates. Alternatively, these cases may represent reactivity to an agent similar to the murine mammary tumour viruses, which are known to cross-react with human benign and malignant proliferative breast lesions (Black et al., 1975; Muller et al., 1976).

The diagnostic accuracy of the AgARFC assay in the detection of resectable breast cancer by the criteria employed in this study was 95.3 per cent for stage I lesions and was 85.1 per cent for stages I and I1 combined. These percentages compare favourably with those reported for other in vitro assays, as well as with the detection rates which have been reported with mammography (Feig et al., 1977). Circulating T lymphocyte sensitization to breast

Breast tumour antigen-sensitive T lymphocytes 851

tumour associated antigen was observed with signifi- cantly lower frequency in patients with advanced mammary carcinoma, non-mammary carcinoma or benign mastopathy. The incidence of BTA-sensitive T cells detected in patients with fibrocystic disease was higher than that in patients with fibroadenomas, by a factor which is in agreement with the increased breast cancer risk rate associated with fibrocystic disease. The data from this study suggest that the AgARFC assay reliably detected early invasive breast carcinoma, at the stage when it carries the most favourable prognosis. Longitudinal study of the patients assayed by the AgARFC method will assess the relevance of the detection of T cell sensitization to breast tumour associated antigens in determining prognosis of patients with breast carcinoma, and evaluate the possibility that detection of BTA- sensitive T cells in patients with fibrocystic disease reflects early malignant transformation.

Acknowledgements We are grateful to the Clark Foundation for the support which made this study possible. We would also like to thank our colleagues who referred patients to this study: Drs Warren B. Burrows, Antoine Munther, Alexander J. Swistel and Duk Lee who contributed extensively to various phases of this research effort. Our thanks go also to Mrs Hermosa Paderon and Mr Cheng Yue Lin for their excellent technical assistance and to Mrs Peggy Mulrooney for her help in the preparation of this manuscript.

References AMERICAN JOINT COMMITTEE FOR CANCER STAGING AND END-

RESULTS REPORTING (1977) Manual for Cancer Staging. Chicago, Ill., pp. 101-107.

AVIS F., AVIS I., NEWSOME J. F. et al. (1976) Antigenic cross- reactivity between adenocarcinoma of the breast and fibrocystic disease of the breast. J. Nut1 Cancer Inst. 56,

AVIS F., MOSONOV I. and HAUGHTON G. (1974) Antigenic cross- reactivity between benign and malignant neoplasms of the human breast. J. Natl Cancer Inst. 52, 1041-1049.

BLACK M. M. (1976) Structural antigenic and biological charac- teristics of precancerous mastopathy. Cancer Res. 36, 2596-2604.

BLACK M. M. and LEIS H. P. JUN. (1971) Cellular responses to autologous breast cancer tissue. Cancer 28, 263-273.

BLACK M. M., ZACHRAN R. E., SHORE B. et al. (1975) Prog- nostically favorable immunogens of human breast cancer tissue: antigenic similarity to murine mammary tumor virus. Cancer 35, 121-128.

BLACK M. M., ZACHRAN R. E., SHORE B. et al. (1976) Biological considerations of tumor-specific and virus-associated antigens of human breast cancers. Cancer Res. 36,769-114.

BRUNSCHWIG A., SOUTHAM c. M. and LEVIN A. G . (1965) Host resistance to cancer: clinical experiments by homotrans- plants, autotransplants and admixture of autologous leukocytes. Ann. Surg. 162,416-425.

BURROWS w. B., PIERCE J. c., HASHIM G. A. et al. (1978) Cell mediated immunity to donor antigens in renal allograft recipients. Surgery 83, 741-745.

CANEVARI s., FOSSATI G. and DELLA PORTA G. (1976) Cellular immune reaction to human malignant melanoma and breast carcinoma cells. J. Nut1 Cancer Inst. 56,705-709.

CANNON G. B., MCCOY J. L., JEROME L. J. et al. (1978) Immuno- logic relationship between breast carcinoma and benign breast disease as detected by the leukocyte migration inhibition assay. J. Nut1 Cancer Inst. 61, 1181-1186.

CARDIFF R. D., WELLINGS s. R. and FAULKIN L. J. (1977) Biology of breast preneoplasia. Cancer 39,2734-2736.

DAVIS H. H., SIMONS M. and DAVIS J. B. (1964) Cystic disease of the breast: relationship to carcinoma. Cancer 17, 957- 978.

17-25.

852 W. G. Ramey et al.

DEAN J. H., CONNOR R., HERBERMAN R. B. et al. (1977) The relative proliferation index as a more sensitive parameter for evaluating lymphoproliferative responses of cancer patients to mitogens and alloantigens. Int. J. Cancer 20,

FARROW J. H. (1970) Current concepts in the detection and treatment of the earliest of breast cancers. Cancer 25, 468477.

FEIG s. A., SHABER G. s., SCHWARTZ G. F. et al. (1977) Thermo- graphy, mammography and clinical examination in breast cancer screening. Review of 16 000 studies. Radiology 122, 123-127.

FOSSATI G., CANEVARI s., DELLA PORTA G. et al. (1972) Cellular immunity to human breast carcinoma. Znt. J . Cancer 10, 391-396.

FUJISAWA T., WALDMAN s. R. and YONEMOTO R. H. (1977) Leuko- cyte adherence inhibition by soluble tumor antigens in breast cancer patients. Cancer 39, 506-513.

GOLUB s. H., O’CONNELL T. x. and MORTON D. L. (1974) Correla- tion of in vivo and in vitro assays of immunocompetence in cancer patients. Cancer Res. 34, 1833-1837.

GROSSER N. and THOMSON D. M. (1975) Cell mediated immunity in breast cancer patients evaluated by antigen-induced leukocyte adherence inhibition in test tubes. Cancer Res.

GROSSER N. and THOMSON D. M. (1976) Tube leukocyte (mono- cyte) adherence inhibition assay for the detection of antitumor immunity. 11. Monocyte reacts with tumor antigen via cytophilic antitumor antibody. Znt. J . Cancer

HAAGENSEN c. D. (1956) Diseuses of the Breast. Philadelphia, Saunders, pp. 156-66.

HARRIS 3. E. and SINKOVICS J. G. (1976) The Immunology of Malignant Disease. St Louis, Mosby, pp. 48-51, 55,480-4.

HASHIM G. A, , LEE D. and PIERCE J. c. (1977) Antigen stimulated rosette formation by sensitized T cells in experimental allergic encephalomyelitis. Neurochem. Res. 2, 99-109.

HASHIM G. A., LEE D., PIERCE J. c. et al. (1978) Myelin basic protein stimulated rosette-forming T cells in multiple sclerosis. Neurochem. Res. 3, 3 7 4 8 .

HASHIM G. A., YEE N. and RAMEY w. G. (1979) The kinetics of T- lymphocyte subpopulations in guinea-pigs sensitized with allogeneic transplantation antigens. (Submitted for publication.)

HERBERMAN R. H. (1976) Immunologic approaches to the diagnosis of cancer. Cancer 37, 549-562.

HOFFKEN K., MEREDITH I. D., ROBINS R. w. et al. (1977) Circulating immune complexes in patients with breast cancer. Br. Med. J . 2, 218-220.

HUGHES L. E. and LYTTON B. (1964) Antigenic properties of human tumours: delayed cutaneous hypersensitivity reactions. Br. Med. J . 1, 209-212.

HUMPHREY L. J., ESTES N. c., MUSE P. A. JUN. et al. (1974) Serum antibody in patients with mammary disease. Cancer 34,

JEEJEEBHOY H. F. (1977) Comparison of autochthonous and allogeneic breast tumour cells in tests for lymphocyte immunity to human tumours. Br. J. Cancer 35,161-169.

JERRELLS T. R., DEAN I. H. and HERBERMAN R. B. (1978a) Relationship between T lymphocyte levels and lympho- proliferative responses to mitogens and alloantigens in lung and breast cancer patients. Znt. J. Cancer 21,282-290.

of suppressor cells in depression of in uitro lympho- proliferative responses of lung cancer and breast cancer. J. Natl Cancer Inst. 61, 1001-1009.

KERN w. H. and BROOKS R. N. (1969) Atypical epithelial hyper- plasia associated with breast cancer and fibrocystic disease. Cancer 24, 668-675.

MATTHEWS N. and WHITEHEAD R. H. (1976) Inhibition of K cell function by human breast cancer sera. Br. J . Cancer 34,

3 59-3 70.

35, 2571-2579.

18, 48-57.

1516-1520.

JERRELLS T. R., DEAN J. H., RICHARDSON 0. L. et al. (1978b) Role

635-640.

MULLER M., ZOTTER s. and KEMMER c. (1976) Specificity of human antibodies to intracytoplasmic type A particles of the murine mammary tumor virus. J. Natl Cancer Inst. 56,

NORDQUIST R. E., ANGLIN J. H. and LERNER M. P. (1977) Anti- body induced antigen redistribution and shedding from human breast cancer cells. Science 197, 366-367.

OLDHAM R. K., DIEU J. Y., CANNON G. B. et al. (1975) Cellular microcytotoxicity in human tumor systems: analysis of results. J. Natl Cancer Inst. 55, 1305-1318.

ORDER s. E. (1977) Antigens, immunoglobulins and immunity during the course of breast cancer. Prog. Clin. Bid. Res.

RAMEY w. G., HASHIM G. A., BURROWS w. B. et al. (1979a) Cell- mediated immunity to breast cancer: tumor antigen sensitive circulating T-lymphocytes. Fed. Proc. 38,993.

RAMEY W. G., HASHIM G. A., BURROWS w. B. et al. (1979b) Detection of breast tumour-antigen-sensitive circulatory T lymphocytes by antigen-stimulated active rosette formation.Cancer Res. 39 (In the press.)

RAMEY w. G., HASHIM G. A., PIERCE J. G. et al. (1977) A new assay for T cell mediated immunity to transplantation antigens. Surgery 81, 640-645.

REISFELD R. A. and KAHAN B. D. (1970) Transplantation antigens. Adv. Zmmunol. 12, 117-200.

RIECHE K., ARNDT A. and PASTERNAK G. (1976) Cellular immunity in mammary cancer patients as measured by the leukocyte migration test. A follow up study. Int. J. Cancer

ROBERTS M. M. and BATHGATE E. M. (1975) The immune reaction to human breast cancer: a comparison of skin and leucocyte migration tests. Br. J . Surg. 62, 870-872.

SANNER T., BRENNHOVD I., CHRISTENSEN I. et al. (1979) Cellular antitumor immune response in women with risk factors for breast cancer. Cancer Res. 39, 654-657.

SJOGREN H. o., HELLSTROM I., BAUSAL s. c. et al. (1971) Suggestive evidence that the ‘blocking’ antibodies of tumor-bearing individuals may be antigen-antibody complexes. Proc. Nut1 Acad. Sci. 68, 1372-1375.

SOUTHAM c. M., BRUNSCHWIG A. and DIZON Q. (1962) Auto- logous and homologous transplantation of human cancer. In: Henry Ford Hospital International Symposium: Biological hteractions in Normal and Neoplasiic Growth. Boston, Little, Brown.

STEWARD A. M. (1977) The modulation of tumour versus host response in lung, colon and breast cancer patients: implications for adjuvant immunochemotherapy. Aust. NZ J. Surg. 47, 642-647.

STEWART T. H. M. and ORIZAGA M. (1971) The presence of delayed hypersensitivity toward cellular extracts of malignant tumors. The frequency and cross-reactivity of this phenomenon in breast cancer and its correlation with survival. Cancer 28, 1472-1478,

mediated inhibition of lymphocyte stimulation by autoch- thonous human tumors. J. Natl Cancer Inst. 47,

VERONESI u. and PIZZOCARA G. (1968) Breast cancer in women subsequent to cystic disease of the breast. Surg. Gynecol. Obstet. 126, 529-532.

WARREN s. (1940) The relation of ‘chronic mastitis’ to carcinoma of the breast. Surg. Gynecol. Obstet. 71, 257-273.

WATERS H. (1978) The Handbook of Cancer Immunology, Vol. I : Basic Cancer-related Immunology. New York, Garland STPM Press.

WHITEHEAD R. H., THATCHER J. and TEASDALE c. (1976) T and B lymphocytes in breast cancer: stage relationship and abrogation of T lymphocytes depression by enzyme treatment in uitro. Lancet 1, 330-333.

immunity to breast cancers. Prog. Clin. Biol. Res. 12, 75-89.

295-303.

12, 119-130.

17.212-287.

VANKY F., STJENSWARD J., KLEIN G. et 231. (1971) Serum-

95-103.

ZIMMERMAN A., RUCHTI C., ROOS B. et al. (1977) The problem Of