[Tamoxifen adjuvant delays early breast cancer. Results of a cooperative randomized trial]

Adjuvant tamoxifen (TAM) has been proved to reduce recurrence and mortality in early breast cancer, nevertheless many patients did not receive TAM as adjuvant therapy after local treatment. In order to study the efficacy of delayed TAM therapy in patients who were not given immediate adjuvant hormonal treatment, a multicenter randomized trial has been conducted by the French National Cancer Centers (FNCLCC). According to eligibility criterias all women with breast cancer who received curative local treatment at least 2 years before (surgery +/- radiotherapy) with or without adjuvant chemotherapy but no hormonal treatment could have been included. Between September 1986 and October 1989, 494 women were randomized to receive either TAM 30 mg/day for 5 years or no treatment. Patients' characteristics such as age, tumoral stage, number of positive nodes, receptors status and time from local treatment were equally distributed in the 2 groups. An improvement in the disease free survival in the TAM treated patients can be observed with a significative difference (p = 0.05), nevertheless the overall survival is not improved in the TAM group. In the same way, in nodes positive patients although no significative improvement in the overall survival can be observed, a significative improvement in the disease free survival (p = 0.05) can be noted. In estradiol receptors positive patients tamoxifen gives a significative reduction in the odds of death (p = 0.04) and recurrence (p = 0.03). The disease free improvement seems to be limited to 50 and more years old patients. The first results of this trial lead to prescribe tamoxifen to all postmenopausal women previously treated for an early breast cancer without adjuvant tamoxifen treatment.

The modulatory role of estrogens on corticotropin-releasing factor gene expression in the hypothalamic paraventricular nucleus of ovariectomized rats: role of the adrenal gland

In vitro, estrogens stimulate corticotropin-releasing factor (CRF) gene transcription. In ovariectomized (OVX) rats, estrogens have a negative effect on CRF mRNA levels in the hypothalamic paraventricular nucleus. This suggests that the stimulatory influence of estrogens found in vitro may be masked in vivo by an inhibitory effect mediated through neural and/or humoral factors. Glucocorticoids may be involved in this phenomenon since estrogens increase circulating corticosterone levels in OVX rats. We studied the effect of 7-day 17 beta-estradiol (E2) treatment on CRF gene expression in the paraventricular nucleus of 2-week OVX mature rats after sham operation or adrenalectomy (ADX) with or without corticosterone replacement. In sham ADX animals, E2 administration increased plasma corticosterone concentrations, did not change the binding capacity or the affinity of circulating corticosteroid binding globulin, and decreased hypothalamic CRF gene expression. Following ADX, CRF mRNA levels increased and were normalized by corticosterone treatment. Estrogens induced a significant build up in CRF mRNA concentrations in both ADX or ADX corticosterone-replaced animals. Our data demonstrate that the positive effect of estrogens on CRF gene expression found in vitro or in vivo after ADX is antagonized, in vivo when the adrenal glands are intact, by a negative influence. They strongly suggest that the increased circulating corticosterone levels induced by E2 administration mediate the inhibitory effect of estrogens on CRF mRNA levels. These observations may explain the gender related differences in the basal and stress-induced hypothalamo-pituitary-adrenal activity.

8-OH-DPAT regulates the amplitude and the phase of LH surge in ovariectomized steroid-primed rats

Precise interactions between ovarian steroids and neurotransmitters are required for the secretion of phasic LH surge. Previous data suggested the existence of an interactive stimulatory effect of progesterone (P) and serotonin (5-HT) on LH release. In the present work the effects of 8-OH-DPAT, a selective 5-HT(1A) agonist, on phasic LH secretion were tested in ovariectomized rats implanted for 6 days with a pellet of 17 β estradiol (OVX-E(2)) and in OVX-E(2) treated with progesterone (OVX-E(2)-P). Intraperitoneal injection of 8-OH-DPAT at 11.00 h in the morning of the expected LH surge had no effect on circadian plasma levels of LH in OVX-E(2) rats, whereas it induced a phase advance and an increase in LH surge in OVX-E(2)-P rats. Administration of the antiprogestin RU 38486 in OVX-E(2)-P rat, totally abolished the combined effects of P and 8-OH-DPAT on phasic LH release. SDZ 216-525, a specific 5-HT(1A) antagonist administered 60 min before 8-OH-DPAT, inhibited the stimulatory effect of the 5-HT(1A) agonist on the amplitude of LH surge. The present data suggest that progesterone is required for the regulation of phasic LH release by 5-HT(1A) agonists and that under this hormonal condition the activation of 5-HT(1A) receptors induces a phase advance and an increase in LH surge.

Estrogens decrease expression of the corticotropin-releasing factor gene in the hypothalamic paraventricular nucleus and of the proopiomelanocortin gene in the anterior pituitary of ovariectomized rats

It is known that estrogens modulate the hypothalamopituitary-adrenal (HPA) axis both under resting conditions and during exposure to stress. Nevertheless, the site of action of estrogens is not still fully elucidated. We sought to determine if estrogens could act on the major hypothalamic ACTH secretagogue: corticotropin-releasing factor (CRF). Mature rats were ovariectomized (OVX) and 2 weeks later implanted with silastic capsules containing 17β-estradiol (E(2)). Animals were sacrificed 7 days later. CRF mRNA in the hypothalamic paraventricular nucleus (PVN) and proopiomelanocortin (POMC) mRNA in the anterior pituitary were measured byin situ hybridization. CRF content in the median eminence was measured by semiquantitative immunocytochemistry. E(2) treatment induced a significant decrease of CRF mRNA levels in the PVN (3.70±0.14vs 4.79±0.15 copies of probe×10(-3)/μm(3) of tissue in OVX rats,P<0.05), an accumulation of immunoreactive CRF in the zona externa of the median eminence (207±36vs 100±15% in OVX rats,P<0.05), and a decrease of POMC mRNA levels in the anterior pituitary (4.6±0.6vs 6.9±0.6 copies of probe ×10(-2)/μm(3) of tissue in OVX rats,P<0.05). These results demonstrate that estrogens have a negative effect on CRF gene expression and secretion and on POMC gene expression. Whether estrogens modulate directly the CRF-synthesizing cells or act through an increase of the glucocorticoid negative feedback remains to be determined.

Stimulatory effects of 5HT1A receptor agonists on luteinizing hormone-releasing hormone release from cultured fetal rat hypothalamic cells: interactions with progesterone

Previous works have suggested an interactive stimulatory effect of progesterone (P) and serotonin (5-HT) on luteinizing hormone release. The purpose of the present study was to determine whether 5-HT via 5-HT1A receptors interacts with P in the process of luteinizing hormone-releasing hormone (LHRH) release. Using fetal hypothalamic neurons in primary cell cultures the first goal of this study was to determine the effects of 5-HT1A receptor agonists on LHRH secretion. 8-Hydroxy-2 (di-n-propylamino) tetralin (8-OH-DPAT) or ipsapirone (10(-5) M) significantly stimulated LHRH release. Pharmacological studies have allowed to rule out the possible involvement of alpha 2- or beta-adrenoreceptors, or 5-HT uptake sites, in the stimulatory effect of 8-OH-DPAT on LHRH release, thus demonstrating the specific involvement of 5-HT1A receptors in the stimulation of LHRH release. The second goal was to test the ability of P to stimulate LHRH release from fetal hypothalamic neurons. P (10(-6) M) applied for 30 or 120 min significantly stimulated LHRH secretion. The maintenance of the stimulation of LHRH release by P after a cycloheximide treatment or by an impermeable analog of P, P-3-BSA, has suggested a nongenomic effect of P on LHRH release. The effects of a pretreatment of cells by P on 8-OH-DPAT-induced LHRH release were tested. While 10(-7) M P alone did not stimulate LHRH release, this concentration of steroid potentiated the LHRH response to 10(-5) M 8-OH-DPAT. These findings led to the conclusion that P acting at the level of the plasma membrane potentiates the stimulatory effect of 5-HT1A receptor agonists on LHRH release.

Impairment of serotoninergic transmission is followed by adaptive changes in 5HT1B binding sites in the rat suprachiasmatic nucleus

Serotonin1B (5-HT1B) receptor binding in the suprachiasmatic nucleus (SCN) following impairment of serotoninergic transmission was studied by quantitative autoradiography. Serotonin (5-HT) denervation with 5,7-dihydroxytryptamine (5,7-DHT) caused a significant increase in the density of 5-HT1B receptors in both the ventral (62%) and dorsal (53%) parts of the SCN as early as 3 days after axotomy. The magnitude of this increase did not differ 3, 15 or 21 days post-lesion. An up-regulation of 5-HT1B receptors with similar magnitude was obtained in the two parts of the SCN after inhibition of 5-HT synthesis by chronic parachlorophenylalanine treatment. In this case, up-regulation was shown to be reversible after restoration of 5-HT synthesis with L-5-hydroxytryptophan. These results indicate that 5-HT1B receptor density in the SCN was inversely correlated with 5-HT levels. These plastic properties exhibited by 5-HT1B receptors in the SCN are discussed in relation to the mode of 5-HT transmission and possible localization of the receptors onto the main chemically defined cell populations of the nucleus.

N-methyl-D-aspartic acid/glycine interactions on the control of 5-hydroxytryptamine release in raphe primary cultures

Glutamic acid and glycine were quantified in cells and medium of cultured rostral rhombencephalic neurons derived from fetal rats. In the presence of 1 mM Mg2+, NMDA (50 microM) significantly stimulated (by 69%) release of newly synthesized 5-[3H]hydroxytryptamine ([3H]5-HT). D-2-Amino-5-phosphonopentanoate (AP-5; 50 microM) blocked the stimulatory effect of NMDA. AP-5 by itself inhibited [3H]5-HT release (by 25%), suggesting a tonic control of 5-HT by glutamate. In the absence of Mg2+, basal [3H]5-HT release was 60% higher as compared with release with Mg2+. AP-5 blocked the increased [3H]5-HT release observed without Mg2+, suggesting that this effect was due to the stimulation of NMDA receptors by endogenous glutamate. Glycine (100 microM) inhibited [3H]5-HT release in the absence of Mg2+. Strychnine (50 microM) blocked the inhibitory effect of glycine, indicating an action through strychnine-sensitive inhibitory glycine receptors. The [3H]5-HT release stimulated by NMDA was unaffected by glycine. In contrast, when tested in the presence of strychnine, glycine increased NMDA-evoked [3H]5-HT release (by 22%), and this effect was prevented by a selective antagonist of the NMDA-associated glycine receptor, 7-chlorokynurenate (100 microM). 7-Chlorokynurenate by itself induced a drastic decrease in [3H]5-HT release, indicating that under basal conditions these sites were stimulated by endogenous glycine. These results indicate that NMDA stimulated [3H]5-HT release in both the presence or absence of Mg2+. Use of selective antagonists allowed differentiation of a strychnine-sensitive glycine response (inhibition of [3H]5-HT release) from a 7-chlorokynurenate-sensitive response (potentiation of NMDA-evoked [3H]5-HT release).

Increase of central 5-HT1B binding sites following 5,7-dihydroxytryptamine axotomy in the adult rat

The effects of selective axotomy of serotoninergic neurons produced by an intracerebroventricular injection of 5,7-dihydroxytryptamine (200 micrograms free base) on 5-HT1B binding sites labeled with S-CM-G-[125I]TNH2 were investigated by quantitative autoradiography in the rat brain. Results show, 21 days after surgery, an upregulation of 5-HT1B receptors in the entorhinal cortex and the dorsomedial and suprachiasmatic nuclei of the hypothalamus. The cellular localization of those 5-HT1B receptors exhibiting post-lesion plastic properties is discussed.

Postmenopausal patients with node-positive resectable breast cancer. Tamoxifen vs FEC 50 (6 cycles) vs FEC 50 (6 cycles) plus tamoxifen vs control--preliminary results of a 4-arm randomised trial. The French Adjuvant Study Group

In 1986 the true benefit of adjuvant medical treatment in postmenopausal patients with pathological node-positive breast adenocarcinoma was still controversial. The French Adjuvant Study Group (FASG) initiated a randomised trial to elucidate the respective roles of adjuvant chemo-and/or hormonotherapy in this group of patients. Of the 776 patients who have been included between 1986 and 1990, 741 were fully eligible for evaluation. Inclusion criteria were postmenopausal patients aged between 50 and 70 years with adenocarcinoma of the breast, positive pathological nodes and no distant metastasis. Patients were randomised to 1 of 4 treatment arms: Group A (n = 192) received tamoxifen 30 mg/day orally for 3 years; Group B (n = 183) received FEC 50 (fluorouracil 500 mg/m2, epirubicin 50 mg/m2 plus cyclophosphamide 500 mg/m2) for 6 cycles; Group C (n = 182) received tamoxifen 30 mg/day orally for 3 years plus FEC 50 for 6 cycles; Group D (n = 184) received no medical adjuvant treatment. Surgery was either modified radical mastectomy (n = 363) or tumorectomy (n = 378), and postoperative irradiation was given to all patients. All major prognostic factors were well balanced between the 4 patient groups. Toxicity was evaluated in 348 patients in Groups B and C who received a total of 1983 chemotherapy cycles. Median epirubicin dose intensity (mg/m2/week) was 15.8 in Group B and 15.7 in Group C. Grade 3 to 4 neutropenia was observed in 4.7% of cycles for Group B and 3.7% for Group C. Grade 3 to 4 nausea/vomiting were seen in 18% of treatment cycles in Group B and 15% in Group C.(ABSTRACT TRUNCATED AT 250 WORDS)

Evidence for the presence of noradrenergic neurons and their inhibitory action on luteinizing hormone-releasing hormone release in cultured fetal rat hypothalamic cells

The control of LHRH release by catecholamine systems during fetal life (embryonic stages) was studied using hypothalamic neurons in primary cell cultures and an attempt was made to characterize the receptor type involved. Phenylephrine and clonidine, respectively alpha 1 an alpha 2 adrenoreceptor agonists, both inhibited LHRH release. These agonist inhibitory induced-effects were antagonized by the respective alpha 1 and alpha 2 adrenoreceptor antagonists (prazosin and rauwolscine). Both prazosin and rauwolscine applied alone induced a marked increase in LHRH release. Similarly, inhibition of catecholamine synthesis obtained by alpha-methyl-para-tyrosine (alpha-MT) led to a significant increase in LHRH release. The stimulatory effects induced by alpha 1 and alpha 2 adrenoreceptor antagonists or by alpha-MT on LHRH release suggest the presence of noradrenergic and/or adrenergic cells in fetal hypothalamic cultures. Therefore, catecholamine contents were measured in fetal hypothalamic cells in culture. Measurable amounts of norepinephrine and dopamine were found in cells, although epinephrine was undetectable. These results show: 1--noradrenergic cells are present in primary culture of fetal hypothalamic cells. 2--This intrinsic hypothalamic noradrenergic system exerts an inhibitory control on LHRH release at an early stage of development through alpha 1 and alpha 2 adrenoreceptors.

Prognostic value of total cathepsin D in breast tumors. A possible role in selection of chemoresistant patients

Evaluation of prognostic factors for breast cancers is important for therapeutic decisions both at the time of surgery and during postoperative surveillance. In 1979, H. Rochefort described an induced protein with a molecular weight of 52,000 Daltons identified as procathepsin D. Total cathepsin D (TCD) (52K + 48K + 34K), expressed in pmol/mg protein, can be measured by an immunoradiometric method commercialized by Cis-Biointernational. Total cathepsin D was assayed in 413 breast cancer tumors from patients who underwent surgery between January 1, 1978, and December 31, 1985. Using a cut-off of 35 pmol/mg protein, patients with an elevated level had a significantly poorer survival than those with a low level (p = 0.03). This difference was not found for node-negative patients but was very significant for node-positive patients (p less than 0.008). The survival of node-positive patients with a low total cathepsin level was not statistically different from that of node-negative patients. Analysis of the N+ subgroup of patients who did not receive adjuvant chemotherapy revealed that TCD no longer had any prognostic value, whereas it was still important for the N+ subgroup who received an adjuvant treatment. Cox multivariate analysis of prognostic value for survival placed total cathepsin D in third position, after nodal invasion and progesterone receptor status, for the entire population, and in first position before progesterone receptor status for the node-positive population. The association of a low cathepsin level and positive progesterone receptors characterized the subgroup of patients with the longest survival. TCD levels played the same role for prediction of the outcome of metastasis.(ABSTRACT TRUNCATED AT 250 WORDS)

Population-specific modulation of 5-HT expression in cultures of embryonic rat rhombencephalon

This study aimed at analyzing the regulation of in vitro serotonin expression by neurons taken from different regions of the embryonic rat rhombencephalon. We studied the influence of co-culture with alarplate tissue using immunocytochemical and biochemical methods. Computer-assisted densitometry was used to estimate the co-culture effects on the serotonin content of the cell bodies. The more dynamic aspects of serotonin expression, such as synthesis and release, were studied by measuring (3H)serotonin newly synthesized from (3H)tryptophan. The density of the immunostaining was significantly decreased in B1,B2 cells by co-culture with both caudal and rostral alar-plate tissue. For B4-B9 cells, only co-culture with rostral alar-plate tissue produced a significant decrease. The de novo synthesis of serotonin was significantly decreased in B1,B2 neurons co-cultured with caudal alar-plate tissue only. Once again, the B4-B9 cells proved to be less influenced by the experimental conditions, as co-culture with both types of alar-plate tissue produced no significant effect. We concluded that the in vitro expression of serotonin can be modulated by environmental factors, but the relative influence of these factors is very different in rostral versus caudal serotonin expressing cell populations.

Alternate day treatment and late effects: the concept of an effective dose per fraction

Although most institutions treat all fields each day, some radiotherapists continue to adopt an alternate day schedule. The resulting daily variations of the dose per fraction in laterally located targets have been analyzed using the linear-quadratic model. Patients with breast carcinoma treated with definitive radiotherapy in 1974-1975 with one field a day were studied. An effective dose per fraction was derived, with a value higher than the average dose per fraction received by the reference point. The greater the fluctuations between the doses per fraction on successive days, the higher the effective dose per fraction. The corresponding cell survival due to alternate treatment as compared to survival with daily treatment depends on the alpha/beta ratio. For a late effect with low alpha/beta ratio, an alternate treatment may lead to almost 10-fold increase in cell kill in these lateral targets such as those responsible for subcutaneous sclerosis as compared to daily treatment of all fields with the same total dose. Taking the average effective dose per fraction in our series, the increase in cell kill was 4-fold. Acute effects would suffer less damage due to alternate treatment because of a high alpha/beta ratio. Treatment on an alternate schedule should be restricted to palliative radiotherapy.

Regulation by estradiol of GABAA and GABAB binding sites in the diencephalon of the rat: an autoradiographic study

Using in vitro quantitative autoradiography we studied the in vivo effects of estradiol on GABAA and GABAB receptors in the rat brain. In all the areas studied (suprachiasmatic nucleus, medial preoptic area, striatum, frontal cortex), estradiol failed to significantly affect the GABAA receptor density. Chronic treatment with estradiol led however in the suprachiasmatic nucleus and the striatum to a decrease in the density of GABAB receptors. GABAB receptor regulation by estradiol was found to be area-specific within the hypothalamus since it was not observed in the medial preoptic area. The down regulation of GABAB receptors in the suprachiasmatic nucleus induced by estradiol treatment might thus explain the inhibitory effect of the steroid on the GABA control of serotonin metabolism we recently reported.

Factors affecting relapse in node-negative breast cancer. A multivariate analysis including the labeling index

Between 1975 and 1982, 167 patients with carcinoma of the breast without axillary lymph node metastases were studied. The thymidine labeling index (LI), representing the percentage of cells in the DNA synthesis phase, was measured in all these patients. High LI values were more frequently encountered in young patients (P = 0.05), in low estrogen receptor (ER) tumor content (P = 0.007) and in high grade tumors (P = 0.0002). The overall 8-year relapse-free survival (RFS) was 68%. Univariate analysis demonstrated that RFS was influenced by histological grading (P = 0.03), ER (P = 0.03), PR (P = 0.02) and LI (P = 0.01). Multivariate analysis using the Cox regression model selected the LI as the single significant prognostic factor with regard to RFS (P = 0.037). These results emphasize the important role of cell proliferation kinetics in defining node-negative breast cancer patients with a high risk of relapse.

Development and use of a computer system in a radiotherapy department: SISGRAD

SISGRAD, the interactive computer system of the Antoine-Lacassagne Cancer Center Radiotherapy Department, has been operational since January 1982. It completes the computerized dosimetry system installed several years earlier and is fully integrated with the institution's central network. SISGRAD is in charge of surveillance of the radiotherapy treatments given by the Center's three radiotherapy units (1400 patients per year); it is also used for administrative purposes in the Department and physically connects all of the Department's operating stations. SISGRAD consists of a series of microcomputers connected to a common mass memory; each microcomputer is used as an intelligent console. SISGRAD was developed to guarantee that the treatments comply with prescriptions, to supply extemporaneous dosimetric data, to improve administrative work, and to supply banks with data for statistical analysis and research. SISGRAD actively intervenes to guarantee treatment quality and helps to improve therapy-related security factors. The present text describes the results of clinical use over a 4-year period. The consequences of integration of the system within the Department are analyzed, with special emphasis being placed on SISGRAD's role in the prevention and detection of errors in treatment prescription and delivery.

The DNA labelling index: a prognostic factor in node-negative breast cancer

The DNA labelling index (LI), representing the fraction of S-phase cells, was studied in 76 patients operated on for breast cancer from 1975 to 1979. No patient had lymph node involvement following axillary dissection (N-), and no adjuvant medical treatment was given. Patients were classified in one of two groups according to the median LI. Patient distribution by age, tumour size, and receptor status was identical in both groups. A higher frequency of grade 3 tumours was noted in the group with an LI above the median value. The cell proliferation rate was an important discriminative factor for metastatic potential. The probability of survival at 8 years for patients with a high LI was significantly lower than that of patients with low LI (36% versus 100%; p less than 0.001). Relapse-free survival at 8 years was respectively 56% and 83% (p less than 0.02). At 2 years, these values were 75% and 100%, indicating the early occurrence of metastases. By contrast, the LI had no prognostic value concerning loco-regional disease recurrence or survival after the appearance of metastasis. No relationship was found in this study between survival and other prognostic factors, namely tumour size, histological grade, or hormone receptor level. The LI currently appears to be the best prognostic factor for N- breast cancer. High risk patients identified by this method could thus be offered adjuvant medical treatment.

CT measurement of lung density: the role of patient position and value for total body irradiation

To obtain more precise data on pulmonary doses in preparation for total body irradiation, the lung density of patients was systematically analyzed in treatment positions using data obtained by computed tomography (CT). With the patient supine, the lung density was not significantly different for the right and left lungs. In contrast, considerable differences were noted between the two lungs in lateral decubitus positions owing to variations in ventilation and perfusion. The relative electron density of lung was also found to decrease with age, dropping to pe = 0.160 at 71 years.

[Adriamycin in low weekly doses. Latest therapy of advanced cancer of the breast]

Thirty-four patients with heavily pretreated advanced breast cancer received adriamycin in weekly doses of 12 mg/m2. Twenty-two patients (18 assessable) had a performance status greater than or equal to 3 (WHO scale). There were 5 (28%) partial responses, 7 (39%) minimal responses, 3 cases (17%) with no change and 3 (17%) with progressive disease; the mean duration of response was superior to 9 months (range: 3-12 months). In the remaining 12 patients, who had resisted a previous multidrug regimen containing adriamycin, there were 3 (25%) partial responses, 2 (17%) minimal responses, 4 cases (33%) with no change and 3 (25%) with progressive disease; the mean duration of response was superior to 5 months (range: 3-11 months). More than 600 weekly injections were administered with only minor blood and digestive tract toxicities. No alopecia was recorded. The total cumulative dose of adriamycin was superior to 900 mg/m2 in 8 patients; none had clinically evident cardiac toxicity. Adriamycin extravasation with subsequent tissue necrosis occurred on 10 occasions due to the frequency of administration; this problem was solved by the systematic installation of central venous catheters.

Value of ultrasound in localizing the internal mammary vessels

The authors discuss the value of ultrasound in determining the depth of the internal mammary vessels und their distance from the medial line. Real-time imaging with a high frequency transducer and a pulsed Doppler system gave highly accurate results. Depending on the intercostal space, the depth of internal mammary vessels varies from 17 to 22 mm. The distance of the internal mammary vessels from the medial line is a function of the intercostal space and the body side, and varies from 25 to 33 mm. Results were found to vary as a function of patient height and weight; likewise, mastectomy was seen responsible for a modification of 2 mm in measurements. Precise localization of the internal mammary vessels allows optimization of radiotherapy, and this rapid examination should be performed just prior to the start of irradiation.

Conservative treatment (chemotherapy/radiotherapy) of locally advanced breast cancer

Once treated almost exclusively by radical surgery, locally advanced breast cancers (Stages III, IV; MO) were later treated solely by irradiation, but local results (cosmesis and recurrences) remained poor. Since 1977, we have used induction chemotherapy in an attempt to treat subclinical metastatic disease while allowing better quality mammary conservation in a greater number of patients. From 1977 to 1980, 25 patients (Stages III, IV; MO) were treated by a sequential association of chemotherapy (Adriamycin [doxorubicin], vincristine, cyclophosphamide, 5-fluorouracil; 3 cycles) followed by irradiation (45 Gy to the breast and nodal areas; 15-30 Gy mammary boost dose). Responses of 50% or more were observed in 86% of the breast lesions and in 80% of nodal lesions. This sequential treatment was always well tolerated. Local recurrences occurred in six patients (24%) who underwent mastectomy without any complications. The survival rate at 4 years is 55%.