- Email: [email protected]
Multidisciplinary treatment approach to locally advanced non-inflammatory breast cancer using chemotherapy and radiotherapy with or without surgery
Radiotherapy and Oncology, 25 (1992) 167-175
© 1992 Elsevier Science Publishers B.V. All rights reserved. 0167-8140/92/$05.00
167
RADION 01066
Multidisciplinary treatment approach to locally advanced non-inflammatory breast cancer using chemotherapy and radiotherapy with or without surgery E m m a n u e l T o u b o u l a, Jean-Pierre Lefranc b, J e a n B l o n d o n b, M a h m u t O z s a h i n ~, Serge M a u b a n c, L a u r e n t H. S c h w a r t z a, Michel Schlienger a, Alain Laugier a a n d R o b e r t A. G u e r i n d Service de Canc~rologie-Radioth~rapie, c Unit~ de Statistiques et lnformatique, Hdpital Tenon, b Service de Chirurgie Gyn~cologique, and d Service de CancOrologie-Radiothdrapie, Groupe Hospitalier PitiO-Salp6tribre, Paris, France
(Received 13 February 1992, revision received 10 June 1992, accepted 9 July 1992)
Key words. Locally advanced breast cancer; Combined modality treatment
Summary Between April 1982 and December 1987, 82 locally advanced non-metastatic and non-inflammatory breast cancers were treated (42 stage IIIA, 40 stage IIIB). The median follow-up is 70 months from the beginning of the treatment. The initial treatment consisted of 4 courses of chemotherapy (doxorubicin, vincristine, cyclophosphamide, 5-fluorouracil) followed by irradiation (45 Gy to the breast and nodal area). A fifth course of chemotherapy was given after radiation therapy. Three different locoregional approaches were proposed depending on the tumoral response. In 32 patients (39%) with residual tumor larger than 3 cm in diameter or located behind the nipple or with multifocal tumors, mastectomy and axillary dissection were performed. Fifty other patients (61%) benefited from conservative treatment: 32 patients (39%) achieved complete remission and received a boost to the initial tumor bed; 18 patients (22%) who had a residual mass less than or equal to 3 cm in diameter were treated by tumorectomy and axillary dissection followed by a boost to the tumorectomy site. After completion of local therapy, all patients received a sixth course of chemotherapy. A maintenance adjuvant chemotherapy regimen without anthracycline was prescribed (12 monthly cycles). Three- and 5-year disease-free survival rates were 81.7% and 72% respectively. Five-year locoregional relapse rate (with or without other sites of failure) was 8.8%. In a multivariate analysis, disease-free survival was significantly influenced by the N-stage ( p < 0.0001), initial tumor size (p = 0.01), and tumor response after initial chemotherapy (p = 0.02). Five-year breast conservation probability was 58.4%. Primary chemotherapy followed by radiotherapy in preoperative doses might permit the selection of some patients for conservative treatment instead of radical surgery but following axillary dissection and conservative treatment, the locoregional complications were not infrequent. Arm lymphedema was noted in 20% (10/ 50) of cases treated with axillary dissection and in 3% (1/32) without axillary dissection. Cosmetic results were satisfactory in only 61.5 },;, of patients following tumorectomy and radiotherapy and in 54.5 ~; of patients treated by radiotherapy alone, There was a substantial incidence of failures occurring more than 5 years after primary treatment and the impact of this aggressive multidisciplinary treatment on long-term survival rate remains to be established.
Introduction The definition of locally advanced breast cancer (LABC) includes primary tumors greater than 5 cm in diameter, or primary tumors of any size involving the skin (ulceration, edema, and/or satellite nodules) or the
chest wall structures. Patients with fixed axillary lymph nodes, and those with clinically detectable internal mammary lymph node metastases are also considered to have LABC [20]. According to a more restrictive definition, T 3 tumors are included only if there is a clinical axillary lymph node involvement, whether fixed
Address for correspondence: Dr Emmanuel Touboul, Service de Canc6rologie-Radioth6rapie, H6pital Tenon, 4 rue de la Chine, 75970 Paris 20,
France.
168 or not [2]. Inflammatory breast cancer is a distinct subset of LABC. Most of these patients have subclinical metastases and the prognosis after local therapy alone is poor [25]. Following mastectomy and axillary lymph node dissection [ 1,18,19], radiotherapy alone [5,6,9,14,22,39], or a combination of preoperative radiotherapy and mastectomy with axillary lymph node dissection [3,32,44], the 5-year disease-free survival is 16-38~o and a locoregional relapse rate of 22-50~o in various series. Over the past 15 years, combined modality therapy has been used to improve local and/or systemic control. In some institutions, chemotherapy was applied as first therapeutic modality, followed by local treatment consisting of either radiotherapy alone or combined with surgery, or surgery alone [4,8,12,23,24,27,29,37,41]. The aim of this combined modality is to treat subclinical metastatic disease with fewer resistant cells [16], and to obtain more breast conservation. We report the results of a prospective nonrandomized study. The treatment includes chemotherapy and radiotherapy with or without surgery in patients suffering from non-inflammatory LABC, without evidence of distant dissemination.
CEA, CA 15-3, gamma-GT levels, liver sonography, bone scans and, if necessary bone marrow biopsy. The range of follow-up was 35-107 months, the average and median follow-up were 70 months from the beginning of the treatment. Age distribution was 26-66 years with an average of 50 years. Thirty-six patients were postmenopausal. Histological classification of tumor biopsies obtained before treatment included 63 infiltrating ductal carcinomas, 17 lobular carcinomas, and 2 mucosecreting carcinomas. Histological grading according to the classification of Scarff, Bloom and Richardson consisted of 6 grade 1, 32 grade 2, 24 grade 3 and 20 non-classified patients. Estrogen and progesteron receptors were not evaluable in most patients. Distribution according to the UICC 1979 classification [26] showed a majority of T 3 and T 4 tumors: 76/82 (92 ~). We have included nine patients with supraclavicular or subclavicular node involvement, which were staged a s N 3 (Table I) according to the UICC 1979 classification [26] because it was used during the period of this study. T 3 tumors were included only if there was clinical axillary lymph node involvement (Table I). The distribution according to tumor size showed that 57~o (47/82) of the patients had local tumors larger than or equal to 6 cm.
Patients and methods
Treatment plan Patients Eighty-two patients with non-inflammatory LABC without distant metastases were consecutively treated between April 1982 and December 1987. The staging procedure included the following: chest X-ray, serum
TABLE
All patients received primary chemotherapy including 4 cycles repeated every 28 days. The combination chemotherapy consisted of doxorubicin 45 mg/m 2 i.v. on day 1; vincristine 1.5 mg/m 2 i.v. on day 2; 5-fluorouracil 500 mg/m2 i.v. per day and cyclophosphamide 300 mg/
I
Clinical staging (TNM Clinical
classification; n
stage grouping
IIIA a
IIIB
a T3No patients
UICC
1979): 82 cases
[26].
TNM
Tumor
size (cm)
classification
42
40
1-2
2.1-3
3.1-5
3
3
.
T2N 2
2
-
-
2
-
-
T3N 1
34
-
-
-
31
3
T3N 2
3
-
-
-
2
1
T1N3 b
1
1
.
T3N 3
3
-
-
-
3
-
TaN o
19
-
-
7
8
4
TaN 1 TaN 2
14
-
-
-
8
6
2
-
-
-
1
1
T4N 3
1
.
were excluded.
b N3 r e f l e c t s t h e p r e s e n c e o f s u b c l a v i c u l a r
lymph
node involvement.
.
> 8
T1N 2
.
.
5.1-8
.
.
.
.
.
.
1
169 m 2 i.v. per day on days 2, 3 and 4. Complete blood counts were performed on day 1 of each cycle of chemotherapy. Three weeks after the last cycle of chemotherapy all patients received external irradiation using 6°Co (1.25 MV photons). The whole breast, chest wall and ipsilateral regional lymph nodes (supraclavicular, axillary and internal mammary nodes) were irradiated to a total dose of 45 Gy in 23 fractions over 31 days. The patient was positioned supine with the ipsilateral arm elevated at least to form a right angle to the torso. The irradiation of the whole breast and the whole thickness of the thoracic wall was accomplished with parallel opposed medial and lateral tangential beams. The dose was specified at the intersection of the beam axes in the central plane. The upper field border included the axillary tail of the breast and avoided the medial surface of the arm in abduction. Inferiorly, the fields extended to a level of 1-2 cm below the margin of the breast. In order to reduce the irradiated lung volume, the collimator was rotated and the beams were tilted 5-10 degrees from 180 ° in order to achieve a straight line of the borders of the field at the lung. Each field was treated every day. Wedges were used to improve dose homogeneity. Minimum and maximum dose in the central plane was greater than 95% and inferior to 110~o respectively. The supraclavicular and axillary areas were treated with a single anterior photon beam. Part of the shoulder joint and humerus was protected by blocks. Blocks were also used to protect spinal cord. The supraclavicular dose was calculated at a depth of 3 cm. The axillary dose from this anterior portal was calculated at the mid-plane of the A-P diameter. In order to supplement the dose to the midaxillary, a smaller posterior field was used. At mid-plane, the axillary dose was 1.50 Gy from anterior field and 0.50 Gy from posterior field per fraction. A separate internal mammary lymph node field was used. The internal mammary lymph nodes were irradiated with an anterior portal. Usually, a field extending 5 cm ipsilaterally from the midline and 1 cm controlaterally over the first five intercostal spaces covered this region with target depth at 3 cm. Four weeks after the end of irradiation, a fifth chemotherapy cycle was done with anthracyclin at a dose of 30 mg/m 2. Four weeks after the fifth cycle (29 weeks after the beginning of treatment), three different locoregional therapeutic approaches were proposed depending on the clinical, mammographic and sonographic response status after initial chemotherapy and radiotherapy, on the site of the tumor, on its multifocality, and on the volume of the breast compared to the residual
tumor. Decisions about local therapy were made 8 weeks after irradiation. Local treatment was planned as follows. (1) When residual tumor size was greater than 3 cm (largest diameter) after treatment, when it had a retroaerolar localization, or when initial tumor was plurifocal, Patey type mastectomy and axillary dissection was performed. The axillary dissection was limited to levels I and II lymph nodes. (2) When the residual tumor was less than or equal to 3 cm in its largest diameter, the therapeutic approach consisted in a conservative surgical excision with large tumorectomy and axillary dissection limited to levels I and II lymph nodes, followed by an interstitial implantation of the initial tumor site and the surrounding tissues, using Pierquin's technique [36]. A boost of 20 Gy was administered, calculated on the 85 ~o reference isodose [ 15,31,35]. Brachytherapy was done using wire implants of 192Ir sources with linear activity of 0.81.2 mCi/cm. The dose-rate on the reference isodose was between 0.55 and 0.80 Gy/h. (3) When the primary tumor and palpable node had disappeared, radiation therapy was completed by a booster dose on the initial tumor bed. It was administered either by external irradiation using 6°Co (1.25 MV) to a dose of 30 Gy in 15 fractions given in 19 days, or by interstitial brachytherapy to a dose of 25 Gy on the 85 To reference isodose volume. The purpose ofbrachytherapy was to reduce the dose delivered to normal mammary tissues and to try to obtain better cosmetic results, by decreasing the risk of cutaneous sequelae. A boost of 15 Gy (6°Co) was delivered by a direct axillary field in axillary lymph nodes in 8 fractions over 10 days, at 3 cm depth to the skin along the beam axis. Following completion of local therapy (3 weeks after the end of radiotherapy or after surgery), each patient received a sixth cycle of polychemotherapy. The dose of anthracyclin was limited to 30 mg/m ~. An adjuvant chemotherapy composed of vincristine 1.5 mg/m 2 i.v. on day 1; 5-fluorouracil 500 mg/m z i.v. and cyclophosphamide 300 mg/m z i.v. on days 2, 3 and 4 was prescribed (12 monthly cycles). After surgery, if there was an involvement of axillary lymph nodes, or if the diameter of the residual tumor was greater than 2 cm, the adjuvant treatment was reinforced by addition ofmethotrexate 24 mg/m2/day i.m. on days 2 and 3. All patients completed the entire program of chemotherapy.
Statistical analysis Proportions and means were compared by the chisquare test and the Student's t-test respectively. Overall
170 and disease-free survival rates from the beginning of treatment, were estimated according to the KaplanMeier method [ 28 ]. Confidence intervals (CI) were calculated from standard errors. The comparison of the overall and disease-free survival rates between the two groups were done using the Mantel method [30]. The analysis of the prognostic factors was done by univariate analyses and then by multiple regression analysis (Cox model) based on disease-free survival probability
[10]. Results
At the end of the fourth chemotherapy cycle Primary tumor Induction chemotherapy produced a significant clinical tumor regression (>/50 ~o in its largest diameter) in 55 ~o (45/82) of the cases. All evidence of tumor disappeared in 10~o (8/82) of the cases. Progressive disease was not observed in any of the 82 patients during the four initial courses of chemotherapy. None of the factors tested (tumor size, histological type, histological grade, menopausal status, and age) had a statistically significant influence on the tumor response following chemotherapy. Palpable nodes Of the 63 patients having a palpable lymph node, a complete clinical nodal remission was obtained for 73 ~o (46/63), and a partial remission for 11 ~o (7/63). Nodal response following chemotherapy was not significantly influenced by the N-stage. Following 45 Gy of external irradiation given in a period of 4.5 weeks Primary tumor Complete tumor regression was achieved in 42.5 % (35/ 82) of the cases, and partial regression in 34~o (28/82), while the other 19 tumors remained stable. None of the above mentioned tested parameters showed a statistically significant influence on the tumor response after radiotherapy. Palpable nodes Complete clinical nodal regression was observed in 87 ~o (55/63) of the cases, and a partial nodal regression in 8~o (5/63) of the patients. The N-stage did not significantly influence the nodal response rate after radiotherapy.
Following primary chemotherapy and radiotherapy of 45 Gy in 4.5 weeks A conservative locoregional treatment could be proposed to 50 patients (61 ~o). For 32 patients (39 ~o) who achieved complete remission, a booster irradiation dose was given by external radiotherapy to 14 patients (44 ~o) and by interstitial implantation to 18 patients (56~o). For the other 18 patients (22~o), who had a residual mass ~ 3 cm in diameter, a tumorectomy with axiUary dissection was done followed by a booster dose of radiotherapy in the tumorectomy site and surrounding tissues, using interstitial implantation. In 32 patients (39~o), mastectomy and axillary dissection were performed. Twenty-nine cases had a residual mass > 3 cm in diameter, whereas the other three achieved complete tumor regression before surgery. These patients had the following tumor types: 1 bifocal tumor, 1 retro-aerolar large tumor, and 1 tumor with large nodes classified a s N 3 (subclavicular node) without substantial nodal reduction. The possibility of breast conservation was significantly related to the initial size of the primary tumor. Out of 32 mastectomy patients, 25 (78~o) had a tumor >~6 cm in diameter, versus 22 (44~o) of 50 conservative treatments (chi-square -- 7.9, p ---0.005).
Analysis of the relationship between clinically assessed tumoral and nodular response, and histopathological findings on the 50 operated patients Primary tumor Before surgery (17 weeks after the fourth cycle of primary chemotherapy and 9 weeks after irradiation of 45 Gy in 4.5 weeks), 47 patients did not achieve complete clinical remission. The pathological assessment showed a residual tumor greater than 1 cm in diameter in 32 cases, a residual tumor ~<0.5 cm in diameter in 9 cases, and no evidence of tumor in 6 cases. For the other 3 cases, the pathological examination confirmed complete clinical remission before surgery. There was good correlation between the tumor response and the histopathological findings after surgery (R = - 0.47, p<0.001). Axillary dissection In 13 cases clinically classified as No, four had involved nodes at pathological examination. In 29 cases with palpable nodes before treatment, and considered as complete clinical remissions at the time of surgery, 13 showed nodal involvement; three patients out of 13 had 4 involved nodes or more. Among eight other cases with palpable nodes, who did not achieve complete
171 nodal remission, seven had tumoral involvement; four out of seven had more than four nodes infiltrated.
Survival and locoregional control Survival The 3- and 5-year overall survival rates calculated from the first day of treatment were 85.3 ~o (70 patients, 95 % CI: 79-91) and 81.3% (47 patients, 95~o CI: 73-90) respectively. The 3- and 5-year disease-free survival rates were 81.7% (67 patients, 95~o CI: 73-90) and 72% (44 patients, 95~o CI: 62-82) respectively (Fig. la). No patient died of intercurrent disease. In univariate analyses (Table II), 8 factors were found to have significant influence on survival: N-stage, initial tumor size, histological grade, tumor response after four initial cycles of chemotherapy, tumor response following chemotherapy and radiotherapy, and for the operated patients the size of residual tumor, axillary nodal status at pathological examination and number of involved axillary lymph nodes. Disease-free survival was influenced by 7 factors: N-stage, histological grade, tumor response after initial chemotherapy, tumor response following chemotherapy and radiotherapy, and for the operated patients the size of residual tumor, axillary nodal status and the presence of 4 or more positive axillary lymph nodes.
¢7~ 8 2
100 ~" 80
"
7
Overall survival Disease-freesurvival Locoregionalrelapse rate
~
~ ~
3
60 ~
2
59
'+~
38
m
40-
~w............. •
23 .,,.,......,......:
20 -
43.i?~ t
0
24
48 72 months
96
120
b
% 100-
RT alone (n=32) Tumorectomy+ RT (n=18) Mastectomy(n=32)
806040-
P > 0.05
20O'
"
0
"
24
-
i
48
•
i
72
•
i
96
-
i
120
months
Fig. 1. (a) Overall survival, disease-freesurvival (DFS) and locoregionat relapse rate in all patients (n = 82). (b)Locoregionalrelapse rate accordingto locoregionaltreatment.
In the multivariate analysis (Table III), two prognostic factors had a significant impact on overall survival: N-stage and initial tumor size. Disease-free survival was significantly influenced by the N-stage, initial tumor size and tumor response after initial chemotherapy.
Locoregional control Whatever the follow-up, following conservative treatment by tumorectomy and axillary dissection and radiotherapy (n = 18), 5 patients had a relapse (3 local, 2 distant metastases). Following radiotherapy alone (n--32), 11 patients relapsed (1 local, 3 distant metastases, 3 local with distant metastases, 1 axillary nodal with distant metastases and 3 local with axillary nodal recurrence and distant metastases). In the nonconservative group (n = 32), there were 12 relapses (10 distant metastases, 1 local with distant metastases and 1 local with axillary nodal recurrence and distant metastases). Five patients developed axillary relapses: they were always associated with distant metastases. None of the patients failed in the supraclavicular area. The 5-year actuarial local relapse rate without regional and distant failure was 6.5% (47 patients, CI: 1-13). We observed 4 isolated local tumor recurrences following conservative treatment (at 30, 30, 54 and 84 months after the beginning of treatment); 3 of them were salvaged after mastectomy with no evidence of disease at 6, 12 and 40 months following mastectomy. The fourth developed distant metastases 40 months after mastectomy. Actuarial locoregional relapse curves with or without distant failure according to the locoregional treatment modality are shown in Fig. lb. The 5-year actuarial overall locoregional relapse rate was 8.8~o (CI: 3-15); it was 12.6~o (CI: 1-24) following radiotherapy alone, 5.9~o (CI: 3-9) following tumorectomy and radiotherapy, and 6.3 % (CI: 0-14) following mastectomy. The 5-year locoregional relapse risk (with or without distant failure) was not influenced significantly by the locoregional treatment ( p > 0.05). In univariate analyses, the 5-year local relapse risk with and without regional or distant failure was influenced significantly by two prognostic factors: N-stage (p=0.01) and histological grade (p = 0.002), whereas the N-stage was the only significant prognostic factor in the multivariate analysis (p = 0.05). In our experience, the 5-year probability of breast conservation was 58.4% (34 patients, 95~o CI: 48-69). After non-conservative surgery, breast reconstruction with prosthesis associated myo-cutaneous flaps was performed in two cases at 4 and 5 years after the treatment without post-operative complication. These
172 TABLE II Influence of prognostic factors on the outcome-univariate analyses. Factor
Age Menopausal status Stage Tumor size (cm) N-stage
Histological type Histological grade (SBR) Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Residual tumor after surgery (cm) AxiUary nodal status N+
Group
< 50 /> 50 premenopausal postmenopausal IliA IIIB <6 >-6 No N1 N2 N3 ductal lobular 1+ 2 3 CR + PR NR CR + PR NR < 1 >- 1 NN+ 1-3 nodes >-4 nodes
n
35 47 45 37 42 40 35 47 19 48 10 5 63 17 38 24 53 29 63 19 18 32 26 24 9 15
Overall survival 5-year survival
Patients at risk
Yo
(n)
83 80 82 80 81 82 94 72 100 81 70 40 79 88 88 66 86 72 87 63 100 72 92 71 78 68
(22) (26) (25) (23) (25) (23) (24) (24) (15) (27) (7) (1) (33) (15) (19) (9) (35) (13) (42) (6) (14) (13) (15) (12) (6) (7)
DFS p-Value
0.53 0.59 0.69 0.01
0.01
0.58 0.002 0.04 0.005 0.047 0.03 0.09
5-year survival
Patients at risk
%
(n)
68 75 69 75 76 68 82 64 88 71 70 20 72 77 78 58 83 52 82 38 87 62 89 53 67 44
(20) (25) (22) (23) (25) (20) (23) (22) (14) (25) (7) (1) (31) (14) (17) (8) (34) (11) (40) (5) (13) (13) (15) (11) (5) (6)
p-Value
0.23 0.33 0.68 0.13
0.003
0.72 0.006 0.006 0.0004 0.02 0.004 0.03
DFS, disease-free survival; SBR, Scarff, Bloom and Richardson; CR, complete remission; PR, partial remission; NR, non-responder.
two patients are living with no evidence of disease at 2 years after reconstruction.
Cancer of the contro-lateral breast Cancer developed the contro-lateral breast in three cases (two patients after non-conservative treatment at 24 months of follow-up and one case after tumorectomy and radiotherapy at 5.5 years of follow-up). Out of these three patients with contro-lateral cancer, two had conservative treatment whereas the last one had radical surgery, These patients are living with no evidence of disease at 2, 2.5 and 3 years, respectively, after the treatment of contro-lateral breast.
Complications Tolerance to treatment was good. Although nausea, vomiting and alopecia were experienced by all patients,
a moderate hematologic toxicity was a constant finding following induction chemotherapy (grade 1 according to the W H O classification) [33]. On the other hand, following completion of local treatment, we observed three chronic leukopenias of grade 2 (~<3x 10 9 leukocytes/1), and six grade 1 paresthesias of the extremities following adjuvant chemotherapy requiring the discontinuation of vincristine. Locoregional complications were not infrequent. Arm lymphedema was observed in 20 ~o (10/50) of the patients treated with axillary dissection and in one patient out of 32 (3 To) treated without axillary dissection. Limitation of shoulder movements was noted in 22 ~o (7/32) of the patients treated with mastectomy and axillary dissection, but it was not observed following tumorectomy and axillary dissection, or radiotherapy without surgery. None of the patients developed congestive heart failure, extended pulmonary fibrosis, brachial plexopathy, or rib fracture.
173 TABLE III Influence of prognostic factors on the outcome-multivariate analysis. Factor
DFS
Overall survival fl-Coefficient
p-Value
/~-Coefficient
p-Value
All patients #7 = 82)
N-stage Tumor size Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Histological grade (SBR)
1.15 0.21 -0.76
0.0003 0.005 0.22
1.14 0.17 - 1.17
< 0.0001 0.01 0.02
0.003 - 0.17
0.99 0.45
0.04 - 0.04
0.90 0:79
1.51 0.07 1.79
0.02 0.67 0.27
1.11 0.02 - 1.59
0.005 0.82 0.08
- 0.92 0.19 - 0.09 - 0.004
0.51 0.31 0.64 0.92
- 0.63 0.17 0.13 0.05
0.42 0.25 0.23 0.87
Operated patients (n = 50)
N-stage Tumor size Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Residual tumor Positive nodes Histological grade (SBR)
DFS, disease-free survival; SBR, Scarff. Bloom and Richardson.
Cosmetic results
Cosmetic results after conservative treatment at 5 years from the beginning of the treatment were assessed according to Harris et al. [21]. For the present study, 35 patients treated by conservative treatment were available. Following tumorectomy and radiotherapy, results were good in 61.5 ~o (8/13), fair in 23 ~o (3/13) and poor in 15.5 }'o (2/13). Following radiotherapy without surgery, results were good in 54.5~, (12/22), fair in 27~,,, (6/22) and poor in 18.5 (4/22) of the patients. Discussion
The present study concerns a series of 82 patients with non-inflammatory LABC, without evidence of distant metastases, treated by primary chemotherapy and radiotherapy with or without surgery and adjuvant chemotherapy after completion of local therapy. The overall tumor response rate after initial chemotherapy was 65~o and complete tumor response rate was 10}o. These results are comparable to other published reports using similar chemotherapy regimens, with complete remission rates between 5 and 18}'o [4,12,24,29]. Following aggressive multidisciplinary approach as applied in the present study, 5-year overall and disease-free survival rate was 81.3 }'o and 72}o respectively. There are studies suggesting possible improved 5-year overall and disease-free survival following a combined modality approach using initial cyto-reductive chemotherapy,
surgery, and/or radiotherapy in the treatment of LABC [8,12,17,25,27,41]. On the other hand, some authors consider that the influence of these aggressive therapeutic approaches in this category of patients is not proved [ 7,13,40,42 ]. Three randomized studies have shown that the addition of adjuvant chemotherapy in patients with stage III breast cancer receiving local radiation or local radiation and surgery, had no effect on the survival [13,38,401. In our experience, a multidisciplinary treatment approach with primary chemotherapy and radiotherapy in preoperative doses (45 Gy in 4.5 weeks) might permit the selection of some patients for conservative treatment. In our series, the 5-year locoregional relapse rate was low (8.8 }'o) with a 5-year breast conservation rate of 58.4}'o. In the experience of Hortobagyi [24], following neo-adjuvant chemotherapy, and radiotherapy or surgery, local recurrences were observed in 21 }'o of the patients. Jacquillat [27], following initial chemotherapy and radiotherapy, reported that the 5-year actuarial locoregional relapse rate was 18~, for T3, and 19}'o for T4 tumors. In series using locoregional treatment alone, the 5-year local relapse rate appears to be high, from 22 to 50~o [1,5,9,14,22,39]. In our multivariate analyses, the most significant and independent factors influencing disease-free survival were the N-stage, initial tumor size, and overall tumor response after initial chemotherapy. Five-year overall survival was 100}'o for No, 81.1%£, for N t, 70°0 for N2,
174 and 40~o for N 3 tumors (p = 0.01, univariate analysis). The N-stage influenced not only the survival rate but also the local control rate. Five-year overall survival was 93.8~o for tumors less than 6 c m in diameter, whereas it was 72.2~o for tumors whose diameter was 6 cm or more (p = 0.01, univariate analysis). In most studies, N-stage and initial tumor size are reliable prognostic factors [8,14,23,26,43,44]. Our results showed that tumor regression following primary chemotherapy was a significant prognostic factor. Five-year diseasefree survival was 82.5~o for complete and partial responders, whereas it was 52.1~o for non-responders (p = 0.006, univariate analysis). This prognostic factor is also reported in other series [8,24,27,34]. The analysis of the locoregional complications and the late sequalae after axillary dissection and the fair or poor cosmetic results seemed to be more frequent than series using radiotherapy alone or preoperative radiation therapy [3,14,39,43,44]. Some authors consider that adjuvant chemotherapy increases the late locoregional complications [ 11 ]. Hematological toxicity and cardiotoxicity of our treatment modality was relatively
moderate. Others, whose treatment modalities are similar, reported more severe hematological toxicity rates in 3-32~o [3,12,24], cardiotoxicity in 1.8-7.5~o [4,8,12,24], and pulmonary fibrosis in 6-9~o [4,12] of the cases. These complications were observed after a total irradiation dose greater than or equal to 70 Gy in 7 weeks to the whole breast, and after a cumulative doxorubicin dose less than or equal to 600 mg/m 2.
Conclusion A multidisciplinary treatment approach for locally advanced noninflammatory breast cancer using primary chemotherapy followed by radiotherapy in preoperative doses may permit the selection of some patients for conservative treatment, but locoregional complications after axillary dissection followed by external irradiation and fair or poor cosmetic results after conservative treatment are not infrequent. The impact of this treatment modality on long-term survival and local control remains to be proved.
References 1 Adair, F., Berg, J., Joubert, L. and Robbins, G . F . Long-term follow-up of breast cancer patients the 30-year report. Cancer 33: 1145-1150, 1974. 2 American Joint Committee for 'Cancer Staging and End Results Reporting Staging of Cancer of the Breast. In: Manual of Staging of Cancer, Whiting Press, Chicago, 1983. 3 Bedwineck, J. M., Rat, D. V., Perez, C. A., Lee, J. and Fineberg, B. Stage III and localized stage IV breast cancer: irradiation alone vs. irradiation plus surgery. Int. J. Radiat. Oncol. Biol. Phys. 8: 31-36, 1982. 4 Bedwineck, J. M., Ratkin, G. A., Philpott, G.W., Wallak, M. and Perez, C.A. Concurrent chemotherapy and radiotherapy for non-metastatic stage IV breast cancer. Am. J. Clin. Oncol. 6: 159-165, 1983. 5 Bruckman, J. E., Harris, J. R., Levene, M. B., Chaffey, J. T. and Hellman, S. Results of treating stage III carcinoma of the breast by primary radiation therapy. Cancer 43: 985-993, 1979. 6 CaUe, R., Pilleron, J. P , Schlienger, P. and Vilcoq, J.R. Conservative management of operable breast cancer. Cancer 42: 2045-2053, 1978. 7 Casper, E. S., Guidera, C. A., Bosi, G.J., Hakes, T. B., Kaufman, R. J., Shurgot, B. and Kinne, D.W. Combined modality treatment of locally advanced breast cancer: adjuvant combination chemotherapy with or without doxorubicin. Breast Cancer Res. Treat. 9: 39-44, 1987. 8 Chauvergne, J., Durand, M., Mauriac, L., David, M., Pigneux, J., Dilhuydy, M. H., Avril, A., Den6poux, R., Mar6e, D. and Bonichon, F. Traitement combin6 de 270 cancers mammaires localement &endus: r~sultats d'un programme th~rapeutique contr616. In: Neoadjuvant Chemotherapy, 169, pp. 225-230. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Ltd, Paris, 1988.
9 Chu, A. M., Cope, O., Doucette, J. and Curran, B. Non metastatic locally advanced cancer of the breast treated with radiation. Int. J. Radiat. Oncol. Biol. Phys. 10: 2299-2304, 1984. 10 Cox, D.R. Regression models and life table. J. Roy. Statist. Soc. 34: 187-220, 1972. 11 Danoff, B. F., Goodman, R. L., Glick, J. H., Hailer, D. G. and Pajak, T . F . The effect of adjuvant chemotherapy on cosmesis and complications in patients with breast cancer treated by definitive irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9: 1625=_ 1630, 1983. 12 De Lena, M., Zucali, R., Viganotti, G., Valagussa, P. and Bonadonna, G. Combined chemotherapy-radiotherapy approach in locally advanced (T3b-T4) breast cancer. Cancer Chemother. Pharmacol. 1: 53-59, 1978. 13 Derman, D. P., Browde, S., Kessel, I. L., De Moor, N. G., Lange, M., Dansey, R., Seymour, L. and Bozwoda, W.R. Adjuvant chemotherapy (C.M.F.) for stage III breast cancer: a randomized trial. Int. J. Radiat. Oncol. Biol. Phys. 17: 257-261, 1989. 14 Dubois, J. B., Salomon, A., Gary-Bobo, J., Pourquier, H. and Pujol, H. Exclusive radical radiation therapy in breast carcinoma. Radiother. Oncol. 20: 24-29, 1991. 15 Dutreix, A., Marinello, G., Pierquin, B., Chassagne, D. and Moulard, J.P. Drveloppement actuel du syst~me de Paris. Deuxi~me partie: les dispositifs en triangles. J. Radiol. Electrol. 60: 319-325, 1979. 16 Goldie, J. H. and Coldman, A.J. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat. Rep. 63: 1727-1733, 1979. 17 Graham, M. V., Perez, C. A., Kuske, R. R., Garcia, D. M. and Fineberg, B. Locally advanced (non inflammatory) carcinoma of the breast: results and comparison of various treatment modalities. Int. J. Radiat. Oncol. Biol. Phys. 21: 311-318, 1991.
175 18 Haagensen, C. D. and Stout, A.P. Carcinoma of the breast II: criteria of operability. Ann. Surg. 118: 859-870, 1943. 19 Haagensen, C . D . Treatment of curable cancer of the breast. Int. J. Radiat. Oncol. Biol. Phys. 2: 975-980, 1977. 20 Haagensen, C.D. Clinicalclassification of the stage of advancement of breast carcinoma. In: Disease of Breast, pp. 851-863. Editor: C.D. Haagensen. W.B. Saunders, Philadelphia, PA, 1986. 21 Harris, J. R., Levene, M. B., Svensson, G. and Hellman, S. Analysis of cosmetic results following primary radiation therapy for stage I and II carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 5: 257-261, 1979. 22 Harris, J. R., Sawicka, J., Gelman, R. and HeUman, S. Management of locally advanced carcinoma of the breast by primary radiation therapy. Int. J. Radiat. Oneol. Biol. Phys. 9: 345-349, 1983. 23 Hery, M., Namer, M., Moro, M., Boublil, J. L. and Lalanne, C. Conservative treatment (chemotherapy/radiotherapy) of locally advanced breast cancer. Cancer 57: 1744-1749, 1986. 24 Hortobagyi, G. N., Blumenschein, G. R., Spanos, W., Montague, E. D., Buzdar, A. U., Yap, H. Y. and Schell, F. Multimodal treatment of locoregionally advanced breast cancer. Cancer 51: 763-768, 1983. 25 Hortobagyi, G . N . Comprehensive management of locally advanced breast cancer. Cancer 66: 1387-1391, 1990. 26 International Union Against Cancer (UICC). TNM Classification of Malignant Tumours, Fourth Edition. UICC, Geneva, 1979. 27 Jacquillat, C., Weill, M., Baillet, F., Borel, C., Auclerc, G., De Maublanc, M. A., Housset, M., Forget, G., Thill, L., Soubrane, C. and Khayat, D. Results of neo-adjuvant chemotherapy and radiation therapy in the breast-conserving treatment of 250 patients with all stages &infiltrative breast cancer. Cancer 66:119129, 1990. 28 Kaplan, E. L. and Meier, P. Non-parametric estimation from incomplete observations. J. Am. Statist. Assoc. 53: 457-481, 1958. 29 Lesnick, G., Perloff, M., Korzun, A., Chu, F., Holland, J. F., Thirwell, M., Ellison, R., Carey, R., Leone, L., Weinberg, V., Rice, M. and Wood, W. Neoadjuvant chemotherapy for stage Ill breast cancer: 5-year report of CALGB 7784. In: Neoadjuvant Chemotherapy, 169, pp. 185-188. Editors: C, Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Lid, Paris, 1988. 30 Mantel, N. and Haenszel, W. Chi-square tests with one degree of freedom: extensions of Mantel-Haenszel procedure. J. Am. Statist. Assoc. 58: 690-700, 1963. 31 Marinello, G., Dutreix, A., Pierquin, B. and Chassagne, D. D~veloppement actuel du syst6me de Paris. J. Radiol. Electrol. 59: 621-626, 1978. 32 Martijn, H., Caspers, R. J. L., Welvaart, K., Van De Velde, C., Hermans, J. and Leer, J. W, H. Locally advanced breast cancer: radiotherapy alone versus surgery and radiotherapy. Radiother. Oncol. 16: 23-29, 1989.
33 Miller, A.B., Hoogstraten, B., Staquet, M. and Winkler, A. Reporting results of cancer treatment. Cancer 47: 207-214, 1981. 34 Namer, M., H6ry, M., Abbes, M., Frenay, M, and Ramaioli, A. Influence of the response rate obtained by neoadjuvant chemotherapy on the survival of patients with locally advanced breast cancer. In: Neoadjuvant Chemotherapy, 169, pp. 177-181. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque. I.N.S.E.R.M., John Libbey Eurotect Ltd, Paris, 1988. 35 Pierquin, B., Dutreix, A., Paine, C. H., Chassagne, D., Marinello, G. and Ash, D. The Paris system in interstitial radiation therapy. Acta Radiologica 17: 33-48, 1978. 36 Pierquin, B., Owen, R., Maylin, C., Otmezquin, Y., Raynal, M., Mueller, W. and Hannoun, B. Radical radiation therapy of breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 6: 17-24, 1980. 37 Ragaz, J., Manji, M., Plenderleith, I. H., Knowling, M., Olivotto, L., Basco, V., Hardman, J., Worth, A. and Kuusk, V. Analysis of the British Columbia study of preoperative (neo-adjuvant) therapy of breast cancer in stage III disease. In: Neoadjuvant Chemotherapy, 169, pp. 249-256. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Ltd. Paris, 1988. 38 Rubens, R. D., Bartelink, H., Engelsman, E., Hayward, J. L., Rotmensz, N., Sylvester, R., van der Shueren, E., Papadiamantis, J., Vassilaros, S. D., Wildier, J. and Winter, P.J. Locally advanced breast cancer: the contribution of cytotoxic and endocrine treatment to radiotherapy. An EORTC Breast Cancer Cooperative Group Trial (10792). Eur. J. Cancer Clin. Oncol. 25: 667-678, 1989. 39 Sheldon, T., Hayes, D.F., Cady, B., Parker, L., Osteen, R., Silver, B., Recht, A., Come, S., Henderson, C. and Harris, J.R. Primary radiation therapy for locally adwmced breast cancer. Cancer 60: 1219-1225, 1987. 40 Shaake-Koning, C., Van Der Linden, E.H., Hort, G. and Engelsman, E. Adjuvant chemo and hormonal therapy in locally advanced breast cancer: a randomized clinical study. Int. J. Oncol. Biol. Phys. 11: 1759-1763, 1985. 41 Sponzo, R. W., Cunningham, T. J. and Caradonna, R.R. Management ofnon-resectable (stage III) breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 5: 1475-1478, 1979. 42 Strom, E.A., Mc Neese, M.D., Fletcher, G . H . , Ramsdahl, M. A., Montague, E. D. and Oswald, M.J. Results of mastectomy and post-operative irradiation in the management oflocoregionally advanced carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 21: 319-323, 1991. 43 Vilcoq, J. R., Fourquet, A., Jullien, D., Gautier, C., Calle, R. and Ghossein, N.A. Prognosis significance of clinical model involvement in patients treated by radical radiotherapy for a locally advanced breast cancer. Am. J. Clin. Oncol. 6: 625-628, 1984. 44 Zucali, R., Uslenghi, C., Kenda, R. and Bonadonna, G. Natural history and survival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer 37: 1422-1431, 1976.
© 1992 Elsevier Science Publishers B.V. All rights reserved. 0167-8140/92/$05.00
167
RADION 01066
Multidisciplinary treatment approach to locally advanced non-inflammatory breast cancer using chemotherapy and radiotherapy with or without surgery E m m a n u e l T o u b o u l a, Jean-Pierre Lefranc b, J e a n B l o n d o n b, M a h m u t O z s a h i n ~, Serge M a u b a n c, L a u r e n t H. S c h w a r t z a, Michel Schlienger a, Alain Laugier a a n d R o b e r t A. G u e r i n d Service de Canc~rologie-Radioth~rapie, c Unit~ de Statistiques et lnformatique, Hdpital Tenon, b Service de Chirurgie Gyn~cologique, and d Service de CancOrologie-Radiothdrapie, Groupe Hospitalier PitiO-Salp6tribre, Paris, France
(Received 13 February 1992, revision received 10 June 1992, accepted 9 July 1992)
Key words. Locally advanced breast cancer; Combined modality treatment
Summary Between April 1982 and December 1987, 82 locally advanced non-metastatic and non-inflammatory breast cancers were treated (42 stage IIIA, 40 stage IIIB). The median follow-up is 70 months from the beginning of the treatment. The initial treatment consisted of 4 courses of chemotherapy (doxorubicin, vincristine, cyclophosphamide, 5-fluorouracil) followed by irradiation (45 Gy to the breast and nodal area). A fifth course of chemotherapy was given after radiation therapy. Three different locoregional approaches were proposed depending on the tumoral response. In 32 patients (39%) with residual tumor larger than 3 cm in diameter or located behind the nipple or with multifocal tumors, mastectomy and axillary dissection were performed. Fifty other patients (61%) benefited from conservative treatment: 32 patients (39%) achieved complete remission and received a boost to the initial tumor bed; 18 patients (22%) who had a residual mass less than or equal to 3 cm in diameter were treated by tumorectomy and axillary dissection followed by a boost to the tumorectomy site. After completion of local therapy, all patients received a sixth course of chemotherapy. A maintenance adjuvant chemotherapy regimen without anthracycline was prescribed (12 monthly cycles). Three- and 5-year disease-free survival rates were 81.7% and 72% respectively. Five-year locoregional relapse rate (with or without other sites of failure) was 8.8%. In a multivariate analysis, disease-free survival was significantly influenced by the N-stage ( p < 0.0001), initial tumor size (p = 0.01), and tumor response after initial chemotherapy (p = 0.02). Five-year breast conservation probability was 58.4%. Primary chemotherapy followed by radiotherapy in preoperative doses might permit the selection of some patients for conservative treatment instead of radical surgery but following axillary dissection and conservative treatment, the locoregional complications were not infrequent. Arm lymphedema was noted in 20% (10/ 50) of cases treated with axillary dissection and in 3% (1/32) without axillary dissection. Cosmetic results were satisfactory in only 61.5 },;, of patients following tumorectomy and radiotherapy and in 54.5 ~; of patients treated by radiotherapy alone, There was a substantial incidence of failures occurring more than 5 years after primary treatment and the impact of this aggressive multidisciplinary treatment on long-term survival rate remains to be established.
Introduction The definition of locally advanced breast cancer (LABC) includes primary tumors greater than 5 cm in diameter, or primary tumors of any size involving the skin (ulceration, edema, and/or satellite nodules) or the
chest wall structures. Patients with fixed axillary lymph nodes, and those with clinically detectable internal mammary lymph node metastases are also considered to have LABC [20]. According to a more restrictive definition, T 3 tumors are included only if there is a clinical axillary lymph node involvement, whether fixed
Address for correspondence: Dr Emmanuel Touboul, Service de Canc6rologie-Radioth6rapie, H6pital Tenon, 4 rue de la Chine, 75970 Paris 20,
France.
168 or not [2]. Inflammatory breast cancer is a distinct subset of LABC. Most of these patients have subclinical metastases and the prognosis after local therapy alone is poor [25]. Following mastectomy and axillary lymph node dissection [ 1,18,19], radiotherapy alone [5,6,9,14,22,39], or a combination of preoperative radiotherapy and mastectomy with axillary lymph node dissection [3,32,44], the 5-year disease-free survival is 16-38~o and a locoregional relapse rate of 22-50~o in various series. Over the past 15 years, combined modality therapy has been used to improve local and/or systemic control. In some institutions, chemotherapy was applied as first therapeutic modality, followed by local treatment consisting of either radiotherapy alone or combined with surgery, or surgery alone [4,8,12,23,24,27,29,37,41]. The aim of this combined modality is to treat subclinical metastatic disease with fewer resistant cells [16], and to obtain more breast conservation. We report the results of a prospective nonrandomized study. The treatment includes chemotherapy and radiotherapy with or without surgery in patients suffering from non-inflammatory LABC, without evidence of distant dissemination.
CEA, CA 15-3, gamma-GT levels, liver sonography, bone scans and, if necessary bone marrow biopsy. The range of follow-up was 35-107 months, the average and median follow-up were 70 months from the beginning of the treatment. Age distribution was 26-66 years with an average of 50 years. Thirty-six patients were postmenopausal. Histological classification of tumor biopsies obtained before treatment included 63 infiltrating ductal carcinomas, 17 lobular carcinomas, and 2 mucosecreting carcinomas. Histological grading according to the classification of Scarff, Bloom and Richardson consisted of 6 grade 1, 32 grade 2, 24 grade 3 and 20 non-classified patients. Estrogen and progesteron receptors were not evaluable in most patients. Distribution according to the UICC 1979 classification [26] showed a majority of T 3 and T 4 tumors: 76/82 (92 ~). We have included nine patients with supraclavicular or subclavicular node involvement, which were staged a s N 3 (Table I) according to the UICC 1979 classification [26] because it was used during the period of this study. T 3 tumors were included only if there was clinical axillary lymph node involvement (Table I). The distribution according to tumor size showed that 57~o (47/82) of the patients had local tumors larger than or equal to 6 cm.
Patients and methods
Treatment plan Patients Eighty-two patients with non-inflammatory LABC without distant metastases were consecutively treated between April 1982 and December 1987. The staging procedure included the following: chest X-ray, serum
TABLE
All patients received primary chemotherapy including 4 cycles repeated every 28 days. The combination chemotherapy consisted of doxorubicin 45 mg/m 2 i.v. on day 1; vincristine 1.5 mg/m 2 i.v. on day 2; 5-fluorouracil 500 mg/m2 i.v. per day and cyclophosphamide 300 mg/
I
Clinical staging (TNM Clinical
classification; n
stage grouping
IIIA a
IIIB
a T3No patients
UICC
1979): 82 cases
[26].
TNM
Tumor
size (cm)
classification
42
40
1-2
2.1-3
3.1-5
3
3
.
T2N 2
2
-
-
2
-
-
T3N 1
34
-
-
-
31
3
T3N 2
3
-
-
-
2
1
T1N3 b
1
1
.
T3N 3
3
-
-
-
3
-
TaN o
19
-
-
7
8
4
TaN 1 TaN 2
14
-
-
-
8
6
2
-
-
-
1
1
T4N 3
1
.
were excluded.
b N3 r e f l e c t s t h e p r e s e n c e o f s u b c l a v i c u l a r
lymph
node involvement.
.
> 8
T1N 2
.
.
5.1-8
.
.
.
.
.
.
1
169 m 2 i.v. per day on days 2, 3 and 4. Complete blood counts were performed on day 1 of each cycle of chemotherapy. Three weeks after the last cycle of chemotherapy all patients received external irradiation using 6°Co (1.25 MV photons). The whole breast, chest wall and ipsilateral regional lymph nodes (supraclavicular, axillary and internal mammary nodes) were irradiated to a total dose of 45 Gy in 23 fractions over 31 days. The patient was positioned supine with the ipsilateral arm elevated at least to form a right angle to the torso. The irradiation of the whole breast and the whole thickness of the thoracic wall was accomplished with parallel opposed medial and lateral tangential beams. The dose was specified at the intersection of the beam axes in the central plane. The upper field border included the axillary tail of the breast and avoided the medial surface of the arm in abduction. Inferiorly, the fields extended to a level of 1-2 cm below the margin of the breast. In order to reduce the irradiated lung volume, the collimator was rotated and the beams were tilted 5-10 degrees from 180 ° in order to achieve a straight line of the borders of the field at the lung. Each field was treated every day. Wedges were used to improve dose homogeneity. Minimum and maximum dose in the central plane was greater than 95% and inferior to 110~o respectively. The supraclavicular and axillary areas were treated with a single anterior photon beam. Part of the shoulder joint and humerus was protected by blocks. Blocks were also used to protect spinal cord. The supraclavicular dose was calculated at a depth of 3 cm. The axillary dose from this anterior portal was calculated at the mid-plane of the A-P diameter. In order to supplement the dose to the midaxillary, a smaller posterior field was used. At mid-plane, the axillary dose was 1.50 Gy from anterior field and 0.50 Gy from posterior field per fraction. A separate internal mammary lymph node field was used. The internal mammary lymph nodes were irradiated with an anterior portal. Usually, a field extending 5 cm ipsilaterally from the midline and 1 cm controlaterally over the first five intercostal spaces covered this region with target depth at 3 cm. Four weeks after the end of irradiation, a fifth chemotherapy cycle was done with anthracyclin at a dose of 30 mg/m 2. Four weeks after the fifth cycle (29 weeks after the beginning of treatment), three different locoregional therapeutic approaches were proposed depending on the clinical, mammographic and sonographic response status after initial chemotherapy and radiotherapy, on the site of the tumor, on its multifocality, and on the volume of the breast compared to the residual
tumor. Decisions about local therapy were made 8 weeks after irradiation. Local treatment was planned as follows. (1) When residual tumor size was greater than 3 cm (largest diameter) after treatment, when it had a retroaerolar localization, or when initial tumor was plurifocal, Patey type mastectomy and axillary dissection was performed. The axillary dissection was limited to levels I and II lymph nodes. (2) When the residual tumor was less than or equal to 3 cm in its largest diameter, the therapeutic approach consisted in a conservative surgical excision with large tumorectomy and axillary dissection limited to levels I and II lymph nodes, followed by an interstitial implantation of the initial tumor site and the surrounding tissues, using Pierquin's technique [36]. A boost of 20 Gy was administered, calculated on the 85 ~o reference isodose [ 15,31,35]. Brachytherapy was done using wire implants of 192Ir sources with linear activity of 0.81.2 mCi/cm. The dose-rate on the reference isodose was between 0.55 and 0.80 Gy/h. (3) When the primary tumor and palpable node had disappeared, radiation therapy was completed by a booster dose on the initial tumor bed. It was administered either by external irradiation using 6°Co (1.25 MV) to a dose of 30 Gy in 15 fractions given in 19 days, or by interstitial brachytherapy to a dose of 25 Gy on the 85 To reference isodose volume. The purpose ofbrachytherapy was to reduce the dose delivered to normal mammary tissues and to try to obtain better cosmetic results, by decreasing the risk of cutaneous sequelae. A boost of 15 Gy (6°Co) was delivered by a direct axillary field in axillary lymph nodes in 8 fractions over 10 days, at 3 cm depth to the skin along the beam axis. Following completion of local therapy (3 weeks after the end of radiotherapy or after surgery), each patient received a sixth cycle of polychemotherapy. The dose of anthracyclin was limited to 30 mg/m ~. An adjuvant chemotherapy composed of vincristine 1.5 mg/m 2 i.v. on day 1; 5-fluorouracil 500 mg/m z i.v. and cyclophosphamide 300 mg/m z i.v. on days 2, 3 and 4 was prescribed (12 monthly cycles). After surgery, if there was an involvement of axillary lymph nodes, or if the diameter of the residual tumor was greater than 2 cm, the adjuvant treatment was reinforced by addition ofmethotrexate 24 mg/m2/day i.m. on days 2 and 3. All patients completed the entire program of chemotherapy.
Statistical analysis Proportions and means were compared by the chisquare test and the Student's t-test respectively. Overall
170 and disease-free survival rates from the beginning of treatment, were estimated according to the KaplanMeier method [ 28 ]. Confidence intervals (CI) were calculated from standard errors. The comparison of the overall and disease-free survival rates between the two groups were done using the Mantel method [30]. The analysis of the prognostic factors was done by univariate analyses and then by multiple regression analysis (Cox model) based on disease-free survival probability
[10]. Results
At the end of the fourth chemotherapy cycle Primary tumor Induction chemotherapy produced a significant clinical tumor regression (>/50 ~o in its largest diameter) in 55 ~o (45/82) of the cases. All evidence of tumor disappeared in 10~o (8/82) of the cases. Progressive disease was not observed in any of the 82 patients during the four initial courses of chemotherapy. None of the factors tested (tumor size, histological type, histological grade, menopausal status, and age) had a statistically significant influence on the tumor response following chemotherapy. Palpable nodes Of the 63 patients having a palpable lymph node, a complete clinical nodal remission was obtained for 73 ~o (46/63), and a partial remission for 11 ~o (7/63). Nodal response following chemotherapy was not significantly influenced by the N-stage. Following 45 Gy of external irradiation given in a period of 4.5 weeks Primary tumor Complete tumor regression was achieved in 42.5 % (35/ 82) of the cases, and partial regression in 34~o (28/82), while the other 19 tumors remained stable. None of the above mentioned tested parameters showed a statistically significant influence on the tumor response after radiotherapy. Palpable nodes Complete clinical nodal regression was observed in 87 ~o (55/63) of the cases, and a partial nodal regression in 8~o (5/63) of the patients. The N-stage did not significantly influence the nodal response rate after radiotherapy.
Following primary chemotherapy and radiotherapy of 45 Gy in 4.5 weeks A conservative locoregional treatment could be proposed to 50 patients (61 ~o). For 32 patients (39 ~o) who achieved complete remission, a booster irradiation dose was given by external radiotherapy to 14 patients (44 ~o) and by interstitial implantation to 18 patients (56~o). For the other 18 patients (22~o), who had a residual mass ~ 3 cm in diameter, a tumorectomy with axiUary dissection was done followed by a booster dose of radiotherapy in the tumorectomy site and surrounding tissues, using interstitial implantation. In 32 patients (39~o), mastectomy and axillary dissection were performed. Twenty-nine cases had a residual mass > 3 cm in diameter, whereas the other three achieved complete tumor regression before surgery. These patients had the following tumor types: 1 bifocal tumor, 1 retro-aerolar large tumor, and 1 tumor with large nodes classified a s N 3 (subclavicular node) without substantial nodal reduction. The possibility of breast conservation was significantly related to the initial size of the primary tumor. Out of 32 mastectomy patients, 25 (78~o) had a tumor >~6 cm in diameter, versus 22 (44~o) of 50 conservative treatments (chi-square -- 7.9, p ---0.005).
Analysis of the relationship between clinically assessed tumoral and nodular response, and histopathological findings on the 50 operated patients Primary tumor Before surgery (17 weeks after the fourth cycle of primary chemotherapy and 9 weeks after irradiation of 45 Gy in 4.5 weeks), 47 patients did not achieve complete clinical remission. The pathological assessment showed a residual tumor greater than 1 cm in diameter in 32 cases, a residual tumor ~<0.5 cm in diameter in 9 cases, and no evidence of tumor in 6 cases. For the other 3 cases, the pathological examination confirmed complete clinical remission before surgery. There was good correlation between the tumor response and the histopathological findings after surgery (R = - 0.47, p<0.001). Axillary dissection In 13 cases clinically classified as No, four had involved nodes at pathological examination. In 29 cases with palpable nodes before treatment, and considered as complete clinical remissions at the time of surgery, 13 showed nodal involvement; three patients out of 13 had 4 involved nodes or more. Among eight other cases with palpable nodes, who did not achieve complete
171 nodal remission, seven had tumoral involvement; four out of seven had more than four nodes infiltrated.
Survival and locoregional control Survival The 3- and 5-year overall survival rates calculated from the first day of treatment were 85.3 ~o (70 patients, 95 % CI: 79-91) and 81.3% (47 patients, 95~o CI: 73-90) respectively. The 3- and 5-year disease-free survival rates were 81.7% (67 patients, 95~o CI: 73-90) and 72% (44 patients, 95~o CI: 62-82) respectively (Fig. la). No patient died of intercurrent disease. In univariate analyses (Table II), 8 factors were found to have significant influence on survival: N-stage, initial tumor size, histological grade, tumor response after four initial cycles of chemotherapy, tumor response following chemotherapy and radiotherapy, and for the operated patients the size of residual tumor, axillary nodal status at pathological examination and number of involved axillary lymph nodes. Disease-free survival was influenced by 7 factors: N-stage, histological grade, tumor response after initial chemotherapy, tumor response following chemotherapy and radiotherapy, and for the operated patients the size of residual tumor, axillary nodal status and the presence of 4 or more positive axillary lymph nodes.
¢7~ 8 2
100 ~" 80
"
7
Overall survival Disease-freesurvival Locoregionalrelapse rate
~
~ ~
3
60 ~
2
59
'+~
38
m
40-
~w............. •
23 .,,.,......,......:
20 -
43.i?~ t
0
24
48 72 months
96
120
b
% 100-
RT alone (n=32) Tumorectomy+ RT (n=18) Mastectomy(n=32)
806040-
P > 0.05
20O'
"
0
"
24
-
i
48
•
i
72
•
i
96
-
i
120
months
Fig. 1. (a) Overall survival, disease-freesurvival (DFS) and locoregionat relapse rate in all patients (n = 82). (b)Locoregionalrelapse rate accordingto locoregionaltreatment.
In the multivariate analysis (Table III), two prognostic factors had a significant impact on overall survival: N-stage and initial tumor size. Disease-free survival was significantly influenced by the N-stage, initial tumor size and tumor response after initial chemotherapy.
Locoregional control Whatever the follow-up, following conservative treatment by tumorectomy and axillary dissection and radiotherapy (n = 18), 5 patients had a relapse (3 local, 2 distant metastases). Following radiotherapy alone (n--32), 11 patients relapsed (1 local, 3 distant metastases, 3 local with distant metastases, 1 axillary nodal with distant metastases and 3 local with axillary nodal recurrence and distant metastases). In the nonconservative group (n = 32), there were 12 relapses (10 distant metastases, 1 local with distant metastases and 1 local with axillary nodal recurrence and distant metastases). Five patients developed axillary relapses: they were always associated with distant metastases. None of the patients failed in the supraclavicular area. The 5-year actuarial local relapse rate without regional and distant failure was 6.5% (47 patients, CI: 1-13). We observed 4 isolated local tumor recurrences following conservative treatment (at 30, 30, 54 and 84 months after the beginning of treatment); 3 of them were salvaged after mastectomy with no evidence of disease at 6, 12 and 40 months following mastectomy. The fourth developed distant metastases 40 months after mastectomy. Actuarial locoregional relapse curves with or without distant failure according to the locoregional treatment modality are shown in Fig. lb. The 5-year actuarial overall locoregional relapse rate was 8.8~o (CI: 3-15); it was 12.6~o (CI: 1-24) following radiotherapy alone, 5.9~o (CI: 3-9) following tumorectomy and radiotherapy, and 6.3 % (CI: 0-14) following mastectomy. The 5-year locoregional relapse risk (with or without distant failure) was not influenced significantly by the locoregional treatment ( p > 0.05). In univariate analyses, the 5-year local relapse risk with and without regional or distant failure was influenced significantly by two prognostic factors: N-stage (p=0.01) and histological grade (p = 0.002), whereas the N-stage was the only significant prognostic factor in the multivariate analysis (p = 0.05). In our experience, the 5-year probability of breast conservation was 58.4% (34 patients, 95~o CI: 48-69). After non-conservative surgery, breast reconstruction with prosthesis associated myo-cutaneous flaps was performed in two cases at 4 and 5 years after the treatment without post-operative complication. These
172 TABLE II Influence of prognostic factors on the outcome-univariate analyses. Factor
Age Menopausal status Stage Tumor size (cm) N-stage
Histological type Histological grade (SBR) Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Residual tumor after surgery (cm) AxiUary nodal status N+
Group
< 50 /> 50 premenopausal postmenopausal IliA IIIB <6 >-6 No N1 N2 N3 ductal lobular 1+ 2 3 CR + PR NR CR + PR NR < 1 >- 1 NN+ 1-3 nodes >-4 nodes
n
35 47 45 37 42 40 35 47 19 48 10 5 63 17 38 24 53 29 63 19 18 32 26 24 9 15
Overall survival 5-year survival
Patients at risk
Yo
(n)
83 80 82 80 81 82 94 72 100 81 70 40 79 88 88 66 86 72 87 63 100 72 92 71 78 68
(22) (26) (25) (23) (25) (23) (24) (24) (15) (27) (7) (1) (33) (15) (19) (9) (35) (13) (42) (6) (14) (13) (15) (12) (6) (7)
DFS p-Value
0.53 0.59 0.69 0.01
0.01
0.58 0.002 0.04 0.005 0.047 0.03 0.09
5-year survival
Patients at risk
%
(n)
68 75 69 75 76 68 82 64 88 71 70 20 72 77 78 58 83 52 82 38 87 62 89 53 67 44
(20) (25) (22) (23) (25) (20) (23) (22) (14) (25) (7) (1) (31) (14) (17) (8) (34) (11) (40) (5) (13) (13) (15) (11) (5) (6)
p-Value
0.23 0.33 0.68 0.13
0.003
0.72 0.006 0.006 0.0004 0.02 0.004 0.03
DFS, disease-free survival; SBR, Scarff, Bloom and Richardson; CR, complete remission; PR, partial remission; NR, non-responder.
two patients are living with no evidence of disease at 2 years after reconstruction.
Cancer of the contro-lateral breast Cancer developed the contro-lateral breast in three cases (two patients after non-conservative treatment at 24 months of follow-up and one case after tumorectomy and radiotherapy at 5.5 years of follow-up). Out of these three patients with contro-lateral cancer, two had conservative treatment whereas the last one had radical surgery, These patients are living with no evidence of disease at 2, 2.5 and 3 years, respectively, after the treatment of contro-lateral breast.
Complications Tolerance to treatment was good. Although nausea, vomiting and alopecia were experienced by all patients,
a moderate hematologic toxicity was a constant finding following induction chemotherapy (grade 1 according to the W H O classification) [33]. On the other hand, following completion of local treatment, we observed three chronic leukopenias of grade 2 (~<3x 10 9 leukocytes/1), and six grade 1 paresthesias of the extremities following adjuvant chemotherapy requiring the discontinuation of vincristine. Locoregional complications were not infrequent. Arm lymphedema was observed in 20 ~o (10/50) of the patients treated with axillary dissection and in one patient out of 32 (3 To) treated without axillary dissection. Limitation of shoulder movements was noted in 22 ~o (7/32) of the patients treated with mastectomy and axillary dissection, but it was not observed following tumorectomy and axillary dissection, or radiotherapy without surgery. None of the patients developed congestive heart failure, extended pulmonary fibrosis, brachial plexopathy, or rib fracture.
173 TABLE III Influence of prognostic factors on the outcome-multivariate analysis. Factor
DFS
Overall survival fl-Coefficient
p-Value
/~-Coefficient
p-Value
All patients #7 = 82)
N-stage Tumor size Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Histological grade (SBR)
1.15 0.21 -0.76
0.0003 0.005 0.22
1.14 0.17 - 1.17
< 0.0001 0.01 0.02
0.003 - 0.17
0.99 0.45
0.04 - 0.04
0.90 0:79
1.51 0.07 1.79
0.02 0.67 0.27
1.11 0.02 - 1.59
0.005 0.82 0.08
- 0.92 0.19 - 0.09 - 0.004
0.51 0.31 0.64 0.92
- 0.63 0.17 0.13 0.05
0.42 0.25 0.23 0.87
Operated patients (n = 50)
N-stage Tumor size Tumor response after induction chemotherapy Tumor response after primary chemotherapy and radiotherapy Residual tumor Positive nodes Histological grade (SBR)
DFS, disease-free survival; SBR, Scarff. Bloom and Richardson.
Cosmetic results
Cosmetic results after conservative treatment at 5 years from the beginning of the treatment were assessed according to Harris et al. [21]. For the present study, 35 patients treated by conservative treatment were available. Following tumorectomy and radiotherapy, results were good in 61.5 ~o (8/13), fair in 23 ~o (3/13) and poor in 15.5 }'o (2/13). Following radiotherapy without surgery, results were good in 54.5~, (12/22), fair in 27~,,, (6/22) and poor in 18.5 (4/22) of the patients. Discussion
The present study concerns a series of 82 patients with non-inflammatory LABC, without evidence of distant metastases, treated by primary chemotherapy and radiotherapy with or without surgery and adjuvant chemotherapy after completion of local therapy. The overall tumor response rate after initial chemotherapy was 65~o and complete tumor response rate was 10}o. These results are comparable to other published reports using similar chemotherapy regimens, with complete remission rates between 5 and 18}'o [4,12,24,29]. Following aggressive multidisciplinary approach as applied in the present study, 5-year overall and disease-free survival rate was 81.3 }'o and 72}o respectively. There are studies suggesting possible improved 5-year overall and disease-free survival following a combined modality approach using initial cyto-reductive chemotherapy,
surgery, and/or radiotherapy in the treatment of LABC [8,12,17,25,27,41]. On the other hand, some authors consider that the influence of these aggressive therapeutic approaches in this category of patients is not proved [ 7,13,40,42 ]. Three randomized studies have shown that the addition of adjuvant chemotherapy in patients with stage III breast cancer receiving local radiation or local radiation and surgery, had no effect on the survival [13,38,401. In our experience, a multidisciplinary treatment approach with primary chemotherapy and radiotherapy in preoperative doses (45 Gy in 4.5 weeks) might permit the selection of some patients for conservative treatment. In our series, the 5-year locoregional relapse rate was low (8.8 }'o) with a 5-year breast conservation rate of 58.4}'o. In the experience of Hortobagyi [24], following neo-adjuvant chemotherapy, and radiotherapy or surgery, local recurrences were observed in 21 }'o of the patients. Jacquillat [27], following initial chemotherapy and radiotherapy, reported that the 5-year actuarial locoregional relapse rate was 18~, for T3, and 19}'o for T4 tumors. In series using locoregional treatment alone, the 5-year local relapse rate appears to be high, from 22 to 50~o [1,5,9,14,22,39]. In our multivariate analyses, the most significant and independent factors influencing disease-free survival were the N-stage, initial tumor size, and overall tumor response after initial chemotherapy. Five-year overall survival was 100}'o for No, 81.1%£, for N t, 70°0 for N2,
174 and 40~o for N 3 tumors (p = 0.01, univariate analysis). The N-stage influenced not only the survival rate but also the local control rate. Five-year overall survival was 93.8~o for tumors less than 6 c m in diameter, whereas it was 72.2~o for tumors whose diameter was 6 cm or more (p = 0.01, univariate analysis). In most studies, N-stage and initial tumor size are reliable prognostic factors [8,14,23,26,43,44]. Our results showed that tumor regression following primary chemotherapy was a significant prognostic factor. Five-year diseasefree survival was 82.5~o for complete and partial responders, whereas it was 52.1~o for non-responders (p = 0.006, univariate analysis). This prognostic factor is also reported in other series [8,24,27,34]. The analysis of the locoregional complications and the late sequalae after axillary dissection and the fair or poor cosmetic results seemed to be more frequent than series using radiotherapy alone or preoperative radiation therapy [3,14,39,43,44]. Some authors consider that adjuvant chemotherapy increases the late locoregional complications [ 11 ]. Hematological toxicity and cardiotoxicity of our treatment modality was relatively
moderate. Others, whose treatment modalities are similar, reported more severe hematological toxicity rates in 3-32~o [3,12,24], cardiotoxicity in 1.8-7.5~o [4,8,12,24], and pulmonary fibrosis in 6-9~o [4,12] of the cases. These complications were observed after a total irradiation dose greater than or equal to 70 Gy in 7 weeks to the whole breast, and after a cumulative doxorubicin dose less than or equal to 600 mg/m 2.
Conclusion A multidisciplinary treatment approach for locally advanced noninflammatory breast cancer using primary chemotherapy followed by radiotherapy in preoperative doses may permit the selection of some patients for conservative treatment, but locoregional complications after axillary dissection followed by external irradiation and fair or poor cosmetic results after conservative treatment are not infrequent. The impact of this treatment modality on long-term survival and local control remains to be proved.
References 1 Adair, F., Berg, J., Joubert, L. and Robbins, G . F . Long-term follow-up of breast cancer patients the 30-year report. Cancer 33: 1145-1150, 1974. 2 American Joint Committee for 'Cancer Staging and End Results Reporting Staging of Cancer of the Breast. In: Manual of Staging of Cancer, Whiting Press, Chicago, 1983. 3 Bedwineck, J. M., Rat, D. V., Perez, C. A., Lee, J. and Fineberg, B. Stage III and localized stage IV breast cancer: irradiation alone vs. irradiation plus surgery. Int. J. Radiat. Oncol. Biol. Phys. 8: 31-36, 1982. 4 Bedwineck, J. M., Ratkin, G. A., Philpott, G.W., Wallak, M. and Perez, C.A. Concurrent chemotherapy and radiotherapy for non-metastatic stage IV breast cancer. Am. J. Clin. Oncol. 6: 159-165, 1983. 5 Bruckman, J. E., Harris, J. R., Levene, M. B., Chaffey, J. T. and Hellman, S. Results of treating stage III carcinoma of the breast by primary radiation therapy. Cancer 43: 985-993, 1979. 6 CaUe, R., Pilleron, J. P , Schlienger, P. and Vilcoq, J.R. Conservative management of operable breast cancer. Cancer 42: 2045-2053, 1978. 7 Casper, E. S., Guidera, C. A., Bosi, G.J., Hakes, T. B., Kaufman, R. J., Shurgot, B. and Kinne, D.W. Combined modality treatment of locally advanced breast cancer: adjuvant combination chemotherapy with or without doxorubicin. Breast Cancer Res. Treat. 9: 39-44, 1987. 8 Chauvergne, J., Durand, M., Mauriac, L., David, M., Pigneux, J., Dilhuydy, M. H., Avril, A., Den6poux, R., Mar6e, D. and Bonichon, F. Traitement combin6 de 270 cancers mammaires localement &endus: r~sultats d'un programme th~rapeutique contr616. In: Neoadjuvant Chemotherapy, 169, pp. 225-230. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Ltd, Paris, 1988.
9 Chu, A. M., Cope, O., Doucette, J. and Curran, B. Non metastatic locally advanced cancer of the breast treated with radiation. Int. J. Radiat. Oncol. Biol. Phys. 10: 2299-2304, 1984. 10 Cox, D.R. Regression models and life table. J. Roy. Statist. Soc. 34: 187-220, 1972. 11 Danoff, B. F., Goodman, R. L., Glick, J. H., Hailer, D. G. and Pajak, T . F . The effect of adjuvant chemotherapy on cosmesis and complications in patients with breast cancer treated by definitive irradiation. Int. J. Radiat. Oncol. Biol. Phys. 9: 1625=_ 1630, 1983. 12 De Lena, M., Zucali, R., Viganotti, G., Valagussa, P. and Bonadonna, G. Combined chemotherapy-radiotherapy approach in locally advanced (T3b-T4) breast cancer. Cancer Chemother. Pharmacol. 1: 53-59, 1978. 13 Derman, D. P., Browde, S., Kessel, I. L., De Moor, N. G., Lange, M., Dansey, R., Seymour, L. and Bozwoda, W.R. Adjuvant chemotherapy (C.M.F.) for stage III breast cancer: a randomized trial. Int. J. Radiat. Oncol. Biol. Phys. 17: 257-261, 1989. 14 Dubois, J. B., Salomon, A., Gary-Bobo, J., Pourquier, H. and Pujol, H. Exclusive radical radiation therapy in breast carcinoma. Radiother. Oncol. 20: 24-29, 1991. 15 Dutreix, A., Marinello, G., Pierquin, B., Chassagne, D. and Moulard, J.P. Drveloppement actuel du syst~me de Paris. Deuxi~me partie: les dispositifs en triangles. J. Radiol. Electrol. 60: 319-325, 1979. 16 Goldie, J. H. and Coldman, A.J. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat. Rep. 63: 1727-1733, 1979. 17 Graham, M. V., Perez, C. A., Kuske, R. R., Garcia, D. M. and Fineberg, B. Locally advanced (non inflammatory) carcinoma of the breast: results and comparison of various treatment modalities. Int. J. Radiat. Oncol. Biol. Phys. 21: 311-318, 1991.
175 18 Haagensen, C. D. and Stout, A.P. Carcinoma of the breast II: criteria of operability. Ann. Surg. 118: 859-870, 1943. 19 Haagensen, C . D . Treatment of curable cancer of the breast. Int. J. Radiat. Oncol. Biol. Phys. 2: 975-980, 1977. 20 Haagensen, C.D. Clinicalclassification of the stage of advancement of breast carcinoma. In: Disease of Breast, pp. 851-863. Editor: C.D. Haagensen. W.B. Saunders, Philadelphia, PA, 1986. 21 Harris, J. R., Levene, M. B., Svensson, G. and Hellman, S. Analysis of cosmetic results following primary radiation therapy for stage I and II carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 5: 257-261, 1979. 22 Harris, J. R., Sawicka, J., Gelman, R. and HeUman, S. Management of locally advanced carcinoma of the breast by primary radiation therapy. Int. J. Radiat. Oneol. Biol. Phys. 9: 345-349, 1983. 23 Hery, M., Namer, M., Moro, M., Boublil, J. L. and Lalanne, C. Conservative treatment (chemotherapy/radiotherapy) of locally advanced breast cancer. Cancer 57: 1744-1749, 1986. 24 Hortobagyi, G. N., Blumenschein, G. R., Spanos, W., Montague, E. D., Buzdar, A. U., Yap, H. Y. and Schell, F. Multimodal treatment of locoregionally advanced breast cancer. Cancer 51: 763-768, 1983. 25 Hortobagyi, G . N . Comprehensive management of locally advanced breast cancer. Cancer 66: 1387-1391, 1990. 26 International Union Against Cancer (UICC). TNM Classification of Malignant Tumours, Fourth Edition. UICC, Geneva, 1979. 27 Jacquillat, C., Weill, M., Baillet, F., Borel, C., Auclerc, G., De Maublanc, M. A., Housset, M., Forget, G., Thill, L., Soubrane, C. and Khayat, D. Results of neo-adjuvant chemotherapy and radiation therapy in the breast-conserving treatment of 250 patients with all stages &infiltrative breast cancer. Cancer 66:119129, 1990. 28 Kaplan, E. L. and Meier, P. Non-parametric estimation from incomplete observations. J. Am. Statist. Assoc. 53: 457-481, 1958. 29 Lesnick, G., Perloff, M., Korzun, A., Chu, F., Holland, J. F., Thirwell, M., Ellison, R., Carey, R., Leone, L., Weinberg, V., Rice, M. and Wood, W. Neoadjuvant chemotherapy for stage Ill breast cancer: 5-year report of CALGB 7784. In: Neoadjuvant Chemotherapy, 169, pp. 185-188. Editors: C, Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Lid, Paris, 1988. 30 Mantel, N. and Haenszel, W. Chi-square tests with one degree of freedom: extensions of Mantel-Haenszel procedure. J. Am. Statist. Assoc. 58: 690-700, 1963. 31 Marinello, G., Dutreix, A., Pierquin, B. and Chassagne, D. D~veloppement actuel du syst6me de Paris. J. Radiol. Electrol. 59: 621-626, 1978. 32 Martijn, H., Caspers, R. J. L., Welvaart, K., Van De Velde, C., Hermans, J. and Leer, J. W, H. Locally advanced breast cancer: radiotherapy alone versus surgery and radiotherapy. Radiother. Oncol. 16: 23-29, 1989.
33 Miller, A.B., Hoogstraten, B., Staquet, M. and Winkler, A. Reporting results of cancer treatment. Cancer 47: 207-214, 1981. 34 Namer, M., H6ry, M., Abbes, M., Frenay, M, and Ramaioli, A. Influence of the response rate obtained by neoadjuvant chemotherapy on the survival of patients with locally advanced breast cancer. In: Neoadjuvant Chemotherapy, 169, pp. 177-181. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque. I.N.S.E.R.M., John Libbey Eurotect Ltd, Paris, 1988. 35 Pierquin, B., Dutreix, A., Paine, C. H., Chassagne, D., Marinello, G. and Ash, D. The Paris system in interstitial radiation therapy. Acta Radiologica 17: 33-48, 1978. 36 Pierquin, B., Owen, R., Maylin, C., Otmezquin, Y., Raynal, M., Mueller, W. and Hannoun, B. Radical radiation therapy of breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 6: 17-24, 1980. 37 Ragaz, J., Manji, M., Plenderleith, I. H., Knowling, M., Olivotto, L., Basco, V., Hardman, J., Worth, A. and Kuusk, V. Analysis of the British Columbia study of preoperative (neo-adjuvant) therapy of breast cancer in stage III disease. In: Neoadjuvant Chemotherapy, 169, pp. 249-256. Editors: C. Jacquillat, M. Weill and D. Khayat. Colloque I.N.S.E.R.M., John Libbey Eurotect Ltd. Paris, 1988. 38 Rubens, R. D., Bartelink, H., Engelsman, E., Hayward, J. L., Rotmensz, N., Sylvester, R., van der Shueren, E., Papadiamantis, J., Vassilaros, S. D., Wildier, J. and Winter, P.J. Locally advanced breast cancer: the contribution of cytotoxic and endocrine treatment to radiotherapy. An EORTC Breast Cancer Cooperative Group Trial (10792). Eur. J. Cancer Clin. Oncol. 25: 667-678, 1989. 39 Sheldon, T., Hayes, D.F., Cady, B., Parker, L., Osteen, R., Silver, B., Recht, A., Come, S., Henderson, C. and Harris, J.R. Primary radiation therapy for locally adwmced breast cancer. Cancer 60: 1219-1225, 1987. 40 Shaake-Koning, C., Van Der Linden, E.H., Hort, G. and Engelsman, E. Adjuvant chemo and hormonal therapy in locally advanced breast cancer: a randomized clinical study. Int. J. Oncol. Biol. Phys. 11: 1759-1763, 1985. 41 Sponzo, R. W., Cunningham, T. J. and Caradonna, R.R. Management ofnon-resectable (stage III) breast cancer. Int. J. Radiat. Oncol. Biol. Phys. 5: 1475-1478, 1979. 42 Strom, E.A., Mc Neese, M.D., Fletcher, G . H . , Ramsdahl, M. A., Montague, E. D. and Oswald, M.J. Results of mastectomy and post-operative irradiation in the management oflocoregionally advanced carcinoma of the breast. Int. J. Radiat. Oncol. Biol. Phys. 21: 319-323, 1991. 43 Vilcoq, J. R., Fourquet, A., Jullien, D., Gautier, C., Calle, R. and Ghossein, N.A. Prognosis significance of clinical model involvement in patients treated by radical radiotherapy for a locally advanced breast cancer. Am. J. Clin. Oncol. 6: 625-628, 1984. 44 Zucali, R., Uslenghi, C., Kenda, R. and Bonadonna, G. Natural history and survival of inoperable breast cancer treated with radiotherapy and radiotherapy followed by radical mastectomy. Cancer 37: 1422-1431, 1976.