Inflammatory breast cancer outcome with epirubicin-based induction and maintenance chemotherapy

Ten-Year Results from the French Adjuvant Study Group GETIS 02 Trial

Abstract

BACKGROUND.

The authors evaluated the long-term efficacy and side effects in patients with nonmetastatic, unilateral, inflammatory breast cancer (IBC) who received homogeneous treatment with intensive induction chemotherapy followed by a maintenance regimen.

METHODS.

One hundred twenty patients were randomized to receive high-dose fluorouracil, epirubicin, and cyclophosphamide (FEC-HD) (fluorouracil 750 mg/m² on Days 1 to 4, epirubicin 35 mg/m² on Days 2 to 4, and cyclophosphamide 400 mg/m² on Days 2 to 4 for 4 cycles every 21 days) with or without lenograstim. Locoregional treatment consisted of surgery and/or radiotherapy. Maintenance chemotherapy was FEC 75 (fluorouracil 500 mg/m², epirubicin 75 mg/m², and cyclophosphamide 500 mg/m² on Day 1 every 21 days for 4 cycles). No hormone treatment was allowed.

RESULTS.

The safety of the FEC-HD regimen was described previously. Among 102 patients who underwent surgery, a pathologic complete response (pCR) was achieved by 23.5% of patients with breast tumors and by 31.4% of patients with involved axillary lymph nodes. The overall pCR rate was 14.7%. One hundred nine patients received FEC 75. After a median 10 years of follow-up, the disease-free survival (DFS) and overall survival (OS) rates were 35.7% and 41.2%, respectively. The median DFS was 39 months (95% confidence interval [95% CI], 25–53 months), and the median survival was 61 months (95% CI, 43–79 months). Five patients developed a temporary decrease in left ventricular ejection fraction without congestive heart failure. In the lenograstim group, 1 patient developed acute myeloblastic leukemia M2, and 1 patient developed myelodysplastic syndrome.

CONCLUSIONS.

FEC-HD induction chemotherapy followed by FEC 75 maintenance regimen had moderate and acute long-term toxicities and lead to high DFS and OS rates in patients with IBC. Cancer 2006. © 2006 American Cancer Society.

Inflammatory breast cancer (IBC) represents from 1% to 6% of breast cancer patients.1 Epidemiologic data showed that this incidence increased from 0.3 to 0.7 cases per 100,000 person-years, which was much larger than the increase in non-IBC during the same period.2 The main issues for patients with IBC are the aggressiveness of the disease and the poor outcome in terms of disease-free survival (DFS) and overall survival (OS). With either surgery, or radiotherapy, or both, <15% of patients survive at 5 years.3 The introduction of a combined-modality approach that includes chemotherapy has led to a significant improvement in patient survival.4, 5 The University of Texas M. D. Anderson Cancer Center experience showed that the achievement of a pathologic response to induction chemotherapy was an excellent indicator of prolonged DFS and OS.5 Unfortunately, this is achieved in <20% of patients.6, 7 Thus, the use of more intensive induction chemotherapy regimens may yield to higher rates of pathologic complete response (pCR).

Anthracyclines are among the most active drugs in the treatment of breast cancer.8 Experience in the metastatic setting has showed that, at equimolar doses, epirubicin (E) was as effective and less hematotoxic and cardiotoxic than doxorubicin (A).9, 10 The activity of combined E and cyclophosphamide (C) has been reported previously in patients with IBC.11 Finally, high response rates have been described with continuous fluorouracil (F) infusion in patients with pretreated metastatic breast cancer.12

Based on these results, we previously conducted a pilot study of intensive chemotherapy using a high-dose FEC regimen (FEC-HD), which demonstrated high clinical activity and a good rate of pCR.13 However, this efficacy was at the expense of significant toxicity in terms of febrile neutropenia. The primary objective of the current study was to demonstrate the use of lenograstim support to decrease the hematologic toxicity of the FEC-HD regimen. These results, including the preoperative chemotherapy regimens, were published previously by Chevallier et al. in 1995 and confirmed the effectiveness of lenograstim in significantly reducing the morbidity associated with FEC-HD.14 Here, we report the long-term efficacy results from an investigation of the role of maintenance chemotherapy with a conventional-dose FEC regimen after FEC-HD induction chemotherapy and the incidence of long-term side effects.

MATERIALS AND METHODS

Study Population

Between February 1990 and September 1992, 120 women with IBC who were recruited from 9 institutions in France were enrolled to the study. Inclusion and exclusion criteria have been described previously.14 The protocol was reviewed and approved by the Ethics Committee/Institutional Review Board, and a written informed consent was obtained before patients were included.

Treatment Protocol and Assessment

All patients received induction chemotherapy as follows: continuous infusion of F (Roche SA, Basel, Switzerland) 750 mg/m² on Days 1 to 4, bolus infusion of E (Pfizer, Milan, Italy) 35 mg/m² on Days 2 to 4, and bolus infusion of C (Asta Medica AG, Frankfurt, Germany) 400 mg/m² on Days 2 to 4 intravenously every 21 days for 4 cycles (FEC-HD). At study entry, before the initiation of induction chemotherapy, patients were randomized to receive either lenograstim (recombinant human granulocyte-colony stimulating factor [G-CSF]; Chugai, Antony, France) 5 μg/kg daily on Days 6 to 15 or placebo in a double-blind manner. Retreatment at Day 21 was allowed if the patient had a neutrophil count >1 × 10⁹/L, a platelet count >75 × 10⁹/L, and residual nonhematologic toxicities Grade <2 according to World Health Organization (WHO) criteria. Treatment was postponed by 1 week in other patients. A prophylaxis with antibiotics was prohibited. Antiemetic treatment was prescribed routinely before each cycle. Clinical response rates were estimated after the second cycle and at the end of induction chemotherapy, taking into account breast tumor, axillary lymph nodes, and inflammatory signs.

Locoregional treatment consisted of surgery and/or radiotherapy (with the choice left to the discretion of the physician). Surgery was either modified radical mastectomy or quadrantectomy in patients who had a good response to induction chemotherapy with an adequate breast volume. All patients who underwent surgery also underwent homolateral axillary clearance. Locoregional radiotherapy consisted of 60 grays (Gy) delivered to the chest wall and axillary area, 50 Gy delivered to the internal mammary chain, and 46 Gy delivered to the supraclavicular area. From 4 to 6 weeks after the first irradiation, a 20-Gy boost was delivered to the breast in patients who underwent breast-conserving surgery. Patients who underwent surgery commenced radiotherapy within 30 days after the last maintenance chemotherapy cycle. Pathologic response was assessed only for patients who underwent surgery.13

Maintenance chemotherapy was administered only if no breast cancer progression occurred, no remaining toxicity of induction chemotherapy was registered, and no cardiac dysfunction disallowed the reintroduction of E. The maintenance chemotherapy used the FEC 75 combination (F 500 mg/m², E 75 mg/m², and C 500 mg/m² intravenously on Day 1 every 21 days for 4 cycles), and no hormone treatment was allowed. A prophylaxis with G-CSF and antibiotics was prohibited. An absolute granulocyte count <2 × 10⁹/L or a platelet count <100 × 10⁹/L on Day 21 led to a treatment interruption of at least 1 week. Treatment was stopped if hematologic recovery took more than 3 weeks. The E dose was reduced by 50% if serum bilirubin levels were from 35 μmol/L to 50 μmol/L, and treatment was stopped if bilirubin levels exceeded 50 μmol/L. The tolerability of chemotherapy was evaluated before each cycle: an electro cardiogram and an absolute blood count were performed on Day 21, and nonhematologic toxicity was evaluated during the period between each cycle according to WHO criteria. It was recommended to assess the left ventricular ejection fraction (LVEF) within 3 to 4 weeks after the last chemotherapy cycle.

Patients underwent clinical and biochemical assessments every 4 months during the 2 years of follow-up, every 6 months for the next 3 years, and yearly thereafter with clinical examinations, chest X-rays, and biochemical screenings. Liver ultrasound, bone scan, and mammography studies were performed once a year.

Statistical Analysis

All assessable patients were entered into an intention-to-treat analysis using SPSS software (SPSS, Inc., Chicago, IL). The study data base was been reviewed and, thus, explained some changes in the values compared with the initial publication.14 The chi-square test was used to compare baseline categorical variables and the incidence of adverse events in both groups.15 Continuous variables were compared by using an analysis of variance.16 DFS was defined as the time from random allocation until first recurrence (local, regional, or distant). Contralateral breast cancer was considered a new primary malignancy. OS was defined as the time from random assignment until death from any cause. DFS and OS rates were computed according to the Kaplan–Meier method, and survival curves were compared by using the log-rank test or the stratified log-rank test, as appropriate.17, 18 The prevalence of the following prognostic factors was analyzed: lenograstim, menopausal status, surgery, baseline inflammatory signs, and pathologic response in breast and lymph nodes. The Cox regression method was used to determine the respective role of each factor in disease recurrence.19

RESULTS

Patient and Treatment Characteristics

One hundred twenty patients were enrolled on the study, including 59 patients who were assigned randomly to the placebo group and 61 patients who were assigned to the lenograstim group. All patients were eligible for and were entered on the intention-to-treat analysis. Baseline characteristics were described previously and were balanced well between both arms.14 The therapeutic flow chart is provided in Figure 1. During induction chemotherapy, 5 patients stopped treatment prematurely because of disease progression (n = 3 patients), cutaneous toxicity (n = 1 patient), or protocol violation (n = 1 patient). Six patients did not receive maintenance chemotherapy because of disease progression (n = 1 patient), persistent neutropenia (n = 2 patients), LVEF decrease (n = 1 patient), patient refusal (n = 1 patient), or protocol violation (n = 1 patient). Finally, 1 patient did not receive radiotherapy after maintenance chemotherapy because of a local recurrence.

Figure 1. This flow chart illustrates the therapeutic sequences that were used in the current study.

Treatment characteristics are described in Table 1. There were no significant differences according to G-CSF induction chemotherapy support. All patients received at least 2 cycles of FEC-HD, and 7 patients did not receive 4 cycles (because of disease progression in 1 patient, cutaneous toxicity in 1 patient, general status impairment in 1 patient, patient refusal in 1 patient, and unknown reasons in 3 patients). The mean cumulative E dose was similar between both groups (413 mg/m² vs. 402 mg/m²; P = .18). At the end of induction chemotherapy, 2 patients presented with persistent Grade 3 neutropenia (1 per treatment group), and patients who received G-CSF support had significantly more residual Grade 1 to Grade 3 anemia (68.8% vs. 49.1%; P = .0015).

Table 1. Treatment Characteristics

Treatment	No. of patients (%)
	Lenograstim (N = 61)	Placebo (N = 59)	Overall (N = 120)
Induction chemotherapy	61 (100)	59 (100)	120 (100)
4 Cycles delivered	59 (96.7)	54 (91.5)	113 (94.2)
Mean epirubicin dose (mg/m^2)	413	402	407
Maintenance chemotherapy	54 (88.5)	55 (93.2)	109 (90.8)
4 Cycles delivered	51 (83.6)	51 (86.4)	102 (85)
Mean epirubicin dose (mg/m^2)	284	283	283
Overall chemotherapy	61 (100)	59 (100)	120 (100)
8 Cycles delivered	51 (83.6)	47 (79.7)	98 (81.7)
Mean epirubicin dose (mg/m^2)	664	665	665
Surgery	50 (82)	52 (88.1)	102 (85)
Quadrantectomy	12 (19.7)	15 (25.4)	27 (22.5)
Modified mastectomy	38 (62.3)	37 (62.7)	75 (62.5)
Radiotherapy	57 (93.4)	57 (96.6)	114 (95)
Chest wall	55 (90.2)	57 (96.6)	112 (93.3)
Axillary area	41 (67.2)	36 (61)	77 (64.2)
Supraclavicular area	55 (90.2)	52 (88.1)	107 (89.2)
Internal mammary chain	50 (82)	47 (79.7)	97 (80.8)
Boost	16 (26.2)	15 (25.4)	31 (25.8)

Overall, 102 patients underwent surgery, which consisted of quadrantectomy in 27 patients (26.5%), and 114 patients received radiotherapy. Thirteen patients received radiotherapy alone as locoregional treatment, including 6 patients who underwent breast biopsy and/or axillary dissection.

The FEC 75 maintenance chemotherapy regimen was delivered to 109 patients (90.8%), including 83 patients before radiotherapy and 26 patients after radiotherapy. Four cycles were delivered to 102 patients (93.6%), and 7 patients stopped FEC 75 before Cycle 4 for Grade 3 neutropenia (n = 1 patient), Grade 3 infection (n = 1 patient), LVEF decrease (n = 3 patients), patients refusal (n = 1 patient), and an unknown reason (n = 1 patient). Toxicities related to maintenance chemotherapy were those observed classically with FEC 75 regimen and did not differ according to G-CSF induction chemotherapy support, except for Grade 3 or 4 neutropenia, which was significantly more frequent in the lenograstim group (40.7% vs. 23.6%; P = .05) (Table 2). Ninety-eight patients (81.7%) received the overall induction and maintenance chemotherapies (mean cumulative E dose received, 665 mg/m²).

Table 2. Acute Toxicity of Maintenance Chemotherapy (Combined Fluorouracil 500 mg/m², Epirubicin 75 mg/m², and Cyclophosphamide 500 mg/m² on Day 1 Every 21 Days for 4 Cycles) According to Granulocyte-Colony Stimulating Factor Support during Induction Chemotherapy

Toxicity	No. of patients (%)
	Lenograstim (N = 54)	Placebo (N = 55)	Overall (N = 109)
Neutropenia
Grade 1–2	17 (31.5)	22 (40)	39 (35.8)
Grade 3–4	22 (40.7)	13 (23.6)	35 (32.1)
Infection
Grade 1–2	13 (24.1)	9 (16.4)	22 (20.2)
Grade 3	2 (3.7)	1 (1.8)	3 (2.7)
Anemia
Grade 1–2	25 (46.3)	25 (45.4)	50 (45.9)
Nausea/emesis
Grade 1–2	40 (74.1)	32 (58.2)	72 (66.1)
Grade 3–4	9 (16.7)	8 (14.5)	17 (15.6)
Stomatitis
Grade 1–2	14 (25.9)	16 (29.1)	30 (27.5)
Alopecia
Grade 1–2	7 (13)	7 (12.7)	14 (12.8)
Grade 3	36 (66.7)	33 (60)	69 (63.3)

Five patients (4.2%) presented with LVEF decrease, including 2 patients who had undetermined baseline values before the initiation of chemotherapy (Table 3). In 3 patients, the LVEF decrease occurred after induction chemotherapy; and, in 1 patient, FEC 75 maintenance chemotherapy was not delivered. In the other 2 patients, the LVEF decrease occurred during maintenance chemotherapy. None of these patients developed congestive heart failure (CHF), and no cardiac symptoms were diagnosed during follow-up.

Table 3. Summary of Cardiac Events and Left Ventricular Ejection Fraction Decreases

Patient no.	Baseline LVEF (%)	Induction chemotherapy			Maintenance chemotherapy			Outcome
		No. of cycles	E cumulative dose (mg/m^2)	LVEF (%)	No. of cycles	E Cumulative dose (mg/m^2)	LVEF (%)
LVEF indicates left ventricular ejection fraction; E, epirubicin; ND, not done; NA: not applicable.
307	60	4	420	44	ND	NA	NA	No further signs
310	55	4	350	48	3	571	43	No further signs
1001	ND	4	420	40	4	683	44	LVEF = 54%
1101	ND	4	420	60	3	634	53	No further signs
1106	57	4	420	56	2	574	47	No further signs

Tumor Response and Characteristics

After induction chemotherapy, there was an overall clinical response rate of 91.1%, as reported previously.14 In the intent-to-treat population, 24 of 120 patients (20%) achieved a pCR of the breast. In patients who underwent quadrantectomy or modified mastectomy, a pCR (i.e., Grade 1 and Grade 2 response) was achieved in 23.5% of patients with breast tumors and in 31.4% of patients with involved axillary lymph nodes (Table 4). The overall pCR rate (breast and axillary lymph nodes) was 14.7%. Among 13 patients who received radiotherapy alone as locoregional treatment, 3 patients underwent a breast biopsy, which indicated 1 Grade 1 response and 2 Grade 4 responses; 2 patients underwent axillary dissection with Grade 4 responses in both patients; and 1 patient underwent a breast biopsy plus axillary dissection, which showed a Grade 4 response in the breast and a Grade 1 response in axillary lymph nodes.

Table 4. Tumor Characteristics and Pathologic Response in 102 Patients who Underwent Surgery

Variables	Surgical patients (N = 102)
	No. of patients	%
* Pathologic response was classified as follows: Grade 1, the disappearance of all tumor on both macroscopic and microscopic examination; Grade 2, the presence of in situ carcinoma of the breast, no invasive tumor, and no tumor found in lymph nodes; Grade 3, the presence of invasive carcinoma with stromal alterations, such as sclerosis or fibrosis; Grade 4, no or few alterations in tumor appearance.
Pathologic tumor size
0 mm	13	12.7
≤20 mm	21	20.6
>20 mm	34	33.3
Not available	34	33.3
Breast response*
Grade 1	13	12.7
Grade 2	11	10.8
Grade 3	37	36.3
Grade 4	40	39.2
Not assessable	1	1.0
No. of axillary lymph nodes involved
0	31	30.4
1 to 3	27	26.5
>3	31	30.4
Not available	13	12.7
Lymph node response*
Grade 1	31	30.4
Grade 2	1	1.0
Grade 3	16	15.7
Grade 4	41	40.2
Not assessable	13	12.7
Overall response*
Grade 1	8	7.8
Grade 2	7	6.9
Grade 3	23	22.5
Grade 4	58	56.9
Not assessable	6	5.9

DFS and OS

The median follow-up was 120 months (range, 30 to 140 months). Because no difference was observed between the lenograstim group and the placebo group, efficacy results are presented for the whole population (Table 4). Overall, 76 patients (63.3%) developed recurrent disease. The 10-year DFS rate was 35.7%, and the median time to recurrence was 39 months (95% confidence interval [95% CI], 25–53 months). Of 76 patients who developed recurrent disease, 53 patients (69.7%) presented with 1 involved site. The pattern of recurrence was as follows: local recurrence (26.3%), soft tissues (9.2%), lymph nodes (9.2%), bone (36.8%), lung (17.1%), liver (26.3%), and other sites (17.1%). Fifteen of 45 patients with ipsilateral breast disease (33.3%) who did not undergo mastectomy developed recurrent disease, including 10 of 18 patients (55.6%) who did not undergo surgery, and 5 of 27 patients (18.5%) who underwent breast-conserving surgery. Among the patients who underwent modified mastectomy, 5 patients developed local chest wall recurrences (6.7%). Sixty-eight of 76 patients (89.5%) developed recurrent disease during the first 4 years after diagnosis. In the patients who were treated according to the protocol (n = 108 patients), the recurrence rate was similar between those who received maintenance chemotherapy before radiotherapy or after radiotherapy (59.8% vs. 53.8%, respectively); however, this rate increased dramatically in patients who received only radiotherapy as locoregional treatment (84.6%) (Fig. 2). The univariate analysis for prognostic factors showed that no surgery, diffuse inflammatory signs, or no pCRs were correlated significantly with disease recurrence (Table 5). When a Cox proportional hazards model was performed, there was no significant prognostic factor for recurrence (Table 5).

Figure 2. This graph illustrates the 10-year disease-free survival curves according to therapeutic sequences: Prob indicates probability; S-C-R, surgery-chemotherapy-radiotherapy (40.3% of patients); S-R-C, surgery-radiotherapy-chemotherapy (42.3% of patients); and R-C, radiotherapy-chemotherapy (11.5% of patients).

Table 5. Prognostic Factors of Recurrence

Factor	Univariate analysis		Multivariate analysis
	HR (95% CI)	P	HR (95% CI)	P
HR indicates hazard ratio; 95%CI, 95% confidence interval; G-CSF, granulocyte-colony stimulating factor; NS, not significant; pCR, pathologic complete response.
G-CSF (no vs. yes)	0.92 (0.47–1.37)	.70		NS
Menopausal status (premenopausal vs. postmenopausal)	1.01 (0.55–1.47)	.98		NS
Surgery (no vs. yes)	2.02 (1.45–2.59)	.01	1.40 (0.00–3.48)	.75
Quadrantectomy vs. modified mastectomy	0.58 (0.00–1.20)	.08	1.20 (0.36–2.04)	.67
Inflammatory signs (limited vs. diffuse)	0.48 (0.00–0.97)	.003	0.73 (0.12–1.34)	.31
Breast pCR (no vs. yes)	2.36 (1.68–3.04)	.01	1.79 (0.32–3.26)	.44
Lymph nodes pCR (no vs. yes)	2.52 (1.91–3.13)	.003	1.93 (1.09–2.77)	.12
Overall pCR (no vs. yes)	3.03 (2.12–3.94)	.02	1.04 (0.00–2.89)	.97

Seventy patients died (58.3%), and the 10-year OS rate was 41.2% (Fig. 3). The median survival was 61 months (95% CI, 43–79 months). Five deaths were not related to breast cancer progression, including 1 death in a patient with M2 acute myeloblastic leukemia (AML) according to French–American–British classification, 1 death from myelodysplastic syndrome (MDS), 1 death from septic shock after recurrence, 1 death from lung cancer, and 1 death from hemorrhage caused by a gastroduodenal ulcer.

Figure 3. This graph illustrates the 10-year overall survival curve for patients in the current study. Prob indicates probability; S-C-R, surgery-chemotherapy-radiotherapy; S-R-C, surgery-radiotherapy-chemotherapy; and R-C, radiotherapy-chemotherapy.

Second Malignancies

Six patients (5.0%) developed contralateral breast cancer. Second malignancies occurred in 6 patients (5.0%). Thirty-four months after the end of maintenance chemotherapy, 1 patient in the lenograstim group developed an M2 AML after receiving a cumulative E dose of 680 mg/m² and a cumulative C dose of 6.5 g/m² and died of her AML 28 months after it was diagnosed. One patient in the lenograstim group developed MDS (del 5q) 65 months after maintenance chemotherapy after receiving a cumulative E dose of 650 mg/m² and a cumulative C dose of 5.8 g/m² and died of her MDS 6 months after it was diagnosed. The overall incidence of treatment-related AML/MDS was 3.3% (2 of 60 in patients) for patients who received concurrent lenograstim support. The remaining 4 second malignancies were lung cancer in 2 patients, anal duct squamous cell carcinoma in 1 patient, and spinocellular carcinoma in 1 patient.

DISCUSSION

The current trial involved the recruitment over 2 years of a large number of patients with IBC who were treated homogeneously. It is noteworthy that, when the study was initiated in 1990, we did not use adjuvant tamoxifen, and the omission of endocrine therapy probably had a negative impact on hormone receptor-positive patients. After 10 years of follow-up, we obtained consolidated results regarding survival data. With a DFS rate of 35.7% and an OS rate of 41.2%, FEC-HD induction chemotherapy followed by conventional FEC maintenance therapy yielded 1 of the higher long-term survival rates described in the literature (Table 6).5, 7, 20–30 However, the comparison of clinical trials is difficult, because treatment sequencing, surgical procedures, cytotoxic agents used, duration of treatment, use of maintenance chemotherapy, and hormone treatments are very heterogeneous.

Table 6. Summary of Survival Results from Chemotherapy Trials in Patients with Inflammatory Breast Cancer

Reference	Induction CT	Maintenance CT	Tamoxifen	No. of patients	Median survival months	Follow-up months	Survival rate
CT indicates chemotherapy; CMF, cyclophosphamide, methotrexate, and fluorouracil; AVCF, doxorubicin, vincristine, cyclophosphamide, and fluorouracil; FAC, fluorouracil, doxorubicin, and cyclophosphamide; VAC, vincristine, doxorubicin, and cyclophosphamide; NA, not applicable; GETIS, Groupe d´Etude des Tumeurs Inflammatoires su Sein; FEC, fluorouracil, epirubicin, and cyclophosphamide; HD, high dose; FEC 75, fluorouracil 500 mg/m^2, epirubicin 75 mg/m^2, and cyclophosphamide 500 mg/m^2 on Day 1 every 21 days for 4 cycles; VCF, vincristine, cyclophosphamide, and fluorouracil; NR, not reported; CAP, cyclophosphamide, doxorubicin, and cisplatin; M, methotrexate; EC, epirubicin and cyclophosphamide; Vds, vindesine; A, anthracycline; CCb, cyclophosphamide and carboplatin.
Anthracycline vs. CMF
Chevallier et al., 19934	CMF		No	64	25
	AVCF/FAC/VAC	No	No	83	45.7	NA	NA
	FEC-HD		Yes	31	32.6
Bauer et al., 199521	CMF			28	18
	FAC	No	No	10	30	NA	NA
Anthracycline-based chemotherapy
GETIS 02	FEC-HD × 4	FEC 75 × 4	No	120	61	10	41.2
Attia-Sobol et al., 199322	AVCF × 6	AVCF × 6 then VCF × 6	No	109	70	10	NR
Palangie et al., 199423	AVCF	No	No	223	NR	10	32
Colozza et al., 199624	CAP × 6	CMF in responders	No	31	48.7	6	29
Maloisel et al., 19907	FAC × 3	FAC × 6 in responders	No	43	NR	5	75
Thomas et al., 199525	AVCF-M × 8	No	No	125	NR	5	50
Tubiana-Mathieu et al., 200126	EC-Vds	FEC + M-HD × 6	Yes	48	Not reached	3	71
Brun et al., 198827	AVCF × 2	AVCF × 4	No	26	31	NA	NA
High-dose chemotherapy plus stem cell support
Bertucci et al., 200428	PolyCT with A
	Conventional	No	Yes	20	NR	5	18
	HD			54			50
Arun et al., 199929	CCb	Immunotherapy	No	24	NR	2	73

The optimal sequencing of combined modalities has not been determined to date. In the current study, postoperative radiotherapy delivered before or after maintenance chemotherapy did not appear to modify disease outcome. In the study by Harris et al, approximately 66% of patients received preoperative irradiation, and no apparent difference in survival was observed between those patients and patients who received postoperative radiotherapy.20

Currently, radiotherapy is delivered systematically, and surgical procedures are more varied and discussed. In the current study, 84.6% of nonsurgical patients developed recurrent disease. The absence of surgery seems to be deleterious to disease outcome, regardless of response to induction chemotherapy. Except in 1 trial that demonstrated no advantage of mastectomy in responsive or stable disease,31 the addition of mastectomy to chemoradiation therapy improved local control and DFS.20, 32, 33 Moreover, surgery allows precise staging and may permit better prognostic stratification. If surgery is to be an important component in improving survival, then does it mean that mastectomy should be the only surgical approach for patients with IBC? Greater than 25% of our patients had undergone breast-conserving surgery, resulting in 5 local recurrences (18.5%), which was not significantly different from patients who underwent mastectomy.

Currently, induction chemotherapy is considered the standard of care in patients with IBC. Numerous analyses have demonstrated that the clinical response to neoadjuvant chemotherapy is predictive of survival.5, 7, 20, 21, 23, 34, 35 Moreover, in a study from The University of Texas M. D. Anderson Cancer Center that involved patients with IBC and with locally advanced breast cancer, the results indicated that patients who had mastectomy specimens with no macroscopic residual disease had a significantly higher 5-year survival rate compared with patients who had a less marked response to therapy (P<.01).6 In our current study, >90% of patients achieved an objective clinical response, and 23.5% of those patients had a pCR of breast tumor. Although 2 studies did not demonstrate the superiority of anthracycline-based induction chemotherapy over methotrexate-based chemotherapy,20, 36 the use of an anthracycline improved survival, and anthracyclines should be considered pivotal drugs for neoadjuvant therapy.4, 21, 37 High-dose, E-based induction chemotherapy without stem cell support has been investigated in other trials, and a clinical response was achieved in 93.5% and 93.6% of patients, for a 3-year OS rate of 71%.4, 26 When it is used at conventional doses, A-based induction chemotherapy lead to a lower clinical response rate.4, 22, 27, 34 More recently, high-dose chemotherapy with bone marrow or peripheral stem cell transplantation has been used.28, 29, 38, 39 The encouraging results obtained need to be confirmed in randomized trials.

Will the introduction of new drugs, such as taxanes, improve the outcome of patients with IBC? The answer remains unclear. One trial that tested the role of paclitaxel after F, A, and C (FAC) demonstrated a survival improvement in estrogen receptor-negative patients with IBC compared with FAC alone.40 In patients with locally advanced breast cancer, the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-27 trial demonstrated an improvement in clinical and pathologic response when 4 cycles of docetaxel were added to 4 cycles of AC as induction chemotherapy.41 However, in those trials, the question was not resolved: Was the improvement related to the contribution of taxanes, to the duration of chemotherapy, or to both? Finally, what is the role of maintenance chemotherapy? After local therapy, investigators at The University of Texas M. D. Anderson Cancer Center recommend the use of adjuvant chemotherapy, because the risk of disease recurrence remains high.37 We used the same drugs at a lower dosage as maintenance treatment, but it may prove be interesting to change to a noncross-resistant chemotherapy regimen. In patients with locally advanced breast cancer, the Aberdeen trial showed that patients who responded to combined C, A, vincristine, and prednisone (CVAP) who received 4 cycles of docetaxel instead of 4 additional cycles of CVAP had better clinical and pathologic responses.42

Because few series have reported on long-term outcomes for patients with IBC, long-term side effects were not described well. In our trial, the use of a high cumulative E dose could lead to long-term toxicities, such as CHF and secondary leukemia. Although 5 patients presented with LVEF decrease, none of those patients developed CHF, and no cardiac symptoms were diagnosed during follow-up. This confirmed that, below a cumulative E dose of 900 mg/m², the risk of developing CHF remains very low. Conversely, we identified 1 patient with M2 AML and 1 patient with MDS in the lenograstim group. In those patients, the respective roles of high cumulative doses E and C and of G-CSF were difficult to distinguish. Secondary, M2 AML and preleukemic phase are related more frequently to alkylating agents.43 If the correlation between secondary leukemia and C dose is established, then the correlation remains unclear with anthracyclines. Moreover, in the high-dose regimens with A or E, confusing factors, such as the use of G-CSF, must be take into consideration.44, 45 Considering the use of G-CSF, we highlighted a surprising issue in terms of residual anemia at the end of induction chemotherapy and an increased risk of neutropenia during maintenance chemotherapy in patients who had received primarily lenograstim. To date, there is no clear explanation for this observation; however, it is reasonable to hypothesize that an impoverishment in medullary reserve related to G-CSF stimulation may be involved in this decline of mature blood cells.

In our patients, a multivariate analysis did not demonstrate any classic prognostic factor of recurrence. Unlike previous findings, the achievement of pCR after induction chemotherapy was not associated with an improvement in DFS or OS. It is possible that the addition of 4-cycle maintenance chemotherapy acted as salvage treatment in patients who did not achieve a pCR after 4 cycles of induction therapy. Eight cycles of induction chemotherapy, as demonstrated in the NSABP B-27 trial, resulted in a higher rate of pCR than 4 cycles in patients with locally advanced breast cancer.41 Therefore, it can be argued that a longer duration of either induction chemotherapy or maintenance therapy delivered to patients who do not achieve a pCR after 4 cycles of induction therapy may be therapeutic options.

Clearly, further advances in the understanding and treatment of IBC are needed. IBC often is characterized by a high pathologic grade and the presence of molecular markers of aggressive disease. Recent works on genetic determinants identified genes that are involved in the development and progression of the disease, leading the way toward a development of new therapies. Among prognostic markers, is has been observed that p53 is overexpressed in 70% of IBC tumors.46 Its role in IBC seems to differ from common gene mutations, and it is believed that its regulation by Mdm2 and PTEN are lost.47 In addition, the use of therapy that acts directly on the p53 protein may be of interest. Another interesting target is Rho C GTPase, which is involved in the angiogenesis process and is overexpressed in 90% of IBC tumors.48 For instance, farnesyl transferase inhibitors (FTIs) reverse the invasive phenotype of Rho C GTPase-overexpressing cell lines and may be a good candidate for managing IBC.49

In the current trial, the FEC-HD induction chemotherapy followed by conventional FEC maintenance therapy in 120 patients who were treated homogeneously provided a solid basis for further trials in patients with IBC. The schedule resulted in a high rate of long-term survival without pronounced, long-term side effects. In the near future, the addition of targeted therapies, such as p53 modulators and FTIs, may improve the outcome for patients with this particularly aggressive disease.

Acknowledgements

We sincerely thank our friend Bernard Chevallier for all of his work and for having initiated this study. Isabelle Chapelle-Marcillac provided editorial assistance in the preparation of the article.