|Home | About | Journals | Submit | Contact Us | Français|
Inflammatory Breast Cancer (IBC) is a rare clinico-pathological entity with a poor prognosis, lagging far behind any other form of non metastatic breast cancer. Since the advent of systemic chemotherapy over 35 years ago, only minimal progress has been made in long term outcome. Although multiple randomized trials of high-dose chemotherapy and autologous progenitor cell transplantation (ASCT) for the treatment of breast cancer have yielded disappointing results, these data are not necessarily relevant to IBC, a distinct clinical and pathological entity. Therefore, the optimal multimodality therapy for IBC is not well established and remains unsatisfactory. We treated 21 women with non metastatic IBC with a multi-modality strategy including high-dose melphalan / etoposide and ASCT. The treatment was overall tolerated with acceptable morbidity and no post ASCT 100-day mortality. With a median potential follow-up of approximately 8 years, the estimated PFS, EFS and OS at 6 years from on-study date are: 67%, 55% and 69% respectively. These results from a small phase II study are among the most promising of mature outcome data for IBC. They strongly suggest, along with results of several already published phase II trials, that ASCT could play a significant role in the first line treatment of IBC.
Inflammatory Breast Cancer (IBC) is a rare clinico-pathological entity  (1-2% of all breast carcinoma) with a very poor prognosis. Historically, IBC response to conventional treatment consisting of surgery or radiation therapy alone was short lived with a time to progression of 12-15 months with long-term overall survival (OS) rarely reaching 15%. The mortality rate in IBC after a local recurrence is close to 100% and most patients with local recurrence die with metastatic disease. The advent of successful combination chemotherapy regimens, along with local irradiation of the breast and regional lymphatics has increased the 5-year disease-free survival rate to 30-35% [2-4], but the long term survival is still significantly inferior to what is being achieved in other high-risk, non-metastatic breast cancers (BC) . The most recent SEER data estimates the median survival of women with stage IIIB IBC at 2.9 years vs 6.4 years with other locally advanced BC . The optimal multimodality therapy remains unestablished and current therapy unsatisfactory . Outside of selected tertiary care centers, the prognosis is even poorer, mostly because of the aggressive nature of IBC and of later diagnosis .
High-dose chemotherapy followed by autologous progenitor blood cell transplantation (ASCT) has been extensively studied in high-risk non-metastatic BC and is a particularly attractive modality for IBC because of the diffuse and aggressive nature of the disease and its propensity to early micrometastasis. From 14 published randomized trials evaluating ASCT efficacy in high-risk non-metastatic BC [9-23] has emerged a de facto general consensus that ASCT does not provide a substantial therapeutic advantage for high-risk non-metastatic BC although two separate meta-analyses of these trials support a modest (13 to 15%) improvement in EFS (but not OS) with ASCT [24,25]. No such assertion of consensus, however, can be made for IBC from these studies since they either specifically excluded subjects with IBC or lacked power for meaningful subset analysis (a total of 30 out of 6063 enrolled patients may have had IBC).
The only outcome data on large numbers of IBC patients treated with ASCT originate from the transplant registries of the CIBMTR and the EBMT Solid Tumors Working Party. They are suggestive of a beneficial effect of ASCT but have not been formally reported. In its 2000 summary report, the CIBMTR reported a 57% 3-year OS for non-metastatic IBC, data based on 811 women who underwent ASCT between 1991and 1997 (no EFS is available). The EBMT Solid Tumors Working Party briefly reported a median PFS of 57 months on 537 transplanted IBC patients . These data should be interpreted with caution as they arise from unavoidably heterogeneous populations, both in disease status and specific treatment modality.
Several phase II studies of ASCT in non-metastatic IBC have been conducted, almost invariably suggesting a substantial benefit over conventional therapy but most also report on short follow up of 2 to 3 years. Here, we report mature data of the NCI experience with ASCT in the treatment of IBC and review the available literature.
Between September 1996 and September 2008, 21 patients with non-metastatic IBC were enrolled onto NCI study 96-C-0104 to evaluate the role of paclitaxel and cyclophosphamide (TC) followed by high-dose melphalan / etoposide (ME) and ASCT in the treatment of IBC. All patients were required to have a diagnosis of carcinoma of the breast, histologically confirmed by the Laboratory of Pathology of the National Cancer Institute. The diagnosis of IBC was based on the classical clinical syndrome including erythema and edema with peau d'orange appearance. The presence of dermal lymphatic involvement with tumor cells was not a requirement for diagnosis; however, patients without the typical clinical signs but with evidence of dermal lymphatic invasion on skin biopsy were also included (2 of the 21 patients).
In order to be eligible, all patients treated for their disease before enrollment on study (chemotherapy and / or definitive surgery) were required to have not failed this therapy and to have no delay between the prior therapy and therapy on study. Other eligibility requirements included: Karnofsky Performance Status >70%, creatinine clearance > 60 ml/min, AST and ALT < 3 times or bilirubin <1.5 times the upper limit of normal, absolute neutrophil count (ANC) >1000/mm3, platelet count >90,000/mm3, cardiac ejection fraction >45% at rest and DLCO>50% of the predicted value. This study was conducted with the approval of the NCI Institutional Review Board. All patients gave written informed consent.
The following TC regimen was given every 4 weeks for 3 to 7 cycles, to achieve maximum clinical response: paclitaxel: 53.3 mg/m2/day continuous i.v. infusion for 3 consecutive days (total dose over 72 hours: 160mg/m2) through a permanent central venous access device, cyclophosphamide: 900mg/m2/day i.v. over one hour, daily for 3 days (total dose 2700 mg/m2) and mesna: daily dose of 30% of the cyclophosphamide daily dose. Premedication for paclitaxel, (dexamethasone, cimetidine and diphenhydramine), standard anti-emetics (5HT3 antagonists) and hydration pre-cyclophosphamide were administered to all patients. G-CSF: 5 μ/kg/day s.c. was started on day 5 of each cycle and continued until ANC >1000 cells/mm3; during cycle 2 (PBSC mobilization), the dose was increased to 5μ/kg twice daily until the last day of apheresis.
Additionally, all patients received an anthracycline-based regimen, either prior to enrollment on study or as part of the pre-transplant induction chemotherapy on study. Patients who had not received prior anthracycline received, following TC chemotherapy, 4 cycles of doxorubicin: 60 mg/m2 i.v. rapid infusion and cyclophosphamide: 600 mg/m2 intravenously on day 1 (AC) every 3 weeks.
PBSC were collected and cryo-preserved after the 2nd TC cycle. When the white blood cell count was between 2000/mm3 and 5000/mm3, a peripheral CD34+ count was obtained daily. Once the CD34+ count was >20/μl, daily 15 to 25 liter apheresis began. Apheresis could also be started for WBC >5000/mm3. The target CD34+ cell dose was 4.0 × 106 cells/kg of body weight with a required minimum total of 2.0 × 106 CD34+ cells/kg in order to proceed with high-dose chemotherapy and ASCT.
Thus, prior to ASCT, all patients had received an anthracycline-based regimen and a minimum of three cycles of TC. Patients receiving TC in the neoadjuvant setting may have received additional cycles (maximum, 7 cycles) until maximum response.
Before proceeding with the preparative regimen, a minimum of 21 days since the last cycle of chemotherapy, complete hematologic recovery defined as an ANC of >500/mm3 and absence of non-hematologic toxicity greater than grade 1 (including a cardiac ejection fraction > 45%) were required. ME was given on days −6, -5 & -4: melphalan 53.3 mg/m2 i.v. over 30 minutes daily for 3 days (160 mg/m2 total dose) and etoposide 600 mg/m2 i.v. over 8 hours daily for 3 days (1800 mg/m2 total dose) starting 1 hour after melphalan infusion completion. PBSC were infused on Day 0. G-CSF 5μ/kg/day started on day 0 after the PBSC infusion and continued until the ANC was ≥ 1000/mm3. HSV seropositive patients received acyclovir prophylaxis until discharge from the hospital then, following hematopoietic recovery, all patients received pneumocystis jerovecii prophylaxis for 6 months.
Loco-regional therapy was assessed individually. Surgery (modified radical mastectomy) was performed either prior to entry on study or following TC. All patients received radiation therapy starting 6 weeks after ASCT. Patients usually received 5000 cGy with an additional 1000 cGy chest wall boost. Patients with HR expressing tumors received tamoxifen 20mg or anastrazole 1mg daily for five years post-transplant starting after completion of radiation therapy. Patients with Her-2 over-expressing tumors did not receive specific antibody therapy.
At entry on study, metastatic disease was excluded in all patients with a CT scan of chest, abdomen & pelvis, a bone scan and a head CT scan or MRI. Tumor markers were not routinely obtained. All patients were restaged clinically and radiologically after the first three and every two subsequent TC cycles. Prior to ASCT, the restaging also included a repeat brain CT scan or MRI. Patients underwent a clinical re-evaluation 6 weeks following ASCT, then every 3 months for two years, every 6 months for one year and yearly thereafter. Imaging re-evaluations were performed at 6 weeks, then 6, 12, 18 & 24 months post ASCT routinely, then only as clinically indicated. Disease response was evaluated as follows: complete response (CR): disappearance of all clinical and radiological disease and no new lesion; partial response (PR): > 50% disease reduction in existing measurable disease and no new lesion; stable disease (SD): < 25% change in existing measurable disease; progressive disease: > 25 % increase in existing measurable disease or appearance of new lesions.
Patients with PD at any re-evaluation or patients with less than PR at the re-evaluation immediately before ASCT were considered treatment failures and taken off study. All toxicities for TC and ME were recorded using the NCI Common Terminology Criteria 2.0 version.
The durations of progression-free survival (PFS), event-free survival (EFS), and overall survival (OS) were calculated from the date the patient went on-study, as well as the date of ASCT, until the date of disease progression (PFS), the date of an event defined as either the date of disease progression or death of any cause (EFS), the date of death from any cause (OS), or last follow-up as appropriate. The probabilities of these outcomes as a function of time were determined by the Kaplan-Meier method. The statistical significance of the difference between two Kaplan-Meier curves was determined by a two-tailed log-rank test; all p-values are reported without adjustment for multiple comparisons. The median potential follow-up was calculated as the median of the intervals from on-study date as well as transplant date until the date of analysis and provides a reasonable measure of the maturity of the trial.
From September 1996 to September 2008, 21 patients with non metastatic IBC were enrolled on study. Patients characteristics and outcome are summarized in Tables 1 & 2. The mean age at entry on study was 50.3 years (range 35 to 67) and 11 patients (50 %) were post-menopausal. Fifty two percent and 60% of the tumors were HR positive and Her-2 over-expressed, respectively. Ten patients (48%) had received some form of treatment prior to entry on study (neo-adjuvant chemotherapy and / or definitive surgery). Twelve of 21 patients had disease evaluable for the TC induction chemotherapy: there were 11 (92%) responses (7 CR: 58% and 4 PR: 33%) and 1 SD (9%). There was 1 pathological CR (9 %). The only patient with SD after TC reached a PR with the subsequent AC regimen. Therefore, all evaluable patients achieved the minimum required response and 20 of 21 patients proceeded to the high-dose chemotherapy. The only patient not undergoing ASCT died of sepsis after delaying seeking treatment during neutropenia following a TC cycle.
The toxicity of the treatment strategy (TC followed or preceded by AC, followed by high-dose ME and ASCT) is within generally acceptable limits and was previously reported in a larger cohort of patients . The hematologic toxicity of the TC regimen was noteworthy but manageable with brief hospitalization for empiric antibiotic therapy of febrile neutropenia occurring in 55% of the cycles. All 21 patients had adequate PBSC collections. Mortality within 100 days of ASCT was zero. Three of the 8 deaths were not related to disease progression: a 51 year-old woman, previously mentioned died of sepsis prior to ASCT; a 67-year old woman died 6 months post-ASCT from progressive multifocal leuko-encephalopathy in the absence of BC recurrence (confirmed at autopsy) and a 66 year-old patient died of pneumonia, disease-free, 47 months post-ASCT. Seven patients had a disease recurrence; 3 patients relapsed at 9 months and one each at 12, 20, 21 and 78 months following ASCT. It is noteworthy that not achieving a pathological CR following TC chemotherapy was not predictive of a shorter EFS; 5 of the 10 patients (50%) achieving a pathological PR are still clinically free of disease (EFS: one patient at 8+ months and four patients from 58+ to 123+ months) while the patient with a pathological CR prior to ASCT has a 66+ month EFS.
The median follow-up was 8.4 years from on-study date (n=21) and 8.3 years from transplant date (n=20). The probabilities of PFS, EFS and OS were 67%, 60% and 75 % respectively at 3 years and 67%, 55% and 69% respectively at 6 years from the on-study date (Figure 1a). The survival estimates are detailed in the Supplemental Table on line. No statistically significant difference in any of these three outcome measures was seen between HR positive and HR negative patients. Statistically significant differences were seen in PFS and EFS (but not in OS) from on-study date between the Her-2 over-expression and non over-expression groups with the following six-year probabilities and corresponding two-tailed log-rank test results for the comparisons: PFS: 80% vs 50%, p= 0.034; EFS: 73% vs 43%, p= 0.050; OS: 83% vs 54%, p= 0.15 (Figure 1b, Supplemental Table on line).
Since the addition of systemic chemotherapy to loco-regional treatment, now over 30 years ago, and in spite of some progress in response rate, no substantial progress has been made in the long term outcome of IBC, neither by the addition of taxanes nor by varying the drug combinations[4,28]. Newer therapies are providing invaluable insight in the disease  but have yet to demonstrate a decidedly superior therapeutic potential in a disease that lags so far behind other non metastatic BC in long term survival.
Although high-dose chemotherapy followed by ASCT has not been found to be beneficial in the treatment of high-risk non-metastatic BC and has been largely abandoned for this indication, no data from randomized clinical trials exist to reach the same conclusion for IBC. Data from two transplant registries suggest some efficacy of ASCT in treating IBC although neither registry has formally reported their results.
Several published phase II studies have found outcome benefit from ASCT in IBC patients [30-36]. Published studies are summarized in Table 3 along with our data. Viens et al.  reported on 17 consecutive IBC patients with a median follow-up of 36 months (range 17-52) with a DFS 58.8% from diagnosis, then subsequently reported a collaborative study in 100 women with non metastatic IBC (Pegase 2 trial; ) with 3-year estimated RFS and OS of 44% and 70%, respectively. Cagnoni et al. reported a cohort of 30 IBC patients with a median follow up of 2 years , subsequently expanded to 56 patients with longer follow-up  with DFS of 65% and OS of 70% at 7 years. Ayash et al.  reported on 46 women with IBC with a 30-month DFS of 64%. The 30-month DFS was estimated at 100% for patients in pathologic CR, 70% with microscopic or 38% with gross disease after ASCT. Adkins et al. reported on 47 consecutive IBC patients : at a mean follow up of 30 months, the 30-month estimated DFS and OS were 57% and 59%, respectively. In the study by Schwartzberg et al., 56 IBC patients were treated : with a median follow-up of 47 months the 3-year estimated EFS and OS were 53% and 72%. Dazzi et al.  investigated the usefulness of a neo-adjuvant high-dose anthracycline containing chemotherapy in 21 patients. Although response rate and outcome were encouraging there was significant toxicity, cardiac in particular. It noteworthy that, since the recurrence free interval in IBC is historically considerably shorter than non IBC, the shorter median follow-up in these studies still suggests an encouraging outcome.
We report here on 21 patients with non metastatic IBC following a treatment regimen successfully piloted in our institution which includes dose intensive paclitaxel / cyclophosphamide induction  then ASCT following high-dose melphalan/etoposide . As previously reported, the hormone receptor status does not appear to be of added prognostic value in our series, while, unlike what is reported for standard dose chemotherapy trials, achieving a pathological CR with standard chemotherapy does not appear to be of prognostic significance in our series which further argues in favor of an added therapeutic benefit from the high-dose chemotherapy following the determination of disease status pathologically. To our knowledge of the published mature literature on IBC outcome, encompassing mostly standard dose chemotherapy, these data represent the most favorable outcome in the first line treatment of this disease and significantly contributes to the body of evidence in support of ASCT for IBC, not only by its outcome data but also by its maturity.
In conclusion, a significant body of phase II data has now accumulated and matured indicating that ASCT may offer a significant benefit in the first line treatment of non metastatic IBC, and in the unfortunate absence of data from randomized trials to confirm or invalidate it, it collectively represents the best outcome data for this disease. Our data therefore confirm as well as extend available findings; both aspects are of substantial value in the field of ASCT and IBC therapy as it now stands. As none of the newer therapeutic approaches presently under investigation has outcome data yet approaching ASCT results (either in treatment efficacy or data maturity), we believe that substantial collaborative efforts should be devoted to validate these results in a well designed and adequately powered phase III randomized trial.
Financial disclosure: the authors have no conflict of interest in the publication of this manuscript
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.