Fax: (011) 97143527776
Differences in survival among women with stage III inflammatory and noninflammatory locally advanced breast cancer appear early
A large population-based study
Article first published online: 29 NOV 2010
Copyright © 2010 American Cancer Society
Volume 117, Issue 9, pages 1819–1826, 1 May 2011
How to Cite
Dawood, S., Ueno, N. T., Valero, V., Woodward, W. A., Buchholz, T. A., Hortobagyi, G. N., Gonzalez-Angulo, A. M. and Cristofanilli, M. (2011), Differences in survival among women with stage III inflammatory and noninflammatory locally advanced breast cancer appear early. Cancer, 117: 1819–1826. doi: 10.1002/cncr.25682
- Issue published online: 19 APR 2011
- Article first published online: 29 NOV 2010
- Manuscript Revised: 9 SEP 2010
- Manuscript Accepted: 31 AUG 2010
- Manuscript Received: 12 JUN 2010
- inflammatory breast cancer;
- locally advanced breast cancer;
- multidisciplinary management;
Significant improvements in the survival of women with breast cancer have been observed and are attributed to a multidisciplinary approach and the introduction of polychemotherapy and endocrine regimens. The objective of this population-based study was to determine whether women with inflammatory breast cancer (IBC) who received treatment in a modern era had a poorer survival compared those with non-IBC locally advanced breast cancer (LABC).
The Surveillance, Epidemiology, and End Results program registry was searched to identify women with stage IIIB/C breast cancer diagnosed between 2004 and 2007 who had undergone surgery and radiotherapy. Patients were categorized as either having IBC or non-IBC LABC according the sixth edition of the American Joint Committee on Cancer (AJCC) criteria. Breast cancer-specific survival (BCS) was estimated using the Kaplan-Meier product limit method and compared across groups using the log-rank statistic. Cox models were then fitted to compare the association between breast cancer type and BCS after adjusting for patient and tumor characteristics.
A total of 828 (19.2%) women and 3476 (80.8%) women had stage IIIB/C IBC and non-IBC LABC, respectively. The median follow-up was 19 months. The 2-year BCS rate was 90% (95% confidence interval [95% CI], 88%-91%) for the entire cohort and 84% (95%CI, 80%-87%) and 91% (95%CI, 90%-91%) among women with IBC and non-IBC LABC, respectively. In the multivariable model, patients with IBC were found to have a 43% increased risk of death from breast cancer compared with patients with non-IBC LABC (hazard ratio, 1.43; 95%CI, 1.10-1.86 [P = .008]).
In the era of multidisciplinary management and anthracycline-based and taxane-based polychemotherapy regimens, women with IBC continue to have worse survival outcomes compared with those with non-IBC LABC. Cancer 2011. © 2010 American Cancer Society.
Inflammatory breast cancer (IBC) is a rare and aggressive type of breast cancer that accounts for approximately 1% to 5% of newly diagnosed breast cancers in the United States.1 The diagnosis of this disease depends on a combination of pathological confirmation of invasive carcinoma and a set of clinical criteria including diffuse erythema, edema involving more than two-thirds of the breast, warmth, tenderness, and rapid enlargement of the breast.2, 3 The nonspecificity of the clinical criteria has resulted in variability of case definition that has been a limiting factor in comparing results of different studies that have examined the prognostic outcomes of women with IBC.4
A multidisciplinary approach to management incorporating preoperative chemotherapy, surgery, radiotherapy, and hormone therapy when indicated has transformed this disease from one that was associated with <5% of patients surviving past 5 years2, 5 to one in which 15-year survival rates of 20% to 30% have been reported.6, 7 Studies comparing IBC with non-IBC locally advanced breast cancer (LABC) have demonstrated inferior prognostic outcomes associated with IBC compared with non-IBC LABC.8-11 However, these studies are limited by the finding that a large percentage of patients were diagnosed in the 1970s, 1980s, and 1990s, when taxanes were not a standard part of treatment. Recent data have indicated that a polychemotherapy regimen comprised of both anthracyclines and taxanes may further improve survival outcomes of women with IBC.12
In 2004, the Surveillance, Epidemiology, and End Results (SEER) registry began incorporating staging based on the sixth edition of the American Joint Committee on Cancer (AJCC) that classifies IBC as T4d. Using data derived from this database and using the standard AJCC definition of IBC, we sought to evaluate whether differences in survival still exist among women with stage IIIB/C IBC and non-IBC LABC in the era of anthracycline-based and taxane-based polychemotherapy regimens. We also hypothesized that because of the known aggressive nature of IBC, any differences in survival that may exist would be observed early in the course of disease.
MATERIALS AND METHODS
The data used for this study were derived from the National Cancer Institute's SEER program, which is a comprehensive source of population-based data that covers approximately 26% of the population in the United States.13 For this study, data were derived from the limited-use SEER 17 registry databases that were released in April 2010. Search criteria were restricted to patients with pathologically confirmed invasive stage IIIB and IIIC breast cancer diagnosed between 2004 and 2007 who had undergone radiotherapy and surgery. Patients with stage IIIA breast cancer and those who had >1 primary malignancy based on coding were excluded from the analyses. The AJCC sixth edition criterion was used to define stage IIIB and IIIC disease. IBC was also defined according to the AJCC sixth edition14 and was categorized and coded within the SEER registry database as T4d disease. According to the AJCC sixth edition, IBC is defined as being characterized by diffuse erythema and edema (peau d'orange) of the breast, frequently without an underlying palpable mass. According to the AJCC sixth edition, nonmetastatic IBC is categorized as either stage IIIB or IIIC depending on the number and type of involved lymph nodes.
Patient and tumor characteristics were tabulated and stratified by breast cancer type (IBC vs non-IBC LABC) and compared across groups with the chi-square test when appropriate. The follow-up cutoff for this dataset was December 31, 2007. Breast cancer-specific survival was calculated from the time of diagnosis to the time of death from breast cancer, with patients who died of causes other than breast cancer censored at the time of their death. Overall survival was calculated from the time of diagnosis to the time of death from any cause or the time of last follow-up. Patients still alive at the time of last follow-up were censored. The Kaplan-Meier product limit method was used to compute both survival estimates and compared across groups using the log-rank statistic. Cox proportional hazards models were fitted to determine the association between type of breast cancer (IBC compared with non-IBC LABC) and the 2 survival endpoints after controlling for patient and tumor characteristics. The final multivariable model included variables that were based on clinical significance rather than statistical significance and included age, race, tumor grade, surgery type, stage of disease, hormone receptor status, and marital status. The final model was also adjusted for the 17 SEER registries. The multivariable models were then run again stratified by breast cancer type to determine prognostic factors associated with each breast cancer type. The proportional hazards assumption of the final model was assessed visually using plots of the model residuals. P values were 2-sided and a value ≤.05 was considered to be statistically significant. All analyses for this study were performed using SAS statistical software (version 9.1; SAS Institute Inc, Cary, NC).
Patient and Tumor Characteristics
Between 2004 and 2007, 4304 women with stage IIIB and IIIC breast cancer who met eligibility criteria were included in the final analysis. Table 1 summarizes the patient and tumor characteristics stratified by breast cancer type. A total of 828 (19.2%) women had IBC and 3476 (80.8%) women had non-IBC LABC. The median year of diagnosis among women with IBC and those with non-IBC LABC was 2005 (range, 2004-2007). A total of 2739 (63.6%) women had hormone receptor-positive disease, 1607 (37.3%) had stage IIIB disease, and 2607 (60.6%) had stage IIIC disease. A total of 482 (11.2%) women were categorized as being of black race and 3813 (88.6%) women were categorized as being of white/other race.
|No. (%)||No. (%)|
|Hormone receptor status|
The median follow-up among all patients was 19 months (range, 0-47 months). At the time of the analyses, 533 (12.4%) women had died of all causes and 459 (10.7%) women had died of breast cancer. Table 2 summarizes the 2-year overall and breast cancer-specific survival estimates. The 2-year overall and breast cancer-specific survival rates for the entire cohort were 88% (95% confidence interval [95% CI], 87%-89%) and 90% (95% CI, 88%-91%), respectively (Fig. 1 Top). The 2-year breast cancer-specific survival rate among women with IBC was 84% (95% CI, 80%-87%) and was 91% (95% CI, 90%-92%) among women with non-IBC LABC, with an absolute difference of 7% at 2 years, a difference that was statistically significant (P<.0001) (Fig. 1 Bottom). After adjusting for several patient and tumor characteristics (Table 3), women with IBC were found to have a 43% increased risk of death from breast cancer compared with women with non-IBC LABC (hazard ratio [HR], 1.43; 95% CI, 1.10-1.86 [P = .008]). Because the use of segmental mastectomy is not considered the standard of care among women with IBC tumors, the models were run again excluding these women. The results were found to be similar, with the risk of death from breast cancer among women with IBC still significantly higher than among women with non-IBC LABC (HR, 1.44; 95% CI, 1.10-1.88).
|Breast Cancer-Specific Survival||Overall Survival|
|2-Year||(95% CI)||P||2-Year||(95% CI)||P|
|Type of cancer|
|Hormone receptor status|
|Breast Cancer-Specific Survival||Overall Survival|
|HR||(95% CI)||P||HR||(95% CI)||P|
|IBC vs non-IBC LABC||1.426||(1.097-1.854)||.0079||1.387||(1.087-1.770)||.0086|
|Black vs white/other||1.691||(1.260-2.271)||.0005||1.671||(1.269-2.200)||.0003|
|Grade III vs grade I/II||1.922||(1.469-2.515)||<.0001||1.878||(1.468-2.404)||<.0001|
|Mastectomy vs segmental||1.444||(1.103-1.890)||.0075||1.386||(1.081-1.775)||.0099|
|Stage IIIC vs. IIIB||1.250||(0.995-1.571)||.0552||1.157||(0.937-1.428)||.1747|
|Age <55 y vs ≥55 y||0.813||(0.665-0.994)||.0430||0.766||(0.635-0.924)||.0053|
|Hormone receptor status (positive vs negative)||0.384||(0.309-0.477)||<.0001||0.409||(0.335-0.500)||<.0001|
|Marital status (yes vs no)||0.984||(0.799-1.212)||.8796||0.924||(0.761-1.121)||.4210|
Table 4 summarizes the results of the multivariable models for breast cancer-specific survival stratified by breast cancer type. Factors found to be significantly associated with worse breast cancer-specific survival among women with IBC included black race, stage IIIC disease, and negative hormone receptor disease. Factors found to be significantly associated with worse breast cancer-specific survival among women with non-IBC LABC included black race, nuclear grade III disease, and hormone receptor-negative disease. Furthermore, women with non-IBC LABC who underwent mastectomy were found to have worse breast cancer-specific survival compared with those who underwent segmental resection.
|HR||(95% CI)||P||HR||(95% CI)||P|
|Black vs white/other||2.688||(1.420-5.089)||.0024||1.504||(1.073-2.109)||.0179|
|Nuclear grade III vs grade I/II||1.651||(0.934-2.916)||.0842||1.963||(1.443-2.669)||<.0001|
|Mastectomy vs segmental||1.765||(0.545-5.716)||.3430||1.432||(1.084-1.891)||.0115|
|Stage IIIC vs IIIB||2.143||(1.367-3.360)||.0009||1.077||(0.832-1.393)||.5736|
|Age <55 y vs ≥55 y||0.706||(0.465-1.070)||.1011||0.881||(0.698-1.111)||.2839|
|Hormone receptor status (positive vs negative)||0.386||(0.245-0.608)||<.0001||0.382||(0.298-0.490)||<.0001|
|Marital status (yes vs no)||1.000||(0.653-1.531)||.9997||0.963||(0.756-1.228)||.7633|
The goal of this population-based study was to determine whether differences in survival still exist among women with stage IIIB/C IBC and non-IBC LABC in an era in which such patients would be recommended to receive multidisciplinary care that includes a polychemotherapy regimen comprised of an anthracycline and taxane. Several interesting observations were demonstrated in the current study. First, differences in survival continued to be apparent. Patients with IBC had a 43% higher risk of breast cancer death compared with women with non-IBC LABC. Second, the differences in survival were observed to appear early, with a statistically significant absolute difference in breast cancer-specific survival of 7% at 2 years, favoring the group of women with non-IBC LABC.
The results of the Early Breast Cancer Trialists' Collaborative Group's15 15-year update demonstrated that 6 months of anthracycline-based polychemotherapy regimens reduced the annual breast cancer death rate by 38% among women aged <50 years and by 20% among women ages 50 to 69 years. Similarly, Ueno et al7 demonstrated the benefit of adjuvant anthracycline in a cohort of 178 women with IBC, with the authors reporting 5-year and 10-year overall survival rates of 40% and 33%, respectively, with an impressive 28% of patients reported to be alive and without disease at 15 years of follow-up. The benefit of the addition of taxanes to therapy for women with high-risk nonmetastatic breast cancer has also been shown in prospective studies and meta-analyses.16 Cristofanilli et al12 retrospectively compared 178 women with IBC who were treated with an anthracycline-only regimen compared with 62 women with IBC who received both an anthracycline and taxanes and reported that the group receiving taxanes had a higher pathological complete response rate and higher progression-free and overall survival compared with the group that did not. With such advances made in the treatment of IBC and non-IBC LABC, one could potentially begin to question whether survival differences still exist between the 2 groups. Thus, using data derived from the SEER registry, we retrospectively attempted to answer this question among women diagnosed with stage IIIB/C IBC and non-IBC LABC between 2004 and 2007. The current study data indicated that the survival differences still existed, with women diagnosed with IBC having a statistically significant 43% increased risk of death from breast cancer compared with women with non-IBC LABC (HR, 1.43; 95% CI, 1.10-1.86 [P = .008]).
The question that now comes to mind is that despite such advances, what factors are contributing to the observed differences? First, we made the assumption that all women were treated with an anthracycline-based and taxane-based regimen because SEER does not specifically code for the type of chemotherapy administered. The homogeneity of the treatment assumption may not be entirely correct. The use of taxanes among women with lymph node-positive breast cancer was approved by the US Food and Drug Administration (FDA) in October 1999 and since then, the general consensus has been to use taxanes for the treatment of women with high-risk breast cancer. Using the SEER-Medicare database, Giordano et al17 studied the impact of the use of taxanes among women with breast cancer after the May 1998 American Society of Clinical Oncology presentation of the Cancer and Leukemia Group B (CALGB) 9344 study (which to our knowledge was the first randomized trial to evaluate the treatment effect of adjuvant paclitaxel in women with lymph node-positive breast cancer),18 but before the approval of taxanes for this particular indication by the FDA. The authors reported an increase in the use of taxanes from 5.2% before May 1998 to 23.6% after the May 1998 presentation. Extrapolating from these data and the finding that the official publication of the CALGB 9344 °Ccurred in October 2003, we felt strong in our assumption that women with stage IIIB/C IBC and non-IBC LABC who were diagnosed between 2004 and 2007 would have a high probability of being treated with an anthracycline and a taxane.
Second, although our analysis was restricted only to women who had received radiotherapy, important information regarding dose and fields covered by radiotherapy could not be determined because this information is not coded in SEER. Recent research has focused on IBC-specific approaches using hyperfractionated radiotherapy or daily bolus application to increase skin dose to improve locoregional control rates among women with IBC.19-22 Although data indicate that postmastectomy accelerated hyperfractionated radiation to 66 grays with radiation delivered to chest wall as well as the axillary, infraclavicular, supraclavicular, and internal mammary lymph node regions improves locoregional, distant metastases-free, and overall survival among women with IBC, this is not standard practice in the community. Similarly, once-daily regimens with tailored daily bolus regimens to increase dose intensity in the skin for women with IBC has been reported to yield excellent locoregional control but is not standard practice in the community.22 It is therefore likely that delivered adjuvant radiotherapy to women with IBC coded within SEER is not homogenous and may thus contribute to the survival differences observed.
Third, a simpler and more direct explanation for the survival differences observed is that IBC is truly a unique entity separate from non-IBC LABC. Loss of WNT1-inducible-signaling pathway protein 3 (WISP3)23 and overexpression of Ras homolog gene family, member C (Rhoc) GTPase24 have been demonstrated in approximately 80% to 90% of IBC tumors compared with 20% to 30% of non-IBC tumors. Using cDNA microarrays, distinct gene expression signatures of IBC have also been identified.25 At the biological level, several studies have also demonstrated increased angiogenesis in IBC.26 As such, there are currently several prospective studies evaluating inhibitors of angiogenesis (eg, bevacizumab) for the treatment of IBC.27 A higher frequency of overexpression of human epidermal growth factor receptor 2 (HER2) among IBC tumors compared with non-IBC tumors has also been reported,28, 29 making HER2-directed targeted therapy with adjuvant trastuzumab important. Women diagnosed in 2004 and 2005, before the approval of trastuzumab in the adjuvant setting, were also included in the cohort studied, which may have contributed to the survival differences observed.
In conclusion, in this large population-based study, we demonstrated that in the era of the use of anthracyclines and taxanes, survival differences between women with stage IIIB/C IBC and those with non-IBC LABC are still present, with differences observed early. Despite all the inherent biases that are usually associated with retrospective analyses and the limitations specifically associated with SEER as outlined earlier, several important strengths of the current study must also be acknowledged. First, IBC is a rare disease and to make significantly meaningful comparisons, large numbers of patients are needed, which is afforded by SEER. Second, the results of the analysis reflect the outcomes observed among women treated in the community rather than single-specialty institutions from which the majority of the published IBC data has been derived to date. Third, such large population-based studies are hypothesis generating. The results observed herein reveal that inferior prognostic outcomes continue to be associated with IBC and that newer targeted agents are required to improve survival outcomes. Furthermore, the early differences observed further serve to strengthen the theory that not only is IBC a more aggressive and unique phenotype but also indicate that perhaps consideration should be given to the more aggressive follow-up of these women. Indeed, we recently made the important observation that a high early incidence of brain metastases was associated with IBC, making it an ideal cohort in which to consider studying early detection and the prevention of disease recurrence to the brain.30 Regardless, it is clear that further study into the biology and treatment of IBC is needed to improve its associated prognostic outcome.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 2Inflammatory carcinoma of the breast. Surg Gynecol Obstet. 1924; 39: 580-595.,
- 3Diseases of the Breast. 2nd ed. Philadelphia: WB Saunders; 1971: 576-584.
- 14Breast. In: GreenFL, PageDL, FlemingID, et al, eds. AJCC Cancer Staging Manual. 6th ed. New York: Springer- Verlag; 2002: 225-281.