Sage Journals: Discover world-class research

Abstract

The survival after the diagnosis of inflammatory breast cancer (IBC) has been steadily improving for the past few decades. This has been due to advances in the knowledge of IBC in a number of fields, including epidemiology, molecular biology, and medical management. In this review we summarize some of the most important recent advances in these fields and suggest possible opportunities for continued improvement.

Keywords

Inflammatory breast cancer epidemiology advances

1. Background

Considerable progress has been made in defining the risk factors for developing IBC and discovering what determines long term survival. Eight years ago we noted the obstacles to obtaining appropriate data on this disease, focusing on the lack of a consistent case definition and the relatively small population of patients [1]. Now there is more agreement on identifying IBC patients with the emphasis on recent onset and rapid clinical changes [2]. There has also been an increase in understanding the molecular biology and the effect on survival leading to an increasing acceptance of inflammatory breast cancer (IBC) as a unique entity and no longer a subcategory of another category of breast cancer such as locally advanced breast cancer (LABC) [3–5]. Furthermore, international communication has improved, expanding the patient population available for evaluation. Additionally, the improved correlation of clinical and epidemiological information with laboratory findings has added to the clarification of what IBC is.

2. Resolving the case definition

When IBC was first identified, a unified case definition did not exist. The two major groups working on standardizing the case definition have been the American Joint Committee on Cancer (AJCC), responsible for developing a uniform staging system for cancer for clinicians, and the surveillance, epidemiology and end results (SEER) Program of the National Cancer Institute which uses highly qualified regional population-based cancer registries for epidemiologic studies. The AJCC definition was primarily a clinical diagnosis. SEER based the diagnosis primarily on histopathology. For both groups, the diagnosis of IBC has evolved over the years, both of them becoming closer but they are still not identical.

At the same time that IBC was being defined as an important entity in the U.S., French clinicians were developing a different approach to IBC. Denoix at the Institut Gustave Roussy in Paris recognized IBC as an extremely aggressive form of breast cancer and coined the term pousee evolutive (PEV) meaning rapidly progressing breast cancer [6]. He described three phases:

PEV3 with clinical involvement covering more than half of the breast, clearly the equivalent of IBC as defined by the AJCC.

PEV2 with clinical signs limited to less than half the breast, recognized in France as an earlier manifestation of IBC but not recognized as IBC in the US.

PEV1 with no skin manifestations of IBC but with rapid tumor growth identified first by the patient and confirmed by the physician. It had the same poor outcome as PEV2 and PEV3, but not been recognized as IBC in the US.

The term PEV0 came to be used for non-IBC breast cancer, which does not have rapid growth.

The PEV classification was also in Tunisia, where IBC affected approximately half of the breast cancer patients seen in the Institut Salah Azaiz, the country’s National Cancer Institute [7]. A comparison of the PEV case definitions with the case definitions of AJCC and SEER is shown in Table 1.

Table 1
PEV categories compared to AJCC and SEER definitions

Category Description

PEV3 Clinical and historical evidence of rapid growth, redness, warmth and edema involving more than half of the breast AJCC+ SEER + ¹ only if there are path findings (dermal lymphatic invasion) (DLI)

PEV2 Clinical and historical evidence of rapid growth, redness, warmth and edema involving less than half of the breast AJCC+ SEER−

PEV1 A history of rapid growth without clinical findings of redness, warmth or edema AJCC− SEER + ¹ only if there are path findings (DLI)

PEV0 The absence of clinical or historical evidence of rapid growth AJCC− SEER¹ uncertain if path findings (DLI) are noted

¹At the time of our IBCR development, the uncertainty of a case categorized by the PEV classification as being classified by SEER as IBC depended on the patient record documenting DLI.

Category	Description
PEV3	Clinical and historical evidence of rapid growth, redness, warmth and edema involving more than half of the breast	AJCC+ SEER + ¹ only if there are path findings (dermal lymphatic invasion) (DLI)
PEV2	Clinical and historical evidence of rapid growth, redness, warmth and edema involving less than half of the breast	AJCC+ SEER−
PEV1	A history of rapid growth without clinical findings of redness, warmth or edema	AJCC− SEER + ¹ only if there are path findings (DLI)
PEV0	The absence of clinical or historical evidence of rapid growth	AJCC− SEER¹ uncertain if path findings (DLI) are noted

Because of the challenge of IBC and the need for a uniform case definition, in 2002 we developed the IBC Registry (IBCR) and a biospecimen repository with funding from the Department of Defense to “clarify the epidemiology and biology of these tumors” [1]. We established a registry of 181 patients from the United States and Canada who had been given the diagnosis of IBC by a clinician and who provided their histories, medical records and tissue blocks for laboratory studies to determine if specific diagnostic markers could be identified that would define the disease. For the purpose of coordinating the clinical, epidemiological and laboratory data, each of our cases was classified using six categories based on AJCC, SEER, and PEV classifications but also including “secondary IBC” described by Taylor and Meltzer [8] and “occult” IBC described by Saltzstein [9].

Our subsequent laboratory studies focused on a classification which allowed a comparison of cases categorized as IBC by either AJCC, SEER or PEV, as shown in Table 2. Category 1, for example, included cases that would be classified as IBC by all three groups whereas Categories 2–4 would not be called IBC by all three. At the time our laboratory studies started, Category 4 cases were not classified as IBC by SEER or AJCC because both required that more than half of the breast be clinically involved but subsequent revisions of their classifications have reduced the size allowing the diagnosis of IBC in more cases. We did not use Category 5 in our analyses because only SEER could have included pathologic findings of dermal lymphatic invasion (DLI) alone and as the study continued, none of the cases, all referred by clinicians, fit that category. Category 6, secondary IBC, was not included in our studies because of the absence of biospecimens to evaluate.

Table 2

IBC Registry categories compared to AJCC, SEER, and PEV definitions

Category	Description
Category 1	Classical history and physical findings, pathological confirmation (invasion of dermal lymphatics)	AJCC+ SEER+ PEV3
Category 2	Classical history and physical findings, no pathological confirmation	AJCC+ SEER− PEV3
Category 3	Clinical findings involving less than half of the breast, pathological confirmation	AJCC− SEER+ PEV2
Category 4	Clinical findings involving less than half of the breast, no pathological confirmation	AJCC− SEER− PEV2
Category 5	Pathologic findings without clinical features (“occult IBC”)	AJCC− SEER + ² only if there are path findings (DLI) PEV0
Category 6	IBC features in recurrent breast cancer (secondary IBC)	not applicable

²At the time of our IBCR development, the uncertainty of a case categorized by the PEV classification as being classified by SEER as IBC depended on the patient record documenting DLI.

Category 1, termed “classic IBC”, included those cases with more than half the breast clinically involved, which met the criteria for PEV3 and AJCC, with DLI required by SEER. Category 2 had the clinical features of Category 1 but did not have DLI and therefore did not meet the SEER case definition of IBC. Category 3 had clinical findings involving less than half the breast, comparable to PEV2, and had DLI and therefore met the SEER criteria for IBC but not AJCC criteria. Category 4 had less than half the breast involved and no DLI comparable to Category 2 but did not meet the AJCC or SEER criteria for IBC. Our laboratory studies in part focused on whether patients diagnosed as IBC by their doctors meeting our Category 4 of IBC (PEV2) were the same disease as Categories 1–3.

Laboratory studies have been very important in helping to show the relevance of the different case definitions. Because a consensus document from an international expert panel on IBC [2] developed minimum criteria for the diagnosis of IBC which required rapid onset of breast erythema, edema and/or peau d’orange as well as pathological confirmation of invasive carcinoma, we decided not to include “occult” IBC in our laboratory studies. “Secondary” IBC did not provide sufficient tumor material for these studies and therefore most of our laboratory studies included only Categories 1–4. Two early studies utilized the PEV classification. The first involved hormone receptors [10] and showed that the mean level of estrogen receptors in PEV1 tumors (19.0 +∕− 6.6) were similar to PEV3 tumors (classic AJCC IBC) (17.9 +∕− 6.1) and both were significantly lower than the mean level in PEV0 tumors (no rapid growth) (30.4 +∕− 9.2). The second study, focused on microvessel density [11], also showed no difference between PEV1, 2 and 3 but markedly greater microvessel density than PEV0. These studies suggest that PEV1 could be a sub-clinical form of IBC.

One subsequent laboratory study [4] compared lymphangiogenic factors, vascular endothelial growth factor D (VEGF-D), and e-cadherin in 100 IBC cases (atypical IBC and classic IBC cases) from the IBC registry with 107 non-IBC LABC cases from the National Cancer Institute’s Cooperative Breast Cancer Tissue Resource (CBCTR). The atypical IBC cases (IBCR Category 4) were cases diagnosed by clinicians but not meeting the case definitions of AJCC or SEER because the clinical features of IBC involved less than half of the breast and there was no documented invasion of dermal lymphatics (no pathological proof).

In another study, Raghav et al. [5] did an analysis of the histological subtypes of 659 IBC patients at MD Anderson Cancer Center. Thirty of the patients with IBC (4.6%) had invasive lobular tumors, 37 (5.6%) had mixed invasive ductal and lobular tumors, and 592 (89.8%) had invasive ductal tumors. They found that the histology did not have a significant effect on survival outcomes in IBC patients, unlike in non-IBC breast cancer patients [12,13], thereby suggesting that IBC is distinct at the molecular level.

The overall findings showed major differences between IBC versus LABC, confirming the previous epidemiologic report showing these were distinct entities [3] and showed the important finding that “atypical IBC”, Category 4, resembled “classic IBC” (Categories 1–3) in all three quantitative tests (E-cadherin, VEGF-D, and LVD) rather than LABC. This study, therefore, suggests that even at the present time the arbitrary definition of clinical involvement of more than 1∕3 of the breast used by AJCC and the international IBC panel [2] may omit cases of IBC. Another outcome of this study was the decision to include all cases of IBC Categories 1–4 in subsequent laboratory studies using IBC biospecimens [14,15].

3. Etiologic studies

Considerable information regarding the causes of IBC indicates that the primary contributors are environmental rather than genetic. It is apparent from population-based registries that the relative incidence of IBC in the US is between 1% and 3% of breast cancer cases [3,16,17] depending on the case definitions used, and it is apparent that different ethnic/racial groups have different incidence rates. Studies have shown that Black women have significantly higher rates of IBC than other groups, and some studies showed that Asian women had the lowest rates [18,19]. These differences could have a genetic or environmental basis but geographic variation independent of racial/ethnic differences strongly support the more important environmental influences. Such differences, first suggested by early reports in Tunisia showing an urban-rural difference [20], have been confirmed over time with the development of a population-based registry in Tunisia, a country with a relative incidence rate of IBC among the highest in the world. The population-based studies in 1994 [21] and 2008 [22] had similar results with the proportion of breast cancer patients with IBC (PEV2 and 3) being 23.2% in 1994 and 24.3% in 2008. The population-based data indicated that the relative proportion of cases between PEV2 and PEV3 showed PEV2 more common (17% and 17.6% in the two series) than PEV3 (6.2% and 6.7%). Support for the hypothesis that the rural predominance in the early study [20] related to socioeconomic status was provided by the predominance of IBC patients in the public sector hospitals versus the private sector hospitals. T4b (comparable to PEV2) comprised 19.4% of breast cancer cases in the public sector hospitals versus 7.7% in the private. T4d (comparable to PEV3) comprised 7.1% in the public versus 3.2% in the private.

The hospital-based single institution studies [7,23] at the Institute Salah Azaiz (ISA), the National Cancer Institute of Tunisia, provided important data over 40 years. The studies found interesting differences as well as similarities in regard to clinical presentation. In the first study [7], 283 of the 581 breast cancer patients (48.7%) between January 1969 to December 1974, were PEV3, and 38 were PEV2 (6.5%). In the subsequent study [23] of 3360 patients between January 1975 and December 1996, 419 (12.5%) were classified as PEV2 or 3, less than the percentage in the first study and clearly higher than the 1–3% reported in the United States. Population based studies show a much lower percentage of PEV3 cases (6.2%) than the 48.7% reported at ISA and there was no difference between the percentages in 1994 and 2004 [22] but the combined percentage of cases of PV 2 and 3 (24.3%) was significantly higher than the 1–3% in the United States.

In summary, both the population and hospital studies indicate that IBC is significantly higher in Tunisia than the United States, but the decrease over time seen in the hospital-based studies is not confirmed in population-based data. However, the importance of lower socioeconomic status as reflected in a prominence of rural versus urban residence and public sector versus private sector hospitals emphasizes lower socioeconomic status as an important risk factor.

3.1. Socioeconomic factors

Descriptions of patterns of IBC suggest that a lower socioeconomic status is a factor in the cause of the disease. As noted above, in Tunisia lower socioeconomic status in the rural populations correlated with a rural predominance of IBC [20] and the apparent national decline of IBC occurred with improved socioeconomic conditions [24]. In the United States, Scott et al. [25] suggested important socioeconomic influences in a study of spatial clustering of county-based IBC rates drawn from the United States Cancer Statistics database. Geographic high rate IBC clusters tended to be more urban than rural, have a higher proportion of Black women, and have a higher percent of the population in poverty. Evidence of spatial clustering was statistically significant, the average rates of IBC in the high rate clusters being approximately 12 times the rates in the low rate clusters. These data were supported by a study using SEER data on the county level [26] which showed that incidence rates for IBC increased as socioeconomic status decreased. Denu et al. [27] also noted that IBC patients were more likely to come from areas of higher poverty. In a nested case-control study comparing 617 IBC cases, 1151 LABC cases, 7600 other breast cancer cases, and 93,654 control subjects, all matched by age and year of diagnosis, Schairer et al. [28] found a higher level of education was associated with a decreased risk of IBC.

In contrast, non-IBC breast cancer have the opposite pattern; higher income women have a higher breast cancer incidence than low income women, linked at least in part to delayed first pregnancy [29].

3.2. Hormonal risk factors

Hormonal factors play a major role in premenopausal IBC, and some of the hormonal risk factors are different for IBC than for non-IBC breast cancer. The most important ones are early age at first pregnancy, obesity and possibly prolonged breast feeding. Other less well documented factors are late menarche and oral contraceptives.

3.2.1. Early age at birth of first child

The importance of early age at birth of first child was strongly suggested by the Tunisian studies [20] which noted that 14 of 15 pre-menopausal women who gave birth at the age of 18 or younger were diagnosed with IBC. This observation was partially confirmed in a subsequent study by Chang et al. [30], who noted that age at birth of first child was lower in the IBC group compared to the non-IBC breast cancer group and the group with other cancers, but the difference was not statistically significant. In a study on risk factors for aggressive breast cancer [31], women who had their first child before the age of 20 had a 3.2 increased odds of having a higher grade tumor, which is a sign of aggressive breast cancer. This is in marked contrast to non-IBC breast cancer where pregnancy has a protective effect. Schairer et al. [28] found that there was a reduction in the risk of ER-IBC with older age at first birth. Atkinson et al. [32] found that women whose age of first pregnancy was <26, had a higher risk of triple-negative IBC.

3.2.2. Obesity

A number of studies have documented obesity as defined by BMI as a major risk factor for pre-menopausal IBC in contrast to non-IBC pre-menopausal breast cancer [28,30].

3.2.3. Prolonged breast feeding

Two studies, Le et al. [33] and Mejri et al. [34] found that prolonged breast feeding was significantly associated with IBC. Le et al. compared 49 French IBC patients with 139 non-IBC controls. They found that women who breast fed 25 months or more had a 4 times greater risk of IBC compared to those who breastfed 6 months or less, and that those who breast fed for 7--24 months, had an almost 2 times greater risk.

Mejri et al. had similar findings. The Mejri study compared 160 Tunisian IBC patients with 580 non-IBC Tunisian patients. They found that women who had breastfed 12 months or more had a 4.6 times greater risk for IBC.

A report by Atkinson et al. [32] noted that a history of breast feeding was associated with a lower risk of triple negative and luminal IBC, but they did not investigate prolonged breast feeding as in the Le and Mejri studies.

Numerous reports including two recent met analyses [35,36] showed that breast feeding was shown to decrease the risk of non-IBC breast cancer proportional to the length of time, thereby indicating another risk factor markedly different in IBC and non-IBC breast cancers.

3.2.4. Oral contraceptives (OC)

A significant contribution of OC use was noted in a study by Moslehi et al. [37] when IBC patients were compared to two large groups of healthy women. OC use was statistically significantly higher among the IBC cases compared to controls in the Women's Health Initiative (WHI) and controls from a study by the University of Toronto.

3.3. Non-hormonal risk factors

Among the non-hormonal risk factors, there are some, such as delay in diagnosis and clustering, which are more prominent in IBC compared to non-IBC breast cancer. Others, such as genetics, are similar in both IBC and non-IBC breast cancer.

3.3.1. Delay in diagnosis

Although both pre-menopausal and post-menopausal IBC patients had a significantly longer delay in diagnosis than non-IBC breast cancer patients, in post-menopausal patients it is a leading risk factor for IBC. Tabbane et al. [7] speculated that in view of the rapid transition observed from PEV1 and 2 to PEV3, there may also be a sudden transition from PEV0 to PEV1-3. Therefore, it is reasonable to assume that the longer a PEV0 tumor exists untreated, the longer the patient has the risk of developing PEV1-3, or IBC.

The delay in diagnosis could explain why rural women in Tunisia have higher rates of IBC than urban women. In a study of the mean tumor size in the population-based Tunisian Cancer Registry in 2004, Maalej et al. [22] found that in 1437 new cases of IBC, tumors seen in public clinics were larger (mean size = 42.5 mm) than those seen in private clinics (mean size = 32.3 mm). The larger tumor size indicates a delay in diagnosis, as the tumor has had more time to grow. Since rural women are more likely than urban women to use public clinics because of their lower socioeconomic status, they are more likely to have larger tumors and therefore higher rates of IBC.

Mourali et al. [20] found that late menarche was associated with post-menopausal IBC in Tunisia, even though women in urban Tunisia tend to have menarche at a later age than women in rural Tunisia, and rural residence in Tunisia is shown to be associated with increased risk of IBC, thus supporting delayed diagnosis as having an influence on the emergence of IBC. Mourali et al. [20] also found delay in diagnosis was associated with post menopausal patients.

3.3.2. Genetics

Overall, studies to date indicate that genetics may have a role in IBC etiology but to no greater extent than non-IBC breast cancer. In a study by Schairer et al. [28] 617 IBC patients, 1151 LABC patients, and 7600 breast cancer patients were compared to 93,654 controls with no cancer, matched on age and year at diagnosis. In the multivariable rate analyses, they found that the risk ratios of persons with a family history of breast cancer for each breast cancer group did not vary much among the three breast cancer groups but did differ from the controls. The percentage of patients who had a family history of breast cancer was 19.6% for IBC, 18.3% for LABC, and 19.6% for breast cancer, and the percentage for controls was 14.1%.

In another study conducted by Moslehi et al. [37], 141 IBC patients were compared to 178 non-randomized non-IBC patients (adjusted for age, parity, BMI, regular alcohol use, whether they had ever used OCs), they found that 17% of the IBC cases had a first-degree breast cancer family history, whereas 24.4% of the non-inflammatory breast cancer patients had a family history. They also compared the IBC patients with 465 post-menopausal breast cancer cases from the WHI trial, unselected for family history, and found that 16.9% of the breast cancer patients in the WHI trial had a family history, the same percentage as in the IBC patients. They also compared the IBC patients with 9317 healthy controls from the WHI trial with no personal history of breast cancer and found that 12.6% of the healthy controls had a family history. In these several studies, they found that the IBC patients were no more likely than non-IBC breast cancer patients to have a family history of breast cancer, but were more likely than controls to have a family history of breast cancer.

In a study at M.D. Anderson in Texas by Chang et al. [30], they compared 68 histologically confirmed IBC cases with 143 women with non-IBC breast cancer (age-matched), and with 134 women with non-breast cancers. They found a slightly non-significant higher incidence of familial breast cancer in the IBC patients (13%), when compared to the non-IBC breast cancer cases (8%).

Thus far there has been no study conclusively linking IBC to any genetic markers identified in the laboratory. That such markers could eventually be detected is suggested by a Tunisian study [20], where 43% of women with IBC had blood type A+ compared to 32% of women with non-IBC breast cancer and 33% of a large population of healthy blood donors. In a subsequent independent study [38], the association with blood type A was confirmed, but HLA typing for A, B and DRW antigens revealed no specific IBC-associated antigen. A relationship between blood type A and IBC has not been evaluated in other populations.

3.3.3. Local trauma to the breast

Observational studies have suggested various triggers to IBC, one being local trauma to the breast. Taylor and Meltzer [8] first described two forms of IBC, the primary form where the clinical characteristics are apparent from the outset, and secondary IBC, where the clinical features appear subsequent to treatment for non-IBC breast cancer. The classic rash and pathologic confirmation of dermal lymphatic invasion are typical of secondary IBC and suggest that the trauma of surgery in the area of non-IBC breast cancer caused prominent angiogenesis whichstimulated the transformation to IBC. The suggestion that trauma could precipitate aggressive cancer has been noted previously [39]. In our IBC Registry, nine patients of secondary IBC and three other women attributed the onset of IBC to significant trauma. IBC appeared immediately after a seat belt trauma in one woman, after (less than one month) a painful ductogram in a 63-year-old woman with fibrocystic disease, and shortly after an elective nipple piercing in a healthy 33-year-old woman.

3.3.4. Acute exposures

Time--space cancer clusters, defined as a number of new cases appearing in a short period of time, are always of great interest because they can suggest a specific environmental trigger of the disease in the area of the cluster. Unlike the clustering study described above, time--space clustering can identify a specific agent for the cluster that may not be important in another cluster; the trigger for one could be infectious and for another it could be a toxic spill [40].

There are three characteristics that are critical to indicate a cancer cluster that is worth investigating for an environmental trigger.

The cluster should consist of only one type of cancer.

The cancer should be a rare cancer.

The cancer should have a short time period between the causative trigger and the appearance of the cancer (latent period).

The last is important because most cancers have a long latent period and the cells divide slowly and at a variable rate so a time--space cluster is not a likely result. IBC fits all of these criteria. It is a relatively rare cancer and has a rapid growth with a short doubling time of cancer cells.

We have reported several clusters of IBC [40,41] and noted the possible contribution of chemical and infectious triggers. The term ``triggering agent'' is used because it indicates an environmental agent that has exposed a community but there are apparently other factors, such as genetic susceptibility, that determine which people in the exposed community get the disease.

3.3.4.1. Chemical exposure

In one IBC cluster we reported [41], there was a confluence of genetics and environmental exposures that could have caused the IBC that developed within a 10-month period in three women who worked in an office of 24 long-term employees. The risk factors identified in the women included office location (a part of the office with poor air and water quality), exposures to herbicides and pesticides, hormone replacement therapy at the time of diagnosis, family history of breast cancer, and obesity. None of the three had all risk factors and conclusions cannot be drawn from three individuals, but each woman had some of the risk factors.

3.3.4.2. Infectious agents

Some evidence for infection as an environmental triggering agent comes from our study showing that the onset of IBC in two patient populations was less per month in winter months than during the rest of the year [42].

3.3.5. Non-acute exposures

3.3.5.1. HMTV virus

One possible non-acute risk factor that has been considered for IBC is a human virus similar to the mouse mammary tumor virus (MMTV), which a group at Mount Sinai Medical Center calls the human mammary tumor virus (HMTV). Support for the presence of this virus comes from five studies from three different laboratories [43--47] and studies of human breast milk suggest that as in mice, breast feeding is the usual mode of transmission [44]. The possible association with IBC comes from the observation of a higher prevalence of the mouse-like gp52 antigen in Tunisian breast cancer patients compared to U.S. breast cancer patients [45] and the high prevalence of HMTV-sequences in U.S. IBC patients compared to U.S. non-IBC breast cancer patients [46,47]. It is not known whether the increased replication of the suspected virus is a result of more rapid tumor growth or the cause.

3.3.5.2. Alcohol

In the study of heredity and selected environmental factors reported by Moslehi et al. [37], a regular alcohol consumption (greater than one drink per day) increased the risk for IBC almost two times when IBC patients were compared to the WHI controls and compared to controls from the University of Toronto.

3.3.6. Familial/household aggregation in unrelated individuals

Familial aggregation is usually due to genetic predisposition but can occur from environmental exposures as well. There have been anecdotal reports of IBC occurring in a husband and wife but we have found none in the literature. A case of familial IBC reported to the IBC Registry that is clearly unrelated to genetics involves a husband and wife with IBC, the husband being diagnosed with IBC 5 years after the onset in his wife.

4. Survival studies

4.1. Response to chemotherapy

An epidemiologic study relevant to survival showed in a prospective cohort study from 155 patients enrolled in the IBC Registry, that immediate response to neoadjuvant chemotherapy, the standard form of primary treatment for IBC, was the most important determinant for prolonged survival [48]. The patients in this study received combination chemotherapy, usually with three drugs including Adriamycin and cyclophosphamide. Chemotherapy response was significantly associated with observed survival; women not responding to chemotherapy had a significantly higher risk of death compared with women with complete or partial response. These data, confirmatory of a previous clinical trial at MD Anderson [49], are important as they represent the population of patients seen in private practice outside of clinical trials.

4.2. Speed of diagnosis

The barriers to rapid diagnosis and treatment have been shown to result from delays by either patients or providers [48,50,51]. Delay in diagnosis, although not showing as great an effect on survival as response to chemotherapy, did show a trend with a progressive decrease in survival associated with longer delays in diagnosis [48,50]. The major obstacle to early diagnosis in young women was the resemblance of IBC to acute infection. Common comments from physicians include ``breast cancer doesn't hurt'' and ``you are too young to get breast cancer'' [51].

4.3. Socioeconomic status

Socioeconomics may also play an important role in determining outcome. In a study of 7624 cases of invasive carcinoma identified by cancer registries in seven states, Denu et al. [27] observed that in the 170 IBC patients, worse outcomes were seen in patients with Medicaid, patients from urban areas, and patients from areas of higher poverty and lower education. The IBC Registry data noted that factors contributing to delayed diagnosis included lack of medical insurance and attention to other priorities, including co-morbidities that obscured the seriousness of the problem. Denu et al. also noted the importance of co-morbidities, which were significantly higher in number and severity than in non-IBC breast cancer patients. Significantly more comorbidities were also seen in African-American patients than Caucasian patients in a smaller study from Emory University in Atlanta (30 versus 25 IBC patients, respectively), and although there was no difference in treatment adherence, Caucasians experienced a higher 3-year distant metastasis-free survival (60% versus 40% respectively, p = 0.19) as well as observed 5 year survival (p = 0.09) [52]. The small numbers in these studies make it difficult to determine the relative impact of purely socioeconomic from other racial/ethnic factors as has been done in other diseases.

5. Summary

While IBC continues to be an aggressive malignancy with high morbidity and mortality, the picture has improved dramatically in recent years with improvements in early recognition and treatment. Continued research in various areas, including epidemiology, may not only uncover more opportunities to control IBC but may also uncover factors that determine the aggressiveness of the cancer, one of the major determinants of survival.

Footnotes

Funding

Grant #DAMD 17-01-0244 with the Department of Defense.

Conflict of interest

The authors declare that there are no conflicts of interests.

References

Levine

, Zolfaghari

, Young

, Hafi

, Cannon

, Ganesan

, Veneroso

, Brem

, Sherman

, What is inflammatory breast cancer? Revisiting the case definition, Cancers (Basel) , 2: 143–152, 2010. doi:10.3390/cancers2010143.

Dawood

, Merajver

, Viens

, Vermeulen

, Swain

, Buchholz

, Dirix

, Levine

, Lucci

, Krishnamurthy

, Robertson

, Woodward

, Yang

, Ueno

, Cristofanilli

, International expert panel on inflammatory breast cancer: Consensus statement for standardized diagnosis and treatment, Ann Oncol , 22: 515–523, 2011. doi:10.1093/annonc/mdq345.

Hance

, Anderson

, Devesa

, Young

, Levine

, Trends in inflammatory breast carcinoma incidence and survival: The surveillance, epidemiology, and end results program at the National Cancer Institute, J Natl Cancer Inst , 97: 966–975, 2005. doi:10.1093/jnci/dji172.

Levine

, Portera

, Hoffman

, Yang

, Takikita

, Duong

, Hewitt

, Swain

, Evaluation of lymphangiogenic factors, vascular endothelial growth factor D and E-cadherin in distinguishing inflammatory from locally advanced breast cancer, Clin Breast Cancer , 12: 232–239, 2012. doi:10.1016/j.clbc.2012.04.005.

Raghav

, French

, Ueno

, Lei

, Krishnamurthy

, Reuben

, Valero

, Ibrahim

, Inflammatory breast cancer: A distinct clinicopathological entity transcending histological distinction, PLoS One , 11: 1–11, 2016. doi:10.1371/journal.pone.0145534.

Denoix

. Treatment of Malignant Breast Cancer. Recent Results in Cancer Research . . (Berlin), Springer/Verlag; 92–94. 1970.

Tabbane

, Muenz

, Jaziri

, Cammoun

, Belhassen

, Mourali

, Clinical and prognostic features of a rapidly progressing breast cancer in Tunisia, Cancer , 40: 376–382, 1977. doi:10.1002/1097-0142(197707)40:1<376::aid-cncr2820400153>3.0.co;2-y.

Taylor

, Meltzer

, “Inflammatory carcinoma” of the breast, Am J Cancer , 33: 33–49, 1938.

Saltzstein

, Clinically occult inflammatory carcinoma of the breast, Cancer , 34: 382–388, 1974. doi:10.1002/1097-0142(197408)34:2<382::aid-cncr2820340223>3.0.co;2-m.

10.

Levine

, Tabbane

, Muenz

, Kamaraju

, Das

, Polivy

, Belhassen

, Scholl

, Bekesi

, Mourali

, Hormone receptors in rapidly progressing breast cancer, Cancer , 54: 3012–3016, 1984. 10.1002/1097-0142(19841215)54:12<3012::aid-cncr2820541232>3.0.co;2-r. PMID: 6498775.

11.

McCarthy

, Yang

, Linnoila

, Merino

, Hewitt

, Parr

, Paik

, Steinberg

, Hartmann

, Mourali

, Levine

, Swain

, Microvessel density, expression of estrogen receptor alpha, MIB-1, p53, and c-erbB-2 in inflammatory breast cancer, Clin Cancer Res , 8: 3857–3862, 2002.

12.

, Moe

, Daling

, Risk of mortality by histological type of breast cancer among women aged 50 to 79 years, Arch Intern Med , 163: 2149–2153, 2003. doi:10.1001/archinte.163.18.2149.

13.

Korkala

, De Vries

, Fridlyand

, Hwang

, Estep

, Chen

, Differentiation of lobular versus ductal breast carcinomas by expression microarray analysis, Cancer Res , 63: 7167–7175, 2003.

14.

Levine

, Hoffman

, MacNeil

, Hashmi

, Yang

, Hewitt

, Golen

, Swain

, Prognostic value of lymphocyte vascular density and e-cadherin in inflammatory breast cancer, J Cancer Therapy , 5: 1380–1387, 2014. doi:10.4236/jct.2014.514139.

15.

Jhaveri

, Teplinsky

, Silvera

, Valeta-Magara

, Arju

, Giashuddin

, Sarfraz

, Alexander

, Darvishian

, Levine

, Hashmi

, Zolfaghari

, Hoffman

, Singh

, Goldberg

, Hochman

, Formenti

, Esteva

, Moran

, Schneider

, Hyperactivated mTOR and JAK2/STAT3 pathways: Molecular drivers and potential therapeutic targets of inflammatory and invasive ductal breast cancers after neoadjuvant chemotherapy, Clin Breast Cancer , 16: 113–122e111, 2016. doi:10.1016/j.clbc.2015.11.006.

16.

Chang

, Parker

, Pham

, Buzdar

, Hursting

, Inflammatory breast carcinoma incidence and survival: The surveillance, epidemiology, and end results program of the National Cancer Institute, 1975–1992, Cancer , 82: 2366–2372, 1998.

17.

Anderson

, Schairer

, Chen

, Hance

, Levine

, Epidemiology of inflammatory breast cancer (IBC), Breast Dis , 22: 9–23, 2005. doi:10.3233/bd-2006-22103.

18.

Goldner

, Behrendt

, Schoellhammer

, Lee

, Chen

, Incidence of inflammatory breast cancer in women, 1992–2009, United States, Ann Surg Oncol , 21: 1267–1270, 2014. doi:10.1245/s10434-013-3439-y.

19.

Hirko

, Soliman

, Banerjee

, Ruterbusch

, Harford

, Chamberlain

, Graff

, Merajver

, Schwartz

, Characterizing inflammatory breast cancer among Arab Americans in the California, Detroit and New Jersey Surveillance, Epidemiology and End Results (SEER) registeries (1988–2008), Springerplus , 2: 3, 2013. doi:10.1186/2193-1801-2-3.

20.

Mourali

, Muenz

, Tabbane

, Belhassen

, Bahi

, Levine

, Epidemiologic features of rapidly progressing breast cancer in Tunisia, Cancer , 46: 2741–2746, 1980. 10.1002/1097-0142(19801215)46:12<2741::aid-cncr2820461234>3.0.co;2-w.

21.

Maalej

, Frikha

, Ben Salem

, Daoud

, Bouaouina

, Ben Abdallah

, Ben Romdhane

, Breast cancer in Tunisia: Clinical and epidemiological study, Bull Cancer , 86: 302–306, 1999.

22.

Maalej

, Hentati

, Messai

, Kochbati

, El May

, Mrad

, Romdhane

, Ben Abdallah

, Zouari

, Breast cancer in Tunisia in 2004: A comparative clinical and epidemiological study, Bull Cancer , 95: E59, 2008. doi:10.1684/bdc.2008.0584.

23.

Boussen

, Bouzaiene

, Ben Hassouna

, Gamoudi

, Benna

, Rahal

, Inflammatory breast cancer in Tunisia: Reassessment of incidence and clinicopathological features, Semin Oncol , 35: 17–24, 2008. doi:10.1053/j.seminoncol.2007.11.017.

24.

Boussen

, Bouzaiene

, Ben Hassouna

, Dhiab

, Khomsi

, Benna

, Gamoudi

, Mourali

, Hechiche

, Rahal

, Levine

, Inflammatory breast cancer in Tunisia: Epidemiological and clinical trends, Cancer , 116: 2730–2735, 2010. doi:10.1002/cncr.25175.

25.

Scott

, Mobley

, Il’yasova

, Geospatial analysis of inflammatory breast cancer and associated community characteristics in the United States, Int J Environ Res Public Health , 14: 404, 2017. doi:10.3390/ijerph14040404.

26.

Schlichting

, Soliman

, Schairer

, Banerjee

, Rozek

, Schottenfeld

, Harford

, Merajver

, Association of inflammatory and noninflammatory breast cancer with socioeconomic characteristics in the Surveillance, Epidemiology, and End Results database, 2000–2007, Cancer Epidemiol Biomarkers Prev , 21: 155–165, 2012. doi:10.1158/1055-9965.EPI-11-0833.

27.

Denu

, Hampton

, Currey

, Anderson

, Cress

, Fleming

, Lipscomb

, Wu

, Wilson

, Trentham-Dietz

, Racial and socioeconomic disparities are more pronounced in inflammatory breast cancer than other breast cancers, J Cancer Epidemiol,\space 7574946, 2017. doi:10.1155/2017/7574946.

28.

Schairer

, Li

, Frawley

, Graubard

, Wellman

, Buist

, Kerlikowske

, Onega

, Anderson

, Miglioretti

, Risk factors for inflammatory breast cancer and other invasive breast cancers, J Natl Cancer Inst , 105: 1373–1384, 2013. doi:10.1093/jnci/djt206.

29.

Lehrer

, Green

, Rosenzweig

, Affluence and breast cancer, Breast J , 22: 564–567, 2016. doi:10.1111/tbj.12630.

30.

Chang

, Buzdar

, Hursting

, Inflammatory breast cancer and body mass index, J Clin Oncol , 16: 3731–3735, 1998. doi:10.1200/JCO.1998.16.12.3731.

31.

Veneroso

, Siegel

, Levine

, Early age at first childbirth associated with advanced tumor grade in breast cancer, Cancer Detect Prev , 32: 215–223, 2008. doi:10.1016/j.cdp.2008.04.001.

32.

Atkinson

, El-Zein

, Valero

, Lucci

, Bevers

, Fouad

, Liao

, Ueno

, Woodward

, Brewster

, Epidemiological risk factors associated with inflammatory breast cancer subtypes, Cancer Causes Control , 27: 359–366, 2016. doi:10.1007/s10552-015-0712-3.

33.

, Arriagada

, Bahi

, Pfeiffer

, Cammoun

, Tabbane

, Rubino

, Are risk factors for breast cancer similar in women with inflammatory breast cancer and in those with non-inflammatory breast cancer? Breast , 15: 355–362, 2006. doi:10.1016/j.breast.2005.08.018.

34.

Mejri

, El Benna

, Rachdi

, Labidi

, Benna

, Daoud

, Hamdi

, Abdelhak

, Boussen

, Reproductive risk factors of inflammatory breast cancer according to luminal, HER2-overexpressing, and triple-negative subtypes: A case comparison study, Oncol Res Treat , 43: 6, 2020. doi:10.1159/000506691.

35.

Bernier

, Plu-Bureau

, Bossard

, Ayzac

, Thalabard

, Breastfeeding and risk of breast cancer: A metaanalysis of published studies, Hum Reprod Update , 6: 374–386, 2000. doi:10.1093/humupd/6.4.374.

36.

Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and breastfeeding: Collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50,302 women with breast cancer and 96,973 women without the disease, Lancet, 360: 187–195, 2002. doi:10.1016/S0140-6736(02)09454-0.

37.

Moslehi

, Freedman

, Zeinomar

, Veneroso

, Levine

, Importance of hereditary and selected environmental risk factors in the etiology of inflammatory breast cancer: A case-comparison study, BMC Cancer , 16: 334, 2016. doi:10.1186/s12885-016-2369-z.

38.

Levine

, Mourali

, Tabbane

, Loon

, Terasaki

, Tsang

, Bekesi

, Studies on the role of cellular immunity and genetics in the etiology of rapidly progressing breast cancer in Tunisia, Int J Cancer , 27: 611–615, 1981. doi:10.1002/ijc.2910270507.

39.

Benish

, Ben-Eliyahu

, Surgery as a double-edged sword: A clinically feasible approach to overcome the metastasis-promoting effects of surgery by blunting stress and prostaglandin responses, Cancers (Basel) , 2: 1929–1951, 2010. doi:10.3390/cancers2041929.

40.

Levine

, Hashmi

, Minaei

, Veneroso

, Inflammatory breast cancer clusters: A hypothesis, World J Clin Oncol , 5: 539–545, 2014. doi:10.5306/wjco.v5.i3.539.

41.

Duke

, Jahed

, Veneroso

, Da Roza

, Johnson

, Hoffman

, Barsky

, Levine

, A cluster of inflammatory breast cancer (IBC) in an office setting: Additional evidence of the importance of environmental factors in IBC etiology, Oncol Rep , 24: 1277–1284, 2010. doi:10.3892/or_00000983.

42.

Levine

, Liu

, Veneroso

, Hashmi

, Cristofanilli

, Seasonal variation in inflammatory breast cancer, Int J Virol Stud Res , 4: 17–21, 2016. doi:10.19070/2330-0027-160003.

43.

Ford

, Tran

, Deng

, Ta

, Rawlinson

, Lawson

, Mouse mammary tumor virus-like gene sequences in breast tumors of Australian and Vietnamese women, Clin Cancer Res , 9: 1118–1120, 2003.

44.

Nartey

, Moran

, Marin

, Arcaro

, Anderton

, Etkind

, Holland

, Melana

, Pogo

, Human Mammary Tumor Virus (HMTV) sequences in human milk, Infect Agent Cancer , 9: 20, 2014. doi:10.1186/1750-9378-9-20.

45.

Levine

, Mesa-Tejada

, Keydar

, Tabbane

, Spiegelman

, Mourali

, Increased incidence of mouse mammary tumor virus-related antigen in Tunisian patients with breast cancer, Int J Cancer , 33: 305–308, 1984. doi:10.1002/ijc.2910330305.

46.

Levine

, Pogo

, Klouj

, Coronel

, Woodson

, Melana

, Mourali

, Holland

, Increasing evidence for a human breast carcinoma virus with geographic differences, Cancer , 101: 721–726, 2004. doi:10.1002/cncr.20436.

47.

Pogo

, Holland

, Levine

, Human mammary tumor virus in inflammatory breast cancer, Cancer , 116: 2741–2744, 2010. doi:10.1002/cncr.25179.

48.

Hoffman

, Khan

, Ajmera

, Zolfaghari

, Schenfeld

, Levine

, Initial response to chemotherapy, not delay in diagnosis, predicts overall survival in inflammatory breast cancer cases, Am J Clin Oncol , 37: 315–321, 2012. doi:10.1097/COC.0b013e318271b34b.

49.

Cristofanilli

, Buzdar

, Hortobagyi

, Update on the management of inflammatory breast cancer, Oncologist , 8: 141–148, 2003. doi:10.1634/theoncologist.8-2-141.

50.

Hashmi

, Zolfaghari

, Levine

, Does secondary inflammatory breast cancer represent post-surgical metastatic disease? Cancers (Basel) , 4: 156–164, 2012. doi:10.3390/cancers4010156.

51.

Hashmi

, Levine

, Barriers to early diagnosis of inflammatory cancer, Insights of Breast Cancer , 1: 1, 2017.

52.

Andic

, Godette

, O’Regan

, Zelnak

, Liu

, Rizzo

, Gabram

, Torres

, Treatment adherence and outcomes in women with inflammatory breast cancer: Does race matter? Cancer , 117: 5485–5492, 2011. doi:10.1002/cncr.26187.

Recent advances in the epidemiology of inflammatory breast cancer

Abstract

Keywords

1. Background

2. Resolving the case definition

3.1. Socioeconomic factors

3.2. Hormonal risk factors

3.2.1. Early age at birth of first child

3.2.2. Obesity

3.2.3. Prolonged breast feeding

3.2.4. Oral contraceptives (OC)

3.3. Non-hormonal risk factors

3.3.1. Delay in diagnosis

3.3.2. Genetics

3.3.3. Local trauma to the breast

3.3.4. Acute exposures

3.3.4.1. Chemical exposure

3.3.4.2. Infectious agents

3.3.5. Non-acute exposures

3.3.5.1. HMTV virus

3.3.5.2. Alcohol

3.3.6. Familial/household aggregation in unrelated individuals

4. Survival studies

4.1. Response to chemotherapy

4.2. Speed of diagnosis

4.3. Socioeconomic status

5. Summary

Footnotes

Funding

Conflict of interest

References