Angiogenesis, Lymphangiogenesis, Growth Pattern and Tumor Emboli in Inflammatory Breast Cancer: a Review of the Current Knowledge

Review oncology

Angiogenesis, lymphangiogenesis, growth pattern and tumor emboli in inflammatory breast cancer: a review of the current knowledge

Authors P. Vermeulen, K. van Golen, L. Dirix

Key words Inflammatory breast cancer; angiogenesis; lymphangiogenesis; tumor emboli

Summary features of the biology of inflammatory breast This review is based on a presentation at the cancer (IBC): angiogenesis, lymphangiogenesis “First International Inflammatory Breast Cancer and the formation of tumor emboli. Information Conference”, MD Anderson Cancer Center, Dec derived from animal models of IBC as well as 2008, Houston, TX, and aims at providing the from translational studies using tissue samples of reader with a comprehensive summary of the patients with IBC will be discussed literature related to 3 important and interrelated (BJMO 2009;Vol 3;2:49-56)

Introduction differences in diagnostic criteria as well as lack of Inflammatory breast cancer (IBC) is a form of lo- attention to the clinical symptoms, sometimes only cally advanced breast cancer characterized by a spe- deductable from anamnesis. cific set of symptoms at diagnosis. The diagnosis is clinical (AJCC T4d) and is supported by pathologi- cal findings that will be discussed in this review. Angiogenesis in animal models of IBC Critical to the diagnosis is the rapid onset of symp- The first, and by now well-known animal model toms and accelerated local growth, usually within 3 for IBC is the MARY-X mouse model, a stable se- to 6 months. Symptoms are swelling of the breast rial transplantable xenograft derived from a 45-year- accompanied by skin edema, also described as old female patient with IBC.4 At least during initial “peau d’orange”, and erythema in more than 50% transplant generations, there is exclusive intravas- of the breast. There may also be exquisite pain. At cular growth, both in the primary subcutaneous presentation, the large majority of the patients has tumor and in the lung metastases, implying that lymph node metastases and distant metastases are MARY-X is a model suitable to study both vascu- often present. Considering cell-of-origin subtypes, larisation of IBC and embolus formation. Com- most cases of IBC belong to the basal, luminal-B pared to non-IBC xenografts MDA-MB-231 and or HER2-overexpressing subtype.1 Outcome is un- MDA-MB-468, MARY-X has the largest stromal favorable, but long survivors are reported after ex- compartment, which is in line with histomorpho- tensive multimodality treatment in which neo-ad- metrical analyses showing low tumor to stroma ra- juvant chemotherapy is an essential element.2 About tio’s in IBC patients.5 Although the initial hypothe- 2 to 5% of all invasive breast carcinomas are IBCs sis was that stimulation of angiogenesis occurred and the incidence seems to be increasing.3 The true followed by vascular homing, more recent studies incidence is hard to determine because of regional showed stem cell or tumor cell vasculogenesis (also

49 v o l . 3 i s s u e 2 - 2 0 0 9 BELGIAN JOURNAL OF MEDICAL ONCOLOGY related receptors were not upregulated. A third animal model of IBC is the naturally occur- ring canine inflammatory mammary cancer.13,14 In one third of cases, endothelial-like neoplastic cells, in- dicative of vasculogenic mimicry, are observed in this model.15 These tumors display elevated levels of both COX-2 mRNA and COX-2 protein. Vascular channels formed by carcinoma cells have also been described in COX-2 overexpressing non-IBC in patients.16

Histomorphometrical studies of angiogenesis in human IBC samples Microvessel density (MVD) was measured us- Figure 1. Principal component analysis using a list of 18,182 ing the Weidner method in a group of 22 informative genes was performed to obtain global views of the patients with non-IBC, of which 86% with high variation in gene expression among the different breast cancer histological grade, and 45 patients of IBC from samples (A). Inflammatory breast cancer samples are colour- Tunisia with PEV-scores 2 to 3 (score 3 correspon- coded red, non-IBC T1 or T2 tumors are colour-coded blue ding to AJCC T4d and score 2 corresponding to and non-IBC T3 or T4 tumors are colour-coded green. The first a clinical history of IBC but with erythema limited principal component is represented by the X-axis, whereas the to less than 50% of the breast). Essentially, blood Y-axis represents the second principal component. Inflamma- vessels were counted in areas of elevated density tory breast cancer and non-IBC samples are separated along spotted at low magnification, and with antibodies the first principal component. Unsupervised hierarchical com- directed at CD31.17 Whether graded or expressed plete linkage clustering was performed using 250 informative as a continuous variable, MVD was significantly genes having the greatest standard deviation. higher in the IBC samples: 51% of IBC samples versus only 14% of non-IBC samples belonged to the mod- erate-to-high MVD category (p= 0.02). The median called vasculogenic mimicry) encircling solid tumor MVD was 25.5 in IBC samples versus 6.5 in non-IBC cell nests.6 This passive formation of tumor emboli samples (p= 0.009). In a more elaborate study the has also been described in a non-IBC murine breast relative vascular area was measured with the Chalk- carcinoma model, MCH66, in which blood vessel ley point-overlap method (CD34) and in addition, the envelope around tumor cell nests.7 A prerequisite for endothelial cell proliferation fraction (CD34-PCNA), passive embolus formation appears to be the strong blood vessel maturity (CD34-alpha smooth muscle homotypic cohesion of tumor cells forming large actin), fibrin deposition (T2G1) and hypoxia (CAIX) aggregates. In IBC, this is due to overexpressed and were measured after (double-)immunostaining in functional E-cadherin.8-10 a group of 35 consecutive patients with IBC (AJCC The second animal model of IBC is WIBC-9, again T4d) and 104 patients with T1-T2 N0 non-IBC.18 The a stable serial transplantable xenograft mouse mod- choice of a non-stage-matched control group is sup- el, derived from a 50-year-old female patient with ported by gene expression profiling results and princi- IBC.11,12 In this model, vascularisation is the result pal-component-analysis showing a greater resemblance of peripheral angiogenesis with high CD31-positive of IBC with T1-T2 breast carcinomas than with local- vessel density and central vasculogenic mimicry. ly-advanced non-IBC (Figure 1).19 The strongest argu- When isolating the tumor cells by laser-assisted ment for increased angiogenesis in IBC is the elevated microdissection and/or by applying species-specific fraction of proliferating endothelial cells in this study, antibodies for immunohistochemistry or probes 15% versus 9% in non-IBC (p= 0.05).18 Chalkley for PCR, these cells were shown to express vascu- relative vascular area was also higher in IBC. About lar markers such as alpha-v-beta-3 integrins, the 90% of blood vessels were immature, i.e. lacking VEGF-receptors flt1 and KDR, and the angiopoi- pericyte coverage, in both IBC and non-IBC cases. etin-receptor tie-2. An elevated expression of hu- About half of the tumors in both groups expressed man and murine angiogenic factors (VEGF, bFGF, CAIX. Abundant stromal fibrin deposition was ANG1 and IL8) was observed in tumor tissue and present in a higher fractions of IBC cases compared to serum. However, lymphangiogenic factors and their non-IBC cases, 26% versus 8% respectively (p= 0.02).

BELGIAN JOURNAL OF MEDICAL ONCOLOGY v o l . 3 i s s u e 2 - 2 0 0 9 50 Review oncology

Molecular studies of angiogenesis in human Anti-VEGF treatment in patients with IBC samples of IBC Given the elevated angiogenic activity in IBC, mRNA expression of angiogenic factors and their re- clinical trials based on a strategy to inhibit the ceptors has been quantified by real-time RT-PCR in VEGF signal have been developed. In a first study, 16 IBC cases and 20 non-IBC cases (stage-matched 18 patients were treated with a combination of and non-stage-matched).5 Relative gene expression a small-molecule tyrosine kinase inhibitor of levels of flt1, KDR, ANG1, TIE1, TIE2, COX2 and VEGFR-2 and conventional chemotherapy (doxo- bFGF were significantly higher in IBC samples. There rubicin).22 The clinical response rate was as high was no difference in expression levels of VEGF and as 90% and plasma VEGF levels increased with ANG2. None of the factors in this study were elevated treatment. Core biopsies were taken before treat- in the non-IBC group. Higher relative gene expression ment, after 2 cycles and at the time of mastectomy: of COX2 and bFGF in IBC was confirmed by im- MVD (“Weidner” method, Factor VIII-related an- munohistochemistry on a tissue microarray (TMA). tigen as endothelial cell marker) was lower after A positive correlation was observed for gene expres- treatment, with the percentage of reduction sig- sion levels vascular endothelium markers (flt1, KDR, nificantly related to overall survival. In a second ANG1, TIE1, TIE2; p-values < 0.01). The relative gene study, 20 patients with IBC, and one patient with expression of VEGF was positively associated with the LABC, were treated with bevacizumab, a human- tumor to stroma-ratio (r= 0.4; p= 0.04) while the rela- ized monoclonal antibody directed at all isoforms tive gene expression of bFGF was negatively associated of VEGF-A, alone during cycle 1 and combined with this ratio (r= -0.5; p= 0.006) suggesting a differ- with doxorubicin and docetacel during cycles 2 ence in cell types expressing these factors in IBC. to 7.23-24 Fourteen patients had a clinical partial Genome-wide expression profiling studies contain in- response, 5 patients had stable disease and 2 pa- formation regarding angiogenesis in IBC. Comparing tients had progressive disease. Tumor biospies after 37 patients with IBC, defined as T4d and/or dermal the first cycle showed effects of bevacizumab on lymphatic invasion, with 44 patients with non-IBC both tumor cells and endothelial cells. Automated (T1-2: n= 30; T3-4c: n= 14) applying home-made 8K quantification of the expression levels of phosphor- cDNA microarrays, unsupervised hierarchical cluster- ylated VEGFR-2, with two different phosphoryla- ing resulted in a group enriched with IBC samples.20 tion sites assessed by IHC, demonstrated a 67% re- In this group, genes belonging to a vascular cluster, i.e. duction in tumor cells (p= 0.004), combined with genes expressed by human umbilical vein endothelial a highly significant increase in tumor cell apopto- cells were overexpressed. Supervised analysis resulted sis and no change in tumor cell proliferation. There in an IBC-related gene expression signature with 85% was a trend towards decrease of the endothelial cell accuracy containing overexpressed genes upstream of proliferation fraction, measured after double im- the MAPK pathway, related to RhoC-induced ang- munostaining with antibodies directed at CD31 iogenesis, and, overexpression of ARNT (beta-subu- and Ki67 (p= 0.06) and a significant decrease in nit of hypoxia-inducible factor 1; HIF1). In another the level of CD31 expression in endothelial cells study, real-time quantitative RT-PCR of more than (41%; p= 0.007). No changes were observed in 500 genes known to be associated with angiogenesis MVD, VEGF-A expression and total VEGFR-2 and inflammation was performed in samples of 36 expression. This study also aimed to determine patients with IBC (AJCC T4d) and 22 patients with predictive factors for response to bevacizumab in non-IBC LABC.21 This resulted in the identifica- baseline biopsies. Protein expression levels by IHC tion of 27 genes significantly upregulated in the IBC of CD31 in the vasculature, of PDGFR-beta in the group. One third of these genes was related to angio- vasculature and in tumor cells, and of VEGF-A in genesis (VEGF-A, TBXA2R, PTGS2/COX2, THBD/ the tumor cells were higher in responders (respec- thrombomodulin, ANGPT2/angiopoietin 2, CCL3, tive p values: 0.0004, 0.01 and 0.04). These IHC CCL5, CCR5, IL-6 (HIF1A (p=0.06)). There was no data were confirmed by gene expression analysis difference in expression of VEGF2, VEGF3, VEGF4, with Gene Ontology (GO) categories containing VEGFR1, VEGFR2 and VEGFR3 and IL-8. In con- genes related to “VEGFR activity”, e.g. PDGFR- trast to earlier reports, no difference in expression of alpha and PDGFR–beta, and genes related to “cell WISP3, RhoC and E-cadherin was found either. motility, locomotion and localization”, e.g. CD31.

51 v o l . 3 i s s u e 2 - 2 0 0 9 BELGIAN JOURNAL OF MEDICAL ONCOLOGY Lymphangiogenesis in animal models of IBC Tumor emboli and growth pattern in IBC Lymphangiogenesis has not been explicitely investi- Although not a prerequisite for diagnosis when gated in the animal models of IBC. In studies of the adopting the AJCC T4d criteria, numerous emboli, MARY-X model, Factor VIII-related antigen is used both in the tumor parenchyma and in the dermis, as a marker for vascularisation.4 This is one of the are often described in the pathology report. They most specific markers for blood vessel endothelium tend to be larger than in non-IBC and are believed with a mutually exclusive vascular expression pat- to be responsible for the skin edema (no clear evi- tern when compared to the lymph vessel endothe- dence is available). In non-IBC, emboli are mainly lium marker D2-40.25 In studies using the WIBC-9 located in lymph vessels. This has not been studied xenograft, CD31 is used as a vascular marker. CD31 thoroughly in IBC. is also expressed on lymph vessels, although weaker than on blood vessels.11 As a cosequence, part of the Lessons from animal models of IBC and confron- vessels described in this model might have been lym- tation with human IBC samples phatic. In this model, mRNA levels of genes coding In the MARY-X model, intravascular growth of IBC for lymphangiogenic factors and receptors (human coincides with strong circumferential expression of VEGF-C, human VEGF-D, and murine VEGFR-3) E-cadherin in tumor cells, and, an intact E-cadher- were lower compared to non-IBC xenografts.12 in-catenin axis.9 This “conserved” E-cadherin expression predicts the IBC phenotype with an odds Histomorphometrical studies of lymphangiogen- ratio of 5.6 in a study containing 83 IBC samples. A esis in human IBC samples tissue microarray study with 34 IBC samples and 41 When comparing 29 samples of patients with IBC non-IBC samples has confirmed E-cadherin overex- with 56 samples of patients with non-IBC by dou- pression in IBC.27 In a whole section immunohisto- ble-immunostaining of tumor tissue sections with chemical study focussing on tumor emboli in IBC, antibodies directed against D2-40 and Ki67 and E-cadherin expression was equally strong in the computer-aided histomorphometrical analysis, the emboli compared to the invasive parenchymal com- peri-tumoral fraction of proliferating lymphatic ponent in 27/35 samples, and stronger in 6/35 sam- endothelial cells was 3-fold higher in IBC versus ples.18 Expression of E-cadherin was never weaker non-IBC (p = 0.005).25 This strongly suggests on- in the intravascular emboli. Two IBC cases with going lymphangiogenesis, at a higher level in IBC. lobular carcinoma were E-cadherin-negative. Even Intra-tumoral lymph vessel area and lymph vessel in non-IBC, increase of the E-cadherin expression perimeter were also larger in IBC samples (respec- has been observed in intravascular tumor emboli, tive p-values: 0.01 and 0.07). The non-IBC control suggesting a biological function of E-cadherin relat- group consisted of patients with T1 to T4a breast ed to embolus formation: strong homotypical inter- carcinoma’s. actions might be necessary for stroma-independent intravascular growth.28 Molecular studies of lymphangiogenesis in human The heterotypical interaction with endothelial cells IBC samples is weak in the emboli in the MARY-X model due mRNA levels of the genes coding for the lymphang- to the lack of functional, although overexpressed, iogenic factors VEGF-C VEGF-D, their receptor MUC1.10 The absence of sialyl groups (Lewis-X and VEGFR-3, and the lymphatic endothelial markers –A) is responsible for this dysfunction.9 In human PROX-1 and LYVE-1 were significantly (all p-values IBC samples, overexpression of glycosylated cyto- < 0.02) elevated in 16 IBC samples compared to 20 plasmic MUC1, but not of sialated MUC1 has been non-IBC samples.20 In contrast with these results, a described, confirming the animal model study.27 comparison of 36 IBC samples with 22 non-IBC lo- Passive embolus formation in the animal models, as cally advanced breast cancer samples uing real-time described earlier in this review, has not been solidly RT-PCR did not show overexpression of VEGF-C, confirmed in human samples of IBC or non-IBC. VEGF-D and VEGFR-3.21 However, in this study In the MARY-X model, tumor cells express mark- RhoC, WISP-3 and E-cadherin, which were repeat- ers that can be related to stem cell behaviour.29 edly shown to be differentially expressed, were not Furthermore, The IBC cell line MDA-IBC-1, de- differentially expressed either.8,26 veloped by the MD Anderson team, also displays

BELGIAN JOURNAL OF MEDICAL ONCOLOGY v o l . 3 i s s u e 2 - 2 0 0 9 52 Review oncology

stem cell characteristics.30 A small study comparing mass. This suggests that one mechanism for IBC 25 human IBC samples with 25 non-IBC samples for fast local growth might be the use of blood and suggests that an enrichment of tumor cells with lymph vessels as a route of lower resistance. stem cell characteristics takes place in emboli, cre- Tumor emboli in IBC have been compared to blas- ating a stem cell niche.29 More IBC samples than tocysts, because of the high content of cells with non-IBC samples showed expression of CD133, stem cell features in both conditions.29 Another notch-3 receptor intracellular domain and ALDH1 condition that resembles the intravascular emboli of (70 to 90% versus 0 to 40%, respectively) both in IBC is ductal carcinoma in situ (DCIS). Hypoxia the parenchymal compartment and in the embolus and necrosis are indeed present in IBC emboli and compartment. A paper by Van Laere and collaegues centrally in high-grade DCIS (comedo-necrosis). discusses evidence for stem cell behaviour of IBC This creates a stem cell-friendly environment: in tumor cells gathered through genome-wide gene ex- DCIS tumor-initiating cells with stem cell features pression profiling.31 have been shown to be present in large numbers.33 Intravascular growth of emboli, as clearly present in Histomorphometrical analysis of tumor emboli the MARY-X model and probably also in the human and growth pattern in IBC situtaion, mimicks what is called “extensive intra- Histological features of the local growth of IBC in ductal carcinoma” in which large parts of the breast the breast gland and of the tumor emboli are des- are occupied by DCIS with only small and scattered cribed in a study comparing 35 consecutive patients invasive areas. Both IBC emboli and DCIS are ex- with IBC with 104 patients with T1-T2 N0 breast amples of stroma-independent growth, limited, re- carcinomas.18 Necrosis was not a prominent charac- spectively, by an endothelial cell layer or by a basal teristic of the parenchymal component of IBC, with membrane with an intact or incomplete myoepithe- only 7 cases with small necrotic foci. None of the lial cell layer. In both conditions this might be based IBC tumors contained a fibrotic focus (FF) (a cen- on the strong cohesion of the cancer cells by elevated tral scar-like area that has replaced or is replacing E-cadherin expression and diminished expression of necrosis) in contrast to 53% of the non-IBC sam- carbohydrates responsible for heterotypic interac- ples. Hypoxia, detected by immunohistochemical tions.34-35 demonstration of CA-IX expression, was present in a comparable fraction of IBC samples and non- IBC samples (46% and 55%, respectively). All this Conclusions is accordant with the smaller tumor to stroma ra- In animal models of IBC (and non-IBC) evidence tio observed in IBC versus non-IBC (in about 50% is in favor of tumor vascularisation and embolus of IBC cases areas with a tumor to stroma ratio of formation being the consequence of vasculogenesis less than 10% [16% of non-IBC’s; p< 0.05]).5 Clini- with endothelial (or endothelial-like) cells originat- cal examination reveals diffuse involvement of the ing from tumor cells (“vasculogenic mimicry”) or breast, often without a palpable nodus (although by bone marrow-derived stem cells. The vascular chan- radiology a main tumor mass can be demonstrated nels formed in this way connect to capillaries origi- in nearly all (98%) cases).32 Histologically, IBC has nating from peripheral angiogenesis. Histomor- large invasive carcinoma-free areas that alternate phometrical and molecular studies of human IBC with invasive carcinoma growing in a diffuse rather samples provide evidence of increased angiogenesis, than in a nodular fashion. and lymphangiogenesis. In IBC this is mainly based Almost all IBC tumors contained lymphovascular on endothelial cell proliferation fraction assessment. emboli (34/35 cases), and, in 26% of IBC cases with It is less clear whether vasculogenesis is really in- emboli, the tumor cells in the emboli expressed CA volved. Angiogenesis has been shown to be a valu- IX. In 23% of the IBC cases with emboli, necrosis able target for the treatment of patients with IBC was present in the emboli. Although not investigat- in 2 major clinical trials with a dual effect on both ed in this study, necrosis in lymphovascular emboli tumor and endothelial cells. is a very rare observation in non-IBC, and probably The growth pattern of IBC has implications for relates to the larger size of the emboli in IBC. A pe- molecular analyses based on tissue aggregates. culiar hallmark of the growth pattern of IBC is the The implications of the diffuse, as opposed to a colocalization of tumor emboli and small islands of nodular growth fashion are a low tumor to stroma invasive carcinoma at a distance of the main tumor ratio in IBC. This for instance leads to gene expres-

53 v o l . 3 i s s u e 2 - 2 0 0 9 BELGIAN JOURNAL OF MEDICAL ONCOLOGY Key messages for clinical practice

1. Based on animal model data and histomorphometrical and molecular studies of tissue samples of patients with IBC, it is firmly established that this breast cancer phenotype has more ongoing angiogenesis and lymphangiogenesis than non-IBC.

2. The former observations make IBC a good model to study anti-angiogenic and vascular target- ing agents. Clinical trials in this context have been conducted, with promising results.

3. Genome-wide expression profiling studies confirm the RhoC-induced angiogenesis as observed in cell line and animal model studies.

4. The growth pattern of IBC is different from non-IBC, with numerous and large intravascular emboli, a higher stroma content and diffuse spread with large tumor-free skip areas.

5. This topographical heterogeneity of IBC has an impact on molecular studies, especially genome-wide gene expression profiling, when whole tissue samples are used: a selective approach of the different tumor components is necessary, e.g. by laser-assisted microdissection.

6. There is an obvious need for international collaboration in order to enlarge the sample size of the translational studies and to achieve a consensus on the diagnostic criteria of IBC: important steps in this direction were taken at the “First International Inflammatory Breast Cancer Conference”, MD Anderson Cancer Center, Dec 2008, Houston, TX.

sion profiles reflecting the cancer cells, the host re- of inflammatory breast cancer. Semin Oncol 2008;35:64-71. active stroma and the intravascular emboli without 3. Hance KW, Anderson WF, Devesa SS, Young HA, the possibility to recover this component-specific in- Levine PH. Trends in inflammatory breast carcinoma incidence formation. A recent study taking the distinct com- and survival: the surveillance, epidemiology, and end results ponents into account seem to confirm this view.36 program at the National Cancer Institute. J Natl Cancer Inst Tumor emboli which are usually large, often con- 2005;97:966-75. tain necrosis and are present in high numbers in 4. Alpaugh ML, Tomlinson JS, Shao Z-M, Barsky SH. A Novel IBC, might be involved in accelerated local growth, Human Xenograft Model of Inflammatory Breast Cancer. Can- in addition to epithelial-mesenchymal transition cer Res 1999;59:5079-84. and tumor cell motility. 5. Van der Auwera I, Van Laere S, Van den Eynden G, et al. In- Relatively small studies are presented throughout creased angiogenesis and lymphangiogenesis in inflammatory this review and this obviously leads to loss of infor- versus noninflammatory breast cancer by real-time reverse mation that can be solved by more intense interna- transcriptase-PCR gene expression quantification. Clin Cancer tional collaboration and the installation of an inter- Res 2004;10:7965-71. national IBC register and related tissue repository, 6. Barsky S.H. et al. Proc Amer Assoc Cancer Res 2005. as initiated at the First International Inflammatory 7. Sugino T, Kusakabe T, Hoshi N, et al. An Invasion-Inde- Breast Cancer Conference in Houston last year. pendent Pathway of Blood-Borne Metastasis. A New Murine Mammary Tumor Model. Am J Pathol 2002;160:1973-80. 8. Kleer CG, van Golen KL, Braun T, Merajver SD. Persistent References E-cadherin expression in inflammatory breast cancer. Mod 1. Van Laere SJ, Van den Eynden GG, Van der Auwera I, et al. Pathol 2001;14:458-64. Identification of cell-of-origin breast tumor subtypes in in- 9. Alpaugh ML, Tomlinson JS, Ye Y, Barsky SH. Relationship of flammatory breast cancer by gene expression profiling. Breast Sialyl-Lewisx/a Underexpression and E-Cadherin Overexpres- Cancer Res Treat 2006;95:234-55. sion in the Lymphovascular Embolus of Inflammatory Breast 2. Dawood S, Ueno NT, Cristofanilli M. The medical treatment Carcinoma. Am J Pathol 2002;161:619-28.

BELGIAN JOURNAL OF MEDICAL ONCOLOGY v o l . 3 i s s u e 2 - 2 0 0 9 54 Review oncology

10. Alpaugh ML, Tomlinson JS, Kasraeian S, Barsky SH. Co- breast cancer and prediction of response to chemotherapy. operative role of E-cadherin and sialyl-Lewis X/A-deficient Cancer Res 2004;64:8558-65. MUC1 in the passive dissemination of tumor emboli in inflam- 21. Bieche I, Lerebours F, Tozlu S, et al. Molecular profiling of matory breast carcinoma. Oncogene 2002;21:3631-43. inflammatory breast cancer: identification of a poor-prognosis 11. Shirakawa K, Tsuda H, Heike Y, et al. Absence of endothelial gene expression signature. Clin Cancer Res 2004;10:6789-95. cells, central necrosis, and fibrosis are associated with aggres- 22. Overmoyer B, Fu P, Hoppel C, et al. Inflammatory breast sive inflammatory breast cancer. Cancer Res 2001;61:445-51. cancer as a model disease to study tumor angiogenesis: re- 12. Shirakawa K, Kobayashi H, Sobajima J, et al. Vasculogenic sults of a phase 1B trial of combination SU5416 and doxorubi- mimicry and its hemodynamics of an inflammatory breast cin. Clin Cancer Res 2007;13:5862-8. cancer xenograft model. Breast Cancer Res 2003;5:136-39. 23. Wedam SB, Low JA, Yang SX, et al. Antiangiogenic and an- 13. Pena L, Perez-Alenza MD, Rodriguez-Bertos A, Nieto A. titumor effects of bevacizumab in patients with inflammatory Canine inflammatory mammary carcinoma: histopathology, and locally advanced breast cancer. J Clin Oncol 2006;24:769-77. immunohistochemistry and clinical implications of 21 cases. 24. Yang SX, Steinberg SM, Nguyen D, Wu TD, Modrusan Z, Breast Cancer Res Treat 2003;78:141-8. Swain SM. Gene expression profile and angiogenic markers 14. Queiroga FL, Perez-Alenza MD, Silvan G, Pena L, Lopes C, correlate with response to neoadjuvant bevacizumab followed Illera JC. Cox-2 levels in canine mammary tumors, includ- by bevacizumab plus chemotherapy in breast cancer. Clin ing inflammatory mammary carcinoma: clinicopathologi- Cancer Res 2008;14:5893-9. cal features and prognostic significance. Anticancer Res 25. Van der Auwera I, Van den Eynden GG, Colpaert CG, et al. 2005;25:4269-75 Tumor lymphangiogenesis in inflammatory breast carcinoma: 15. Pena L. Abstract “First International Inflammatory Breast a histomorphometric study. Clin Cancer Res 2005;11:7637-42. Cancer Conference”, Dec 2008, MD Anderson Cancer Center, 26. Van Golen KL, Davies S, Wu ZF, et al. A novel putative low- Houston, Texas. affinity insulin-like growth factor-binding protein, LIBC (lost 16. Basu GD, Liang WS, Stephan DA, et al. A novel role for in inflammatory breast cancer), and RhoC GTPase correlate cyclooxygenase-2 in regulating vascular channel formation with the inflammatory breast cancer phenotype. Clin Cancer by human breast cancer cells. Breast Cancer Res 2006;8:R69. Res 1999;5:2511-19. 17. McCarthy NJ, Yang X, Linnoila IR, et al. Microvessel den- 27. Charafe-Jauffret E, Tarpin C, Bardou VJ, et al. Immunophe- sity, expression of estrogen receptor alpha, MIB-1, p53, and notypic analysis of inflammatory breast cancers: identification c-erbB-2 in inflammatory breast cancer. Clin Cancer Res of an “inflammatory signature”. J Pathol 2004;202:265-73. 2002;8:3857-62. 28. Tomlinson JS, Alpaugh ML, Barsky SH. An intact overex- 18. Colpaert C, Vermeulen P, Benoy I, et al. Inflammatory pressed E-cadherin/alpha, beta-catenin axis characterizes the breast cancer shows angiogenesis with high endothelial pro- lymphovascular emboli of inflammatory breast carcinoma. liferation rate and strong E-cadherin expression. Br J Cancer Cancer Res 2001;61:5231-41. 2003;88:718-25. 29. Xiao Y, Ye Y, Yearsley K, Jones S, Barsky SH. The lymphovas- 19. Van Laere S, Van der Auwera I, Van den Eynden G, et al. cular embolus of inflammatory breast cancer expresses a stem Distinct molecular phenotype of inflammatory breast cancer cell-like phenotype. Am J Pathol 2008;173:561-74. compared to non-inflammatory breast cancer using Affyme- 30. MDA-IBC-1: Abstract “First International Inflammatory trix-based genome-wide gene-expression analysis. Br J Can- Breast Cancer Conference”, Dec 2008, MD Anderson Cancer cer 2007;97:1165-74. Center, Houston, Texas. 20. Bertucci F, Finetti P, Rougemont J, et al. Gene expres- 31. Van Laere S et al. Abstract “First International Inflam- sion profiling for molecular characterization of inflammatory matory Breast Cancer Conference”, Dec 2008, MD Anderson

55 v o l . 3 i s s u e 2 - 2 0 0 9 BELGIAN JOURNAL OF MEDICAL ONCOLOGY Cancer Center, Houston, Texas. C orrespondence address 32. Yang-WT, Le-Petross HT, Macapinlac H, et al. Inflammatory breast cancer: PET/CT, MRI, mammography, and sonographic findings. Breast Cancer Res Treat 2008;109:417-26. Authors: P. Vermeulen 1, K. van Golen 2, L. Dirix 1 33. Farnie G, Clarke RB, Spence K, et al. Novel cell culture tech- 1 Translational Cancer Research Group & Pathology nique for primary ductal carcinoma in situ: role of Notch and departments, Cancer Center of the Sint-Augustinus epidermal growth factor receptor signaling pathways. J Natl Hospital & University of Antwerp, Wilrijk, Belgium; Cancer Inst 2007;99:616-27. 2 Department of Biological Sciences, University of 34. Götte M, Kersting C, Radke I, Kiesel L, Wülfing P. An expres- Delaware, Newark, DE sion signature of syndecan-1 (CD138), E-cadherin and c-met is associated with factors of angiogenesis and lymphangio- Please send all correspondence to: genesis in ductal breast carcinoma in situ. Breast Cancer Res Dr. P. Vermeulen 2007;9:R8. Pathology department 35. Jeschke U, Mylonas I, Shabani N, et al. Expression of sialyl Sint Augustinus Hospital lewis X, sialyl Lewis A, E-cadherin and cathepsin-D in human Oosterveldlaan 24 breast cancer: immunohistochemical analysis in mammary 2610 Wilrijk carcinoma in situ, invasive carcinomas and their lymph node Belgium metastasis. Anticancer Res 2005;25:1615-22. [email protected] 36. Boersma BJ, Reimers M, Yi M, et al. A stromal gene signature associated with inflammatory breast cancer. Int J Cancer Conflicts of interest: the authors have nothing to dis- 2008;122:1324-32. close and indicate no potential conflicts of interest.

Geen tijd om bij elke grote internationale bijeenkomst aanwezig te zijn? Toch behoefte om het nieuws van de recentste ontwikkelingen direct tot u te nemen?

Meldt u dan nu aan voor onze digitale congresmailing Oncologie, een nieuwe service van het Nederlands Tijdschrift voor Oncologie.

Digitale Congresmailing Oncologie

U kunt zich aanmelden via Voor meer informatie kunt u zich wenden tot http://congresmailing.ariezmp.nl Ariez Medical Publishing, 0031-20-5612050

BELGIAN JOURNAL OF MEDICAL ONCOLOGY v o l . 3 i s s u e 2 - 2 0 0 9 56