Angiogenesis in triple-negative adenoid cystic carcinomas of the breast

Transcription

1 Virchows Arch (2011) 459: DOI /s ORIGINAL ARTICLE Angiogenesis in triple-negative adenoid cystic carcinomas of the breast Semir Vranic & Snjezana Frkovic-Grazio & Nurija Bilalovic & Zoran Gatalica Received: 15 June 2011 /Revised: 19 August 2011 /Accepted: 22 August 2011 /Published online: 4 September 2011 # Springer-Verlag 2011 Abstract We compared microvascular density (MVD), lymph vessel density (LVD), and the expression of hypoxia pathway-associated proteins between primary triple-negative adenoid cystic carcinoma of the breast (TN-ACC) and gradematched triple-negative breast carcinomas of no special type (TNBC). Twelve TN-ACC and 15 TNBC were investigated immunohistochemically for CD31, podoplanin (D2-40), von Hippel Lindau protein (pvhl), and hypoxia-inducible factor-1alpha (HIF-1α) protein. All cases were lymph node negative (pn0). The study revealed a median MVD (CD31) of 34 vessels/mm 2 (mean ± SD, 41.33±6.5/mm 2 )inthetn- ACCsubgroupandamedianof55microvessels(mean± SD, 54.9±6.3/mm 2 ) in the TNBC subgroup. The median S. Vranic (*) : N. Bilalovic Department of Pathology, Clinical Center of the University of Sarajevo, Bolnička 25, BA Sarajevo, Bosnia and Herzegovina semir.vranic@gmail.com S. Vranic : Z. Gatalica Department of Pathology, Creighton University School of Medicine, Omaha, NE, USA S. Frkovic-Grazio Department of Gynecological Pathology and Cytology, University Clinical Hospital Ljubljana, Ljubljana, Slovenia S. Frkovic-Grazio Department of Obstetrics and Gynecology, University Clinical Hospital Ljubljana, Ljubljana, Slovenia Z. Gatalica Caris Life Sciences, Phoenix, AZ, USA LVD (D2-40) was 10.5/mm 2 (mean ± SD, 11.9±1.5/mm 2 )in the TN-ACC subgroup and 15.0/mm 2 (mean ± SD, 16.9± 2.5/mm 2 ) lymph vessels in the TNBC subgroup. The differences were not statistically significant (P=0.93, P= 0.67, respectively). pvhl was detectable in all TN-ACCs whereas two cases of TNBC had less than 5% of the positive cells. HIF-1α protein expression was significantly higher in the tumor cell population than in adjacent normal cells in both subgroups (P=0.009 for TNBC and P=0.028 for TN- ACC, respectively), but there was no significant difference between the two tumor groups. Up-regulation of the hypoxia-induced signaling is seen in both TN-ACC and grade-matched TNBC. Despite its perceived low malignant potential, TN-ACC of the breast does not differ in the number of blood and lymphatic vessels in comparison with the grade-matched TNBC. The reported biologic differences between TN-ACC and TNBC do not appear to result from neoangiogenesis. Keywords Adenoid cystic carcinoma. Triple-negative carcinoma. Breast. Angiogenesis. Hypoxia Introduction Angiogenesis is one of the hallmarks of cancer [1, 2]. Tumor growth is angiogenesis dependent, and therefore, the forming of new blood vessels represents an essential step for survival in a hypoxic tumor environment [3, 4]. Additionally, the density of blood and lymphatic vessels correlates with aggressiveness and metastatic potential of breast carcinomas [5]. Many solid tumors adapt to hypoxia, defined as tissue oxygen level lower than 7% of concentration observed in normal and well-vascularized tissues [6]. The key regulator

2 378 Virchows Arch (2011) 459: of hypoxia-induced adaptation is hypoxia-inducible factor- 1α (HIF-1α) [6]. Under normal oxygenation, HIF-1α is hydroxylated and slated for degradation via von Hippel Lindau (VHL)-dependent ubiquitination pathway. However, under hypoxic or pseudohypoxic conditions, prolyl hydroxylase enzymes (PHD) are inhibited, and HIF-1α escapes degradation. Stable HIF-1α then translocates to the nucleus where it can heterodimerize with HIF-1β (ARNT) to form an active HIF transcription complex leading to expression of genes that can enhance survival in hypoxic conditions: glucose metabolism and angiogenesis, or reduce apoptosis [7]. VHL gene is an important regulator of HIF-1α and consequently tumor angiogenesis [8]. Its genetic alterations have been rarely seen in breast cancer [9 11]. Triple-negative, basal-like breast carcinoma represents a subtype of breast carcinoma linked to an aggressive clinical course and poor prognosis due to its high metastatic potential [12, 13]. In contrast, adenoid cystic carcinoma of the breast is a special type of triple-negative basal-like breast carcinoma associated with excellent prognosis [14 16]. The objectives of this study were to investigate microvascular density (MVD) and the hypoxia pathway proteins in triple-negative adenoid cystic carcinomas (TN-ACC) and compare the results with grade-matched triple-negative breast carcinomas of no special type (TNBC). Materials and methods Breast tissue samples Formalin-fixed, paraffin-embedded tissues of 12 primary TN-ACC and 15 grade-matched TNBC were retrieved from the pathology archives of the authors' affiliated institutions. Both tumor types were graded using the Nottingham histologic grading as recommended by the 7th American Joint Committee on Cancer recommendations [17]. Both cohorts were previously characterized for various biologic characteristics and recently reported [16, 18, 19]. Immunohistochemistry Endothelium of the tumor microvessels was highlighted using CD31 antibody (Table 1). MVD was quantified using Weidner et al. criteria [20]. Briefly, each slide was first scanned at low power to identify the areas with the highest number of blood vessels ( hot spot areas). Three areas with the highest number of blood vessels were counted using a graticule that measured 1 mm 2. Assessment of lymph vessel density (LVD) was performed in sections immunostained with podoplanin (D2-40, Table 1) using the same approach as in the assessment of MVD. Table 1 Clones, dilutions, sources of antibodies, and types of tissue pretreatment used in the study Antibody (clone) pvhl was calculated as a percentage of positive tumor cells. Only membranous/cytoplasmic staining was considered specific. For HIF-1α scoring, a median HIF-1α expression in normal/benign glands served as a cutoff point for evaluation in both subgroups. Only nuclear HIF-1α expression was considered specific [21]. Immunohistochemical assays for estrogen receptor (ER; clone 6F11, Ventana Medical Systems, Tucson, AZ; clone SP1, Lab Vision, Fremont, CA), progesterone receptor (PR; clone 16, Ventana Medical Systems; clone PgR636, Dako, Glostrup, Denmark), and Her-2/neu protein (Clone CB11, Ventana Medical Systems) were previously done andreported[16, 18]. Positive and negative controls for each marker were used according to the supplier's data sheets. Statistical analysis Mann Whitney U test was used to compare the means of lymph vessel density and microvessel density between the subgroups. The continuous variables (CD31, D2-40, pvhl, HIF-1α) were divided into two categories according to the median values, and the associations between these variables and other clinicopathologic parameters were done using 2 2 contingency tables and Fisher's exact test. A P value was set at <0.05. All statistical analyses were performed using the Statistical Package for Social Sciences software (v. 17.0; SPSS Inc., Chicago, IL). Results Patients' characteristics Dilution Source Antigen retrieval (water bath) CD31 (JC70A) 1:80 DakoCytomation ph 9.0 EDTA (20 min) Podoplanin (D2-40) 1:5 DakoCytomation ph 9.0 EDTA (20 min) VHL (Ig33) 1:100 LabVision ph 6.0 citrate buffer (20 min) HIF-1α (H1α 67) 1:1,000 AbCam ph 6.0 citrate buffer (20 min) All patients were women. The mean age of the TN-ACC patients was 59.6 years (range, years) and 58.5 years (range, years) for the TNBC patients.

3 Virchows Arch (2011) 459: Histopathologic and immunohistochemical characterization of TN-ACC and TNBC of the breast Both cohorts have been previously studied for various biologic characteristics and recently reported [16, 18, 19]. Clinicopathologic characteristics of both cohorts are summarized in Table 2. Briefly, the subgroups were gradematched triple-negative breast carcinomas (ER/PR/Her-2/ neu negative), and all were lymph node negative (pno). Two of the TN-ACC cases had known distant metastases (to kidneys) [16, 22]. Status of the distant metastases was not known for the TNBC group. Lymph vessel density and microvessel density in TN-ACC and TNBC of the breast Evaluation of MVD revealed a median of 34 blood vessels/ mm 2 (mean ± SD, 41.33±6.5) in the TN-ACC group and a median of 55 microvessels/mm 2 (mean ± SD, 54.9±6.3) in the TNBC group (Fig. 1a, b). The median LVD was 10.5/ mm 2 (mean ± SD, 11.9±1.5) in the TN-ACC group and 15.0/ mm 2 (mean ± SD, 16.9±2.5) in the TNBC group (Fig. 1c, d). The differences for both variables were not statistically significant between the groups (P=0.93 for MVD and P= 0.67 for lymph vessel density, Mann Whitney U test). Unequivocal lymphatic or vascular invasion was not evident by either CD31 or D2-40 staining in any of the TN- ACC cases including two cases with distant metastases. In contrast, two TNBC cases exhibited unequivocal lymphatic and vascular invasion of which one had extensive peritumoral vascular invasion. Table 2 Clinicopathologic characteristics of both cohorts Characteristic TN-ACC TNBC Tumor size <2 cm 3 (25%) 2 (18.2%) a 2 5 cm 7 (58.3%) 6 (54.5%) >5 cm 2 (16.7%) 3 (27.3%) Histologic grade b 1 3 (25%) 2 (13.3%) 2 6 (50%) 8 (53.3%) 3 3 (25%) 5 (33.3%) Axillary lymph nodes Axillary dissection 8 (66.7%) 12 (80%) Sentinel lymph node biopsy 4 (33.3%) 3 (20%) Type of surgery Mastectomy 8 (66.7%) 11 (73.3%) Breast-conserving surgery 4 (33.3%) 4 (26.7%) a Data were not available for four cases b Both tumor types were graded using the Nottingham histologic grading Cellular hypoxia pathway-associated proteins in TN-ACC and TNBC of the breast pvhl expression was detectable in all TN-ACC cases (Fig. 2a). It was also demonstrable in the majority of the TNBC cases (Fig. 2b) with only two cases having less than 5% of the positive cells. The mean percentage of the pvhl positive cancer cells was 84% in TN-ACC and 71% in TNBC (P=0.42). Adjacent ductal epithelium retained pvhl expression in all cases when available for evaluation. HIF-1α protein nuclear expression was rarely observed in normal breast epithelium with an average of 2% of the positive cells. In contrast, this protein was observed in all tumors with the mean percentage of positive tumor cells of 15% in TN-ACC and 17% in TNBC (Fig. 2c, d). This was statistically significant in both groups (P=0.009 for TNBC and P=0.028 for TN-ACC, respectively). Correlation between the biologic variables and clinicopathologic parameters HIF-1α protein expression was strongly associated with the tumor grade in the TNBC subgroup (P=0.025), while a trend towards a significant positive association was observed in the TN-ACC subgroup (P=0.08, Fisher's exact test). No other significant correlation was seen among the tested variables. Discussion The tumor growth and progression are directly dependent on the development of new blood vessels (angiogenesis) [3, 5]. It increases the opportunity of the tumor cells to enter the circulation and develop metastases, the most common cause of cancer-related mortality [5]. It is usually measured by microvessel density that correlates well with tumor aggressiveness [5]. Triple-negative basal-like breast carcinomas tend to have an aggressive behavior and poor outcome mainly due to their high metastatic potential [13]. This subtype is also characterized by higher microvessel density compared with the non-basal and the non-triple-negative breast carcinomas [13]. On the other hand, TN-ACC of the breast is well known to have an indolent clinical course with favorable outcome [14 16] despite molecular and gene expression profiling studies that revealed an unfavorable gene signature and molecular portrait compatible with that seen in highly aggressive triple-negative, basal-like breast carcinomas [23, 24]. Our study for the first time highlighted that the number of blood and lymphatic vessels did not differ significantly in the TN-ACC subgroup in comparison with grade-matched triple-negative breast carcinomas of no special type. These results indicate that the discrepancy in

4 380 Virchows Arch (2011) 459: Fig. 1 Endothelium of the tumor microvessels highlighted using CD31 antibody in a case of triple-negative adenoid cystic carcinoma of the breast (a) and a case of triple-negative breast carcinoma (b); lymph vessel density measured by podoplanin (D2-40) in triple-negative adenoid cystic carcinoma (c) and triple-negative breast carcinoma (d) ( 20) A B C D clinical behavior and aggressiveness between TN-ACC and TNBC probably cannot be attributed to the blood and lymphatic vessel density. Angiogenesis is closely regulated by cellular hypoxia [25]. Hypoxia occurs in many disease processes, and it is common in solid tumors due to the tumor outgrowing the existing vasculature [25]. Neoplasms surviving in the hypoxic environment are usually associated with aggressive behavior, increased tumor cell motility, metastasis, and poor outcome [26, 27]. Central to the hypoxic response of the Fig. 2 pvhl was demonstrable in the majority of TN-ACCs (a) and triple-negative breast carcinomas (b); HIF-1α protein nuclear expression with moderate intensity in TN-ACC case (c) and triple-negative breast carcinoma (d) ( 20) A B C D

5 Virchows Arch (2011) 459: tumor cells is the transcription factor HIF-1α [28]. Nuclear translocation of HIF-1α protein leads to the up-regulation of transcription of the genes involved in cell proliferation, angiogenesis, and glucose metabolism and is associated with tumor progression in various human tumors [7, 29]. In breast cancer, its deregulation is also associated with increased proliferation and angiogenesis, down-regulation of estrogen receptor, and shorter survival in patients with negative lymph nodes [27, 30 32]. Our study revealed that HIF-1α protein is commonly up-regulated in both TN-ACC and TNBC which is in line with the study of Rakha et al. that revealed activation of hypoxia-related proteins in TNBC [33]. Earlier studies also showed that up to 62% of all breast carcinomas are hypoxic if HIF-1α protein is used as a marker of hypoxia [6, 34]. The percentage of HIF-1α positivity in breast cancer tumor cells however varies between the studies due to the various cutoff values the investigators used [35, 36]. We observed low HIF-1α protein expression in benign/normal glands (~2%), which is in agreement with previous reports [28], and both of our tumor cohorts exhibited significantly higher percentages of positive cells. Concordant with our IHC results, we observed consistent overexpression of HIF-1α mrna in three cases of breast ACC in various metastatic sites and overexpression of VEGFα mrna in one case, using whole genome expression array (HumanHT-12 v4 beadchip, Illumina Inc., San Diego, CA) (Gatalica Z, unpublished observation). Another important regulator of angiogenesis is the VHL gene (3p26-p25) [8]. Its product, VHL protein (pvhl), is a component of an Skp1 Cdc53 F-box (SCF)-like ubiquitin ligase complex that targets the α-subunits of the hypoxiainducible factor for poly ubiquitylation and proteasomal degradation [37]. Genetic alterations of the VHL gene have been linked to hereditary and sporadic clear cell renal cell carcinomas [38]. Earlier studies reported that VHL gene and pvhl are not commonly involved in breast cancer carcinogenesis [9 11]. Consistent with those studies, all TN-ACC and a majority of TNBC cases in our study retained pvhl expression. Furthermore, no 3p locus loss was observed in a series of breast ACC recently reported [39]. We conclude that up-regulation of the hypoxia-induced signaling is commonly seen in both TN-ACC and gradematched TNBC when compared to the surrounding normal tissues. Despite its low malignant potential, TN-ACC of the breast does not differ in the number of blood and lymphatic vessels in comparison with the grade-matched TNBC. Acknowledgments We are grateful to Ms. Mindee C. Kurtis, Creighton Medical Laboratories, Creighton University Medical Center (Omaha, NE) for excellent technical assistance. Conflict of interest We declare that we have no conflict of interest. References 1. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144: Hanahan D, Weinberg RA (2000) The hallmarks of cancer. Cell 100: Folkman J (2007) Is angiogenesis an organizing principle in biology and medicine? J Pediatr Surg 42: Weinberg RA (2007) The biology of cancer. Garland Science, Taylor & Francis Group, LLC, New York 5. Weidner N (2004) The importance of tumor angiogenesis. Am J Clin Pathol 122: Lundgren K, Holm C, Landberg G (2007) Hypoxia and breast cancer: prognostic and therapeutic implications. Cell Mol Life Sci 64: Haase VH (2006) The VHL/HIF oxygen-sensing pathway and its relevance to kidney disease. Kidney Int 69: Calzada MJ (2010) Von Hippel-Lindau syndrome: molecular mechanisms of the disease. Clin Transl Oncol 12: Lantzsch T, Hefler L, Koelbl H, Lampe D (2002) Expression of von Hippel-Lindau gene protein in breast cancer tissue. Gynecol Oncol 84: Sourvinos G, Miyakis S, Liloglou TL, Field JK, Spandidos DA (2001) Von Hippel-Lindau tumour suppressor gene is not involved in sporadic human breast cancer. Tumour Biol 22: Corless CL, Kibel AS, Iliopoulos O, Kaelin WG Jr (1997) Immunostaining of the von Hippel-Lindau gene product in normal and neoplastic human tissues. Hum Pathol 28: Mohammed RA, Ellis IO, Mahmmod AM et al (2011) Lymphatic and blood vessels in basal and triple-negative breast cancers: characteristics and prognostic significance. Mod Pathol 24: Sabatier R, Jacquemier J, Bertucci F et al (2011) Peritumoural vascular invasion: a major determinant of triple-negative breast cancer outcome. Eur J Cancer Mar 8 doi: /j. ejca Khanfir K, Kallel A, Villette S et al (2011) Management of adenoid cystic carcinoma of the breast: a rare cancer network study. Int J Radiat Oncol Biol Phys (in press) May 11. doi: /j.ijrobp Ghabach B, Anderson WF, Curtis RE, Huycke MM, Lavigne JA, Dores GM (2010) Adenoid cystic carcinoma of the breast in the United States (1977 to 2006): a population-based cohort study. Breast Cancer Res 12:R Vranic S, Frkovic-Grazio S, Lamovec J et al (2010) Adenoid cystic carcinomas of the breast have low Topo IIα expression but frequently overexpress EGFR protein without EGFR gene amplification. Hum Pathol 41: Edge SB, Byrd DR (2010) AJCC cancer staging manual, 7th edn. Springer, Berlin 18. Zhang XT, Kang LG, Ding L, Vranic S, Gatalica Z, Wang ZY (2011) A positive feedback loop of ER-α36/EGFR promotes malignant growth of ER-negative breast cancer cells. Oncogene 30: Vranic S, Gurjeva O, Frkovic-Grazio S, Palazzo J, Tawfik O, Gatalica Z (2011) IMP3, a proposed novel basal phenotype marker, is commonly overexpressed in adenoid cystic carcinomas but not in apocrine carcinomas of the breast. Appl Immunohistochem Mol Morphol Mar 22 (in press) doi: / PAI.0b013e Weidner N, Semple JP, Welgh WR, Folkman J (1991) Tumor angiogenesis and metastasis correlation in invasive breast carcinoma. N Engl J Med 324: Yan M, Rayoo M, Takano EA, KConFab Investigators, Fox SB (2009) BRCA1 tumours correlate with a HIF-1alpha phenotype

6 382 Virchows Arch (2011) 459: and have a poor prognosis through modulation of hydroxylase enzyme profile expression. Br J Cancer 101: Vranic S, Bilalovic N, Lee LM, Kruslin B, Lilleberg SL, Gatalica Z (2007) PIK3CA and PTEN mutations in adenoid cystic carcinoma of the breast metastatic to kidney. Hum Pathol 38: Weigelt B, Geyer FC, Reis-Filho JS (2010) Histological types of breast cancer: How special are they? Mol Oncol 4: Weigelt B, Horlings HM, Kreike B et al (2008) Refinement of breast cancer classification by molecular characterization of histological special types. J Pathol 216: Otrock ZK, Hatoum HA, Awada AH et al (2009) Hypoxiainducible factor in cancer angiogenesis: structure, regulation and clinical perspectives. Crit Rev Oncol Hemat 70: Voss MJ, Moller MF, Powe DG, Niggemann B, Zanker KS, Entschladen F (2011) Luminal and basal-like breast cancer cell show increased migration induced by hypoxia, mediated by an autocrine mechanism. BMC Cancer 11: Bos R, Zhong H, Hanrahan CF et al (2001) Levels of hypoxiainducible factor-1 alpha during breast carcinogenesis. J Natl Cancer Inst 93: Kimbro KS, Simons JW (2006) Hypoxia-inducible factor-1 in human breast and prostate cancer. Endocr Relat Cancer 13: Carroll V, Ashcroft M (2006) Role of hypoxia-inducible factor (HIF)-1A versus HIF-2A in the regulation of HIF target genes in response to hypoxia, insulin-like growth factor-i, or loss of von Hippel-Lindau function: implications for targeting the HIF pathway. Cancer Res 66: Bos R, van Diest PJ, de Jong JS, van der Groep P, van der Valk P, van der Wall E (2005) Hypoxia-inducible factor-1alpha is associated with angiogenesis, and expression of bfgf, PDGF-BB, and EGFR in invasive breast cancer. Histopathology 46: Bos R, van der GP, Greijer AE et al (2003) Levels of hypoxiainducible factor-1alpha independently predict prognosis in patients with lymph node negative breast carcinoma. Cancer 97: Ryu K, Park C, Lee Y (2011) Hypoxia-inducible factor 1 alpha represses the transcription of the estrogen receptor alpha gene in human breast cancer cells. Biochem Biophys Res Commun 407: Rakha EA, Elsheikh SE, Aleskandarany MA et al (2009) Triplenegative breast cancer: distinguishing between basal and nonbasal subtypes. Clin Cancer Res 15: Vleugel MM, Bos R, Buerger H et al (2004) No amplifications of hypoxia-inducible factor-1alpha gene in invasive breast cancer: a tissue microarray study. Cell Oncol 26: Trastour C, Benizri E, Ettore F et al (2007) HIF-1alpha and CA IX staining in invasive breast carcinomas: prognosis and treatment outcome. Int J Cancer 120: van der Groep P, Bouter A, Menko FH, van der Wall E, van Diest PJ (2008) High frequency of HIF-1alpha overexpression in BRCA1 related breast cancer. Breast Cancer Res Treat 111: Kaelin WG Jr (2002) Molecular basis of the VHL hereditary cancer syndrome. Nat Rev Cancer 2: Lin F, Shi J, Liu H et al (2008) Immunohistochemical detection of the von Hippel-Lindau gene product (pvhl) in human tissues and tumors: a useful marker for metastatic renal cell carcinoma and clear cell carcinoma of the ovary and uterus. Am J Clin Pathol 129: Lopez-Garcia MA, Wetterskog D, Lambros MB et al (2011) Genomic profiling of adenoid cystic carcinomas of the breast. Mod Pathol 24(suppl1):52A (meeting abstract 205)