1 Introduction
Breast cancer is both the most common malignancy and the leading cause of cancer death in the women's world. In 2015 alone, 2.4 million new cases and 523,000 cancer-related deaths occurred.[1] In some Western countries, the mortality rate is declining because of early detection and adjuvant systemic therapies. However, in developing countries, breast cancer is more often diagnosed at a later stage, and long-term survival is poorer.[2,3] Although progress has been made in the treatments of breast cancer, its prognosis is still poor. Five-year survival is 85% in many Northern and Western European countries, while it is 60% in developing countries.[4] In the development, progression, and prognosis of breast cancer, it is very heterogeneous and complex. Despite there are a few biomarkers that have been developed, including circulating tumor cells (CTCs),[5] estrogen receptor (ER), progesterone receptor (PR), and p53,[6] novel tissue- or serum-based biomarkers that can be measured repeatedly to provide prognosis information are still needed.
Hyaluronic acid (HA) is the major glycosaminoglycan (GAG) in the extracellular matrix (ECM). D-glucuronate residue is linked through a β-(1→3) glycosidic bond with N-acetyl-D-glucosamine residue to form disaccharide. HA is formed with more than 2000 repeating disaccharides, which are linked together through a β−(1→4) glycosidic bond to form a giant polysaccharide.[7] Because HA has no sulfate ester groups in other GAGs,[8,9] its molecule structure is relatively simple. However, several reports demonstrated that HA actively regulates cell adhesion, migration, proliferation,[10] and inflammatory diseases.[11–13] In addition, the high HA level in several types of solid tumors, such as bladder, prostate, breast, lung, and colon, is considered to be a malignant marker in clinic.[14–20] HA regulates the tumor microenvironment by interacting with specific receptors and intracellular signal transduction that promotes a malignant phenotype. Therefore, high levels of HA are identified in various cancer types (such as breast cancer, prostate cancer, lung cancer, and ovarian cancer).[21–23] Besides, several clinical studies have shown that HA promotes breast cancer cell invasion, motility and EMT by triggering various signaling pathways and by upregulating the expression of collagen degrading enzymes.[24–30] Recent accumulated evidence supports that a relationship between HA tumor expression and poor outcome in ovarian, gastric and breast carcinomas.[31,32] However, some of these studies have failed to confirm that HA is a potential prognostic biomarker for cancer. As a result, these researches in different tumor types remains controversial, and the consistency and magnitude of the prognostic value of HA expression remains enigmatic.[33–35] To verify its clinical relevance, we performed this meta-analysis systematically by integrating all published evidence to determine whether HA can predict poor survival in breast cancer patients.
2 Materials and methods
2.1 Search strategy and study selection
According to the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement and methods,[36] we carefully performed a systematical search of PubMed, EMBASE, Web of Science, the Chinese National Knowledge Infrastructure and Wanfang data for relevant articles up to September 6, 2019. The Medical Subjects Headings (MeSH) and the appropriate corresponding keywords were used: “breast cancer” or “breast tumor” or “breast neoplasm,” and “hyaluronan” or “hyaluronic acid.” The included studies met the following criteria:
- 1. the diagnosis of breast cancer was confirmed by pathological or mammography examination;
- 2. the level of HA in either tumor tissue or plasma was examined;
- 3. full text publication described the correlation between the HA level and overall survival (OS) and/or disease-free survival (DFS) and/or recurrence-free survival (RFS) and/or progression free survival (PFS); and
- 4. the publication language was confined to English.
The exclusive criteria were as follows: conference abstracts, letters, comments, reviews, or basic research articles; studies with duplicate data and repeat analyses and those without hazard ratio (HR), 95% confidence interval (95% CI), and P value, or useful data for the calculation from the data or graphical survival plots.[37–39] Duplicates were first removed from the search results. The qualified references were identified by two reviewers independently in accordance to the inclusion criteria outlined above. Any disagreements over study were resolved by a third assessor until a consensus was reached.
2.2 Quality assessment and data extraction
The quality of all the studies included were assessed systematically based on a critical review checklist of the Dutch Cochrane Centre proposed by MOOSE.[40] The required information were extracted from included studies by two investigators independently, including the following characteristics: publication information (i.e., first author, publication year, and country); patients characteristics (i.e., ethnicity, sample site, number of patients, median age, stage of disease, histological type, and follow-up); data for prognosis meta-analysis (i.e., HA detection method, cut-off value, the HRs of the HA for OS, DFS, PFS, RFS, and P values). If the Kaplan–Meier curves alone were available, we extracted survival data from the graphical survival plots using the software Engauge Digitizer 10.8 to estimate the HRs.[41]
2.3 Statistical analysis
The HR with its variance estimates (95% CI) was estimated to evaluate the correlations between HA expression and breast cancer prognosis. The high expression and low expression of HA was defined according to the cut-off value provided in the article. The inter-study heterogeneity of the combined HRs was evaluated with Cochran's Q test and the I-squared statistic. P < .05 or I2 ≥ 50% indicated statistically heterogeneity, which was analyzed with the random-effects model.[41,42] Otherwise, the fixed-effects model was employed.[43] Subgroup analyzes were conducted to figure out the source of existing heterogeneity among results of different studies. HR values >1 indicated significant associations with poor prognosis. Odds ratios (ORs) with 95% CI were used to examine the correlation between elevated expression of HA and the clinical variables in breast cancer.[44] Publication bias was assessed using Egger linear regression test (P < .05 was considered as significant heterogeneity).[45] Finally, sensitivity analysis was carried out by evaluating result stability after sequential omission of each study on the pooled outcomes. All statistical tests were conducted with STATA V12.0 software (Stata Corporation, College Station, TX).
2.4 Ethics
This study was prepared according to PRISMA guidelines. We used publicly available data without involving human participants. Therefore, it does not require formal human research ethics committee review.
3 Results
3.1 Search results and main characteristics of included studies
The detailed flow diagram of screening process is showed in Figure 1. A total of 1855 records were initially identified in the databases of PubMed, EMBASE, Web of Science, the Chinese National Knowledge Infrastructure and Wanfang data, and 678 of them were excluded due to duplication. After titles and abstracts screened for eligibility, 669 articles were removed according to the exclusion criteria. Of the remaining 9 candidate articles enrolled for full-text articles screened, two articles did not provide the sufficient information for the extraction of HRs and 95% CI.[46,47] As a result, 7 eligible articles, including 12 studies for OS, 10 studies for DFS/RFS/PFS, respectively, were considered in the meta-analysis.[33–35,48–51] The included studies investigated a total of 2664 cases from Spain, Italy, Finland, and Germany, which composed with mixed histological type of breast cancer patients. The main characteristics of the eligible studies are summarized in Tables 1 and 2. Period of inclusion ranged from 2000 to 2016. Nine studies detected HA in stroma or tumor by immunohistochemical staining, while one study tested HA in tumor by immunoradiometric analysis, and two studies tested HA in plasma by enzyme linked immunosorbent assay. OS was reported in 6 articles, while DFS or RFS was reported only in 2 articles. Besides, in discovery and validation cohort, Peng et al examined the HA level in plasma of breast cancer patients for different model adjustment (age or clinical pathological factors),[33] thus we got four HR estimates for OS and PFS respectively in one article. HR was reported directly in 4 articles and estimated indirectly in the other 3 articles.
;)
Figure 1: Flow diagram of included study.
Table 1: Main characteristics of enrolled studies.
Table 2: Main results of enrolled studies.
3.2 Correlation between HA and survival outcome with subgroup analysis
According to the heterogeneity of 12 data with OS result (I2 = 66.3%, P = .001), we chose a random model to examine the pooled HR and its corresponding 95% CI. The result showed there was a significant association between a high HA level and poor OS (HR = 1.86, 95% CI: 1.28–2.71, P = .001) (Fig. 2A). Similarly, the heterogeneity of ten studies with DFS/RFS/PFS results was statistically significant (I2 = 66.4%, P = .002), and the random-effects model was used to calculate the pooled HR and its 95% CI, which was significantly different (HR = 1.63, 95% CI: 1.14–2.33, P = .007) (Fig. 2B). The result showed that higher HA level in breast cancer was related to a worse DFS/RFS/PFS.
Figure 2: The pooled HR of HA for overall survival and disease-free survival of breast cancer.
To determine whether the sample type has an effect on the prognostic role of HA in breast cancer, we performed the subgroup analysis for OS by the sample type (Fig. 3). After divided into two subgroups, the heterogeneity for each group was solved (I2 = 0%, P = .582 for stroma group and I2 = 0%, P = .700 for plasma group). Moreover, a high HA level in stroma (HR = 1.63, 95% CI: 1.06–2.51, P = .025) and plasma (HR = 3.26, 95% CI: 2.25–4.73, P < .001) both significantly predicted poor OS. Details are shown in Table 3. However, the HR of HA in plasma is higher than that of HA in stroma.
Figure 3: The pooled HR of HA for overall survival of breast cancer by subgroup analysis.
Table 3: Subgroup analyses of pooled hazard ratios for overall survival with elevated HA expression.
3.3 Correlation between HA level and clinicopathologic parameters
The relationship between HA level and some clinical characteristics was evaluated through the ORs derived from each available study, including tumor grade, tumor size, lymph node metastasis, distant metastases, ER, and PR status (Table 4). There were no significant relationships between high HA level and clinicopathologic parameters. However, a tendency shows that lymph node metastasis (OR = 1.55, 95% CI: 0.96–2.49, P = .07) and tumor grade (OR = 2.10, 95% CI: 0.89–4.96, P = .089) were statistically significant in the correlation study of HA level on the clinical characteristics of patients (Fig. 4).
Table 4: Meta-analyses of HA expression classified by clinicopathologic parameters.
Figure 4: The pooled OR of HA for lymph node metastasis and tumor grade of breast cancer.
3.4 Publication bias and sensitivity analysis
At last, the publication bias was evaluated by Egger's assays. In both OS and DFS/RFS/PFS studies, the funnel plots were showed almost symmetrically (Fig. 5). The P-values for these tests were .209 and .648, respectively, indicating that there was no publication bias among the included studies. Meanwhile, sensitivity analysis was performed to assess whether the pooled results could be influenced by any individual study. As showed in Figure 6, no significant change was found in the results when any single study was excluded, confirming the robustness and reliability of meta-analysis results.
Figure 5: The publication bias was evaluated through Egger's tests.
Figure 6: Sensitivity analysis for OS (A) and DFS/RFS/PFS (B).
4 Discussion
Because of high mortality and poor prognosis, breast cancer is the leading cancer in females. As mentioned previously, significantly increased levels of HA in carcinomas is considered as a biomarker for some solid tumor types, such as bladder, prostate, breast, lung, and colon.[14–20] As a major glycosaminoglycan in ECM, HA can modulate tumor microenvironment to promote the malignant phenotype, which may be explained by the biological characteristics of HA. Besides, some clinical studies have indicated that HA enhances the capacity invasion, motility, and EMT of breast cancer cell, by triggering various signaling pathways and through upregulation of collagen-degrading enzymes expression.[24–30]
Recently, a series of quantitative studies reported the prognostic role of HA in various cancers. Sun et al demonstrated that high-level HA in stroma predicts unfavorable recurrence-free survival (HR = 2.712, 95% CI:1.323–5.560, P = .006) and OS (HR = 1.736, 95% CI:1.050–2.871, P = .032) in HPV-negative oral cavity cancer.[52] In colorectal cancer, Llaneza et al also reported that high tumor cytosolic HA levels indicates shorter relapse-free survival and OS periods (both P < .05).[53] Consistence with Llaneza's result, Kobel's report also showed the pooled RR of OS 5.06 (95%CI: 1.18–21.57, P = .028.[54] However, insignificant or opposite results were also reported in some studies. Florian et al found that presence of HA alone is not of prognostic importance in pancreatic ductal adenocarcinoma.[55] Corte et al showed that high HA intratumoral levels were significantly associated with longer RFS in the subgroup of patients with ductal histological type tumors (P = .01).[34] Overall, the relationship between HA and prognosis in breast cancer patients remains conflicting and inconsistent. Therefore, we conducted this meta-analysis to comprehensively assess the relationship between HA and survival in breast cancer patients based on 7 articles.
Our comprehensive meta-analysis from 7 articles with a cohort of 2664 patients showed the high HA is associated with poor survival (both OS and DFS/RFS/PFS) in breast cancer patients. As known, breast tumor is a heterogeneous cancer and its prognosis is mainly affected by molecule subtype and tumor's status.[56] Therefore, we selected a random-effect model to minimize the influence of these differences, which may confuse the result of meta-analysis. Simultaneously, subgroup analyses were conducted according to the three sites where HA come from (tumor, stroma, and plasma). The heterogeneity diminished significantly in the stroma and plasma subgroup but not in tumor subgroup. Due to limited number of clinicopathologic parameters studies, though there was no significant correlation between HA level and tumor size, distant metastases, ER, and PR status, a tendency shows that tumor grade and lymph node metastasis were significant in the association study of HA level and the clinical characteristics of patients. The latter is consistent with Kettunen's result,[57] which showed that the positive correlations of peritumor/tumor apparent diffusion coefficient ratio with lymph node positivity and high HA content.
Although we are the first to conduct a meta-analysis on the predictive effect of HA for survival in breast cancer, our study still has some limitations. First, the obvious heterogeneity of the enrolled study in the OS group is caused by the different sample sites, methods, the cut-off values of HA and the follow-up duration. Second, although no significant publication bias was detected in the meta-analysis, all the studies being retrospective and usually single center with small sample size cohorts, which was relatively insufficient to determine the conclusion. Third, several HRs which were calculated from the survival curve may be slightly erroneous. Moreover, although some papers from developing countries were researched about relationship between HA and breast cancer diagnosis or metastasis,[58,59] the patients included in this study are only from developed countries. Finally, some cases included in the pooled analyses of OS and DFS/RFS/PFS were analyzed by univariate analysis, which is less reliable than multivariate analysis because compositive factors are considered in multivariate analysis. Moreover, different covariates were adjusted in each study, thus leading bias to the pooled results. Despite these limitations, this meta-analysis has demonstrated the correlation between HA level and breast cancer.
In addition, more research is needed to actually exploit the prognostic value of HA. First, only one article detected the HA levels in plasma, while others used tumor tissue or stroma samples. It is well known that plasma detection can conveniently provide timely information about host responses. Therefore, researchers should evaluate the prognostic value of HA in plasma for breast cancer. Second, the lack of a gold standard for HA threshold limits, its application to routine clinical management. Therefore, the HA cut-off value should be clearly defined based on the global population.
In conclusion, although large-scale and multicenter prospective studies are needed to confirm these findings, our meta-analysis provides statistical evidence that the high level of HA can predict unfavorable breast cancer outcomes. The HA could be a new prognostic biomarker helpful for the selection of individual therapeutic strategy for breast cancer patients.
Author contributions
All authors have read and approved the final version of the manuscript.
Conceptualization: Jun Zhang.
Data curation: Weiping Wu, Lifen Chen, Jing Jin.
Funding acquisition: Jun Zhang, Yanzhong Wang.
Methodology: Lifen Chen.
Software: Weiping Wu.
Supervision: Jun Zhang.
Validation: Yanzhong Wang, Jun Zhang.
Writing – original draft: Weiping Wu.
Writing – review & editing: Yanzhong Wang, Jun Zhang.
References
[1]. Fitzmaurice C, Allen C, Barber RM, et al. Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: a Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol 2017;3:524–48.
[2]. Siegel R, Ma J, Zou Z, et al. Cancer statistics, 2014. CA Cancer J Clin 2014;64:9–29.
[3]. Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010;127:2893–917.
[4]. Allemani C, Weir HK, Carreira H, et al. Global surveillance of cancer survival 1995-2009: analysis of individual data for 25,676,887 patients from 279 population-based registries in 67 countries (CONCORD-2). Lancet 2015;385:977–1010.
[5]. Wallwiener M, Hartkopf AD, Baccelli I, et al. The prognostic impact of circulating tumor cells in subtypes of metastatic breast cancer. Breast Cancer Res Treat 2013;137:503–10.
[6]. Sturgeon CM, Duffy MJ, Stenman UH, et al. National Academy of Clinical Biochemistry laboratory medicine practice guidelines for use of tumor markers in testicular, prostate, colorectal, breast, and ovarian cancers. Clin Chem 2008;54:e11–79.
[7]. Velesiotis C, Vasileiou S, Vynios DH. A guide to hyaluronan and related enzymes in breast cancer: biological significance and diagnostic value. FEBS J 2019;286:3057–74.
[8]. Afratis N, Gialeli C, Nikitovic D, et al. Glycosaminoglycans: key players in cancer cell biology and treatment. FEBS J 2012;279:1177–97.
[9]. Karamanos NK, Piperigkou Z, Theocharis AD, et al. Proteoglycan chemical diversity drives multifunctional cell regulation and therapeutics. Chem Rev 2018;118:9152–232.
[10]. Turley EA, Noble PW, Bourguignon LY. Signaling properties of hyaluronan receptors. J Biol Chem 2002;277:4589–92.
[11]. Sironen RK, Tammi M, Tammi R, et al. Hyaluronan in human malignancies. Exp Cell Res 2011;317:383–91.
[12]. Petrey AC, de la Motte CA. Hyaluronan, a crucial regulator of inflammation. Front Immunol 2014;5:101.
[13]. Heldin P, Lin CY, Kolliopoulos C, et al. Regulation of hyaluronan biosynthesis and clinical impact of excessive hyaluronan production. Matrix Biol 2019;78-79:100–17.
[14]. Tammi MI, Oikari S, Pasonen-Seppänen S, et al. Activated hyaluronan metabolism in the tumor matrix—causes and consequences. Matrix Biol 2019;78–79:147–64.
[15]. Heldin P, Basu K, Olofsson B, et al. Deregulation of hyaluronan synthesis, degradation and binding promotes breast cancer. J Biochem 2013;154:395–408.
[16]. Rizzardi AE, Vogel RI, Koopmeiners JS, et al. Elevated hyaluronan and hyaluronan-mediated motility receptor are associated with biochemical failure in patients with intermediate-grade prostate tumors. Cancer 2014;120:1800–9.
[17]. Schwertfeger KL, Cowman MK, Telmer PG, et al. Hyaluronan, inflammation, and breast cancer progression. Front Immunol 2015;6:236.
[18]. Arai E, Nishida Y, Wasa J, et al. Inhibition of hyaluronan retention by 4-methylumbelliferone suppresses osteosarcoma cells in vitro and lung metastasis in vivo. Br J Cancer 2011;105:1839–49.
[19]. Misra S, Hascall VC, Markwald RR, et al. Interactions between hyaluronan and its receptors (CD44, RHAMM) regulate the activities of inflammation and cancer. Front Immunol 2015;6:201.
[20]. Pasonen-Seppänen S, Takabe P, Edward M, et al. Melanoma cell-derived factors stimulate hyaluronan synthesis in dermal fibroblasts by upregulating HAS2 through PDGFR-PI3K-AKT and p38 signaling. Histochem Cell Biol 2012;138:895–911.
[21]. Karousou E, D’Angelo ML, Kouvidi K, et al. Collagen VI and hyaluronan: the common role in breast cancer. Biomed Res Int 2014;2014:606458.
[22]. Rilla K, Siiskonen H, Tammi M, et al. Hyaluronan-coated extracellular vesicles—a novel link between hyaluronan and cancer. Adv Cancer Res 2014;123:121–48.
[23]. Theocharis AD, Karamanos NK. Proteoglycans remodeling in cancer: underlying molecular mechanisms. Matrix Biol 2019;75–76:220–59.
[24]. Montgomery N, Hill A, McFarlane S, et al. CD44 enhances invasion of basal-like breast cancer cells by upregulating serine protease and collagen-degrading enzymatic expression and activity. Breast Cancer Res 2012;14:R84.
[25]. Bourguignon LY, Wong G, Earle C, et al. Hyaluronan-CD44 interaction promotes c-Src-mediated twist signaling, microRNA-10b expression, and RhoA/RhoC up-regulation, leading to Rho-kinase-associated cytoskeleton activation and breast tumor cell invasion. J Biol Chem 2010;285:36721–35.
[26]. Li L, Qi L, Liang Z, et al. Transforming growth factor-β1 induces EMT by the transactivation of epidermal growth factor signaling through HA/CD44 in lung and breast cancer cells. Int J Mol Med 2015;36:113–22.
[27]. Hamilton SR, Fard SF, Paiwand FF, et al. The hyaluronan receptors CD44 and Rhamm (CD168) form complexes with ERK1,2 that sustain high basal motility in breast cancer cells. J Biol Chem 2007;282:16667–80.
[28]. Fieber C, Baumann P, Vallon R, et al. Hyaluronan-oligosaccharide-induced transcription of metalloproteases. J Cell Sci 2004;117(Pt 2):359–67.
[29]. Chang-I K, Cheng-Yi C, Chih-Chang Y, et al. Enhanced membrane-type 1 matrix metalloproteinase expression by hyaluronan oligosaccharides in breast cancer cells facilitates CD44 cleavage and tumor cell migration. Oncol Rep 2012;28:1808–14.
[30]. Gialeli C, Theocharis AD, Karamanos NK. Roles of matrix metalloproteinases in cancer progression and their pharmacological targeting. FEBS J 2015;278:16–27.
[31]. Setälä LP, Tammi MI, Tammi RH, et al. Hyaluronan expression in gastric cancer cells is associated with local and nodal spread and reduced survival rate. Br J Cancer 1999;79:1133–8.
[32]. Anttila MA, Tammi RH, Tammi MI, et al. High levels of stromal hyaluronan predict poor disease outcome in epithelial ovarian cancer. Cancer Res 2000;60:150–5.
[33]. Peng C, Wallwiener M, Rudolph A, et al. Plasma hyaluronic acid level as a prognostic and monitoring marker of metastatic breast cancer. Int J Cancer 2016;138:2499–509.
[34]. Corte MD, Gonzalez LO, Lamelas ML, et al. Expression and clinical signification of cytosolic hyaluronan levels in invasive breast cancer. Breast Cancer Res Treat 2006;97:329–37.
[35]. Auvinen P, Tammi R, Parkkinen J, et al. Hyaluronan in peritumoral stroma and malignant cells associates with breast cancer spreading and predicts survival. Am J Pathol 2000;156:529–36.
[36]. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. J Clin Epidemiol 2009;62:1006–12.
[37]. Parmar MK, Torri V, Stewart L. Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints. Stat Med 1998;17:2815–34.
[38]. Williamson PR, Smith CT, Hutton JL, et al. Aggregate data meta-analysis with time-to-event outcomes. Stat Med 2002;21:3337–51.
[39]. Tierney JF, Stewart LA, Ghersi D, et al. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials 2007;8:16.
[40]. Stroup DF, Berlin JA, Morton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008–12.
[41]. Higgins JP, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–60.
[42]. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7:177–88.
[43]. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959;22:719–48.
[44]. Cornfield J. A method of estimating comparative rates from clinical data; applications to cancer of the lung, breast, and cervix. J Natl Cancer Inst 1951;11:1269–75.
[45]. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
[46]. Ponting J, Howell A, Pye D, et al. Prognostic relevance of serum hyaluronan levels in patients with breast cancer. Int J Cancer 1992;52:873–6.
[47]. Wernicke M, Pineiro LC, Caramutti D, et al. Breast cancer stromal myxoid changes are associated with tumor invasion and metastasis: a central role for hyaluronan. Mod Pathol 2003;16:99–107.
[48]. Suwiwat S, Ricciardelli C, Tammi R, et al. Expression of extracellular matrix components versican, chondroitin sulfate, tenascin, and hyaluronan, and their association with disease outcome in node-negative breast cancer. Clin Cancer Res 2004;10:2491–8.
[49]. Casalini P, Carcangiu ML, Tammi R, et al. Two distinct local relapse subtypes in invasive breast cancer: effect on their prognostic impact. Clin Cancer Res 2008;14:25–31.
[50]. Karihtala P, Soini Y, Auvinen P, et al. Hyaluronan in breast cancer: correlations with nitric oxide synthases and tyrosine nitrosylation. J Histochem Cytochem 2007;55:1191–8.
[51]. Auvinen P, Tammi R, Kosma VM, et al. Increased hyaluronan content and stromal cell CD44 associate with HER2 positivity and poor
prognosis in human breast cancer. Int J Cancer 2013;132:531–9.
[52]. Sun S, Won HS, Hong SA, et al. Prognostic implications of stromal hyaluronic acid protein expression in resected oropharyngeal and oral cavity cancers. Korean J Intern Med 2020;35:408–20.
[53]. Llaneza A, Vizoso F, Rodríguez JC, et al. Hyaluronic acid as prognostic marker in resectable colorectal cancer. Br J Surg 2000;87:1690–6.
[54]. Kobel M, Weichert W, Cruwell K, et al. Epithelial hyaluronic acid and CD44v6 are mutually involved in invasion of colorectal adenocarcinomas and linked to patient
prognosis. Virchows Arch 2004;445:456–64.
[55]. Florian G, Marius K, Guido S, et al. Is hyaluronan deposition in the stroma of pancreatic ductal adenocarcinoma of prognostic significance? PLoS One 2017;12:e0178703.
[56]. Carey LA, Perou CM, Livasy CA, et al. Race, breast cancer subtypes, and survival in the Carolina Breast Cancer Study. JAMA 2006;295:2492–502.
[57]. Kettunen T, Okuma H, Auvinen P, et al. Peritumoral ADC values in breast cancer: region of interest selection, associations with hyaluronan intensity, and prognostic significance. Eur Radiol 2020;30:38–46.
[58]. Wu M, Cao M, He Y, et al. A novel role of low molecular weight hyaluronan in breast cancer metastasis. FASEB J 2015;29:1290–8.
[59]. Zhang B, Cao M, He Y, et al. Combination of plasma HA and circulating M2-like monocytes may serve as a diagnostic marker for breast cancer. J Cancer 2017;8:3522–30.
