OSTEOPONTIN EXPRESSION IS INDEPENDENT FACTOR FOR POOR SURVIVAL IN ORAL SQUAMOUS CELL CARCINOMA: A COMPUTER ASSISTED ANALYSIS ON TMA SECTIONS Department of Pathology, Rijeka University School of Medicine, Rijeka, Croatia Manuela Avirović, Koviljka Matušan-Ilijaš ,Giuseppe Damante, Dora Fabrro, Robert Cerović, Mirna Juretić, Blaženka Grahovac, Nives Jonjić, and Ksenija Lučin Figure 1. Immunohistochemical staining for osteopontin (OPN) in oral squamous cell carcinoma. A: Normal oral mucosa showing OPN staining of low intensity in basal cells. B: Tumor cells and lymphocites are commonly negative for OPN, while intraepithelial granulocytes are strongly positive. C: Invasive tumor front with moderate to strong OPN expression in tumor cells, and moderate to strong OPN staining in stromal macrophages and granulocytes. D: Moderate to strong OPN expression in tumor cells, and weak OPN expression in stromal fibroblasts and myofibroblasts, respectively. Magnification x200. Figure 2. Kaplan-Meier cumulative survival analysis according to staining for OPN. The log-rank test showed significantly shorter overall survival in patients with high tumoral OPN expression (a, p=0.022), while patients with high stromal OPN expression showed a trend toward longer survival (b, p=0.161). Oral squamous cell cancer (OSCC) is the fifth most common type of cancer worldwide which is characterized by a high degree of local invasiveness and a high rate of locoregional metastasis. Ostepontin (OPN) is noncollagenous extracellular matrix phosphoglycoprotein involved in various physiological and pathological events including tumor progression. It is synthesized by a variety of tissue types including fibroblasts, osteoblasts, osteocytes, dendritic cells, macrophages, smooth muscle cells, endothelial cells and some epithelial cells. The main functional domains of OPN consist of GRDGS motif which binds to αv and α5β1 integrins, and thrombin cleaved epitope SVVYGLR (OPN-R, a non-RGD domain) which binds to integrin receptors α4β1, α9β1, and α9β4, present on a number of immune cells such as macrophages, neutrophils and T lymphocytes. The hyaluronan receptor CD44 has also been identified as a receptor for OPN. The binding of OPN to these surface receptors can elicit extensive changes in cells functions, such as enhanced mobility and adhesion, accelerated growth and division, prolongation of cell survival, and angiogenesis. OPN has been studied as a secreted, extracellular protein, but recently an intracellular isotype of OPN (iOPN) has been described in fibroblasts, macrophages and dendritic cells, which has biological functions distinct from secreted OPN. The aim of this study was to analyze the expression of OPN in normal oral mucosa and oral squamous cell carcinoma (OSCC), and to assess its prognostic significance. Tissue mixcroarray (TMA): For each representative tumor sample 2 cores from superficial part of the tumor, 2 cores from deep invasive front, 2 cores from lymph node with metastasis, if present, and 1 core from normal oral mucosa were taken and placed into recipient paraffin block using a manual tissue microarrayer. Immunohistochemistry: Tumor samples were processed for immunohistological analysis to determinate OPN and Ki-67 protein expression. Staining results for tumor-derived OPN (t- OPN) and Ki-67 proliferation index were assessed by using computer-assisted analysis. The results of stromal-derived OPN (s-OPN) were analysed in TMA sections from invasive tumor front by standard, i.e. non-computer assisted, scoring system. This semiquantitative system takes into consideration the proportion of positive cells (scored on a scale from 1-3 (1-<33%, 2-34-66%, 3->66%) and staining intensity (scored on a scale from 0-3; 0-negative, 1-weak positivity, 2- moderate positivity, and 3-strong positivity). The proportion and intensity were than summed to produce total score of 0 to 6. A score of 0-2 was regarded as low s-OPN expression, score of 3-4 as moderate, and score of 4-6 as strong s-OPN expression. For the purpose of survival statistics, tumors were divided into 2 categories with low (0-3) or high (4-6) s-OPN expression. Association of OPN expression, clinicopathological variables and clinical outcome The immunohistochemical expression of t-OPN was compared with the usual clinicopathological prognostic factors, including pT, histological grade, nodal stage, WHO clinical stage and Ki-67 proliferation index. We found statistically significant association between t-OPN protein and nodal stage (p=0.045), and WHO clinical stage (p=0.033) (Table1). The relationship between OPN level and clinical outcome was investigated in whole group of 86 patients. The median follow up was 29 months (range 1-60). Patients were grouped as having low or high t-OPN levels based on the cut off t-OPN level (OPN<71 versus OPN>71) for the entire tumor cell population, i.e. for both superficial and deep parts. The cumulative 5-year overall survival rates of patients who had low t-OPN expression (54.3%) was significantly higher than those of patients who had high t-OPN expression (24.4%, p=0.022, log-rank test) (Fig 2a). In univariate analysis pT stage, pN stage, and clinical stage were also prognostic factors for survival with p<0.05. When s-OPN expression was compared to clinicopathological parameters and clinical outcome, no significant associations were found. However, opposite observation was found between s-OPN expression and cumulative 5-year overall survival. Lower survival rates (29.4%) were observed in the low s-OPN group, reverse to higher survival rates (39.7%) in the group with high s-OPN expression (p=0.161, log-rank test) (Figure 2b). OPN is upregulated in tumor and stromal OSCC cells. Tumor cell-derived OPN is involved in tumor progression and can independently predict the clinical outcome. Stromal- derived OPN probably has a different function compared to OPN secreted from tumor cells. INTRODUCTION: METHODS: RESULTS: CONCLUSION: In multivariate Cox proportional hazard analysis, OPN from deep, invasive front emerged as independent prognostic factor for survival (p=0.049), while overall OPN expression showed only a trend to influence survival (p=0,122). Literature: 1. Forastiere A et al:, Head and neck cancer, N.Engl J med. 2. Denhardt DT et al, Osteopontin as a means to cope with enviromental insults,:regulation of inflammation, tissue remodeling, and cell survival:, J Clin Invest. 3. Weber GF et al, Receptor ligand interaction between CD44 and Osteopontin, Science 4. Inoue M et al, Intracellular OPN (iOPN) end immunity, Immunol Res 5. Weber GF et al, Osteopontin is a marker of cancer aggressiveness and patient survival, Br J Cancer 6. Barne L et al, WHO Classification of tumors, pathology and genetics of head and neck tumors, Lyon IARC 7. Petrik D et al, Plasma Osteopontin is a independent prognosttic marker for head and neck cancers, J Clin Oncol. 8. Kon S et al, Antibodies to different peptides in Osteopontin reveal complexities in the various secreted forms, J Cell Biochem. 9. Rittling SR, Osteopontin in macrophage function, Exp Rev Mol Med. Multivariate analysis of prognostic factors in oral squamous cell carcinoma Variable Relative risk 95% confidence interval p value Lymph nodes 1.4 0.5-3.2 0.485 Pathological stage 2.6 1.3-5 0.006 TNM 1.4 0.4-4.6 0.559 t-OPN 1.8 1-3.2 0.049 A B C D Table 1. Comparison of clinicopathological parameters with tOPN values using χ² test and survival proportion by univariate log rank test Variables Number of cases N=86 Tumor OPN cut off 71 (0) (1) p value Cumulative 5-year survival % p value Gender Male Female 72 14 0.199 34 38 4 10 0.407 50.0 34.0 Age ≤ 60 ≥ 60 33 53 0.280 17 16 21 32 0.954 36.4 38.3 WHO PT T1 and T2 T3 and T4 57 29 0.709 26 31 12 17 ≤0.001 50.0 14.3 Nodal Stage Negative Positive 35 51 0.045 20 15 18 33 0.020 51.6 28.6 WHO TNM I and II III and IV 26 60 0.033 16 10 22 38 ≤0.001 68.2 25.9 Histol. Grading Well Moderate Poor 9 55 18 0.494 5 4 27 28 8 10 0.906 42.9 36.5 36.8 Ki 67 proliferation index ≤ 20 ≥ 20 40 45 0.614 16 20 24 24 0.529 32.4 40.5 Stromal OPN 0 1 17 63 0.034 12 5 28 39 0.161 29.4 39.7 Tumoral OPN 0 1 38 48 0.022 54.3 24.4