- Email: [email protected]
Diagnostic and prognostic value of serum vitronectin levels in human glioma
Journal of the Neurological Sciences 371 (2016) 54–59
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Diagnostic and prognostic value of serum vitronectin levels in human glioma☆ Mao-Hua Chen a, Chuan Lu a, Jun Sun a, Xian-Dong Chen a, Jun-Xia Dai a, Jian-Yong Cai a, Xiang-Lin Chen b,⁎ a
Department of Neurosurgery, The Central Hospital of Wenzhou City, Wenzhou Medical University Dingli Clinical College, 32 Dajian Lane, Wenzhou 325000, Zhejiang Province, China Department of Cerebrovessel, The People's Hospital of Qingyuan, The Sixth Affiliated Hospital of Guangzhou Medical University, B24 Xinchengyinquan Road, Qingcheng District, Qingyuan 511518, Guangdong Province, China
b
a r t i c l e
i n f o
Article history: Received 6 August 2016 Received in revised form 30 September 2016 Accepted 14 October 2016 Available online 15 October 2016 Keywords: Serum vitronectin Glioma Prognosis Diagnosis
a b s t r a c t Objective: Vitronectin is an extracellular matrix protein, the synthesis of which by glioma cells correlates with tumor grade. The current study was designed to investigate the relationship between serum vitronectin levels and clinicopathological characteristics, diagnosis and prognosis in glioma patients. Methods: In a prospective observatory study, a total of 98 glioma patients, 98 healthy controls, 98 other non-glioma brain tumors, and 98 other non-tumor neurological diseases were recruited. Following univariate analyses, multivariate analyses were performed to explore the associations of serum vitronectin levels with survival and clinicopathological parameters. Receiver operating characteristic curve analysis was done to assess its diagnostic and prognostic predictive value. Results: Serum vitronectin levels were significantly elevated in glioma patients as compared with other groups. High Wealth Health Organization grade was independently associated with high vitronectin levels. Serum vitronectin levels could significantly distinguish glioma patients from other groups and discriminate highgrade glioma from low-grade glioma. Vitronectin levels markedly predicted 5-year progression and 5-year mortality. Moreover, serum vitronectin was identified as an independent predictor for 5-year overall survival and 5year progression-free survival as well as 5-year mortality and 5-year progression. Conclusion: Serum vitronectin may be a promising diagnostic and prognostic biomarker that can be detected in the peripheral blood of patients with glioma. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Glioma is the most common among adult primary intracranial tumors and constitutes N40% of primary neoplasms in central nervous system [1–3]. Up to date, histopathology of tumor specimens has been identified as a gold standard for the grading and prognostic assessment of glioma [4,5]. However, blood biomarkers are gaining increasing interest for diagnostic and prognostic evaluation in cancer patients [6–9]. Currently, no biomarkers in peripheral blood have been established for the routine clinical management of glioma patients. Hence, it is important to identify objective markers, particularly biomarkers that can be determined noninvasively in peripheral blood, for glioma diagnosis and prognostic prediction.
Abbreviations: AUC, area under curve; CI, confidence interval; HR, hazard ratio; KPS, Karnofsky performance score; OR, odds ratio; ROC, receiver operating characteristic; WHO, World Health Organization. ☆ Institution at which the work was performed: The Central Hospital of Wenzhou City. ⁎ Corresponding author. E-mail address: [email protected] (X.-L. Chen).
http://dx.doi.org/10.1016/j.jns.2016.10.022 0022-510X/© 2016 Elsevier B.V. All rights reserved.
Vitronectin is an adhesive protein normally found in serum and plasma, also termed as serum spreading factor and complement S-protein. Generally, vitronectin plays important roles in inflammation, cell adhesion, cell necrosis, and blood coagulation [10–12]. Interestingly, vitronectin has been shown to promote the adhesion of a wide variety of cell types in vitro. Also, as a major component or master organizer of extracellular matrix, vitronectin has been found to regulate vascular endothelial growth factor-induced angiogenesis, a key step in tumor formation [13–15]. Importantly, expression of the vitronectin gene was detected in human glioma by in situ hybridization [16]; vitronectin protected in vitro glioma cells from apoptotic death [17]; glioma cell migration was induced by vitronectin in human serum and cerebrospinal fluid [18]. Moreover, vitronectin tissue expression correlated with tumor grade of glioma [19]. It is suggested that vitronectin should be associated with the occurrence and progression of cancers and also have close relation to glioma stage. Subsequently, we postulated (1) circulating vitronectin levels might be elevated in glioma patients; (2) circulating vitronectin levels might be associated with pathological severity of glioma; (3) circulating vitronectin levels might have the potential to discriminate glioma from other diseases and identify the glioma patients at risk of poor outcome. Here, vitronectin level in serum from a
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
group of glioma patients was investigated and further was identified as a candidate circulating biomarker for the early detection and prognosis of gliomas. 2. Materials and methods 2.1. Study population This prospective, observatory study included a group of newly, histologically confirmed glioma patients between January 2007 and January 2010 at the Department of Neurosurgery, the Central Hospital of Wenzhou City, China. We excluded those patients who underwent radiotherapy or chemotherapy before admission and had prior history of any type of cancer. Additionally, immediate family members are excluded from the history of glioma. Control groups were composed of three subgroups: healthy subjects, other non-glioma brain tumors, and other non-tumor neurological diseases. We excluded those controls with other systemic diseases including autoimmune diseases, uremia, liver cirrhosis, malignancy, and chronic heart or lung disease. This study was approved by the ethics Committee at our hospital (NO. CHWC2006008). Written informed consents were obtained from the study participant or a close relative.
55
Kolmogorov-Smirnov test was carried out to investigate data distribution. The continuous variables (i.e. age and serum vitronectin levels) were normally distributed and were presented as mean ± standard. KPS score and tumor size were dichotomized based on 80 and 5 cm respectively and therefore identified as the categorical variables. Also, the categorical variables were reported as counts (percentage). Chi-square tests (or Fisher exact tests), t-tests and one-way ANOVA tests were performed for intergroup comparisons as appropriate. Post-hoc analysis was done using Least Significant Difference test. Receiver operating characteristic (ROC) curve was performed to assess discriminatory ability with estimated area under curve (AUC). A suitable cutoff value was selected from ROC curve to obtain the optimal sensitivity and specificity. A binary logistic regression analysis was performed to identify risk factors, as well as odds ratio (OR) values and related 95% confidence interval (CI) were estimated. Survival curve were made by using the KaplanMeier method, and the log-rank test was used to evaluate survival difference between the groups. Cox regression analysis was used to assess predictors related to overall survival. The hazard ratio (HZ) values and corresponding 95% CI were calculated. Statistical significance was accepted as a P-value of b0.05.
3. Results 2.2. Assessment 3.1. Clinical characteristics of study population In accordance with World Health Organization (WHO) classification, gliomas are divided to low-grade ones (WHO Grade I or II) and highgrade ones (WHO Grade III or IV) [4]. Karnofsky performance score (KPS) was utilized to assess routine life ability [20]. We also recorded some related information, for instance age, gender, tumor diameter and tumor location. The diagnosis of glioma was confirmed by histological examination. Patients were treated by standard radiotherapy and chemotherapy [21]. All patients were followed up in 3-month intervals for the first 2 years and every 6 months thereafter for up to 5 years. Overall survival time was defined as the period from the date of treatment to the date of death or last date of follow-up. Death was directly related to glioma. This excluded death by indirect but related causes, e.g. accidental trauma. In accordance with RANO criteria [22], progression was defined as an increase in 25% of the product of perpendicular diameters of enhancing lesions, a significant increase in the T2/FLAIR non-enhancing component, appearance of new lesions, and clinical deterioration not attributable to causes other than the tumor or a reduction in the corticosteroid dose. Progression-free survival was calculated from the date of initial treatment to that of disease progression or last follow-up. 2.3. Determination
This study finally enrolled a total of 98 glioma patients, 98 healthy controls, 98 other non-glioma brain tumors, and 98 other non-tumor neurological diseases. This group of glioma patients included 59 males and 39 females and had a mean age of 54.5 ± 12.0 years (range, 32–75 years). 98 healthy controls comprised 56 males and 42 females as well as had a mean age of 52.3 ± 10.3 years (range, 34–72 years). 98 patients with other non-glioma brain tumors (22 vestibular schwannomas, 42 meningiomas and 34 pituitary adenomas) consisted of 54 males and 44 females and had a mean age of 50.9 ± 11.8 years (range, 31–74 years). 98 patients with other non-tumor neurological diseases (traumatic brain injury, 42 patients; spontaneous intracerebral hemorrhage, 26 patients; aneurysmal subarachnoid hemorrhage, 30 patients) were composed of 55 males and 43 females and had a mean age of 49.2 ± 10.7 years (range, 29–71 years). All individuals were Chinese. Intergroup differences of gender and age were not statistically significant between patients and other groups. Among this group of patients consisting of 59 males and 39 females, a total of 36 patients had an age of N 60 years and 62 had an age of ≤60 years; Sixty-six patients had a KPS score of ≥80, and the remaining 32 had a score of b80. Tumor size was ≤5 cm in 65 patients and N5 cm in 33 patients. In total, 81 supratentorial gliomas and 17 infratentorial
Peripheral venous blood was obtained from the patients within 24 h before surgery and from the controls at study entry. The blood samples were immediately placed into sterile test tubes and centrifuged at 3000g for 30 min at 4 °C to collect serum. Serum was stored at −70 °C until assayed. Serum vitronectin (total 65 + 75 kDa polypeptide) levels were measured by using an enzyme-linked immunosorbent assay (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions [23]. Intra-assay precision revealed a coefficient of variation of 4.8%, with an accuracy of 99%. The coefficient of variation of interassay precision was 5.6%, with an accuracy of 97%. The vitronectin levels were calculated by a standard curve. Each serum sample was tested in duplicate and the result was interpreted as μg/mL. All determinations were performed by laboratory technicians blinded to all clinical data. 2.4. Statistical analysis Statistical analysis was done using SPSS 19.0 (SPSS Inc., Chicago, IL, USA) and MedCalc 9.6.4.0 (MedCalc Software, Mariakerke, Belgium).
Fig. 1. Comparisons of serum vitronectin levels between glioma patients and healthy controls, other non-glioma brain tumors as well as other non-tumor neurological diseases.
56
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Fig. 2. Distinguishing ability of serum vitronectin between glioma patients and healthy controls, other non-glioma brain tumors in addition to other non-tumor neurological diseases as well as between high-grade and low-grade gliomas based on receiver operating characteristic curve.
ones was found, including 54 low-grade gliomas (WHO grade I and II) and 44 high-grade gliomas (WHO grade III and IV). 3.2. Serum vitronectin level as a potential diagnostic biological marker Serum vitronectin levels were markedly elevated in all glioma patients as compared to healthy controls, other non-glioma brain tumors, or other non-tumor neurological diseases in Fig. 1. Using ROC curves, we identified the optimal serum vitronectin levels that distinguished glioma from healthy controls, other non-glioma brain tumors, or other non-tumor neurological diseases with high sensitivities and specificities, as well as significantly large AUCs (Fig. 2). 3.3. Association of serum vitronectin levels with clinicopathological features of human glioma In terms of mean value of serum vitronectin levels, the glioma patients were divided to high level group (Nmean value) and low level group (b mean value). In Table 1, as compared with serum vitronectin level b126.1 μg/mL group, serum vitronectin level N 126.1 μg/mL group had higher WHO grade, lower KPS score and larger tumor size.
The multivariate analysis demonstrated that high WHO grade was an independent risk factor for serum vitronectin level N 126.1 μg/mL (OR, 26.623; 95% CI, 7.751–91.444; P b 0.001). Also, in Fig. 3, high-grade gliomas had significantly high serum vitronectin levels, as compared to low-grade gliomas. ROC curve analysis demonstrated that serum vitronectin level discriminated high-grade gliomas from low-grade gliomas significantly (Fig. 2).
3.4. Serum vitronectin identified as a prognostic marker in patients with human glioma Tables 2 and 3 indicate that 5-year overall survival, 5-year progression-free survival, 5-year mortality or 5-year progression was correlated with serum vitronectin level, WHO grade, KPS score and tumor size. The multivariate analyses selected high WHO grade and serum vitronectin level N126.1 μg/mL as the independent predictors for them (Table 4). Fig. 4 illustrates, compared with patients with serum vitronectin level b126.1 μg/mL, those with serum vitronectin level N 126.1 μg/mL had significantly shorter 5-year overall survival time and 5-year progressionfree survival time; also, serum vitronectin levels significantly predicted 5-year mortality and 5-year progression.
Table 1 The factors associated with serum vitronectin levels in human glioma. Serum vitronectin levels
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) Karnofsky performance score (b80/≥80) World Health Organization grade (III + IV/I + II)
N126.1 μg/mL
b126.1 μg/mL
P value
Odds ratio (95% CI)
P value
35/20 23/32 46/9 25/30 23/32 41/14
24/19 13/30 35/8 8/35 9/34 3/40
0.432 0.238 0.771 0.005 0.029 b0.001
1.385 (0.613–3.130) 1.659 (0.714–3.854) 1.168 (0.409–3.335) 3.646 (1.433–9.274) 2.715 (1.094–6.741) 39.048 (10.422–146.294)
0.564 0.332 0.772 0.004 0.026 b0.001
All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group. CI indicates confidence interval.
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Fig. 3. Comparison of serum vitronectin levels between high-grade and low-grade gliomas.
4. Discussion To my best knowledge, this is the first study to measure the circulating vitronectin levels in patients with glioma. The main findings of this study were that: 1) serum vitronectin levels were significantly increased in glioma patients, as compared to healthy controls, other non-glioma brain tumors and other non-tumor neurological diseases and moreover, this biomarker had obvious ability to distinguish gliomas from other groups based on ROC curves; 2) high serum vitronectin levels had an independent relation to high-grade glioma using a binary logistic regression analysis and even, it could significantly discriminate high-grade gliomas from low-grade gliomas in accordance with ROC curve; 3) serum vitronectin was associated independently with 5-year overall survival time and 5-year progression-free survival according to Cox regression analysis as well as 5-year progression and 5-year mortal-
57
ity using binary logistic regression analysis, and serum vitronectin levels had high prognostic value for 5-year mortality and 5-year progression based on ROC curve. These findings substantialize serum vitronectin as a potential prognostic and diagnostic biomarker for glioma. Vitronectin is a glycoprotein, which is a member of the integrin family. It changes in the sera and tissues of people, and is associated with the occurrence and progression of tumor [24–27]. Up to date, there are a few papers regarding circulating vitronectin. Recent studies showed that serum vitronectin levels had significant diagnostic and prognostic value for melanoma and breast cancer [23,28,29]. It has been confirmed that vitronectin could protect glioma cells from apoptotic death [17] and vitronectin from human serum and cerebrospinal fluid induced glioma cell migration [18]. Of note, vitronectin had been localized on cultured human glioma cells [30] and human glioma tissues [16]. Interestingly, this protein expression in glioma tissues has been found to be associated with clinical grade of glioma [19]. Thus, it is postulated that serum vitronectin might be a prognostic and diagnostic biomarker for human glioma. This was designed to first assess the discriminatory ability of serum vitronectin levels in human glioma. We recruited three subgroups of controls, i.e. healthy controls, other non-glioma brain tumors and other non-tumor neurological diseases. According to ROC curve analysis and intergroup comparisons of serum vitronectin levels, it was found that serum vitronectin levels possessed high discriminatory ability for distinguish glioma from other subgroups, indicating serum vitronectin might have significant value as an diagnostic indicator for human glioma. Second, we screened the risk factors associated with serum vitronectin levels identified as categorical variable, using univariate analysis followed by a binary logistic regression analysis; it was demonstrated that high WHO grade was independently associated with high serum vitronectin levels. Also, serum vitronectin levels, emerging as a continuous variable, were compared between high grade and low grade glioma based WHO scale. Meantime, a ROC curve analysis was
Table 2 The factors associated with 5-year overall survival and 5-year progression-free survival in human glioma. 5-year overall survival
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) KPS score (b80/≥80) WHO grade (III + IV/I + II) Serum vitronectin level N126.1 μg/mL
5-year progression-free survival
Hazard ratio (95% CI)
P value
Hazard ratio (95% CI)
P value
1.414 (0.828–2.415) 1.455 (0.866–2.446) 0.684 (0.346–1.353) 2.115 (1.263–3.540) 1.865 (1.105–3.149) 2.878 (1.716–4.828) 2.823 (1.603–4.970)
0.204 0.157 0.278 0.004 0.020 b0.001 b0.001
1.153 (0.719–1.851) 1.400 (0.873–2.246) 0.583 (0.312–1.087) 2.328 (1.433–3.782) 2.074 (1.258–3.419) 3.372 (2.039–5.577) 2.915 (1.742–4.877)
0.554 0.163 0.089 0.001 0.004 b0.001 b0.001
Cox regression analysis was used to assess predictors related to survival. The hazard ratio values and corresponding 95% CI were calculated. CI indicates confidence interval; WHO, World Health Organization; KPS, Karnofsky performance score. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group.
Table 3 The factors associated with 5-year mortality and 5-year progression in human glioma. 5-year mortality
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) KPS score (b80/≥80) WHO grade (III + IV/I + II) Serum vitronectin level N126.1 μg/mL
5-year progression
Odds ratio (95% CI)
P value
Odds ratio (95% CI)
P value
2.164 (0.937–5.000) 1.431 (0.597–3.431) 0.797 (0.250–2.544) 3.184 (1.211–8.370) 2.780 (1.054–7.333) 5.220 (2.051–13.288) 4.778 (1.988–11.484)
0.071 0.422 0.701 0.019 0.039 0.001 0.002
1.263 (0.503–3.168) 1.970 (0.702–5.530) 0.693 (0.179–2.689) 3.460 (1.077–11.121) 3.082 (1.156–9.929) 4.588 (1.554–13.545) 4.339 (1.642–11.467)
0.619 0.198 0.596 0.037 0.029 0.006 0.003
A univariate binary logistic regression analysis was performed to identify risk factors, as well as odds ratio values and related 95% CI were estimated. CI indicates confidence interval; WHO, World Health Organization; KPS, Karnofsky performance score. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1μg/mL) from the glioma group.
58
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Table 4 Multivariate analysis of risk factors associated with 5-year poor outcome in human glioma. Serum vitronectin level N126.1 μg/mL
5-year overall survival 5-year progression-free survival 5-year mortality 5-year progression
WHO grade (III + IV/I + II)
Odds ratio (95% CI)
P value
Odds ratio (95% CI)
P value
2.502 (1.381–4.511) 2.702 (1.511–4.891) 3.399 (1.191–8.588) 3.102 (1.112–7.351)
0.004 0.002 0.007 0.009
2.612 (1.454–4.545) 2.892 (1.561–5.130) 3.509 (1.445–10.521) 3.416 (1.296–9.832)
0.002 0.001 0.005 0.006
CI indicates confidence interval; WHO, World Health Organization. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group.
performed to assess discriminatory ability of vitronectin levels for different clinical grades. Actually, elevated vitronectin levels had close relation to poor clinical grade, suggesting serum vitronectin might have the potential to assess glioma malignancy degree. At last, we evaluated the prognostic value of serum vitronectin in terms of 5-year overall survival, 5-year progression-free survival, 5-year mortality and 5-year progression. Intriguingly, serum vitronectin emerged as an independent predictor for the abovementioned clinical endpoints. ROC curve analysis also showed an interesting result that serum vitronectin levels predicted 5-year mortality and 5-year progression with high sensitivities and specificities. These results substantialized serum vitronectin as a useful prognostic biomarker. Actually, it is impossible that any biomarker is only produced in a type of cell or tissue. Vitronectin is not exceptional. In this study, this biomarker had high predictive or diagnostic value based on areas under curve. Taken together, serum vitronectin might act as a diagnostic and prognostic biomarker for human glioma. However, the current study had two limitations. At first, all patients were Chinese, thus the results may not be more generally applicable to populations of other ethnic origins. Second, clinical parameters, such as pathological diagnosis, are variable between institutions and/or individual clinicians, and hence the results with this small cohort may reflect biases inherent in the acquisition of such clinical data. Clearly, these results remain to be validated in a multicenter clinical trial.
5. Conclusions Serum vitronectin levels are elevated in human glioma, as well as are independently associated with pathologic grade based on WHO classification and 5-year poor prognosis. Moreover, this biomarker significantly discriminates gliomas from controls, other non-glioma brain tumors and other non-tumor neurological diseases, and distinguishes highgrade gliomas from low-grade gliomas, as well as predicts 5-year progression and mortality. Thus, serum vitronectin has the potential to be a promising diagnostic and prognostic biomarker for glioma.
Conflict of interest The authors report no conflicts of interest concerning the materials or methods used in this study or the findings specified in this article.
Acknowledgements The authors thank all staffs in Department of Neurosurgery, The Central Hospital of Wenzhou City (Wenzhou, China) for their technical support. This study was supported financially by Zhejiang Province Medical and Health Project (No. 2014RCA026, 2015ZDA028, 2016RCA026).
Fig. 4. Survival curves of glioma patients in accordance with serum vitronectin levels and predictive value of serum vitronectin levels for 5-year poor outcome based on receiver operating characteristic curve.
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
References [1] P.Y. Wen, S. Kesari, Malignant gliomas in adults, N. Engl. J. Med. 359 (5) (2008) 492–507. [2] Y. Iwadate, T. Sakaida, T. Hiwasa, Y. Nagai, H. Ishikura, M. Takiguchi, et al., Molecular classification and survival prediction in human gliomas based on proteome analysis, Cancer Res. 64 (7) (2004) 2496–2501. [3] A. Omuro, L.M. DeAngelis, Glioblastoma and other malignant gliomas: a clinical review, JAMA 310 (17) (2013) 1842–1850. [4] D.N. Louis, H. Ohgaki, O.D. Wiestler, W.K. Cavenee, P.C. Burger, A. Jouvet, et al., The 2007 WHO classification of tumours of the central nervous system, Acta Neuropathol. 114 (2) (2007) 97–109. [5] M.J. Riemenschneider, J.W. Jeuken, P. Wesseling, G. Reifenberger, Molecular diagnostics of gliomas: state of the art, Acta Neuropathol. 120 (5) (2010) 567–584. [6] S.J. Lee, J.W. Seo, Y.S. Chae, J.G. Kim, B.W. Kang, W.W. Kim, et al., Genetic polymorphism of miR-196a as a prognostic biomarker for early breast cancer, Anticancer Res. 34 (6) (2014) 2943–2949. [7] X. Chen, Y. Ba, L. Ma, X. Cai, Y. Yin, K. Wang, et al., Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases, Cell Res. 18 (10) (2008) 997–1006. [8] V. Zigmont, A. Garrett, J. Peng, M. Seweryn, G.A. Rempala, R. Harris, et al., Association between Prediagnostic serum 25-Hydroxyvitamin D concentration and glioma, Nutr. Cancer 67 (7) (2015) 1120–1130. [9] M.B. Nijaguna, V. Patil, A.S. Hegde, B.A. Chandramouli, A. Arivazhagan, V. Santosh, et al., An eighteen serum cytokine signature for discriminating glioma from normal healthy individuals, PLoS One 10 (9) (2015), e0137524. [10] D.I. Leavesley, A.S. Kashyap, T. Croll, M. Sivaramakrishnan, A. Shokoohmand, B.G. Hollier, et al., Vitronectin–master controller or micromanager? IUBMB Life 65 (10) (2013) 807–818. [11] K.T. Preissner, U. Reuning, Vitronectin in vascular context: facets of a multitalented matricellular protein, Semin. Thromb. Hemost. 37 (4) (2011) 408–424. [12] O.B. Ekmekçi, H. Ekmekçi, Vitronectin in atherosclerotic disease, Clin. Chim. Acta 368 (1–2) (2006) 77–83. [13] K.T. Preissner, D. Seiffert, Role of vitronectin and its receptors in haemostasis and vascular remodeling, Thromb. Res. 89 (1) (1998) 1–21. [14] S.A. Mousa, Vitronectin receptors in vascular disorders, Curr. Opin. Investig. Drugs 3 (8) (2002) 1191–1195. [15] J. Nip, P. Brodt, The role of the integrin vitronectin receptor, alpha v beta 3 in melanoma metastasis, Cancer Metastasis Rev. 14 (3) (1995) 241–252. [16] C.L. Gladson, D.A. Cheresh, Glioblastoma expression of vitronectin and the alpha v beta 3 integrin. Adhesion mechanism for transformed glial cells, J. Clin. Invest. 88 (6) (1991) 1924–1932.
59
[17] J.H. Uhm, N.P. Dooley, A.P. Kyritsis, J.S. Rao, C.L. Gladson, Vitronectin, a glioma-derived extracellular matrix protein, protects tumor cells from apoptotic death, Clin. Cancer Res. 5 (6) (1999) 1587–1594. [18] Y. Fukushima, M. Tamura, H. Nakagawa, K. Itoh, Induction of glioma cell migration by vitronectin in human serum and cerebrospinal fluid, J. Neurosurg. 107 (3) (2007) 578–585. [19] P.M. Menon, J.A. Gutierrez, S.A. Rempel, A study of SPARC and vitronectin localization and expression in pediatric and adult gliomas: high SPARC secretion correlates with decreased migration on vitronectin, Int. J. Oncol. 17 (4) (2000) 683–693. [20] J.M. Milstein, M.E. Cohen, L.F. Sinks, The influence and reliability of neurologic assessment and Karnofsky performance score on prognosis, Cancer 56 (Suppl. 7) (1985) 1834–1836. [21] M. Weller, M. van den Bent, K. Hopkins, J.C. Tonn, R. Stupp, A. Falini, et al., EANO guideline for the diagnosis and treatment of anaplastic gliomas and glioblastoma, Lancet Oncol. 15 (9) (2014) e395–e403. [22] D. Brandsma, M.J. van den Bent, Pseudoprogression and pseudoresponse in the treatment of gliomas, Curr. Opin. Neurol. 22 (6) (2009) 633–638. [23] M. Kadowaki, T. Sangai, T. Nagashima, M. Sakakibara, H. Yoshitomi, S. Takano, et al., Identification of vitronectin as a novel serum marker for early breast cancer detection using a new proteomic approach, J. Cancer Res. Clin. Oncol. 137 (7) (2011) 1105–1115. [24] S. Hess, S.M. Kanse, C. Kost, K.T. Preissner, The versatility of adhesion receptor ligands in haemostasis: morpho-regulatory functions of vitronectin, Thromb. Haemost. 74 (1) (1995) 258–265. [25] B. Felding-Habermann, D.A. Cheresh, Vitronectin and its receptors, Curr. Opin. Cell Biol. 5 (5) (1993) 864–868. [26] K.T. Preissner, D. Jenne, Vitronectin: a new molecular connection in haemostasis, Thromb. Haemost. 66 (2) (1991) 189–194. [27] K.T. Preissner, Structure and biological role of vitronectin, Annu. Rev. Cell Biol. 7 (1991) 275–310. [28] I. Ortega-Martínez, J. Gardeazabal, A. Erramuzpe, A. Sanchez-Diez, J. Cortés, M.D. García-Vázquez, et al., Vitronectin and dermcidin serum levels predict the metastatic progression of AJCC I-II early-stage melanoma, Int. J. Cancer 139 (7) (2016) 1598–1607. [29] F. Tas, S. Karabulut, E. Bilgin, D. Tastekin, D. Duranyildiz, Clinical significance of serum fibronectin and vitronectin levels in melanoma patients, Melanoma Res. 24 (5) (2014) 475–479. [30] C.L. Gladson, J.N. Wilcox, L. Sanders, G.Y. Gillespie, D.A. Cheresh, Cerebral microenvironment influences expression of the vitronectin gene in astrocytic tumors, J. Cell Sci. 108 (Pt 3) (1995) 947–956.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Diagnostic and prognostic value of serum vitronectin levels in human glioma☆ Mao-Hua Chen a, Chuan Lu a, Jun Sun a, Xian-Dong Chen a, Jun-Xia Dai a, Jian-Yong Cai a, Xiang-Lin Chen b,⁎ a
Department of Neurosurgery, The Central Hospital of Wenzhou City, Wenzhou Medical University Dingli Clinical College, 32 Dajian Lane, Wenzhou 325000, Zhejiang Province, China Department of Cerebrovessel, The People's Hospital of Qingyuan, The Sixth Affiliated Hospital of Guangzhou Medical University, B24 Xinchengyinquan Road, Qingcheng District, Qingyuan 511518, Guangdong Province, China
b
a r t i c l e
i n f o
Article history: Received 6 August 2016 Received in revised form 30 September 2016 Accepted 14 October 2016 Available online 15 October 2016 Keywords: Serum vitronectin Glioma Prognosis Diagnosis
a b s t r a c t Objective: Vitronectin is an extracellular matrix protein, the synthesis of which by glioma cells correlates with tumor grade. The current study was designed to investigate the relationship between serum vitronectin levels and clinicopathological characteristics, diagnosis and prognosis in glioma patients. Methods: In a prospective observatory study, a total of 98 glioma patients, 98 healthy controls, 98 other non-glioma brain tumors, and 98 other non-tumor neurological diseases were recruited. Following univariate analyses, multivariate analyses were performed to explore the associations of serum vitronectin levels with survival and clinicopathological parameters. Receiver operating characteristic curve analysis was done to assess its diagnostic and prognostic predictive value. Results: Serum vitronectin levels were significantly elevated in glioma patients as compared with other groups. High Wealth Health Organization grade was independently associated with high vitronectin levels. Serum vitronectin levels could significantly distinguish glioma patients from other groups and discriminate highgrade glioma from low-grade glioma. Vitronectin levels markedly predicted 5-year progression and 5-year mortality. Moreover, serum vitronectin was identified as an independent predictor for 5-year overall survival and 5year progression-free survival as well as 5-year mortality and 5-year progression. Conclusion: Serum vitronectin may be a promising diagnostic and prognostic biomarker that can be detected in the peripheral blood of patients with glioma. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Glioma is the most common among adult primary intracranial tumors and constitutes N40% of primary neoplasms in central nervous system [1–3]. Up to date, histopathology of tumor specimens has been identified as a gold standard for the grading and prognostic assessment of glioma [4,5]. However, blood biomarkers are gaining increasing interest for diagnostic and prognostic evaluation in cancer patients [6–9]. Currently, no biomarkers in peripheral blood have been established for the routine clinical management of glioma patients. Hence, it is important to identify objective markers, particularly biomarkers that can be determined noninvasively in peripheral blood, for glioma diagnosis and prognostic prediction.
Abbreviations: AUC, area under curve; CI, confidence interval; HR, hazard ratio; KPS, Karnofsky performance score; OR, odds ratio; ROC, receiver operating characteristic; WHO, World Health Organization. ☆ Institution at which the work was performed: The Central Hospital of Wenzhou City. ⁎ Corresponding author. E-mail address: [email protected] (X.-L. Chen).
http://dx.doi.org/10.1016/j.jns.2016.10.022 0022-510X/© 2016 Elsevier B.V. All rights reserved.
Vitronectin is an adhesive protein normally found in serum and plasma, also termed as serum spreading factor and complement S-protein. Generally, vitronectin plays important roles in inflammation, cell adhesion, cell necrosis, and blood coagulation [10–12]. Interestingly, vitronectin has been shown to promote the adhesion of a wide variety of cell types in vitro. Also, as a major component or master organizer of extracellular matrix, vitronectin has been found to regulate vascular endothelial growth factor-induced angiogenesis, a key step in tumor formation [13–15]. Importantly, expression of the vitronectin gene was detected in human glioma by in situ hybridization [16]; vitronectin protected in vitro glioma cells from apoptotic death [17]; glioma cell migration was induced by vitronectin in human serum and cerebrospinal fluid [18]. Moreover, vitronectin tissue expression correlated with tumor grade of glioma [19]. It is suggested that vitronectin should be associated with the occurrence and progression of cancers and also have close relation to glioma stage. Subsequently, we postulated (1) circulating vitronectin levels might be elevated in glioma patients; (2) circulating vitronectin levels might be associated with pathological severity of glioma; (3) circulating vitronectin levels might have the potential to discriminate glioma from other diseases and identify the glioma patients at risk of poor outcome. Here, vitronectin level in serum from a
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
group of glioma patients was investigated and further was identified as a candidate circulating biomarker for the early detection and prognosis of gliomas. 2. Materials and methods 2.1. Study population This prospective, observatory study included a group of newly, histologically confirmed glioma patients between January 2007 and January 2010 at the Department of Neurosurgery, the Central Hospital of Wenzhou City, China. We excluded those patients who underwent radiotherapy or chemotherapy before admission and had prior history of any type of cancer. Additionally, immediate family members are excluded from the history of glioma. Control groups were composed of three subgroups: healthy subjects, other non-glioma brain tumors, and other non-tumor neurological diseases. We excluded those controls with other systemic diseases including autoimmune diseases, uremia, liver cirrhosis, malignancy, and chronic heart or lung disease. This study was approved by the ethics Committee at our hospital (NO. CHWC2006008). Written informed consents were obtained from the study participant or a close relative.
55
Kolmogorov-Smirnov test was carried out to investigate data distribution. The continuous variables (i.e. age and serum vitronectin levels) were normally distributed and were presented as mean ± standard. KPS score and tumor size were dichotomized based on 80 and 5 cm respectively and therefore identified as the categorical variables. Also, the categorical variables were reported as counts (percentage). Chi-square tests (or Fisher exact tests), t-tests and one-way ANOVA tests were performed for intergroup comparisons as appropriate. Post-hoc analysis was done using Least Significant Difference test. Receiver operating characteristic (ROC) curve was performed to assess discriminatory ability with estimated area under curve (AUC). A suitable cutoff value was selected from ROC curve to obtain the optimal sensitivity and specificity. A binary logistic regression analysis was performed to identify risk factors, as well as odds ratio (OR) values and related 95% confidence interval (CI) were estimated. Survival curve were made by using the KaplanMeier method, and the log-rank test was used to evaluate survival difference between the groups. Cox regression analysis was used to assess predictors related to overall survival. The hazard ratio (HZ) values and corresponding 95% CI were calculated. Statistical significance was accepted as a P-value of b0.05.
3. Results 2.2. Assessment 3.1. Clinical characteristics of study population In accordance with World Health Organization (WHO) classification, gliomas are divided to low-grade ones (WHO Grade I or II) and highgrade ones (WHO Grade III or IV) [4]. Karnofsky performance score (KPS) was utilized to assess routine life ability [20]. We also recorded some related information, for instance age, gender, tumor diameter and tumor location. The diagnosis of glioma was confirmed by histological examination. Patients were treated by standard radiotherapy and chemotherapy [21]. All patients were followed up in 3-month intervals for the first 2 years and every 6 months thereafter for up to 5 years. Overall survival time was defined as the period from the date of treatment to the date of death or last date of follow-up. Death was directly related to glioma. This excluded death by indirect but related causes, e.g. accidental trauma. In accordance with RANO criteria [22], progression was defined as an increase in 25% of the product of perpendicular diameters of enhancing lesions, a significant increase in the T2/FLAIR non-enhancing component, appearance of new lesions, and clinical deterioration not attributable to causes other than the tumor or a reduction in the corticosteroid dose. Progression-free survival was calculated from the date of initial treatment to that of disease progression or last follow-up. 2.3. Determination
This study finally enrolled a total of 98 glioma patients, 98 healthy controls, 98 other non-glioma brain tumors, and 98 other non-tumor neurological diseases. This group of glioma patients included 59 males and 39 females and had a mean age of 54.5 ± 12.0 years (range, 32–75 years). 98 healthy controls comprised 56 males and 42 females as well as had a mean age of 52.3 ± 10.3 years (range, 34–72 years). 98 patients with other non-glioma brain tumors (22 vestibular schwannomas, 42 meningiomas and 34 pituitary adenomas) consisted of 54 males and 44 females and had a mean age of 50.9 ± 11.8 years (range, 31–74 years). 98 patients with other non-tumor neurological diseases (traumatic brain injury, 42 patients; spontaneous intracerebral hemorrhage, 26 patients; aneurysmal subarachnoid hemorrhage, 30 patients) were composed of 55 males and 43 females and had a mean age of 49.2 ± 10.7 years (range, 29–71 years). All individuals were Chinese. Intergroup differences of gender and age were not statistically significant between patients and other groups. Among this group of patients consisting of 59 males and 39 females, a total of 36 patients had an age of N 60 years and 62 had an age of ≤60 years; Sixty-six patients had a KPS score of ≥80, and the remaining 32 had a score of b80. Tumor size was ≤5 cm in 65 patients and N5 cm in 33 patients. In total, 81 supratentorial gliomas and 17 infratentorial
Peripheral venous blood was obtained from the patients within 24 h before surgery and from the controls at study entry. The blood samples were immediately placed into sterile test tubes and centrifuged at 3000g for 30 min at 4 °C to collect serum. Serum was stored at −70 °C until assayed. Serum vitronectin (total 65 + 75 kDa polypeptide) levels were measured by using an enzyme-linked immunosorbent assay (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions [23]. Intra-assay precision revealed a coefficient of variation of 4.8%, with an accuracy of 99%. The coefficient of variation of interassay precision was 5.6%, with an accuracy of 97%. The vitronectin levels were calculated by a standard curve. Each serum sample was tested in duplicate and the result was interpreted as μg/mL. All determinations were performed by laboratory technicians blinded to all clinical data. 2.4. Statistical analysis Statistical analysis was done using SPSS 19.0 (SPSS Inc., Chicago, IL, USA) and MedCalc 9.6.4.0 (MedCalc Software, Mariakerke, Belgium).
Fig. 1. Comparisons of serum vitronectin levels between glioma patients and healthy controls, other non-glioma brain tumors as well as other non-tumor neurological diseases.
56
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Fig. 2. Distinguishing ability of serum vitronectin between glioma patients and healthy controls, other non-glioma brain tumors in addition to other non-tumor neurological diseases as well as between high-grade and low-grade gliomas based on receiver operating characteristic curve.
ones was found, including 54 low-grade gliomas (WHO grade I and II) and 44 high-grade gliomas (WHO grade III and IV). 3.2. Serum vitronectin level as a potential diagnostic biological marker Serum vitronectin levels were markedly elevated in all glioma patients as compared to healthy controls, other non-glioma brain tumors, or other non-tumor neurological diseases in Fig. 1. Using ROC curves, we identified the optimal serum vitronectin levels that distinguished glioma from healthy controls, other non-glioma brain tumors, or other non-tumor neurological diseases with high sensitivities and specificities, as well as significantly large AUCs (Fig. 2). 3.3. Association of serum vitronectin levels with clinicopathological features of human glioma In terms of mean value of serum vitronectin levels, the glioma patients were divided to high level group (Nmean value) and low level group (b mean value). In Table 1, as compared with serum vitronectin level b126.1 μg/mL group, serum vitronectin level N 126.1 μg/mL group had higher WHO grade, lower KPS score and larger tumor size.
The multivariate analysis demonstrated that high WHO grade was an independent risk factor for serum vitronectin level N 126.1 μg/mL (OR, 26.623; 95% CI, 7.751–91.444; P b 0.001). Also, in Fig. 3, high-grade gliomas had significantly high serum vitronectin levels, as compared to low-grade gliomas. ROC curve analysis demonstrated that serum vitronectin level discriminated high-grade gliomas from low-grade gliomas significantly (Fig. 2).
3.4. Serum vitronectin identified as a prognostic marker in patients with human glioma Tables 2 and 3 indicate that 5-year overall survival, 5-year progression-free survival, 5-year mortality or 5-year progression was correlated with serum vitronectin level, WHO grade, KPS score and tumor size. The multivariate analyses selected high WHO grade and serum vitronectin level N126.1 μg/mL as the independent predictors for them (Table 4). Fig. 4 illustrates, compared with patients with serum vitronectin level b126.1 μg/mL, those with serum vitronectin level N 126.1 μg/mL had significantly shorter 5-year overall survival time and 5-year progressionfree survival time; also, serum vitronectin levels significantly predicted 5-year mortality and 5-year progression.
Table 1 The factors associated with serum vitronectin levels in human glioma. Serum vitronectin levels
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) Karnofsky performance score (b80/≥80) World Health Organization grade (III + IV/I + II)
N126.1 μg/mL
b126.1 μg/mL
P value
Odds ratio (95% CI)
P value
35/20 23/32 46/9 25/30 23/32 41/14
24/19 13/30 35/8 8/35 9/34 3/40
0.432 0.238 0.771 0.005 0.029 b0.001
1.385 (0.613–3.130) 1.659 (0.714–3.854) 1.168 (0.409–3.335) 3.646 (1.433–9.274) 2.715 (1.094–6.741) 39.048 (10.422–146.294)
0.564 0.332 0.772 0.004 0.026 b0.001
All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group. CI indicates confidence interval.
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Fig. 3. Comparison of serum vitronectin levels between high-grade and low-grade gliomas.
4. Discussion To my best knowledge, this is the first study to measure the circulating vitronectin levels in patients with glioma. The main findings of this study were that: 1) serum vitronectin levels were significantly increased in glioma patients, as compared to healthy controls, other non-glioma brain tumors and other non-tumor neurological diseases and moreover, this biomarker had obvious ability to distinguish gliomas from other groups based on ROC curves; 2) high serum vitronectin levels had an independent relation to high-grade glioma using a binary logistic regression analysis and even, it could significantly discriminate high-grade gliomas from low-grade gliomas in accordance with ROC curve; 3) serum vitronectin was associated independently with 5-year overall survival time and 5-year progression-free survival according to Cox regression analysis as well as 5-year progression and 5-year mortal-
57
ity using binary logistic regression analysis, and serum vitronectin levels had high prognostic value for 5-year mortality and 5-year progression based on ROC curve. These findings substantialize serum vitronectin as a potential prognostic and diagnostic biomarker for glioma. Vitronectin is a glycoprotein, which is a member of the integrin family. It changes in the sera and tissues of people, and is associated with the occurrence and progression of tumor [24–27]. Up to date, there are a few papers regarding circulating vitronectin. Recent studies showed that serum vitronectin levels had significant diagnostic and prognostic value for melanoma and breast cancer [23,28,29]. It has been confirmed that vitronectin could protect glioma cells from apoptotic death [17] and vitronectin from human serum and cerebrospinal fluid induced glioma cell migration [18]. Of note, vitronectin had been localized on cultured human glioma cells [30] and human glioma tissues [16]. Interestingly, this protein expression in glioma tissues has been found to be associated with clinical grade of glioma [19]. Thus, it is postulated that serum vitronectin might be a prognostic and diagnostic biomarker for human glioma. This was designed to first assess the discriminatory ability of serum vitronectin levels in human glioma. We recruited three subgroups of controls, i.e. healthy controls, other non-glioma brain tumors and other non-tumor neurological diseases. According to ROC curve analysis and intergroup comparisons of serum vitronectin levels, it was found that serum vitronectin levels possessed high discriminatory ability for distinguish glioma from other subgroups, indicating serum vitronectin might have significant value as an diagnostic indicator for human glioma. Second, we screened the risk factors associated with serum vitronectin levels identified as categorical variable, using univariate analysis followed by a binary logistic regression analysis; it was demonstrated that high WHO grade was independently associated with high serum vitronectin levels. Also, serum vitronectin levels, emerging as a continuous variable, were compared between high grade and low grade glioma based WHO scale. Meantime, a ROC curve analysis was
Table 2 The factors associated with 5-year overall survival and 5-year progression-free survival in human glioma. 5-year overall survival
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) KPS score (b80/≥80) WHO grade (III + IV/I + II) Serum vitronectin level N126.1 μg/mL
5-year progression-free survival
Hazard ratio (95% CI)
P value
Hazard ratio (95% CI)
P value
1.414 (0.828–2.415) 1.455 (0.866–2.446) 0.684 (0.346–1.353) 2.115 (1.263–3.540) 1.865 (1.105–3.149) 2.878 (1.716–4.828) 2.823 (1.603–4.970)
0.204 0.157 0.278 0.004 0.020 b0.001 b0.001
1.153 (0.719–1.851) 1.400 (0.873–2.246) 0.583 (0.312–1.087) 2.328 (1.433–3.782) 2.074 (1.258–3.419) 3.372 (2.039–5.577) 2.915 (1.742–4.877)
0.554 0.163 0.089 0.001 0.004 b0.001 b0.001
Cox regression analysis was used to assess predictors related to survival. The hazard ratio values and corresponding 95% CI were calculated. CI indicates confidence interval; WHO, World Health Organization; KPS, Karnofsky performance score. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group.
Table 3 The factors associated with 5-year mortality and 5-year progression in human glioma. 5-year mortality
Gender (male/female) Age (N60 years/≤60 years) Glioma location (supratentorial/infratentorial) Tumor size (N5 cm/≤5 cm) KPS score (b80/≥80) WHO grade (III + IV/I + II) Serum vitronectin level N126.1 μg/mL
5-year progression
Odds ratio (95% CI)
P value
Odds ratio (95% CI)
P value
2.164 (0.937–5.000) 1.431 (0.597–3.431) 0.797 (0.250–2.544) 3.184 (1.211–8.370) 2.780 (1.054–7.333) 5.220 (2.051–13.288) 4.778 (1.988–11.484)
0.071 0.422 0.701 0.019 0.039 0.001 0.002
1.263 (0.503–3.168) 1.970 (0.702–5.530) 0.693 (0.179–2.689) 3.460 (1.077–11.121) 3.082 (1.156–9.929) 4.588 (1.554–13.545) 4.339 (1.642–11.467)
0.619 0.198 0.596 0.037 0.029 0.006 0.003
A univariate binary logistic regression analysis was performed to identify risk factors, as well as odds ratio values and related 95% CI were estimated. CI indicates confidence interval; WHO, World Health Organization; KPS, Karnofsky performance score. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1μg/mL) from the glioma group.
58
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
Table 4 Multivariate analysis of risk factors associated with 5-year poor outcome in human glioma. Serum vitronectin level N126.1 μg/mL
5-year overall survival 5-year progression-free survival 5-year mortality 5-year progression
WHO grade (III + IV/I + II)
Odds ratio (95% CI)
P value
Odds ratio (95% CI)
P value
2.502 (1.381–4.511) 2.702 (1.511–4.891) 3.399 (1.191–8.588) 3.102 (1.112–7.351)
0.004 0.002 0.007 0.009
2.612 (1.454–4.545) 2.892 (1.561–5.130) 3.509 (1.445–10.521) 3.416 (1.296–9.832)
0.002 0.001 0.005 0.006
CI indicates confidence interval; WHO, World Health Organization. All glioma patients were dichotomized to two groups based on the mean value of serum vitronectin levels (126.1 μg/mL) from the glioma group.
performed to assess discriminatory ability of vitronectin levels for different clinical grades. Actually, elevated vitronectin levels had close relation to poor clinical grade, suggesting serum vitronectin might have the potential to assess glioma malignancy degree. At last, we evaluated the prognostic value of serum vitronectin in terms of 5-year overall survival, 5-year progression-free survival, 5-year mortality and 5-year progression. Intriguingly, serum vitronectin emerged as an independent predictor for the abovementioned clinical endpoints. ROC curve analysis also showed an interesting result that serum vitronectin levels predicted 5-year mortality and 5-year progression with high sensitivities and specificities. These results substantialized serum vitronectin as a useful prognostic biomarker. Actually, it is impossible that any biomarker is only produced in a type of cell or tissue. Vitronectin is not exceptional. In this study, this biomarker had high predictive or diagnostic value based on areas under curve. Taken together, serum vitronectin might act as a diagnostic and prognostic biomarker for human glioma. However, the current study had two limitations. At first, all patients were Chinese, thus the results may not be more generally applicable to populations of other ethnic origins. Second, clinical parameters, such as pathological diagnosis, are variable between institutions and/or individual clinicians, and hence the results with this small cohort may reflect biases inherent in the acquisition of such clinical data. Clearly, these results remain to be validated in a multicenter clinical trial.
5. Conclusions Serum vitronectin levels are elevated in human glioma, as well as are independently associated with pathologic grade based on WHO classification and 5-year poor prognosis. Moreover, this biomarker significantly discriminates gliomas from controls, other non-glioma brain tumors and other non-tumor neurological diseases, and distinguishes highgrade gliomas from low-grade gliomas, as well as predicts 5-year progression and mortality. Thus, serum vitronectin has the potential to be a promising diagnostic and prognostic biomarker for glioma.
Conflict of interest The authors report no conflicts of interest concerning the materials or methods used in this study or the findings specified in this article.
Acknowledgements The authors thank all staffs in Department of Neurosurgery, The Central Hospital of Wenzhou City (Wenzhou, China) for their technical support. This study was supported financially by Zhejiang Province Medical and Health Project (No. 2014RCA026, 2015ZDA028, 2016RCA026).
Fig. 4. Survival curves of glioma patients in accordance with serum vitronectin levels and predictive value of serum vitronectin levels for 5-year poor outcome based on receiver operating characteristic curve.
M.-H. Chen et al. / Journal of the Neurological Sciences 371 (2016) 54–59
References [1] P.Y. Wen, S. Kesari, Malignant gliomas in adults, N. Engl. J. Med. 359 (5) (2008) 492–507. [2] Y. Iwadate, T. Sakaida, T. Hiwasa, Y. Nagai, H. Ishikura, M. Takiguchi, et al., Molecular classification and survival prediction in human gliomas based on proteome analysis, Cancer Res. 64 (7) (2004) 2496–2501. [3] A. Omuro, L.M. DeAngelis, Glioblastoma and other malignant gliomas: a clinical review, JAMA 310 (17) (2013) 1842–1850. [4] D.N. Louis, H. Ohgaki, O.D. Wiestler, W.K. Cavenee, P.C. Burger, A. Jouvet, et al., The 2007 WHO classification of tumours of the central nervous system, Acta Neuropathol. 114 (2) (2007) 97–109. [5] M.J. Riemenschneider, J.W. Jeuken, P. Wesseling, G. Reifenberger, Molecular diagnostics of gliomas: state of the art, Acta Neuropathol. 120 (5) (2010) 567–584. [6] S.J. Lee, J.W. Seo, Y.S. Chae, J.G. Kim, B.W. Kang, W.W. Kim, et al., Genetic polymorphism of miR-196a as a prognostic biomarker for early breast cancer, Anticancer Res. 34 (6) (2014) 2943–2949. [7] X. Chen, Y. Ba, L. Ma, X. Cai, Y. Yin, K. Wang, et al., Characterization of microRNAs in serum: a novel class of biomarkers for diagnosis of cancer and other diseases, Cell Res. 18 (10) (2008) 997–1006. [8] V. Zigmont, A. Garrett, J. Peng, M. Seweryn, G.A. Rempala, R. Harris, et al., Association between Prediagnostic serum 25-Hydroxyvitamin D concentration and glioma, Nutr. Cancer 67 (7) (2015) 1120–1130. [9] M.B. Nijaguna, V. Patil, A.S. Hegde, B.A. Chandramouli, A. Arivazhagan, V. Santosh, et al., An eighteen serum cytokine signature for discriminating glioma from normal healthy individuals, PLoS One 10 (9) (2015), e0137524. [10] D.I. Leavesley, A.S. Kashyap, T. Croll, M. Sivaramakrishnan, A. Shokoohmand, B.G. Hollier, et al., Vitronectin–master controller or micromanager? IUBMB Life 65 (10) (2013) 807–818. [11] K.T. Preissner, U. Reuning, Vitronectin in vascular context: facets of a multitalented matricellular protein, Semin. Thromb. Hemost. 37 (4) (2011) 408–424. [12] O.B. Ekmekçi, H. Ekmekçi, Vitronectin in atherosclerotic disease, Clin. Chim. Acta 368 (1–2) (2006) 77–83. [13] K.T. Preissner, D. Seiffert, Role of vitronectin and its receptors in haemostasis and vascular remodeling, Thromb. Res. 89 (1) (1998) 1–21. [14] S.A. Mousa, Vitronectin receptors in vascular disorders, Curr. Opin. Investig. Drugs 3 (8) (2002) 1191–1195. [15] J. Nip, P. Brodt, The role of the integrin vitronectin receptor, alpha v beta 3 in melanoma metastasis, Cancer Metastasis Rev. 14 (3) (1995) 241–252. [16] C.L. Gladson, D.A. Cheresh, Glioblastoma expression of vitronectin and the alpha v beta 3 integrin. Adhesion mechanism for transformed glial cells, J. Clin. Invest. 88 (6) (1991) 1924–1932.
59
[17] J.H. Uhm, N.P. Dooley, A.P. Kyritsis, J.S. Rao, C.L. Gladson, Vitronectin, a glioma-derived extracellular matrix protein, protects tumor cells from apoptotic death, Clin. Cancer Res. 5 (6) (1999) 1587–1594. [18] Y. Fukushima, M. Tamura, H. Nakagawa, K. Itoh, Induction of glioma cell migration by vitronectin in human serum and cerebrospinal fluid, J. Neurosurg. 107 (3) (2007) 578–585. [19] P.M. Menon, J.A. Gutierrez, S.A. Rempel, A study of SPARC and vitronectin localization and expression in pediatric and adult gliomas: high SPARC secretion correlates with decreased migration on vitronectin, Int. J. Oncol. 17 (4) (2000) 683–693. [20] J.M. Milstein, M.E. Cohen, L.F. Sinks, The influence and reliability of neurologic assessment and Karnofsky performance score on prognosis, Cancer 56 (Suppl. 7) (1985) 1834–1836. [21] M. Weller, M. van den Bent, K. Hopkins, J.C. Tonn, R. Stupp, A. Falini, et al., EANO guideline for the diagnosis and treatment of anaplastic gliomas and glioblastoma, Lancet Oncol. 15 (9) (2014) e395–e403. [22] D. Brandsma, M.J. van den Bent, Pseudoprogression and pseudoresponse in the treatment of gliomas, Curr. Opin. Neurol. 22 (6) (2009) 633–638. [23] M. Kadowaki, T. Sangai, T. Nagashima, M. Sakakibara, H. Yoshitomi, S. Takano, et al., Identification of vitronectin as a novel serum marker for early breast cancer detection using a new proteomic approach, J. Cancer Res. Clin. Oncol. 137 (7) (2011) 1105–1115. [24] S. Hess, S.M. Kanse, C. Kost, K.T. Preissner, The versatility of adhesion receptor ligands in haemostasis: morpho-regulatory functions of vitronectin, Thromb. Haemost. 74 (1) (1995) 258–265. [25] B. Felding-Habermann, D.A. Cheresh, Vitronectin and its receptors, Curr. Opin. Cell Biol. 5 (5) (1993) 864–868. [26] K.T. Preissner, D. Jenne, Vitronectin: a new molecular connection in haemostasis, Thromb. Haemost. 66 (2) (1991) 189–194. [27] K.T. Preissner, Structure and biological role of vitronectin, Annu. Rev. Cell Biol. 7 (1991) 275–310. [28] I. Ortega-Martínez, J. Gardeazabal, A. Erramuzpe, A. Sanchez-Diez, J. Cortés, M.D. García-Vázquez, et al., Vitronectin and dermcidin serum levels predict the metastatic progression of AJCC I-II early-stage melanoma, Int. J. Cancer 139 (7) (2016) 1598–1607. [29] F. Tas, S. Karabulut, E. Bilgin, D. Tastekin, D. Duranyildiz, Clinical significance of serum fibronectin and vitronectin levels in melanoma patients, Melanoma Res. 24 (5) (2014) 475–479. [30] C.L. Gladson, J.N. Wilcox, L. Sanders, G.Y. Gillespie, D.A. Cheresh, Cerebral microenvironment influences expression of the vitronectin gene in astrocytic tumors, J. Cell Sci. 108 (Pt 3) (1995) 947–956.