Post-surgical resection prognostic value of combined OPN, MMP7, and PSG9 plasma biomarkers in hepatocellular carcinoma

Abstract

Biomarkers for hepatocellular carcinoma (HCC) following curative resection are not currently sufficient for prognostic indication of overall survival (OS) and disease-free survival (DFS). The aim of this study was to investigate the prognostic performance of osteopontin (OPN), matrix metalloproteinase 7 (MMP7), and pregnancy specific glycoprotein 9 (PSG9) in patients with HCC. A total of 179 prospective patients with HCC provided plasma before hepatectomy. Plasma OPN, MMP7, and PSG9 levels were determined by enzyme-linked immunosorbent assay. Correlations between plasma levels, clinical parameters, and outcomes (OS and DFS) were overall analyzed. High OPN (≥149.97 ng/mL), MMP7 (≥2.28 ng/mL), and PSG9 (≥45.59 ng/mL) were prognostic indicators of reduced OS (P<0.001, P<0.001, and P=0.007, respectively). Plasma PSG9 protein level was an independent factor in predicting OS (P=0.008) and DFS (P=0.038). Plasma OPN+MMP7+PSG9 elevation in combination was a prognostic factor for OS (P<0.001). OPN was demonstrated to be a risk factor-associated OS in stage I patients with HCC and patients with low α-fetoprotein levels (<20 ng/mL). These findings suggested that OPN, MMP7, PSG9 and their combined panels may be useful for aiding in tumor recurrence and mortality risk prediction of patients with HCC, particularly in the early stage of HCC carcinogenesis.

Keywords

biomarkers / OPN / MMP7 / PSG9 / HCC / prognosis

Introduction

Hepatocellular carcinoma (HCC) is the most common primary liver tumor, having an incidence virtually identical to mortality, resulting in up to 700 000 new diagnoses and 600€000 deaths per annum worldwide [¹]. In many countries, HCC is the most common type of cancer, in part due to relatively high hepatitis B virus (HBV) infection rates and limited availability of recent antiviral agents. Although the mechanistic role of chronic HBV infection in HCC remains unclear, the relationship between HCC and chronic HBV infection is well documented [²]. Meanwhile, surgical resection can be an effective treatment for patients with HCC, and the efficacy of such treatment is limited in patients with late stage HCC diagnoses [³]. The reported 5-year disease-free survival (DFS) rate is only 25%–30%, and recurrence often leads to early mortality [⁴,⁵]. Current post-surgical prognostic techniques do not have sufficient ability to accurately predict patient outcomes. Thus, clinically useful biomarkers of survival and metastatic recurrence in HCC are urgently needed.

Recently, several studies have proposed α-fetoprotein (AFP) as a biomarker of post-surgical outcomes in HBV-associated HCC (HBV-HCC); the 5-year survival correlated with low (1.4–4.1 ng/mL) AFP levels, and thereafter, outcomes progressively declined with increasing AFP levels [⁶–⁸]. Clinical usefulness of plasma AFP detection is limited by the practical datable range (>20 ng/mL), particularly in early-stage HCC, because even minute differences in AFP within the normal range can have an effect on prognosis and progression in HBV-HCC [⁸,⁹]. Numerous serum protein markers, such as lectin-bound AFP (AFP-L3), des-γ carboxyprothrombin (DCP), glypican-3, Dickkopf-1, and Golgi protein 73 (GP73), have been identified as alternative prognostic biomarkers for prediction of HCC diagnosis or prognosis [¹⁰–¹²]. Nonetheless, few prognostic markers have been widely adopted in regular clinical practice over the last decade although isolated examples exist on adoption of tumor serologic markers AFP-L3 and DCP used for clinical diagnostic purpose, particularly in Japan where such practice is endorsed by the Japan Society of Hepatology [¹³]. Meanwhile, AFP-L3% exceeding 5% can be used in the detection of small tumors and early stage HCC, and neither AFP-L3 nor DCP has been found to have greater prognostic value than AFP in early detection and recurrence prediction [¹⁰], necessitating development of several sensitive prognostic methods for clinical practice. HCC development has also been linked to OPN, a glycoprotein secreted into the extracellular matrix [¹⁴–¹⁷]. OPN has been shown to play an important role in tumor progression, especially in metastasis as an inducer of tumor-cell proliferation, migration, and extracellular matrix invasion [¹⁸]. Furthermore, downstream signaling pathways and effectors, including PI 3′-kinase/Akt, nuclear factor-kB, matrix metalloproteinases (MMPs), and urokinase-type plasminogen activator (uPA), are activated by OPN, which has been linked to the promotion of tumor metastasis in other cancer types [¹⁸]. Thus, OPN may be a promising candidate as a diagnostic and prognostic biomarker and may be an interesting therapeutic target for limiting metastasis.

Previously, the authors reported a novel approach for the detection of human blood proteins based on the secretome of cancerous tissues [¹⁹], a finding that may enable sensitive tumor-biomarker detection. Using existent data in an HCC secretome database of 1365 proteins, osteopontin (OPN), matrix metalloproteinase 7 (MMP7), and pregnancy-specific β-1-glycoprotein 9 (PSG9) were found to be the key participants in the transforming growth factor β pathway, a pivotal pathway that acts as a tumor suppressor in normal or dysplastic cells and a tumor promoter in advanced cancers [²⁰]. Furthermore, plasma levels of OPN, MMP7, and PSG9 are readily assessed in patients with primary HCC, potentially providing a clinically applicable biomarker set.

Materials and methods

Patients and specimens

Patients with HCC who underwent hepatectomy were enrolled in this study. All subjects had definite postoperative pathological diagnosis and confirmed as HCC. The patients whose Child score was A underwent hepatectomy in this study. Plasma was collected prospectively from 179 patients with HCC (40 females, mean age 54.71±11.24 years) prior to hepatectomy at the Cancer Hospital of Peking Union Medical College and Chinese Academy of Medical Sciences from March 2009 to June 2011. Peripheral blood samples were collected by venipuncture and preserved in EDTA-coated tubes. Samples were centrifuged at 4 °C for 10 min at 1000 g to separate plasma from blood cells. Supernatants were collected, divided into aliquots, and stored at −80 °C until use [²¹]. All procedures related to human samples were approved by the Review Board of Chinese Academy of Medical Sciences Cancer Institute. All sample subjects provided informed consent, and the study was conducted under the approval of the Institutional Ethics Committee.

HCC diagnosis

HCC was diagnosed based on abdominal ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and/or biochemical profiles (AFP serology and liver function enzymes). Diagnoses were post-surgically confirmed by histopathology according to the criteria proposed by the American Association for the Study of Liver Disease guidelines [²²]. HCC stage was defined according to the American Joint Committee on Cancer/Cancer Staging Manual (7th edition) by using tumor node metastasis (TNM) staging criteria, where stage I (single tumor, no vascular invasion) HCC was classified as early stage [²³]. Diagnosis of chronic hepatitis B (CHB) and liver cirrhosis were conducted as previously described by Shen et al. [²⁴].

Follow-up

Patients were followed until September 2015. Overall survival (OS) was defined as the interval between surgery and death. Only the patients died of tumor progression were enrolled. For patients suspected of recurrent cancers, recurrence or metastasis was verified by CT and/or MRI examinations. Recurrence or metastasis diagnoses were based on routine imaging and elevated AFP levels. Cause of death (recurrence, metastasis, or complicated liver cirrhosis) was recorded. DFS was defined as the interval between surgery and recurrence. If recurrence was not diagnosed, the date of death or last follow-up was recorded.

Plasma OPN, MMP7, and PSG9 assessment

OPN, MMP7, and PSG9 plasma protein concentrations were assessed using enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instructions. ELISA kits for OPN and MMP7 were purchased from R&D Systems Inc. (Minneapolis, MN, USA). ELISA kits for PSG9 were purchased from BlueGene (Shanghai, China). In brief, 100 mL of diluted plasma for OPN (1:20) or MMP7 (1:2) was added to each well in the antibody precoated microtiter plate. For PSG9, 50 mL of diluted plasma (1:2) was added to each well. Then, conjugated antibodies were added to each well and mixed. The results were washed to remove excess plasma, and color development was achieved. Absorbance was measured immediately using a microplate reader (Bio-Rad Laboratory, Hercules, CA, USA). AFP levels were tested using a commercial electrochemiluminescent immunoassay kit (Roche Diagnostics, Mannheim, Germany) at the Clinical Diagnostic Laboratories at the Cancer Hospital of the Chinese Academy of Medical Sciences.

Statistical analysis

All statistical analyses were performed using SPSS software, version 17.0 (SPSS Inc., IL, USA) except for the time-dependent receiver operating characteristic (ROC) analysis, which was performed with R3.1.1. Continuous data for OPN, MMP7, and PSG9 plasma levels were presented as mean±standard deviations, and relationships between plasma levels of the three proteins and clinical parameters were analyzed by Mann–Whitney test. DFS and OS rates by plasma levels of the selected proteins were assessed by log-rank test and the Kaplan–Meier curves. Variables determined by univariate analysis were entered in a multivariate analysis performed using a Cox multivariate analysis hazards model with forward stepwise selection. Optimal cutoff for OPN, MMP7, and PSG9 was determined with time-dependent ROC curve analysis, as previously described by Lambert et al. [²⁵]. P-values less than 0.05 were considered statistically significant (P<0.05).

Results

Plasma OPN, MMP7, and PSG9 levels and clinical parameters

Plasma OPN, MMP7, and PSG9 levels in patients with HCC were found to be 170.37±145.44, 3.29±10.40, and 30.69±93.99 ng/mL, respectively. Plasma OPN and MMP7 levels were high in elderly patients, particularly in patients aged>55 years (P=0.007 and P=0.046, respectively). Plasma OPN levels were significantly elevated in the patients with high TNM stages of II–IV (P=0.006). OPN plasma levels were significantly low in patients with tumor size<3 cm (P<0.001), possibly indicative of a role in tumor growth. Furthermore, MMP7 plasma levels in patients with multiple tumors were significantly high (P=0.002). The ELISA results and the patient characteristics are summarized in Table 1.

Prognostic value of plasma OPN, MMP7, and PSG9 levels

At the end of the follow-up period (5–76 months, mean 30 months), tumor recurrence was identified in 77 (43.02%) patients, and 29 (16.20%) patients died following tumor progression. Prognostic significance of plasma OPN, MMP7, PSG9, and AFP levels was determined by time-dependent ROC curve, where area under the ROC curve (AUC) corresponded to OS time (Fig. 1A). OPN AUC was highest, whereas PSG9 exhibited the lowest AUC. These findings indicated that AFP is a suboptimal prognostic biomarker for HCC. Cutoff values were calculated from AUC based on time-dependent ROC curves, with OPN, MMP7, PSG9 and AFP intervals at the 30th month (Fig. 1A). Based on the time-dependent ROC curve, cutoff values were determined for OPN (149.97 ng/mL), MMP7 (2.28 ng/mL), PSG9 (45.59 ng/mL), and AFP (95.97 ng/mL) (Fig. 1B–1E). AUC values for OPN, MM7, PSG9, and AFP were 0.704, 0.574, 0.542, and 0.571, respectively.

Comparison of Kaplan–Meier curves indicated low OS rates in patients with high OPN (≥149.97 ng/mL), MMP7 (≥2.28 ng/mL), and PSG9 (≥45.59 ng/mL) (Fig. 2A–2C).

Two patient subgroups were readily identified in results, including (1) low, all three protein plasma levels were below cutoff (n=86, 48.04%; OPN<149.97 ng/mL, MMP7<2.28 ng/mL, PSG9<45.59 ng/mL) and (2) high, at least one of the three protein plasma level was above cutoff (n=93, 51.96%). Low group (n=86): median OS time was 30 months, and median DFS time was 26 months. High group (n=93): median OS time was 21 months, and median DFS time was 17 months. Kaplan–Meier curves indicated low OS rates in the high group (Fig. 2D).

OPN (P<0.001), MMP7 (P=0.009), PSG9 (P=0.011), and their combinations (P<0.001) were potential predictive biomarkers of OS according to Cox univariate analyses. OPN (P<0.001) and MMP7 (P=0.013), but not PSG9 (P=0.059), were potential biomarkers for DFS. Multivariate analyses of clinical parameters demonstrated that PSG9 was an independent prognostic factor for both OS (P=0.008) and DFS (P=0.038), and differentiation was an independent prognostic factor for DFS, P=0.019 (Table 2).

OPN, MMP7, and PSG9 for prognostic prediction in early-stage HCC

An additional analysis was conducted on a subgroup of 86 early-stage (stage I) patients with HCC. Patients with high plasma OPN levels had significantly shorter OS compared with patients with low OPN levels (Fig. 3A). Patients of the group “high” had a lower OS compared with patients of group “low.” Notably, OPN was demonstrated to be a risk factor associated with OS in patients with AFP<20 ng/mL (Fig. 3B). Furthermore, although plasma protein levels above the cutoff exhibited short OS time in patients with AFP<20 ng/mL, the combination does not show better performance than OPN.

Discussion

The current study examined plasma levels of OPN, MMP7, and PSG9, revealing that some or all these indicators may be elevated in patients with poor OS. Furthermore, the combination of these three biomarkers could be used to predict early recurrence and survival, demonstrating to be effective in patients with early-stage (stage I) HCC. These findings suggested that combined plasma OPN, MMP7, and PSG9 levels may be useful targets in the development of prognostic panels that can aid in tumor recurrence and mortality-risk-prediction patients with HCC, particularly in the early stage of HCC carcinogenesis.

In clinical treatment of HCC, relatively high rates of metastasis and recurrence contribute to overall poor survival in these patients, and few treatments are currently available to limit metastasis-related recurrence [²⁶]. One of the key limitations to improving clinical care is the inability to distinguish between patients that will exhibit recurrence after surgery, particularly in patients with early-stage disease without significant vascular invasion and regional or distant recurrence. As Cheng et al. [⁵] suggested, identifying risk factors for morality is important for effective surveillance and selection of adjunctive therapies in the postoperative period in patients with liver cancer. In the current study, 16.20% of patients died because of tumor progression during the follow-up period. In studies that have examined HCC over long time periods of up to 5 years, the OS rate progressively declines with time as tumor recurrence increases [²⁷]. Meanwhile, similar initial OS rates were observed, and long follow-up would be necessary to demonstrate the sustained prognostic ability of these serum markers throughout the months and years of clinical surveillance following resection. Notably, the current study identified PSG9 as independent prognostic biomarkers for OS and DFS in HCC, a combination not previously described. Furthermore, these findings indicated that combined detection of these three proteins can improve prognostic detection of DFS and OS, and that detection of high OPN, MMP7, and/or PSG9 levels could be useful in detecting poor OS and recurrence.

Within the limited follow-up window of the present study, 43.02% of patients with HCC exhibited recurrence after curative resection surgery, with a large portion exhibiting elevated OPN levels. Previously, Sieghart et al. [²⁷] reported that OPN expression could serve as a predictor of HCC recurrence following liver transplantation in patients with HCC. The results of this study confirmed that elevated plasma levels of OPN are associated with great recurrence and poor OS, particularly in patients with early stage cancer and AFP levels below 20 ng/mL. In addition, these findings were well aligned with the results in other tumor types that indicate OPN may be linked recurrence and reduced OS in gastric, breast, lung, colon, and ovarian cancers [²⁸]. The current study results provided novel evidence that plasma OPN levels may be indicators for tumor size, proliferation, and progression (Table 1), which are well aligned with prior mechanistic descriptions on the role of OPN in metastatic processes of tumor cell proliferation and migration [²⁷–²⁹]. Despite its potential as a prognostic marker, genetic heterogeneity provides reasonability to expect that abnormal OPN levels in the plasma will not always be apparent in early stage HCC for all patients.

Researchers have begun to propose combination methods for determining prognostic in patients with HCC that may be able to increase efficacy of these methods because no single serum biomarker elevation is detectable in all HCC cases [³⁰]. In this study, serum OPN, MMP7, and PSG9 combinations were selected based on information from a previous study, providing evidence that these markers are routinely elevated in patients with HCC [²⁰]. As demonstrated in the present study, previous studies of liver disease linked the MMP7 to capsular infiltration, portal vein invasion, and disease progression and recurrence although MMP7 levels may not always immediately be detectable in the peripheral blood supply at early disease stages [³¹,³²]. MMP7 has also been identified as a regulator of liver matric turnover [³³] and of metastasis-related genes [³⁴], but it has rarely been reported in the context of HCC progression. According to the authors’ previous work, MMP7 may also be a potential diagnostic marker for hepatic cirrhosis patients with HBV, consistent with the current study results [²⁰].

Similarly, the carcinoembryonic antigen, PSG9, is a highly glycosylated protein known to be deregulated during colorectal carcinogenesis and active in regulating immunoreaction and angiogenesis, consistent with the current study findings [³⁵,³⁶]. PSG9 protein level was found to be an independent prognostic factor for OS and DFS in patients with HCC, a result not previously described and possibly mechanistically related to tumor cell invasion and migration. Notably, serum PSG9 elevation (≥45.59 ng/mL) was observed in only 17 patients with HCC, and further research in large HCC patient cohorts would be necessary to verify trends in PSG9 and elucidate its full mechanism of action.

Our result showed that combined panels may be useful for aiding in tumor recurrence and mortality risk prediction of patients with HCC, particularly in the early stage of HCC carcinogenesis, because 60%–70% of patients with HCC still displayed low AFP levels during early stage disease, which can confound prognosis [³⁷,³⁸]. Additionally, some studies suggested that two biologically distinct recurrence mechanisms exist in HCC, responsible for early and late recurrence [³⁹,⁴⁰]. Discrimination between such recurrence mechanisms may not be apparent because this study only examined a relatively short follow-up period. Thus, further analysis is needed to validate these biomarkers in large-scale cohorts inclusive of robust follow-up periods. Moreover, the data of this study remain preliminary and will need to be re-evaluated by further studies.

In conclusion, high plasma levels of OPN, MMP7, and PSG9 in the peripheral blood were associated with high incidence of early recurrence and poor OS after curative resection for HCC. The combination of these three plasma protein biomarkers has superior prognostic predictive ability. Notably, this study is a valuable preliminary step toward achieving a conclusive evidence for combination serum biomarker prognostic use in HCC. Results may lead to the development of clinically viable techniques for detecting serum OPN, MMP7, and PSG9 levels as prognostic biomarkers of HCC.

References

[1] Torre LA, Bray F, Siegel RL, Ferlay J, Lortet-Tieulent J, Jemal A. Global cancer statistics, 2012. CA Cancer J Clin 2015; 65(2): 87–108

[2] Lu T, Seto WK, Zhu RX, Lai CL, Yuen MF. Prevention of hepatocellular carcinoma in chronic viral hepatitis B and C infection. World J Gastroenterol 2013; 19(47): 8887–8894

[3] Kumada T, Toyoda H, Tada T, Kiriyama S, Tanikawa M, Hisanaga Y, Kanamori A, Tanaka J, Kagebayashi C, Satomura S. High-sensitivity Lens culinaris agglutinin-reactive α-fetoprotein assay predicts early detection of hepatocellular carcinoma. J Gastroenterol 2014; 49(3): 555–563

[4] Kornberg A, Witt U, Matevossian E, Küpper B, Assfalg V, Drzezga A, Hüser N, Wildgruber M, Friess H. Extended postinterventional tumor necrosis-implication for outcome in liver transplant patients with advanced HCC. PLoS One 2013; 8(1): e53960

[5] Cheng J, Wang W, Sun C, Li M, Wang B, Lv Y. Meta-analysis of the prognostic and diagnostic significance of serum/plasma osteopontin in hepatocellular carcinoma. J Clin Gastroenterol 2014; 48(9): 806–814

[6] Tateishi R, Shiina S, Yoshida H, Teratani T, Obi S, Yamashiki N, Yoshida H, Akamatsu M, Kawabe T, Omata M. Prediction of recurrence of hepatocellular carcinoma after curative ablation using three tumor markers. Hepatology 2006; 44(6): 1518–1527

[7] Lugassy C, Wadehra M, Li X, Corselli M, Akhavan D, Binder SW, Péault B, Cochran AJ, Mischel PS, Kleinman HK, Barnhill RL. Pilot study on “pericytic mimicry” and potential embryonic/stem cell properties of angiotropic melanoma cells interacting with the abluminal vascular surface. Cancer Microenviron 2013; 6(1): 19–29

[8] Blank S, Wang Q, Fiel MI, Luan W, Kim KW, Kadri H, Mandeli J, Hiotis SP. Assessing prognostic significance of preoperative α-fetoprotein in hepatitis B-associated hepatocellular carcinoma: normal is not the new normal. Ann Surg Oncol 2014; 21(3): 986–994

[9] Shimada K, Sakamoto Y, Esaki M, Kosuge T, Morizane C, Ikeda M, Ueno H, Okusaka T, Arai Y, Takayasu K. Analysis of prognostic factors affecting survival after initial recurrence and treatment efficacy for recurrence in patients undergoing potentially curative hepatectomy for hepatocellular carcinoma. Ann Surg Oncol 2007; 14(8): 2337–2347

[10] Marrero JA, Feng Z, Wang Y, Nguyen MH, Befeler AS, Roberts LR, Reddy KR, Harnois D, Llovet JM, Normolle D, Dalhgren J, Chia D, Lok AS, Wagner PD, Srivastava S, Schwartz M. α-fetoprotein, des-γ carboxyprothrombin, and lectin-bound α-fetoprotein in early hepatocellular carcinoma. Gastroenterology 2009; 137(1): 110–118

[11] Fu SJ, Qi CY, Xiao WK, Li SQ, Peng BG, Liang LJ. Glypican-3 is a potential prognostic biomarker for hepatocellular carcinoma after curative resection. Surgery 2013; 154(3): 536–544

[12] Mao Y, Yang H, Xu H, Lu X, Sang X, Du S, Zhao H, Chen W, Xu Y, Chi T, Yang Z, Cai J, Li H, Chen J, Zhong S, Mohanti SR, Lopez-Soler R, Millis JM, Huang J, Zhang H. Golgi protein 73 (GOLPH2) is a valuable serum marker for hepatocellular carcinoma. Gut 2010; 59(12): 1687–1693

[13] Kobayashi M, Hosaka T, Ikeda K, Seko Y, Kawamura Y, Sezaki H, Akuta N, Suzuki F, Suzuki Y, Saitoh S, Arase Y, Kumada H. Highly sensitive AFP-L3% assay is useful for predicting recurrence of hepatocellular carcinoma after curative treatment pre- and postoperatively. Hepatol Res 2011; 41(11): 1036–1045

[14] Liu J, Xu K, Chase M, Ji Y, Logan JK, Buchsbaum RJ. Tiam1-regulated osteopontin in senescent fibroblasts contributes to the migration and invasion of associated epithelial cells. J Cell Sci 2012; 125(Pt 2): 376–386

[15] Syn WK, Agboola KM, Swiderska M, Michelotti GA, Liaskou E, Pang H, Xie G, Philips G, Chan IS, Karaca GF, Pereira TA, Chen Y, Mi Z, Kuo PC, Choi SS, Guy CD, Abdelmalek MF, Diehl AM. NKT-associated hedgehog and osteopontin drive fibrogenesis in non-alcoholic fatty liver disease. Gut 2012; 61(9): 1323–1329

[16] Sun J, Xu HM, Zhou HJ, Dong QZ, Zhao Y, Fu LY, Hei ZY, Ye QH, Ren N, Jia HL, Qin LX. The prognostic significance of preoperative plasma levels of osteopontin in patients with TNM stage-I of hepatocellular carcinoma. J Cancer Res Clin Oncol 2010; 136(1): 1–7

[17] Shang S, Plymoth A, Ge S, Feng Z, Rosen HR, Sangrajrang S, Hainaut P, Marrero JA, Beretta L. Identification of osteopontin as a novel marker for early hepatocellular carcinoma. Hepatology 2012; 55(2): 483–490

[18] Philip S, Kundu GC. Osteopontin induces nuclear factor kappa B-mediated promatrix metalloproteinase-2 activation through IκBα/IKK signaling pathways, and curcumin (diferulolylmethane) down-regulates these pathways. J Biol Chem 2003; 278(16): 14487–14497

[19] Xiao T, Ying W, Li L, Hu Z, Ma Y, Jiao L, Ma J, Cai Y, Lin D, Guo S, Han N, Di X, Li M, Zhang D, Su K, Yuan J, Zheng H, Gao M, He J, Shi S, Li W, Xu N, Zhang H, Liu Y, Zhang K, Gao Y, Qian X, Cheng S. An approach to studying lung cancer-related proteins in human blood. Mol Cell Proteomics 2005; 4(10): 1480–1486

[20] Yang L, Rong W, Xiao T, Zhang Y, Xu B, Liu Y, Wang L, Wu F, Qi J, Zhao X, Wang H, Han N, Guo S, Wu J, Gao Y, Cheng S. Secretory/releasing proteome-based identification of plasma biomarkers in HBV-associated hepatocellular carcinoma. Sci China Life Sci 2013; 56(7): 638–646

[21] Zhang Y, Xu B, Liu Y, Yao H, Lu N, Li B, Gao J, Guo S, Han N, Qi J, Zhang K, Cheng S, Wang H, Zhang X, Xiao T, Wu L, Gao Y. The ovarian cancer-derived secretory/releasing proteome: a repertoire of tumor markers. Proteomics 2012; 12(11): 1883–1891

[22] Bruix J, Sherman M; American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update. Hepatology 2011; 53(3): 1020–1022

[23] Edge SB, Compton CC. The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM. Ann Surg Oncol 2010; 17(6):1471–1474

[24] Shen Q, Fan J, Yang XR, Tan Y, Zhao W, Xu Y, Wang N, Niu Y, Wu Z, Zhou J, Qiu SJ, Shi YH, Yu B, Tang N, Chu W, Wang M, Wu J, Zhang Z, Yang S, Gu J, Wang H, Qin W. Serum DKK1 as a protein biomarker for the diagnosis of hepatocellular carcinoma: a large-scale, multicentre study. Lancet Oncol 2012; 13(8): 817–826

[25] Lambert J, Chevret S. Summary measure of discrimination in survival models based on cumulative/dynamic time-dependent ROC curves. Stat Methods Med Res 2016; 25(5): 2088–2102

[26] Zhu W, Guo L, Zhang B, Lou L, Lin Z, Zhu X, Ren N, Dong Q, Ye Q, Qin L. Combination of osteopontin with peritumoral infiltrating macrophages is associated with poor prognosis of early-stage hepatocellular carcinoma after curative resection. Ann Surg Oncol 2014; 21(4): 1304–1313

[27] Sieghart W, Wang X, Schmid K, Pinter M, König F, Bodingbauer M, Wrba F, Rasoul-Rockenschaub S, Peck-Radosavljevic M. Osteopontin expression predicts overall survival after liver transplantation for hepatocellular carcinoma in patients beyond the Milan criteria. J Hepatol 2011; 54(1): 89–97

[28] Zhang H, Ye QH, Ren N, Zhao L, Wang YF, Wu X, Sun HC, Wang L, Zhang BH, Liu YK, Tang ZY, Qin LX. The prognostic significance of preoperative plasma levels of osteopontin in patients with hepatocellular carcinoma. J Cancer Res Clin Oncol 2006; 132(11): 709–717

[29] Chen RX, Xia YH, Cui JF, Xue TC, Ye SL. Osteopontin, a single marker for predicting the prognosis of patients with tumor-node-metastasis stage I hepatocellular carcinoma after surgical resection. J Gastroenterol Hepatol 2010; 25(8): 1435–1442

[30] Wang M, Mehta A, Block TM, Marrero J, Di Bisceglie AM, Devarajan K. A comparison of statistical methods for the detection of hepatocellular carcinoma based on serum biomarkers and clinical variables. BMC Med Genomics 2013; 6 (Suppl 3):S9

[31] Ishii Y, Nakasato Y, Kobayashi S, Yamazaki Y, Aoki T. A study on angiogenesis-related matrix metalloproteinase networks in primary hepatocellular carcinoma. J Exp Clin Cancer Res 2003; 22(3): 461–470

[32] Kessenbrock K, Plaks V, Werb Z. Matrix metalloproteinases: regulators of the tumor microenvironment. Cell 2010; 141(1): 52–67

[33] Liang B, Chen R, Wang T, Cao L, Liu Y, Yin F, Zhu M, Fan X, Liang Y, Zhang L, Guo Y, Zhao J. Myeloid differentiation factor 88 promotes growth and metastasis of human hepatocellular carcinoma. Clin Cancer Res 2013; 19(11): 2905–2916

[34] Hung TM, Chang SC, Yu WH, Wang YW, Huang C, Lu SC, Lee PH, Chang MF. A novel nonsynonymous variant of matrix metalloproteinase-7 confers risk of liver cirrhosis. Hepatology 2009; 50(4): 1184–1193

[35] Salahshor S, Goncalves J, Chetty R, Gallinger S, Woodgett JR. Differential gene expression profile reveals deregulation of pregnancy specific β1 glycoprotein 9 early during colorectal carcinogenesis. BMC Cancer 2005; 5:66

[36] Lisboa FA, Warren J, Sulkowski G, Aparicio M, David G, Zudaire E, Dveksler GS. Pregnancy-specific glycoprotein 1 induces endothelial tubulogenesis through interaction with cell surface proteoglycans. J Biol Chem 2011; 286(9): 7577–7586

[37] Qin LX, Tang ZY. Recent progress in predictive biomarkers for metastatic recurrence of human hepatocellular carcinoma: a review of the literature. J Cancer Res Clin Oncol 2004; 130(9): 497–513

[38] Osaki A, Suda T, Kamimura K, Tsuchiya A, Tamura Y, Takamura M, Igarashi M, Kawai H, Yamagiwa S, Aoyagi Y. A safe and effective dose of cisplatin in hepatic arterial infusion chemotherapy for hepatocellular carcinoma. Cancer Med 2013; 2(1): 86–98

[39] Sherman M. Recurrence of hepatocellular carcinoma. N Engl J Med 2008; 359(19): 2045–2047

[40] Wu JC, Huang YH, Chau GY, Su CW, Lai CR, Lee PC, Huo TI, Sheen IJ, Lee SD, Lui WY. Risk factors for early and late recurrence in hepatitis B-related hepatocellular carcinoma. J Hepatol 2009; 51(5): 890–897

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature