Oligosaccharide modification is dramatically altered during tumorigenesis. Cancer-specific glycosylation is a potential biomarker for cancer diagnosis. Fucosylation is one of the most important oligosaccharide modifications involved in cancer and inflammation . Several fucosylated proteins have been reported to be potential cancer biomarkers [2–4]. In particular, fucosylated α-fetoprotein (AFP-L3) has been widely examined as a cancer biomarker for hepatocellular carcinoma (HCC) . AFP-L3 is formed by the attachment of a fucose residue to asparagine 232 in the amino acid sequence of AFP. AFP and AFP-L3 are distinguished by the presence of α1-6 fucosylation in AFP-L3. Lens culinaris agglutinin (LCA) or lectin recognizes specific carbohydrates, including α1-6 fucose. Lectin-dependent fractionation of AFP was originally described by Breborowicz  and Taketa . Aoyagi et al. evaluated the clinical utility of AFP-L3, particularly as a diagnostic tool for HCC . AFP-L3 has been approved as an HCC-specific biomarker. However, the availability of the AFP-L3 diagnostic assay is slightly different between USA and Japan. Marrero et al. reported that a new cut-off value for AFP yielded higher sensitivity than AFP-L3 or protein induced by vitamin K antagonist II (PIVKA-II) . This finding was the result of multiple analyses of HCC serum samples. However, there remains insufficient evidence to measure AFP alone as a diagnostic marker for early HCC.
We reported that fucosylated haptoglobin (Fuc-Hpt) is a novel marker for pancreatic cancer and colorectal cancer [10–12]. Further, we investigated the molecular mechanisms underlying its production in patients with pancreatic cancer, and found that increased interleukin-6 (IL-6) production by pancreatic cancer cells induced the production of Fuc-Hpt in the liver . The majority of Hpt found in sera is produced from the liver. However, the liver expresses quite a low level of Fuc-Hpt. To examine Fuc-Hpt as a clinical cancer biomarker, we previously developed a lectin antibody enzyme-linked immunosorbent assay (ELISA) system .
The application of Fuc-Hpt as a biomarker of hepatic diseases is complicated by changes in Hpt production due to alteration of liver function. For example, liver cirrhosis (LC) leads to the reduced production of Hpt , whereas hepatic inflammation due to chronic hepatitis (CH) causes a dramatic increase in Hpt production. To overcome this issue, we recently reevaluated the sensitivity of our lectin antibody ELISA kit. We found that Fuc-Hpt could be detected in sera that were diluted 25–625 times . In the present study, we examined the clinical utility of serum Fuc-Hpt as a biomarker for HCC. We investigated serum samples from patients with chronic liver diseases and serum samples from patients with HCC before and after diagnosis and therapy. Conventional cancer biomarkers, such as AFP, AFP-L3, and PIVKA-II, were compared with Fuc-Hpt.
Materials and methods
Ninety-two patients with HCC, 81 patients with LC, 145 patients with CH caused by hepatitis B or C infection, and 242 normal volunteers were enrolled. These patients were treated at Osaka National Hospital, Osaka Rosai Hospital, or Osaka University Hospital from 1996 to 2010. A normal volunteer was characterized by the presence of normal liver biochemistry and no history of liver disease or alcohol abuse. LC was diagnosed based on liver histology or clinical, laboratory, and imaging data. The diagnosis of HCC was made either by histopathology or by a combination of imaging tests [ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), or angiography]. Liver biopsy was obtained to confirm the diagnosis in some cases. Nineteen patients with HCC whose sera were collected at three time points during 1996–2003 were defined as long-term follow-up LC–HCC patients. Most of the long-term follow-up patients were treated with transcatheter arterial embolization and/or percutaneous ethanol injection therapy. Thirteen patients with HCC whose sera were collected several times from 2009 to 2012 were defined as short-term follow-up LC–HCC patients. All of the short-term follow-up patients had stage I HCC with tumors that were <2 cm in size and were treated with surgery. Patients with LC were divided into two groups, which included 40 cases that did not develop HCC for at least 5 years (the non-HCC group) and 41 cases that developed HCC (the HCC group). A subset of liver biopsies from patients with CH was analyzed for hepatitis activity. Fuc-Hpt was previously examined in healthy volunteers . In follow-up studies (long- and short-term follow-up), to investigate the usefulness of Fuc-Hpt as a HCC biomarker, the Fuc-Hpt values before and after the diagnosis of HCC were used in the analyses. Sera were maintained at –80 °C. The protocol and informed consent were approved by Institutional Review Boards at Osaka University Hospital, and the study was conducted in accordance with the Helsinki Declaration.
Lectin antibody ELISA for Fuc-Hpt
The Fab fragment of anti-human Hpt IgG (Dako, Carpinteria, CA, USA) was coated onto the bottom of a 96-well ELISA plate, because IgG has the fucosylated oligosaccharide in its Fc portion. The antibody-coated plate was blocked with phosphate-buffered saline (PBS) containing 3% bovine serum albumin for 1 h followed by washing in PBS containing 0.1% Tween 20 (PBS-T). A 50-μL aliquot of sera, which had been diluted by a factor of 125, was added to each well and incubated for 1 h at room temperature. To detect Fuc-Hpt, 1/1000 diluted biotinylated AAL (Aleuria aurantia) was added into each well. As is well known, AAL recognizes both core and outer-arm fucosylation. Therefore, our lectin antibody ELISA system measures both core and outer-arm Fuc-Hpt. The data are presented as ratios relative to the Fuc-Hpt produced by PK8 cells transfected with an expression vector of the Hpt gene, as described previously . Hemoglobin inhibited this lectin antibody ELISA as we reported previously ; we therefore excluded hemolysis samples in this study. Limit of the detection is 1.09–70 U/mL in this ELISA kit. Detailed procedures were previously described [13, 14, 16]. The Fuc-Hpt elevation rate was defined as a change in Fuc-Hpt values at two time points during a 10-year follow-up period.
Determination of total haptoglobin, AFP, AFP-L3, and PIVKA-II
Total Hpt levels were measured using a Haptoglobin ELISA kit (ASSAYPRO, St. Charles, MO, USA). A 25-μL aliquot of serum from each patient enrolled in the study was examined according to the manufacturer’s protocol. Serum levels of AFP and PIVKA-II were determined using a chemiluminescent enzyme immunoassay (CLEIA). We used the on-chip immunoassay to measure AFP-L3 levels using the micro Total Analysis System of Wako Diagnostics (μ-TAS WAKO) according to the manufacturer’s protocol .
Statistical analysis was conducted using JMP Pro 10.0 software (SAS Institute Inc., Cary, NC, USA). Kruskal-Wallis and Wilcoxon tests were used to assess any significant differences in continuous clinical or serological characteristics between groups. Differences between groups in the Kaplan-Meier analysis were tested with Wilcoxon’s test. A receiver operating characteristic (ROC) curve was generated by plotting sensitivity versus 1 – specificity for every possible cut-off score, and area under the ROC curve (AUC) was calculated. p-Values <0.05 were considered significant.
Serum levels of Fuc-Hpt in patients with chronic liver diseases
Serum Fuc-Hpt levels measured using lectin antibody ELISA were 1314±1574 U/mL in patients with CH, 2433±3262 U/mL in patients with LC, and 3260±3511 U/mL in patients with HCC (Figure 1A). Serum levels of Fuc-Hpt in healthy volunteers were 363±482 U/mL. Levels of Fuc-Hpt increased with disease progression. Fuc-Hpt levels were significantly higher in patients with CH and LC compared to those in normal volunteers. Further, Fuc-Hpt levels in LC patients were significantly higher than those in CH patients (p<0.005). However, there were no significant differences in Fuc-Hpt levels between patients with LC and HCC (p=0.061), although the median level of Fuc-Hpt was slightly higher in HCC patients. ROC analysis indicated that the diagnostic performance of Fuc-Hpt as the HCC biomarker was generally good (HV vs. HCC; AUC 0.84, sensitivity 75.0%, specificity 79.0%, HV, CH and LC vs. HCC; AUC 0.76, sensitivity 40.4%, specificity 76.2%) (Supplementary Data, Figure 1, that accompanies the article at http://www.degruyter.com/view/j/cclm.2015.53.issue-1/issue-files/cclm.2015.53.issue-1.xml). Both graphs indicated that the AUC, sensitivity, and specificity were lower in HCC compared with those observed in pancreatic cancer patients reported in our previous study . Collectively, Fuc-Hpt seems to be a unique biomarker that can help discriminate pathological changes in liver diseases.
Next, we investigated whether Fuc-Hpt levels are associated with the development of HCC from LC. LC patients were divided into two groups. The non-HCC group included LC patients who did not develop HCC during a 5-year follow-up period. The HCC group included all the remaining LC patients. Fuc-Hpt levels were lower in patients in the non-HCC group than those of patients in the HCC group, although the differences were not statistically significant (Figure 1B). A few patients in the HCC group had extremely high Fuc-Hpt levels, none of the patients in the non-HCC group had the high Fuc-Hpt levels that were observed in these few outlying cases in the HCC group.
Comparison of serum Fuc-Hpt among patients with LC and CH
Serum Fuc-Hpt levels were investigated in patients with CH who underwent liver biopsy to assess liver histology (n=62). CH patients who underwent liver biopsy were stratified according to hepatitis pathological activity. The activity of CH was determined by pathological analysis according to the METAVIR scoring system and divided into A0 and A1-3 . Fuc-Hpt levels in patients with CH (A0) were significantly lower than those in patients with CH (A1-3) (Figure 2A). In addition, we compared the histological scores and clinical data of our subjects (Supplementary Data, Table 1A, B and Figure 2). We found that serum Fuc-Hpt levels tended to correlate with activity scores, but did not correlate with fibrosis stage scores divided by METAVIR scoring system (F0-4) . Interestingly, the ratio of Fuc-Hpt to total Hpt correlated significantly with activity scores. As Hpt is an acute-phase protein produced in the liver, we compared Fuc-Hpt and total Hpt levels. However, there was no correlation between total Hpt and Fuc-Hpt levels (Figure 2B), suggesting that secretion mechanisms of Hpt and Fuc-Hpt were different, and measuring the Fuc-Hpt level could be valuable instead of measuring Hpt level alone. Next, we compared serum Hpt levels in chronic liver disease patients (Figure 2C). We found that serum Hpt levels were significantly decreased in LC patients compared to levels in CH patients as previously reported . Hpt levels were significantly higher in HCC patients than in LC patients. We also investigated the relationships between serum Fuc-Hpt levels and several variables (alanine aminotransferase, albumin, total bilirubin, prothrombin time [%]) in limited numbers of patients from whom we could obtain clinical data (Supplementary Data, Figure 3). We found no significant relationships between serum Fuc-Hpt levels and several variables in both LC and HCC patients. The Fuc-Hpt/Hpt ratio was significantly increased in LC and HCC patients compared to that in CH patients (Figure 2D).
Fuc-Hpt in LC–HCC follow-up patients
Serum Fuc-Hpt levels were examined at three time points during a 10-year period in 19 patients who progressed from LC to HCC (Table 1, Supplementary Data, Figure 4). Interestingly, 80% of cases showed increased Fuc-Hpt levels after HCC was detected. However, we were unable to determine the cut-off value for the development of HCC. More than 70% of patients showed a 3- to 5-fold increase in Fuc-Hpt. We evaluated Fuc-Hpt as a diagnostic marker for HCC compared to other conventional cancer biomarkers for HCC, such as AFP, AFP-L3, and PIVKA-II. The AUC values for HCC prediction by Fuc-Hpt, PIVKA-II, AFP, and AFP-L3 were 0.840, 0.696, 0.602, and 0.530, respectively (Table 2). These results indicated that the diagnostic ability of Fuc-Hpt was superior to that of other classical HCC diagnostic markers. The cut-off values for each marker were determined by ROC analysis. Combined analysis of Fuc-Hpt and other tumor markers did not increase the AUC value for HCC diagnosis. Next, we investigated the Fuc-Hpt elevation rate in these patients to determine if increased Fuc-Hpt is predictive for HCC. The Fuc-Hpt elevation rate, which was defined as a change in Fuc-Hpt values at two time points during a 10-year follow-up period, was significantly higher in HCC(+) patients with LC than in HCC(–) patients with LC (Figure 3A). ROC analysis indicated that a Fuc-Hpt elevation rate of more than 498.2% showed 100% specificity for HCC detection (Figure 3B). We found in this study that the Fuc-Hpt elevation rate was significant for the prediction of HCC occurrence in each patient.
Which cancer biomarkers are useful for early diagnosis of HCC?
Fuc-Hpt levels were investigated in 13 patients with early-stage HCC who underwent surgery. Fuc-Hpt was measured every 6 months before and after HCC detection for 2–3 years. Results are summarized in Table 3. The AUC values of Fuc-Hpt, PIVKA-II, AFP, and AFP-L3 were 0.618, 0.658, 0.546, and 0.555, respectively. Although AUC values of Fuc-Hpt and PIVKA-II were higher than those of AFP and AFP-L3, none of the cancer biomarkers was sufficient to make an early diagnosis of HCC. Combination analysis of Fuc-Hpt and PIVKA-II, AFP, or AFP-L3 was also not effective for HCC prediction. In contrast, approximately 40% of HCC patients showed gradual increases in Fuc-Hpt and PIVKA-II. Representative cases that showed the utility of cancer biomarkers for detecting early HCC are summarized in Supplemental Data. Among 13 patients with HCC, cases 1–5 showed gradual increases in Fuc-Hpt before HCC was detected by ultrasound or CT. Peak increases in Fuc-Hpt were observed 1–1.5 years before HCC detection in cases 6–10 followed by decreased Fuc-Hpt at the time of HCC diagnosis. Fuc-Hpt did not show utility as a cancer biomarker in cases 11–13. PIVKA-II was useful as a biomarker for early HCC in five cases (1, 4, 6, 7, and 11). AFP was useful as a biomarker for early HCC in cases 5, 9, and 13. In contrast, AFP-L3 was useful for early HCC detection only in case 13. Although the specificity of HCC prediction by AFP was very high, the sensitivity was only 33.3%. In addition, gradual increases in AFP were observed in 3/13 cases (23%). Quantitative analyses of AFP-L3 (concentration) and AFP-L3% (data not shown) were not effective for early HCC prediction. Although HCC patient numbers investigated in the present study were limited, our study demonstrated that Fuc-Hpt is a more useful biomarker for early diagnosis of HCC than other conventional biomarkers.
In the present study, we have demonstrated that Fuc-Hpt is a novel and unique biomarker that can discriminate between pathological changes occurring in chronic liver diseases. Serum Fuc-Hpt levels increased with the progression of liver disease, suggesting that Fuc-Hpt levels may be dependent on liver functions and/or fibrosis. Our previous study already demonstrated that hepatoma cell lines (Hep3B and HepG2) produce Fuc-Hpt . In addition, the present retrospective analysis showed increased levels of Fuc-Hpt over time, indicating its potential utility as a predictive biomarker of HCC. Surprisingly, this potential utility of Fuc-Hpt was superior to other classical HCC biomarkers (AFP, AFP-L3, PIVKA-II) (Table 2). In particular, Fuc-Hpt elevation rates were useful for the diagnosis of HCC development, and a Fuc-Hpt elevation rate of more than 498.2% showed 100% specificity for HCC detection (Figure 3). Although none of these HCC biomarkers (Fuc-Hpt, AFP, AFP-L3, PIVKA-II) were sufficiently able to aid in the early diagnosis of HCC in this study, Fuc-Hpt showed gradual increases in five of 13 (38%) patients prior to imaging HCC diagnosis (Table 3, Supplementary Data, Figure 5). In addition, Fuc-Hpt levels were significantly higher in patients with chronic active hepatitis than in those with inactive hepatitis (Figure 1B, Supplementary Data, Table 1A), suggesting that inflammation may be involved in Fuc-Hpt induction. Consistent with this hypothesis, inflammation-related cytokines, such as IL-6, have been shown to induce Fuc-Hpt production . However, total Hpt and Fuc-Hpt levels were not correlated, suggesting that the production of Hpt is not always dependent on serum Fuc-Hpt levels. As we reported previously, secretion of fucosylated proteins in the liver is selectively regulated by cellular polarity [19, 20], which may also affect serum Fuc-Hpt levels in liver diseases.
Indeed, serum Hpt levels are increased with inflammation because IL-6 receptor stimulates the production of Hpt in the liver. However, in cases with LC patients, the production of Hpt in the liver is decreased due to liver dysfunction as described previously . Hpt levels were significantly higher in HCC patients than in LC patients. In this study, the mean value of Hpt in HCC patients was higher than that in LC patients and lower than that in CH patients. Therefore, these results demonstrated that serum Hpt levels are not suitable markers for the prediction of HCC occurrence in CH patients. This study demonstrated that serum Fuc-Hpt levels were also correlated with liver inflammation in patients who had undergone liver biopsy. The mean value of serum Fuc-Hpt levels increased with the progression of liver disease. We would like to investigate the precise relationships between clinical data (including histological data) and serum Fuc-Hpt levels in future studies.
Routine follow-up of patients with liver diseases includes imaging by ultrasonography or CT every 3–6 months. Patients with LC should be closely followed to monitor HCC development. The present study demonstrates the utility of Fuc-Hpt and conventional cancer biomarkers, such as AFP and PIVKA-II. However, as previously noted, it is difficult to apply biomarkers for the early diagnosis of HCC. Although the AUC values of Fuc-Hpt and PIVKA-II were <0.6, these biomarkers detected HCC before imaging diagnosis in only 30%–40% of patients with early-stage HCC (Supplementary Data, Figure 5). Interestingly, increased levels of AFP showed the highest specificity in short-term follow-up analyses. Gradual increases in AFP are recognized as a sign of HCC development, consistent with the results of this study. Serum Fuc-Hpt levels >1655.4 U/mL and Fuc-Hpt elevation rates >498.2% showed high specificity (87.5% and 100%) for diagnosing HCC in patients with LC. The high specificity suggests that Fuc-Hpt may be produced from HCC cells but not from hepatocytes in the cirrhotic liver. Moreover, fucosylation is a possible signal for the polarized secretion of fucosylated glycoproteins into bile ducts in the liver, as we have previously reported . Therefore, the deformity of hepatocyte polarity in cancer cells would cause increased production or secretion of Fuc-Hpt into the serum in HCC patients.
Fuc-Hpt may be a useful predictive marker for HCC. The non-HCC group of LC patients with low levels of Fuc-Hpt did not develop HCC during a 5-year follow-up. In summary, serum Fuc-Hpt is a biomarker that can discriminate between pathological changes in liver diseases and showed value as a potential predictive biomarker for HCC. Based on these promising data, we have initiated a prospective study with a larger group of patients with LC to validate Fuc-Hpt as a novel diagnostic marker for HCC.
Financial support: This study was supported by a Grant-in-Aid for Scientific Research (A), No. 21249038 from the Japan Society for the Promotion of Science and performed as a research program of the Project for Development of Innovative Research on Cancer Therapeutics (P-Direct), Ministry of Education, Culture, Sports, and Science.
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Financial Support: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.
Wang M, Long RE, Comunale MA, Junaidi O, Marrero J, Di Bisceglie AM, et al. Novel fucosylated biomarkers for the early detection of hepatocellular carcinoma. Cancer Epidemiol Biomarkers Prev 2009;18:1914–21.CrossrefWeb of ScienceGoogle Scholar
Hashimoto S, Asao T, Takahashi J, Yagihashi Y, Nishimura T, Saniabadi AR, et al. alpha1-acid glycoprotein fucosylation as a marker of carcinoma progression and prognosis. Cancer 2004;101:2825–36.Google Scholar
Debruyne EN, Vanderschaeghe D, Van Vlierberghe H, Vanhecke A, Callewaert N, Delanghe JR. Diagnostic value of the hemopexin N-glycan profile in hepatocellular carcinoma patients. Clin Chem 2010;56:823–31.Web of ScienceGoogle Scholar
Breborowicz J, Mackiewicz A, Breborowicz D. Microheterogeneity of alpha-fetoprotein in patient serum as demonstrated by lectin affino-electrophoresis. Scand J Immunol 1981;14:15–20.PubMedCrossrefGoogle Scholar
Taketa K, Izumi M, Ichikawa E. Distinct molecular species of human alpha-fetoprotein due to differential affinities to lectins. Ann N Y Acad Sci 1983;417:61–8.Google Scholar
Marrero JA, Feng Z, Wang Y, Nguyen MH, Befeler AS, Roberts LR, et al. Alpha-fetoprotein, des-gamma carboxyprothrombin, and lectin-bound alpha-fetoprotein in early hepatocellular carcinoma. Gastroenterology 2009;137:110–8.Google Scholar
Okuyama N, Ide Y, Nakano M, Nakagawa T, Yamanaka K, Moriwaki K, et al. Fucosylated haptoglobin is a novel marker for pancreatic cancer: a detailed analysis of the oligosaccharide structure and a possible mechanism for fucosylation. Int J Cancer 2006;118:2803–8.Google Scholar
Nakano M, Nakagawa T, Ito T, Kitada T, Hijioka T, Kasahara A, et al. Site-specific analysis of N-glycans on haptoglobin in sera of patients with pancreatic cancer: a novel approach for the development of tumor markers. Int J Cancer 2008;122:2301–9.Google Scholar
Takeda Y, Shinzaki S, Okudo K, Moriwaki K, Murata K, Miyoshi E. Fucosylated haptoglobin is a novel type of cancer biomarker linked to the prognosis after an operation in colorectal cancer. Cancer 2012;118:3036–43.Google Scholar
Narisada M, Kawamoto S, Kuwamoto K, Moriwaki K, Nakagawa T, Matsumoto H, et al. Identification of an inducible factor secreted by pancreatic cancer cell lines that stimulates the production of fucosylated haptoglobin in hepatoma cells. Biochem Biophys Res Commun 2008;377:792–6.Web of ScienceGoogle Scholar
Matsumoto H, Shinzaki S, Narisada M, Kawamoto S, Kuwamoto K, Moriwaki K, et al. Clinical application of a lectin-antibody ELISA to measure fucosylated haptoglobin in sera of patients with pancreatic cancer. Clin Chem Lab Med 2010;48:505–12.Web of ScienceGoogle Scholar
Kormoczi GF, Saemann MD, Buchta C, Peck-Radosavljevic M, Mayr WR, Schwartz DW, et al. Influence of clinical factors on the haemolysis marker haptoglobin. Eur J Clin Invest 2006;36: 202–9.Google Scholar
Kamada Y, Kinoshita N, Tsuchiya Y, Kobayashi K, Fujii H, Terao N, et al. Reevaluation of a lectin antibody ELISA kit for measuring fucosylated haptoglobin in various conditions. Clin Chim Acta 2013;417:48–53.Web of ScienceGoogle Scholar
Kagebayashi C, Yamaguchi I, Akinaga A, Kitano H, Yokoyama K, Satomura M, et al. Automated immunoassay system for AFP-L3% using on-chip electrokinetic reaction and separation by affinity electrophoresis. Anal Biochem 2009;388:306–11.Web of ScienceGoogle Scholar
Bedossa P, Poynard T. An algorithm for the grading of activity in chronic hepatitis C. The METAVIR Cooperative Study Group. Hepatology 1996;24:289–93.Google Scholar
Nakagawa T, Uozumi N, Nakano M, Mizuno-Horikawa Y, Okuyama N, Taguchi T, et al. Fucosylation of N-glycans regulates the secretion of hepatic glycoproteins into bile ducts. J Biol Chem 2006;281:29797–806.Google Scholar
Nakagawa T, Takeishi S, Kameyama A, Yagi H, Yoshioka T, Moriwaki K, et al. Glycomic analyses of glycoproteins in bile and serum during rat hepatocarcinogenesis. J Proteome Res 2010;9:4888–96.Web of ScienceCrossrefGoogle Scholar
The online version of this article (DOI: 10.1515/cclm-2014-0427) offers supplementary material, available to authorized users.
About the article
Published Online: 2014-07-23
Published in Print: 2015-01-01