Cancer Research Folkman AM No Date
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yoon, S. K.
Right arrow Articles by Kim, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yoon, S. K.
Right arrow Articles by Kim, J. W.
[Cancer Research 64, 5434-5441, August 1, 2004]
© 2004 American Association for Cancer Research

Immunology

The Human Cervical Cancer Oncogene Protein Is a Biomarker for Human Hepatocellular Carcinoma

Seung Kew Yoon1, Nam Kyu Lim2, Seon-Ah Ha2, Yong Gyu Park3, Jong Young Choi1, Kyu Won Chung1, Hee Sik Sun1, Myung Ja Choi5, Junho Chung6, Jack R. Wands7 and Jin Woo Kim2,4

1 Department of Internal Medicine and WHO Collaborating Center on Viral Hepatitis, 2 Molecular Genetic Laboratory, 3 Departments of Biostatistics, and 4 Obstetrics & Gynecology, The Catholic University of Korea, Seoul, Korea; 5 Korea Institute of Science and Technology, Bioanalysis and Biotransformation Research Center, Seoul, Korea; 6 Department of Biochemistry and Molecular Biology, Cancer Research Institute, Seoul National University College of Medicine, Seoul, Korea; and 7 Liver Research Center and Department of Medicine, Brown Medical School, Providence, Rhode Island


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Human cervical cancer oncogene (HCCR) was identified and appeared to function as a negative regulator of p53 gene. The objective of this study was to validate HCCR expression as a candidate marker for human hepatocellular carcinoma. HCCR epitope was identified as Y355LGTRR360. According to immunofluorescence study, HCCR was predominantly localized in the plasma membrane and cytoplasm of hepatocellular carcinoma. HCCR proteins were overexpressed in the tumorous compared with the nontumorous cirrhosis tissues. However, HCCR was not detected in normal liver tissue. Concentration of HCCR protein in the serum was measured in a total of 570 subjects, and comparisons were made to {alpha}-fetoprotein. Serological studies revealed 78.2% sensitivity of HCCR (cutoff value, 15 µg/ml), which was significantly higher than 64.6% of {alpha}-fetoprotein (P = 0.0098) and 95.7% specificity for hepatocellular carcinoma. Forty of 52 (76.9%) patients with carcinoma negative for {alpha}-fetoprotein showed positive values for HCCR. A positive rate of 69.2% in carcinoma patients with tumor sizes <2 cm was found to be a higher rate than measurement of {alpha}-fetoprotein. Furthermore, HCCR expression was also detected in liver cirrhosis at an intermediate level between carcinoma and normal groups, which gave 88.1% sensitivity and 79.0% specificity using 8 µg/ml as a cutoff value. In summary, the HCCR assay may have an advantage over the {alpha}-fetoprotein assay in that it is elevated according to disease progression from liver cirrhosis to carcinoma, and it is more frequently positive in patients with early, small hepatocellular carcinoma.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Hepatocellular carcinoma (HCC) has been reported as the fourth most common malignancy worldwide (1 , 2) . The prognosis of HCC is dependent on the size at presentation. Thus, early detection may offer hope for a more favorable prognosis.

To date, both ultrasonography and serum {alpha}-fetoprotein (AFP) assay are the principle methods of screening for HCC, but the diagnostic sensitivity of AFP for HCC is reported to range from 39% to 64% and specificity from 76% to 91%, respectively (3, 4, 5) . In small tumors (<3 cm) sensitivity of AFP levels varied from 26.9% to 39.3% (6 , 7) . In addition, Des-r-carboxyl prothrombin or serum protein induced by vitamin K absence or antagonist-II (PIVKA-II) and Lens culinaris agglutinin-A AFP (AFP-L3) have been reported to represent highly specific tumor markers when compared with AFP for the early detection of HCC (8 , 9) .

Recently, we identified two new oncogenes associated with human cervical cancer that are also overexpressed in other human tumors including those of gastrointestinal origin (10) . The human cervical cancer oncogene (HCCR) is classified into two species, HCCR-1 and HCCR-2 according to molecular characteristics. HCCR-2 (GenBank accession number AF315598) lacks exon 1 of HCCR-1 (GenBank accession number AF195651). These two proteins are normal alternative splicing variants. Previous work suggests that cells expressing HCCR-1 and HCCR-2 are tumorigenic in nude mice. Their functional role in tumorigenesis may reside as a negative regulator of the p53 tumor suppressor gene (10) . In the present study we investigated whether HCCR is overexpressed in HCC, detectable in serum, and potentially useful as a biomarker for assessing the presence of HCC.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Polyclonal Antibody Production.
The COOH-terminal portion of human HCCR-1 cDNA encoding a polypeptide from the 167th to the 360th amino acid residue was cloned into the pMAL-p2X (New England Biolabs, Beverly, MA) vector, which was then expressed in the Escherichia coli strain BL21 DE3. Recombinant HCCR-1 protein was purified from bacteria using amylose resin. The fusion protein was cleaved with factor Xa protease, and the resulting HCCR-1 protein was then used to generate polyclonal antibodies.

Immunization and Establishment of Hybridoma Lines.
BALB/c mice were immunized twice by s.c. injection. The first immunizations were with 50 µg of HCCR-1167–360 in Freund’s complete adjuvant. Two weeks later, the second immunizations were begun with HCCR-1167–360. Mice with the highest antibody titer received a boost of HCCR-1167–360 3 days before splenic removal. The splenocytes were fused to P3-X63-Ag8 myeloma cells using the procedure described perviously by Kennett (11) .

Phage Libraries Displaying Random Peptides Expression.
Anti-HCCR-1 IgG fraction was coated on Costar 3590 96-well assay plates (Corning) after being diluted to a final concentration of 10 µg/ml. The phage library in 10 µl of Tris-buffered saline with 50% glycerol was then mixed with 40 µl of 6% BSA in PBS and then added to the wells. After incubation, the plates were washed extensively with PBS containing 0.05% Tween 20. The bound phage was eluted by incubating the plate with 100 mM glycine (pH 2.2) with 0.1% BSA. Eluted phages were immediately neutralized (to pH 7) and then added to exponentially growing E. coli ER2537 cells. After the infection, the cells were grown at 37°C overnight. The phages present in the growth medium were purified by polyethylene glycol 8000/NaCl precipitation as described previously (12) . Five more cycles of this biopanning procedure were then performed. In each round of panning, the number of phages applied to the well and those finally eluted were determined. Individual clones were isolated as output phages of the last round and infected to E. coli ER2537 cells. After overnight culture the growth medium was stored and used for ELISA.

ELISA.
The wells of a Costar ELISA plate were coated with anti-HCCR-1 IgG by applying 50 µl of the immunoglobulin fraction at 10 µg/ml in PBS. After the wells had been blocked with 6% BSA in PBS, phage amplified from individual clones were added and incubated. The wells were then washed 5 times with PBS Tween. Horseradish peroxidase-conjugated anti-M13 monoclonal antibody (Pharmacia, Sweden) was added after being diluted in 6% BSA in PBS. A plate was incubated and washed with PBST followed by 50 µl of 1-Step 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulforic acid) diammonium salt peroxidase substrate solution (Pierce, IL) added to each well. After incubation, the reaction was stopped by adding 50 µl of 3 N HCl. The absorbance of each well was measured at 405 nm with an ELISA reader (EL 312e, Bio-Tek Instruments Inc., Winooski, VT).

DNA Sequence Analysis.
Plasmid DNA was isolated from cells infected with individual phage clones, and the DNA templates were subjected to automatic dideoxy-DNA sequencing. The oligonucleotide primer 5'-CCCTCATAGTTAGCGTAACG-3'with a fluorescence dye at position 96 downstream of the region encoding the random peptides in the library was synthesized and used for sequencing.

Subcellular Localization of HCCR.
Hep3B and Huh-7 cells were seeded in chamber slides. Cells were fixed in absolute methanol, and the fixed cells were blocked with 2% BSA in PBS. Cells were subsequently incubated with a murine monoclonal anti-HCCR-1167–360 antibody at a dilution of 1:200 followed by FITC-conjugated AffiniPure Donkey antimouse IgG (Jackson Immnoresearch). Cells were then washed and mounted with Universal mount. As a negative control, both cell lines were immunostained with nonrelevant antihepatitis C virus core antibody (1:500). After the immunostaining procedure, cells were then observed by confocal microscopy (Bio-Rad MRC 1024).

Western Blot Analysis.
Normal liver, nontumorous and HCC tissues were lysed in Laemmli sample buffer. Equivalent volumes of lysates containing 20 µg of total protein were loaded on 10% SDS-polyacrylamide gels. The membranes were incubated with 1:1000 dilution of rabbit polyclonal anti-HCCR-1167–360 serum in Tris-buffered saline. Next, the membranes were washed and incubated with a horseradish peroxide-conjugated goat antirabbit secondary antibody in Tris-buffered saline. Proteins were detected with the use of an enhanced chemiluminescence Western blot detection kit (Amersham Pharmacia Biotech., Sweden).

Immunohistochemical Staining.
For analysis of liver tissues, paraffin sections including normal, nontumorous and HCC tissues were dewaxed in xylenes, rehydrated, and retrieved in the microwave with 10 mM citrate buffer (pH 6.0). Then they were quenched in 90% methanol containing 3% H2O2 and blocked with normal serum corresponding to the animal source of the second antibody. The sections were incubated with mouse monoclonal anti-HCCR-1168–360 antibody, and the sections were treated with the avidin-biotin horseradish peroxidase complex (Vector ABC Elite kit; Vector Laboratories, Burlingame, CA). 3,3'-Diaminobenzidine was used as the substrate for developing the reaction.

Subjects.
During the study period from December 2001 to December 2002, a total of 570 subjects were enrolled at the Kangnam St. Mary’s hospital at the Catholic University of Korea. There were 147 individuals with histologically proven or clinically diagnosed HCC, 59 with liver cirrhosis, 143 with biopsy proven chronic hepatitis, and 11 with biopsy-proven nonalcoholic steato hepatitis (NASH). As controls, 72 pregnant women and 138 normal volunteers who were undergoing a routine health care check were included. All patients and controls were subjected to the analysis with individual consent for the study. The use of blood samples was approved by the Ethics Committee of our institution. The clinical diagnosis of HCC was on the basis of the following criteria: hepatic space occupying lesion suggestive of HCC on computed tomography, magnetic resonance imaging and angiography, serum AFP level exceeding 20 ng/ml, and serum positive for hepatitis B surface antigen or antihepatitis C virus antibody. All liver cirrhosis patients had low platelet counts (<100,000/mm3) and a history of ascites and/or esophageal variceal bleeding suggestive of decompensated liver disease and portal hypertension. The clinical characteristics of the study groups are presented in Table 1Citation .


View this table:
[in this window]
[in a new window]

  		Table 1 Baseline characteristics of the enrolled subjects

 
Detection of HCCR in Serum by Indirect ELISA.
The HCCR levels were measured using an indirect ELISA. Serial dilutions of purified HCCR-1167–360 were coated on 96-well Maxisorp microtiter plates in 100 µl of 10 mM phosphate buffer and incubated. The standard curve was constructed with serial dilutions of HCCR-1167–360 ranging from 20 to 320 ng/ml. Using 20, 40, 80, 160, and 320 ng/ml of HCCR-1167–360 protein, a calibration curve was prepared. The microtiter plate was then saturated with PBS containing 2 mg/ml BSA. Purified polyclonal primary antibody (100 ng/ml/well) was prepared in PBS containing 0.1% BSA and 0.05% Tween 20 and incubated.

Serum samples, diluted at 1:100, were analyzed in duplicate. A 100-µl sample was dispensed into the wells. Then, it was washed with PBS containing 0.05% Tween 20, and the primary antibody (100 ng/well) was added to each well and incubated. After washing, the second antibody (goat antirabbit antibody conjugated with horseradish peroxidase), diluted at 1:10,000, was added to the wells. The microtiter plate was washed, and a 100-µl aliquot of the color-developing reagent (tetramethylbenzidine) was dispensed into the wells to measure the horseradish peroxidase activity. The reaction was stopped by addition of 100 µl of 2 N sulfuric acid to each well and A450nm was read using a SpectraMax 250 microplate reader (Molecular Devices, Sunnyvale, CA); the data were processed with the SoftMax software. A standard curve was prepared, and a reference sample was included in each test run. The concentration of HCCR in each serum sample was calculated by reference to the standard curve and expressed as nanograms of HCCR per milliliter of serum. The reference ranges were obtained by receiver operating characteristic (ROC) curve analyses; values of 8 and 15 µg/ml were used as cut-offs for liver cirrhosis and HCC, respectively.

Statistical Analysis.
Serum levels (log value, µg/ml) in six different groups are presented as mean ± SE and evaluated using the Scheffe multiple comparison. To determine positive responses, we used the value of 20 ng/ml as the cutoff for AFP (Abbott Laboratories), and the cutoff values for HCCR were determined by ROC curve analyses. To confirm the validity of cutoff values, further diagnostic analysis was performed with 60 HCC, 66 liver cirrhosis, and 50 normal controls. Positive response rates of AFP and HCCR were compared by the McNemar test between cirrhosis, HCC, and pregnancy patient group, respectively. We did not have AFP levels for 24 patients with acute hepatitis. So only 119 patients were used in McNemar test, and positive response rates were presented in terms of marginal proportions. Sensitivity, specificity, and accuracy of HCCR were estimated, and their 95% confidence intervals are presented for liver cirrhosis and HCC with regard to normal group.

After dividing the subjects with liver cancer into two groups, based on a tumor size of 2 cm, {chi}2 and Fisher’s exact test were used to detect the differences of positive rates for HCCR levels between groups, and the McNemar test was used to compare the positive rates of AFP and HCCR within each group. The same procedures were performed for tumor stages (stage I, II versus III, IV). The two-tailed test, significance level 0.05, and SAS release 6.12 were used for all statistical analyses.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Generation of Antibodies to HCCR.
The recombinant HCCR-1 protein was used as an immunogen for polyclonal and monoclonal antibody production. Western blot analysis showed strong bands corresponding to HCCR-1 and HCCR-2 in HCC tissue; recombinant HCCR-1 served as a positive control (Fig. 1A)Citation . These bands disappeared when the antiserum was preabsorbed with recombinant HCCR-1 protein (Fig. 1A)Citation . Western blot analysis using 20-day-old rats and adult murine renal tissue revealed the presence of homologous proteins in these species as well (Fig. 1B)Citation . It is noteworthy that HCCR-1 and HCCR-2 proteins have also been found in normal human kidney tissue (10) .



View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. Expression of HCCR-1 and HCCR-2 proteins. A, Western blot analysis using polyclonal antiserum showed strong bands corresponding to recombinant HCCR-1 (about Mr 66,000), HCCR-1 (about Mr 42,000), and HCCR-2 (about Mr 36,000) as a positive control. These bands disappeared when the antiserum was preabsorbed with recombinant HCCR-1. B, Western blot analysis using 20-day-old rat and adult mouse kidney tissues revealed the presence of homologous proteins in these species, as well. HCCR, human cervical cancer oncogene.

 
Monoclonal antibodies directed against HCCR-1167–360 were generated. One hundred and five wells were positive for the presence of anti-HCCR-1167–360 antibodies using an antibody capture ELISA.

An Immunogenic Epitope Mapping on the HCCR Protein by Biopanning.
To define an immunogenic part of HCCR, we selectively enriched clones from the combinatorial phage display peptide library through repetitive rounds of panning on ELISA plates coated with the IgG fraction prepared from the antisera of immunized rabbits. In this context, four from random hexamer library, six from disulfide constraint random hexamer library, and six clones from random dodecamer library were found to be selectively enriched by panning and showed strong reactivity to the antibodies in phage ELISA although not reacting to normal rabbit sera. Two clones from random hexamer library and one clone from disulfide constraint hexamer library showed the consensus sequence of YLG(D/T)XR, which was very closely matched to the COOH-terminal sequence YLGTRR of HCCR-1 and HCCR-2 (Fig. 2A)Citation . Other clones from these two libraries and random dedecamer library revealed no consensus sequences corresponding to HCCR protein sequences. Therefore, a linear epitope on HCCR-1 and HCCR-2 was identified as the YLGTRR sequence (Fig. 2B)Citation .



View larger version (32K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Determination of an epitope on HCCR-1 and HCCR-2 using phage display of a combinatorial peptide library. A, HCCR-1 encodes a polypeptide of 360 amino acids with a predicted Mr 42,000, whereas HCCR-2 encodes a polypeptide of 304 amino acids with an Mr 36,000. HCCR-2 is an alternative splicing variant and lacks exon 1 of HCCR-1 (written in italic). The COOH-terminal portion of human HCCR-1 from the 167th amino acid residue to the 360th amino acid residue (underlined) was expressed in E. coli and used as an immunogen. B, biopanning of phage display of combinatorial peptide library on ELISA plate coated with immunoglobulin G fraction prepared from the sera of immunized rabbits yielded two clones from random hexamer library and one clone from disulfide constraint random hexamer library, which showed the consensus sequence of YLG(D/T)XR, which was very closely matched to the COOH-terminal sequence YLGTRR of HCCR-1 and HCCR-2. HCCR, human cervical cancer oncogene.

 
Immunolocalization of HCCR in Huh-7 and Hep3B Cells.
To characterize localization of HCCR protein in HCC cells, immunostaining using the anti-HCCR-1168–360 monoclonal antibody was performed. Confocal immunofluorescence analysis revealed that the HCCR-1 protein was mainly expressed on the plasma membrane and cytoplasm, although excluding the nucleus of HCC cell lines (Fig. 3, A and B)Citation . However, no staining of HCC cell lines was observed with a nonrelevant antihepatitis C virus core antibody used as a negative control (data not shown).



View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Localization of HCCR in Huh-7 and Hep3B cells. Confocal immunofluorescence analyses show that HCCR protein is mainly expressed on the plasma membrane and cytoplasm but not in the nucleus of both hepatocellular carcinoma (HCC) cell lines, Huh-7 (A) and Hep3B (B) cells. However, no staining of HCC cells was observed with a nonrelevant antihepatitis C virus core antibody (data not shown). HCCR, human cervical cancer oncogene.

 
Western Blot and Immunohistochemical Analyses.
Western blot analysis was performed using one normal liver tissue obtained from a liver transplantation procedure and four different tumors and their surrounding nontumorous tissues containing liver cirrhosis. Increased expression of HCCR-1 and HCCR-2 relative to liver cirrhosis was detected in HCC (Fig. 4)Citation . In the normal liver tissue, no signal was observed.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 4. Comparison of HCCR protein expression in one normal liver tissue obtained during liver transplantation. Four hepatocellular carcinoma (HCC) tissues and their corresponding nontumorous counterparts containing liver cirrhosis by Western blot analysis. N, normal liver; LC, liver cirrhosis tissues. HCCR, human cervical cancer oncogene.

 
By immunohistochemical staining, HCCR was overexpressed in the HCC (Fig. 5A)Citation compared with liver cirrhosis (Fig. 5B)Citation . The immunoreactivity was observed mainly in HCC tumor cells and, to a less extent, in preneoplastic hepatocytes of cirrhosis with a predominant cytoplasmic localization. In normal liver (Fig. 5C)Citation , there was no expression of HCCR.



View larger version (102K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 5. Immunohistochemical staining analyses. A, the immunoreactivity was observed mainly in the tumor cells of Hepatocellular carcinoma(s) (HCCs) with predominant cytoplasmic staining. B, however, the immunohistochemical staining of liver cirrhosis revealed weak staining of hepatocytes compared with HCC tumor cells. C, normal liver tissue showed no staining.

 
Distributions of Serological Levels of HCCR.
The distributions of serum HCCR levels (log value, µg/ml) in six different groups are presented in Fig. 6Citation and classified into three significantly different (P < 0.001) groups according to their values. The serum level of HCCR was highest in patients with HCC (mean ± SE, 23.82 ± 0.94) and also detected in a moderate level in liver cirrhosis (16.47 ± 1.02). In contrast, it was decreased to the basal level of normal controls (7.12 ± 0.23) in chronic hepatitis (8.24 ± 0.23), pregnancy (7.24 ± 0.31), and NASH groups (3.17 ± 0.08).



View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 6. Distributions of serum HCCR levels. The distributions of serum HCCR levels measured by ELISA in six different groups including normal, pregnancy, nonalcoholic steato hepatitis (NASH), chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma are shown. Data are represented as mean ± SE. *P < 0.001 compared with all other groups by Scheffe’s multiple comparison.

 
Cutoff Values.
To discriminate among three groups, we performed two ROC curve analyses using liver cirrhosis and normal liver cirrhosis and HCC groups, separately. The ROC curve for cirrhosis and normal groups that plotted sensitivity and false-positive fraction according to various cutoff values was presented in Fig. 7Citation . Among many values, we presented cutoff values, sensitivities, specificities, and total accuracies for two competitive points, A and B. Considering sensitivity and specificity as equally important, a value of 8 µg/ml was chosen as a more efficient cutoff level for subsequent diagnostic evaluation between normal and liver cirrhosis groups.



View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 7. Receiver operating characteristic curve for determination of cutoff value for liver cirrhosis. The receiver operating characteristic curve that plotted sensitivities and false-positive fractions according to two cutoff values (A and B) are shown.

 
ROC curve for liver cirrhosis and HCC patients was presented in Fig. 8Citation . Value of 17.3 µg/ml was the best cutoff point by logistic regression analysis considering sensitivities, specificities, and total accuracies. However we chose the value of 15 µg/ml as more convenient cutoff for diagnostic evaluation from a practical viewpoint.



View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 8. Receiver operating characteristic curve for determination of cutoff value (A) for hepatocellular carcinoma.

 
Diagnostic Results.
Diagnostic results using two distinct cutoff values were presented in Table 2Citation . With the cutoff value of 8 µg/ml, positive rate of HCCR in patients with liver cirrhosis was 88.1%, whereas the negative rate in normal controls was 79.0%. Total diagnostic accuracy of HCCR for liver cirrhosis was 81.7%. We also obtained 78.2% of positive rate for HCC group, 45.8% of negative rate for liver cirrhosis group, and 68.9% of total accuracy using cutoff value of 15 µg/ml.


View this table:
[in this window]
[in a new window]

  		Table 2 Diagnostic results of HCCR a

 
Positive response rates for each group with two cutoff values are shown in Table 3Citation . If we used cutoff value of 8 µg/ml, sensitivities of HCC, liver cirrhosis, hepatitis, pregnancy, and NASH were 100%, 88.1%, 48.3%, 33.3%, and 0%, respectively, whereas negative rate for normal was 79%. Instead, the cutoff value of 15 µg/ml gave positive rates of 78.2% and 54.2% for HCC and liver cirrhosis and a negative rate of 95.7% for normal group. Positive rates for other groups were very low. Therefore, after diagnosing patients with serum level of HCCR ≥ 15 µg/ml as cirrhosis or HCC, we might further diagnose those patients with serum level of HCCR between 8 and 15 µg/ml as cirrhosis. If we consider only three groups (normal, cirrhosis, and HCC), the overall diagnostic accuracy of two cutoff values was 70.9% (Table 4)Citation .


View this table:
[in this window]
[in a new window]

  		Table 3 Positive rates for each group using two cut-offs

Number (%).

 

View this table:
[in this window]
[in a new window]

  		Table 4 Overall accuracy of HCCRa using two cut-off values

Overall accuracy = 244/344 = 70.9%

 
Cross-Validation of Cutoff Values.
To confirm the validity of both cutoff levels, additional analysis was performed with the prospective cohort, those composed of 60 HCC, 66 liver cirrhosis, and 50 normal controls. The serum levels of HCCR were 19.13 ± 0.76 (µg/ml, mean ± SE) in HCC, 16.03 ± 0.39 in liver cirrhosis, and 7.75 ± 0.52 in normal controls, respectively, which were similar to those of previous results.

With the cutoff value of 8 µg/ml, the positive rate of HCCR in patients with liver cirrhosis was 100%, and the negative rate in normal control was 54%. Total diagnostic accuracy for liver cirrhosis and normal control was 80.2%. We also obtained 75% of positive rate for HCC, 42.4% of negative rate for liver cirrhosis, and 57.9% of total accuracy using cutoff value of 15 µg/ml.

Prospective cohort gave somewhat different values of sensitivity and specificity from previous data; however, total accuracies were similar in both data with 8 µg/ml. Cutoff value of 15 µg/ml gave similar diagnostic abilities in both data. Therefore, we could use both cutoff values derived from a retrospective analysis and compare them with AFP levels.

Diagnostic Abilities for Cancer Comparison Using AFP.
We used the value of 20 ng/ml as the cutoff for AFP and the cutoff value of 15 µg/ml for HCCR to diagnose HCC. One hundred and fifteen (78.2%) cancer patients were correctly detected by HCCR, whereas 95 (64.6%) patients were correctly detected by AFP, which was significantly different (P = 0.0098; Table 5Citation ). Particularly, 40 (76.9%) of 52 patients who were diagnosed as negative by AFP were correctly identified by HCCR. In addition, nine patients with metastatic lesions (six with lung and mediastinum nodules, one with skin mass, and two with bone metastases) who were negative for AFP also showed positive values for HCCR (data not shown).


View this table:
[in this window]
[in a new window]

  		Table 5 Diagnostic result of AFP and HCCR in HCC, number(%)

P = 0.0098, by McNemar test.

 
Positive rates of AFP for other groups were presented in Table 6Citation . Comparing these values with those (cutoff value of 15 µg/ml) in Table 3Citation , AFP showed significantly higher false-positive rates in all groups other than HCCR (P < 0.0001) except NASH.


View this table:
[in this window]
[in a new window]

  		Table 6 Positive rates of AFPa in each group

Number (%).

 
Among 72 women during gestation who were positive for AFP, no women (0%) showed positive responses for HCCR. In patients who were diagnosed with chronic hepatitis, the positive rate of the HCCR was 3 (2.1%) of 143 patients, and AFP was elevated in 31 (26.1%) of 119 patients. All 11 patients with NASH were diagnosed as negative both by AFP and HCCR.

Relationship of HCCR Levels to Tumor Size.
Although the positive rate of AFP was increased to 20.5% for the large sized tumor (≥2 cm; 66.7%) compared with the small sized tumor (<2 cm; 46.2%), there was no significant difference between two tumor sizes (P = 0.1395; Table 7Citation ). Positive rate of HCCR in large-sized tumor was significantly higher than that of AFP (P = 0.0295). Difference between these two diagnostic methods was 23% in small-sized tumor and 12.1% in large sized tumor, suggesting the potential usefulness of HCCR levels in detecting early HCC.


View this table:
[in this window]
[in a new window]

  		Table 7 Positive rates of AFPa and HCCR for each tumor size in HCCb

 
HCCR Levels and Tumor Stages of HCC.
Although the elevated levels of AFP increased to approximately 4% for advanced stages (to tumor lymph node metastasis stage III and IV) compared with the early stage (to tumor lymph node metastasis stage I and II), there was no significant difference between tumor stages (P = 0.7379; Table 8Citation ). The positive rates of HCCR were somewhat higher than those of AFP in both stages, and the difference between these two diagnostic methods was 11.8% in the early stage and 5.4% in the advanced stage, respectively.


View this table:
[in this window]
[in a new window]

  		Table 8 Positive rates of AFPa and HCCR for each tumor stage in HCCb

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
From analysis of mRNA species using a differential display reverse transcription-PCR approach, we identified a new human oncogene HCCR that was overexpressed in various human tumors including leukemia, lymphoma, and gastrointestinal cancer (10) . Although the HCCR oncogene was originally identified in human cervical cancer, its serum levels were found to be elevated in patients with HCC but not with cervical cancer (data not shown), implying its close relationship with hepatocarcinogenesis.

We defined an immunogenic part present in the extracellular domain of HCCR-1 using a phage display combinatorial peptide library. Of 16 clones specific to the anti-HCCR-1 immunoglobulin fraction, three clones were found to comprise a consensus sequence of YLG(D/T)XR, which is very closely related to the COOH-terminal sequence YLGTRR of HCCR. Three other clones, one from disulfide constraint hexamer library and two from the dodecamer library, showed the consensus sequence of SSL, which could not be allocated to any sequence present in HCCR. There was no significant sequence homology in the remaining clones. It suggests that the COOH-terminal sequence YLGTRR appears to be a major linear epitope of HCCR recognized by anti-HCCR antibodies to be used to detect serological levels of HCCR. This is consistent with the observation that the polysera reacted with both HCCR-1 and HCCR-2, which share the same COOH-terminal region. Therefore, epitope mapping using a biopanning technique validated that the anti-HCCR antibodies used for diagnosis of HCC actually reacted to the sequence found in HCCR protein. It seems that YLGTRR is one of the critical epitopes for immunogenicity of HCCR proteins and thereby might affect the diagnostic efficiency for HCC. It has yet to be clarified whether other epitopes uncovered by this panning procedure still exist.

With respect to histopathological changes in hepatocarcinogenesis, previous work has suggested that the morphological definition of preneoplastic lesions and early HCC has yet to be clarified, although the morphological transition from hyperplastic nodules to overt HCC has been observed (13) . In this regard, development of serological tumor markers associated with tumorigenesis will facilitate early clinical diagnosis of small HCC in the absence of histopathology related to early HCC. Serum AFP levels have been used as a tumor marker for the diagnosis of HCC. In addition, several other tumor markers including PIVKA-II, AFP-L3, transforming growth factor ß1, and hepatocyte growth factor are now available for early detection or improvement of diagnostic accuracy of HCC (6, 7, 8, 9 , 14) . From our study, we found that HCCR levels in serum were significantly higher in patients with HCC than in those with liver cirrhosis. The cutoff value of 15 µg/ml was found to be able to discriminate between HCC and liver cirrhosis. Sensitivity and specificity of HCCR for HCC (versus liver cirrhosis) were 78.2% and 45.8%, respectively, and the specificity for normal was 95.7%. It indicates that this tumor marker could discriminate liver cirrhosis from HCC based on the degree of HCCR expression level in the serum.

On the other hand, liver cirrhosis itself was assigned to a group with concentration of HCCR ranging from 8 to 15 µg/ml, which is discernible from the normal, chronic hepatitis, and pregnancy groups. This observation suggests that patients with intermediate concentration of HCCR between 8 and 15 µg/ml could be assigned to the potential carriers of liver cirrhosis. Sensitivity and specificity of HCCR for liver cirrhosis were 88.1 and 79.0%, respectively.

Using two different cutoff values, 8 and 15 µg/ml, 244 (70.9%) of 344 subjects were correctly classified (Table 4)Citation . In this context, patients with liver cirrhosis may be in a highly carcinogenic state, and serum HCCR level above 15 µg/ml may be a risk factor for the development of HCC.

AFP is an oncofetal glycoprotein, and its level increases in maternal blood during gestation (15) . In our study, all 72 pregnant women showed, as expected, elevated values of AFP above 20 ng/ml in serum whereas none (0%) had abnormal HCCR values above 15 µg/ml (P < 0.0001). This observation implies that HCCR level does not correlate with AFP in most instances.

Although AFP has been used as a tumor marker for HCC, its low sensitivity and specificity make early diagnosis of HCC difficult (3, 4, 5) . AFP is also elevated during hepatocyte regeneration following liver damage. Previous reports revealed that serum AFP was abnormally elevated to 31–52% in acute hepatitis, 15–58% in chronic hepatitis and 11–47% in liver cirrhosis (16 , 17) . The diagnostic comparison of AFP and HCCR in chronic hepatitis is shown in Table 3Citation and 6Citation . There was significant difference in diagnostic accuracy between the two methods (P < 0.0001) and false-positive rate of AFP is consistent with previous reports (16 , 17) .

Because of the poor sensitivity and specificity of AFP, early diagnosis of small HCC is difficult with this marker alone (18 , 19) . Therefore, another tumor marker like PIVKA-II or ultrasonography has been used for confirmatory diagnosis of HCC. PIVKA-II has been reported as a more sensitive and specific tumor marker for early diagnosis of HCC (6 , 20) . Furthermore, a recent report demonstrated that the combined assessment of AFP and PIVKA-II increased sensitivity, although there was no significant correlation between AFP and PIVKA-II levels (5 , 21) . AFP-L3 has been reported to be a useful marker in the early diagnosis of HCC. Its diagnostic accuracy varied from 17% to 42% in small HCC of <2 cm (22, 23, 24, 25) . Previous work has suggested that simultaneous measurement of serum des-{gamma}-carboxy prothrombin and AFP-L3 in patients with small HCC increased the sensitivity to 54.1%; specificity and accuracy increased to 97.8% and 84.1%, respectively (26) . Thus, the markers are complementary and useful for the diagnosis and evaluation of small HCC when measured simultaneously. In our results, positive rate of HCCR in small HCC < 2 cm in diameter was 69.2%, whereas that of AFP was 46.2% as shown in Table 7Citation , suggesting HCCR may be a promising tumor marker in the early diagnosis of small HCC.

It is difficult to distinguish between dysplastic nodules and early HCC by imaging alone because dysplastic nodules may develop into HCC. In particular, small and early HCC (<2 cm) is not easily revealed by imaging, serological tumor marker analysis, or both. Previous work has suggested that three-phase helical dynamic computed tomography was relatively insensitive for detection of HCC and dysplastic nodules in cirrhotic liver and especially difficult for dysplastic nodules and HCC < 2 cm (27) .

We compared the positive rates of HCCR and AFP in 110 HCC patients according to tumor lymph node metastasis. The positive rates of HCCR were higher, but not significant, than those of AFP in both early (76.5% versus 64.7%) and advanced (68.8% versus 74.2%) stages of HCC (Table 8)Citation .

In conclusion, a newly identified oncogene and its corresponding serum protein appear promising as a new biomarker for hepatocarcinogenesis.


   FOOTNOTES
 
Grant support: Korea Research Foundation Grant (KRF-2002-005-E00011).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Jin Woo Kim, Molecular Genetic Laboratory, College of Medicine, The Catholic University of Korea, 505 Banpo-dong, Seocho-ku, Seoul, 137-040, Korea. Phone: 82-2-590-2389; Fax: 82-2-593-2389; E-mail: jinwoo{at}catholic.ac.kr

Received 11/23/03. Revised 5/12/04. Accepted 5/26/04.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. National Cancer Institute. Screening for hepatocellular cancer—screening/detection-health professionals. Bethesda, MD: NIH; 2001 [cited June 2004]. Available from cancernet.nci.nih.gov.
  2. Sherman M. Surveillance for hepatocellular carcinoma. Semin Oncol, 28: 450-9, 2001.[Medline]
  3. Sherman M, Peltekian KM, Lee C. Screening for hepatocellular carcinoma in chronic carriers of hepatitis B virus: incidence and prevalence of hepatocellular carcinoma in a North American urban population. Hepatology, 22: 432-8, 1995.[CrossRef][Medline]
  4. Oka H, Tamori A, Kuroki T, Kobayashi K, Yamamoto S. Prospective study of alpha-fetoprotein in cirrhotic patients monitored for development of hepatocellular carcinoma. Hepatology, 19: 61-6, 1994.[CrossRef][Medline]
  5. Ishii M, Gama H, Chida N, et al Simultaneous measurements of serum alpha-fetoprotein and protein induced by vitamin K absence for detecting hepatocellular carcinoma. South Tohoku District Study Group. Am J Gastroenterol, 95: 1036-40, 2000.[Medline]
  6. Kuromatsu R, Tanaka M, Shimauchi Y, et al Usefulness of ED036 kit for measuring serum PIVKA-II levels in small hepatocellular carcinoma. J Gastroenterol, 32: 507-12, 1997.[Medline]
  7. Okuda H, Nakanishi T, Takatsu K, et al Measurement of serum levels of des-gamma-carboxy prothrombin in patients with hepatocellular carcinoma by a revised enzyme immunoassay kit with increased sensitivity. Cancer (Phila), 85: 812-8, 1999.
  8. Yamashita F, Tanaka M, Satomura S, Tanikawa K. Prognostic significance of Lens culinaris agglutinin A-reactive alpha-fetoprotein in small hepatocellular carcinomas. Gastroenterology, 111: 996-1001, 1996.[CrossRef][Medline]
  9. Song BC, Chung YH, Kim JA, et al Transforming growth factor-beta1 as a useful serologic marker of small hepatocellular carcinoma. Cancer (Phila), 94: 175-80, 2002.
  10. Ko J, Lee YH, Hwang SY, et al Identification and differential expression of novel human cervical cancer oncogene HCCR-2 in human cancers and its involvement in p53 stabilization. Oncogene, 22: 4679-89, 2003.[CrossRef][Medline]
  11. Kennett RH. Cell fusion. Methods Enzymol, 58: 345-59, 1979.[Medline]
  12. Yi KS, Chung JH, Lee YH, et al Inhibition of EGF-induced activation of phospholipase C-gamma 1 by a single chain antibody fragment. Oncogene, (4 Pt 1);20: 7954-64, 2001.[CrossRef]
  13. Okuda K. Hepatocellular carcinoma. J Hepatol, 32(Suppl 1): 225-37, 2000.
  14. Yamagamim H, Moriyama M, Matsumura H, et al Serum concentrations of human hepatocyte growth factor is a useful indicator for predicting the occurrence of hepatocellular carcinomas in C-viral chronic liver diseases. Cancer (Phila), 95: 824-34, 2002.[CrossRef]
  15. Adinolfi A, Adinolfi M. Alpha-fetoprotein during development and in disease. J Med Genet, 12: 138-51, 1975.[Abstract/Free Full Text]
  16. Lok AS, Lai CL. Alpha-fetoprotein monitoring in Chinese patients with chronic hepatitis B virus infection: role in the early detection of hepatocellular carcinoma. Hepatology, 9: 110-5, 1989.[Medline]
  17. Johnson PJ. The role of serum alpha-fetoprotein estimation in the diagnosis and management of hepatocellular carcinoma. Clin Liver Dis, 5: 145-59, 2001.[CrossRef][Medline]
  18. Oka H, Kurioka N, Kim K, et al Prospective study of early detection of hepatocellular carcinoma in patients with cirrhosis. Hepatology, 12: 680-7, 1990.[Medline]
  19. Okuda K, Kojiro M. Small hepatocellular carcinoma Okuda K Ishak KG K. G. eds. . Neoplasm of the liver, p. 215-26, Verlag Tokyo 1987.
  20. Mita Y, Aoyagi Y, Yanagi M, Suda T, Suzuki Y, Asakura H. The usefulness of determining des-gamma-carboxy prothrombin by sensitive enzyme immunoassay in the early diagnosis of patients with hepatocellular carcinoma. Cancer (Phila), 82: 1643-8, 1998.
  21. Tanaka Y, Kashiwagi T, Tsutsumi H, et al Sensitive measurement of serum abnormal prothrombin (PIVKA-II) as a marker of hepatocellular carcinoma. Hepatogastroenterology, 46: 2464-8, 1999.[Medline]
  22. Kuromatsu R, Tanaka M, Tanikawa K. Serum alpha-fetoprotein and lens culinaris agglutinin-reactive fraction of alpha-fetoprotein in patients with hepatocellular carcinoma. Liver, 13: 177-82, 1993.[Medline]
  23. Taketa K, Endo Y, Sekiya C, et al A collaborative study for the evaluation of lectin-reactive alpha-fetoproteins in early detection of hepatocellular carcinoma. Cancer Res, 53: 5419-23, 1993.[Abstract/Free Full Text]
  24. Sato Y, Nakata K, Kato Y, et al Early recognition of hepatocellular carcinoma based on altered profiles of alpha-fetoprotein. N Engl J Med, 328: 1802-6, 1993.[Abstract/Free Full Text]
  25. Shiraki K, Takase K, Tameda Y, Hamada M, Kosaka Y, Nakano T. A clinical study of lectin-reactive alpha-fetoprotein as an early indicator of hepatocellular carcinoma in the follow-up of cirrhotic patients. Hepatology, 22: 802-7, 1995.[CrossRef][Medline]
  26. Sassa T, Kumada T, Nakano S, Uematsu T. Clinical utility of simultaneous measurement of serum high-sensitivity des-gamma-carboxy prothrombin and Lens culinaris agglutinin A-reactive alpha-fetoprotein in patients with small hepatocellular carcinoma. Eur J Gastroenterol Hepatol, 11: 1387-92, 1999.[Medline]
  27. Lim JH, Kim CK, Lee WJ, et al Detection of hepatocellular carcinomas and dysplastic nodules in cirrhotic livers: accuracy of helical CT in transplant patients. Am J Roentgenol, 175: 693-8, 2000.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Clin. Cancer Res.Home page
S. S. Jung, H. S. Park, I. J. Lee, H. Namkoong, S. M. Shin, G. W. Cho, S.-A. Ha, Y. G. Park, Y. S. Lee, J. Ko, et al.
The HCCR Oncoprotein as a Biomarker for Human Breast Cancer
Clin. Cancer Res., November 1, 2005; 11(21): 7700 - 7708.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yoon, S. K.
Right arrow Articles by Kim, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yoon, S. K.
Right arrow Articles by Kim, J. W.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online