Tumour-specific induction of immune complexes: DCP-IgM in hepatocellular carcinoma

European Journal of Clinical Investigation Vol 38 571

DOI: 10.1111/j.1365-2362.2008.01985.x

Blackwell Publishing LtdORIGINAL ARTICLESIgM immune complexes for HCC detectionL. Beneduce et al. ORIGINAL ARTICLE

Tumour-specific induction of immune complexes: DCP-IgM in hepatocellular carcinomaL. Beneduce*, G. Pesce*, A. Gallotta*, F. Zampieri*, A. Biasiolo†, N. Tono‡, N. Boscato†, A. Gatta†, P. Pontisso† and G. Fassina*

*XEPTAGEN SpA, Marghera Venezia, Italy, †University of Padova, Italy, ‡Istituto Oncologico Veneto-I.O.V. (IRCCS), Padova, Italy

ABSTRACT

Background In the sera of liver, colorectal and prostate cancer patients, several biomarkers may be detected as IgM immune complexes. To determine whether the presence of immune complexes was correlated to an increase of IgMs, we measured the IgM content in the sera of patients with hepatocellular carcinoma (HCC) and cirrhosis, and evaluated the occurrence of des-gamma-carboxy prothrombin (DCP) as immune complexes (DCP-IgM) compared to the levels of DCP and alpha-fetoprotein (AFP).

Patients and methods Serum samples from 31 patients with cirrhosis, 33 untreated HCC patients diagnosed by ultrasound, computed tomography and/or magnetic resonance and confirmed by histopathology, when indicated, and 30 healthy controls were analysed. Concentrations of IgM and DCP-IgM were determined by ELISAs.

Results Circulating IgM in patients with HCC (median level = 1·79 mg mL–1) and cirrhosis (1·09 mg mL–1) were not significantly different (P = 0·1376) while DCP-IgM were significantly higher in HCC patients (median level = 2171·2 AU mL–1) than in those with cirrhosis (1152 AU mL–1, P = 0·0047). No correlation was found between DCP-IgM and IgM in HCC (r = 0·227) and cirrhosis patients (r = 0·475). DPC-IgM was positive in 55% (18/33) of HCC patients and in 26% (8/31) of cirrhosis patients compared to 39% and 26% for DCP and 48% and 13% for AFP. DCP-IgM, DCP and AFP tests had 100% specificity in healthy controls.

Conclusions DCP-IgM in HCC patients was not associated with an increase in IgM concentration. DCP-IgM was more frequently detected in HCC patients than DCP and AFP, strengthening the diagnostic role of IgM immune complexes for liver cancer.

Keywords Des-gamma-carboxy prothrombin, hepatocellular carcinoma, IgM immune complexes, immunesurveillance, natural immunity.

Eur J Clin Invest 2008; 38 (8): 571–577

Introduction

The immune system has three primary strategies to prevent tumorigenesis: protection of the host from virus-induced tumours, prevention or elimination of the inflammatory processes which could cause tumorigenesis and eventually, identification and specific elimination of tumour cells on the basis of their expression of tumour-specific antigens [1]. The concept of cancer immunosurveillance was first formulated by Paul Ehrlich in the early 1900s [2], but it was almost abandoned in the 1980s since studies conducted in nude mice lacking T cells and B cells did not demonstrate an increase in tumour development compared to wild type mice. Nowadays, a large number of studies in animals [3] and clinical observations in humans [4,5] have provided evidence supporting and refining the cancer immunosurveillance theory now updated as immunoediting [1,3,5].

This current model of cancer immunoediting is divided into three phases: elimination, equilibrium and escape [1,3,5].

The elimination phase acts when the immune system detects tumour cells and completely or partially eliminates them. In the latter case of partial elimination, the immunoediting predicts a temporary equilibrium state between the immune system and the tumour. In this phase the tumour growth is under the control of the immune system and it continues to exert a selective suppression on the transformed cells. If the neoplastic growth process selects the tumour cells variants that are able to evade the anti-tumour immune response, the tumour escapes the immune system control and it becomes free to continue growing [1,3,5].

In humans, tumours develop a series of strategies to evade immunosurveillance, presumably unrelated to the other

www.ejci-online.com L. BENEDUCE ET AL.

572 © 2008 The Authors. Journal Compilation © 2008 Blackwell Publishing Ltd

characteristics of carcinogenesis, resulting from the selective pressure exerted by the immune system [6].

The immune response might be too inefficient to prevent the development of cancer, either due to selection of tumour cells escaping the immune response or to induction of tumour-antigen specific immunotolerance [6]. Tumour immunology investigation has led to the identification of a series of tumour associated antigens [7,8], providing evidence that most tumours trigger an immunogenic response, such as in the case of prostate cancer [9] where the detection of serum autoantibodies may help the diagnosis of prostate cancer [9–11].

In our previous studies we were able to demonstrate that tumour biomarkers lead to the formation of circulating tumour associated antigens-IgM complexes, such as in the case of prostate cancer [11], colorectal cancer [12] and liver cancer [13–16]. We have also shown that the assessment of these circulating immune complexes provided a better diagnostic performance than the analysis of the corresponding free biomarker.

The evidence of the diagnostic relevance of IgM immune complexes in prostate cancer (PC) was provided by the assessment of serum levels of the prostate specific antigen (PSA)-IgM complex [11] while in colorectal cancer (CRC) by the evaluation of the serum levels of carcino-embryonic antigen (CEA)-IgM complex [12]. The measurements of biomarker-IgM immune complexes in the sera of CRC patients or in the sera of PC patients allowed the identification of a much higher number of patients compared to free CEA or free PSA [11,12]. In liver cancer, both alpha-fetoprotein (AFP), which is the conventional serological marker for the detection of hepatocellular carcinoma (HCC), and squamous cell carcinoma antigen (SCCA), a novel HCC biomarker [13], when detected as IgM complexes improved the diagnosis and prognosis of liver cancer [14–16].

In this study, to prove that the IgM immune complexes occurrence in liver cancer was due to a tumour specific response and not to a non-specific IgM increase, we measured, in parallel, the concentration of total IgM and des-gamma-carboxy prothrombin (DCP) as IgM immune complexes (DCP-IgM) in the sera of patients with HCC and cirrhosis.

DCP has been found elevated in the plasma of patients with HCC [17] as a result of an acquired defect in the post translational carboxylation of the prothrombin precursor in neoplastic cells [18]. Several clinical observations reported the usefulness of the assessment of serum DCP to discriminate the liver cancer from non-malignant chronic liver disease [19–22]. Therefore in this study we compared the diagnostic relevance of the three biomarkers DCP-IgM, free DCP and AFP.

Patients and methods

Serum samples from 38 patients with HCC (mean age ± SD = 61 ± 12 years, M/F ratio = 2/1) and from 31 patients with cirrhosis (mean age ± SD = 55 ± 9 years, M/F ratio = 2/1) all

with HCV infection, were collected at the Liver Unit of the Department of Clinical and Experimental Medicine, University of Padua, according to the institutional approved procedures. Serum samples from 30 healthy donors were used as controls (mean age + SD = 57 + 7 years, M/F ratio = 2/1). All patients gave fully informed consent authorizing blood use for research purposes before sample collection. The diagnosis of HCC was based on the presence of hepatic focal lesion > 2 cm detected by liver ultrasound and confirmed by computed tomography and/or magnetic resonance as imaging techniques. The final diagnosis was confirmed by histopathological analysis on ultrasound-assisted fine-needle biopsy, when indicated. Blood samples from patients with HCC were taken prior to initiation of any HCC treatment. Patients with cirrhosis underwent regular liver ultrasound screening to exclude the occurrence of hepatic nodules.

Gel filtration

One hundred microlitres of pooled sera were analysed using a gel filtration column BioSep SEC S-4000 (Phenomenex, Macclesfield, Cheshire, UK) as previously described (13). Fractions were collected every 30 s and immunoreactivity for DCP-IgM was tested by ELISA. Briefly, ELISA plates were coated overnight at 4 °C with 100 μL per well of mouse antihuman DCP antibody (US-biological, Swampscott, MA, USA) at a concentration of 10 μg mL–1 in 50 mm sodium phosphate buffer, pH 7·2, containing 150 mm sodium chloride (PBS) and then blocked for 2 h with 1% bovine serum albumin (BSA) in PBS. After blocking, 100 μL of collected fractions diluted 1 : 2 in PBS containing 1% BSA and 0·05%. Tween 20 (PBS-B-T) were incubated for 1 h at room temperature. The DCP-IgM complex was revealed using peroxidase-conjugated antihuman IgM (Sigma Aldrich, Milan, Italy) and developed with 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) (Sigma Aldrich) and hydrogen peroxide as substrates.

DCP-IgM assay

Ninety-six-well ELISA plates were coated with 1 μg of mouse antihuman DCP antibody in 100 μL of PBS pH 7·2 per well at 4 °C overnight and then blocked for 2 h with 1% BSA in PBS. After blocking, 100 μL of serially diluted reference standards and samples in PBS-B-T were incubated for 1 h at room temperature. The DCP-IgM complex was revealed using peroxidase-conjugated antihuman IgM and developed with ABTS and hydrogen peroxide as substrates. The amount of DCP-IgM was expressed in arbitrary units/mL (AU mL–1) by using purified DCP-IgM calibrators obtained from five patients with HCC, as previously described (15). The assay was standardized and its analytical performance evaluated: coefficient of variation intra-assay and interassay was less than 10%.

IGM IMMUNE COMPLEXES FOR HCC DETECTION

European Journal of Clinical Investigation Vol 38 573

AFP and free DCP measurement

Serum free DCP levels were determined in parallel in each sample using ASSERHACROME PIVKA-II (Diagnostica Stago, Asnieres Sur Seine, France) according to the manufacturer’s instructions.

Serum AFP levels were also determined in parallel in each sample using AFP ELISA kit (DRG Diagnostics, Marburg, Germany) according to the manufacturer’s instructions.

Immunoblotting analysis

Gel-filtration fractions containing DCP-IgM, as assessed by ELISA, were pooled and concentrated by precipitation with polyethylene glycol 6000 then resuspended in reducing Laemmli buffer. The protein samples were separated by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to Hybond ECL nitrocellulose membrane (GE Healthcare, Milan, Italy). The membrane was blocked in 5% Non Fat Milk Solution in PBS containing 0·05% Tween 20 for 1 h at room temperature. Blots were then incubated with peroxidase-conjugated goat antihuman IgM (μ-chain specific, Sigma Aldrich) or, alternatively, with biotinylated mouse antihuman DCP followed by peroxidase-conjugated streptavidin (GE Healthcare). Protein bands were revealed utilizing enhanced chemiluminescence (ECL) reagents (Millipore, Milan, Italy) and Kodak CCD instrumentation (Eastman Kodak Company, Rochester, NY, USA).

Total IgM measurement

Ninety-six well ELISA plates were coated with 1 μg of rabbit antihuman IgM antibody (DakoCytomation, Milan, Italy) in 100 μL of PBS pH 7·2 per well at 4 °C overnight and then blocked for 2 h with 1% BSA in PBS. After blocking, 100 μL of serially diluted IgM calibrators (DakoCytomation) and samples in PBS-B-T were incubated for 1 h at room temperature. The IgM level was revealed using peroxidase-conjugated antihuman IgM (DakoCytomation) and developed with ABTS and hydrogen peroxide as substrates. The amount of IgM was expressed in mg mL–1 by interpolation of samples absorbance on the calibration curves plotted with IgM calibrators. The assay was standardized and its analytical performance evaluated: coefficient of variation intra-assay and interassay was less than 10%.

Statistical analysis

Statistically significant differences between the groups were determined by the Mann–Whitney U-test. A 2-tailed P-value of < 0·01 was used to determine statistical significance. All analyses were performed using Analyse-it® software (Analyse-it software, Leeds, UK).

Results

A total of 64 HCV positive patients suffering from chronic liver disease and 30 healthy controls were analysed. Hepatitis C virus

infection was the most recurrent aetiologic agent among 33 HCC patients and 31 patients with cirrhosis.

The serum concentrations of IgM expressed in mg mL–1 were determined by interpolation of samples absorbance on the calibration curves plotted with IgM calibrators. The calibration curve ranged from 3 × 10–6 mg mL–1 to 2 × 10–4 mg mL–1.

The amount of IgM was not significantly higher in the group of liver cancer patients (median level = 1·79 mg mL–1, concentration range: 0·02–5·00 mg mL–1) compared to those with cirrhosis (median level = 1·09 mg mL–1, concentration range: 0·09–5·07 mg mL–1; Mann–Whitney U-test, P = 0·1376).

To characterize circulating DCP-IgM, serum samples from five patients with HCC were grouped and subjected to gel-filtration analysis. Fractions collected from the column every 30 s were tested for the presence of DCP-IgM. A strong reactivity was observed for DCP-IgM in the fraction eluting at high molecular weight (> 2000 kDa) thus providing the first evidence of the occurrence of circulating DCP-IgM complexes in patients’ sera (Fig. 1). The three fractions (from 13·5 to 14·5 min) representing the elution peak of DCP-IgM immune complexes, according to the highest reactivity for DCP-IgM ELISA, were pooled and analysed by SDS-PAGE followed by Western blotting.

The immunoblotting analysis confirmed that the sample contained components migrating at the expected molecular weight for reduced IgM and DCP (70 kDa) (Fig. 1 inset). The sera of five healthy donors were also grouped and tested as a negative control. There was no apparent difference in the gel filtration profile but the test for the presence of DCP-IgM indicated an absence of significant reactivity (Fig. 2).

The ELISA for DCP-IgM assay was developed by using a standard reference purified DCP-IgM obtained from the gel filtration of the pooled sera from five HCC patients [15]. The serum concentrations of DCP-IgM, expressed in Arbitrary Units/mL (AU mL–1), were determined by interpolation of samples absorbance on the two calibration curves plotted with purified calibrators. The first calibration curve ranged from 21·5 AU mL–1

to 325 AU mL–1, the second from 325 to 1300 AU mL–1.The DCP-IgM cut-off was calculated as mean + 2 SD

(2030 AU mL–1) of serum levels of healthy controls. The entire group of the 30 healthy controls had levels of DCP-IgM below the cut-off.

Circulating immune complexes of DPC-IgM were above the cut-off in 18 of the 33 HCC patients (55% of diagnostic sensitivity; SE) and in 8 of the 31 patients with cirrhosis (74% of diagnostic specificity; SP), compared to 13 out of 33 HCC patients (39% SE) and 8 out of 31 (74% SP) patients with cirrhosis with DCP levels above the 2 ng mL–1 cut-off (Fig. 3, Table 1).

The serum levels of AFP were above the 20 ng mL–1 cut-off in 48% (SE; 16/33) of HCC patients and in four 4 of 31 patients with cirrhosis (87% SP) (Fig. 3, Table 1). The specificity of AFP and free DCP tests was 100% (30/30) in healthy controls.

www.ejci-online.com L. BENEDUCE ET AL.

574 © 2008 The Authors. Journal Compilation © 2008 Blackwell Publishing Ltd

Figure 1 Upper panel: Gel filtration profile of hepatocellular carcinoma (HCC) patients’ sera and immunoblotting analysis of fractions eluting at 13·5–14·5 min (inset) stained with anti-human des-gamma-carboxy prothrombin (DCP) (lane 1) and anti-human IgM (lane 2). Lower panel: Enzyme-linked immunosorbent assay (ELISA) characterization of collected fractions to detect DCP as immune complexes (DCP-IgM).

Figure 2 Gel filtration profile of healthy donors sera (upper panel). ELISA characterization of collected fractions to detect DCP-IgM (lower panel).

IGM IMMUNE COMPLEXES FOR HCC DETECTION

European Journal of Clinical Investigation Vol 38 575

The distribution of serum levels of DCP-IgM was significantly higher in HCC patients (median level = 2171·2 AU mL–1, concentration range: 187·2–26176·0 AU mL–1) than in those with cirrhosis (median level = 1152 AU mL–1, concentration range: 0–7696·0 AU mL–1, Mann–Whitney U-test P = 0·0047), while the distribution of serum levels of free DCP and AFP did not reach statistical significance in the two groups (Mann–Whitney U-test, P > 0·01). No correlation was found between DCP-IgM and IgM levels in patients with HCC and in patients with cirrhosis (HCC, r = 0·227; cirrhosis r = 0·475). Table 1 shows the indexes of diagnostic accuracy and the range of the observed levels in the sera of the HCC and cirrhotic patients for each biomarker analyzed.

The diagnostic accuracy measured as the area under the receiver-operating characteristics (ROC) curve (AUC) showed that DCP-IgM (AUC = 0·706) performed better than free DCP (AUC = 0·650) and AFP (AUC = 0·675) in differentiating liver cancer from cirrhosis (Fig. 4).

Table 1 Comparison of specificity, sensitivity, and range of concentration in HCC patients (n = 33) and cirrhosis group (n = 31) for DCP-IgM, free DCP and AFP

Biomarker Sensitivity Specificity Range in HCC Range in cirrhosis

DCP-IgM (2030 AU mL−1) 55% 74% 187·2–26176·0 (AU mL−1) 0–7696·0 (AU mL−1)

DCP (2 ng mL−1) 39% 74% 0·1–753·3 (ng mL−1) 0–642·9 (ng mL−1)

AFP (20 ng mL−1) 48% 87% 0–10072·4 (ng mL−1) 0–86 (ng mL−1)

Figure 3 Levels of IgM, DCP-IgM, DCP and alpha-fetoprotein (AFP) above their cut-off in patients with cirrhosis and HCC. In order to display the different units of measurement on the same graph, the values were standardized to have mean 0 and variance 1.

Figure 4 Receiver-operating characteristic (ROC) curve analysis of DCP-IgM, free DCP and AFP levels in patients with liver cancer and patients with cirrhosis.

www.ejci-online.com L. BENEDUCE ET AL.

576 © 2008 The Authors. Journal Compilation © 2008 Blackwell Publishing Ltd

Discussion

Multivalent IgMs are typically considered the main component of the innate immunity with the binding capacity of a wide range of tumour antigens [23]. It is well established that natural IgMs play an important role in the first line of defence against infectious agents, in the regulation of proliferation of immune cells and in the immunosurveillance against tumoural cells growth [24,25].

Tumour-specific antigens may be recognized by native or adaptive immune effectors and, in most cases, the cancer cells are destroyed before they become clinically apparent. However, the immune response often might be too inefficient to prevent the development of cancer, either because tumour cells that evade the immune response are selected, or because tumour antigen specific immuno tolerance is induced [1–8,26].

We have recently demonstrated the occurrence of tumour biomarkers associated with IgM thus forming circulating immune complexes in different human tumours, including AFP and SCCA in liver cancer, CEA in colorectal cancer and PSA in prostate cancer [11,12,14–16].

In these cases the immunity effectors have been shown acting without IgG intrusion, since only IgMs were found bound to the tumour antigen, forming circulating immune complexes [15].

The overall results obtained in our previous studies on the analysis of IgM immune complexes for cancer diagnosis [11,12,14–16] led us to speculate that IgM-biomarker immune complexes could be involved in cancer immunoediting, likely reflecting host immune protective mechanisms to apply selective pressure on neoplastic cells trying to overcome the tumour. Natural human polyreactive IgM autoantibodies in the immune complexes could bind the tumour antigens with low affinity but with pentameric avidity [27].

The results of the present study help to clarify the role of IgM in the formation of specific biomarker immune complexes associated with tumour growth since, for the first time, we have provided evidence that the increase of the immune complexes in liver cancer was not associated with an increase of total IgMs, supporting a tumour specific induction of immune complexes.

The results also extended the occurrence of biomarker-IgM complexes in cancer, with the first description of the presence of DCP-IgM immune complexes in the sera of HCC patients.

To date, in liver cancer, two diagnostic biomarkers under the form of IgM immune complexes, namely AFP-IgM and SCCA-IgM, were identified.

To evaluate the diagnostic relevance of this newly identified IgM immune complex in liver cancer we further compared the determination of DCP-IgM to free DCP and AFP.

An analysis of HCC and the sera of cirrhosis patients confirmed the presence of specific DCP-IgM associated only with the liver diseases, whereas all control sera from healthy subjects were negative (100% specificity).

DCP-IgM assessment improved the diagnostic sensitivity for the identification of HCC patients when compared to the DCP or AFP assays (55% vs. 39% or 48%, respectively). The specificity of DCP-IgM for liver cancer, assessed as the percentage of negative patients with cirrhosis, was 74%, compared to 74% and 87% obtained with DCP and AFP determination, respectively. In addition, the diagnostic accuracy measured as AUC was higher for DCP-IgM (0·706) than for free DCP (0·650) and AFP (0·675).

In conclusion, the observed enhancement of diagnostic indexes of DCP-IgM suggests that DCP-IgM immune complexes could be a useful serological marker for HCC, providing additional data for the clinical management of liver cancer by using tumour specific biomarker immune complexes.

Acknowledgements

This study was supported in part by FIRB project 2003, Nanosized Cancer Polymarker Biochip, project code RBLA03S4SP_007.

Address

XEPTAGEN SpA, Marghera Venezia, Italy (L. Beneduce, G. Pesce, A. Gallotta, F. Zampieri, G. Fassina); Department of Clinical and Experimental Medicine, University of Padova, Italy (A. Biasiolo, N. Boscato, A. Gatta, P. Pontisso); Istituto Oncologico Veneto-I.O.V. (IRCCS), Padova, Italy (N. Tono).Correspondence to: Luca Beneduce PhD, XEPTAGEN S.p.A., Via delle Industrie 9, 30175 Marghera Venezia, Italy, Tel.: +39 041 5093910; fax: +39 041 5093884; e-mail: [email protected]

Received 11 January 2008; accepted 14 May 2008

References

1 Swann BJ, Smyth MJ. Immune surveillance of tumours. J Clin Invest 2007;117:1137–45.

2 Ehrlich P. About the current state of cancer research. Ned Tijdschr Geneeskd 1909;5:273–90.

3 Dunn GP, Bruce AT, Ikeda H, Old LJ, Schreiber RD. Cancer immunoediting: from immunosurveillance to tumor escape. Nat Immunol 2002;3:991–8.

4 Buell JF, Gross TG, Woodle ES. Malignancy after transplantation. Transplantation 2005;80:S254–64.

5 Bhardwaj N. Harnessing the immune system to treat cancer. J Clin Invest 2007;117:1130–6.

6 Zitvogel L, Tesniere A, Kroemer G. Cancer despite immunosurveillance: immunoselection and immunosubversion. Nat Rev Immunol 2006;10:715–27.

7 Vollmers HP, Brandlein S. The ‘early birds’: natural IgM antibodies and immune surveillance. Histol Histopathol 2005;20:927–37.

8 Vollmers HP, Brandlein S. Natural IgM antibodies: from parias to parvenus. Histol Histopathol 2006;21:1355–36.

9 Wang X, Yu J, Sreekumar A, Varambally S, Shen R, Giacherio D et al. Autoantibody signatures in prostate cancer. N Engl J Med 2005;353:1224–35.

IGM IMMUNE COMPLEXES FOR HCC DETECTION

European Journal of Clinical Investigation Vol 38 577

10 Nesterova M, Johnson N, Cheadle C, Cho-Chung YS. Autoantibody biomarker opens a new gateway for cancer diagnosis. Biochim Biophys Acta 2006; 1762:398–403.

11 Beneduce L, Prayer-Galletti T, Marcello Grimani Giustinian A, Gallotta A, Betto G, Pagano F et al. Detection of prostate-specific antigen coupled to immunoglobulin M in prostate cancer patients. Cancer Detect Prev 2007;31:402–7.

12 Castaldi F, Marino M, Beneduce L, Belluco C, De Marchi F, Mammano E et al. Detection of circulating CEA-IgM complexes in early stage (stage 1) colorectal cancer. Int J Biol Markers 2005;20:204–8.

13 Pontisso P, Calabrese F, Benvegnù L, Lise M, Belluco C, Ruvoletto MG et al. Overexpression of squamous cell carcinoma antigen variants in hepatocellular carcinoma. Br J Cancer 2004;90:833–7.

14 Beneduce L, Castaldi F, Marino M, Tono N, Gatta A, Pontisso P et al. Improvement of liver cancer detection with simultaneous assessment of circulating levels of free alpha-fetoprotein (AFP) and AFP-IgM complexes. Int J Biol Markers 2004;19:155–9.

15 Beneduce L, Castaldi F, Marino M, Quarta S, Ruvoletto M, Benvegnù L et al. Squamous cell carcinoma antigen-IgM complexes as novel biomarkers for hepatocellular carcinoma. Cancer 2005;103:2558–65.

16 Pontisso P, Quarta S, Caberlotto C, Beneduce L, Marino M, Bernardinello E et al. Progressive increase of SCCA-IgM immune complexes in cirrhotic patients is associated with development of hepatocellular carcinoma. Int J Cancer 2006;119:735–40.

17 Liebman HA, Furie BC, Tong MJ, Blanchard RA, Lo KJ, Lee SD et al. Des-γ-carboxy (abnormal) prothrombin as a serum marker of primary hepatocellular carcinoma. N Eng J Med 1984;310:1427–31.

18 Ono M, Ohta H, Ohhira M, Sekiya C, Namiki M. Measurement of immunoreactive prothrombin precursor and vitamin-K dependent gamma-carboxylation in human hepatocellular carcinoma tissues: decreased carboxylation of prothrombin precursor as a cause of des-gamma-carboxyprothrombin synthesis. Tumor Biol 1990;11:319–26.

19 Okuda H, Nakanishi T, Takatsu K, Saito A, Hayashi N, Takasaki K et al. Serum levels of des-γ-carboxyprothrombin measured using the revised enzyme immunoassay kit with increased sensitivity in relation to clinicopathologic features of solitary hepatocellular carcinoma. Cancer 2000;88:544–9.

20 Marrero JA, Su GL, Wei W, Emick D, Conjeevaram HS, Fontana RJ et al. Des-gamma carboxythrombin can differentiate hepatocellular carcinoma from non malignant chronic liver disease in American patients. Hepatology 2003;37:1114–21.

21 Volk ML, Hernandez JC, Su GL, Lok AS, Marrero JA. Risk factors for hepatocellular carcinoma may impair the performance of biomarkers. A comparison of AFP, DCP, AFP-L3. Cancer Biomark 2007;3:79–87.

22 Toyoda H, Kumada T, Osaki Y, Oka H, Kudo M. Role of tumor markers in assessment of tumor progression and prediction of outcomes in patients with hepatocellular carcinoma. Hepatol Res 2007;37(Suppl. 2):S166–71.

23 Lacroix-Desmazes S, Kaveri SV, Mouthon L, Ayouba A, Malanchere E, Coutinho A et al. Self-reactive antibodies (natural autoantibodies) in healthy individuals. J Immunol Meth 1998;216:117–37.

24 Varambally S, Bar-Dayan Y, Bayry J, Lacroix-Desmazes S, Horn M, Sorel M et al. Natural human polyreactive IgM induce apoptosis of lymphoid cell lines and human peripheral blood mononuclear cells. Int Immunol 2004;16:517–24.

25 Vollmers HP, Brandlein S. Natural antibodies and cancer. J Autoimmun 2007;29:295–302.

26 Schreiber RD. Cancer vaccines 2004 opening address: The molecular and cellular basis of cancer immunosurveillance and immunoediting. Cancer Immun 2005;5(Suppl. 1):1–8.

27 Vollmers HP, Brandlein S. Natural IgM antibodies: The orphaned molecules in immune surveillance. Adv Drug Deliv Rev 2006;58:755–65.