The contributions of cytokines to the development and progression of HCC is well documented [2]. Cytokine production is thought to play a control role in regulating the recruitment of tumor associated inflammatory cells, in the induction of angiogenesis and in the direct modulation of tumor cell proliferation [3-5]. Tumor necrosis factor-α (TNF-α) is a 17 KDa protein produced primarily by mononuclear phagocytes and lymphocytes [6]. It has been reported that TNF-α is involved in the pathogenesis of a diversity of liver diseases including viral hepatitis and HCC [7]. The immunomodulatory effect of TNF-α is achieved via binding to specific cellular receptors (TNF-Rs); TNF-R55 and TNFR75 [8]. The TNF-TNFRs system has been implicated in a wide range of biological functions including cytotoxicity against tumors and virus-infected cells, direct antiviral and proinflammatory activities as well as stimulation of many immune effector cells [9]. TNF-α plays a central role in up-regulating cell surface adhesion molecules such as intracellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), E-selectin and P-selectin, thereby promoting adhesion of leukocytes to and subsequent transmigration through the endothelial barrier [10].

The expression of adhesion molecules, which lead to the interaction of particular sets of cells, may be the primary signal for the activation and extravasation of immune cells into the set of inflammation [11]. P-selectin (CD62, GMP- 140, PADGEM) belongs to the selectin family of adhesion molecules. P-selectin acts as a receptor that supports binding of leukocytes to activated platelets and endothelium. Pselectin- mediated adhesive interaction operates in conjunction with cell-cell interactions directed by related molecules and are likely to be important in both hemostatic and inflammatory processes [12]. The physiologic role of Pselectin might be the mediation of initial leukocyte adhesion to activated endothelium during acute inflammation. It may work in concert with E-selectin to direct early, regionally specific adherence of neutrophils and monocytes at sites of acute inflammation. A soluble form of P-selectin found in serum and plasma has been described which might represent a proteolytic fragment or more likely a soluble splice variant lacking the transmembrane domain [13]. Pselectin performs an important role in tumor formation and metastasis since malignant cells were shown to express receptors for P-selectin [14]. Soluble adhesion molecules (sP-selectin) like their membrane bound counterparts may affect cell-cell interactions. Thus identification of soluble forms of these molecules, provide the potential for further clinical study of their role in the monitoring of the inflammatory and malignant diseases [15].

Gamma-glutamyl transferase (GGT, Enzyme commission number 2.3.2.2.) is a membrane-bound ectoenzyme that catalyzes the degradation of glutathione and other γ- glutamyl compounds [16]. Serum GGT activity is a sensitive marker of hepatobiliary disorders. It is generally accepted as the most sensitive marker of cholestasis and pancreatic disease or enzymatic induction by alcohol and drugs [17]. Serum GGT levels are affected by several other factors and a variety of clinical conditions; that measurement of serum total GGT lacks specificity, and its value in the differential diagnosis is limited because of great overlap in patients with benign or malignant liver diseases [18]. GGT mediates the uptake of glutathione, by breaking down extracellular glutathione and making its amino acid component available to the cell [19].

Glutathione S-transferase-Pi (GST-Pi) is an acidic, 25 KDa dimeric protein. It is a detoxifying isoenzyme that catalyzes the nucleophilic addition of glutathione at electrophilic centers of wide variety of compounds, hence removal of such xenobiotics from the body becomes possible. The amount of GST-Pi in primary hepatomas are several ten-fold higher than in normal liver [1,20,21].

Alphafetoprotein (AFP) is a well-established cell differentiation and tumor marker. Activity transcribed by fetal and postnatal hepatocytes, AFP is almost undetectable in adult liver parenchymal cells, and its production can resume in the liver under pathologic conditions, such as primary hepatocellular carcinoma (HCC). An elevated AFP level is a sensitive marker for the diagnosis of HCC. However, higher false positivity and false negativity of the serum AFP marker also may be found in the diagnosis of HCC [22]. An elevated concentration may be found in certain benign liver diseases [e.g. schistosomal hepatic fibrosis (SHF)]; 20% to 30% of patients with HCC had a negative AFP result, while some cases of chronic hepatitis and cirrhosis showed a positive AFP result [22].

The aim of the present study was to determine the levels of serum TNF-α, serum sP-selectin, serum GGT, plasma GST-Pi and serum AFP in patients with HCC before and after 3 cycles of single intravenous agent chemotherapy (adriamycin) to determine their value as tumor markers in HCC patients for diagnosis and follow up.

The HCC patients were subjected to proper and full history recording, thorough clinical examination, routine laboratory investigations including complete blood count (CBC) and liver function tests, ultrasonography (USG) of the liver, chest X-ray (when needed) and USG-guided needle liver biopsy to establish the pathological diagnosis.

Serum samples from the control and benign groups and the malignant group [before and during chemotherapy; i.e. after 3 cycles of single agent chemotherapy (50 mg adriamycin/ m² of body surface area, given intravenously, every three weeks)] were used for determination of TNF-α, sPselectin, GGT, GST-Pi and AFP [23].

Determination of serum TNF-α and sP-selectin was carried out using a solid phase sandwich enzyme immunosorbent assay (ELISA) kit (Immuno Tech, France for TNF-α and Bender MedSystems for sP-selectin). Briefly, serum samples were added to the wells of microtitre plate precoated with specific monoclonal antibodies for TNF-α and sP-selectin. After incubation at room temperature and washing of unbound antigen, the enzyme-linked polyclonal antibodies specific for TNF-α and sP-selectin were added to the wells. After washing to removal all unbound enzyme, the substrate was added to induce a colored reaction product. The color reaction was stopped and the intensity was measured at 450 nm. The intensity of this color is directly proportional to the concentration of TNF-α or sP-selectin present in the sample. Results are expressed as pg/ml for TNF-α and ng/ml for sP-selectin.

Determination of serum GGT was determined by the method of Rosalki and Tarlow [24], where the almost colorless L-γ-glutamyl-p-nitroaniline used as substrate from which the enzyme γ-glutamyl transferase liberates yellow p-nitroaniline so that the increase in color gives a measure of its activity, glycylglycine was used as glutamyl acceptor.

GST-Pi was labeled with iodine-125 using the chloramine- T reaction by modification of the method of Hayes et al. [25]. In brief, GST-Pi (Sigma Chemical Company Lts; St. Louis, MO 63178 USA, 10 µl; 5 µg) and 1251-labeled sodium iodide (MDS Nordion S.A; B-6220 Fleurus, Belgium, 5 µl; 500 µCi, carrier free) were added to a reactivial (Pierce Co, USA), followed by chloramine-T (10 µl; 16 µg). After 30 seconds the reaction was stopped by cysteine (100 µl; 56 µg), potassium iodide (10 µl; 100 µg) and elution buffer (250 µl). The contents of the reactivial were then equilibrated with the elution buffer and transferred to a (1.6 cm±35 cm) gel filtration chromatographic column packed with sephadex G-25 (fine). The labeling reaction mixture was applied to the column and was eluted with the elution buffer and the fractions with the highest both radioactivity (CPM) and protein content (Abs280, Lowry method) were used as the tracer for our improved RIA kit [26]. The serum GST-Pi levels in different study groups were measured using a modified RIA procedure of Fan et al. [27] in which the first antibody was rabbit anti-human GST-Pi (DAKO A/S; produktionsvej 42, DK-2600 Glostrup, Denmark, diluted 1:25), the second antibody was swine anti-rabbit immunoglobulin antibody (DAKO A/S, diluted 1:10). GST-Pi was expressed as µg/l.

Serum AFP was measured using a ready-to-use-IRMA (Immunoradiometric assay) kit (Diagnostic Products Corp. Los Angeles, U.S.A) [1]. Results are expressed as IU/ml.

Statistical method
Data were subjected to an analysis of variance using the general model procedure (SAS Institute, 1994). Variables having a significant F-test (p < 0.05) were compared using the least significant difference (LSD) test [28].

Table 1: Results of serum TNF-α, sP-selectin, GGT, GST-Pi and AFP in the control, benign (SHF) and malignant (HCC) groups before and after 3 cycles of chemotherapy

In follow up study for the malignant group, the levels of TNF-α, sP-selectin and GST-Pi showed a significant decrease after 3 cycles of single intravenous agent chemotherapy with adriamycin.

No correlation was reported between the studied parameters and tumor size in HCC patients group (Table 2).

Table 2: The spearman correlation (r) and its P value for all studied parameters related to the tumor size in the HCC group

In malignant group (Table 3), the serum levels of GGT and GST-Pi were significantly higher in HCC patients with poorly differentiated tumors than in patients with well and moderately differentiated tumors. While serum levels of TNFα, sP-selectin and AFP were insignificantly higher in HCC patients with poorly differentiated tumors than in well and moderately differentiated tumors.

Table 3: Relationship between the studied parameters and the pathological differentiation in the HCC patients group

In our study, the level of all studied parameters were significantly increased in HCC patients before chemotherapy compared to SHF group and control group and this revealed that TNF-α, sP-selectin, GGT, GST-Pi and AFP levels may have diagnostic role in HCC patients.

The levels of TNF-α, sP-selectin and GST-Pi were significantly decreased after 3 cycles of treatment with adriamycin compared to patients before treatment. The noticeable reduction in all parameters during treatment mirrored the response to chemotherapy and may be due to the direct effect of adriamycin on cells or hepatic retention or both. This may indicate the validity of these parameters in the follow up of HCC patients treated with adriamycin.

According to the pathologic differentiation of the HCC group it was found that GGT and GST-Pi were significantly increased, while TNF-α, sP-selectin and AFP levels were insignificantly increased in patients with poorly differentiated tumors than those with well and moderately differentiated tumors and that revealed a good correlation between the levels of both GGT and GST-Pi and pathological differentiation.

The actual stimulus for increased TNF-α production in patients is not clear. It is generally accepted that hepatitis B and/or C viruses are major causes of chronic liver disease such as chronic hepatitis and cirrhosis as well as HCC [29]. Larrea et al. [30], have shown that hepatitis infection is associated with increased transcriptional expression of the TNF-α gene in the liver with high serum levels of TNF-α. Several reports postulated that the production of TNF-α in the liver take place not only in the non-parenchymal cells but also in hepatocytes [30]. In addition, Chisari [31] observed that in hepatitis B and/or C infected patients the cytotoxic T-lymphocytes (CTLs) control the viral replication not only by direct cytotoxic activity of infected cells but also via the production of IFN-γ and TNF-α. It seems also possible that oxidative distress might be involved in the induction of TNF-α in HCC or SHF patients [32]. Another possibility related to the increased levels of TNF-α in patients is its release from malignant cells or from other non-malignant cells (as T-cells or macrophages upon their activation) [33].

In our study, elevated levels of sP-selectin in both benign and malignant groups may be attributed to many factors. Some of these factors were studied in this work such as TNF-α which induces the expression of P-selectin, affects the release of P-selectin from cells and seems to be important for hepatic injury in HCC or SHF patients. Also elevated levels of serum sP-selectin may indicate the activation and/or damage of platelets and endothelial cells in these patients groups. However, it is not known whether the sP-selectin is secreted rather than shed from platelets or endothelial cells [34]. The increased levels of TNF-α and sp-selectin in the HCC patients more than in SHF patients can attribute to the more intense inflammation and severity of the disease which may be occurring in HCC patients.

In our study, the possibility of diagnosis and follow up values of GST-Pi were confirmed by observations of both Fan et al. [27] and Niitsu et al. [21] The presence of a significant direct strong correlation between the GST-Pi and GGT levels and the pathological differentiation in HCC patients and the absence of correlation between both GGT and GST-Pi and the tumor size in the same patients can be interpreted in that the tumor production of GST-Pi and GGT is related to the biological characteristics of the tumor cells but not to the tumor size.

The diagnostic value of serum AFP that was reported in this study was confirmed by observations of Abd El- Moneim et al. [35] and Chen and Sung [36]; on the other hand, these results were encountered by the observation of Chen et al. [22]. Our results of the non-usefulness of using serum AFP to follow up HCC patients was confirmed by observation of Johnson and Williams [37]. The absent correlations we found between the serum AFP level and both the pathological differentiation and the tumor size in HCC patients were typically confirmed by observation of Chen et al [22].

In conclusion, all the studied parameters may be used in HCC diagnosis, while only serum TNF-α, sP-selectin and GST-Pi are valuable in monitoring chemotherapy. A study including larger number of cases and different therapeutic modalities is mandatory. In addition, serum GGT and GST-Pi would correlate with HCC grades.

1) Morsi MI, Abd El-Moneim N, Hassan AAM, et al. Radioimmunoassay of plasma glutathione S-transferase-Pi as a tumor marker in hepatocellular carcinoma. Bull Alex Fac Med 2001;XXXVII:379-84.

2) Pardoll DM. Paracrine cytokine adjuvants in cancer immunotherapy. Annu Rev Immunol 1995;13:399-405.

3) Aref S, Menessy A. Correlation of soluble IL-2R and tumor necrosis factor α receptor (TNF-αR) levels with severity of chronic hepatitis C liver injury. The Egypt J Hematol 1997;22:327-40.

4) El-Aggan H, Hegab SE, Baddour NM, et al. Angiogenin in hepatocellular carcinoma: Correlation with tumor vascularity and proliferative activity. Egypt J Immunol 1998;5:167-79.

5) Quentmeier H, Driks WG, Fleckenstein D, et al. Tumor necrosis factor-α induced proliferation requires synthesis of granulocyte macrophage colony-stimulating factor. Exp Hepatol 2000;28:1008-15.

6) Tracy KJ, Vlassara H, Cerami A. Cachectin/tumor necrosis factor. Lancet 1989;11:1122-5.

7) Romieu R, Lacabanne V, Kayibanda M, et al. Critical stages of tumor growth regulation in transgenic mice harboring a hepatocellular carcinoma revealed by distinct patterns of tumor necrosis factor-a and transforming growth factor-β mRNA production. Intl Immunol 1997;9:1405-13.

8) Wallch D, Varfolomeev EE, Malinin NL, et al. Tumor necrosis factor receptor and Fas signaling mechanisms. Annu Rev Immunol 1999;17:331-7.

9) Schumann J, Ties G. Pathophysiological mechanisms of TNF during introxication with natural and man-made toxins. Toxicol 1999;138:1453-9.

10) Wolf D, Hallmann R, Sass G, et al. TNF-α induced expression of adhesion molecules in the liver is under the control of TNFR1-relevance for concanavalin-A-induced hepatitis. J Immunol 2001;166:1300-7.

11) Silva-Teixeira DN, Doughty BL, Goes AM. Human schistosomiasis: Modulation of in vitro granulomatous hypersensitivity in lymphocyte proliferative response by macrophages undergoing differentiation. Infect Immunity 1996;44:552-9.

12) McEver RP. Leukocyte interactions mediated by P-selectin. In: Lipsky PE, Rothlein R, Kishimoto TK, Faanes RB, Smith CW, editors. Structure, function and regulation of molecules involved in leukocyte adhesion. Springer Verlag, 1991;135- 50.

13) Ushiyama S, Laue TM, Moore KL, et al. Structural and functional characterization of monomeric soluble P-selectin and comparison with membrane P-selectin. J Biol Chem 1993;268:15229-37.

14) Lotan R, Raz A. Endogenous lectins as mediators of tumor cell adhesion. J Cell Biochem 1988;37:107-17.

15) De Rose V, Oliva A, Messore B, et al. Circulating adhesion molecules in cystic fibrosis. Am J Resp Critic Med 1998;157:1234-9.

16) Nemesanszky E, Lott JA. Gamma-glutamyltransferase and its isoenzymes: progress and problems. Clin Chem 1985;31:797-803.

17) Whitfield JB. Gamma-glutamyl transferase. Crit Rev Clin Lab Sci 2001;38:263-355.

18) Bellini M, Tumino E, Giordani R, et al. Serum gammaglutamyl- transpeptidase isoforms in alcoholic liver disease. Alcohol Alcohol 1977;32:259-66.

19) Yao DF, Huang ZW, Chen SZ, et al. Diagnosis of hepatocellular carcinoma by quantitative detection of hepatoma-specific bands of serum g-glutamyltransferase. Am J Clin Chem 1998;110:743-9.

20) Wu JW, Meng XY, Xu KC, et al. Simultaneous determination of multiple markers of primary liver cancer: diagnostic significance. J Gastroenterol Hepatol 1988;3:29-35.

21) Niitsu Y, Takahashi Y, Saito T, et al. Serum glutathione Stransferase- Pi as a tumor marker for gastrointestinal malignancies. Cancer 1989;63:317-23.

22) Chen DS, Sung JL, Shen JC, et al. Serum alpha-fetoprotein in the early stage of human hepatocellular carcinoma. Gastroenterol 1984;86:1404-9.

23) Lowitz B, Casciato D. Gastrointestinal tract cancer. In: Casciato D and Lowitz B, editors. Manual of clinical oncology. 2nd ed. Boston, Toronto: Little Brown and Company, 1988;129-49.

24) Rosalki SB, Tarlow D. Optimized determination of gammaglutamyltransferase by reaction-rate analysis. Clin Chem 1974;20:1121-4.

25) Hayes JD, Giligan D, Chapman B, et al. Purification of human hepatic glutathione S-transferase and the development of a radioimmunoassay for their measurement ion plasma. Clin Chim Acta 1983;134:107-21.

26) Ohnishi S, Barr J. A simplified method of quantitating protein using the Biuret and Phenol reagents. Anal Biochem 1978;86:193-200.

27) Fan KC, Huang YC, Li CH. Radioimmunoassay for plasma glutathione S-transferase-Pi and its clinical application in gastrointestinal cancer. Cancer 1995;76:1363-7.

28) SAS institute (1994). SAS/SATAT® User’s guide: Statistics. Ver. 6.04, 4th ed. SAS Institute Inc., Cary, N.C.

29) Shoji K, Shuhei N, Kazuhiro H, et al. Clinical significance of prior hepatitis B virus infection in patients with hepatitis C virus-related hepatocellular carcinoma. Am Cancer Soc 1999;20:793-8.

30) Larrea E, Garcia N, Qian C, et al. Tumor necrosis factor alpha gene expression and the response to interferon in chronic hepatitis C. Hepatol 1996;23:210-7.

31) Chisari FV. Cytotoxic T-cell and viral hepatitis. J Clin Invest 1997;99:1472-7.

32) Adamson GM, Billing RE. Tumor necrosis factor induced oxidative stress in isolated mouse hepatocytes. Arch Biochem Biphys 1992;294:223-9.

33) Qin LX, Tang ZY. The prognostic molecular markers in hepatocellular carcinoma. World J Gastroenterol 2002;8:385- 92.

34) Vasoux R, Hoffman PA, Tomita JK, et al. Cloning and characterization of a new intracellular adhesion molecule ICAM-3. Nature 1992;360:485-8.

35) Abd El-Moneim N, Awad S, Morsi M, et al. Alpha-fetoprotein and hepatitis B virus markers in relation of liver cirrhosis and to hepatocellular carcinoma during chemotherapy. Bull Alex Fac Med 1997;23:223-30.

36) Chen D, Sung J. Serum alpha-fetoprotein in hepatocellular carcinoma. Cancer 1997;40:779-83.

37) Johnson P, Williams R. Serum alpha-fetoprotein estimations and doubling time in hepatocellular carcinoma: influence of therapy and possible value in early detection. JNCI 1980;64;1329-32.