Alphafetoprotein values were frequently elevated in patients with HCC. AFP values >200 ng/ml was seen in 51.7% (31 cases) compared to controls. This is significantly lower than the 75% and 85% obtained by Alpert
] and Fakunle
]. It validates previous studies that have found that not all patients with hepatocellular carcinoma elaborate this tumour marker
]. Values for SCCA were similar in both cases and control with no significant statistical difference (p
In this study AFP had an AUC of 0.916 which makes it a good screening test for the diagnosis of hepatocellular carcinoma. This was quite higher than the 0.717 found by Gianelli
] in his own study which looked at the immunohistochemistry of liver tissues with hepatocellular carcinoma.
AUC for SCCA was noted to be 0.525 with a 95% confidence interval of 0.389-0.661. This study did not find a significant level of discriminatory power of SCCA between HCC patients and healthy control groups unlike previous studies
]. This may be explained by the fact that previous studies involved tissue specimens of both primary nodules of HCC and peritumoural areas. They performed Immunohistochemistry quantification on tissue specimens. SCCA antigen appeared strongly stained in tumoral tissues
]. However, no correlation was observed between serum and tissue levels of SCCA antigen (r
]. SCCA is mainly expressed in cytosol and is not associated with membrane bound organelles, therefore circulating serum levels are likely due to cell lysis rather than a secretory process. This may account for why serum SCCA may not be ideal discriminatory marker for HCC diagnosis.
SCCA sensitivity was 75%, compared to AFP sensitivity of 51.7%. It compares favorably with work done by Gianlugi Giannelli et al
] where SCCA sensitivity was 84.2%. However the specificity of SCCA in our study was poor (26.7%), compared to AFP specificity of 100%. The specificity was also poor in the study by Gianluigi Gianneli (48.9%). This antigen is noted to be common in the skin, saliva and sweat of apparently normal people and this may account for the values that were found in the controls.
There are therapeutic techniques for the treatment of hepatocellular carcinoma in the early stage; hence, emphasis is now being placed on early detection in patients at risk who are presently asymptomatic. Detection of HCC at an early stage may reduce mortality significantly. Mass screening may be justified as the population at-risk can be identified easily and tumour ablation or resection at an early stage can increase survival. However, massive screening should be justified only when sensitive and specific diagnostic procedures are available.
Des- gamma- carboxy- prothrombin (DCP) is another promising tool with limited expense and wide accessibility. A bivariate meta- analysis of the diagnostic performance of DCP by Gao et al
] found a sensitivity of 67%, specificity of 92%, positive likelihood ratio of 92% and an AUC of 0.89. They however noted significant heterogeneity and recommended more prospective studies. DCP assay was not one of the objectives in this present study, however, it is recommended for future considerations.
Presently, SCCA is still a research tool and has not been approved for widespread clinical use in the screening of patients with hepatocellular carcinoma. It is also very expensive. Unfortunately as shown in this study, it also lacks good discriminatory power between with HCC and apparently healthy controls and may not be useful in this subset of patients. We noted that all the cases recruited into this study presented with advanced hepatocellular carcinoma and died within 3 months of presentation, reflecting the poor prognosis at this stage. Present recommendation focuses on 6 monthly ultrasound surveillance and measurement of serum Alphafetoprotein in patients with chronic hepatitis and or cirrhosis.
Thirty one cases (51.7%) were found to be HBsAg positive in this study. This represents a significant association between HCC and HBV as an aetiological agent in this study.
Previous study by Ndububa
] noted a prevalence of 61% of HBsAg in HCC. Our study did not find a significant difference between HCV infection in cases and control. Only 2 cases and 1 apparently healthy control were positive for HCV. While the risk was not significant in this study, it has however been well proven that HCV is an important aetiological risk factor for the development of hepatocellular carcinoma. This is especially true in Western Europe and Asia
] found that 55% of HBsAg negative cirrhotic patients with HCC were anti-HCV positive. The reduced prevalence found in our study suggests that HCV infection may not be a strong risk factor compared to hepatitis B in our patients.
More work need to be done in search of appropriate tumour markers for early detection of hepatocellular carcinoma among Nigerians.
The scope of this study was limited majorly by funds. However, future studies will consider inclusion of controls with cirrhosis and metastatic liver disease. Another major limitation was the inability to study the efficacy of SCCA in patients with early HCC due to late presentation in our environment. This is reflected in our study where all the patients presented with advanced HCC.