Biomarkers in hepatocellular carcinoma

Abstract

Hepatocellular carcinoma (HCC) is the most common primary liver cancer. Globally in 2020, liver cancer were the seventh most common cancer form and the second most common form of cancer related death. Early HCCs are often asymptomatic and hard to detect, however, early discovery is crucial for curative treatment. Cirrhosis patients, having an increased risk of HCC development, are therefore offered ultrasonography twice annually to increase the chances of early detection of HCC. However, ultrasonography has its drawbacks. It is operator dependent, time consuming, and the visibility can be hampered by obesity or adjacent organs. Additionally, ultrasonography has a low sensitivity in finding small HCCs. Biomarkers as substitution or addition to ultrasonography is an alternative, however, no biomarker with sufficient performance exists today. HCCs detected in early stages can be offered curative treatment. While liver transplantation has the best prognosis, many patients are still offered liver resection or local ablation where the recurrence rates are higher, due to the shortage of organs. As of today, there is no method to accurately predict recurrence. Biomarkers could play a role here too. Finally, portal vein thrombosis (PVT) is a serious complication to HCC. Neutrophil extracellular traps (NETs) could offer a pathophysiological link between HCC and thrombosis, but this association has not been evaluated.

The aims of the preceding studies are to investigate new potential biomarkers for HCC, either for screening, prognosis, or thrombosis. The focus were plasma proteins, micro-RNAs (miRNAs) and NETs.

The aim of the first study was to perform a wide scan of circulating plasma proteins to identify potential new screening biomarkers for HCC. Proteins were screened with suspension bead array (SBA) and proximity extension assay (PEA). Confirmation was performed with enzyme linked immunosorbent assay (ELISA), a Luminexâ platform or an electrochemiluminescence (ECL) platform. We focused on the target group for HCC screening, i.e. patients with liver cirrhosis. In the verification cohort of 160 patients with cirrhosis or HCC, fibroblast growth factor 21 (FGF21) and thioredoxin reductase 1 (TXNRD1) were significantly correlated to HCC, together with alpha-fetoprotein (AFP) and des- gamma carboxyprothrombin (DCP). Models using different combinations of the mentioned proteins, together with sex and age, were tested. In the end, TXNRD1 was found to be a potential new screening biomarker for HCC when added to AFP, DCP, sex and age.

In the second study, the aim was to identify circulating miRNAs as potential new screening biomarkers for HCC in patients with cirrhosis. RNA sequencing (RNAseq) was used to find upregulated miRNA in HCC tissue as compared to surrounding non-tumorous liver tissue. These upregulated miRNAs were then analyzed in plasma to evaluate which of these that were transferred to plasma. In the last step, the selected miRNAs were analyzed and compared between HCC and cirrhosis patients in a cohort of 200 patients. Eleven miRNAs were found to be increased in plasma from patients with HCC as compared to patients with cirrhosis. While two miRNA, miR-93-5p and miR-151a-3p, were significantly correlated to HCC in stepwise logistic regression, however, no miRNA was found to surpass or increase the performance of AFP, DCP, sex and age in combination.

In the third study, we aimed to explore an association between the miRNA profiles in HCC tissue with the occurrence of tumor recurrence after curative intent liver resection. We found ten differentially expressed miRNAs in HCC tissue when comparing patient with and without recurrence. The ten-miRNA signature could improve a predictive model for recurrence containing the best clinical and histological parameters (AFP, tumor size, microvascular invasion, satellite lesions) in our cohort of 67 patients. Results were verified in a cohort from The Cancer Genome Atlas (TCGA) with 161 patients.

In the last study, we asked whether NETs levels in plasma were increased in patients with liver cirrhosis, and if a concomitant HCC would further increase these levels. NETs were measured as citrullinated histone H3-DNA (H3Cit-DNA), a specific marker of NET formation, in a cohort of 82 HCC patients, 95 cirrhosis patients and 50 healthy controls. H3Cit-DNA were significantly elevated in patients with Child-Pugh B or C as compared to healthy controls and patients with Child-Pugh A. H3Cit-DNA was elevated in cirrhosis and HCC as compared to healthy controls, but no difference was seen between the latter two groups. Also, we found no difference in H3Cit-DNA between patients with or without thrombosis, neither future nor historic. We conclude that H3Cit-DNA is correlated to the degree of cirrhosis, but not to HCC or thrombosis.

To conclude, finding new biomarkers for HCC faces steep challenges. We could not find any circulating miRNAs in plasma that could improve the screening of HCC in cirrhotic patients. Likewise, although H3Cit-DNA as a marker for NET formation was increased in advanced cirrhosis, plasma levels were not correlated to neither HCC nor PVT. However, we found circulating TXNRD1 as a potential candidate that could improve the screening of HCC in cirrhotic patients, a finding which prompts further validation. Furthermore, we could determine a ten-miRNA signature in HCC tissue that improved the prediction of HCC recurrence after curative-intent liver resection. We suggest that all these results should be further validated in independent cohorts.