Liver cancer is one of the most common types of cancer worldwide, with the highest incidence rates reported in Asia and Africa. Hepatocellular carcinoma (HCC) mainly includes hepatocellular liver cancer, cholangiocarcinoma and mixed cell carcinoma (1). Of these, HCC accounts for 75-85% of all liver cancers (2). The annual worldwide incidence of HCC is increasing by 3-9% annually (3). Liver cancer has a poor prognosis and is the second leading cause of cancer-related deaths (4), with a 5-year overall survival (OS) rate of <10% (5). Treatment options for early stage liver cancer include liver resection, radio frequency ablation (RFA) and liver transplantation. However, due to its insidious onset, >50% of patients are diagnosed during the middle or late disease stages of the disease, missing the opportunity for curative treatment. Consequently, transcatheter arterial chemoembolization (TACE) has become the first-line treatment for intermediate and advanced-stage liver cancer (6).

Survival times among patients receiving TACE exhibit significant differences due to the heterogeneity of the disease. This is a challenging task for the duplication or cessation of space therapy (7-9). To provide the best individualized treatment to patients with cancer, it is necessary to identify biomarkers that can effectively predict the survival outcomes. Currently, alpha-fetoprotein (AFP) is the most commonly used marker for predicting the onset and recurrence of liver cancer. However, ~1/3 of patients with liver cancer are AFP-negative (10). Various studies have attempted to develop risk prediction models to predict treatment outcomes, including specially designed nomograms and post-TACE prognostic scoring systems (11-13). These models revealed the influence of liver function and the baseline tumor characteristics on the survival of TACE-treated patients with liver cancer. Tumor characteristics such as pathological type, differentiation, tumor size, number of tumors and vascular invasion are the main indicators for predicting prognosis (14-16). Biochemical indicators such as liver function, serum gamma-glutamyl transferase, serum vascular endothelial growth factor, C-reactive protein and novel metabolism-related gene have also attracted significant attention and research (17-19). However, these biomarkers are insufficient for accurately predicting prognosis, with reported inaccuracies in 45% of cases (12,20,21). Furthermore, the scoring system is complex, and its clinical application is difficult to promote. Therefore, such biomarkers are not widely used in clinical practice at present (22).

Serum ferritin (SF) is an iron storage protein composed of 24 subunits and was first discovered by the French scientist Laufberger in 1937(23). SF is the oldest known protein involved in iron metabolism and plays essential roles in cell proliferation, angiogenesis, immunosuppression and iron transport (24). Abnormal SF levels have been shown to be closely associated with tumor progression and poor prognosis (25). Some studies have suggested that SF levels may reflect the extent of liver inflammation and fibrosis, and SF may be a poor prognostic risk factor for survival and recurrence after percutaneous RFA in patients with HCC (26). Furthermore, preoperative SF is an independent prognostic factor for liver cancer after liver resection (27). Despite this, there is limited research on the impact of ferritin levels on the prognosis of patients with HCC, and the current results are conflicting. In addition, the prognostic value of SF in HCC patients undergoing TACE is unclear. Therefore, the present study aimed to investigate the impact of preoperative SF levels on the survival outcomes of patients with HCC undergoing TACE treatment, and to determine whether preoperative SF levels can serve as an independent prognostic biomarker for these patients.

Liver cancer is the second leading cause of cancer-related deaths in China. For most patients with unresectable or inoperable HCC, TACE is considered the first-line treatment option. TACE is considered to cause tumor necrosis by creating a hypoxic environment and producing cytotoxic effects on tumor cells by concentrating high doses of chemotherapy drugs locally on the tumor (28). TACE can improve the quality of life and extend the survival of patients in intermediate or advanced HCC stages (29,30).

SF is a group of proteins that play an important role in iron storage, and is primarily found in the liver, spleen and bone marrow. Under normal physiological conditions SF is mainly composed of light (L) chains; however, in numerous malignant tumors the ratio of heavy (H) ferritin and H/L ferritin increases (31). The reasons for the increase in SF in liver cancer are as follows (32): i) Liver cancer cells can synthesize and secrete ferritin or hetero-ferritin; ii) the uptake and clearance of ferritin in liver cancer tissue are affected; and iii) hepatocyte damage and necrosis cause the release of stored ferritin in the hepatocyte cytoplasm into the bloodstream. Elevated SF levels have also been reported in malignant tumors of the blood system (33), as well as non-tumor diseases, including hemochromatosis, chronic kidney disease, diabetes (34-36), rheumatoid arthritis and adult Still's disease (37). Multiple pathological factors influence the levels of SF, and its instability leads to a lack of specificity. Therefore, predicting prognosis based on SF is challenging.

In the present study, it was determined that 274 ng/ml was the cut-off point for SF. By contrast, Wu et al (27) used 267 ng/ml as the optimal SF cut-off point. The cut-off point for SF in a Korean study cohort was 150 ng/ml, whereas an Italian study reported that the optimal prognostic threshold for SF was 244 ng/ml (26,38). These variations suggested that the normal range of SF may be influenced by factors such as differences in laboratory equipment, region and ethnicity. Thus, the currently published data on SF as a prognostic tool for liver cancer lack generalizability and applicability. Furthermore, SF levels can also be affected by the batch of experimental reagents and equipment. Thus, the accuracy of preoperative SF levels in predicting the prognosis of liver cancer may be compromised.

The correlation between SF levels and clinicopathological variables was then studied, and it was found that HBV infection, AST, ALT and GGT were significantly correlated with preoperative SF levels, while being unrelated to other laboratory parameters. Increased preoperative SF levels were also positively correlated with sex and cirrhosis. AST, ALT and GGT are indicators of liver cell injury. The presence of HBV infection and cirrhosis suggests the impairment of liver function. The destruction of normal liver cells and the presence of liver cancer cells can both lead to the release of ferritin into the bloodstream, resulting in increased SF levels. The correlation between SF levels and clinicopathological variables was also investigated. Wu et al (27) reported that in HCC patients undergoing liver resection TNM and BCLC stages closely correlated with preoperative SF levels while remaining unrelated to other clinicopathological variables. By contrast, Facciorusso et al (26) reported that in HCC patients undergoing RFA treatment, no significant correlation was found between SF levels and other prognostic factors. It was inferred that the different treatments employed may explain this inconsistency in the results.

In the present study, univariate analysis revealed that the presence of extrahepatic metastasis, vascular invasion, cirrhosis, and AST, GGT, AFP and total bilirubin levels were predictors of OS. However, the multivariate Cox analysis refined the number of predictors for OS, focusing on the presence of extrahepatic metastasis and vascular invasion. The presence of extrahepatic metastasis and vascular invasion both indicate tumor progression and are associated with increased mortality. However, the present findings indicated that preoperative SF levels are not an independent predictor of mortality in patients with HCC undergoing TACE. A recent study found limited prognostic value for SF in patients with decompensated cirrhosis, suggesting it may not be an independent predictor of mortality (39). Consequently, the value of SF for liver disease prognosis remains controversial.

The present study had certain limitations that should be acknowledged. First, the analysis did not include changes in SF levels after TACE. Second, no distinction was made between various interventional embolization methods, although research has revealed that drug-eluting bead-transarterial chemoembolization has no advantage over conventional transarterial chemoembolization in patients with unresectable HCC (40). Furthermore, the extended time span of the study means there are no mature guideline for earlier cases as a reference. According to BCLC, the study might not have been suitable candidates for TACE. Fourth, the present study was designed as a retrospective single-group analysis, with a relatively small sample size, which could introduce bias; therefore, its conclusions may require further validation through randomized controlled trials or large-scale prospective cohort studies. Finally, imaging follow-up data were unavailable for numerous patients, resulting in cases where only OS data were available without corresponding disease-free survival data.

In conclusion, this single-center study demonstrated that preoperative SF levels in patients with HCC undergoing TACE was not significantly correlated with prognosis. The present findings indicated that SF has limited utility as a prognostic indicator for patients with HCC.