The retrospective analysis in this study was conducted at Guangxi Medical University Cancer Hospital, between January 2014 to December 2022. Serum samples for proteomics analysis and metabolomics analysis were collected at Guangxi Medical University Cancer Hospital between September 2022 and February 2023. This study adhered to the ethics outlined in the Declaration of Helsinki and received approval from the Ethics Committee of Guangxi Medical University Cancer Hospital.

A total of 2462 patients diagnosed with HCC underwent curative resection. The inclusion criteria for retrospective study were as follows: (1) HCC confirmed by postoperative pathological analysis, (2) no history of previous anti-cancer therapy, (3) absence of concurrent malignant tumors, (4) availability of comprehensive laboratory, and pathological. On the other hand, the exclusion criteria consisted of the following: (1) individuals with a history of anti-tumor therapy such as radiotherapy, or chemotherapy (n=205), (2) cases without a clear pathological diagnosis (n=106), (3) patients with other tumor diseases (n=221), (4) recurrence of HCC (n=196), (5) unavailability of comprehensive laboratory and pathological data (n=613).

The diagnostic criteria for clonorchiasis were as follows, with any one of the subsequent conditions deemed sufficient for establishing a diagnosis (19). (1) Intraoperative or postoperative pathological examination revealing the presence of adult C. sinensis in the liver or gallbladder. (2) Preoperative fecal examination showing the presence of C. sinensis eggs. Based on above criteria, a total of 1121 patients were included in this study, among whom 118 exhibited concomitant C. sinensis infection along with HCC. Therefore, the patients were divided into two groups: the HCC group (without C. sinensis infection) and the HCC_Cs group (HCC combined with C. sinensis infection).

The data collection process encompassed multiple aspects, including: (1) General information: gender and age. (2) Hematological tests: including various tumor markers such as alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), and hepatitis B surface antigen (HBsAg). (3) Pathological indicators: liver cirrhosis, number of tumors, tumor differentiation degree (based on the Edmondson-Steiner histological grading system) and the presence of microvascular invasion (MVI). After excluding those without follow-up data (n=216), a total of 905 patients with HCC were included in the calculation of overall survival (OS) and recurrence-free survival (RFS).

Excluding those with no follow-up data (n=216), a total of 905 patients with HCC were subjected to the calculation of OS and RFS. The exclusion process for participants in this study was graphically displayed in Figure 1 .

The concentration of alanine aminotransferase (ALT), aspartate aminotransferase (AST), lipase (LPS), apolipoprotein A (APO-A), apolipoprotein B1 (APO-B1), high density lipoprotein cholesterol (HDL-C), triglyceride (TG), apolipoprotein A1/apolipoprotein B1 (A1/B1), transferrin (TRF), and blood ammonia (AMM) were determined utilizing a Siemens ADVIA 2400 chemistry analyzer (Siemens, Munich, Germany). Meanwhile, the levels of AFP and CEA were quantified employing an Abbott I2000SR analyzer (Abbott, Illinois, USA).

All patient follow-up information was diligently managed by professionals. Disease status or date of death was determined through telephone contact or an outpatient monitoring system. Tumor recurrence was identified by analyzing radiological observations from CT or MRI scans, with a focus on discerning characteristic enhancement patterns that indicate of intrahepatic recurrence. For extrahepatic tumors or those with atypical HCC imaging traits, verification was secured through biopsy. Patients underwent systematic monitoring at specific time intervals subsequent to their surgical procedures. OS was calculated by determining the time between the date of liver resection and the date of death or last follow-up until August 30, 2023. RFS was defined as the time interval between the date of liver resection and the date of death or last follow-up also until August 30, 2023.

For proteomic analyses, 8 serum samples were selected from HCC and C. sinensis positive HCC (HCC_Cs) patients, respectively. All the serum samples were obtained from patients with HCC diagnosed on first admission, with no history of tumor treatment and no concomitant tumors. The proteins of both groups were detected using 4D Label Free Quantitative proteomics technology. EASY-nLC 1000 system (Thermo, Massachusetts, USA) and a timsTOF Pro2 mass spectrometer (Bruker, Karlsruhe, Germany) were used to solubilize peptides and subsequently analyzed by Untargeted liquid chromatography-tandem mass spectrometry (LC-MS/MS). The MS/MS scan range was set from 100 to 1700 m/z. Data acquisition was performed using the parallel accumulation serial fragmentation (PASEF) acquisition mode. Raw data were searched using MaxQuant software. All data were analyzed through the online cloud platform of Majorbio (https://www.majorbio.com/). DEPs were classified based on three categories: biological process (BP), cellular component (CC), and molecular function (MF) using Gene Ontology (GO, http://geneontology.org/) annotation. Pathway enrichment analysis of DEPs was conducted using the Kyoto Encyclopedia of Gene and Genomes (KEGG) pathway database (http://www.genome.jp/kegg/). The STRING protein interaction database (https://string-db.org/) was used to analyze the protein-protein interaction (PPI) network.

12 serum samples were selected from the HCC and HCC_Cs patient groups, respectively. All the serum samples were obtained from patients with HCC diagnosed on first admission, with no history of tumor treatment and no concomitant tumors. LC-MS/MS was performed to detect metabolites in both groups. The subsequent analytical procedures were executed by Majorbio Bio-Pharmm Technology Co., Ltd. (Shanghai, China) according to the reference method described below (20): briefly, after the samples were processed accordingly, the tests were performed on a UHLC-Q Active HF-X system (Thermo, Massachusetts, USA). The samples were separated by HSS T3 column (Waters, Milford, USA) and then detected by mass spectrometry. Mass spectrometry signals were acquired in both positive and negative ion scanning modes. The parameters were set as follows: spray voltage of 3.5 kV and -3.5 kV, a scanning range of 70-1050 m/z, normalized collision energy set at 20-40-60V, and primary and secondary mass spectrometry resolutions of 60000 and 7500, respectively. Data was collected using the DDA mode. Subsequently, the raw metabolic data was processed using Progenesis QI (Waters Corporation, Milford, USA) with multiple analytical techniques to generate a comprehensive data matrix for further analysis, including relative standard deviation (RSD) distribution, Orthogonal Partial Least-Squares Discriminant Analysis (OPLS-DA) analysis, Venn diagram analysis, identification of differentially expressed metabolites (DEMs), screening of significant metabolites, cluster analysis of DEMs, and KEGG pathway analysis.

Correlation analyses were conducted on the Majorbio cloud platform to examine the relationships between DEPs and DEMs based on the proteomics and metabolomics datasets. The analyses included two-way orthogonal partial least squares (O2PLS) analysis, functional enrichment analysis, Venn diagram analysis, and expression correlation analysis.

Statistical analysis was performed using IBM SPSS Statistics software version 26.0 and R version 4.2.1. Intergroup differences for categorical data, presented as ratios, were compared using either the Chi-square test or Fisher’s exact test. The Mann-Whitney U test was used to compare non-normal continuous data. Survival curves were generated using the Kaplan-Meier method, and intergroup comparisons of OS and RFS rates were performed using the log-rank test. For proteomic analyses, Student’s t-test (two-tailed) was used with a P value < 0.05 and |FC| ≥ 2 as the filtering criteria. For metabolomics data, FC analysis and T test/non-parametric test were used to analyze the differences between the two groups of samples. Pearson’s correlation coefficient and Fisher’s exact test were used for correlation analysis of DEPs and DEMs, as well as correlation analysis of proteomics and metabolomics pathway enrichment, respectively. All statistical tests were two-sided, and P < 0.05 was considered to be statistically significant.

A total of 1121 patients were included in the study, comprising 988 males and 133 females, with an average age of 51 ± 10 years old. Detailed clinical baseline data for each group were displayed in Table 1 . Statistical analysis revealed significant differences in gender, age, BCLC stage, liver cirrhosis, MVI, AFP, and HBsAg between the HCC_Cs group and C. sinensis negative HCC group (P < 0.05, Table 1 ). The prevalence of males was significantly higher in the HCC_Cs group compared to the HCC group (97.5% vs. 87.0%, P=0.001). The BCLC stage was more advanced in the HCC_Cs group than the HCC group (BCLC_B~C 56.8% vs. 40.4%, P=0.001). Moreover, the HCC_Cs group exhibited higher rates of cirrhosis (66.1% vs. 53.8%) and MVI (58.5% vs. 46.3%) (all P < 0.05). The AFP (P=0.004) and HBsAg (P=0.018) values of patients in the HCC_Cs group were significantly higher and lower than those in the HCC group, respectively. However, no statistically significant differences were observed in tumor size, number of tumors, Edmondson grade, CEA, and HCV ( Table 1 ).

A total of 813 patients with HCC and 92 patients with HCC_Cs were included in the study for the analysis of OS and RFS. The median follow-up time for OS was 77 months (range 1-118 months), while the median follow-up time for RFS was 16 months (range 1-118 months). At the last follow-up, there were 509 (62.6%) and 43 (46.7%) patients alive in the HCC group and HCC_Cs group, respectively. Moreover, among the patients who underwent RFS follow-up, 177 patients (21.7%) in the HCC group and 11 patients (11.9%) in the HCC_Cs group did not experience recurrence. The HCC_Cs patients had worse OS compared to the HCC patients, with three-year survival rates of 51.2% vs. 64.9% and five-year survival rates of 40.8% vs. 55.9% (P=0.016, Figure 2A ). Similarly, compared with HCC patients, patients with HCC_Cs had lower RFS rates (three years: 34.2% vs. 22.5%; five years: 18.7% vs. 10.7%, P=0.003, Figure 2B ).

Serum biochemical analysis revealed that the levels of LPS, APO-A, HDL-C, A1/B1, and TRF were significantly lower in the HCC_Cs group compared to the HCC group (P < 0.05, Figures 3C-E, G, H ). Conversely, the levels of triglyceride (TG) were significantly higher (P < 0.05, Figure 3F ). Although there was no significant difference in the levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and blood ammonia (AMM) between the two groups, the levels of ALT and AMM were increased in the HCC_Cs group (P > 0.05, Figures 3A, B, I ). Furthermore, the serum samples (12 cases in each group) subsequently used for proteomic and metabolomic analyses also exhibited similar trends in these indicators ( Supplementary Figure S1 ).

To further clarify the DEPs between HCC and HCC_Cs patient serum, totally 16 serum samples of patients (8 cases in each group) were conducted proteomic analyses. The Venn diagram revealed that the HCC group and the HCC_Cs group together shared 588 proteins, and individually contained 36 and 59 proteins, respectively ( Figure 4A ). Additionally, clustering analysis results showed that samples within group have high correlation and good sample repeatability ( Figure 4B ). The volcano plot displayed 63 up-regulated DEPs and 43 down-regulated DEPs (FC ≥ 2 or FC ≤ 0.5, P < 0.05), and representative up-/down-regulated DEPs were marked in pink ovals and blue ovals, respectively ( Figure 4C ).

The main enriched GO terms included proteasome core complex, intracellular protein-containing complex, proteasome core complex, beta-subunit complex, nucleoplasm, phagocytosis, cellular nitrogen compound biosynthetic process, proteasomal protein catabolic process, proteasome complex, catalytic activity, acting on a nucleic acid, and regulation of inflammatory response to antigenic stimulus. The primarily KEGG pathways involved were proteasome, spinocerebellar ataxia (SCA), Alzheimer disease (AD), Parkinson disease (PD), pathways of neurodegeneration - multiple diseases, amyotrophic lateral sclerosis (ALS), arginine biosynthesis, type I diabetes mellitus, Huntington disease (HD), and Th17 cell differentiation. The DEPs such as PSMB3, PSMB4, PSMB5, PSMA1, PSMA3, ILF2, HSP90 β, PARK7, OLFM4, ASS, and CNN2 were primarily participated in the aforementioned GO terms and KEGG pathways ( Figures 4D, E ).

The results of the expression pattern cluster analysis confirmed that the expression trends of the DEPs in the samples within groups were relatively similar ( Figure 5A ). 20 important DEPs were further selected, clustered and analyzed their expression, including up-regulated CNN2, CDH6, ILF2, CRN, OLFM4, and DBNL, as well as down-regulated TRF, PARK7, HSP90β, PSMB3, IGFBP1, and ASS, etc. ( Figure 5B ). To explore the interaction among DEPs, the PPI network of DEPs was constructed using the STRING database. The proteins at the center of the network were identified as L1CAM, CLSTN1, CSPG4, and AGT ( Figure 5C ).

To investigate the differences in metabolites between the HCC group and HCC_Cs group, untargeted LC-MS/MS was performed on the serum samples from both groups. The reliability of the experimental data was confirmed by the RSD results of the quality control (QC) samples ( Figures 6A, B ). The OPLS-DA model showed small intra-group sample differences and large inter-group sample differences ( Figures 6C, D ). The metabolites detected under the negative and positive ion modes were presented in Figures 6E, F , respectively. A total of 100 DEMs were detected, with 37 up-regulated and 63 down-regulated. Representative up- and down- regulated DEMs were marked with yellow rectangles and green rectangles, respectively ( Figure 7A ). The classification diagram of DEMs was shown in Figure 7B , and the cluster analysis in Figure 7C illustrated the expression changes of the DEMs. Variable importance in projection (VIP) analysis identified the top 35 DEMs that significantly contributed to grouping (VIP ≥ 1, P < 0.05). The first 5 metabolites were spermidine, LysoPA(16:0/0:0), 12(R)-HETE, L-4-Hydroxyglutamate semialdehyde, and taurodeoxycholic acid (TDCA) ( Figure 7D ).

KEGG enrichment analysis confirmed that C. sinensis infection significantly affected pathways including glycerophospholipid metabolism, carbohydrate digestion and absorption, mineral absorption, AGE-RAGE signaling pathway in diabetic complications, diabetic cardiomyopathy, parathyroid hormone synthesis, secretion and action, galactose metabolism, insulin secretion, glycine, serine and threonine metabolism, and HIF-1 signaling pathway. The DEMs involved in these KEGG enrichment pathways were L-serine, PE(20:3(8Z,11Z,14Z)/P-16:0), DG(15:0/18:3(6Z,9Z,12Z)/0:0), D-galactose, D-glucose, and aminoacetone ( Figure 7E ).

The O2PLS method was utilized to evaluate the intrinsic correlation between the two datasets. These findings showed a high degree of overlap between the two groups of substances, indicating a strong similarity between them ( Figure 8A ). Functional enrichment analysis and Venn analysis identified a total of 75 pathways co-enriched by proteomics and metabolomics ( Figure 8B ). Moreover, Figure 8C displayed 20 crucial co-enriched pathways, encompassing immune and inflammation-related pathways such as Th17 cell differentiation, Th1 and Th2 cell differentiation, and NF-κB signaling pathway, cancer-related pathways like HCC and Ras signaling pathway, neurological diseases related pathways including SCA, PD, and pathways of neurodegeneration - multiple diseases, and diabetes and cardiovascular disease-related pathways such as insulin resistance (IR), type II diabetes mellitus, and diabetic cardiomyopathy. Metabolic pathways closely associated with HCC progression, such as galactose metabolism, carbohydrate digestion and absorption, sphingolipid metabolism, and arginine and proline metabolism, were also co-enriched ( Figure 8C ). The correlation analysis results of DEPs and DEMs unveiled that proteins such as CRN, TRF, MHC I, and IgD formed a complex mutual regulatory network with metabolites such as LysoPA(16:0/0:0), LysoPA(0:0/16:0), 12(R)-HETE, D-glucose, D-galactose, and L-serine ( Figure 8D ). The detailed data of the network list of DEPs and DEMs can be found in Supplementary Table S1 .