This case-control study was conducted from September 2018 to February 2020 in the Department of Biochemistry, in collaboration with the Department of Surgical Gastroenterology at the Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Puducherry, India. Ethical approval was obtained from the Institute Ethics Committee (JIP/IEC/2018/0282), and the study adhered to the Indian Council of Medical Research (ICMR) biomedical guidelines and the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment.

Consecutive patients were recruited. Both male and female participants aged 20–60 years were enrolled and divided into two groups: healthy controls (n = 40) and HCC patients (n = 40). Controls were selected based on normal liver function tests and the absence of any recent infections or tumors. HCC patients meeting the inclusion criteria in line with the European Association for the Study of the Liver (EASL) criteria. Moreover, critically ill patients and those with rare variants like fibrolamellar carcinoma were excluded, as this subtype can complicate HCC diagnosis. Non-invasive imaging techniques were used to distinguish fibrolamellar carcinoma from classical HCC. HCC diagnosis was based on ultrasonography and confirmed through CT or MRI, supported by elevated alpha-fetoprotein (AFP) levels or histopathological analysis. Tumor characteristics, including lesion count, vascular invasion, and metastasis, were assessed via CT or MRI and reviewed by an independent radiologist at JIPMER. Confirmation of HCC was done by biopsy or ultrasound-guided fine needle aspiration cytology (FNAC), interpreted by a blinded pathologist. Disease severity was evaluated using the Child-Pugh score, and staging was conducted according to the Barcelona Clinic Liver Cancer (BCLC) system.

All patients’ demographic and clinical data were collected at admission. Furthermore, approximately 5 mL of blood was collected from both control subjects and HCC patients into clot activator tubes (BD Vacutainer, REF 369032) and allowed to stand for one hour. The samples were then centrifuged at 1,000 × g for 10 min at 4°C to separate the serum. Routine biochemical analyses were performed immediately using the fresh serum, while the remaining samples were aliquoted and cryopreserved at –80°C for further analysis.

Liver tumor tissues and adjacent non-tumorous tissues (histologically confirmed as controls) were obtained from HCC patients undergoing hepatectomy as part of their treatment protocol. Each resected tissue sample was divided: one portion was fixed in 10% neutral buffered formalin (NBF) for histological examination, and the other was snap-frozen in liquid nitrogen and stored at –80°C for molecular studies.

Liver function markers, including bilirubin, total protein, albumin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and gamma-glutamyltransferase (γGT), were measured using commercially available kits (Beckman Coulter Ireland, Inc.) on a Beckman Coulter AU5800 autoanalyzer (USA), following the manufacturer’s instructions. Prothrombin time (PT) and international normalized ratio (INR) were assessed using a chemiluminescence immunoassay system (Siemens ADVIA Centaur CP). Serum AFP levels were quantified by chemiluminescence using the UniCel DxI 600 Access Immunoassay System (Beckman Coulter, USA; 33210).

Serum PXR and IL-1β concentrations were estimated using sandwich ELISA kits from FineTest, Wuhan Fine Biotech Co., Ltd, Wuhan, China (PXR: Cat. No. EH1200; IL-1β: Cat. No. EH0185), and serum TNF-α was assessed using the human TNF-α ELISA kit (Cat. No. KB1145; Krishgen Biosystems, USA) as per the manufacturer’s instructions. Briefly, standards and serum samples were added to microplate wells pre-coated with target-specific antibodies. After incubation and washing to remove unbound substances, a biotinylated detection antibody and streptavidin-HRP conjugate were added sequentially. The resulting colorimetric reaction, following substrate addition, was measured at the appropriate wavelength using a microplate reader (Spectra max Plus 384; Serial No: AMNR06890; Thermo Fisher Scientific Inc.). Concentrations of PXR, TNF-α, and IL-1β were calculated by comparing absorbance values to a standard curve generated for each target analyte.

ROC curve analysis was conducted using GraphPad Prism version 10 (San Diego, CA, USA) and MedCalc (version 23.3.7; MedCalc Software Ltd) to evaluate the diagnostic performance of AFP and PXR in HCC patients. The cut-offs for PXR, AFP were chosen based on ROC analysis and Youden’s index, providing strong sensitivity and specificity for distinguishing HCC from non-HCC individuals.

Total protein was extracted from tumorous and non-tumorous liver tissues of HCC patients using ice-cold RIPA buffer (Cat. # 786-489; G-Biosciences, USA) containing a protease inhibitor cocktail and PMSF (Cat. # P8340 and 11359061001, respectively; Sigma-Aldrich, USA). Protein concentrations were determined using the Pierce Coomassie Plus (Bradford) protein assay kit (Cat. # 23236; Thermo Fisher Scientific, USA). Protein samples (~30 μg) were separated by SDS-PAGE on 7.5% and 10% polyacrylamide gels and transferred to nitrocellulose membranes (S045A330R; Advantec, Japan). Membranes were blocked and incubated with rabbit polyclonal anti-PXR (Cat. # A1583; ABclonal, USA) and rabbit monoclonal phosphorylated anti-NFκB-p65 (Cat. # 3033 1:1,000; Cell Signaling, UK), and mouse monoclonal anti-β-actin (Cat. # AC004; ABclonal, USA) antibodies at recommended dilutions. After overnight incubation, membranes were probed with HRP-conjugated goat anti-rabbit or goat anti-mouse secondary antibodies (Cat. # 405306; BioLegend, USA). Protein bands were visualized using an enhanced chemiluminescent substrate (Cat. # 32209; Thermo Fisher Scientific, USA) and imaged on a ChemiDoc XRS+ system (Cat. # 1708265; Bio-Rad, USA). Densitometry analysis was performed using Image Lab software (V 6.1.0; Bio-Rad, USA), and target bands were normalized to β-actin.

Paraffin blocks of 10% NBF-fixed tumorous and non-tumorous liver tissues were prepared. Serial 5 μm sections were obtained using an automated microtome (Cat. # 149AUTO00C1; Leica Biosystems) and stained with hematoxylin and eosin (H&E). Histological characteristics of the HCC tumor and non-tumorous tissues (control liver) were assessed by an independent, blinded pathologist from the Department of Pathology, JIPMER. For immunohistochemistry (IHC), tissue sections were placed on silane-coated slides, deparaffinized in xylene, and rehydrated. Endogenous peroxidase activity was blocked with 3% H₂O₂ (Cat. # 88597; Sigma-Aldrich, USA) and rinsed with phosphate-buffered saline (PBS). Antigen retrieval was performed by incubating sections in citrate buffer (pH 6.0) (Cat. # ALF-J63950-AP; Thermo Fisher Scientific, USA) at 110°C for 10 min. The sections were then blocked with 2.5% normal horse serum for 20 min and incubated overnight with rabbit polyclonal anti-PXR (1:100, Cat. # A1583; ABclonal, USA) at 4°C. After washing with PBS, sections were incubated with ImmPRESS Universal Polymer Reagent (MP-7500; Vector Labs, USA) for 30 min at room temperature. PXR localization was detected using the ImmPACT DAB kit (SK4105; Vector Labs, USA), and sections were counterstained with hematoxylin. Negative controls were included using a no-primary antibody condition. Finally, sections were dehydrated, mounted with Histomount (Cat. # 008030; Thermo Fisher Scientific, USA), and photographed using an EVOS FLc imaging system (Cat. # AME3300; Life Technologies, USA).

Data were analyzed using GraphPad Prism version 10 (San Diego, CA, USA) and MedCalc (version 23.3.7; MedCalc Software Ltd). Qualitative data are presented as counts and percentages, while quantitative data are expressed as mean ± standard error (SE). A two-tailed unpaired t-test or Mann-Whitney U test was used to compare cases and controls. Spearman’s correlation analysis assessed associations between parameters. ROC curve analysis for AFP and PXR was performed to evaluate diagnostic accuracy in HCC participants. P values < 0.05 were considered statistically significant.

Table 1 shows the clinical and tumor radiological characteristics of HCC patients. The Child-Pugh score was used to assess the disease severity and was distributed as Child class A (60%), Child class B (32.5%), and Child class C (7.5%). BCLC staging was also done for HCC participants and accordingly found to be in stages A (12.5%), B (27.5%), C (55%), and D (5%). Among the HCC participants, 65% had one lesion, 20% had two lesions, 7.5% had three, and 7.5% had multifocal lesions. Hepatic and portal vein invasions were observed in 27.5% and 37.5% of HCC participants, respectively. 22.5% were hepatitis B surface antigen (HBsAg) positive and 17.5% were HCV positive. Cirrhosis was present in 62.5% of HCC participants.

Serum levels of liver function parameters are given in Table 2. Levels of total and direct bilirubin, AST, ALT, γGT, ALP, PT, and INR increased significantly, while serum albumin level was decreased in HCC cases compared to controls. Furthermore, higher AFP levels (P < 0.0001) were observed in HCC cases compared to healthy controls.

PXR, TNF-α, and IL-1β serum concentrations were increased (P < 0.0001) in HCC cases compared to control subjects (Figure 1A–C). Furthermore, serum levels of PXR showed a positive correlation with TNF-α and IL-1β (r = 0.456, P < 0.001; r = 0.476, P < 0.0001 respectively; Figure 1D and E). Moreover, a positive correlation was noted between PXR and elevated γGT levels (r = 0.490, P < 0.0001; Figure 1F and G). Notably, subgroup analysis among HCC cases revealed that PXR concentration was higher (P < 0.05) in alcoholic HCC as compared to non-alcohol-related HCC (Figure 1H). We also observed PXR expression according to BCLC staging in HCC patients (Figure 1I). PXR expression was significantly higher in stage B and stage C compared to stage A (P = 0.009 and P = 0.001, respectively). Although PXR levels were elevated in stage C compared to stage B, this difference was not statistically significant.

Figure 2 illustrates the ROC curve analysis of AFP and PXR for evaluating their diagnostic performance in HCC participants. AFP at a cut-off value > 18.9 ng/mL achieved an AUC of 0.883, with 83% sensitivity and 88% specificity, yielding a positive likelihood ratio (LR⁺) of 6.6 and a negative likelihood ratio (LR^–) of 0.2 (Figure 2A). PXR at a cut-off value > 2.014 ng/mL demonstrated an AUC of 0.902, with 83% sensitivity and 90% specificity, corresponding to an LR⁺ of 8.3 and an LR^– of 0.19 (Figure 2B). Figure 2C shows a pairwise comparison of ROC curves between PXR and AFP (ng/mL) using DeLong’s test [14]. The comparison between PXR and AFP ROC curves shows a small difference in AUCs (0.0191), with a non-significant P-value of 0.7616. The 95% confidence interval for the difference includes zero (–0.104 to 0.142), and the z statistic is 0.303, indicating no statistically significant difference in diagnostic performance between PXR and AFP.

Figure 3A shows the H&E-stained liver section of non-tumorous (control liver) and tumorous tissue obtained from HCC participants. The control liver showed typical hepatic lobular architecture characterized by densely stained hepatocyte cytoplasm with prominent granularity, round nuclei, and nucleoli arranged in regular cords or plates separated by sinusoids. The portal zone included connective tissue, hepatic artery, portal vein, and bile duct. The HCC liver tissue showed classical features of HCC characterized by the presence of malignant cells, thick trabecular cords, well-differentiated and interdigitated with scant basophilic cytoplasm, nuclear overcrowding, and hyperchromatic nuclei.

Figure 3B shows immunostaining of PXR on sections of the control liver and HCC liver. Control liver tissue showed moderate cytoplasmic and focal nuclear PXR positivity, while HCC liver tissue demonstrated strong cytoplasmic and focal nuclear PXR positivity. Figure 3C shows hepatic PXR protein expression by western blot in the control liver and HCC tumor tissues. PXR protein expression was increased (P < 0.05) in HCC tumor tissue compared to control liver tissue. Similarly, hepatic phosphorylated NFκB expression was significantly increased in the HCC liver compared to the control liver (Figure 3D).