Abstract

Despite recent developments, it is still very difficult to differentiate hepatocellular carcinoma (HCC) from intrahepatic cholangiocarcinoma (iCCA). Clinically available methods such as CT, MRI, and contrast-enhanced ultrasound require the highest level of investigator experience to reliably differentiate between HCC and iCCA. Here, we used the surface analysis interferometric and fluorescence imaging methodology for small extracellular vesicles (small EVs) analysis in liver cancer category (LI-RADS) M (LR-M) patients. We evaluated their clinical performance and proposed a 5-point system using serological markers to distinguish HCC from iCCA.

Small EVs from 24 patients’ sera were isolated by size exclusion chromatography (SEC). Particle counts were measured by nanoparticle tracking analysis (NTA). Small EVs were captured on a microarray chip coated with CD9, CD63 and CD81 and surface stained with fluorescent labelled anti-CD133/2 antibody. Surface analysis of small EVs on a single particle level was performed with ExoView^® R100 (NanoView Biosciences, Boston, USA). Data analysis was performed using ExoView^® Analyzer 3.2. Additionally, thrombocytes, ALP, CRP were assessed.

1) Whole sera particle concentrations isolated from iCCA and HCC differ significantly (p=0.0364) with a mean particle concentration/mL ± SD of 1.8x10¹⁵ ± 1.4x10¹⁵in iCCA and 8.2x10¹⁴ ± 5.3x10¹⁴ in HCC (AUROC 0.74, sensitivity and specificity 75% and 58.3%, respectively. 2) Individual small EV subpopulations as characterized by CD9⁺CD133/2⁺ and CD81⁺CD133/2⁺ were significantly elevated in iCCA (CD9⁺C133/2⁺: p=0.0315, AUROC 0.74 and CD81⁺CD133/2⁺: p=0.0244, AUROC 0.826). 3) Thrombocytes, ALP and CRP were elevated in iCCA compared to HCC. (p=0.0238 and AUROC 0.76; p=0.0023 and AUROC 0.85; p=0.0092 and AUROC 0.88, respectively). 4) On an individual basis, each biomarker such as particles, small EVs, thrombocytes, ALP, and CRP did not achieve sensitivity or specificity greater than 80%. However, when each biomarker was converted into a binary categorical variable and a final score was calculated by assigning a point for each factor to the biomarker's cut-off value below it, samples could be scored from 0 to 5. Samples with a score of 3 to 5 were categorized as HCC and 0 to 2 points as iCCA, resulting in a diagnostic performance with a sensitivity of 100% and a specificity of 83.3%.