Quantitative ctDNA Detection in Hepatoblastoma: Implications for Precision Medicine

Abstract

Hepatoblastoma is characterized by driver mutations in CTNNB1, making it an attractive biomarker for a liquid biopsy approach utilizing circulating tumor DNA (ctDNA). This prospective observational study sought to ascertain the feasibility of ctDNA detection in patients with hepatoblastoma and explore its associations with established clinical indicators and biomarkers, including serum Alpha-fetoprotein (AFP). We obtained 38 plasma samples and 17 tumor samples from 20 patients with hepatoblastoma. These samples were collected at various stages: 10 at initial diagnosis, 17 during neoadjuvant chemotherapy, 6 post-operatively, and 5 at disease recurrence. Utilizing a bespoke sequencing assay we developed called QUENCH, we identified single nucleotide variants and deletions in CTNNB1 ctDNA. Our study demonstrated the capability to quantitate ctDNA down to a variant allele frequency of 0.3%, achieving a sensitivity of 90% for patients at initial diagnosis, and a specificity of 100% at the patient level. Notably, ctDNA positivity correlated with tumor burden, and ctDNA levels exhibited associations with macroscopic residual disease and treatment response. Our findings provide evidence for the utility of quantitative ctDNA detection in hepatoblastoma management. Given the distinct detection targets, ctDNA and AFP-based stratification and monitoring approaches could synergize to enhance clinical decision-making. Further research is needed to elucidate the interplay between ctDNA and AFP and determine the optimal clinical applications for both methods in risk stratification and residual disease detection.

Keywords: AFP; NGS; ctDNA; ddPCR; hepatoblastoma; liquid biopsy.

Figure 1

CtDNA positivity correlates with tumor burden (a) and levels correlate with clinical time point (b). (a) Each row represents a different patient. Colored rectangles represent tested cfDNA samples: grey—ctDNA-negative, indigo—ctDNA-positive in localized cases, purple—ctDNA-positive in metastatic cases. (b) Scatter plot of ctDNA-positive samples at different VAF ranges, color coded according to time point of collection. Dx—initial diagnosis—cyan; NACT—during neoadjuvant chemotherapy (the number of previously administered cycles shows in brackets)—green (post 1 cycle)/orange (post 2 or more cycles); Rec—disease recurrence—purple. CtDNA VAF (for the scatter plot) and % of samples in each VAF range presented in bar chart.

Figure 2

QUENCH verification—concordance with ddPCR. Matched ddPCR and NGS results from 8 patients (color-coded according to each patient). (a) High concordance (R² = 0.9606) observed in samples with QUENCH VAF > 0.3%. ddPCR shows higher sensitivity for VAF < 2.2%. (b) Sensitivity, specificity, and concordance of variant detection evaluated by QUENCH and ddPCR.

Figure 3

CtDNA moderately correlates with AFP. (a) VAF of QUENCH-positive samples and AFP are moderately concordant (R² = 0.6713). Grey—ctDNA-negative, indigo—ctDNA-positive, salmon—ctDNA-positive outlier not included in the linear regression analysis. (b) Sensitivity, specificity, and concordance of variant detection evaluated by QUENCH and AFP and ddPCR and AFP.

Figure 4

AFP levels correlate with tumor size at diagnosis (a), but only smaller tumors show a correlation between ctDNA VAF and tumor size (b). Matched tumor volume, AFP, and NGS results at diagnosis from 10 patients. (a) AFP levels and tumor size are concordant (R² = 0.7923). Indigo—samples included in the analysis, salmon—sample not included in the analysis. (b) CtDNA VAF levels and small tumor size are concordant (R² = 0.8509). Indigo and salmon—samples included in the analysis, grey—samples excluded from the linear regression analysis.

Figure 5

CtDNA correlates with dynamic treatment response. Each graph represents an individual patient. Serial samples with matched ctDNA VAF (evaluated by QUENCH—indigo squares or ddPCR—grey triangles, left Y axis) and AFP levels (salmon, right Y axis), collected at different clinical time points: Dx—initial diagnosis, NACT—during neoadjuvant chemotherapy, post-op—post resection of primary tumor, Rec—disease recurrence. Age-adjusted normal values of AFP presented with a dashed line.

Figure 6

CfDNA levels in hepatoblastoma correlate with disease stage (a) and clinical time point (b). N—control group, HB—hepatoblastoma, Loc—localized disease, Met—metastatic disease, Dx—initial diagnosis, NACT—during neoadjuvant chemotherapy, Rec—disease recurrent disease, post-op—post resection of primary tumor. p-values are presented in asterisks (* = 0.0425, ** = 0.0098, *** = 0.0005, **** < 0.0001, ns—not significant).

Figure 7

Correlation between cfDNA amount in plasma and ctDNA positivity. CfDNA levels (ng/mL) and cfDNA samples’ positivity (Pos) or negativity (Neg) for CTNNB1 mutation. p-value is presented in asterisks (* = 0.0441).