Advancing precision oncology in intrahepatic cholangiocarcinoma

ICCA accounts for 10–15% of primary liver cancers and is notorious for late-stage diagnosis and resistance to chemotherapy and radiotherapy. While surgical resection offers the best survival outcomes, most patients present with advanced disease, highlighting the urgency for early detection and personalized therapies. Emerging research reveals that ICCA's heterogeneity-manifested in histological diversity, genetic mutations, and molecular subtypes-dictates clinical behavior and therapeutic response. Integrating imaging with molecular profiling offers a non-invasive approach to stratify patients and guide treatment, bridging the gap between bench and bedside.

Histological subtypes and imaging correlates

ICCA is classified into two main histological subtypes based on anatomical and morphological criteria:

1. Large Duct Type (LD-ICCA):

   • Originates near the perihilar region; exhibits columnar cells, mucin production, and markers like S100P and MUC5AC.

   • Aggressive: higher rates of lymph node metastasis (LNM), vascular invasion, and recurrence.

   • Imaging: Hypoenhancement on arterial-phase CT/MRI, infiltrative margins, and bile duct dilation (Figs. 3–4).

2. Small Duct Type (SD-ICCA):

   • Peripheral location; cuboidal cells, fibrous stroma, and markers like SPP1 and CD56.

   • Better prognosis: lower metastatic potential.

   • Imaging: Arterial hyperenhancement, round/lobulated contours, and absence of bile duct encasement (Fig. 2).

Key imaging studies:

• CT: Arterial hyperenhancement distinguishes SD-ICCA (specificity: 90.9%) (Nam et al.).

• MRI: Hypoenhancement and bile duct dilation predict LD-ICCA (AUC: 0.91) (Xiao et al.).

• PET/CT: Higher SUV_max in LD-ICCA correlates with poorer survival (Kozaka et al.).

Genetic alterations and radiogenomics

ICCA harbors distinct mutational profiles:

• LD-ICCA: KRAS, TP53, and CDKN2A mutations (poor prognosis).

• SD-ICCA: IDH1/2 mutations and FGFR2 fusions (targetable).

Imaging Predictors of Mutations:

• IDH1/2: Higher CT enhancement ratios (AUC: 0.80) (Zhu et al.).

• KRAS: Elevated metabolic tumor volume on PET/CT (Ikeno et al.).

• LRRK2: CT texture features predict gene expression (AUC: 0.75) (Gu et al.).

Molecular subtypes and therapeutic implications

Multi-omics studies reveal four molecular classifications with clinical relevance:

1. Inflammation Class: Activated STAT3 pathway; potential response to JAK-STAT inhibitors.

2. Proliferation Class: RAS/MAPK activation; sensitive to sorafenib.

3. Immune-Activated Subtype (IG3): High CD8+ T-cell infiltration; benefits from immunotherapy.

4. Mesenchymal Subtype: Fibroblast-rich, worst prognosis.

Imaging Biomarkers:

• MRI texture features predict immune subtypes (AUC: 0.92 for CD8+ density) (Zhang et al.).

• Gadoxetic acid uptake correlates with metabolic subtypes (Jeon et al.).

Challenges and future directions

1. Clinical Translation: Most radiogenomic studies are retrospective; multi-center validation is needed.

2. HBV Paradox: TP53 mutations worsen prognosis in HBV+ patients, yet HBV+ ICCA has better outcomes overall.

3. Therapeutic Gaps: IDH inhibitors (e.g., ivosidenib) show promise, but resistance mechanisms remain unclear.

4. Integrated Approaches: Combining radiomics, spatial transcriptomics, and AI may refine subtype prediction.

Conclusion

ICCA's heterogeneity demands a multifaceted approach. Imaging signatures-when aligned with histological and molecular data-can non-invasively predict subtypes, mutations, and therapeutic vulnerabilities. While challenges persist in standardization and clinical adoption, the synergy of radiomics and molecular profiling heralds a new era of precision oncology for ICCA. Future research must prioritize prospective trials to validate these tools and translate them into actionable clinical strategies.