Actionable Molecular Alterations in Intrahepatic Versus Extrahepatic Cholangiocarcinoma: Frequency Disparities and Assay-Platform Implications

Epidemiologic and Anatomic Framing

Cholangiocarcinoma (CCA) is a biologically heterogeneous malignancy arising along the biliary tree that is broadly stratified into intrahepatic cholangiocarcinoma (iCCA), originating proximal to second-order intrahepatic ducts, and extrahepatic cholangiocarcinoma (eCCA), which encompasses perihilar tumors at the hepatic duct confluence and distal tumors of the common bile duct. This anatomical dichotomy is not merely topographic; it reflects divergent cells of origin, etiologic exposures, and, critically, distinct actionable genomic landscapes 223. Within iCCA itself, a further histologic distinction between small-duct and large-duct subtypes carries molecular significance: small-duct iCCA disproportionately harbors FGFR2 fusions and IDH1 mutations, whereas large-duct iCCA more closely resembles eCCA in its mutational profile, including higher rates of KRAS, TP53, SMAD4, and HER2/ERBB2 alterations 2324.

Adding complexity to prevalence estimates, historical ICD coding errors led to systematic misclassification of perihilar tumors as intrahepatic; ICD-11 now introduces discrete codes for iCCA, perihilar CCA, and distal CCA, but retrospective cohort comparisons must be interpreted cautiously 14. Geographic and etiologic factors—most notably liver fluke (Opisthorchis viverrini) infection prevalent in Southeast Asia—further shape mutational landscapes. Fluke-associated iCCA harbors FGFR2 fusions at a dramatically lower rate (approximately 0.8%) compared with non-fluke-associated iCCA (approximately 11.6%), a difference with direct implications for pretest probability when applying FGFR-targeted testing algorithms in endemic regions 312.

Alteration-Specific Frequency Disparities by Anatomic Subtype

FGFR2 fusions/rearrangements. FGFR2 fusions represent the most robustly established iCCA-specific therapeutic target. In large-scale genomic profiling cohorts, FGFR2 fusions are detected in approximately 10–20% of small-duct iCCA but are substantially rarer in eCCA, where frequencies fall to low single-digit percentages 1524. The fusion landscape is architecturally complex, with over 150 distinct fusion partners identified (e.g., BICC1, TRIM8, ATE1, DBP), all of which are mutually exclusive with IDH1 mutations and other dominant drivers, suggesting these alterations define genuinely distinct oncogenic pathways 37. In contrast, FGFR2 fusions in perihilar and distal eCCA subtypes are comparatively uncommon, underscoring that FGFR-targeted therapeutic strategies are primarily an iCCA endeavor.

IDH1 mutations. IDH1 mutations are the second major iCCA-enriched alteration, occurring in approximately 10–20% of iCCA in most contemporary series and up to ~29.5% in certain Chinese iCCA cohorts, while remaining uncommon in eCCA 415. IDH1 and FGFR2 alterations are mutually exclusive, and both are enriched in the small-duct iCCA subtype, whereas IDH2 mutations are less frequent. The consistent iCCA enrichment signal across ethnically and geographically diverse cohorts supports routine IDH1 testing as a standard component of iCCA workup.

HER2/ERBB2 amplification and overexpression. HER2 alterations follow a distinct distribution, with enrichment in eCCA and gallbladder carcinoma relative to iCCA. Across BTC cohorts broadly, HER2 amplification or overexpression is observed in approximately 10–27% of eCCA/gallbladder-adjacent cancers, compared with approximately 5% amplification frequency in large-duct iCCA and substantially lower rates in small-duct iCCA 1526. Critical diagnostic nuances exist: in intrahepatic cholangiocarcinoma specifically, FISH-based HER2 amplification frequencies differ depending on the threshold applied (e.g., 4.28%, 5.58%, or 9.87% using three validated FISH criteria), and intratumoral heterogeneity is a recognized confounder when interpreting single-core biopsies 26. HER2-amplified ICC also demonstrates higher tumor mutational burden (TMB) and a distinct T-cell–exhausted immune microenvironment, findings that may have implications for combination immunotherapy trial design 26.

BRAF V600E. BRAF V600E mutations occur at low frequency across BTC subtypes, with most series reporting approximately 3–6% prevalence without strong enrichment for either iCCA or eCCA 1527. Importantly, BRAF alterations show mutual exclusivity with FGFR2 fusions and IDH1 mutations in iCCA, defining yet another discrete molecular subset 27. Despite their rarity, BRAF V600E mutations are clinically actionable, making their identification through broad genomic profiling essential rather than optional.

Clinical Actionability and Treatment Relevance

The iCCA enrichment of FGFR2 fusions directly informs second-line treatment selection. Pemigatinib and futibatinib—selective FGFR inhibitors—have demonstrated objective response rates of approximately 37% and 42%, respectively, with median progression-free survival of approximately 6.9 and 9.0 months in FGFR2 fusion–positive iCCA 15. Both agents are endorsed by NCCN 2025 and ESMO guidelines as targeted options in previously treated FGFR2-altered disease 128. Acquired resistance via secondary FGFR2 kinase-domain mutations is well-documented, making dynamic re-profiling at progression—whether via re-biopsy or high-quality ctDNA—an important clinical consideration 715.

For IDH1-mutant iCCA, ivosidenib demonstrated a statistically significant improvement in progression-free survival versus placebo in the randomized phase III ClarIDHy trial, establishing IDH1 inhibition as a standard targeted option in this biomarker-selected population 15. IDH1 testing is endorsed by major guidelines, and the preponderance of IDH1 mutations in iCCA makes routine testing in this subtype particularly impactful.

HER2-directed therapy in BTC has undergone rapid evolution. Zanidatamab, a bispecific HER2 antibody, achieved a confirmed objective response rate of approximately 41%—rising to approximately 52% in IHC 3+ patients—in the phase IIb HERIZON-BTC-01 trial, leading to FDA accelerated approval in 2024 and subsequent EU conditional marketing authorization in 2025 for previously treated, unresectable/metastatic HER2-positive BTC 1531. Trastuzumab deruxtecan (T-DXd) and the tucatinib plus trastuzumab combination have also demonstrated activity in HER2-positive BTC, while the dual-blockade regimen of trastuzumab plus pertuzumab produced an objective response rate of approximately 23% in the MyPathway basket trial 15. Clinical benefit in BTC is currently linked to unequivocal HER2 positivity—IHC 3+ and/or confirmed gene amplification—as HER2-low status has not yielded consistent benefit 1528.

For BRAF V600E–mutant BTC, dual BRAF/MEK inhibition (e.g., dabrafenib plus trametinib) is recognized by guidelines across regions as a targeted option despite the relatively limited, tumor-agnostic and BTC-specific data supporting this approach 628. Broad genomic profiling remains the only reliable mechanism to identify these patients, and co-occurring mutations may influence downstream signaling and resistance context 27.

Assay-Platform Implications

The molecular heterogeneity of CCA demands a multi-platform approach to biomarker assessment, with method selection tightly coupled to the biology of each target.

FGFR2 fusions present the greatest detection challenge. Amplicon-based DNA panels and narrow hybrid-capture designs are prone to false negatives because they may not cover atypical or novel fusion partners, and intronic breakpoints can escape detection 24. External proficiency testing demonstrated that RNA-based, fusion-agnostic NGS approaches outperform amplicon-limited panels, with NGS-only cohorts achieving an 81% pass rate versus 60% for NGS-plus-FISH groups, partly reflecting platform design differences rather than FISH inadequacy per se 24. Comprehensive RNA-based fusion detection or well-designed hybrid-capture DNA+RNA panels are therefore preferred for iCCA. FFPE-related RNA degradation and low tumor cellularity represent key preanalytic limitations; reflex RNA testing should be triggered when DNA-only results are negative but clinical suspicion remains high 8928. Break-apart FISH for FGFR2 retains a role as orthogonal confirmation when equivocal results arise.

IDH1 mutations are readily captured by standard hotspot DNA NGS targeting canonical variants such as R132; IHC for IDH1 R132H detects only a subset of mutations and is insufficient as a standalone test in CCA 32. Low tumor purity on small biopsy specimens remains the primary source of false negatives, reinforcing the need for adequate tissue procurement and tumor content verification.

HER2 assessment requires coordinated multi-modal evaluation. IHC serves as the frontline screen for protein overexpression, with reflex ISH/FISH for IHC 2+ cases and NGS copy-number calls as supportive evidence 152628. Notably, IHC-FISH concordance in ICC is imperfect, and intratumoral heterogeneity across cores can produce discordant results 26. NGS-based copy-number calling thresholds vary by panel design and bioinformatic pipeline, necessitating laboratory-specific validation before clinical use.

BRAF V600E is reliably detected by DNA-based hotspot NGS or validated PCR. VE1 IHC can screen for V600E protein, but molecular confirmation is generally recommended given concordance limitations 32.

ctDNA offers a noninvasive complement to tissue profiling, particularly for monitoring known alterations, detecting emergent resistance mutations (e.g., secondary FGFR2 kinase-domain substitutions), and guiding decisions when tissue is scarce 259. However, ctDNA sensitivity for structural rearrangements such as FGFR2 fusions is variable; biliary obstruction in eCCA can further impair tumor DNA shedding, making tissue-first profiling the reference standard 2528.

Recommended Testing Workflow

For all patients with newly diagnosed advanced or metastatic CCA, early comprehensive tumor-based profiling using a combined DNA+RNA NGS panel is recommended, capturing point mutations, copy-number variants, and fusions across a broad actionable target set including FGFR2, IDH1, HER2/ERBB2, BRAF, NTRK, RET, BRCA1/2, and MSI/dMMR status 11123. This approach aligns with both NCCN 2025 and the ESMO 2025 interim guideline update, both of which mandate upfront broad molecular profiling for advanced BTC 128.

For iCCA, priority targets include FGFR2 fusions (with RNA-based or hybrid-capture capability) and IDH1 mutations; HER2 and BRAF V600E should still be assessed comprehensively to avoid missing actionable events in less common subsets. If DNA-only NGS returns negative for FGFR2 fusions in an iCCA patient, reflex RNA-based fusion testing is warranted before concluding FGFR2-negative status 8924. For eCCA, the emphasis shifts toward HER2 IHC with reflex FISH for IHC 2+ cases, supported by NGS copy-number data, while FGFR2 and IDH1 testing should still be performed given their rare but actionable occurrence in extrahepatic disease 1528.

ctDNA should be integrated as a complement—particularly when tissue is inadequate or contraindicated, or when monitoring treatment response and resistance emergence—but should not replace tissue-based profiling for initial alteration detection 259. Re-profiling at progression is clinically supported by data showing only approximately 49% overall oncogenic concordance between primary and advanced iCCA, with evidence that secondary resistance mutations and clonal shifts can be captured through repeat biopsy or liquid biopsy 27.

Standardized Comparison: Actionable Alterations in iCCA Versus eCCA

Footnotes. iCCA: intrahepatic cholangiocarcinoma; eCCA: extrahepatic cholangiocarcinoma (includes perihilar and distal subtypes); BTC: biliary tract cancer; GBC: gallbladder carcinoma; NGS: next-generation sequencing; FFPE: formalin-fixed paraffin-embedded; IHC: immunohistochemistry; FISH: fluorescence in situ hybridization; ISH: in situ hybridization; CNV: copy-number variant; T-DXd: trastuzumab deruxtecan; PCR: polymerase chain reaction. Frequency estimates are approximate ranges derived from contemporary large-scale profiling studies and consensus guidelines; variability across cohorts reflects differences in histologic subclassification, geographic background (including fluke-endemic regions), sample size, tumor purity, and inclusion or exclusion of gallbladder cancer and ampullary tumors 23141523.

Concluding Perspective

The genomic epidemiology of CCA is definitively subtype-specific: treating iCCA and eCCA as a single molecular entity risks systematic under-detection of targetable alterations and missed therapeutic opportunities. FGFR2 fusions and IDH1 mutations define the dominant actionable architecture of iCCA—particularly the small-duct phenotype—whereas HER2 amplification and overexpression are the principal biomarker-driven targets in eCCA and gallbladder-adjacent BTC. BRAF V600E, though rare, demands inclusion in broad profiling regardless of anatomic subtype. Translating this molecular landscape into clinical benefit requires assay platforms capable of detecting fusion-partner diversity through RNA-capable NGS, coordinated HER2 IHC/ISH confirmation workflows, and the judicious integration of ctDNA as a complement to tissue-based profiling. Implementing these strategies within a multidisciplinary molecular tumor board framework—with explicit attention to iCCA versus eCCA prevalence patterns, geographic context, and reflex-testing algorithms—represents the current standard for optimizing access to approved targeted therapies and supporting enrollment into precision oncology trials 1111528.