Second-Line Targeted Therapy in Cholangiocarcinoma: Pemigatinib, Futibatinib, and Ivosidenib — Efficacy, Resistance, and Rational Sequencing

Introduction

Advanced cholangiocarcinoma (CCA) has historically carried a dismal prognosis, with systemic chemotherapy offering modest benefit. The identification of recurrent, therapeutically actionable genomic alterations—FGFR2 fusions/rearrangements in approximately 10–15% of intrahepatic CCA (iCCA) and IDH1 mutations in approximately 14–20%—has enabled a precision-medicine approach that is now supported by regulatory approvals on three continents 926. Three approved targeted agents currently form the cornerstone of molecular therapy for biomarker-selected advanced cholangiocarcinoma: pemigatinib and futibatinib for FGFR2 fusion/rearrangement-positive disease, and ivosidenib for IDH1-mutant disease, and ivosidenib for IDH1-mutant disease. This review synthesizes pivotal trial data, resistance mechanisms, and sequencing principles to guide clinical decision-making in the second-line and later setting.

Clinical Efficacy and Response Durability

FGFR2-targeted therapy operates through kinase inhibition, producing objective radiological tumor shrinkage in a meaningful proportion of patients. Pemigatinib, a selective reversible FGFR1–3 inhibitor, received accelerated FDA approval in April 2020 based on the phase II FIGHT-202 trial 21. In the final analysis with a median follow-up of 45.4 months, 108 patients with FGFR2 fusion/rearrangement-positive iCCA achieved an objective response rate (ORR) of 37.0%, a median duration of response (DoR) of 9.1 months, a median progression-free survival (PFS) of 7.0 months, and a median overall survival (OS) of 17.5 months 1. Importantly, responders demonstrated markedly superior OS compared with non-responders (46.1 versus 13.7 months), underscoring the prognostic significance of early tumor shrinkage. The disease control rate (DCR) was 82.4%, indicating that stable disease also confers meaningful benefit. Subgroup analyses demonstrated numerically higher ORRs in patients with ECOG performance status 0 and non-metastatic disease; co-alterations in TP53 or PBRM1 were associated with significantly worse OS (hazard ratios 3.33 and 2.46, respectively) 1.

Futibatinib, an irreversible covalent FGFR1–4 inhibitor, was evaluated in the pivotal FOENIX-CCA2 trial, a multicenter single-arm phase II study, demonstrating an ORR of approximately 42%, a median DoR of 9.7 months, and a median PFS of 9.0–11.1 months 1427. Cross-trial comparisons must be interpreted with caution given differences in patient populations, prior treatment burden, and study design, but the higher ORR and numerically longer PFS with futibatinib are consistent with its distinct pharmacological mechanism. Real-world data from 120 US patients receiving pemigatinib confirmed a real-world ORR of 59.2% and median PFS of 7.4 months, broadly consistent with FIGHT-202 but with a higher nominal ORR likely reflecting differences in imaging frequency and radiological assessment 4.

IDH1 inhibition operates through a fundamentally different mechanism. Ivosidenib, a selective reversible inhibitor of mutant IDH1, suppresses production of the oncometabolite (R)-2-hydroxyglutarate ((R)-2HG), thereby restoring cellular differentiation rather than inducing direct cytotoxicity. Consequently, the clinical efficacy pattern is predominantly cytostatic: the phase III ClarIDHy trial reported an ORR of only approximately 2% by RECIST, but a median PFS of 2.7 versus 1.4 months with placebo (hazard ratio 0.37) and a median OS of 10.3–10.8 months 2627. After adjustment for the 68–70% crossover rate, the OS benefit widened to approximately 5.1 months in favor of ivosidenib . At 12 months, the PFS rate was 22% in the ivosidenib arm versus 0% with placebo, highlighting durable disease control in a meaningful subset. Patients and clinicians should recognize that absence of radiological shrinkage does not preclude clinical benefit with ivosidenib.

Safety and Treatment Practicality

FGFR inhibitors share a characteristic, predictable toxicity profile dominated by on-target effects. Hyperphosphatemia—reflecting FGFR1-mediated disruption of phosphate homeostasis—occurs in 58–81% of patients with pemigatinib and futibatinib, respectively, and is predominantly grade 1–2 120. Management involves dietary phosphate restriction, phosphate binders, and dose interruption in refractory cases; it rarely necessitates permanent discontinuation. Ocular toxicity, including corneal opacity and central serous retinopathy, affects 5–20% of patients and mandates baseline ophthalmologic assessment and periodic monitoring. Additional class effects include alopecia, nail changes, stomatitis, and diarrhea. In FIGHT-202, dose interruptions occurred in 42.2% and dose reductions in 13.6% of patients, with only 10.2% discontinuing due to adverse events 1. Real-world data suggest even lower rates of dose modification in community practice, likely reflecting shorter follow-up and evolving management experience 4.

Ivosidenib carries a distinct and generally milder toxicity profile. Common adverse events include fatigue, nausea, diarrhea, and abdominal pain, most of which are grade 1–2. QT prolongation has been observed in a dose-dependent manner and warrants baseline electrocardiographic assessment and electrolyte monitoring, particularly in patients receiving concurrent QT-prolonging agents. Differentiation syndrome, well-described with IDH inhibitors in hematological malignancies, is rare in solid tumors . Elevated liver enzymes warrant periodic monitoring given the hepatobiliary disease context.

On-Target Resistance Mechanisms

Resistance to FGFR inhibitors is predominantly driven by secondary FGFR2 kinase-domain mutations. A meta-analysis of 82 patients demonstrated that 60% developed one or more acquired kinase-domain mutations upon progression; the most frequent affected the molecular-brake residue N550 (in 63% of patients with kinase-domain mutations) and the gatekeeper residue V565 (47%) 2. Reversible ATP-competitive inhibitors, including pemigatinib, show pronounced resistance (>100-fold increase in half-maximal inhibitory concentration) to V565F, V565L, and N550K mutations due to steric clashes with the inhibitor-binding pocket 2. Futibatinib's covalent binding to C492 confers retained activity against most kinase-domain mutations, with the critical exception of V565F (>200-fold resistance) 23. Importantly, C492 disruption—which would abolish futibatinib binding—was observed in only 1 of 42 futibatinib-treated patients, because C492 mutations impair FGFR2 signaling and confer poor clonal fitness, in sharp contrast to cysteine-targeting resistance mechanisms encountered with osimertinib or ibrutinib 2.

Resistance in iCCA is characteristically polyclonal: circulating tumor DNA (ctDNA) analysis reveals up to 13 distinct concurrent FGFR2 mutations per patient at progression, a heterogeneity that single-site tissue biopsy systematically underestimates 233. Off-target resistance mechanisms—including MAPK pathway alterations (KRAS, NRAS) and PI3K/mTOR pathway mutations (PIK3CA, TSC1)—occur in approximately 31–36% of patients progressing on FGFR inhibitors 38.

For ivosidenib, acquired resistance mechanisms include second-site IDH1 mutations (e.g., D279N) that disrupt allosteric inhibitor binding while preserving (R)-2HG production, and IDH isoform switching through acquisition of IDH2 mutations, allowing tumors to maintain oncometabolite-driven epigenetic reprogramming despite IDH1 blockade . These findings support the concept that IDH1-mutant CCA remains (R)-2HG–dependent even after disease progression on ivosidenib, with implications for next-generation IDH inhibitor strategies.

Sequencing Logic and Clinical Decision-Making

For FGFR2 fusion-positive iCCA progressing on first-line gemcitabine/cisplatin or chemoimmunotherapy, both pemigatinib and futibatinib represent guideline-endorsed second-line options (NCCN Category 1/2; ESMO-MCBS scores of 2 and 3, respectively) 2627. In the absence of randomized head-to-head data, futibatinib's higher ORR and numerically superior PFS, combined with its broader resistance coverage through irreversible binding, may favor its use as the first FGFR inhibitor. However, pemigatinib's established safety profile, global approval status (USA, EU, Japan), and reversible mechanism may be preferred in patients with prior retinal pathology or tolerance concerns 7.

When a patient progresses on a reversible FGFR inhibitor, futibatinib can overcome the majority of secondary kinase-domain mutations encountered at resistance, particularly N550K and most V565 variants except V565F 31120. Clinical evidence supporting this sequencing approach includes an ORR of 17.9% (5 of 28 patients) with futibatinib after prior FGFR inhibitor exposure in a phase I expansion cohort 20, and a case report documenting 36 months of sustained FGFR-targeting benefit through sequential reversible-then-irreversible inhibitor therapy 11. Molecular re-profiling by tissue biopsy or, preferably, ctDNA analysis at progression is strongly recommended; detection of V565F signals pan-FGFR inhibitor resistance and should prompt consideration of next-generation agents (lirafugratinib, KIN-3248) 36, clinical trials, chemotherapy re-challenge, or best supportive care, depending on performance status and patient goals 526.

For IDH1-mutant CCA, ivosidenib is the standard second-line choice. Its cytostatic mechanism necessitates clear patient communication: disease stabilization—not radiological shrinkage—is the therapeutic objective. Patients with rapidly progressive disease, poor performance status, or highly symptomatic disease may derive limited benefit. Upon progression, the irreversible IDH1 inhibitor LY3410738 and dual IDH1/IDH2 strategies are under clinical investigation and represent rational approaches based on the isoform-switching resistance mechanism .

Evidence Gaps and Future Directions

Critical evidence gaps constrain current practice. No phase III randomized trial directly compares pemigatinib and futibatinib, and all cross-trial efficacy comparisons remain confounded by differences in patient selection and study design 2524. Prospective sequencing studies—evaluating outcomes for patients transitioning between FGFR inhibitors guided by molecular resistance profiling—are absent from the published literature. Neither NCCN 2025 nor ESMO 2024 guideline documents formally recommend ctDNA monitoring algorithms for real-time resistance detection, despite accumulating evidence for its clinical utility in this disease 2627. Emerging next-generation FGFR inhibitors such as lirafugratinib, KIN-3248 (active against V565F), and INCB126503 (a highly isoform-selective FGFR2/3 inhibitor with reduced hyperphosphatemia) hold promise for overcoming recalcitrant resistance mutations 618. The interaction between modern first-line chemoimmunotherapy regimens (e.g., gemcitabine/cisplatin plus durvalumab from TOPAZ-1) and subsequent targeted therapy efficacy remains an important unanswered question.

Comparative Summary Table