Targeted Therapy in Advanced Thyroid Cancer: Efficacy Benchmarks, Resistance Mechanisms, and Clinical Implications

Clinical Rationale and Biomarker-Driven Treatment Selection

Advanced thyroid cancer encompasses a heterogeneous spectrum of histologies — differentiated thyroid cancer (DTC), poorly differentiated thyroid cancer (PDTC), anaplastic thyroid cancer (ATC), and medullary thyroid cancer (MTC) — each defined by distinct molecular drivers that confer sensitivity to rationally designed, genotype-directed therapies. Three alterations have emerged as the most therapeutically actionable targets in current practice: the BRAF V600E突变 (BRAF V600E mutation), RET融合 (RET fusion), and NTRK融合 (NTRK fusion).

The BRAF V600E substitution constitutes the predominant driver in a subset of papillary thyroid cancers and is detected at high frequency in ATC, a disease characterized by dismal prognosis and limited systemic options prior to the advent of targeted therapy. Dual MAPK pathway blockade with dabrafenib (a selective BRAF inhibitor) and trametinib (a MEK inhibitor) suppresses the constitutive BRAF-to-MEK-to-ERK signaling cascade and mitigates paradoxical ERK reactivation that would otherwise occur with single-agent BRAF inhibition 116.

RET fusions represent oncogenic drivers in a minority of DTCs — particularly in younger patients and those with prior radiation exposure — and activating RET point mutations are the canonical molecular hallmark of MTC. Selpercatinib, a highly selective RET inhibitor, achieves potent RET kinase inhibition with substantially improved tolerability relative to older, multikinase inhibitors (MKIs) such as vandetanib and cabozantinib 4.

NTRK1/2/3 fusions are rare in adult PTC but enriched in pediatric PTC and constitute actionable drivers across thyroid histologies. Larotrectinib, a selective pan-TRK inhibitor, exploits this fusion-dependent oncogenesis in a tumor-agnostic manner, yielding robust responses independent of histologic subtype, with modest activity in ATC where co-existing genomic complexity may limit TRK dependence 7.

Contemporary guidelines from NCCN (2025), ESMO, and ATA (2025) unanimously mandate comprehensive molecular profiling — including BRAF V600E, RET fusion, and NTRK fusion testing — as a prerequisite to first-line systemic therapy decisions in advanced, RAI-refractory disease, prioritizing targeted agents when actionable drivers are identified 171819.

Efficacy Benchmarks

Dabrafenib plus trametinib (BRAF V600E突变): Phase II evidence from the ROAR basket trial established the clinical utility of this combination in BRAF V600E–mutant ATC. In an updated analysis of 36 ATC patients, the investigator-assessed ORR was 56% (95% CI 38.1–72.1%), including 3 complete responses; median duration of response (DOR) was 14.4 months, median PFS 6.7 months, and median OS 14.5 months, with 12-month OS approximately 51.7% and 24-month OS approximately 31.5% — a remarkable outcome benchmark in a disease historically associated with survival measured in weeks to months 16. An earlier analysis of 16 evaluable ATC patients reported an ORR of 69% with 12-month DOR approximately 90%, 12-month PFS approximately 79%, and 12-month OS approximately 80%, reflecting the impact of small sample sizes on point estimates 1. In BRAF V600E RAI-refractory DTC (PHAROS trial data), ORR was approximately 51% (95% CI 37–65%), disease control rate approximately 79%, median PFS approximately 11.3 months, and median DOR approximately 18.9 months 16. Notably, these cross-histology efficacy differences — ATC versus DTC — are partially reflective of patient selection, disease biology, and prior therapies rather than drug potency per se, and indirect cross-trial comparisons should be interpreted cautiously.

Selpercatinib (RET融合): Across the LIBRETTO-001 phase 1/2 program in RET fusion–positive (non-MTC) thyroid cancer, ORR was 95.8% in treatment-naïve patients (n=24) and 85.4% in pretreated patients (n=41), with 2-year PFS of 95.2% in the naïve cohort and median PFS of 27.4 months in the pretreated group 5. EMA-level synthesis of LIBRETTO-001 reported an overall ORR of 79% (95% CI 65–88%) with median DOR approximately 18.4 months; intracranial ORR of approximately 91% was reported in patients with brain metastases from RET fusion–positive tumors 2. In RET-mutant MTC, long-term LIBRETTO-001 data demonstrated ORR 82.5% in MKI-naïve patients and 3-year PFS of 75.2% (naïve) versus 54.6% (MKI-pretreated). The landmark phase 3 LIBRETTO-531 trial established superiority over physician's choice (vandetanib or cabozantinib) in first-line RET-mutant MTC: PFS hazard ratio 0.28 (95% CI 0.16–0.48; P<0.001), median PFS not reached versus 16.8 months, and ORR 69.4% versus 38.8% 6.

Larotrectinib (NTRK融合): Pooled thyroid-specific data from three multicenter phase 1/2 basket trials (n=29 thyroid patients) demonstrated an overall ORR of 71% (95% CI 51–87%), with DTC (PTC/FTC) ORR 86% and ATC ORR 29%, emphasizing the histologic differential in TRK dependency 7. Twenty-four-month DOR was 81%, 24-month PFS 69%, and 24-month OS 76%. Across the broader tumor-agnostic dataset (n=55), ORR was 75% (95% CI 61–85%), DOR ≥6 months in 73% of responders, and the thyroid subset (n=5) yielded an ORR of 100%, albeit in a very small sample 3. Real-world cohort data further corroborate an approximately 75% ORR in thyroid cancer patients 16.

Safety and Practical Clinical Considerations

Dabrafenib plus trametinib carries a well-characterized toxicity profile including pyrexia, fatigue, rash, and metabolic effects. Grade 3/4 events in the ROAR ATC cohort included hyponatremia (19%), pneumonia (13%), and anemia (13%) 14. The standard regimen is dabrafenib 150 mg twice daily plus trametinib 2 mg once daily, with stepwise dose reductions available. A critical pharmacologic consideration is that dabrafenib is a potent CYP3A4 and CYP2C8 inducer, which can substantially reduce exposure to coadministered CYP3A4 substrates, including hormonal contraceptives; clinicians must screen for interactions proactively. Hyperglycemia and pyrexia syndromes require structured monitoring and management 16.

Selpercatinib is dosed at 160 mg twice daily (≥50 kg) or 120 mg twice daily (<50 kg). Clinically relevant grade ≥3 toxicities include hypertension (approximately 15–22%), ALT/AST elevations (6–11%), and QTc prolongation (approximately 4–5%). Dose reductions occurred in approximately 30–39% and permanent discontinuations in approximately 4.7–5% in pivotal trials, markedly less than MKI comparators (26.8% discontinuations) 56. QTcF >500 ms or an increase >60 ms from baseline requires dose interruption. Strong CYP3A4 inhibitors increase selpercatinib exposure and may necessitate dose reduction to 80 mg twice daily; proton pump inhibitors reduce drug exposure, favoring timed antacid use. Rare chylous effusions have been reported in case series 2. Hepatic function monitoring before initiation and throughout therapy is mandatory 16.

Larotrectinib has the most favorable tolerability profile among the three agents. The majority of treatment-emergent adverse events are grade 1–2 (fatigue, myalgia, dizziness, mild transaminase elevations); grade ≥3 neurologic adverse events occur in approximately 6% and hepatotoxicity (any-grade ALT/AST elevation) in approximately 45%, with dose reductions rare and no permanent discontinuations for toxicity in the thyroid-specific cohort 37. Strong CYP3A4 inhibitors increase larotrectinib AUC approximately 4.3-fold; strong inducers reduce exposure by approximately 81%; interaction screening is essential before co-prescribing 16.

Across all three agents, ATA 2025 guidelines emphasize re-differentiation strategies — particularly with BRAF/MEK inhibition — to restore RAI sensitivity in select DTC patients before additional MKI or non-targeted systemic therapy, representing a unique sequencing opportunity 1916.

Resistance Mechanisms

Dabrafenib plus trametinib: Acquired resistance in BRAF V600E–driven thyroid cancers predominantly converges on MAPK pathway reactivation, mediated through secondary RAS mutations (NRAS, KRAS), BRAF amplification, MEK1/2 activating mutations, and BRAF splice variants. Co-occurring mutations prevalent in ATC — TP53, TERT promoter, CDKN2A, and SWI/SNF chromatin remodeling gene alterations — contribute to intrinsic resistance and limited durability. NF1 loss and RTK upregulation (EGFR, HER2) further diversify the resistance landscape 16. Systematic thyroid-specific resistance characterization remains limited in retrieved materials, with much of the mechanistic framework extrapolated from melanoma and other BRAF-driven malignancies.

Selpercatinib: On-target resistance emerges via kinase-domain mutations that impair drug-binding geometry: solvent-front mutations G810C/S/R and hinge-region variants Y806C/N reduce binding affinity while preserving RET kinase activity; V738A has also been described. These mutations confer cross-resistance to pralsetinib in preclinical models 12. Bypass resistance pathways include MET amplification (potentially addressable by combining selpercatinib with a MET inhibitor) and KRAS amplification. Preclinical data from RET/PTC1-positive thyroid cell models implicate Akt–mTOR pathway upregulation as an intrinsic and acquired resistance mechanism, with everolimus demonstrating activity across parental and resistant lines, supporting a rationale for RET plus mTOR combination strategies 13.

Larotrectinib: On-target resistance is driven by solvent-front and gatekeeper mutations in NTRK kinase domains: TRKA G595R (NTRK1) and TRKC G623R, G696A, and F617L (NTRK3), as well as NTRK2 G667C, reduce inhibitor binding. Off-target bypass mechanisms include KRAS amplification, MET amplification, and the emergence of BRAF fusions, indicating clonal evolution toward alternative growth signaling cascades. Next-generation TRK inhibitors — selitrectinib (LOXO-195) and repotrectinib — demonstrate activity against solvent-front resistance mutations and represent an established sequential strategy 3716.

Clinical Interpretation and Future Directions

The convergence of regulatory approvals, randomized phase 3 data (for selpercatinib), and phase 2 basket trial evidence (for dabrafenib–trametinib and larotrectinib) firmly establishes a genotype-first approach in advanced thyroid cancer. Broad next-generation sequencing (DNA and RNA) with fusion calling should be performed at diagnosis of advanced disease and at progression to identify primary drivers and emergent resistance mechanisms. RNA-based assays are particularly important for NTRK3 due to intronic architecture and for RET fusion detection 714.

At disease progression, tissue re-biopsy or circulating tumor DNA (ctDNA) assessment is critical to distinguish on-target resistance (e.g., RET G810/Y806 variants, TRK solvent-front mutations) from bypass mechanisms (e.g., MET amplification, KRAS alteration), as these pathways demand fundamentally different next-line strategies — sequential inhibitor transition for on-target mechanisms, and rational combination or immunotherapy integration for bypass events 1216.

Key unmet needs include: next-generation RET inhibitors retaining potency against solvent-front and hinge-resistance mutations; prospective characterization of BRAF/MEK resistance mechanisms in thyroid-specific ATC cohorts; evidence-based combination regimens targeting bypass pathways (e.g., RET plus mTOR, RET plus MET); and larger datasets in rare molecular subsets such as non-V600E BRAF alterations, non-canonical NTRK partners in ATC, and treatment-naïve RET fusion DTC 121314.

Summary Comparison Table