When to Begin Treatment

According to Dr. Wirth, the clinical imperative is to confirm RAI refractoriness before systemic therapy is initiated. Per the American Thyroid Association (ATA) guidelines, this includes tumors that never concentrate iodine, those that progress despite RAI uptake, or those that have received high cumulative doses without benefit.¹ Additional RAI can lead to long-term harm, including secondary malignancies; a recent analysis estimated a 0.03% risk of leukemia associated with increasing RAI exposure.²

Among patients with asymptomatic, slowly progressing RAI-refractory DTC, immediate treatment may not be necessary. The RIFTOS MKI study compared outcomes in patients who initiated a multikinase inhibitor (MKI) versus those observed without treatment. In the observation cohort, median progression-free survival (PFS) was 13.8 months compared with 18.3 months in the MKI cohort.³ These data support the safety of active surveillance in appropriately selected patients.

Commenting on the evolving debate around the timing of MKI initiation, Dr. Wirth noted: “There aren’t many settings in oncology where we have effective drugs and choose not to give them. But that’s the position we’re often in with DTC.”

Selecting Among Treatment Options

For those who do require treatment, the preferred first-line therapy in the current NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Thyroid Carcinoma is lenvatinib, based on data from the SELECT trial. These recommendations are outlined in the NCCN Guidelines, which summarize systemic therapy options for RAI-refractory DTC, including preferred first-line and second-line regimens (Figure 1).

View Full Size

Principles of Systemic Therapy. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Thyroid Carcinoma, Version 1.2025 [THYR-B].

©2025 National Comprehensive Cancer Network® (NCCN®), All rights reserved. NCCN Guidelines and this illustration may not be reproduced in any form without the express written permission of NCCN. To view the recent and complete version of these NCCN Guidelines, go to NCCN.org.

Citation: Journal of the National Comprehensive Cancer Network 23, Supplement; 10.6004/jnccn.2025.5015

In the phase III SELECT study, lenvatinib improved median PFS from 3.6 months (placebo) to 18.3 months, and demonstrated a 65% objective response rate (ORR). In contrast, sorafenib, examined in the DECISION trial, showed a median PFS of 10.8 months and a 12% ORR. Although both agents are category 1 options, lenvatinib is preferred because of its superior efficacy.^4,5

However, Dr. Wirth reminded, MKIs come with a chronic toxicity burden.⁴ In the SELECT trial, 14% of patients discontinued lenvatinib due to adverse events, and most required dose reductions. The adverse event profile—primarily hypertension, diarrhea, proteinuria, and fatigue—must be weighed carefully, especially given the long median duration of benefit, which exceeds 3 years in some patients, she noted.

The field initially moved toward delayed initiation of MKIs to preserve quality of life. However, retrospective analyses of the SELECT trial findings suggest this may come at the cost of efficacy. According to one retrospective analysis described by Dr. Wirth, patients who started treatment with an ECOG performance status (PS) of 0 had improved ORR, PFS, and overall survival than those with a PS of 1.⁶

Another analysis focused on patients with lung metastases ≥1 cm, and it showed that early treatment with lenvatinib conferred a statistically significant overall survival benefit (median OS, 44.7 vs 33.1 months), despite 89% crossover from placebo.⁷ These results suggest that earlier initiation may offer better disease control and survival, especially in symptomatic or higher-risk patients.

The appropriate starting dose of lenvatinib was evaluated in Study 211, which compared the standard 24 mg daily dose with 18 mg. The lower dose did not meet noninferiority criteria for response rate and showed no significant reduction in grade ≥3 toxicities.⁸ Thus, the 24-mg dose remains the recommended starting dose, with dose reductions as needed for tolerability, Dr. Wirth advised.

In patients who experience disease progression after first-line MKI therapy, cabozantinib is the standard second-line agent. The COSMIC-311 trial enrolled patients previously treated with lenvatinib and/or sorafenib and demonstrated improved PFS (11.0 vs 1.9 months) (hazard ratio [HR], 0.22 [96% CI, 0.15–0.32]; P<.0001) with cabozantinib versus placebo, although the ORR was low (< 15%).⁹

Use of Molecular Profiling in Guiding Appropriate Treatment Selection

At the same time, molecular profiling is reshaping the treatment algorithm. In a large MSK-IMPACT analysis of >10,000 tumors,¹⁰ thyroid cancer had the second-highest rate of potentially actionable alterations among solid tumors, with common alterations in DTC being BRAF V600E, NRAS/HRAS, and RET.

RET fusions occur in approximately 10% of papillary thyroid cancers overall, but they are found in up to 30% of pediatric cases and in tumors associated with prior radiation exposure.^11–13 NTRK fusions are also more common in younger patients and confer distinct histologic features, such as multinodularity, fibrosis, and lymphovascular invasion.¹⁴

For these molecular subtypes, gene-targeted therapies provide efficacy and tolerability. In the LIBRETTO-001 trial, selpercatinib achieved a 96% ORR (including 21% complete responses) in RET fusion–positive DTC among MKI-naïve patients. Median PFS has not yet been reached in this group, with responses ongoing beyond 4 years.¹⁵ “RET fusion–positive thyroid cancer is where we see some of the most dramatic responses, with complete responses more than 20% when selpercatinib is used first line,” said Dr. Wirth, emphasizing the importance of early genomic testing. Even in pretreated patients, the ORR remained at 85%, with a median PFS of 27 months.

Similarly, larotrectinib, a highly selective TRK inhibitor, produced an 88% ORR in NTRK fusion–positive DTC, including a 16% complete response rate and a median duration of response of >32 months (95% CI, 22.8 months–not estimable). Larotrectinib also offers a favorable safety profile and infrequent dose-limiting toxicities, according to Dr. Wirth. This ORR was higher than the 79% ORR reported across all tumor types in the pooled trial data.¹⁵ These agents are now listed in the NCCN Guidelines as first-line options in fusion-positive DTC.

Other TRK inhibitors, including entrectinib and repotrectinib, are also approved in this setting.¹⁶ In addition, pralsetinib remains available for RET fusion–positive tumors, with a 91% ORR demonstrated in the ARROW trial.¹⁷

For tumors harboring BRAF V600E mutations (seen in up to 60% of papillary thyroid cancers), dabrafenib with or without trametinib is approved by the FDA under tissue-agnostic indications. However, in an NCCN-sponsored trial, response rates were modest (30%–35%) with this strategy, and PFS was less than that achieved with lenvatinib, suggesting that BRAF-targeted therapy may be better suited to the second-line setting.¹⁸

Considerations for Sequencing Therapies

According to Dr. Wirth, one of the most pressing questions remains how to optimally sequence therapies. Although the activity of selpercatinib and larotrectinib in pretreated patients is well documented, outcomes are significantly better when these agents are given as initial therapy rather than after MKIs, because the median PFS drops substantially when gene-targeted agents are used after MKIs.

“With these gene-targeted therapies, the argument for delaying treatment becomes a lot weaker—especially when side effects are minimal, and efficacy is high,” Dr. Wirth explained. Given the absence of effective third-line options after treatment resistance, using the best drug first remains a guiding principle for many clinicians.

At Massachusetts General Hospital, next-generation sequencing testing is routinely performed on archival tissue in patients expected to require systemic therapy, even if treatment is not immediately needed. Liquid biopsy currently lacks sensitivity in this setting because of low circulating tumor DNA levels in DTC; thus, molecular testing should prioritize RNA-based panels capable of detecting kinase fusions, particularly those with rare 5′ partners, Dr. Wirth explained.

The management algorithm for RAI-refractory DTC now includes several precision strategies. Patients with actionable fusions should receive gene-specific therapy as first-line treatment. For those without targetable alterations, lenvatinib remains the preferred initial option, with cabozantinib used on disease progression. Dabrafenib with or without trametinib may be considered in BRAF-mutant tumors after MKI failure, as well as for those who are highly symptomatic or experiencing disease progression and have no targetable alterations. Lenvatinib may be combined with pembrolizumab, as supported by limited data showing a 15% ORR in a salvage setting.¹⁹

What Next?

Dr. Wirth posed some clinical questions that are emerging in the treatment of thyroid cancers. Can these targeted agents be used in the neoadjuvant or adjuvant setting? Can kinase inhibitors restore iodine avidity through redifferentiation therapy? And how do we overcome acquired resistance when patients experience disease progression on gene-specific drugs? Trials are underway, but data remain limited given the rarity of eligible populations.

Although these complexities remain unresolved, Dr. Wirth acknowledged, one fact is clear: the molecular landscape of thyroid cancer is no longer optional knowledge. In RAI-refractory DTC, identifying the right patient, at the right time, for the right therapy now means understanding the tumor’s genetic signature—and acting on it with urgency and precision.