Optimizing Treatment Patterns in Relapsed or Refractory Differentiated Thyroid Cancer

Differentiated thyroid cancer (DTC) has a high cure rate, and patients who do not respond to initial standard-of-care treatments represent a significant clinical challenge.¹

“Typically, patients will present with an incidental mass in their thyroid or perhaps in their neck, and sometimes thyroid cancers are identified on imaging studies,” Frances P. Worden, MD, said in a recent OncLive Peer Exchange. “The paradigm for the treatment of these individuals is surgery and neck dissection followed by radioactive iodine [RAI] therapy. Patients are then followed using markers such as thyroglobulin and ultrasounds. If there’s locally advanced disease, they will go on to receive treatment, perhaps with another surgery or more RAI, until they become refractory. That’s typically when these patients are referred to us [head and neck cancer specialists].”

The disease has a 5-year overall survival (OS) rate of approximately 98%,² and patients generally respond well to treatment, moderator Lori J. Wirth, MD, explained. However, when patients develop distant metastatic disease, the median OS caps at 5 years. In the frontline setting, tyrosine kinase inhibitors (TKIs) have served as a backbone treatment; however, novel options have entered the treatment landscape with the advent of next-generation sequencing (NGS) identifying actionable mutations.¹

During the Peer Exchange, a panel of head and neck cancer experts discussed the nuances of managing relapsed or refractory DTC with lenvatinib (Lenvima) or sorafenib (Nexavar), as well as sequencing considerations with targeted therapies in the second-line setting and beyond for patients with relapsed or refractory disease.

Deciding to Initiate Treatment

Patients with DTC who have disease progression and/or loss of RAI uptake after 600 mCi of radioactive iodine I-131 are considered to have refractory disease.^3,4 “Some countries will push it a little farther, but generally that would also be a term that we would use for refractory if the patient has gone beyond that,” Worden said. At that point, treatment algorithms have suggested that 2 courses of action remain: observation or systemic therapy.

The global, noninterventional RIFTOS MKI study (NCT02303444) showed that among 647 patients who either underwent treatment with a multikinase inhibitor (MKI) or were assigned to a watch-and-wait approach, the median time to symptomatic progression (TTSP) was 55.4 months (IQR, 18.6-not estimable [NE]). For patients who received lenvatinib, sorafenib, or another MKI, the median TTSP was 55.4 months (IQR, 15.2-NE). The median TTSP in the cohort of patients who did not initiate MKI treatment at study entry was 51.4 months (IQR, 20.0-NE).⁴

“We followed these individuals for a period and then considered starting treatments in those windows [when the disease is] slowly growing,” Worden said. “No matter how you want to do it—3 months, 6 months—[we did] whatever was appropriate. The goal of the study was to say, ‘This is when you’re probably going to start on therapy.’ However, we couldn’t determine that, but we could figure out that the median TTSP was approximately 55.4 months. I love these data because there is a subset of patients that we can watch. Patients come in and say ‘I have disease that’s incurable. What are we going to do here?’ And I can say, ‘Here are good data.’ I show them the curves from this paper and [note] that we can safely watch [their disease]. That has been helpful, especially if they have small amounts of tumor…. I think that study is important, and I’m glad [its results have been] published.”

In practice, Ezra Cohen, MD, FRCPSC, FASCO, noted that he is hesitant to start patients on MKIs right away, even after presenting the patient with the news that although treatments are available, none are curative. “My conversation with a patient starts with [informing them that they] unfortunately have RAI-refractory disease and evidence of disease, he said. He added that he also informs the patient that no curative treatments for the disease are available. “Secondly, I want to delay as much as possible exposure to agents [that are available]. If [the patient] is feeling well, I can’t make them feel better by starting on one of these drugs, even though we know that the very specific kinase inhibitors are likely to be incredibly well tolerated with very few adverse effects [AEs]. I’m still not going to make an asymptomatic patient feel better by putting them on a drug,” Cohen said.

Should a patient become symptomatic and/ or have progression that necessitates systemic treatment, Cohen noted that in the absence of molecular alterations, lenvatinib is the firstchoice treatment based on data from the SELECT trial (NCT01321554). In the phase 3 study, patients with RAI-refractory DTC were randomly assigned to either lenvatinib (n = 261) or placebo (n = 131). The intervention was associated with a significant improvement in overall response at 64.8% vs 1.5% with placebo (OR, 28.87; 95% CI, 12.46-66.86; P < .001).³ Complete responses were reported in 4 patients treated with lenvatinib and were maintained through the evaluation period.

Further, the median progression-free survival (PFS) with lenvatinib was 18.3 months (95% CI, 15.1-NE) vs 3.6 months (95% CI, 2.2-3.7) with placebo (HR, 0.21; 99% CI, 0.14-0.31; P < .001).³

“[In cases where I don’t know about] molecular alterations or in the absence of a targetable molecular alteration, my first choice is lenvatinib based on the data from the SELECT trial—the high response rate, and let’s not forget some patients had a complete response, which we don’t really see with sorafenib,” Cohen said.

Cohen added that patient characteristics are important to keep in mind when deciding on a course of treatment. For example, results of a post hoc subgroup analysis of SELECT demonstrated that patients with lower volume disease with better performance status performed better with lenvatinib than the general population.⁵

“If for some reason I can’t use lenvatinib, I will use sorafenib or cabozantinib [Cabometyx],” Cohen said. “I try to reserve cabozantinib for TKI-refractory patients based on data from a phase 3 [COSMIC-311] study [NCT03690388].”

Second-line Treatment and Beyond

Although biologically cabozantinib and lenvatinib have similar activity, Cohen said that he reserves the former for lenvatinib-refractory patients “because the level 1 evidence demonstrates benefit there.”

In COSMIC-311, patients with RAI-refractory DTC who previously received VEGFR-targeted therapy were randomly assigned 2:1 to receive cabozantinib or matching placebo. Among 125 patients who received cabozantinib, the objective response rate did not meet statistical significance at 15% (99% CI, 5.8%-29.3%) vs 0% in patients who received placebo (n = 62). The primary end point of PFS was met at interim analysis and was not reached (96% CI, 5.7-NE) among patients who received cabozantinib vs a median of 1.9 months (96% CI, 1.8-3.6) for those who received placebo (HR, 0.22; 96% CI, 0.13-0.36; P < .0001).⁶

The agent was approved by the FDA in September 2021 for the treatment of adult and pediatric patients 12 years and older with locally advanced or metastatic DTC that has progressed following prior VEGFR-targeted therapy and who are ineligible or refractory to RAI.⁷

In an analysis of cabozantinib based on prior duration of lenvatinib treatment, the median PFS was improved vs placebo among patients regardless duration of prior therapy. The HRs for those who received lenvatinib for less than 1 year, 1 to 2 years, or longer than 2 years were 0.32 (95% CI, 0.15-0.69), 0.22 (95% CI, 0.10-0.51), and 0.28 (95% CI, 0.14-0.55), respectively.⁸

Molecular Testing

Cohen added that one of the most important tools for managing DTC is NGS. “I get molecular profiling for all patients,” he said, adding that his institution is equipped to accommodate this protocol. “I get it for all patients because now we know that most patients with papillary thyroid cancer will have a targetable molecular alteration and some of the available agents are incredibly effective. The TRK inhibitors, the RET inhibitors [have] approximate response rates of 70% to 80% and median durations of response that in some trials have not been reached.”

BRAF mutations are the most prevalent in patients with DTC9; however, Marcia S. Brose, MD, PhD, said that in the first-line setting BRAF inhibitors are not the optimal treatment choice. “There is 1 exception in my opinion, which is BRAF, because the BRAF inhibitors aren’t quite as good as the MKIs as far as efficacy,” she said. “They might be about the same, maybe slightly better as far as tolerability, but they really are inferior as far as overall response rate. I hold off on treating with BRAF inhibitors until after lenvatinib in the first-line setting, but for the most part knowing about those NTRK fusions and RET fusions up front is important. So now I do RNA-based NGS prior to treatment.”

Brose added that although she and Cohen both recommend testing for all patients, overtreatment and excessive test ordering should be avoided. “This comes up a lot, especially in countries and places where money is limited,” Brose said. “We don’t want to be doing RNA-based NGS on all patients because 85% of the time we actually won’t need that information, especially if we’re going to cure the patient with surgery and radioactive iodine alone—their genotype does not matter. Additionally, for patients with advanced cancer, we know that [more than] 50% of them will have the BRAF mutation. These tend to be mutually exclusive. If you do 2 steps of testing— where you first test for the BRAF mutation with an inexpensive test and remove another half of the population, then you’re only doing the expensive RNA-based NGS on [a smaller population]. The downside is you use up more tissue and it takes longer.”

In approaching a treatment decision for patients who do have an identified NTRK gene fusion, RET fusion, or BRAF V600E mutation, the panel noted that sequencing, though largely driven by efficacy, is also driven by tolerability.^7,10-16

For example, in discussing use of the approved RET inhibitor selpercatinib (Retevmo), Worden said his experience treating patients with the agent is why it is his first choice. “I was involved with the [early] selpercatinib studies and had a lot of experience treating both patients with medullary [thyroid cancer] and lung cancer and those with RET fusions and DTC,” he said. “In terms of the AE profile, there are similar hematologic issues that can arise with pralsetinib [Gavreto]. The daily dosing [of pralsetinib] does offer an advantage. Interestingly I’ve had issues with [selpercatinib-related] AEs that, despite lowering the dose, adjusting intervals, or even [managing AEs], patients can’t overcome them. I’ve stopped selpercatinib and then put them on pralsetinib. Even vice versa, I’ve seen patients tolerate the other RET inhibitor quite well without the AEs they were experiencing with the initial drug. My go-to is still selpercatinib and I will, if for some reason there’s a tolerability issue, then shift over to pralsetinib.”

Brose echoed this experience in practice, stating that switching was not common for reasons other than toxicity, such as disease progression, because patients did not remain on the second agent for too long. “I only [switched] on 2 different occasions where patients were progressing because I had no other options,” Brose said. “The new agent didn’t last that long. I don’t think there were a lot of patients staying on it for a long time. Toxicity is a better reason for switching between the two, but I do think in terms of sequencing, because of the responses, we tend to now put these [targeted agents] up front.” Brose added that the National Comprehensive Cancer Network guidelines do not state how they should be sequenced because the RET inhibitors and TRK inhibitors were approved based on response rate.¹⁷ “We probably won’t have as much of a head-to-head trial because the [mutations are] still so rare, and at the end of the day, I don’t think we’re going to get those data. It is something you can consider.”

Acquired Resistance and the Future of Sequencing

NGS serves another role in the postRET inhibitor setting as well: identifying mechanisms of resistance. “It’s so important when patients do develop acquired resistance that we do a tissue biopsy to identify the mechanisms of resistance because there are the on-RET solvent front mutations and other on-target acquired resistance mutations for which RET inhibitors may work,” Cohen said. “But we also see bypass pathway activation as mechanisms of acquired resistance, and second-generation RET inhibitors do not have much promise in that setting.”

For those who do progress following systemic first-line treatment, the next step in RET-driven DTC would include MKIs such as lenvatinib or selpercatinib. Citing the LIBRETTO-531 trial (NCT04211337), Wirth noted that its data may inform new pathways forward following progression on RET inhibitors. “The primary end point of that trial is PFS, and worked into the trial design is crossover to selpercatinib from the investigator’s choice arm when patients progress on cabozantinib or vandetanib. We will have good sequencing data in medullary thyroid cancer in the future. But for now, this is an important question that we must wrestle with in the clinic,” Wirth said.

Cohen agreed that although there are no data to support one approach over another, “the absence of data doesn’t mean that it’s negative data.”

Worden added that treatment of patients becomes based on extrapolation of findings from clinical experience and individual treatment circumstances. For example, he noted that for a patient with disease progression whose DTC becomes poorly differentiated, he would likely proceed to cabozantinib or lenvatinib as a second-line treatment. “Although the patients who were treated in the COSMIC study and the SELECT trial weren’t necessarily on directed therapies as frontline therapy because they weren’t around then, I can’t help [but] believe that they couldn’t respond to those 2 agents.”

Brose cited her own patients, adding, “I have a couple of patients whom I’ve treated long enough that they started to progress, and I put them on lenvatinib, and so far, so good. Who knows how long that will last, but I think the agent has enough of a different mechanism that we are getting response.”

References

1. Lieberman L, Worden F. Novel therapeutics for advanced differentiated thyroid cancer. Endocrinol Metab Clin North Am. 2022;51(2):367-378. doi:10.1016/j.ecl.2021.11.019
2. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin. 2022;72(1):7-33. doi:10.3322/caac.21708
3. Schlumberger M, Tahara M, Wirth LJ, et al. Lenvatinib versus placebo in radioiodine-refractory thyroid cancer. N Engl J Med. 2015;372(7):621-630. doi:10.1056/NEJMoa1406470
4. Brose MS, Smit JWA, Lin CC, et al. Multikinase inhibitors for the treatment of asymptomatic radioactive iodine-refractory differentiated thyroid cancer: global noninterventional study (RIFTOS MKI). Thyroid. 2022;32(9):1059-1068. doi:10.1089/ thy.2022.0061
5. Fukuda N, Takahashi S. Clinical indications for treatment with multi-kinase inhibitors in patients with radioiodine-refractory differentiated thyroid cancer. Cancers (Basel). 2021;13(9):2279. doi:10.3390/cancers13092279
6. Brose MS, Robinson B, Sherman SI, et al. Cabozantinib for radioiodine-refractory differentiated thyroid cancer (COSMIC-311): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol. 2021;22(8):1126-1138. doi:10.1016/S14702045(21)00332-6
7. FDA approves cabozantinib for differentiated thyroid cancer. FDA. Updated September 22, 2021. Accessed May 31, 2023. bit.ly/3N9PapO
8. Brose MS, Krajewska JA, Vaiman F, et al. Cabozantinib (C) vs placebo (P) in patients (pts) with radioiodine-refractory (RAIR) differentiated thyroid cancer (DTC) who progressed after prior VEGFR-targeted therapy: outcomes by duration of prior lenvatinib (L) treatment. Ann Oncol. 2022;33(suppl 7):S1297-S1298. doi:10.1016/j.annonc.2022.07.1733
9. Lin Y, Zhang X, He H, et al. Clinical and molecular profiles of radioiodine refractory differentiated thyroid cancer. Ann Oncol. 2022;33(suppl 7):S1300. doi:10.1016/j.annonc.2022.07.1737
10. Vitrakvi. Prescribing information. Bayer HealthCare Pharmaceuticals Inc; 2022. Accessed June 2, 2023. bit.ly/3WKVtU0
11. Rozlytrek. Prescribing information. Genentech Inc; 2022. Accessed June 2, 2023. bit.ly/3NbszcP
12. Retevmo. Prescribing information. Eli Lilly and Company; 2022. Accessed June 2, 2023. bit.ly/43HKTj9
13. Gavreto. Prescribing information. Blueprint Medicines Corp and Genentech Inc; 2022. Accessed June 2, 2023. bit.ly/3oMzC27
14. Tafinlar. Prescribing information. Novartis Pharmaceuticals Corp; 2023. Accessed June 2, 2023. bit.ly/3C9HFZO
15. Mekinist. Prescribing information. Novartis Pharmaceuticals Corp; 2023. Accessed June 2, 2023. bit.ly/3ISVPlO
16. Lenvima. Prescribing information. Eisai Inc; 2022. Accessed June 2, 2023. bit.ly/3qn0N3W
17. NCCN. Clinical Practice Guidelines in Oncology. Thyroid carcinoma, version 2.2023. Accessed May 31, 2023. https://www.nccn.org/ professionals/physician_gls/pdf/thyroid.pdf