Prospects for Redifferentiating Agents in the Use of Radioactive Iodine Therapy for Thyroid Cancer

The ability of thyroid follicular cells to uptake and retain iodide allows the use of radioactive iodine (RAI) for imaging and targeted killing of residual and metastatic thyroid cancer after thyroidectomy. However, ∼30% of patients with distant metastatic differentiated thyroid cancer (DTC) have RAI refractory (RAIR) disease either due to insufficient RAI uptake or resistance to the destructive effects of RAI therapy, and this correlates with a reduced 10-year survival rate of ∼10%. For DTC patients with progressive RAIR disease, sorafenib and lenvatinib have been shown to prolong progression-free survival, with lenvatinib having a better objective response rate than sorafenib when each agent was compared with placebo. The overall survival was not significantly improved by sorafenib or lenvatinib, which may be due to the crossover design of both clinical trials. However, the patient's quality of life is often compromised by a wide spectrum of adverse effects associated with sorafenib and lenvatinib, and some patients have to forego further treatment. Recent success in applying short-term treatment of BRAF and/or MEK inhibitors to restore/enhance RAI accumulation in DTC patients with RAI nonavid metastatic lesions has fueled much enthusiasm in redifferentiation RAI therapy.

Since Na⁺/I⁻ symporter (NIS)-mediated iodide influx is the limiting step for thyroidal RAI accumulation and NIS expression is reduced or absent in malignant thyroid tissues, much effort has been focused on identifying small molecule inhibitors to restore/enhance NIS expression and function. In addition to BRAF/MEK inhibitors, retinoic acid, histone deacetylase inhibitors (trichostatin, tributyrin, LBH589/panobinostat), poly(ADP-ribose) polymerase-1 (PARP-1) inhibitor PJ34, a non-nucleoside reverse transcriptase inhibitor (nevirapine), AMPK inhibitor (compound C), and an inverse agonist of ERRγ (GSK5182, DN200434) have been shown to restore or enhance NIS expression/function in anaplastic thyroid cancer (ATC) cell lines, DTC cell lines, or rat thyroid follicular cell line.

In this issue of Thyroid, Oh et al. (1) identified a novel tyrosine kinase inhibitor 5-(5-{4H, 5H,6H-cyclopenta[b]thiophen-2-yl}-1,3,4-oxadiazol-2-yl)-1-methyl-1,2-dihydropyridin-2-one (CTOM-DHP) that restores RAI uptake in 8505C ATC cells most likely through inhibiting both MAPK and PI3K/Akt signaling pathways. The efficacy of DN200434 and CTOM-DHP in restoring RAI uptake in corresponding ATC cell lines was further validated in nude mice carrying tumor xenografts. Of interest, MEK inhibitor PD-325901 and PI3K inhibitor GDC-0941 that inhibit MAPK and PI3K/Akt signaling pathway, respectively, in combination have been reported to synergistically promote tumor shrinkage with histology reversion in a genetically engineered double mutant BRAF^V600E and PIK3CA^H1047R mouse model of thyroid cancer with phenotypic progression to ATC in some areas (2). However, histology reversion with increased NIS expression was mainly caused by MEK inhibitor PD-32591 but not by PI3K inhibitor GDC-0941 in this ATC mouse model. For mouse model of DTC, MEK/BRAF inhibitors have been shown to enhance RAI uptake in genetically engineered thyroid cancer mouse models carrying BRAF^V600E mutation.

To this date, redifferentiation RAI therapy has not been investigated among ATC patients. However, there are a few studies with small number of DTC patients that have clinically assessed the concept of redifferentiation RAI therapy with short-term treatment of MEK and/or BRAF inhibitors. There are two reasons why redifferentiation therapy with MEK/BRAF inhibitors can be quickly translated into clinical trials for DTC patients. First, ∼85% of DTC patients have driver mutations leading to MAPK activation that results in NIS reduction, and RAI nonavid diseases in DTC patients are often associated with high output of MAPK signaling pathway. Second, BRAF/MEK inhibitors have already been investigated in patients with progressive RAIR disease for antitumor activity and, therefore, redifferentiation RAI therapy with shorter treatment duration of BRAF/MEK inhibitor could be readily implemented. In comparison, “patient selection” among ATC patients whose tumors are genetically complex and “treatment regimen optimization” for other aforementioned compounds for redifferentiation RAI therapy would require more efforts.

Ho et al. (3) studied the use of an MEK inhibitor, selumetinib, in 20 DTC patients with RAIR disease. Twelve out of 20 patients exhibited iodine avidity after 4 weeks of selumetinib treatment. The investigators treated 8 of these 12 patients, who achieved iodine uptake that allowed, through lesional dosimetry calculation, the delivery of a dose of ¹³¹I of 2000 cGy or more. Out of these 8 patients, 5 had partial response (PR) in their structural disease and 3 had stable disease (SD) as evaluated by imaging 6 months after the RAI. It is noted that four out of the five patients with a PR had NRAS mutated poorly DTC. Rothenberg et al. (4) evaluated the use of a BRAF inhibitor, dabrafenib, in 10 patients with papillary thyroid cancer who were considered RAIR. After 4 weeks of treatment, there was demonstration of iodine avidity in known structural disease in 6 (60%) of 10 patients. These 6 patients received RAI treatment and 2 had PR and 4 had SD at 3 months follow-up. Iravani et al. (5) showed that 3 of 3 patients with the BRAF^V600E mutation treated with BRAF/MEK inhibitors and 1 of 3 patients with an NRAS mutation treated with tramentinib demonstrated restoration of RAI uptake and proceeded to RAI therapy after thyroid hormone withdrawal. Among the 4 RAI-treated patients, 2 of 3 patients with the BRAF^V600E mutation and the 1 patient with an NRAS mutation had PR beyond 3 months. To date, it remains unknown why some DTC patients do not have restoration or enhanced RAI uptake after BRAF/MEK inhibitor therapy. It may be informative to compare tumor mutation landscape, tumor transcriptome profile, and germline variants between responsive versus nonresponsive patients as tumor driver mutation alone may not fully reflect tumor genetic makeup and the influence from tumor microenvironments.

The modulation of NIS expression/function appears to be tumor-context dependent, as indicated by selective responsiveness to BRAF/MEK inhibitors among patients with RAI nonavid disease with BRAF^V600E or NRAS ^Q61X tumor driver mutation(s). Fletcher et al. (6) recently reported that NIS cell surface trafficking can be compromised in thyroid cancer overexpressing valosin-containing protein (VCP) and that VCP inhibitors can reverse this process and thereby increase RAI uptake. Furthermore, Lakshmanan et al. (7) reported that RAI uptake increased by MEK or BRAF inhibitor in thyroid cells overexpressing BRAF^V600E oncogene was extensively reduced by TGF-β, a cytokine known to be secreted in the invasive fronts of thyroid cancers. The reduction of RAI uptake by TGF-β was mainly mediated by NIS reduction and could be reversed by apigenin, a plant-derived flavonoid. Moreover, PI3K inhibitor GDC-0941 greatly increases RAI uptake through decreasing RAI efflux. Among patients nonresponsive to MEK/BRAF inhibitor to increase RAI uptake in target lesions, further investigation on their tumor transcriptome is warranted. For patients with tumor in which VCP is overexpressed, addition of an FDA-approved VCP inhibitor ebastine or clotrimazole may be considered. For patients with tumor in which signaling outputs of TGF-β and/or PI3K are high, addition of apigenin and/or GDC-0941, respectively, may make a difference.

The variability in increased iodine avidity after redifferentiation therapy coupled with the mixed clinical responses to RAI therapy in the mentioned studies illustrates the need to investigate this treatment further. It is likely that this variable response reflects the complexity and heterogeneity of RAI nonavid lesions among patients and within the individual patient. Further delineation of tumor context including pertinent molecules and signaling pathways that modulate NIS-mediated RAI uptake and RAI retention is needed to provide better guidance for selecting the optimal therapeutic regimen to achieve maximum RAI uptake in target lesions. In addition, since the efficacy of RAI therapy depends not only on sufficient RAI delivery to metastatic lesions but also on lesions' radiosensitivity, the effect of this redifferentiation therapy may go beyond increased iodine delivery to these lesions.

MEK inhibitor has been reported to radiosensitize tumors derived from several tissue types to external beam radiation therapy. Robb et al. (8) recently reported that thyroid cancer cell lines carrying the BRAF^V600E mutation were associated with resistance to ionizing radiation, and the BRAF inhibitor, vemurafenib, selectively radiosensitized BRAF^V600E tumor cells by inhibiting DNA double-strand break repair. Vemurafenib in combination with radiotherapy resulted in marked and sustained regression of thyroid tumor xenografts carrying BRAF^V600E mutation. Our unpublished data also showed that MEK inhibitor trametinib increased RAI radiosensitivity in RAI-resistant human thyroid cancer cell lines expressing exogenous NIS. Thus, BRAF/MEK inhibitors likely not only enhance RAI delivery but also increase RAI radiosensitivity. To date, differential RAI radiosensitivity among lesions within individual patients or between patients has not been investigated. For a starting point, characterization of germline variants and tumor mutation landscape in patients who have macrometastases that are super-responsive to RAI treatment may help identify mechanisms underlying the differential RAI responsiveness among patients. Uncovering the mechanistic effectors underlying MEK inhibitor-enhanced radiosensitization and identifying gene expression and/or mutation classifiers associated with RAI resistance in thyroid cancer will allow us to maximize RAI responsiveness for thyroid cancer patients by enhancing both RAI delivery and RAI radiosensitivity.

There remain additional areas that warrant further investigation in the clinical application of this therapy. First, the optimal treatment duration for BRAF and/or MEK inhibitors in individual patients has not been vigorously investigated. To help define this individualized optimal treatment window, clinically identifiable markers that demonstrate a “readiness” to proceed to RAI therapy are needed, which may include monitoring serial changes in serum thyroglobulin level and/or the extent of RAI uptake shown by diagnostic ¹²³I or ¹²⁴I scan. Second, similar to the current clinical practice of RAI therapy, there is a disagreement regarding the role of dosimetry versus empiric dose in redifferentiation RAI therapy. While dosimetry aims at choosing the appropriate administered activity that will achieve a safe and effective treatment, there is still the assumption that all lesions among different patients have similar RAI radiosensitivity. Third, the best method of preparing these patients for treatment remains unclear. This includes the method of stimulation whether with recombinant thyrotropin or thyroid hormone withdrawal as well as the possible differences in assessing the extent of RAI avidity among lesions through diagnostic scans using ¹²³I, ¹³¹I, or ¹²⁴I (9). Fourth, these studies only treated patients with RAI if the diagnostic scan showed increased RAI avidity after the redifferentiation therapy and excluded patients from RAI therapy whose scans did not demonstrate increased iodine uptake. Such an approach might be excluding patients who could be responsive to RAI therapy with RAI-avid lesions that are only demonstrated in a posttherapeutic ¹³¹I scan. Finally, should redifferentiation therapy also apply to RAI naive high-risk patients, such as patients with gross extrathyroidal extension, advanced nodal disease of >3 cm, and/or distant metastasis? Among patients with lung metastasis, RAI nonavid disease is one of the strongest independent predictors for poor survival. Could redifferentiation RAI therapy in high-risk RAI naive patients prevent or prolong the time to progression to RAIR disease? It is possible that redifferentiation RAI therapy may reduce the recurrence rate and/or halt disease progression by increasing RAI delivery to target lesions. It may also minimize unwanted adverse effects by reducing RAI delivery to nontarget organs.

While there are only limited clinical data on redifferentiation therapy in DTC patients, there is none in patients with ATC. Treatment options for ATC include surgery, external beam radiation, chemotherapy, and in the case of BRAF^V600E mutated tumors, a combination of BRAF inhibitor dabrafenib and MEK inhibitor trametinib. Resistance to these treatments is frequent and development of new agents by exploring new ideas is needed. ATC is a genetically complex malignancy reflecting its origin from several distinct DTC subtypes, with oncogenic mutations in the MAPK pathway (BRAF, RAS) as the key drivers of DTC with additional genetic alterations in the PI3K pathway, TP53, and the TERT promoter for anaplastic transformation (10). Oh et al. (1) and others have demonstrated redifferentiation by various compounds in human ATC cell lines and in a preclinical genetically engineered mouse model (2). Clinically, however, there are several issues that are unique to ATC that should be emphasized. RAI therapy cannot be used for de novo ATC patients with unresectable local disease who often have a sizable residual normal thyroid tissue that precludes the use of RAI. In addition, since the response to RAI therapy is usually gradual, the efficacy of redifferentiation RAI therapy may not be achieved clinically for ATC patients with rapidly progressive disease. As a matter of fact, studies on redifferentiation RAI therapy in RAIR DTC patients excluded those with symptomatic or rapidly progressive disease.

In view of the fact that the BRAF^V600E mutation occurs in 41% of ATC patients (10) and the high objective response rate of BRAF/MEK inhibitors among ATC patients carrying the BRAF^V600E mutation, it may be worthwhile to start the exploration of the redifferentiation RAI therapy in ATC in these patients. In particular, those BRAF^V600E mutated ATC patients whose tumors demonstrate redifferentiation after treatment with BRAF/MEK inhibitors as demonstrated by a diagnostic RAI scan may benefit by adding RAI treatment. For patients with ATC carrying both BRAF^V600E mutation and additional genetic alterations in the PI3K pathway, addition of the FDA-approved PI3K inhibitor GDC-0941 should be explored not only to further increase antitumor activity but also to further increase RAI uptake through decreasing RAI efflux. Finally, some of the lessons learned from the optimization of redifferentiation RAI therapy for DTC patients as already discussed may be applicable to dabrafenib- and trametinib-treated ATC patients carrying tumors with BRAF^V600E mutation as well.

In summary, RAI therapy is proven to be effective in improving the clinical outcome for high-risk DTC patients with RAI-avid disease. Short-term treatment with BRAF/MEK inhibitors has the potential to improve the efficacy of RAI therapy by increasing RAI delivery into target lesions and by increasing lesions' radiosensitivity to RAI therapy. With several emerging redifferentiating agents in the pipeline, personalized redifferentiation RAI therapy for selected DTC and ATC patients guided by tumor mutation landscape, germline variants, serum thyroglobulin doubling time, and radiomics derived from RAI-SPECT, FDG-PET, and CT images, may be possible. Prospective multicenter trials will be needed to evaluate the efficacy and safety of redifferentiation strategies in a larger cohort of patients.