Mitogen-Activated Protein Kinase Pathway Inhibition for Redifferentiation of Radioiodine Refractory Differentiated Thyroid Cancer: An Evolving Protocol

Patients

Consecutive patients who underwent therapy with an RAI redifferentiation protocol at the Peter MacCallum Cancer Centre were reviewed retrospectively. Inclusion criteria were as follows: (i) histologically proven metastatic DTC; (ii) sufficient bone marrow (hemoglobin equal or above 90 g/L, neutrophil above 1.5 × 10⁹/L, and platelet above 100 × 10⁹/L) and renal function (creatinine clearance at least 40 mL/min); (iii) ECOG score 2 or below; and (iv) presence of a BRAF^V600E or NRAS mutation. RAIR DTC was defined as follows: (i) minimal or no RAI avidity in some or all measurable sites of disease on appropriately performed RAI scans and/or (ii) imaging or biochemical progression despite RAI treatment within the past 12 months. For assessment of RAI avidity, either post-therapy ¹³¹I single-photon emission computed tomography (SPECT)/computed tomography (CT) or pretherapy ¹²³I SPECT/CT or ¹²⁴I positron emission tomography (PET)/CT was allowed. Conditions for appropriately conducted RAI scans included adequate TSH stimulation (>30 mU/L) and no exogenous iodine exposure in the preceding three months.

Each therapy was approved individually by the institutional tumor board under compassionate use criteria (Special Access Scheme—Category A). This study was approved by the Peter MacCallum Cancer Centre Ethics Committee (approval number 19/14R).

Mutation analysis

Amplicon-based next-generation sequencing was performed on DNA extracted from macro-dissected, formalin-fixed, paraffin-embedded tumor tissue followed by targeted sequencing on an Illumina MiSeq, 2 × 150 bp. Aligned reads and called variants were analyzed in PathOS (23). Variants were described according to the Human Genome Variation Society (HGVS) nomenclature version 2 (24).

Redifferentiation schedule

The redifferentiation protocol is detailed as below and summarized in Figure 1.

OPEN IN VIEWER

FIG. 1.

Schematic demonstration of redifferentiation protocol. CT, computed tomography; FDG, fluoro-2-deoxyglucose; PET, positron emission tomography.

Patient preparation and MAPK inhibition regimen

All included patients underwent four weeks of THW unless this was clinically contraindicated (n = 1). TSH stimulation was determined by blood sampling at approximately four weeks, and a TSH level >30 mU/L was considered satisfactory. If satisfactory TSH stimulation was not achieved with THW only (n = 1), and in the patient with THW contraindication, two injections of 0.9 mg rhTSH were administered intramuscularly 24 hours apart before RAI imaging and therapy.

The patients with NRAS mutations were treated with the MEK inhibitor trametinib (2 mg daily) for four weeks. The patients with a BRAF^V600E mutation were treated with a combination therapy, including a BRAF inhibitor and an MEK inhibitor: dabrafenib (150 mg twice daily) and trametinib (2 mg daily), or vemurafenib (960 mg twice daily) and cobimetinib (60 mg daily) for four weeks. The choice of the agents used for the combination therapy was based on the access to these medications at the time of prescription. Treatment began four weeks from the expected date of RAI treatment, concurrent with THW. The patients were reviewed weekly and adverse events were recorded based on the Common Terminology Criteria for Adverse Events (CTCAE) version 4.

¹²⁴I imaging protocol

Following confirmation of satisfactory TSH stimulation, two time point ¹²⁴I PET/CT imagings were used to assess restoration of the iodine concentration at metastatic sites. For this, 1–2 MBq/kg body weight of ¹²⁴I was injected intravenously (mean 120 MBq, range 80–210 MBq). Sequential CT and PET acquisition was then performed on an integrated PET/CT device (GE Discovery PET/CT 690 or GE Discovery 710; GE Healthcare, Milwaukee, WI). The CT portion of the study was acquired in the craniocaudal direction encompassing vertex to midthigh with the patient supine, with a slice thickness of 5 mm, an increment of 1.5 mm, 140 keV, 220 mAs, and 0.6 pitch. PET acquisition was subsequently acquired in the craniocaudal direction to minimize misregistration in the neck and chest as the primary regions of interest. Three-dimensional acquisition was performed with emission data corrected for random, scatter, and decay. Reconstruction was conducted with an ordered subset expectation maximization algorithm with two iterations/eight subsets. Gaussian filtering with 7 mm at full-width at half-maximum kernel was applied to mitigate image noise in delayed acquisitions.

Two time point pretreatment ¹²⁴I PET/CT images were used to predict radiation dosimetry of subsequent ¹³¹I therapy by providing a measure of regional pharmacokinetics. By modeling the uptake and clearance time of ¹²⁴I, an estimate of regional decays per GBq injected (rad per mCi) was calculated considering different physical half-life of ¹³¹I. Disintegrations were converted into an estimate of radiation-absorbed dose based on the OLINDA (v1.1) sphere model dose factors (25). To achieve a dependable estimate of pharmacokinetics, the first ¹²⁴I images were captured at ∼24–48 hours and the second time point was delayed to 72–96 hours. Local tracer uptake in the tumor and bone marrow was calculated. To account for partial volume effect in case of small tumor size, an expanded contour, which captured all of the local activity as well as a true physical contour based on the CT margins (where visible), was utilized to reconcentrate the activity to an accurate tumor volume.

F-18-labeled fluoro-2-deoxyglucose imaging protocol

F-18-labeled fluoro-2-deoxyglucose (FDG) PET/CT scans were performed at baseline, before the commencement of MAPK inhibitor treatment and at approximately three to four weeks following MAPK inhibitor, just before ¹²⁴I injection. FDG PET/CT was used to assess metabolic activity at the metastatic sites and monitor response to treatment.

Baseline and post-MAPK inhibitor treatment FDG PET/CT scans were performed on the abovementioned scanners. Patients fasted for at least six hours, had blood glucose measurements before administration of 3–4 MBq/kg of FDG, and acquisition of PET images was commenced approximately one hour later.

RAI treatment and follow-up

Restoration of RAI concentration following MAPK inhibitor was defined visually by new RAI avidity at metastatic sites. With increasing experience, an estimated radiation-absorbed dose derived from ¹²⁴I two time point imagings was incorporated in the decision-making of suitability for RAI treatment. ¹²⁴I-based dosimetry estimates were regarded as a proof-of-concept approach and the information was used as a general rather than strict guide for choosing the administered activity in conjunction with other aspects of the disease, including organs involved and the volume of metastases. All patients underwent a post-treatment whole-body planar and SPECT/CT of the metastatic region at approximately seven days after administration of RAI therapy.

The patients had follow-up at three months to assess imaging and biochemical response to treatment. Imaging response was assessed based on response evaluation criteria in solid tumors version 1.1 (RECIST v1.1) (26). Biochemical response was assessed by measuring thyroglobulin levels. Ongoing follow-up (median 16 months, range 5–40) until the date of article submission was recorded.

Optimizing MAPK pathway inhibition

The MAPK signaling cascade is a critical pathway for human cancer cell survival, differentiation, dissemination, and resistance to drug therapy (27). The MAPK pathway is activated by upstream genomic events and/or activation of multiple signaling events where information converges at this nodal pathway point. This pathway is highly regulated in normal cells by phosphatases and bidirectional communication with other pathways, such as the AKT/mTOR pathway, which promotes anabolism (27). In cancer cells, this pathway is dysregulated, and its impact predominantly depends on the intensity of the signal and the context or tissue in which the signal is aberrantly activated (28). Understanding the differential nature of activation of the MAPK pathway in each tumor type is critical in developing single and combination regimens, as different tumors have unique mechanisms of primary and secondary signaling and subsequent sensitivity to drugs (28).

Phase I and II studies using different BRAF inhibitors (including dabrafenib and vemurafenib) have shown antitumor activity in patients with progressive BRAF^V600E mutant PTC refractory to RAI therapy (29,30). In patients with BRAF^V600E-mutated metastatic melanoma, combination therapy with a BRAF inhibitor and an MEK inhibitor significantly improved median OS and PFS compared with a BRAF inhibitor alone (31). A randomized phase II study comparing the BRAF inhibitor dabrafenib with and without the MEK inhibitor trametinib demonstrated similar responses but longer PFS with the combination (ClinicalTrials.gov Identifier: NCT01723202). In the study by Rothenberg et al., 10 patients with tumors harboring the BRAF^V600E mutation and RAIR PTC were treated with a single-agent BRAF inhibitor (dabrafenib), with new or enhanced iodine uptake in six patients (18). All these six patients underwent RAI therapy with a partial response based on imaging in two patients six months following this treatment. However, given potentially more potent blockade of the MAPK pathway in the BRAF^V600E mutant cohort, we opted to use a combination therapy with a BRAF and an MEK inhibitor. To the best of our knowledge, this is the first report of using a combination therapy with a BRAF and an MEK inhibitor in conjunction with RAI therapy in the literature. In this series, all three patients with a BRAF^V600E mutation regained sufficient iodine avidity to undergo RAI therapy with satisfactory iodine retention on post-treatment scans, and two of these patients achieved a partial imaging response three months post-treatment.

Although clinical trials of MEK inhibitor monotherapy using selumetinib in NRAS mutant patients have not yet demonstrated robust response rates (32,33), highly favorable results have been reported in a redifferentiation protocol (17). In the study by Ho et al., all five patients with an NRAS mutation treated with selumetinib had successful redifferentiation and underwent RAI therapy, while only one of nine patients with a BRAF^V600E mutation reaching the predefined threshold of iodine uptake restoration to undergo RAI therapy (17). Different levels of MEK inhibitor sensitivity were observed in tumor cell lines with BRAF or RAS mutations (34). Furthermore, in preclinical models, differences in the potency of MEK inhibitors have been observed. For instance, in ovarian and lung cancer cell line models, trametinib has shown more effective inhibition of cellular proliferation than other MEK inhibitors, including selumetinib (34,35). Given a potentially higher potency of trametinib and a lack of access to selumetinib in Australia, we elected to use trametinib in RAS mutant patients. Only one of three NRAS mutant patients achieved sufficient iodine avidity following single-agent MEK inhibitor therapy and was treated with RAI with satisfactory iodine retention. This patient maintained a durable response up to 30 months after therapy with the redifferentiation protocol (Fig. 2) despite previously progressing disease. The optimal single or combination small-molecule inhibitors of the MAPK pathway remain unknown and require further study. Table 4 summarizes the key features of the studies by Ho et al. (17) and Rothenberg et al. (18) compared with the current cohort.

Although in prior studies and in our cohort, at least four weeks of MAPK inhibitor therapy has been administered, the duration of redifferentiation therapies remains unclear. In thyroid cancer cell line models, withdrawal of BRAF oncogenic induction has resulted in an increase in apoptosis within 72 hours, which was no longer detectable by two weeks (16). It was also noted that apoptosis was restricted within two weeks due to the restoration of the follicular structure, thyroid-specific gene expression, and thyroid function. Although therapy with the small-molecule MAPK inhibitor used in the study showed a more attenuated recovery of expression of thyroid-specific genes compared with the withdrawal of oncogenic induction, it is conceivable that even a shorter duration treatment would potentially achieve optimal redifferentiation with optimal blockade of this pathway. Shorter duration of treatment with MAPK inhibitors may also minimize the toxicity associated with these medications, while maximizing the long-term therapeutic effect of RAI therapy.

Appropriate clinical setting and patient selection

Since 2014, two novel systemic treatments have attracted attention and gained regulatory approval in the treatment RAIR DTC, sorafenib and lenvatinib. These two mTKI were approved for use in the United States and the European Union for patients with metastatic RAIR DTC based on phase III randomized placebo-controlled clinical trials, demonstrating improvement in PFS (4,5). Despite promising results from these trials, the OS advantage remains to be determined without evidence of improvement in QOL. In fact, these agents are associated with numerous adverse events likely to negatively impact QOL, necessitating dosage reductions in nearly two-thirds of patients, treatment discontinuation in up to 20% of patients, and therapy-related death in 1.5–2% (4,5,36). Although the crossover trial designs effectively contaminated OS data, the absence of a proven effect of mTKI on OS in these trials may indicate that a delay in initiation of these treatments until observation of RECIST progression among initially placebo-treated patients did not adversely affect OS.

In the absence of definitive data on OS and QOL, an evidence-based recommendation with expert consensus suggests considerable restraint in the use of mTKI. Based on general expert consensus guidelines, consideration of treatment with mTKI should be reserved for structurally progressive, symptomatic, and/or imminently threatening DTC (in which disease progression is expected to require intervention and/or to produce morbidity or mortality in less than six months), that is, RAIR, and not amenable to satisfactory control using directed approaches such as surgery or radiotherapy (37,38).

Therefore, it is important to optimally apply all approaches to therapeutic RAI before embarking on these treatments. Patients with specific therapeutically targetable tumor alterations, such as BRAF or RAS mutations in the MAPK pathway, can also be considered for short-term therapies specifically targeting these alterations to restore iodine avidity. However, given the indolent nature of the metastatic disease in most patients, therapeutic interventions should not be considered for patients with stable, asymptomatic metastatic disease unless agents with a significant likelihood of leading to complete remission, prolongation of survival, or biologic impact such as redifferentiation are available. Compared with conventional chemotherapy or novel agents, the ability to induce iodine uptake in RAIR metastatic DTC has the potential advantage of requiring only a short course of targeted therapy to maximize the potential long-term therapeutic effect from RAI therapy.

In this cohort, we recruited patients with progressive metastatic disease despite treatment with RAI within the previous 12 months. We also treated patients with de novo non-RAI avid metastatic disease to modify the biology of the tumors and render them susceptible to RAI therapy. One patient (patient 6) had never been treated with RAI but had iodine avid residual disease in the neck and iodine-negative distant metastatic disease. While utilization of the redifferentiation protocol maintained the treatment efficacy in the neck, it also restored iodine avidity in the distant metastatic disease enabling all sites of disease to be effectively targeted by RAI therapy. Future trials, however, are needed to clarify the long-term efficacy of this treatment strategy.

Role of dosimetry

The dosimetry approach is not currently used in many institutions, although most available guidelines do mention its feasibility (39,40). The American Thyroid Association (ATA) guidelines classify dosimetry-guided RAI-administered activity to lung and bone metastases as a weak recommendation with low-quality evidence (38). The purpose of dosimetry in these contexts is to limit the whole-body retention to 2.1 GBq (80 mCi) at 48 hours and 2 Gy to the bone marrow. The usefulness of dosimetry should not be assessed only according to patient survival or therapeutic efficacy; the possible avoidance of futile RAI treatments and better characterization of RAIR disease should also be considered. This is especially of critical importance when the degree of restoration of iodine avidity is assessed following redifferentiation therapy with MAPK inhibition.

Ho et al. used ¹²⁴I PET/CT lesional dosimetry to predict which patients were most likely to respond to RAI therapy (17). Based on ¹²⁴I PET/CT dosimetry following redifferentiation therapy, if one or more lesions could be treated with a dose of 20 Gy or more with an ¹³¹I-administered activity of 8.1 GBq (300 mCi), the patient was eligible for RAI therapy. The eligible patients underwent a whole-body and blood standard ¹³¹I dosimetry protocol to determine maximum tolerable activity. Following redifferentiation therapy with selumetinib, 12 of 20 patients had increased iodine uptake, but only 8 patients reached the predefined threshold for treatment. Of those eight patients, six achieved imaging partial response. Rothenberg et al. used ¹³¹I for assessment of restoration of iodine avidity, and any new or increased uptake at the sites of disease were considered suitable for RAI therapy (18). In this study, 6 of 10 patients were treated with RAI with two patients achieving a partial response.

¹²⁴I PET dosimetry is a promising technique for more precise quantitation of RAI uptake (41). The potential advantages of the ¹²⁴I PET/CT over ¹³¹I whole-body scanning include inherent quantitation of PET imaging, greater sensitivity for the detection of small lesions even at several days postadministration, and the ability to predict radiation dosimetry. In this small sample size, we have described a dosimetry approach based on two time point ¹²⁴I PET/CT, which may obviate the need for further pretreatment ¹³¹I dosimetry, hence simplifying this protocol. The question for the future is how to include dosimetry into the patient management framework. A larger sample size using prospective dosimetry is needed to validate this ¹²⁴I dosimetry approach in metastatic DTC.

THW versus rhTSH

No randomized trial comparing THW- to rhTSH-mediated therapy for the treatment of distant metastatic disease has been reported. Importantly, however, rhTSH has not been approved for use in patients with metastatic disease and there is increasing evidence indicating an improvement of the therapeutic index of RAI therapy by THW. Studies have shown that ¹²⁴I PET using rhTSH stimulation does not predict the uptake on ¹³¹I post-therapy scans at sites of metastatic disease following THW (20), but is highly accurate when THW is used for both scans (21). A small intrapatient ¹²⁴I dosimetry comparison also demonstrated higher tumor-to-dose-limiting-organ (bone marrow)-absorbed dose ratio using THW than rhTSH-stimulated administrations (22). One small comparative study showed the median half-life and cumulative activity in tumor tissue was higher in THW than rhTSH-mediated RAI treatment (42).

In contrast to both prior studies using redifferentiation therapy in which rhTSH-mediated RAI therapy was used (17,18), our protocol mandated THW unless it was contraindicated due to medical comorbidities. To the best of our knowledge, this is the first study demonstrating the feasibility of THW in this treatment strategy. THW also pragmatically avoids the expense and logistical challenge of rhTSH stimulation necessary for both ¹²⁴I PET dosimetry and subsequent ¹³¹I therapy.

Predictive markers of response beyond genotype

Baseline metabolic activity and metabolic response to treatment measured by FDG PET have prognostic implication in many cancers, including DTC (43,44). Consequently, the prognostic value of FDG PET as a real-time imaging biomarker in a patient-based rather than lesion-based scale should be used at diagnosis of metastatic DTC and may guide the intensity of therapy aggressiveness.

In BRAF mutant melanoma, metabolic changes early after BRAF and MEK inhibition can be a useful prognostic indicator of response to treatment (45) but are not necessarily predictive of structural imaging response (46).

In our small cohort of three patients with a BRAF^V600E mutation, we have observed a significant metabolic response following four weeks of combination BRAF and MEK inhibition (Fig. 2 and Supplementary Fig. S1). This also translated into the successful restoration of iodine uptake in all three patients, which rendered them suitable to RAI therapy. Surprisingly, two patients with RAS mutations had an increase in metabolic activity following MEK inhibition (Table 4 and Supplementary Fig. S2). This occurred despite a structural response in one of these patients. We hypothesize this may relate to bidirectional communication and deviation from the MAPK pathway to the AKT/mTOR pathway with a resultant increased glycolytic metabolic activity. THW with increased TSH stimulation may also contribute to this phenomenon. However, our preliminary data indicate that future trials may need to incorporate FDG PET in the selection and response assessment of the patients undergoing a redifferentiation protocol.

Limitations

It should be noted that the overall small sample size, retrospective nature of the study, and the short follow-up in some patients preclude further statistical analysis and recommendations regarding the optimal redifferentiation protocol. The aim of this study, however, was to demonstrate the feasibility of utilizing different genotype-selected regimens of MAPK inhibitors, including combination therapy utilizing THW stimulation, and to provide further evidence of potential efficacy in selected patients. Importantly, given concerns regarding the use of rhTSH in metastatic disease, this is the first published series of thyroid redifferentiation therapy using THW TSH stimulation. There is variability in the protocol used among the patients. This is a novel and complex treatment strategy with several uncertainties in terms of optimal protocol. Our protocol did evolve over time with variable access to combination BRAF/MEK inhibitor therapies and increasing experience in treating these patients. We have also demonstrated the feasibility of incorporating ¹²⁴I dosimetry in this protocol. Nonetheless, prospective multicenter trials are needed to further assess the efficacy and safety of this treatment strategy in a larger cohort of patients.