Introduction

Thyroid cancer comprises a diverse range of histologic and clinical entities encompassing well-differentiated papillary and follicular thyroid cancers (PTC and FTC, respectively), medullary thyroid cancer (MTC), poorly differentiated thyroid cancer (PDTC), and anaplastic thyroid cancer (ATC). There were an estimated 44,000 cases of thyroid cancer in the United States in 2010, and the incidence is rising.¹ For most patients, thyroid cancer is effectively treated by surgery. Advanced, differentiated thyroid cancer (DTC), which concentrates iodine, may benefit from treatment with radioiodine.² Despite significant surgical and endocrine advances, challenges remain in the treatment of patients with progressive, metastatic thyroid cancer. Unfortunately, cytotoxic chemotherapy has a limited record of success in treating patients with advanced disease.³ Recently we have witnessed an explosion of clinical trials demonstrating efficacy for novel, targeted therapies in thyroid cancer. The goals of this chapter are to introduce the key concepts in the development of targeted cancer drugs, highlight the relevant targets that are known for thyroid cancer, review the published literature on targeted agents for thyroid cancer, and outline the challenges in ongoing drug development.

Drug Development

Clinical development for cancer therapeutics follows an evolving process of drug development.⁴ New targets are identified through basic science investigation involving cell biology, cancer genomics, and genetic models. Assays have been developed to read out a particular pathway or protein activity. Chemical screens are conducted to identify novel chemical structures that produce a desired biochemical effect. Compounds are optimized to enhance activity, bioavailability, and pharmacokinetic parameters. Eventually these compounds are tested in animals to study pharmacodynamics and toxicity. Promising candidates are then moved on to human studies, phase 1 studies. In cancer drug development, phase 1 trials are generally performed on small numbers of subjects, usually with a range of malignancies, to establish the maximum tolerated dose (MTD) and monitor for adverse effects. Once the MTD is established, phase 2 trials in specific diseases are performed. Phase 2 trials have typically been conducted as single-arm, nonrandomized trials in particular diseases. Typically, such phase 2 trials require a sample size of 20 to 50 patients. In recent years, we have seen the advent of new trial designs, including genotype directed phase 2 studies and randomized phase 2 studies.^5,6 Drugs that show significant activity in phase 2, and appear both safe and tolerable, may progress to phase 3. Phase 3 studies are randomized studies in which patients receive the study drug, or a placebo, or the standard of care. In oncology, the primary outcome of interest is generally either overall survival (OS) or progression-free survival (PFS). Such trials may include a variety of secondary outcomes including safety, quality of life, and correlative laboratory and imaging studies. Phase 3 trials usually require multicenter collaboration, sample sizes of several hundred patients and significantly more resources in terms of data monitoring and oversight.

A core set of principles has guided the successful development of targeted agents in leukemia and solid tumors since the 1990s. Having a mechanistic insight into the pathways that control tumorigenic growth is critical. The most successful anticancer drugs have been developed for diseases in which the molecular genetics of the underlying disease is relatively well understood. The ideal drug-target combination has an assay that can be used to follow the ability of the drug to “hit” the target. This may be in the form of inhibiting the activity of a kinase to phosphorylate a substrate, for example. Novel drugs need to be tolerable at their effective doses. This is often a major challenge with small molecules, which tend to have pleiotropic effects on human physiology. Targeted drugs often have subsets of patients in which they are most effective. This often means that the drug works best in patients of a particular genotype, such as trastuzumab in HER2 positive breast cancer patients. The ability to predict which patients with a particular histology will respond to a targeted agent is a major challenge.

Thyroid Cancer Genetics

The common genetic lesions in thyroid cancer are well described⁷ (see Chapter 17, Molecular Pathogenesis of Thyroid Neoplasia). PTC and FTC commonly harbor mutations, which lead to activation of the MAPK pathway. For PTC, approximately 40% to 60% harbor the somatic mutation in BRAF, BRAF-V600E, which is also common in malignant melanoma.⁸ Ras mutations occur in 10% to 20% of PTC and are mutually exclusive with BRAF mutations.⁹ Ras mutations are more common in FTC, where they can be seen in 40% to 50% of tumors. Follicular thyroid adenomas and carcinomas uniquely harbor PPAR-γ rearrangements in 20% to 60% of cases.¹⁰ Anaplastic thyroid cancers are the least well understood genetically.¹¹ ATCs commonly have TP53 mutations, Ras mutations, and may have mutations in CTNNB1. Medullary thyroid cancer, a C-cell derived neuroendocrine malignancy, may be sporadic or hereditary and is associated with activating mutations in RET, a receptor tyrosine kinase in 30% to 50% of sporadic MTC and 80% to 90% of hereditary MTC.¹² RET gene rearrangements are also seen in papillary thyroid cancer, where RET may undergo translocation with several other genes.

Novel Agents in Thyroid Cancer

Here we review the published reports of clinical trials testing targeted therapies for advanced thyroid cancer. We summarize the clinical efficacy data for 16 drugs that have been tested in differentiated (DTC), medullary (MTC), or anaplastic thyroid cancer (ATC).

Vandetanib (ZD-6474)

Vandetanib is a small molecule inhibitor of VEGF, PDGF, and EGF receptors and RET.¹³ Vandetanib was evaluated at a starting dose of 300 mg/d in a single-arm, phase 2 study of patients with either locally advanced, unresectable, or metastatic hereditary MTC.¹⁴ Progressive disease was not a criterion for study entry. Partial responses were observed in 6/30 patients (20%); 16/30 patients (53%) had stable disease greater than 24 weeks, and 6/30 patients (20%) were stable between 8 weeks and 24 weeks. There was 1/30 patient (3%) with progressive disease. The median PFS was 27.9 months (95% CI, 19.4 to not estimable). The median duration of therapy was 18.8 months. A majority of treated patients had a decline in serum calcitonin and CEA, but there was no definite relationship between the tumor markers and response by RECIST. The most common adverse events were diarrhea, fatigue, rash, and nausea. Other significant adverse events included QTc prolongation and hypertension. Twenty-four patients required dose reductions because of adverse events.

Based on the efficacy demonstrated in the multicenter, phase 2 study, vandetanib has been subjected to an international, randomized, placebo-controlled, phase 3 clinical study, the ZETA trial.¹⁵ In this study, patients with hereditary or sporadic MTC were required to have measurable, progressive disease within 14 months prior to study entry. The primary end point was progression-free survival, and crossover was permitted. This study did not require subjects to have documented progressive disease prior to study entry. The trial enrolled 331 patients, 90% of whom had sporadic MTC and 95% of whom had metastatic disease. Vandetinib was associated with a statistically significant improvement in PFS compared to placebo (HR, 0.45; 95% CI, 0.30 to 0.69). Significant side effects included diarrhea, rash, nausea, hypertension, and headache.

If the Food and Drug Administration (FDA) approves vandetanib on the basis of the data from the ZETA trial, which seems likely, it will be the first targeted agent approved for thyroid cancer and only the second drug FDA approved for thyroid cancer. This raises several important questions, notably whether subsequent experimental drugs will need to be compared to vandetanib directly and whether improvement in PFS will be used as a criterion for approval, or whether improvement in overall survival will be required for FDA approval.

XL-184

XL-184 is a small molecule inhibitor of RET, MET, VEGF receptors, and KIT.¹⁶ XL-184 is under active clinical investigation for several malignancies, including MTC. A multi-institution, phase 1 study of XL-184 across a range of malignancies has recently been reported.¹⁷ In this study, 85 patients were enrolled in a 3 + 3 dose escalation design; 37 of these patients had MTC, and the dose of XL-184 ranged from 75 mg to 175 mg orally daily. The authors reported that 10/34 patients (29%) had confirmed a partial response, and 7 additional patients had unconfirmed responses. Fifteen patients (41%) had stable disease of at least 6 months’ duration; however, progressive disease was not a criterion for eligibility, making it difficult to determine the significance of stable disease during treatment with the study drug. Of the responders to XL-184, several patients had previously progressed on other tyrosine kinase inhibitors. Activating mutations in RET were found in 81% of MTC patients, including M918T (n = 14) and C634Y (n = 1). The majority of MTC patients had sporadic disease.

Based on these findings, an international, double-blinded, placebo-controlled, phase 3 study of XL-184 (175 mg orally daily) in MTC was initiated in which overall survival is the primary end point (NCI NCT00704730). In this ongoing study, patients with hereditary or sporadic MTC are required to have measurable, progressive disease within 14 months prior to study entry. As with the vandetanib study, the primary end point of this trial is progression-free survival. This trial does not, however, allow crossover or unblinding at disease progression, which has generated some controversy. Overall survival is a secondary end point of the study.

Motesanib (AMG-706)

Motesanib diphosphate, an oral inhibitor of VEGF receptors 1, 2, and 3, PDGF receptor, RET, and c-KIT, has been subjected to a multicenter, phase 2 trial in DTC in which 93 patients with locally advanced or metastatic radioiodine resistant DTC were enrolled.¹⁸ Patients were required to have progressive disease, by investigator assessment, within 6 months prior to study entry. Thirty-two patients completed the planned 48-week course of treatment with motesanib at 125 mg once daily. Responses were assessed by a centralized, independent review. Thirteen patients (14%) experienced a partial response by RECIST, 62 patients (67%) had stable disease, 7 patients (8%) had progressive disease, and 11 patients were not assessable (12%). The median duration of response was estimated to be 32 weeks among responders. There was no association between histologic subtype or BRAF-V600E mutation status and response. Treatment was associated with grade 3 toxicities in 51 (55%) patients. There were 2 treatment related deaths resulting from pulmonary hemorrhage in patients with progressive disease. There were 5 (5%) patients with cholecystitis and 2 (2%) patients with cardiac disorders. Hypothyroidism or an increase in thyroid-stimulating hormone (TSH) occurred in 22% of patients.

Motesanib diphosphate was also studied in a parallel single-arm phase 2 study of patients with progressive, or symptomatic MTC.¹⁹ Ninety-one patients with sporadic (84%) or hereditary (16%) MTC were enrolled. Responses were assessed by a centralized, independent review with confirmation performed 4 weeks after the initial study. Two patients (2%) experienced a partial response by RECIST, 74 patients (81%) had stable disease, 7 (8%) patients had progressive disease, and 8 (9%) patients were not assessable. Patients with diarrhea at baseline experienced transient improvement in diarrhea symptoms, which was not sustained. Eight patients (9%) experienced gallbladder toxicity, 3 experienced acute cholecystitis. Thirty-seven patients (41%) experienced elevated TSH levels compared with baseline or hypothyroidism.

Sorafenib (bay 43-9006)

Sorafenib is a small molecule inhibitor of multiple kinases including BRAF, CRAF, VEGF receptors, PDGF receptors, KIT, FLT3, and RET. Sorafenib is FDA approved for the treatment of unresectable hepatocellular carcinoma and renal cell carcinoma. Three phase 2 studies have been reported using sorafenib at dose of 400 mg orally twice daily in thyroid cancer. Two of these have been in DTC and one in MTC.

Sorafenib was studied in a single-arm, phase 2 study of patients with thyroid cancer, including medullary and anaplastic subtypes.²⁰ Patients were required to have progressive disease in the year prior to enrolling in the study. RECIST was used to assess 30/31 patients: partial responses occurred in 7/30 patients (23%), stable disease was seen in 16/30 patients (53%), and there were 2/30 patients (3%) with progressive disease; 2/2 patients with anaplastic thyroid cancer had progressive disease on study. One patient with medullary thyroid cancer had stable disease. There were no significant differences in responses by histologic subtype. The median progression-free survival was 79 weeks. The median duration of treatment was 27 weeks; 47% of patients required dose reductions, and 63% of patients had a break from treatment due to adverse events. Although serum thyroglobulin levels declined in a majority of patients, there was no relationship between Tg levels and the degree or duration of response.

In a second study of sorafenib in thyroid cancer, 58 patients with thyroid cancer were enrolled in a single-arm study of sorafenib.²¹ The majority of patients had PTC (73%) and were naive to cytotoxic chemotherapy (80%). RECIST was used to assess 50/58 patients: 6/56 (11%) had partial responses, 35/56 (63%) had stable disease, and 9/56 (16%) had progressive disease. There were 19/50 (38%) patients with durable stable disease greater than 6 months’ duration. Dose reductions were performed in 52% of patients. The median PFS was 16 months (95% CI, 8 to 27.5 months). There were no partial responses among the patients with Hurthle cell or follicular cell cancers (n = 11 total), although 9/11 (82%) had stable disease. Although serum Tg levels declined in many patients with responses, there was no apparent relationship between Tg levels and the degree or duration of response. A subset of patients underwent imaging to assess tumor perfusion with contrast-enhanced MRI, which revealed decreased tumor perfusion on therapy, consistent with an anti-angiogenic response. A subset of patients underwent imaging using fluorodeoxyglucose positron emission (FDG-PET) scans; there was no correlation between PET response and objective response. A majority of the patients with PTC had activating BRAF mutations (17 of 22, 77%); these patients experienced a range of responses from PR to PD on therapy, suggesting that factors beyond BRAF mutation status are correlated with a response to sorafenib.

Sorafenib has also been explored as a therapy for patients with sporadic and hereditary MTC in a single-arm, phase 2 clinical study that enrolled 21 patients.²² RECIST was used to evaluate 20/21 patients: 2/21 (10%) had partial responses, and 18/21 (86%) had stable disease. In the cohort of patients with sporadic MTC, 8/15 (50%) had durable stable disease for greater than 15 months. In the cohort of patients with sporadic disease (n = 15) the median PFS was 17.9 months (95% CI, 8 to not estimable). RET genotyping was performed on 12/16 patients with sporadic MTC, and 10 of 12 tested patients (83%) were positive for a RET mutation, 9 of whom had M918T. There were partial responses noted for individual patients harboring both C634Y and M918T RET mutant tumors. Decreases in calcitonin and CEA were noted, but neither correlated with response by RECIST.