Medical management of metastatic medullary thyroid cancer

Authors


Abstract

Medullary thyroid cancer (MTC) is an aggressive form of thyroid cancer that occurs in both heritable and sporadic forms. Discovery that mutations in the rearranged during transfection (RET) proto-oncogene predispose to familial cases of this disease has allowed for presymptomatic identification of gene carriers and prophylactic surgery to improve the prognosis of these patients. A significant number of patients with the sporadic type of MTC and even those with familial disease still present with lymph node or distant metastases, making surgical cure difficult. Over the past several decades, many different types of therapy for metastatic disease have been attempted with limited success. Improved understanding of the molecular defects and pathways involved in both familial and sporadic MTC has resulted in new hope for these patients with the development of drugs targeting the specific alterations responsible. This new era of targeted therapy with kinase inhibitors represents a significant step forward from previous trials of chemotherapy, radiotherapy, and hormone therapy. Although much progress has been made, additional agents and strategies are needed to achieve durable, long-term responses in patients with metastatic MTC. This article reviews the history and results of medical management for metastatic MTC from the early 1970s up until the present day. Cancer 2014;120:3287–3301. © 2014 American Cancer Society.

INTRODUCTION

Medullary thyroid cancer (MTC) comprises 5% to 10% of all thyroid cancers.[1] MTC arises from the parafollicular C cells of the thyroid gland, which originate in the neural crest. The disease progresses from C-cell hyperplasia (CCH), often with elevated calcitonin levels, to microscopically invasive carcinoma, then grossly evident carcinoma.[2] Like other neuroendocrine tumors (NETs), MTC can elaborate a variety of products like calcitonin (CT), carcinoembryonic antigen (CEA), serotonin, and chromogranin A, which may cause symptoms such as diarrhea in patients with metastatic disease. In the context of CCH and MTC, the secretion of calcitonin predominates and can be used to confirm the diagnosis,[3] indicate treatment efficacy,[4] and monitor for disease progression or recurrence.[5]

MTC develops sporadically in 60% to 75% of patients[3, 6] or as a result of a germline mutation in the rearranged during transfection (RET) proto-oncogene, as observed in multiple tumors from patients with endocrine neoplasia type 2A (MEN2A) and MEN2B and familial MTC syndrome (FMTC). MTC often progresses in an indolent fashion and has a 15-year rate survival of 85%, but it has a tendency to spread to locoregional lymph nodes early, making surgical cure difficult.[7] Total thyroidectomy and lymphadenectomy result in biochemical cure (normalization of CT and CEA) only 40% of the time.[7, 8] Even when a biochemical cure is achieved, approximately 9% of patients will later develop recurrent disease.[8]

For patients with sporadic MTC, total thyroidectomy with central neck dissection (at a minimum) is performed upon histologic confirmation of the disease. Moley and DeBenedetti suggest that, in addition to central lymph node dissection, bilateral functional neck dissections be performed for patients with initially palpable MTC, because 44% of patients will have at least 1 positive lymph node in the contralateral compartment.[9] Patients with known RET mutations are offered prophylactic thyroidectomy in childhood or upon discovery of the mutation.[10, 11] Unfortunately, among patients with sporadic MTC or those with RET mutations who undergo surgery after MTC develops, only one-third will achieve biochemical cure after their index procedure; thus, many patients require reoperation for recurrent or persistent MTC.[9] It is encouraging to note that reoperation for cervical lymph node disease results in 5-year and 10-year overall survival rates of 97% and 94%, respectively; yet this long survival is not equivalent to a cure, and many patients endure repeated interventions, because they live with persistent recurrent disease.[12] In patients with widespread metastases, surgical debulking to palliate symptoms is recommended, because most patients can achieve long-term, symptom-free survival (median, 8.2 years).[13] Even patients with thoracic or mediastinal metastases may achieve some benefit from metastasectomy, although only small series exist, making it difficult to draw conclusions about the efficacy of this procedure.[14]

Because of the difficulty in achieving an initial surgical cure and the tendency of recurrent disease to progress to a point at which further metastasectomy is inadvisable because of disease extent or the high risk of surgery in the multiply reoperated neck, the need for medical treatment of MTC has been recognized for some time. Effective medical therapy to eliminate residual micrometastatic disease after an initial operation could raise the proportion of patients achieving biochemical cure and palliate those with symptomatic recurrent disease who are not amenable to reoperation. Until recently, no such effective medical therapy existed; and, because the rarity of MTC made clinical trial accrual difficult, much of our knowledge about medical treatment for MTC rested on small prospective series and retrospective reports.

The advent of targeted small-molecule kinase inhibitor drugs has revolutionized medical treatment of MTC. Drugs like vandetanib and cabozantinib produce disease regression in a significant portion of patients and can extend progression-free survival (PFS) in patients with advanced, unresectable MTC.[15, 16] Other multikinase inhibitors, such as sunitinib and sorafenib, also offer hope to patients with MTC who progress on other treatments, and ongoing clinical trials continue to evaluate additional agents. The objective of the current review was to update readers on the recent developments in targeted small-molecule therapies for the medical management of MTC. We also have attempted to provide an overview of the major radioactive and chemotherapeutic regimens that preceded those new therapies—and remain as treatment options in MTC—as well as some of the many other therapies that have been tried with limited success in this previously treatment-refractory disease.

Tyrosine Kinase Inhibitors

The first indication of the promise of small-molecule kinase inhibitors came from the class prototype, imatinib. Targeting the mutant breakpoint cluster region-Abelson (BCR-ABL) tyrosine kinase in chronic myeloid leukemia (CML), imatinib dramatically improved response rates of patients who had CML in blast crisis and significantly forestalled progression from the chronic phase in long-term studies.[17, 18] Imatinib also targets the mutated c-KIT receptor, which is responsible for gastrointestinal stromal tumor (GIST), and the use of imatinib after resection of high-risk GISTs produced similarly impressive results, with the 5-year survival rate improving from 35% to 83%.[19] These encouraging studies suggested a role for small-molecule inhibitors in MTC.

Like CML and GIST, oncogenic transformation in MTC occurs because of a mutation that causes the constitutive activation of a signaling pathway. The causative genetic region for autosomal-dominant MEN2A was mapped by genetic linkage to chromosome 10 in the late 1980s[20, 21]; and, in the early 1990s, it was determined that mutations in the RET gene cause MEN2A, MEN2B, and FMTC.[22-26] In sporadic MTC, somatic RET mutations occur in 40% to 65% of tumors.[16, 27] Although many different RET mutations can lead to MEN2 syndromes, the most prevalent mutations include a cysteine to arginine substitution at position 634 (C634R) in MEN2A and a methionine to thymine substitution at position 918 (M918T) in MEN2B.[28] The M918T mutation also represents the most common somatically occurring mutation in sporadic MTC.[27] RET is a membrane-bound receptor tyrosine kinase involved in renal and enteric nervous development and is activated by any of 4 glial-derived neurotrophic factor (GDNF) molecules.[29] Whereas RET activation principally induces the rat sarcoma (Ras)-rapidly accelerated fibrosarcoma (Raf)-mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK)/mitogen-activated protein kinase (MAPK) pathway, RET also can activate phosphatidylinositol-3-kinase/Akt (PI3K/Akt), Janus-activated kinase/signal transducers and activators of transcription (JAK/STAT), and jun-N terminal kinase (JNK), among other pathways (Fig. 1).[29-31] In MTC, RET mutations lead to substrate-independent dimerization of the receptor, causing constitutive activation, unrestricted signaling, and, ultimately, cancer.[29, 32]

Figure 1.

Receptors and pathways in medullary thyroid cancer (MTC) are illustrated. Kinase inhibitors block the activity of rearranged during transfection (RET), vascular endothelial growth factor receptor (VEGFR), and other receptors, inactivating the mitogen-activated protein kinase (MAPK) and phosphatidylinositol-3-kinase (PI3K) pathways, as well as others. Although the MAPK and PI3K pathways are the principle targets of these receptors in MTC, each receptor also directly or indirectly influences additional pathways, including Janus-activated kinase/signal transducers and activators of transcription (JAK/STAT), focal adhesion kinase (FAK), and jun-N-terminal kinase (JNK). Extensive interpathway cross-talk exists, which helps explain the complex beneficial and adverse effects of multikinase inhibitors. mTOR indicates mammalian target of rapamycin; PIP2, phosphatidylinositol bisphosphate; PIP3, phosphatidylinositol triphosphate; JUN, JUN protein; RET, rearranged during transfection kinase; PDGFR, platelet-derived growth factor receptor; FGFR, fibroblast growth factor receptor; MET, mesenchymal epithelial transition factor/hepatocyte growth factor receptor; EGFR, epidermal growth factor receptor.

Although it could be readily appreciated that MTC shared similar mechanisms of molecular pathogenesis to cancers treatable with imatinib, research into CML, melanoma, and papillary thyroid cancer helped spur the development of MAPK and related pathway inhibitors.[29, 33] In addition to RET, interactors with the MAPK pathway include receptors for vascular endothelial growth factor (VEGF), platelet-derived growth factor (PDGF), and epidermal growth factor (EGF).[34] Gain-of-function mutations in v-Raf murine sarcoma viral oncogene homolog B1 (BRAF) kinase, a downstream target of RET and a key effector in the MAPK pathway, occur in nearly 60% of melanoma cell lines and 45% of human papillary thyroid cancers.[35-37] With aberrant activation of the MAPK and related pathways recognized as causative events in human cancers affecting large numbers of patients, intensive research efforts have produced small-molecule inhibitors with activity at multiple receptors and at multiple steps of these pathways, from receptor to effector kinases (Table 1).[38-56] Although much of this research has focused on papillary thyroid cancer or other cancers, several kinase inhibitors have been evaluated for activity in MTC, and 2, vandetanib and cabozantinib, are currently approved by the US Food and Drug Administration for MTC.

Table 1. Selected Drugs Targeting Pathways Involved in Medullary Thyroid Cancer
DrugAlternative NameTargets (References)Selected Trials (References)
  1. Abbreviations: AXL, AXL receptor tyrosine kinase; BCR-ABL, breakpoint cluster region-Abelson tyrosine-protein kinase; CML, chronic myeloid leukemia; CSF1R, colony-stimulating factor-1 receptor (c-Fms); EGFR, epidermal growth factor receptor; FGFR, fibroblast growth factor receptor; FLT3, fms-related tyrosine kinase 3; KIT, c-Kit/stem cell factor receptor; MET, mesenchymal epithelial transition factor/hepatocyte growth factor receptor; MTC, medullary thyroid cancer; PDGFR, platelet-derived growth factor receptor; RAF, rapidly accelerated fibrosarcoma kinase; RET, rearranged during transfection; SRC, avian sarcoma viral oncogene homolog; VEGFR, vascular endothelial growth factor receptor.

VandetanibZD6474VEGFR2, VEGFR3, EGFR, RET, PDGFR (Wells & Santoro 2009,[29] Carlomango 2002[38])MTC early phase (Wells 2010,[39] Robinson 2010[40]); phase 3: 331 MTC patients (Wells 2012[15])
CabozantinibXL184MET, VEGFR2, RET, KIT, AXL, FLT3 (Yakes 2011[41])MTC phase 1 (Kurzrock 2011[42]); phase 3: 330 MTC patients (Elisei 2013[16])
SorafenibBAY 43-9006RAF, RET, VEGFR1-3, PDGFR, KIT, FLT3 (Hong 2008,[31] Haraldsdottir & Shah 2014[43])Phase 1: 1 MTC patient (Hong 2008[31]), 13 MTC patients (Hong 2011[44]); phase 2: thyroid cancer, 1 MTC patient (Gupta-Abramson 2008[45]), 21 MTC patients (Lam 2010[46])
SunitinibSU11248VEGFR1,2, PDGFR, RET, KIT, FLT3, CSF1R (Shah 2014,[43] Carl 2010[47])Phase 1: thyroid cancer, including 7 MTC patients (Carl 2010[47])
AxitinibAG-013736VEGFR1-3, much lower for KIT, PDGFR (Cohen 2008[48])Phase 2: thyroid cancer, including 11 MTC patients (Cohen 2008[48])
MotesanibAMG 706VEGFR1-3, PDGFR, KIT (Schlumberger 2009[49])Phase 1: 1 MTC patient (Rosen 2007[50]); phase 2: 91 MTC patients (Schlumberger 2009[49])
PonatinibAP24534BCR-ABL, SRC, FLT3, KIT, FGFR1, PDGFR, VEGFR2, RET (De Falco 2013[51])MTC cell culture, animal model (De Falco 2013[51]); MTC cell culture, comparison with other MTC agents (Mologni 2013[52]); phase 2: CML, high rate of arterial thrombotic events (Cortes 2013[53])
ImatinibSTI571BCR-ABL, PDGFR, CSF1R, KIT, lower activity against RET (Frank-Raue 2007,[54] de Groot 2007[55])Phase-II: MTC (Frank-Raue 2007,[54] de Groot 2007[55])
GefitinibZD1839EGFR (Pennell 2008[56])Trial including 4 MTC patients (Pennell 2008[56])

Vandetanib was the first kinase inhibitor approved for treatment of symptomatic metastatic MTC.[57] Initially developed as an orally available VEGF receptor (VEGFR) inhibitor, it was determined that vandetanib prevented the activation of RET receptors with common mutations and blocked MTC tumor growth in mice.[38] Two less common RET mutants, a valine to methionine substitution at position 804 (V804M) and a valine to leucine substitution at position 804 (V804L), exhibit resistance to vandetanib inhibition of RET.[58] In 2010, results from a phase 2 study demonstrated that treatment of with 300 mg of vandetanib daily induced objective responses in 6 of 30 patients (20%) with unresectable locally advanced or metastatic hereditary MTC.[39] Stable disease or objective responses occurred in 22 of 30 patients (73%), and 80% had reductions in serum calcitonin levels. The estimated PFS was 27.9 months. A second study of vandetanib 100 mg daily in patients with advanced, hereditary MTC reported similar results: 13 of 19 patients (68%) had stable disease or partial objective responses.[40] In both trials, diarrhea, rash, and fatigue were the most common adverse events, and QTc prolongation requiring dose reduction occurred.[39, 40]

Supported by the results from those initial trials, investigators initiated a phase 3 randomized, double-blind trial of vandetanib 300 mg daily versus placebo in patients with unresectable, locally advanced or metastatic MTC.[15] Included were 331 patients with sporadic or hereditary MTC. In that trial, vandetanib significantly prolonged PFS: the estimated median PFS was 30.5 months in the vandetanib treatment group compared with 19.3 months in the placebo group (P < .001). Objective radiologic responses occurred in 45% of patients in the treatment group versus 13% of the placebo group (P < .001), and disease control rates were 87% in the treatment group versus 71% in the placebo group (P=.001). No overall survival benefit was noted because of inadequate event numbers, and possibly because 93% of patients who progressed while receiving placebo elected to crossover to open-label vandetanib. In addition, all but 1 objective responses in the placebo group occurred after crossover. Toxicities were common, and 35% of patients required dose reductions because of adverse events. Grade 3 or greater adverse events included diarrhea (11%), hypertension (9%), QTc prolongation (8%), and fatigue (6%). Although not observed in the trial, torsades de pointes has been reported, and QTc monitoring is essential during vandetanib treatment.[43]

A potential problem with VEGFR-inhibitor treatment is the development of resistance or diminished response because of blockade-induced up-regulation of related pathways.[42] Cabozantinib, or XL184, is an orally available multikinase inhibitor with activity against MET, VEGFR2, RET, and others.[41] Its developers proposed that its broader inhibition of important kinases, particularly MET (which can be up-regulated in MTC), offers the potential to avoid treatment failures caused by targeting VEGFR alone.[41, 42] In cell culture and animal models, cabozantinib prevented phosphorylation of its target kinases, reduced cell proliferation, and limited angiogenesis, tumor invasiveness, and metastasis of multiple cancer cell lines.[41] In a phase 1 trial of cabozantinib, of 10 of 37 patients (29%) with advanced or recurrent MTC had an objective partial response, and 68% had either stable disease or an objective response at 6 months.[42]

The efficacy of cabozantinib for prolonging PFS was demonstrated in a phase 3 randomized, double-blind trial of cabozantinib among patients with progressive MTC.[16] In that manufacturer-sponsored trial, 330 patients with advanced or recurrent MTC were randomized to receive either cabozantinib 140 mg daily or placebo until they developed intolerable toxicity or disease progression. Patients who received cabozantinib had significantly longer PFS than those who received placebo (median PFS, 11.2 months vs 4.0 months; P < .001), and this advantage applied to patients both with and without RET mutations. Objective radiographic responses were observed in 28% of patients in the cabozantinib arm versus 0% of patients in the placebo arm (P < .001). Although not enough deaths had occurred to fully evaluate overall survival, despite the large advantage in PFS among treated patients, the investigators observed no difference in overall survival: death occurred in 21 patients (10%) in the cabozantinib arm and in 10 patients (9%) in the placebo arm. This was not because of crossover, because crossover upon progression was forbidden by the study protocol, and patients who were receiving any other additional cancer therapy were censored. Adverse events were frequent with cabozantinib treatment. Grade 3 or 4 adverse events (defined as severe or life-threatening/disabling[59]), including hemorrhage, fistulas, and gastrointestinal perforation, occurred in 69% of patients in the cabozantinib arm versus 33% in the placebo arm, whereas “serious adverse events,” defined as imminently life-threating or resulting in death (which were reported separately for grade 3 and 4 events),[60] occurred in 42.1% of patients in the cabozantinib arm versus 22.9% in the placebo arm. Overall, 79% of cabozantinib-treated patients required dose reductions because of adverse events. Because patients were required to have radiographically evident disease progression for inclusion in this trial, which was not required for the phase 3 trial of vandetanib, patients in the cabozantinib trial may have had more advanced disease, although this also was influenced by crossover (PFS: cabozantinib trial placebo group, 4.0 months; vandetanib trial placebo group, 19.3 months).[15, 16] This difference and the protocol difference regarding whether treatment crossover was allowed complicate direct comparison of the 2 trials, in terms of both outcomes and side-effect profiles.

Current National Comprehensive Cancer Network guidelines now recommend that clinicians consider either vandetanib or cabozantinib as medical therapy for advanced, unresectable MTC (Fig. 2).[61] After failure of these agents, the guidelines recommend considering 2 additional agents—sorafenib or sunitinib—or enrollment in a clinical trial.[61] The evidence for sorafenib and sunitinib in MTC is not as robust as that for the other drugs, but small studies have demonstrated efficacy of both in MTC.

Figure 2.

This graphic provides a summary of medullary thyroid cancer treatment guidelines. PRRT indicates peptide receptor radionuclide therapy; EBRT, external-beam radiation therapy.

Sorafenib is a multikinase inhibitor that is FDA approved for the treatment of renal cell and hepatocellular cancers.[46] Although it was developed for BRAF and CRAF inhibition, sorafenib also potently inhibits several other pathways and kinases including VEGFR, PDGF receptor (PDGFR), and wild-type, mutant, and V804M and V804L resistant-mutant RET.[31] A phase 1 study evaluating treatment with sorafenib and the farnesyl-transferase inhibitor tipifarnib, which affects RAS kinase, demonstrated a dramatic objective response in a patient with sporadic, RET mutation-positive MTC.[31] Later follow-up with 13 MTC patients enrolled in the same study revealed a 38% partial response rate and a 69% overall disease control (partial response plus 6-month stable disease) rate.[44] A phase 2 study evaluated 21 MTC patients who received 400 mg twice-daily sorafenib without tipifarnib. All patients experienced stable disease and some tumor shrinkage, and 2 patients had objective partial responses.[46] Eleven patients (52%) had stable disease that lasted at least 6 months.[46] Adverse events were frequent: 76% of patients required dose reductions, and grade ≥3 complications occurred in 13 patients (62%). The most common serious adverse events were hand-foot-skin reaction, hypertension, diarrhea, and infection.[46] A nonrandomized, retrospective study of Spanish patients with thyroid cancer who received sorafenib treatment demonstrated a 47% response rate among 15 patients with MTC.[62] Additional small studies and case reports support the consideration of sorafenib in MTC, although additional randomized studies are needed.[57, 63, 64]

Sunitinib is a multityrosine kinase inhibitor with activity against VEGFR1 and VEGFR2, c-KIT, fms-related tyrosine kinase 3 (FLT3), PDGFR, and RET.[47] On the basis of its inhibitory spectrum, in 2008, Kelleher and McDermott reported treating a patient with metastatic MTC using sunitinib, which led to a marked reduction in tumor size on imaging studies and improvement in clinical symptoms.[65] Another group reported a patient with advanced MTC whose dramatic response to sunitinib permitted surgical resection of his disease, despite the absence of a detectable RET mutation.[66] A phase 2 trial that tested sunitinib at a dose of 37.5 mg daily included 7 patients with metastatic MTC and reported objective responses in 3 of 6 patients who had radiologically evaluable lesions (50%) and stable disease in 2 patients, for a disease control rate of 71% in this small sample.[47] Greater than 10% of all treated patients experienced grade ≥3 toxicities of fatigue, diarrhea, hand/foot syndrome, and leukopenia/neutropenia.[47] Both sorafenib and sunitinib effectively blocks phosphorylation of the RET V804M mutant, which confers resistance to vandetanib and reduces the effectiveness of cabozantinib.[52]

Other tyrosine kinase inhibitors have been assessed for the treatment of MTC. The class-prototype imatinib has limited inhibitory activity against RET, and 2 open-label trials included a total of 24 MTC patients who received imatinib. In those trials, no objective responses were observed, whereas treatment incurred significant toxicity, and the use of imatinib has stalled in MTC.[54, 55] Gefitinib represents another drug with efficacy in a different cancer type that was explored in MTC. Gefitinib is an EGF receptor (EGFR) inhibitor with activity in EGFR-mutant nonsmall cell lung cancer; however, when it was tested in a phase 2 trial that enrolled 4 patients with MTC, no objective responses were noted.[56]

Axitinib is an orally available kinase inhibitor with relative specificity to VEGFR1, VEGFR1, and VEGFR3.[48] In a phase 2 trial, 60 patients, including 11 with MTC, received axitinib at a dose of 5 mg twice daily. Partial response or stable disease occurred in 5 of those 11 patients (45%). Adverse events in that trial included fatigue and diarrhea, which were common. Twelve percent of all treated patients experienced grade ≥3 hypertension, and no other serious toxicity occurred in >5% of patients. It is noteworthy that the activity of axitinib in MTC occurs without inhibition of RET. Another orally available multikinase inhibitor, motesanib, demonstrates activity against VEGFR1, VEGFR2, VEGFR3, PDGFR, and KIT but does not inhibit mutant RET.[49] A phase 2 trial that enrolled 91 patients with MTC demonstrated that, although only 2 patients achieved objective partial radiographic responses, 76% of treated patients achieved “clinical benefit,” defined as an objective response or durable stable disease.[49] Both RET mutation-positive and mutation-negative patients achieved durable stable disease, and higher rates (62%) were observed for RET mutation-negative patients than for RET mutation-positive patients (42%).[49] Grade 3 and 4 adverse events occurred in 38% and 3% of patients, respectively, and gallbladder toxicity occurred in 8 of 91 patients (9%).[49]

Ponatinib, a newer multikinase inhibitor, demonstrates broad inhibitory activity across a wide range of kinases. In particular, ponatinib produces strong inhibition at low drug concentrations of BCR-ABL and RET; and its anti-RET potency is at least 100 to 1000 times greater than that of vandetanib, cabozantinib, or motesanib.[52] In vitro studies have demonstrated that ponatinib inhibits phosphorylation and signaling through multiple pathways; and, along with BCR-ABL and RET, ponatinib blocks VEGFR2, PDGFRα, avian sarcoma viral oncogene homolog (SRC), KIT, fibroblast growth factor receptor 1 (FGFR1), and FLT3.[51] Two recent studies have supported the potential efficacy of ponatinib in MTC. In cell culture, low concentrations of ponatinib prevent phosphorylation of RET and its downstream target ERK1/ERK2,[52] and mice injected with tumors from an MTC cell line exhibited significantly inhibited tumor growth with ponatinib treatment.[51] Notably, ponatinib exhibits in vitro activity against kinases with common inhibitor-resistance mutations, including the BCR-ABL T315I mutant, the FLT3 F691I mutation, and the RET V804 and Y806 mutations.[51, 52] In addition, although rare, some patients do carry RET V804 or Y806 mutations, which confer resistance to vandetanib and reduce cabozantinib effectiveness by blocking the RET ATP-ase active site where these inhibitors bind.[52] By exploiting a slightly different mechanism of activity—binding and stabilizing the inactive receptor state—ponatinib could offer an alternative after initial treatment options have failed.[52] Although such a broad spectrum of activity theoretically could lead to greater efficacy, recent clinical trials of ponatinib in CML have raised serious safety concerns. In the phase 2 PACE trial, high rates of arterial thrombotic events were observed, including cardiovascular (7.1%), cerebrovascular (3.6%), and peripheral vascular (4.9%) thrombosis.[53] Ponatinib sales and clinical trials were suspended in late 2013 after US Prescribing Information adverse event surveillance identified a 17.1% rate of arterial thrombotic events,[53] and the current drug label contains a black-box warning citing a rate of arterial and venous occlusion of “at least 27%” in ponatinib-treated patients.[67] Although ponatinib once again can be prescribed for CML under a risk evaluation and mitigation strategy, reports of fatal arterial thrombotic events continue, and its future in MTC appears to be doubtful.[68]

Neither vandetanib nor cabozantinib exhibits a definitive correlation between RET mutational status and efficacy.[69] Phase 2 trials of cabozantinib identified patients with and without RET mutations who had tumor shrinkage.[42] In the phase 3 vandetanib trial, patients with sporadic MTC who had the RET M918T mutation had a higher response rate than those who did not (54.5% vs 30.9%), but the investigators' inability to determine the mutational status of 45% of the study patients because of inadequate tissue specimens limited conclusions regarding genotype and response to treatment.[15] Among patients who received motesanib, which functions principally through VEGFR inhibition, higher rates of partial response and stable disease (8% and 62%, respectively) were observed in RET mutation-negative patients than in RET mutation-positive patients (0% and 42%).[49] Further complicating genotype-phenotype correlations, axitinib offers minimal if any inhibition of RET,[48] yet it still produced stable disease or objective responses in 45% of patients with MTC, whereas gefitinib produced no responses through targeting EGFR.[56] Although RET mutations are sufficient to induce neoplasia in patients with MEN2, aberrant activity of VEGFR and MET is observed in MTC.[42, 49] Lack of a strong connection between the presence of an activating RET mutation and response to treatment (or drug activity against RET) highlights the contributions that blockade of additional receptors other than RET, especially VEGFR, make to the clinical efficacy of these drugs.[69] Thus, despite the appeal of “rational” treatment, the pathways representing the most important targets for clinical effectiveness may not be the same as those presumed based on our understanding of MEN2. Further research could improve our understanding of signal transduction pathways when considered as complex networks rather than as discrete and sequentially proceeding pathways, thus improving future efforts to target therapeutics.

Although there are now options for patients with metastatic MTC, the question of optimal medical management likely will remain open for the time being. To our knowledge, no randomized trial has yet demonstrated an overall survival benefit in MTC with targeted agents, and the toxicities and expense of such agents are considerable.[69] Head-to-head studies directly comparing therapies are lacking but could elucidate the relative benefits of different agents. Over the next several years, longer follow-up of patients included in recent trials will become available to determine effects on overall survival, studies of additional agents will accrue, and continuing research will further illuminate the cellular networks involved in response to treatment.

Cytotoxic Chemotherapy

Before the introduction of current kinase inhibitor-based therapies, chemotherapy and radiation formed the mainstay of medical treatment for MTC. Although recent advances in small-molecule therapies have largely supplanted these regimens, they remain in consideration for patients with refractory disease, and research continues to improve the sensitivity of MTC cells to these therapies. Because kinase inhibitors display mostly cytostatic rather than cytocidal effects, a need persists for effective agents to eliminate cancerous cells.

Doxorubicin and dacarbazine

Standard cytotoxic chemotherapy agents have been used in the management of metastatic MTC for more than 30 years with mixed results. Doxorubicin (Adriamycin) and dacarbazine (DTIC) are the 2 agents with the most extensive track record, and DTIC is currently recommended as a treatment option for patients with disseminated, symptomatic disease.[61]

Doxorubicin was first tested in patients with MTC in the 1970s. Two case series[70, 71] and 1 small prospective study[72] documented the response of advanced MTC to doxorubicin administered as a single agent. The results from the case series were split. One reported a robust antitumor effect,[70] but the other reported only minor tumor shrinkage.[71] The prospective study was more encouraging, because 3 of 5 patients with MTC responded to treatment and demonstrated significantly better overall survival than the patients who did not respond.[72]

Thereafter, doxorubicin was studied in combination with several other agents. In 1985, Shimaoka et al combined doxorubicin with cisplatin in a multi-institutional, prospective, randomized controlled trial of advanced thyroid cancer. Ten of 84 patients in that study had MTC: 4 received doxorubicin alone, and 6 received the combination regimen. Disappointingly, only 1 patient with MTC in the doxorubicin-only arm and 2 patients with MTC in the combination arm demonstrated partial tumor responses, defined as a decrease >50% in total tumor volume. There were no complete responses.[73] Sridhar et al also tested doxorubicin in combination with cisplatin, but they included only 1 patient with MTC in a study of neuroendocrine cancers. The patient with MTC did relatively well, demonstrating a partial response that lasted for 58 weeks; but treatment was eventually discontinued because of doxorubicin-mediated cardiac toxicity.[74]

In 1990, Scherubl et al attempted to build on previous work by testing doxorubicin combination regimens in a phase 2 clinical trial. In that study, vincristine and cisplatin were added to doxorubicin and administered to 20 patients with advanced thyroid cancer, 10 of whom had MTC. Of the patients with MTC, 1 had a partial tumor response, 6 had stable disease, and 3 progressed. The authors declared this combination ineffective for the treatment of advanced thyroid cancers.[75] Droz et al used a range of chemotherapeutics in patients with advanced, nonanaplastic thyroid cancers but included only a small number of patients with MTC in each group. Those investigators observed partial tumor responses in the patients with MTC in the doxorubicin-only arm, but these were short lived, and many patients experienced cardiac toxicity.[76] Although doxorubicin induces partial morphologic tumor responses in a reasonable proportion of patients with MTC, its use is severely limited by cardiac toxicity; thus, it is not an ideal chemotherapeutic agent for patients with advanced MTC.

DTIC, an alkylating agent, was first used as a single-agent chemotherapeutic for MTC in the 1980s. The first case reports boasted dramatic reductions in metastatic tumor burdens (although these lasted <1 year) and, thus, encouraged further investigation of this drug's efficacy in advanced MTC.[77, 78] The first DTIC/5-fluorouracil (5-FU) combination trial achieved partial responses in 3 of the 5 patients with MTC who were included in the study. These responses lasted from 8 to 10 months. A fourth patient achieved stable disease. Given the encouraging tumor responses and a limited number of grade 1 and 2 side effects, that study served as the basis for testing DTIC combinations in larger numbers of patients with MTC.[79]

The DTIC combination trials that followed failed to achieve results that would suggest the drug's utility as a standard adjuvant treatment of advanced MTC. DTIC was combined in various ways with cyclophosphamide, vincristine, streptozocin, and epirubicin. Although no trial included more than 20 patients with MTC, most reported moderate success, with 33% to 50% of patients achieving partial tumor responses or stable disease for an average of 1 year.[80-84] Despite the modest results of those trials, of the major antineoplastic agents studied, DTIC is the only 1 currently recommended for patients with disseminated, symptomatic MTC by the National Comprehensive Cancer Network.[61]

Capecitabine

Capecitabine is a 5-FU prodrug that is converted in target tissues by thymidylate phosphorylase and selectively accumulates in tumors, rather than in plasma or muscle tissue. Capecitabine interferes with DNA synthesis by inhibiting thymidylate synthase, thus preventing tumor cell proliferation.[85] Widespread experimentation with 5-FU in other neuroendocrine cancers, the availability of capecitabine by oral administration, and the promise of trials evaluating capecitabine in colorectal and breast cancers have led to the use of this drug in metastatic MTC.[86] To date, only case reports exist; and, in these, the concurrent use of other agents obscures the effects of capecitabine. Nevertheless, disease stabilization has been reported in most patients.[86-88]

External-Beam Radiation

External-beam radiotherapy (EBRT) has been used sporadically in the treatment of MTC for more than 40 years, but it is not currently considered standard treatment. Like other methodologies, there are limited clinical trials to guide decision-making about the indications for EBRT. Most of what is known about the efficacy and safety of EBRT for MTC is derived from small, single-institution experiences.

Before 1990, several small series reported contradictory conclusions about the utility of EBRT as adjuvant treatment for MTC. In 1977, Steinfeld used EBRT to treat 4 patients with MTC, and they achieved local disease control that lasted 3 to 6 years. He recommended using EBRT to treat MTC and further suggested testing the modality in combination with chemotherapy. Samaan et al published a contradictory retrospective study in 1988. In that series, 57 patients with various stages of MTC who underwent total or subtotal thyroidectomy followed by EBRT were compared with patients who underwent surgery alone. The patients who received EBRT had worse survival than those who underwent thyroidectomy alone.[89]

The largest series examining EBRT in MTC was published in 1990 by Jensen et al. That multicenter, retrospective review of 5287 patients with thyroid cancer included 200 patients who had MTC. Approximately 30% of the patients with MTC were offered some sort of adjuvant treatment in the study, although the specific number who received EBRT was not reported. The authors observed that nearly equal proportions of patients with MTC at each disease stage were offered EBRT. An analysis of the entire MTC cohort demonstrated that patients who underwent surgery followed by EBRT had a 100% 5-year survival rate, whereas patients who underwent surgery alone had a 91% 5-year survival rate. When the cohort was divided by disease stage, EBRT conferred a survival advantage in patients who had regional disease, but not in those who had distant metastases.[91] Given the large sample size, the series by Jensen and colleagues serves as 1 of the foundations for the American Thyroid Association recommendation to offer EBRT to patients who have persistently symptomatic MTC post-thyroidectomy and are at high risk for locoregional recurrence.[11]

Another large series that examined EBRT in patients with MTC was a retrospective, single-center series that studied patients who were diagnosed and treated for MTC from 1954 to 1992. That study of 72 patients with MTC included 43 who received EBRT. At that institution, the indications for EBRT in the context of MTC included: 1) residual microscopic disease after thyroidectomy, 2) the presence of tumor within 2 mm of the resection margin, 3) extraglandular invasion or lymph node involvement, and 4) elevated postoperative CT (in the absence of distant metastases). Patients received a variety of radiation protocols. The most important finding in that series was that, in patients who were at high risk for locoregional recurrence, EBRT appeared to confer a significantly longer 10-year relapse-free survival compared with those who underwent surgery alone (P=.049).[91] That study provided support for the report by Jensen et al demonstrating that EBRT is indicated in those patients who are at high risk for locoregional recurrence.

Nearly a decade later, Schwartz et al published a single-institution series that included 34 patients with MTC who underwent total thyroidectomy, appropriate neck dissection, and EBRT between 1995 and 2004. Indications for adjuvant EBRT in that study were gross or microscopic residual disease, soft tissue extension, lymph node disease, or mediastinal involvement. There were 4 locoregional failures in the cohort, all of which occurred within 26 months after the completion of EBRT. In that study, the 5-year disease specific survival rate was 62%, the overall survival rate was 56%, and the 5-year relapse-free rate was 87%. These results supported a role for EBRT in patients at high risk for local recurrence in the adjuvant setting. The authors emphasized that this type of control was especially important in patients with MTC, even those with distant metastases, because they tend to live a long time with disease, and good locoregional control can confer a better quality of life.[93]

The most recent study to examine the use and utility of EBRT in patients with MTC was published in 2010 by Martinez et al. Their analysis was conducted using the National Cancer Institute Surveillance, Epidemiology, and End Results cohort from 1988 to 2004 and included 534 patients with MTC. All patients had undergone total thyroidectomy with at least 1 lymph node excised at the time of surgery. Patients with distant metastases were excluded. It is noteworthy that EBRT did not affect overall survival when all patients (n=66) were examined or when the subset of patients with lymph node-positive disease was analyzed.[94]

Currently, the American Thyroid Association suggests using EBRT postoperatively in patients who are at high risk for locoregional recurrence and offering it to patients who have extensive metastatic (M1) disease as a palliative treatment.[11]

Radioactive Iodine

Radioactive iodine (RAI) has been useful in the treatment of differentiated thyroid cancer since the 1940s[95] but is largely ineffective in MTC. In differentiated thyroid cancers, follicular cells concentrate radioactive 131I through a sodium iodide symporter on the cell surface, permitting damage to the cancer cell by the emitted β particle. Although C cells do not concentrate iodine,[96] they can be damaged by exposure to radioactive 131I by virtue of the bystander effect.[97-100] RAI treatment has been offered to patients who have MTC with positive postoperative thyroid scans or persistently elevated calcitonin, because these are signs that thyroid tissue was left behind after surgery. It is possible that the residual tissue is a mixture of C cells and iodine-concentrating follicular cells; thus, small foci of MTC may be eradicated by the β particles emitted from neighboring follicular cells.

Although initial case reports were promising,[101, 102] the utility of this modality has not been borne out in modern practice. Saad et al published the MD Anderson experience with RAI in MTC in 1983. In their series, there was no difference in disease course, recurrence rates, or overall survival between 15 patients with MTC who received RAI and 84 patients who underwent surgery alone.[103] In 2006, Erdogan et al published a series (n=7) suggesting that RAI may be useful as an adjuvant treatment for MTC in patients who have persistently elevated CT levels and residual microscopic disease. Their conclusions were based on the finding that 2 of 3 patients with localized disease and 1 of 4 patients with locoregional disease achieved a biochemical cure after treatment with RAI. However, their conclusions were limited by the small study population and short follow-up interval, which was a maximum of 5 years.[104]

The most recent RAI series by Meijer et al[105] is the largest and supports the suggestion made 30 years earlier that RAI is not an effective adjuvant treatment for MTC. That multicenter, retrospective analysis compared 232 patients who had local or locoregional MTC and underwent surgery along with 61 matched patients who also received postoperative RAI. The investigators attempted to standardize the surgical treatment received by the patients by only including those patients who underwent surgery according to the American Thyroid Association guidelines set forth in 2009.[11] In these patients, they observed no difference in disease-free or disease-specific survival. Their findings held when they performed the same analysis for the subgroup of patients who had hereditary, clinically apparent MTC. Patients with MTC still occasionally receive RAI when medical therapies like cytotoxic chemotherapy fail, but current evidence indicates this is unlikely to provide much benefit. There has been renewed interest in “resensitizing” thyroid cells that are refractory to treatment with RAI by using kinase inhibitors in follicular-origin thyroid cancer, but those experiments have not yet gained ground in MTC.[106, 107]

Somatostatin Analogue-Based Regimens

Somatostatin is an endogenous peptide that inhibits many secretory or proliferative cellular functions by binding to a somatostatin receptor (SSTR) expressed on cells of neural crest origin. The endogenous peptide has a short half-life, but longer lasting synthetic analogues allow imaging and treatment of a variety of NETs. Octreotide, the most widely used analogue, binds primarily to SSTR type 2 (SSTR2), which is the most commonly expressed receptor subtype on gastrointestinal (GI) NETs and on MTC cells.[107] Octreotide suppresses hormone secretion by NETs, thereby alleviating symptoms, reducing biomarker levels, and occasionally causing tumor stabilization.[109] Given the drug's success in treating patients with GI NETs, it has also been tried in patients with MTC.

In 1990 Mahler et al reported 3 patients with metastatic MTC who received treatment daily with escalating doses of octreotide (1.0 mg, 1.5 mg, or 2.0 mg daily) delivered using a subcutaneous pump. In 2 patients, both with MEN2A, CT levels dropped below baseline and remained at normal levels for 15 and 17 months. In addition, those patients reported an improvement in symptoms, providing support for use of the drug in patients who have MTC with symptoms that are refractory to other treatments.[110] Two years later, Modigliani et al published a trial that included 14 patients who had MTC and persistently elevated CT levels. Those patients were stratified into high-risk or low-risk groups based on their pretreatment CEA levels and received a daily infusion of octreotide 0.5 mg daily for 90 days. In contrast to Mahler et al, this group observed that treatment with octreotide was not associated with a sustained decrease in CT levels, nor did the drug cause significant tumor stabilization or regression. In fact, in the high-risk group of patients, only 1 patient had a slight drop in CT, and all demonstrated progressive disease on imaging studies.[111]

These early disappointing results led investigators to experiment with octreotide combined with other antineoplastic therapies. Two different trials from the same Italian group combined octreotide with recombinant interferon-α2b (IFNα2b), as this adjunct had been tried in other NETs with positive results.[112] In the first trial, which was published in 1996, patients with MTC received treatment for a total of 12 months with 3 daily injections of octreotide and 3 weekly injections of IFNα2b. Two of 8 patients enrolled in that study dropped out because of the side effects from IFNα2b treatment. The remaining 6 patients had no significant tumor response but did have decreases in their CEA levels.[113] The second trial was published in 2000 and included 7 patients with symptomatic and advanced MTC. This time, patients received treatment with lanreotide, another long-acting somatostatin analogue, in combination with IFNα2b. Tumor responses were better in that trial, with 3 patients demonstrating disease stabilization after 6 months of treatment and 2 patients exhibiting minor (<25%) tumor regression. The 2 remaining patients continued to progress throughout treatment. Six of 7 patients had an improvement of their symptoms. Despite the slightly improved outcomes compared with the previous trial, the authors concluded that IFNα2b was not a helpful adjunct to MTC treatment with somatostatin analogues, because it caused unpleasant side effects in most patients and failed to have much effect on tumor growth.[114]

The most recent trial examining the utility of a somatostatin analogue for MTC treatment enrolled 22 patients with advanced MTC and treated all with either subcutaneous octreotide or octreotide LAR (the long-acting formulation). Thirteen patients in that group also received chemotherapy (n=6), EBRT (n=2), or a combination of the 2 (n=5). In the group of 13 patients who received a somatostatin analogue plus additional treatments, 12 had subjective tumor responses, including 5 partial responses and 7 with stable disease. In the group that received somatostatin analogues alone (n=9), 6 patients had objective tumor responses, including 3 partial responses and 3 stabilizations. There was no significant difference in the overall survival reported for the 2 different groups; therefore, the authors reiterated what others had already observed: somatostatin analogues may be helpful for controlling symptoms in advanced MTC, but they do not appear to significantly affect tumor burden.[115]

Peptide Receptor Radionuclide Therapy

Peptide receptor radionuclide therapy (PRRT) was first developed to treat SSTR-expressing tumors but has now expanded to use other receptors, such as cholecystokinin (CCK) receptors A (CCK-A) and CCK-B. This treatment relies on the binding of a radiolabeled ligand to its respective receptor expressed on a tumor's surface to produce a cytotoxic effect. Patients who are eligible for this treatment should have advanced cancer with inoperable metastases.[116] In MTC, PRRT targeting SSTR2 and CCK-A or CCK-B has been tried, because MTC expresses both receptors at high levels.[108, 117]

The only prospective study using CCK-B/gastrin receptor-based PRRT treated 8 patients who had advanced MTC using the radioligand 90Y-DTPT-D-Glu-minigastrin.[118] Patients were injected with the radioligand at 4-week to 6-week intervals and received escalating doses until toxicity limited further increases. Overall, 2 patients developed partial remissions, and 4 had stable disease. Unfortunately, the regimen proved relatively toxic, because 1 patient (the patient who demonstrated the best response) went on to develop CML as well as grade 1 nephrotoxicity, and another patient developed chronic renal failure. Other PRRT trials in patients with MTC have all used radiolabeled somatostatin analogues that bind to SSTR2. In 2004, Bodei et al published their experience with SSTR PRRT in 21 patients with MTC.[119] Those patients received various cumulative doses of Yttrium Y 90 (90Y)-DOTA-tyr3-octerotide (90Y-DOTATOC), but, regardless of the dose, all were pretreated with amino acid solutions to prevent nephrotoxicity. The authors observed that 67% of their patients derived a clinical benefit from PRRT, which manifest as either an objective tumor response or stabilization of tumor burden. Two patients had complete responses, and the duration of clinical benefits ranged from 3 months to 40 months. Twelve patients had stabilization, and 7 failed to respond to treatment. None of these patients suffered permanent renal toxicity. The authors were encouraged by the moderate responses in their patients, but they noted that the patients treated in this study were the least likely to derive benefit from the treatment, because their tumor burdens were too great for an optimal 90Y radiation effect.[119]

In 2007, a group in Basel, Switzerland published results from a phase 2 trial examining the efficacy of 90Y-DOTATOC treatment in 31 patients with stage IV MTC (M1 disease) who had progressed within the last 12 months and had visible radiolabeled somatostatin analogue uptake on an octreotide scan.[120] Patients received treatment with 90Y-DOTATOC until they failed to respond to treatment. Treatment response was gauged based on an alteration of CT doubling time, toxicity, and overall survival. Patients who demonstrated significant lengthening of their CT doubling time were characterized as “responders.” In this cohort, there were 18 responders (58.1%) and 13 nonresponders (41.9%). PRRT responders demonstrated significantly longer survival than nonresponders as measured both from the time of MTC diagnosis (108 months vs 80 months; P=.009) and from the time of 90Y-DOTATOC treatment (74.5 months vs 10.8 months; P=.02). It is noteworthy that there was no correlation between PRRT response and uptake on a pretreatment octreotide scan, suggesting that octreotide scanning may exclude patients from treatment who actually could benefit. Like in other PRRT trials, some patients developed nephrotoxicity as a consequence of treatment, 4 of which were permanent toxicities. Although that study used changes in CT doubling time rather than objective tumor response by imaging to gauge benefit, the difference in survival between responders and nonresponders supports the conclusion that PRRT has a positive effect on advanced MTC and, thus, is worthy of further study.[120]

The most recent report of PRRT for metastatic MTC evaluated both 90Y-labeled and 177Lu-labeled somatostatin analogues.[121] In total, 16 patients with advanced nonradioiodine-avid thyroid cancer were enrolled, including 8 who had metastatic MTC. Patients received up to 5 treatments of PRRT. Of the patients with MTC, 4 maintained stable disease, 1 developed a partial remission, and 1 continued to progress. Two patients were lost to follow-up. The patient who attained partial remission also had decreased CT levels, whereas those who progressed had their CT levels increase. In that trial, only mild hematologic toxicity was reported, and all cases were reversible. Of the entire cohort, the patients with MTC demonstrated the most promising responses to treatment, supporting the Swiss group's conclusion that PRRT is a worthwhile addition to MTC treatment and warrants a phase 3 study.[121]

Experimental Treatments

Several other treatments are in the early stages of development for the treatment of advanced MTC and hopefully will have an impact on future management. Histone deacetylase inhibitors increase 125I accumulation in follicular cells and cause apoptosis. Reports of its use in 3 patients with MTC have been published; and, in each case, only disease stabilization was achieved. It is now being tried as a chemosensitizing agent.[122-129] Thalidomide is known for its antiangiogenic properties and has been tried in 7 patients with MTC in a phase 2 trial. Because of severe side effects, only 5 patients completed the study. A partial tumor response was achieved in only 1 patient.[130-136] Plitidepsin is an antibiotic that can induce apoptosis and has been used to treat 1 patient with MTC. That patient had a partial response to treatment, with a 52% reduction in lymph node disease. A phase 2 trial is now underway.[137] Radioimmunotherapy, a targeted treatment that uses bispecific, anti-CEA monoclonal antibodies that also bind a radiolabeled hapten (131I or 90Y-DOTA), has been reported in approximately 100 patients with MTC. Most patients have stable disease with treatment, but a few minor tumor responses have been reported as well as 1 complete response.[138-145] Vaccination with dendritic cells pulsed with MTC-specific antigens also has been attempted: a trial in 7 patients with MTC resulted in 1 objective tumor response and 2 biologic responses.[146] ,147

Summary

For much of its history, advanced or metastatic MTC has proven refractory to most medical therapies. Although DTIC or doxorubicin can produce tumor responses in a subset of patients with MTC, no large trials demonstrating survival benefits exist. Nevertheless, until recently, DTIC and doxorubicin were the only options that exhibited any effectiveness, and they still may play a role when MTC fails with other treatments. Other therapies like somatostatin analogues may relieve symptoms in some patients but do not alter the disease course. EBRT and RAI are mostly ineffective. Additional chemotherapeutic combinations, novel agents, immunotherapy, and other drugs all have been tried in MTC but were disappointing or remain in the early phases of development. Given the long record of failures in managing this frustrating disease, the high rates of partial response and disease stabilization with small-molecule kinase inhibitors mark an important shift in MTC treatment. Evidence from large randomized trials demonstrating improved PFS supports using drugs like vandetanib and cabozantinib as first-line agents for symptomatic metastatic MTC. Sorafenib and sunitinib also may produce improvement in patients who fail on first-line drugs. Despite better PFS, however, cytostatic rather than cytotoxic effects predominate with these kinase inhibitors, and their impact on overall survival remains to be demonstrated. Moreover, their adverse events are frequent and often serious. Although the challenge posed by metastatic MTC has not been solved by currently available kinase inhibitors, patients have more options than ever before. Research in the next few years will likely add additional kinase inhibitors for patients who progress on or develop resistance to existing drugs, and the search continues for agents with greater efficacy and cytotoxicity for MTC cells.

FUNDING SUPPORT

This work was supported by National Institutes of Health grant 5T32 CA148062-04 (to J.E.M. and S.K.S.)

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary