Medullary thyroid carcinoma (MTC) accounts for <8% of thyroid cancers. The primary treatment of hereditary or sporadic MTC is total thyroidectomy with dissection of ipsilateral and central lymph nodes, extended in some cases to contralateral dissection. Many patients are cured by this surgical intervention, especially those with familial hereditary MTC and who were operated on before tumor metastatic spread. However, a non-negligible number of patients show persistent disease after primary surgery, as documented by measurable serum calcitonin.1 Pentagastrin stimulation makes the test extremely sensitive. Three months after surgery, serum calcitonin level is not detectable in >60% of patients without lymph node involvement, as opposed to <20% of patients with lymph node involvement.2 When a recurrence is localized in the neck or mediastinum, a new surgery can be performed, but is followed by a decrease of calcitonin serum level to undetectable levels in <⅓ of patients.3 In patients with MTC, the rate of overall survival 10 years after primary surgery is 69%, but it decreases to 10% when distant metastases are present.4 Patients with localized residual disease and/or distant metastases may survive for several years or rapidly progress and die of their disease. Thus, highly reliable prognostic factors are needed for an early distinction between high-risk patients who need to be treated and low-risk patients who warrant a watch-and-wait approach. These prognostic factors are very useful in the search for new therapeutic options.
This review will first consider the most appropriate prognostic indicators for the selection of patients in clinical trials evaluating a new treatment modality, and the most efficient imaging techniques aimed at detecting distant metastases. Then, the effectiveness and toxicity of all currently available treatments will be analyzed and compared.
Selection of Patients on the Basis of Prognostic Indicators
Among the various prognostic parameters that could identify high- and low-risk groups, advanced age, advanced stage of the disease, and associated multiple endocrine neoplasia 2B are commonly accepted factors of poor prognosis,5 and have been tested in many multivariate analyses. Recently the presence of lymph node metastasis was confirmed as the most significant factor affecting the outcome of the disease.6 These factors have been used to develop different staging systems. Kebebew et al7 concluded that the European Organization for Research and Treatment of Cancer prognostic scoring system,8 which takes into account age, sex, nature, and stage of disease for all thyroid cancers, had the highest predictive value.
The presence of a somatic mutation in the RET oncogene is a strong risk factor. Mutations are observed frequently in familial disease, but also in 43% of the sporadic cases.9 Higher probability of persistent disease and lower survival rates are associated with these mutations. A correlation between somatic RET mutation and the presence of lymph node metastases at diagnosis has been reported. The Cdc25B phosphatase has also been proposed as a novel marker of aggressive MTC.10 Thus, the molecular characterization of MTC is advancing and will eventually provide prognostic information and new therapeutic approaches. However, at this time, although these prognostic factors and scoring systems predict rather accurately the probability of cure by primary surgery, they remain rather poor in predicting disease outcome for patients who are not cured by surgery.
Tisell et al11 have shown that the tumor cell proliferation index provided by the measure of Ki-67 expression is another prognostic marker for MTC. Measuring calcitonin and/or carcinoembryonic antigen (CEA) doubling time (DT) is another way of looking at tumor progression rate. Calcitonin DT has been shown to be an independent predictor of survival, with a high predictive value, in a population of patients who have not normalized their calcitonin, even after repeated surgery.12 Forty-one patients with calcitonin DT >2 years were still alive at the end of the study, 2.9 to 29.5 years after primary surgery. Eight (67%) patients with calcitonin DT between 6 months and 2 years died of the disease 40 to 189 months after surgery, and all 12 patients with calcitonin DT <6 months died of the disease 6 months to 13.3 years after primary surgery. Although this study was retrospective and involved a limited number of patients, calcitonin DT was used to select patients with progressive disease in 2 radioimmunotherapy (RIT) clinical trials. These RIT trials showed a significant increase in overall survival (OS) as compared with a historical control group matched for calcitonin DT.13 Calcitonin DT was also taken into account in a positron emission tomography (PET) imaging study that concluded that maximum standard uptake value (SUVmax) correlated with calcitonin DT, and that 2-fluoro-2-deoxy-D-glucose (FDG)-PET/computed tomography (CT) could be used for staging of patients with progressive MTC, with possible prognostication by SUV quantification,14 in contrast with a similar study that did not consider the aggressiveness of the disease and concluded that FDG PET scan had low sensitivity and a low prognostic value.15
Imaging Workup Before Systemic Treatment
Two imaging studies have been performed in 2 series of progressive MTC patients.14, 16 In a first series of 35 patients, neck metastases were detected in 81% of patients, mediastinum metastases in 54%, lung metastases in 34%, liver metastases in 43%, and bone involvement in 74%.16 In the second series of 33 patients, metastases were located in the neck in 70% of patients, in the mediastinum in 63%, in the lungs in 31%, in the liver in 56%, in the bone/bone marrow in 69%, and in other sites (brain, heart, and axillary and celiac lymph nodes) in 19%.14 Several imaging methods should be proposed for a patient with an abnormal residual calcitonin level persisting after complete surgery: ultrasonography and CT for neck exploration, and CT for chest, abdomen, and pelvis. Magnetic resonance imaging (MRI) appears to have an advantage over CT for the detection of liver metastases from endocrine tumors.17 Moreover, MRI appears to be a sensitive imaging technique for detecting spread to bone/bone marrow.14 MRI has a higher sensitivity than bone scintigraphy, which detects bone involvement at a relatively advanced stage of tumor infiltration, when an osteoblastic reaction has occurred, but bone scintigraphy has the advantage of evaluating the whole skeleton. 18F-FDG-PET also appears of interest for the detection of disease throughout the body.14 In this study, the sensitivity of FDG-PET/CT in progressive MTC patients was 83% for neck, 85% for mediastinum, 75% for lung, 60% for liver, and 67% for bone metastases; the overall sensitivity was 76%. Moreover, SUVmax correlated significantly with calcitonin DT (P = .011) and with minimal DT (the minimum of CEA and calcitonin DT) (P = .027). 18F-Dihydroxyphenylalanine-PET is another functional whole-body imaging procedure that seems to provide interesting results in neuroendocrine tumors.18
RIT and Pretargeted Radioimmunotherapy
A few MTC patients have been enrolled in early RIT clinical trials assessing radiolabeled anti-CEA monoclonal antibodies in mixed populations of patients with cancers expressing CEA. Then, 1 phase 1/2 study was conducted specifically on MTC treated with the F(ab′)2 fragment of an anti-CEA monoclonal antibody (MoAb) labeled with iodine-131.19 Myelosuppression was the dose-limiting toxicity. Of the 12 assessable patients, 7 showed reduction of tumor markers, and 1 showed a 45% reduction of overall tumor burden. Disease was considered radiologically stable in 11 of the 12 patients for periods ranging from 3 to 26 months. Similar results were obtained in a second study evaluating high-dose RIT combined with autologous hematopoietic stem cell rescue in patients with rapidly progressing metastatic MTC.20 Except for 1 case of grade 3 gastrointestinal toxicity, nonhematologic toxicity was relatively mild up to the 1200 centigray (cGy) dose to critical organs. Of the 12 patients, 1 had a partial remission for 1 year, another had a minor response for 3 months, and 10 had stabilization of disease lasting between 1 and 16 months.
For radioresistant solid tumors such as MTC, pretargeting strategies have been proposed to increase the therapeutic index when compared with RIT using directly labeled MoAb and to increase the absorbed doses delivered to tumor cells.21 An unlabeled antitumor immunoconjugate is injected first. Later, when the immunoconjugate has sufficiently cleared from the circulation, the radionuclide is injected, coupled to a rapidly clearing agent with high affinity for the immunoconjugate prelocalized in the tumor. Interesting clinical results have been obtained with the avidin/biotin system, using 1 or several chase steps to clear excess high-affinity (Kd = 10−15 mol/L) antibody/avidin complexes.22 The Affinity Enhancement System is another pretargeting technique that uses a bispecific antibody (BsmAb) and a radiolabeled bivalent hapten. In this system, the affinity of the hapten for the BsmAb is limited (Kd = 10−8 mol/L), but the bivalent hapten binds avidly to the immunoconjugate bound to the surface of target cells, whereas hapten-BsmAb complexes in the circulation dissociate, and excess hapten is cleared, at least in part, through the kidneys. Increased tumor-to-normal tissue ratios and reduced toxicity, which may be further improved by adding a chase step, have been demonstrated in RIT studies in animals bearing human MTC (TT cells) xenografts.23-25 Affinity Enhancement System RIT using anti-CEA × antidiethylenetriamine pentaacetic acid (anti-DTPA) murine BsmAb (F6-734) and di-DTPA-131I hapten has been evaluated in MTC animal models with interesting results. A study in CEA-expressing MTC xenografts, comparing toxicity and efficacy of iodine-131-labeled F(ab”)2 and pretargeted bivalent hapten, showed that pretargeted RIT was as efficient as the directly labeled antibody and markedly less toxic.23 Repeated treatments with Affinity Enhancement System agents increased efficacy without increasing toxicity.26 These data allowed implementation of a clinical evaluation of this new therapeutic approach.
A first feasibility study was conducted in 5 patients with persistent disease or recurrence of MTC.27 The purpose of this study was to estimate the dose delivered to tumor targets and normal tissues after injections of a murine BsmAb anti-CEA × anti-DTPA and the di-DTPA-131I hapten. Dosimetric results showed that small lesions received potentially tumoricidal irradiation (up to 4.7 cGy/MBq), a dose comparable to that delivered by iodine-131 therapy to neck metastases of differentiated thyroid carcinoma (1.2-3.8 cGy/MBq for lesions of 8-40 g).28
A phase 1/2 clinical trial with escalating doses was implemented in 1996 to evaluate toxicity, pharmacokinetics, dosimetry, and antitumor activity of the Affinity Enhancement System, using the murine bispecific antibody F6-734 and di-DTPA-131I, in 26 patients with recurrence of MTC.29 Twenty to 50 mg of BsmAb and 1.5 to 3.7 GBq of di-DTPA-131I were injected 4 days apart. Immunoscintigraphy showed good targeting of tumor sites suspected by conventional imaging. The biological half-life in tumors ranged from 3 to 95 days, and tumor doses ranged from 0.1 to 5.0 cGy/MBq. The estimated tumor-to-nontumor dose ratios were 43.8 ± 53.4, 29.6 ± 35.3, 10.9 ± 13.6, and 8.4 ± 10.0 for total body, red marrow, liver, and kidneys, respectively. Dose-limiting toxicity was hematological, and maximum tolerated activity was estimated at 1.8 GBq/m2 in the group of patients with suspected bone marrow involvement. Preliminary data analysis showed a significant analgesic effect in 5 cases, 5 with minor tumor responses, and 4 having biological responses. The therapeutic responses were observed mainly in patients with a small tumor burden and after repeated courses of RIT. Nine patients developed human antimouse antibody (HAMA).
Because hematological toxicity was relatively high and immune responses frequent, further optimization of the treatment including the development of chimeric or humanized BsmAb was considered necessary. A prospective phase 1 optimization study was designed to determine optimal BsmAb dose, hapten activity, and pretargeting interval.30, 31 Thirty-four patients with CEA-expressing tumor were enrolled (non-MTC, 25 patients; MTC, 9 patients). These patients received escalating doses of BsmAb hMN-14 (humanized anti-CEA antibody) × m734 (murine anti-DTPA antibody) and escalating activities of radiolabeled di-DTPA-131I 5 or 7 days later. A BsmAb dose of 40 mg/m2 with a pretargeting interval of 5 days appeared to be a good compromise between toxicity and efficacy. The rate of tumor stabilization was 45% in the 1-year assessment. HAMA elevation was observed in 8% of patients and human antihuman antibody in 33%.
Six years after the first pretargeted RIT phase 1/2 study and 3 years after the second, long-term disease stabilization was observed in 53% of the MTC patients, as documented by morphological imaging (CT, MRI) and serial calcitonin and CEA serum measurements. Therefore, a retrospective study was conducted to compare the survival of 29 patients given pretargeted RIT with that of 39 contemporaneous untreated patients for whom data were collected by the French Tumor Endocrine Group.13 A second objective was to examine whether post-pretargeted RIT variations of calcitonin DT could be used as a surrogate marker for survival by comparing, among treated patients, the survival of biological responders and nonresponders, defining a responder as showing at least a 100% increase in calcitonin DT. OS was significantly longer in high-risk (calcitonin DT <2 years) treated than in high-risk untreated patients (median OS, 110 months vs 61 months; P < .030). Forty-seven percent of patients, defined as biological responders, experienced significantly longer survival than nonresponders (median OS, 159 months vs 109 months; P < .035) or untreated patients (median OS, 159 months vs 61 months; P < .010). Treated patients with bone/bone marrow disease had a longer survival than patients without such involvement (10-year OS of 83% vs 14%; P < .023). Toxicity was mainly hematological and related to bone/bone marrow tumor spread.
After the encouraging results obtained in the 2 phase 1/2 studies, a phase 2 pretargeted RIT study was developed, with the purpose of evaluating the response rate, time to progression, and overall survival in progressive MTC patients (calcitonin DT <5 years). Forty-eight patients have been included between April 2004 and January 2008. Forty-five patients have been treated, receiving 40 mg/m2 of hMN-14xm734 BsMAb and 1.8 GBq/m2 131I-di-DTPA 4-6 days apart. Six patients were re-treated. A preliminary analysis of results was performed in September 2008, in 33 evaluable patients (20 males and 13 females) who received 35 treatments, with 15 months follow-up (6 to 36 months). A patient was considered unresponsive if progression according to Response Evaluation Criteria in Solid Tumors, FDG-PET, or biomarkers serum concentration was observed at 3 months post-RIT, or no effect on CEA or calcitonin DT (at least 100% CEA or calcitonin DT variation) was obtained.
In the 33 evaluable patients, the median pre-RIT calcitonin serum level was 2698 pg/mL (range, 193-83,652), the median CEA serum level was 52.2 ng/mL (range, 1.3-3813.5), calcitonin DT was 1.2 years (range, 0.2-5.9), and CEA DT was 1.8 years (range, 0.5-23.8). Tumor targeting was obtained in all cases. The sensitivity of immunoscintigraphy was 92%. Figure 1 shows the good tumor uptake observed in an MTC patient with mediastinal lymph nodes and lung involvement. Allergic reactions were observed during 2 BsMAb perfusions, grade 1 liver toxicity after 3 of 35 (8.5%) injections, and grade 3/4 hematological toxicity after 19 of 35 (54%) injections. Efficacy was observed after 18 of 35 (51%) injections, with a time-to-progression of 18 months (range, 6-36) for Response Evaluation Criteria in Solid Tumors and 15 months (6 to 36) for PET and biomarker levels. In this series, 24% (8 of 33) of patients were considered in the low-risk group (DT, >2 years), 58% (13 of 33) in the intermediate-risk group (DT, 6 months to 2 years), and 18% (6 of 33) in the high-risk group (DT, <6 months). Efficacy was found in 62% of patients in the low-risk group, 53% in the intermediate-risk group, and 50% in the high-risk group. Figure 2 shows an example of the biological response obtained in a progressive MTC patient with diffuse bone marrow involvement.
Other Treatment Modalities
Chemotherapy in MTC
A limited number of clinical studies, enrolling small numbers of patients and using different drugs or combination of drugs, have been performed >10 years ago. Considering a total of 87 patients enrolled in 4 trials and treated with different chemotherapeutic regimens, progression-free survival, reported for only 22 patients, ranged from 4 to 29 months (median, 10 months). The OS reported for 20 patients ranged from 8.5 to 33+ months (median, 17.5 months).13 In the absence of data on pretreatment prognostic indicators of survival for treated patients, it is difficult to draw any valid conclusions on the real treatment effectiveness, because it has been well established that some patients, even with progressive metastatic disease, can benefit from long periods of survival in the absence of treatment. Moreover, severe toxicity has been reported with some combination chemotherapy regimens.32 In this situation, chemotherapy is not considered a valid therapeutic option for patients with advanced disease.
Chemoembolization for liver metastases
Metastatic extension is often disseminated in multiple organs, such as bones, lungs, and liver, but in some patients metastases are predominantly localized in the liver. In this clinical situation, selective intra-arterial chemoembolization using different drugs and a variable number of courses (1-9) has been performed in 23 patients in 2 studies.33, 34 Some transient (median duration, 24 months) partial remission or stabilization was observed in 70% (16 of 23) of cases with good symptom palliation. This interesting effectiveness was mainly observed in patients with limited liver involvement (<30%). Thus, it appears that chemoembolization can be useful for a small percentage of patients who have a metastatic extension limited to a small part of the liver. However, in our MTC patient series, no patient showed metastatic disease limited to the liver.14, 16
Therapy with multikinase inhibitors
Among signal transduction pathways that lead to neoplastic transformation, the RET protein plays a major role in MTC. Somatic mutations are observed in most familial forms and in 30% to 50% of sporadic forms.35 Consequently, RET appeared as a favorable target for molecular therapy of MTC even if a substantial number of patients with sporadic form could not benefit. Other signaling components, including vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and platelet-derived growth factor (PDGF) receptors, can be involved in MTC, and multikinase inhibitors targeting 1 or more of them have been evaluated in preclinical and clinical trials.
Imatinib mesylate, which inhibits RET among other protein-tyrosine kinases, has been evaluated in patients with advanced MTC. Three clinical studies enrolled a total of 30 patients.36-38 Stabilization was observed in 30% (9 of 30) of patients over 6 to 24 months. Severe toxicity, including rash, fatigue, laryngeal mucosal swelling, nausea, and vomiting, was reported in 1 study.37
More recently, various protein-kinase inhibitors have been evaluated in patients with advanced MTC, and some preliminary results are only available in abstract form. Axitinib, an inhibitor of VEGF receptors 1, 2, and 3, was evaluated in 60 patients with advanced thyroid cancers, including 12 patients with MTC.39 Stabilization was observed in 50% of all patients for up to 13 months without any differentiation for patients with MTC. Sorafenib, which selectively inhibits RET tyrosine kinases, was evaluated in 5 patients with metastatic MTC.40 Surprisingly, 1 complete response was observed together with 1 partial response with a 50% decline of calcitonin in all 5 patients, but the initial dose had to be reduced by 50% because of serious side effects. Vandetanib targets RET, VEGF receptor, and EGF receptor tyrosine kinases and was evaluated in 30 patients with locally advanced or metastatic hereditary MTC.41 A partial response was observed in 20% (6 of 30) and stabilization in 30% (9 of 30) of patients for up to 9 months. A biological response was observed in 63% (19 of 30) for at least 6 weeks. Some grade 3 adverse events, including rash and diarrhea, were observed in 3 patients. Finally, motesanib, which targets all known VEGF, PDGF, Kit, and Ret receptors, was evaluated in 83 patients with both hereditary and sporadic MTC.42 A partial response was observed in 2 patients, and stabilization for no longer than 6 months in 43 (52%).
All these studies performed in patients with locally or metastatic advanced MTC and using various multikinase inhibitors have shown some tumor effect documented by a substantial decrease of calcitonin. A transient stabilization of the disease was observed, extending up to 24 months. However, in the absence of data on real tumor growth rate before treatment, it is not possible to draw any valid conclusions regarding survival benefit. Indeed, it is well known that tumor growth rate in patients with metastatic MTC is highly variable and, for some patients, progression can be so slow that it simulates stabilization extending over months and even years.
As mentioned above, we have shown that calcitonin DT is the best prognostic indicator to select patients with rapidly growing metastatic disease before any investigational treatment.12 In patients with calcitonin DT >5 years, life expectancy is very long, and watchful waiting may be the most appropriate strategy. For patients with calcitonin DT<2 years and, all the more, <6 months, the disease is rapidly progressing, with a short life expectancy warranting treatment with an investigational drug.
To the best of our knowledge, there currently is no approved drug for the systemic treatment of metastatic MTC. No real survival benefit has been convincingly documented with chemotherapy, which furthermore was associated with severe toxicity. It is too early to evaluate the potential effectiveness of multikinase inhibitors. Reported results of phase 2 trials have shown some transient stabilization, but more patients with documented rapidly progressive metastatic disease should be included and their survival compared with that of comparable untreated patients of historical studies. It is appreciated that the optimal comparison should be made in randomized prospective trials, but because of the low frequency of MTC, it would take many years to enroll a sufficient number of patients, especially if only patients with short calcitonin DT are included. Additional years would be necessary for survival analyses. Up to now, pretargeted RIT has been the only innovative treatment modality convincingly showing some survival benefit in patients with rapidly progressing metastatic disease. These results have been obtained in a first phase 2 study including 29 patients, and confirmed in a second phase 2 study of 33 patients, for a total number of 62.
CONFLICT OF INTEREST DISCLOSURES
The papers in this supplement represent proceedings of the “12th Conference on Cancer Therapy with Antibodies and Immunoconjugates,” in Parsippany, New Jersey, October 16-18, 2008. Unrestricted grant support for the conference was provided by Actinium Pharmaceuticals, Inc.; Bayer Schering Pharma; Center for Molecular Medicine and Immunology; ImClone Systems Corporation; MDS Nordion; National Cancer Institute; National Institutes of Health; New Jersey Commission on Cancer Research; and PerkinElmer Life & Analytical Sciences. The supplement was supported by an unrestricted educational grant from ImClone Systems Corporation, a wholly-owned subsidiary of Eli Lilly and Company, and by page charges to the authors. Dr. Goldenberg is a director and shareholder in Immunomedics, Inc. and IBC Pharmaceuticals, Inc.