[D1] WHAT ARE DIRECTIONS FOR FUTURE RESEARCH?

[D2] Novel therapies and clinical trials
While surgery and the judicious use of RAI, as described in these guidelines, is sufficient treatment for the majority of patients with DTC, a minority of these patients experience progressive, life-threatening growth and metastatic spread of the disease. The recent explosion of knowledge regarding the molecular and cellular pathogenesis of cancer has led to the development of a range of targeted therapies, now undergoing clinical evaluation. Efficacy has already been demonstrated for several agents in phase II studies, including axitinib, motesanib, sorafenib, pazopanib, and thalidomide, whereas many others are in ongoing trials. Randomized phase III trials to demonstrate improved survival, improved progression free survival, or superiority of one therapy over another have not been performed, however, and none of these drugs have been specifically approved for treatment of metastatic thyroid carcinoma. These therapies can be grouped into a number of categories.

[D3] Inhibitors of oncogenic signaling pathways.   Tyrosine kinase inhibitors of interest in thyroid carcinoma usually target transmembrane tyrosine kinase receptors that initiate signaling through the MAP kinase pathway. This signaling pathway is activated in the majority of PTCs. Inhibitors of RET, RAS, RAF, and MEK kinases target various members of the same signaling pathway. Several of these agents are in development with several clinical trials completed or underway. Specific oncogene targeting for follicular thyroid cancer and Hu¨ rthle cell cancer awaits better understanding of the pathways involved in initiation of these tumor types, although responses in patients with these subtypes have been reported in clinical trials.

[D4] Modulators of growth or apoptosis.   Key components of growth and apoptotic pathways are targeted by PPARγ activators, including COX2 inhibitors; rexinoids, which activate RXR; bortezomib, which inactivates the cancer proteasome; and derivatives of geldanomycin, which target the hsp-90 protein. Clinical trials in thyroid cancer of each of these agents are available.

[D5] Angiogenesis inhibitors.   Targeting of vascular endothelial growth factor (VEGF) receptors and other members of the signaling cascades responsible for neoangiogenesis may limit the growth of cancers by restricting their blood supply. Many of the kinase inhibitors that have been studied to date are very potent inhibitors of the tyrosine kinase of the VEGF receptors. Trials of several of these agents are currently underway in all subtypes of thyroid cancer.

[D6] Immunomodulators. Stimulation of the immune response to cancer may be achieved by augmenting the activity of antigen-presenting dendritic cells. This approach has shown possible benefits in phase I clinical trials, but has not yet been studied in thyroid cancer. The apparent immunogenicity of thyroid cells makes this an attractive approach for future clinical trials.

[D7] Gene therapy.   Preclinical studies have demonstrated some efficacy in thyroid cancer cell lines. Approaches include introducing toxic genes under the control of thyroid-specific promoters, or restoration of the p53 tumor suppressor gene in anaplastic thyroid cancer cell lines. Problems with gene delivery limit the clinical utility of these approaches, which have not yet reached clinical trials in thyroid cancer.
Each of these targeted approaches holds promise for our future ability to treat patients with life-threatening disease unresponsive to traditional therapy. In the meantime, for appropriate patients, entry into one of the available clinical trials may be an attractive option.

[D8] Better understanding of the long-term risks of RAI
With the more widespread use of RAI in the management of thyroid cancer, and the normal life expectancy of most patients with the disease, it is imperative that we have a better understanding of the long-term risks associated with its use. Research that focuses on how to minimize the impact of RAI on the salivary glands in order to prevent sialadenitis and xerostomia would provide a significant benefit to patients. A better understanding of the long-term effects of RAI on reproductive issues in men and women is also an important topic. Finally, while the risk of second malignancies appears small following the usual activities of RAI used for remnant ablation, we need better understanding of the long-term risks for salivary gland tumors, bladder tumors, and colon cancers when repeated doses of RAI are needed in young patients who are potentially long-term survivors of thyroid cancer.

[D9] Clinical significance of persistent low levels of serum Tg
After initial surgery and RAI therapy some patients will have persistently detectable stimulated serum Tg when evaluated 9–12 months later. Most of these patients have stimulated Tg levels in the range of 1–10 ng/mL, levels typically associated with a small volume of tissue. Some of these patients demonstrate a subsequent spontaneous fall in Tg over time, others remain stable, while still others demonstrate rising Tg levels. The optimal management of these patients is unknown. How often should they undergo neck US or stimulated serum Tg testing? Will sensitive Tg assays combined with neck US replace stimulation testing? Which (if any) of these patients should undergo chest CT, PET, or empiric RAI therapy? Can we improve our abilities to predict and monitor which patients are likely to be harmed by their disease as opposed to those who will live unaffected by theirs? Does metastatic disease in small local lymph nodes have the potential to metastasize to distant sites during observation while on TSH suppression therapy? The current impetus to test and treat all of these patients is based on the argument that early diagnosis may lead to early treatment of residual disease when treatment is more likely to be effective, as opposed to less effective treatment when the tumor is more bulky, more extensive, or has spread to inoperable locations. However, there is no current proof that aggressive treatment of minimal residual disease improves patient outcome. This is brought into focus by the fact that only about 5% of all PTC patients die of their disease, yet 15–20% of low-risk PTC patients are likely to have persistent disease based on persistent measurable Tg with stimulation testing.

[D10] The problem of Tg antibodies
Anti-Tg antibodies are a common clinical problem in patients with DTC (305). The presence of these antibodies usually interferes with serum Tg measurement and recovery assays do not appear to accurately predict this interference (305,428). Decreasing antibody levels are correlated with “disease-free” status while increasing levels suggest persistent disease (306,429). However, there are clear exceptions to this “rule.” These patients are therefore a challenge to manage or study because one often can not be certain of their disease status. This problem limits definitive investigation which, in turn, hampers development of evidence-based guidelines such as these to assist clinicians. Measurement of Tg mRNA in the blood may be a sensitive marker for persistent thyroid cells even in the presence of anti-Tg antibodies (430–432), but RNA extraction is not well standardized and some studies question the specificity of this marker (433,434). Future studies optimizing the measurement of Tg mRNA and perhaps other thyroid-related substances in blood from DTC patients with anti-Tg antibodies are needed to better monitor this challenging subgroup of DTC patients. This goal would also be enhanced by development of Tg assays that have limited interference by anti-Tg antibodies and by methods to clear anti-Tg antibodies prior to Tg measurement.

[D11] Small cervical lymph node metastases
The rates of cervical lymph node metastases generally range from about 20% to 50% in most large series of DTC, with higher rates in children or when micrometastases are considered. The location and number of lymph node metastases is often difficult to identify before, during, or after surgery, especially micrometastases. Although postoperative 131I given to ablate the thyroid remnant undoubtedly destroys some micrometastases, the most common site of recurrence is in cervical lymph nodes, which comprise the majority of all recurrences. Future research must consider the dilemma of minimizing iatrogenic patient harm versus preventing cancer morbidity and (perhaps) mortality. Perhaps techniques will be developed to safely remove or destroy small cervical nodal metastases, which in some cases would otherwise progress to overt, clinically significant metastases. Conversely, the clinical significance of very small (<0.5 cm) nodal metastases needs to be clarified by long-term follow-up studies. Development of a cost-effective method to determine which metastases can be safely followed without intervention would be of great benefit.

[D12] Improved risk stratification
Current risk stratification schemes rely almost exclusively on clinical, pathological, and radiological data obtained during the initial evaluation and therapy of the patient. However, none of the commonly used risk stratification schemes adequately incorporate the prognostic implications of the very detailed pathological descriptions that are provided (e.g., various histological subtypes of thyroid cancer, frequent mitoses, areas of tumor necrosis, minor degrees of extrathyroidal extension, or capsular invasion) or the molecular characteristics of the primary tumor. Furthermore, current staging systems are static representations of the patient at the time of presentation and are not easily modifiable over time as new data become available during follow-up. Therefore, a risk stratification system that incorporates all the important information available at presentation and also evolves over time as new data become available would be useful in providing ongoing risk assessments that would optimize management throughout the life of the patient.