TSH stimulates thyroid cell proliferation and thus has long been hypothesized to play a potential role in thyroid cancer recurrence. Consequently, TSH suppression with levothyroxine in doses above those required for ‘replacement’ has been proposed in the management of thyroid cancer. While some conflicting evidence exists, larger studies and meta-analyses have been able to demontrate a link between TSH and a reduction in adverse outcomes [1]. In a study of 2036 patients registered in a multi-institutional thyroid cancer registry, Jonklaas et al. demonstrated improved overall survival in patients with stage II, III or IV disease [4]. Interestingly, whereas a dose-response relationship with the level of TSH suppression was observed in those with more advanced disease (stages III and IV), suppression of the TSH beyond the initial subnormal range in stage II disease did not produce a further incremental benefit. Currently, the ATA recommends aggressive suppression of TSH (<0.1 mU/L) in people with persistent disease but a more conservative approach in those with high-risk (0.1–0.5 mU/L) and low-risk (0.3–2 mU/L) disease in order to prevent recurrence [1]. Clearly, the benefit derived from TSH suppression must be balanced against the potential detrimental effects on cardiac and bone health in each patient.

RAI ablation following surgery is frequently employed in the primary treatment of DTC. Potential benefits include: (i) facilitating cancer surveillance with serum Tg by ablation of remnant thyroid tissue; (ii) potential reduction in risk of recurrence/disease-specific mortality; and (iii) identification and treatment of persistent disease [1]. When persistent/recurrent metastatic disease is identified, RAI is usually the preferred initial treatment as long as the disease remains iodine-avid [1, 7, 12, 18, 22, 30, 31, 33, 34, 58–61]. The published data on this topic, whereas largely retrospective, does suggest that RAI is of utility in guiding prognosis and might confer a survival benefit.

In addition to iodine avidity, a number of other factors appear predictive of the response to RAI. Durante et al. published a series of 444 patients with distant metastatic disease from PTC or FTC [18]. Of these, 295 were found to have iodine-avid metastatic disease. All patients were treated empirically with 100 mCi. In the setting of persistent iodine-avid disease, treatments were then repeated at 3–12-month intervals until negative studies were obtained. No fixed limit on maximum cumulative dose was set. Overall, 127 patients achieved negative imaging studies during follow-up. A large proportion of these occurred later in the course of therapy (>5 years). In this subset of patients, 10-year survival was 92 % compared to 29 % in those with iodine-avid disease and persistently positive imaging. The 10-year survival of patients without ¹³¹I at the outset was just 10 %. The authors examined the characteristics of patients in the most favourable group and found that they were more likely to be young, (<40 years) with PTC, and pulmonary site-only metastases with no findings or micronodular disease on cross-sectional imaging [18]. A number of additional studies have published similar findings supporting the benefit of RAI [6, 7, 10, 15, 22, 34, 58, 61].

Although not universal’ most studies seem to indicate that pulmonary disease responds more favourably to RAI than metastases at other sites [7, 15, 18, 22, 36, 61]. Most studies show that the rate of RAI avidity (~67 %) is similar between lung and bone metastases. However’ bone metastases appear to be more advanced at the time of diagnosis as most are visible on plain radiographs or cross-sectional imaging [18, 36]. Thus’ it would be more reasonable to expect outcomes comparable to those seen in macronodular pulmonary disease. In these subsets’ evidence continues to suggest a survival benefit associated with RAI treatment but complete responses are rare [10, 13, 17, 18, 22, 23, 33, 35, 59–61]. As a result’ additional therapies play a larger role in the treatment of bone disease.

Current recommendations strongly support the use of RAI in patients with micronodular, iodine-avid disease. In these patients repeated doses should be given at 6–12 month intervals while the disease continues to concentrate iodine and until negative results are achieved [1]. In addition macronodular pulmonary and bone disease may be treated as long as objective benefits are observed [1]. A single safe dose limit has not been defined and the decision to terminate RAI treatment should be individualized to the patient.

The use of therapeutic dose RAI in patients with non-avid disease on diagnostic scans has been an area of considerable controversy. A number of studies have been able to demonstrate an increased sensitivity of post-therapeutic (>100 mCi) whole-body scans compared to diagnostic (<6 mCi) studies [37, 40–42, 48]. Pacini et al. demonstrated uptake on 30/42 post-therapeutic scans in this clinical setting [42]. Those 30 patients were given additional RAI treatments. A mean of three treatments per patient were administered (cumulative range of doses 90–500 mCi) over a period of 6.7 years. Unfortunately’ when compared to an untreated group of 28 similar patients’ no significant difference in Tg levels or clinical outcomes were observed. More recently’ a systematic review published by Ma et al. concluded that insufficient evidence existed to support this approach [40]. Consistent with this assertion’ the updated ATA guidelines do not recommend the use of RAI in these groups [1].

The positive outcomes associated with RAI in appropriate patients and the lack of efficacy seen with other treatments for metastatic disease have led to interest in agents that might enhance iodine uptake and retention in non-avid tumours [12, 43–47, 62, 63] Impaired iodine (¹³¹I) uptake appears to be related to diminished cell membrane expression of the sodium-iodide symporter. Retinoic acid’ via its binding to the retinoic acid receptor’ regulates transcription of genes related to cell differentiation. In vitro studies demonstrate some extent of redifferentiation with these compounds [64, 65]. As a result, cells increase their expression of the symporter and take up iodine more efficiently. Unfortunately, clinical studies have failed to show clinical benefit consistently. Courbon et al. reported a series of 11 patients with metastatic DTC and no uptake on whole-body scans [43]. All were treated with 13-cis-retinoic acid. RAI uptake, a surrogate clinical marker of redifferentiation, increased marginally in merely 2 patients. After a mean follow-up of 24 months, 5 patients had died, 5 showed evidence of progression and only one had stable disease. Zhang et al. used another isomer—all-trans-retinoic acid—in 11 patients with advanced-stage DTC and poor RAI uptake [47]. Increase in RAI uptake was seen in 4 patients and diminished uptake in 6. Partial responses were observed in 5 patients and progressive disease was noted in 4 of them. Currently, the use of these compounds outside of a clinical trial cannot be recommended [1].

Lithium has been shown to prolong the duration of RAI uptake by inhibiting its release from the cell [62]. Unfortunately, this effect is not mediated through increase in sodium-iodide symporter expression and thus is only of potential benefit in tumours that are already iodine-avid. Rosiglitazone, a thiazolinedione drug used in patients with diabetes, has shown some early promise in thyroid cancer redifferentiation [44–46]. Rosiglitazone is a peroxisome proliferator-activated receptor gamma agonist. This transcription factor is a regulator of cellular differentiation and has been shown to inhibit tumour development in animal models and human tissue cell lines. Phase II studies have demonstrated some efficacy in increasing RAI uptake in patients with metastastic DTC and negative whole-body scans [44]. Kebebew et al. reported the clinical results of rosiglitazone in 20 patients with DTC and negative RAI whole-body scans [45]. Five patients demonstrated an increase in RAI uptake after treatment. A reduction in Tg level post-therapy was seen in 3 patients. None had a partial response by RECIST criteria and 7 had progressive disease. Thus, whereas this drug shows promise with respect to tumour redifferentiation and increased RAI uptake, significant clinical benefit has yet to be demonstrated. Finally, gene therapy with the introduction of the sodium-iodide symporter gene into tumour cells using an adenovirus vector has been investigated, but its efficacy seems to be limited by short duration of action [63, 66].