In NMTC tumor cells, retinoids exert several re-differentiating effects: induction of 5’ deiodinase (Van Herle et al. 1990; Schreck et al. 1994) increased expression of the differentiation marker, alkaline phosphatase, or decreased expression of CD97, which is highly expressed in anaplastic thyroid carcinoma (Hoang-Vu et al. 1999), as well as stimulation of intercellular adhesion molecule-1 synthesis (Bassi et al. 1995). Interestingly, the promoter of the NIS gene has a retinoic acid response element (Schmutzler et al. 2002). Experimental data demonstrate that retinoids can reinduce NIS mRNA expression (Schmutzler et al. 1997) and radioiodine uptake (Elisei et al. 2005). Although leading to partial rather than complete thyroid cancer cell re-differentiation, these effects have been hoped to render radioiodine uptake and enable treatment of NMTC foci (Fig. 1).

The promising experimental results and their potential therapeutic impact prompted a number of centers including ours to initiate clinical studies to evaluate the effectiveness of 13-cis-retinoic acid (13-CRA) as re-differentiation therapy (RDT) in patients with advanced radioiodine non-avid thyroid cancer. This compound is a ligand for the retinoic acid receptor (RAR). The therapy is generally well tolerated and the most frequent side effects are dryness of skin and mucosa, and hypertriglyceridemia (Coelho et al. 2005). The clinical studies have confirmed that 13-cis retinoic acid therapy leads to restoration of radioiodide uptake (Grunwald et al. 1998a, b; Simon et al. 1998) and in some cases even to tumor regression (Boerner et al. 2002; Simon et al. 2002; Gruning et al. 2003; Haugen 2006; Oyen et al. 2007). Overall induction or increase in I-131 uptake is observed in about 30 % (range 6–62 %) of cases (Table 2). However, this increased uptake did not transfer to clinical results and most studies objective tumor regression was not achieved (Courbon et al. 2006; Handkiewicz-Junak et al. 2009; Kim et al. 2009). It has to be noted, however, that none of these studies used RECIST criteria to assess radiological response. Evaluation of biochemical response (change in serum thyroglobulin level) is complicated by at least two factors. First, if patients are treated under endogenous stimulation, serum Tg, before and after 13-CRA treatment, is measured under confounding TSH factor (first TSH suppression than stimulation) that precludes direct comparison. The second question is, how biochemical response should be defined. Should it be defined as an increase in Tg level directly after 13-CRA treatment, or should it be decrease in Tg level during follow-up. The first situation is regarded as a sign of tumor redifferentiation. The latter if accompanied by tumor regression or stagnation in radiological examination can be regarded as decrease in cancerous cells. The interpretation of an “increased” radioiodine uptake also carries certain limitations since a qualitative assessment of the iodine avidity is prone to a lack of reproducibility and more sophisticated individual dosimetry was in most cases not part of the published studies. All in all the above-mentioned studies have so far failed to translate into a marked clinical effect and response rates of the order of 20–30 % can be expected (Simon et al. 2002; Gruning et al. 2003).

One of the reasons for this lack of success could be the specificity of 13-cis RA. There are some in vitro data (Schmutzler et al. 1997; Haugen et al. 2004) suggesting that a differential expression of both RAR and the retinoid receptor RXR in thyroid carcinoma cell lines and tissues is present (Haugen et al. 2004; Elisei et al. 2005), which corresponded to the responsiveness to ligands for these receptors. A novel drug tested in DTC redifferentiation therapy is Bexarotene, an RXR agonist, which also has affinity to and induces transcriptional activation of RAR. The anti-neoplastic potential has been demonstrated not only in cutaneous T cell lymphoma (Farol and Hymes 2004), but also in other malignant tumors (Rigas and Dragnev 2005). In DTC 6 weeks of Bexarotene treatment was able to reinduce I-131 uptake in 8 out of 11 (72 %) patients with metastatic DTC (Liu et al. 2006); nevertheless, the uptake was only discernable at SPECT and not present in all metastases, visualized by CT scanning. There were no complete or partial responses and after 6-month follow-up most treated patients suffered from disease progression. To complicate the even more the interpretation of the clinical usefulness of RA in thyroid cancer, none of the published studies were randomized, blinded study and effectiveness parameters were not uniformly studied (Table 2). Also, none of the studies analyzed progression-free or overall survival. Therefore, the question on clinical validity of retinoids in differentiated thyroid carcinoma still awaits confirmation.

Controlled clinical trials to further establish criteria for the pretherapeutic selection of patients who will benefit from RA therapy and to quantify tumor doses achieved in relation to clinical outcome are still sparse. For the future of these agents a more sophisticated approach could probably be based on the determination of the individual receptor status, one of the primary research goals to be addressed (Haugen et al. 2004; Schmutzler et al. 2004). However these developments have to be discussed in the presence and future advent of other therapeutic target developments in general oncology impacting on up-to-date thyroid cancer management.