TSH-suppressive therapy (ST) is part of the treatment protocol for differentiated thyroid carcinoma (DTC). There is however little evidence of its clinical effectiveness. On the other side, ST is not free from side effects related to the subclinical hyperthyroidism status induced in patients for a long time period. Because of this, widespread use of ST in all patients has recently been questioned, and individualized treatment based on the characteristics of each particular case has been proposed.
Action of thyroid hormones (THs) depends on their binding to specific nuclear receptors. However, a second pathway mediated by a membrane receptor located in integrin αVβ3 has recently been established. It has been postulated that the proliferative and angiogenic effects attributed to THs would depend on this second mechanism. It is not known whether the tumorigenic action shown in other neoplasms may have an impact on DTC.
The objective of this study was to review the relationship between ST and cancer, particularly as regards tumor evolution and occurrence of second primary neoplasms. One of the future challenges in the field of DTC will be to establish the specific role of TSH and THs in malignant thyroid cells, in order to be able to define more optimized therapeutic schemes.
El tratamiento supresor (TS) de la TSH forma parte del protocolo terapéutico en el carcinoma diferenciado de tiroides (CDT). No obstante, existe poca evidencia sobre su efectividad clínica. Además, esta práctica no está exenta de efectos secundarios relacionados con el estado de hipertiroidismo subclínico a la que se somete al paciente durante un largo periodo de tiempo. Por todo ello, recientemente se ha empezado a poner en duda que la TS deba ser aplicada de forma generalizada en todos los pacientes, proponiéndose un uso individualizado en función de las características de cada caso en concreto.
La acción de las hormonas tiroideas (HT) depende de su unión a receptores nucleares específicos. Pero en los últimos años se ha establecido una segunda vía mediada por un receptor de membrana presente en la integrina αVβ3. Se ha postulado que el efecto proliferativo e inductor de angiogénesis atribuido a las HT dependería de este segundo mecanismo. Se desconoce si la acción tumorogénica demostrada en otras neoplasias pueda tener repercusión en el CDT.
El objetivo del presente trabajo es revisar la relación entre TS y cáncer, especialmente en la evolución del tumor y en la aparición de segundas neoplasias primarias. Uno de los retos de futuro en el campo del CDT será establecer el papel específico de la TSH y de las HT en la célula tiroidea malignizada para poder definir unas pautas terapéuticas más optimizadas.
Treatment for differentiated thyroid carcinoma (DTC) is based on a triad consisting of total thyroidectomy, the ablation of thyroid remnants with 131I, and treatment with high thyroxine doses. This last strategy is what is usually known as “suppressive treatment” (ST), and was proposed as a helpful treatment more than 75 years ago1 when the administration of thyroid extracts was shown to improve the course of DTC metastases. TSH is the main factor in thyroid cell growth and differentiation. The trophic effect is maintained in DTC, and inhibition with high thyroxine doses therefore helps prevent the extension of residual cancer cells not killed after surgery and ablation.
However, this postulate has recently been challenged after the predominant role of TSH as a cell proliferation element, particularly in neoplastic cells, was questioned.2,3 On the other hand, new questions have arisen about the optimal treatment for DTC as the result of advances made in the past decade in our understanding of the molecular bases governing tumorigenesis. One of these questions refers to ST and the possibility that this treatment, apart from its known side effects, may induce proliferation of either DTC itself or second tumors. The objective of this study was to review the current state of ST based on these findings.
A weak scientific basisSince its progressive implementation, ST has been widely accepted and is part of all treatment protocols for DTC. Scientific evidence for its efficacy is however limited, and there are not many studies which support its clinical use.
In 2002, McGriff et al.4 published the first and only meta-analysis to date on the subject. These authors only analyzed the 10 articles they considered adequate for the study, which were otherwise not free from significant methodological limitations. Thus, three of them were from the same research group, so that patients could be included several times.5–7 The same was applied to two other studies.8,9 On the other hand, the Young et al.9 article only analyzed follicular carcinomas (FTC), and the one by Sanders and Rossi7 only included hidden carcinomas, i.e. those diagnosed at the metastatic stage. Moreover, the reported series included quite old DTC cases. For example, the Wanebo et al.10 study ended in 1976, the Cady et al.5 study included subjects treated for DTC between 1931 and 1970, the study by Sanders and Rossi7 analyzed a patient group monitored between 1940 and 1990, and the highly referenced series of Mazzaferri and Jhiang8 reported cases from 1950 to 1993. It is obvious that the diagnostic and therapeutic procedures performed on those patients cannot be extrapolated to current times. Finally, the number of cases in the different series was also small. Sanders and Rossi7 reported 92 patients, and the Pujol et al. study,11 one of the most commonly cited to show the efficacy of ST, was based on 121 subjects. However, based on the characteristics of all of these studies, the meta-analysis by McGriff et al.4 concluded that treatment with high thyroxine doses was effective, but that it was of little significance as regards improving the survival of patients with DTC.
Some of the studies included in the meta-analysis found ST to be effective under certain circumstances. Cooper et al.12 suggested that TSH inhibition should only be used in patients classified as high risk. Patient selection and the recognition that the widespread use of ST should be reconsidered gained strength following the review by McGriff et al. Thus, a 2006 study by Jonklaas et al.13 stratified the efficacy of this therapeutic approach. The authors found that in stage I, survival was not related to the extent of TSH suppression. In stage II, a direct relationship was seen in TSH levels higher than 3mU/L. In stages III and IV, a clear correlation was found between ST and survival. Moreover, Hovens et al.14 more recently defined a value of TSH of approximately 2mU/L as the cut-off point for best discriminating the risk of disease relapse. All of these led to an ST algorithm which was more rational and adapted to the characteristics of each individual patient being proposed.15 This new approach started to be seen in some guidelines and practical recommendations.16,17
The only randomized, prospective study conducted to date which assesses the efficacy of ST has recently been published. Sugitani and Fujimoto18 randomized more than 400 patients undergoing surgery for DTC into two groups. The first group was treated with thyroxine to achieve TSH suppression, while patients in the second group were treated to maintain TSH within the normal range. After a mean follow-up of 7 years, the authors found no significant difference between the two groups in regard to disease-free time, relapse, time of relapse, distant metastases, overall mortality, or specific mortality.
An strategy not devoid of riskTo the weak scientific evidence concerning the actual effectiveness of ST, it should be added that it is not devoid of side effects. There are several reports which analyze this aspect, including the excellent review by Reverter and Colomé19 recently published in this journal. The most significant harmful effects of ST are those derived from the subclinical hyperthyroidism chronically induced in patients, which often leads to a true, symptomatic clinical hyperthyroidism.
Thyroid hormones and cancerThe existence of a direct relationship between thyroid hormones (THs) and cancer was suggested more than a century ago.20 Different studies have found a significant association between hormone levels and the occurrence of various neoplasms, including kidney, pancreas, ovary, and breast tumors.21 In 1984, Brinton et al.22 reported that the risk of breast cancer increased more than 10-fold when replacement therapy with THs was started in hypothyroid women. A large scale epidemiological study conducted in Norway on more than 29,000 people monitored for nine years showed that TSH levels less than 0.5 mU/L were associated with increased cancer occurrence (hazard ratio, 1.34; confidence interval, 1.06–1.69).23 Malignant lung and prostate tumors were most common. By contrast, levels consistent with hypothyroidism did not increase the chance of tumor occurrence.23 In this regard, recent studies suggest that hypothyroidism may improve the effectiveness of anticancer treatments.21 Specifically, an increased progression-free time has been reported in patients with renal cancer who experience hypothyroidism following the administration of sunitinib and sorafenib.24,25 This has led to it being postulated that THs play some role in both tumor proliferation and angiogenesis. However, the pathophysiological basis of such effects may not be established for years.
A new paradigm in thyroid hormone actionIntegrins are a group of integral membrane heterodimers able to interact with various extracellular proteins, growth factors, and certain hormones to generate intracellular responses. More than 20 different integrins, resulting from the combination of various subtypes of the two subunits (alpha and beta) that form them, have been reported to date. In 2005, Bergh et al.26 reported that the integrin known as αVβ3 has a specific site that acts as an HT receptor. This altered the traditional concept that HTs, and specifically triiodothyronine (T3), only acted through nuclear receptors (TRs). On the other hand, various studies had already demonstrated that some TH actions were not mediated by TRs. Thus, effects due to interaction with traditional receptors were called “genomic actions”, and all other effects, “non-genomic actions”27. Studies subsequent to the Bergh et al.26 discovery confirmed the hypothesis that non-genomic actions were due to TH interaction with their surface receptor in integrin αVβ3.28 It should be noted that this integrin has a specific locus for T3 and another different locus for tetraiodothyronine (T4).
Integrin αVβ3 is expressed in endothelial and smooth muscle cells, but shows a particularly strong expression in the cell membrane of a great number of tumors, including breast, prostate, and liver tumors. The Bergh et al. study26 suggested that integrin activation by THs was responsible for the angiogenesis promoter action of THs, and that the T4-αVβ3 complex acted by activating the signaling pathway dependent upon mitogen-activated protein kinase (or MAPK pathway). Other studies have supported this hypothesis, and the relationship between cancer and THs, first suspected more than a century ago, is currently thought to be due to activation of the αVβ3 receptor.29,30 In addition, it has recently been established that T4 plays a crucial role in this phenomenon.
Impact on thyroid cancerAs previously seen, THs may, on the one hand, play a stimulating role in cancer progression, but there is also evidence to suggest that the pathophysiological mechanism is the activation of the MAPK signaling pathway, a key pathway in cell differentiation and proliferation which has been shown to be determinant in the development of papillary thyroid carcinoma (PTC).31 All of these data have led to the impact of ST in the treatment scheme for DTC coming under consideration and suggested a potential new, unknown adverse effect of ST: the possibility that ST is related to either DTC evolution or to the occurrence of a second tumor. Few experimental data are available in this regard.
Hoffmann et al.32 showed in 2005 that normal thyroid tissue expresses αVβ3 and that various DTC cell lines show variable integrin expression patterns. In addition, Illario et al.33 have shown that the T4-αVβ3 complex also activates the MAPK signaling pathway in thyroid cells. More recently, Yalcin et al.34 reported that, in an experimental follicular carcinoma model, T4-αVβ3 blockade resulted in decreases in both angiogenic capacity and tumor mass. In an interesting article, Lin et al.35 found that activation of T4-αVβ3, at physiological T4 levels, caused a proliferative stimulus and a decreased apoptotic capacity in PTC and FTC cell cultures. In their conclusions, the authors postulated that in some patients ST may have a stimulating action on residual tumor growth, even in the absence of TSH.
TSH suppressive therapy and second tumorsPatients with DTC are at greater risk than the general population of developing a second primary tumor (SPT). Studies reported in recent decades, including three large epidemiological studies36–38 and a meta-analysis,39 confirm a risk of SPT 5–31% greater than expected. One of the reasons given for this increase is the carcinogenic effect of 131I ablation therapy.40 De Vathaire et al.41 reported that the increase in colon cancer occurrence was related to the total dose of 131I administered. More recently, Fallahi et al.42 estimated that a cumulative total dose higher than 40GBq (1.08 Ci) was associated with a significant increase in SPTs. Ronckers et al.37 conducted a study based on the patient cohort from the US Surveillance, Epidemiology, and End-results (SEER) program. The authors analyzed the incidence of a subgroup of tumors in tissues with a greater exposure to the radioisotope. These included tumors in the salivary glands, stomach, small bowel, and urinary bladder, as well as in leukemias. The risk of experiencing some of these tumors was two-fold greater in patients with DTC who had received 131I as compared to those given no ablation therapy.
However, other authors found no association between SPT and 131I treatment. Bhattacharyya and Chien43 compared two patient groups with DTC depending on whether or not they had received isotope treatment and found that STP occurred in 6.7% and 4.8% of untreated and treated patients respectively. Similarly, Berthe et al.44 and Verkooijen et al.45 found no influence of the type of treatment used. This fact, combined with the evidence that the inverse relationship is also significant37,38,46 (patients with extrathyroid neoplasms who subsequently develop DTC), has led to other hypotheses being considered. Thus, it has been suggested that a patient may have common risk factors for the occurrence of different tumors, such as some environmental conditions or a genetic predisposition.45,47
No published study has assessed the potential role of ST in the risk of SPT. Although this is fully speculative, it is interesting to note that in a majority of series, the most common SPTs are breast, kidney, and prostate tumors, which are the ones most commonly involved in the association between THs and cancer.
TSH an differentiated thyroid carcinomaThe role of THs and TSH in DTC appears to be even more complex. Although, as already noted, the epidemiological Norwegian study by Hellevik et al.23 related TSH levels less than 0.5mU/L to a high risk of various tumors, there is some evidence that the opposite occurs in DTC, i.e. a direct relationship appears to exist between TSH values and the risk of thyroid cancer. Boelaert et al.48 reported in 2006 that serum TSH levels were an independent predictor of malignancy. Other studies have subsequently shown that presurgical TSH levels are a risk marker for DTC in thyroid nodular disease (TND).49,50 Jin et al.51 found that in TND patients, TSH levels less than 0.9 mU/L were associated with a 10% chance of experiencing DTC, but that the risk increased to 65% with TSH levels higher than 5.5mU/L. Moreover, TSH elevation is also related to DTC which is diagnosed at more advanced stages or is more aggressive.49,52 Our group recently reported a 12% risk of malignancy in patients with TND and subclinical hyperthyroidism, which increased to 20.5% when TSH was within normal limits and to 42% in patients with subclinical hypothyroidism.53 TSH levels were, in turn, correlated to tumor size, so that mean levels were 1.36±1.62mU/L in TND without DTC, 1.71±1.52mU/L in patients with a final diagnosis of DTC less than 1cm in size (microcarcinoma), and 2.42±2.5mU/L in cases with bigger DTCs.
ConclusionsST is usually part of the treatment scheme for DTC. The possibility that subclinical hyperthyroidism associated with ST could cause side effects, particularly at cardiovascular and bone level, has been considered for some time. More recently, it has been established that the proliferative and angiogenesis-promoting effects derived from THs are due to hormone interaction with integrin αVβ3. The influence of this effect on ST on both the course of DCT and the occurrence of second tumors is currently unknown. On the other hand, low TSH levels correlate to an increased risk of extrathyroid tumors, but appear to decrease the risk of DTC in TND.
DTC has traditionally been considered as a group of TSH-dependent tumors, and TSH inhibition by ST was therefore regarded as an effective measure. However, neither the possibility that DTC was TH-dependent nor the direct effect of THs on tumor course was ever taken into consideration. In the future, we will need to identify which tumors are closer to TSH dependence and which have a predominant dependence on THs. This approach may possibly allow us to understand cases that do not respond to standard treatment, or to understand why ST has not shown a universal efficacy. The aim should be to determine the specific role of TSH and THs in DTC occurrence and development, and to individually optimize the most adequate treatment while minimizing any adverse effects.
Conflicts of interestThe author states that he has no conflicts of interest.
Please, cite this article as: Zafon C. Tratamiento supresor de la TSH en el cáncer diferenciado de tiroides. Un dogma en revisión. Endocrinol Nutr. 2012;59:125–30.