The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030
Address all correspondence and requests for reprints to: Steven I. Sherman, M.D., The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Box 435, Houston, Texas 77030. E-mail: . sisherma{at}mdanderson.org
Initial therapy of patients with differentiated thyroid carcinoma generally includes surgical removal of the thyroid gland and any grossly evident locoregional metastatic disease. In the absence of evidence for remaining cancer, postsurgical administration of radioiodine for adjuvant ablation of residual thyroid tissue is often administered to achieve two intertwined primary objectives: 1) to increase the specificity and positive predictive value of subsequent radioiodine scanning and serum thyroglobulin measurements for detection of recurrent or metastatic disease by eliminating sources of false positive test results (i.e. normal follicular cells); and 2) to destroy any residual albeit undetected microscopic foci of malignant disease. The cellular targets of therapy differ for each of these two goals. In the first instance, treatment is aimed at eliminating differentiated functions of normal thyroid follicular cells, which usually is interpreted as indicating death of all such remaining normal cells. In the second instance, treatment is aimed at death or, at a minimum, inhibition of subsequent growth of the malignant cells, which is usually inferred from the absence of detectable serum thyroglobulin and radioiodine uptake. The outcomes of achieving each of these goals may also differ. By reducing false positive test results, adjuvant radioiodine therapy theoretically could improve early detection of recurrent disease, potentially allowing additional treatments that prevent serious morbidity or even death. Paradoxically, such a benefit might yield statistical evidence of a shortened time to detection of recurrence but improved survival as a consequence of this lead time bias. In contrast, by eliminating microscopic residual malignant cells, adjuvant radioiodine could also reduce the actual lifetime risk of a recurrence occurring, along with possibly decreasing the risk of dying due to recurrent disease. Whether these theoretical benefits accrue to differentiated thyroid cancer patients remains a matter of controversy among experts in the field. Reflecting these differences of opinion, some consensus treatment guidelines specifically recommend adjuvant radioiodine therapy be given to most patients with primary tumors greater than 1 cm in size, whereas others are less definitive.
Assuming that adjuvant therapy is beneficial, what factors might influence the effectiveness of radioiodine to achieve these two separate goals? At a most simplistic level, the two primary parameters would be 1) the ability to deliver a sufficient radiation dose to the target tissue, and 2) the inherent radiation sensitivity of that target. Whereas the latter is likely a function of the biology of the particular tissue of interest (be it benign or malignant thyroid cells), the former depends on the tissue, the method of therapy, and numerous patient-specific factors. Optimizing adjuvant radioiodine therapy, therefore, requires manipulation of a complex set of physiological and radiobiological factors, not all of which are well understood (1, 2).
For radioiodine to work, it must first get into thyroid cells, a process that requires presence of functioning sodium iodide symporter (NIS) proteins on the basolateral membrane of the follicular cell. Once in the follicular cell, iodine must be transported across the cell to the apical surface, where it can be excreted into the follicular lumen to be oxidized and organified by thyroid peroxidase along the interface between the cell membrane and the colloid. Iodide returns into the cell as a result of pinocytosis of thyroglobulin-containing colloid, followed by lysosomal cleavage of thyroglobulin into thyroid hormones that can then be either deiodinated or packaged for secretion across the basolateral membrane into blood. Iodine, thus, can exit the follicle as a component of thyroid hormones or as free iodide. When radioiodine is administered to a patient, the amount of isotope that is taken up by the thyroid cells and the kinetics of release of iodide from those cells contribute to the instantaneous and cumulative radiation dose absorbed by the follicular tissue. The choice of a particular isotope is also critical, because the properties of the various radioactive iodines theoretically available (131I, 123I, and 125I) differ with respect to the mixture of types and energy levels of radiation emitted as well as the physical decay half-lives.
Manipulation of follicular cell iodine metabolism is a prerequisite to successful radioiodine therapy. The two central regulators are TSH and iodine itself. As a stimulus to thyroid hormone production, TSH induces numerous steps in intracellular iodine metabolism, including expression of the NIS gene, translocation of NIS protein to the cell membrane, basolateral ATPase required for NIS function, apical iodine efflux, production of thyroglobulin and thyroid peroxidase, pinocytosis, thyroid deiodinase, and finally thyroid hormone export. In contrast, increasing amounts of intracellular iodine counter most of the stimulatory effects of TSH as a form of autoregulation of iodine metabolism. The traditional procedures used in adjuvant radioiodine therapy have been developed to maximize the therapeutic radiation dose while minimizing the dose to normal bystander tissues other than thyroid follicular cells. Maximum TSH elevation is achieved by induction of hypothyroidism, allowing for the gradual decline of either endogenously synthesized or exogenously administered thyroid hormones. Evidence has long suggested that both the absolute height of the TSH concentration as well as the duration of TSH elevation are necessary parameters. The exponential rise in TSH levels after thyroid hormone withdrawal to yield a concentration of at least 30 mU/liter has been considered adequate to provide stimulation of iodine uptake, but whether this level actually maximizes the uptake capacity has been less clear. Groups led by Robbins and Schlumberger (3, 4) independently examined the duration of thyroid hormone withdrawal, and both studies concluded that 3 wk of substitution of triiodothyronine for levothyroxine, followed by an additional 2 wk of complete hormone withdrawal, provided maximal TSH stimulation to radioiodine uptake in both thyroid bed remnants and metastatic foci. These findings appeared consistent with earlier data that radioiodine uptake peaked in normal thyroid tissue 1012 d after cessation of chronic triiodothyronine suppression. Indeed, net uptake, being a result of both trapping and incorporation of iodine into the follicle but minus iodine efflux from the cell in the form of both hormones and free iodine, might be limited by chronic TSH-induced increases in iodine metabolism and clearance. This latter concern about the effects of TSH stimulation of iodine clearance from thyroid cells would play a major role in the delivery of a radiation dose, because increasing clearance rate would lead to a decreased retention of iodine in the thyroid follicles, thus potentially decreasing the actual radiation dose. As for the other differentiated function of interest, thyroglobulin production, little difference was seen in serum thyroglobulin levels after 2 and after 4 wk of triiodothyronine withdrawal (5).
Further limiting understanding of the role of acute and chronic TSH effects has been the absence of useful cellular models, so that most conclusions have been based on the limited studies in the intact human. Analyses of whole body radioiodine kinetics and quantitative radiation dosimetry have until recently only been applied to the situation of chronic withdrawal and TSH elevation. Using 6 wk of levothyroxine withdrawal (with triiodothyronine therapy during the first 3 wk), Maxon et al. (2) demonstrated that delivery of 30,000 cGy from 131I could ablate thyroid bed remnant activity in 81% of patients, whereas only 8,000 cGy appeared necessary to ablate uptake in nodal metastases. These radiation doses required for successful ablation have been widely quoted, but whether they are truly applicable to other experimental models is quite unclear. For example, do the kinetics of TSH stimulation, as well as hypothyroidism itself, alter the radiation sensitivity and the repair mechanisms that prevent cellular death from radioiodine therapy? Furthermore, the definition of successful ablation itself, the absence of visible radioiodine uptake on subsequent 131I scans, might depend on the activity of radioiodine used in scanning and the method of TSH elevation, and makes no reference to ablation of sources of thyroglobulin production. Finally, why markedly lower radiation doses are required to ablate a metastatic focus of malignant cells compared with those needed to destroy normal remnant tissue remains unexplained.
To compound the confusion of the role of acute vs. chronic TSH stimulation have been the various attempts to abet hormone withdrawal by maneuvers to increase TSH acutely. Again, studies decades ago looked at two approaches: exogenous TSH in the form of bovine TSH, and exogenous TRH administered to stimulate endogenous TSH production. Conflicting results have been reported, with some evidence that an acute TSH surge superimposed on chronic endogenous TSH elevation from hormone withdrawal might augment radioiodine uptake; but whether this could translate to a clinical benefit remained quite unclear (6). More recently, the trials of recombinant human TSH (rhTSH; Thyrogen, Genzyme Corp., Cambridge, MA) have suggested that acute TSH elevation might not be quite as effective as chronic hypothyroidism in stimulating radioiodine uptake and lesion visualization on scans. The confirmatory Phase III trial, attempting to find a middle ground between chronic and acute TSH effects, found no difference between prolonged TSH stimulation from three injections each given 72 h apart and that from two injections given 24 h apart (7).
Superimposed on this background of attempting to understand the effects of TSH on the radiobiology of iodine has been a desire to minimize the activity of radioiodine administered for adjuvant therapy. Historically, a major driver in this effort was the United States federal regulation requiring patients who were administered activities of 30 mCi or greater to be hospitalized, a recently removed restriction. Dosimetric analyses demonstrated that many patients could be treated with such low activities with successful ablation of uptake visualized on subsequent diagnostic radioiodine scans, particularly if thyroidectomy reduced 24-h uptake measurements to less than 5% (2). Other studies, including one randomized trial, have also indicated the potential success of "low dose" radioiodine therapy to ablate thyroid bed uptake on subsequent scans, but varying definitions of successful ablation have been applied. Although radioiodine therapy has generally been described as safe, reducing the whole body radiation dose by using lower activities for adjuvant therapy has been assumed to be an even safer approach, but data regarding long-term efficacy as well as safety have been limited (8).
Into this mix of uncertainty regarding the optimal approach to TSH stimulation and the radioiodine activity needed for adjuvant therapy has come the article by Pacini et al. (9), in this issue of the JCEM. In their study, the investigators assigned patients to three nonrandomized cohorts to examine differing methods to raise TSH levels prior to adjuvant therapy with 30 mCi 131I: group 1, who underwent thyroid hormone withdrawal; group 2, who also underwent withdrawal but received additional TSH stimulation from two injections with rhTSH; and group 3, who remained on thyroid hormone and received injections of rhTSH. In the three groups, 24-h radioiodine uptake was lowest in group 3 (2.5%), significantly lower than that seen in both groups 1 (5.8%) and 2 (9.4%). Using an approximation to calculate the initial radiation dose rate from the 30 mCi treatment, group 3 (10.7 Gy/h) patients received the lowest initial dose compared with group 1 (27.1 Gy/h) and group 2 (48.5 Gy/h). Success of ablation was determined by a subsequent 4 mCi 131I scan performed under standard hypothyroid conditions, 610 months after adjuvant therapy. The frequency of successful ablation was 84% in group 1, 78.5% in group 2, and only 54% in group 3. As a secondary endpoint, the frequency of an undetectable TSH-stimulated thyroglobulin level at the time of the follow-up diagnostic scan was also determined for each group. By reworking the data presented in the manuscript, one can calculate the frequencies of an undetectable thyroglobulin level as 74% in group 1, 83% in group 2, and 63% in group 3. Finally, further calculations demonstrate that the frequencies of achieving both remnant ablation on subsequent scanning and undetectable thyroglobulin levels were 70% in group 1, 67% in group 2, and 47% in group 3.
What can be inferred from this study? First, it would seem that use of 30 mCi radioiodine often fails to achieve the primary short-term goal of adjuvant therapy (i.e. eliminating all differentiated function from normal remnants). About 30% of patients treated after standard thyroid hormone withdrawal had either a positive thyroid bed remnant on subsequent 4 mCi scanning or detectable thyroglobulin level during follow-up testing. Based on simply having an elevated thyroglobulin level, which these authors have stated elsewhere would warrant further radioiodine therapy, at least 25% of these patients would require additional treatment (10, 11). Therefore, attempts to improve the effectiveness of adjuvant therapy are well justified.
The second major conclusion would seem to be that adjuvant therapy administered to euthyroid patients following rhTSH stimulation is less effective at complete ablation of visible radioiodine uptake than that administered to hypothyroid patients, despite achieving high levels of TSH in both situations. A number of potential explanations come to mind. One obvious possibility would be that the radioiodine uptake and retention following rhTSH was simply insufficient to provide enough of a radiation dose to kill all of the follicular cells. In the absence of information on the clearance of radioiodine from the remnant, one cannot know if the radiation dose rate is high enough for a long enough period of time to avoid sublethal, reparable cellular damage (12). On the other hand, the ability to repair radiation damage could be dependent on ambient thyroid hormone levels, and thus a higher radiation dose might be required to achieve similar effects on thyroid cells.
In contrast, use of rhTSH to increase TSH levels acutely above the chronic elevation from thyroid hormone withdrawal led to increased uptake and radiation dose rate, but these effects failed to improve ablation efficacy. Interestingly, there was a trend toward a higher frequency of elimination of thyroglobulin production in group 2 patients, suggesting a discordance between the sensitivity to radioiodine therapy for these two separate differentiated follicular cell functions. One explanation for the failure of these initial dose rate calculations to predict ablation efficacy may be failure of critical assumptions implicit in the formula used in the study. First, the nonpenetrating radiation particles that derive from 131I (beta particles and conversion electrons) deliver their energy in a tissue diameter primarily between 5 and 50 mm (13). In contrast, the smaller the thyroid bed remnant below 5 mm in any dimension, the lower the absorbed radiation from treatment. Given the residual tissue mass provided in Table 3, it is likely that many of the patients had remnants smaller than 5 mm in diameter, leading to overestimation of the actual initial dose rate. Second, the formula assumes similar single exponential decay of cellular radioiodine content among all three patient groups, but the differing modes of TSH elevation in the cohorts makes this assumption tenuous. Third, these findings likely cannot be extrapolated to the issue of lethal therapeutic doses to undetected microscopic residual foci of carcinoma, given the likely small diameter and the absence of information about the kinetics of radioiodine uptake and clearance of these tumor deposits.
Adjuvant radioiodine therapy may benefit differentiated thyroid cancer patients in a variety of ways, improving the accuracy of follow-up testing and reducing the overall likelihood of disease recurrence and death. Whereas several studies have supported this contention in high-risk patients, this has been less clear and more controversial in low-risk individuals. If adjuvant therapy is to achieve all of its desired goals, then optimal patient preparation to maximize radiation dose to both remnant and microscopic tumor foci is required. The study by Pacini et al. (9) seems to indicate that 30 mCi 131I is inadequate, even after superimposing rhTSH on thyroid hormone withdrawal-induced hypothyroidism. It is hoped that the ongoing randomized trial of 100 mCi 131I given to euthyroid patients receiving rhTSH compared with standard hypothyroid preparation, with its careful attention to quantitative dosimetry, will provide greater insight into the differences between hypothyroid and euthyroid high-dose radioiodine therapy.
Acknowledgments
Footnotes
Abbreviations: NIS, Sodium iodide symporter; rhTSH, recombinant human TSH.
Received July 16, 2002.
Accepted July 16, 2002.
References