Washington Hospital Center Washington, DC 20010-2975
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
In their respective sections of this Therapeutic Controversy, Sherman and Gopal and Schlumberger and Hay address different but related aspects of the management of patients with DTC. The respective authors are not engaged therefore in a debate. While DTC remains one of the most curable of all cancers, patients with aggressive disease are occasionally seen, and outcomes have been clearly related to a number of variables (4, 5, 6). Given the numerous variables existing in any single individual or in any group of patients, I believe the design of and adherence to an algorithmic approach to the follow-up management of patients with DTC to be both treacherous and possibly overly simplistic. Rather, management of each case should be individualized. Having said that, let us examine some of the recommended approaches proposed by Sherman and Gopal and by Schlumberger and Hay.
Several series suggest that excellent prognosis with cure attends DTC of less than 1.5 cm diameter, even when treated with less than a near total thyroidectomy (5, 6). I would like to focus this discussion on higher risk lesions that are 2 cm or greater and that might have been associated with either regional node metastases or distant metastases. For such patients, the usual management would be total thyroidectomy followed by radioiodine ablation. This initial management has been almost universally adopted since the publication by Mazzaferri et al. (7) demonstrating that less aggressive surgery without radioiodine ablation is associated with higher rates of recurrence and death. Schlumberger and Hay would have us adopt a more "selective" approach to the use of radioiodine. They challenge whether ablation is actually of benefit and point out that the reduced recurrence rates seen in the Mazzaferri series after radioiodine ablation may have been the result of less than complete total thyroidectomies. This conclusion is based upon the finding of comparable recurrence rates with or without radioiodine ablation at the Mayo Clinic, where an ostensibly more complete thyroidectomy was regularly performed.
Schlumberger and Hay cite the results of Simpson et al. (8) as supporting this contention. These workers found that radioiodine ablation benefited survival only in patients with residual microscopic disease. Curiously, the Mayo group had previously published data indicating that frequency of local recurrence was no different between patients having total vs. bilateral subtotal thyroidectomy (9), and survival was not improved by total thyroidectomy in either minimal or higher risk patients with papillary carcinoma (10). Thus, the cited explanation for the variance of the Mayo results with those of Mazzaferri would appear unlikely or at least incomplete, and there must be other differences between the respective patient groups analyzed. I would infer a somewhat different conclusion from the observations of Simpson et al., and that is a resultant clear benefit of radioiodine ablation. The issue is related to the degree of certainty of the treating physician that there is no residual microscopic disease. Many of the reported series do not provide clear evidence of the extent of residual tissue as might be inferred from RAI uptakes or serum Tg levels. I would agree that a low-to-moderate risk patient with a baseline Tg level of less than 5 ng/mL and a postoperative RAI uptake of less than 0.5% would not require ablative therapy. However, if we place ourselves in the shoes of even a "low risk" patient, would we not willingly accept the consequence of a 3060 mCi ablative dose of 131I in exchange for the certainty and peace of mind provided by a subsequent negative scan and undetectable serum Tg level? That radioiodine therapy "wipes the slate clean" and provides the ability to more readily detect recurrence by either 131I TBS or by monitoring serum Tg is not, apparently, a compelling argument to Schlumberger and Hay. They point out that Tg levels adequately monitor such patients, with 93% and 80% having immeasurably low levels while on and off levothyroxine suppression, respectively. These are impressive percentages on a statistical basis but not on an individual patient basis; I have difficulty settling for an 8090% average if radioiodine ablation could bring my awareness of the presence or absence of residual disease closer to 100%. Schlumberger and Hay would simply catch those patients initially missed at a later time, several months later perhaps, as serum Tg began to rise, and they would then further evaluate and treat at that time. Such management would likely suffice and achieve a "cure" in most patients, but again I have concern for those who unpredictably may manifest more aggressive disease. In such patients, cure appears to be best achieved when their disease is caught and treated as early as possible. Mazzaferri and Jiang found that delays in treatment were associated with a higher mortality (4). Moreover, a serum Tg-based system of follow-up for detection of residual disease might be misleading due to false negatives; Schlumbergers group (11) reported patients with little elevation in Tg even in the face of pulmonary metastases, and in the last section of this Controversy, they cite that 20% of patients with known lymph node metastases may have low serum Tg. That Tg might be undetectable in the face of known metastatic disease while patients are on levothyroxine has been known for quite some time (12).
Schlumberger and Hay acknowledge that administering an ablative dose of radioiodine permits a follow-up scan that can detect the presence of unrecognized metastases. This fact also bears on our ability to achieve the salutary effects of treatment by earlier diagnosis. Casara et al. (13) realized remission rates of 78% in patients with negative chest x-rays but positive lung uptake on scan, in contrast to a remission rate of less than 4% in those patients who already had evidence of lung metastases on chest x-ray. Such studies confirm the reciprocal relationship between success of cancer therapy and size and duration of lesions.
Certainly few would argue with Schlumberger and Hay that one should always take advantage of the administration of a treatment dose of 131I by performing a post-treatment scan [a position we share (3)], particularly with large (75150 mCi) treatment doses that might disclose previously undetected metastases. But we can see patients with low serum Tg and significant metastatic disease even when no longer TSH suppressed (14). This phenomenon is related to the whole dilemma of false negative or low serum Tg levels in the face of known residual malignancy. The most likely explanations for this include the presence of interfering anti-Tg antibodies, which can be present in as many as 40% of patients with DTC (15), and the dedifferentiation of the tumor cells, such that they can still trap iodide but can no longer make and release Tg. It should be mandatory for laboratories to measure anti-Tg antibodies in every sample in which Tg is being measured. While assays for serum Tg are improving and becoming more standardized (albeit very slowly), Spencer et al. (15) point out that the presence of antibodies continues to be a problem, and can be associated with either falsely low or falsely high serum Tg measurements. Hence, a largely Tg-based strategy as suggested by Schlumberger and Hay may be problematic at times given the frequency of antibodies in the DTC population.
Schlumberger and Hay question the need for levothyroxine withdrawal for the purpose of a follow-up TBS a year after therapy. At present, this procedure requires allowing the patient to become sufficiently hypothyroid to raise serum TSH to more than 40 mU/L and thereby facilitate 131I uptake and imaging of either residual thyroid bed tissue or malignancy. Such scans are done with doses of 131I that range from 310 mCi. The larger doses are associated with better images and improved sensitivity of tumor detection but may also be associated with "stunning" or with a lower fractional uptake of 131I with the subsequent treatment dose (16). Schlumberger and Hay would have us abandon these TBS doses as either unnecessary in the low risk patient with low serum Tg or problematic in the higher risk patient because of stunning. In the latter group, they would have us opt instead for a much larger dose that would suffice for both imaging and treatment. In this context, I am drawn again to the importance of individualizing management in the decision making process. I fully concur with the approach of Schlumberger and Hay to base the aggressiveness of further diagnostic and therapeutic approaches on the patients risk factors, clinical situation, serum Tg on and off levothyroxine therapy, and other nonisotopic imaging techniques, such as ultrasonography or MRI, to identify residual or recurrent disease. With the imminent availability of recombinant human TSH (rhTSH) (17), we will be able to evaluate patients without having to render them hypothyroid. Indeed, in the absence of anti-Tg antibodies, failure of Tg to increase after rhTSH administration would be a compelling argument for the patient being free of malignancy and therefore having no indication for scanning or further isotope treatment. There will also likely be a role for rhTSH in facilitating the therapy of patients with metastatic disease. I have personally treated one patient with metastatic disease after rhTSH preparation, and another has been reported with salutary results (18). The ability to identify residual or recurrent disease by other scanning modalities that also would not require discontinuing TSH-suppressive levothyroxine therapy is discussed below.
Most endocrinologists dealing with thyroid cancer patients would agree with the therapeutic approach recommended by Schlumberger and Hay of combination 131I and surgery. Their use of 131I to identify lesions by an intraoperative detector probe is innovative. Many centers in the U.S. would operate first on any palpable or otherwise accessible tumor recurrence and follow the surgery with radioiodine therapy. This has the benefit of "debulking" the tumor mass and rendering the radioiodine dose as more effective therapy. As pointed out by Schlumberger and Hay, "small neoplastic foci respond better to 131I than larger ones." I think that it might be preferable to employ 123I to guide the intraoperative gamma probe of the surgeon, an isotope that would be associated with less radiation exposure and less potential for stunning. Then, following surgery, the patient could be treated with a large dose of 131I, particularly if serum Tg was still elevated (19). Surgical excision may not be warranted for the appearance of clinically detectable cervical lymph nodes in the presence of low serum Tg and negative anti-Tg antibodies. Rather, I would first want to demonstrate the presence of DTC metastases in the lymph nodes, which could be done by fine needle aspiration biopsy, with or without ultrasound guidance, and by subjecting the aspirate to either cytologic examination (20), PCR-based genetic analysis (21), or measurement of Tg (22).
Dr. Schlumberger was the first investigator to advocate empiric high dose 131I therapy for patients who are "scan negative, thyroglobulin positive." Drs. Sherman and Gopal advise caution in applying high dose therapy in such patients in the absence of data confirming efficacy and an acceptable risk/benefit ratio, and I must agree with them. With this scenario, one should first attempt to uncover a cause for a possibly false negative scan or a false positive elevation of serum Tg. As mentioned above, the latter can be due to interfering anti-Tg antibodies (15). Explanations for a false negative radioiodine scan include inadequate TSH elevation, stable iodine contamination (e.g. history of recent iodine contrast radiography), dispersed microscopic metastases too small to visualize, or dedifferentiation of the tumor such that it can still produce Tg but has lost its iodide trapping ability. To rule out iodine contamination, serum or urinary iodide can be measured and a repeat TBS 46 weeks after an iodide depletion regimen can be considered (23).
In this setting faced with a decision as to how to proceed, I would
again look at the patient in terms of risk factors, evidence of earlier
metastatic or aggressive disease, and any arguments for employing other
imaging tools such as MRI or ultrasound to visualize possibly occult
disease. For example, given this scenario of "positive Tg and
negative TBS" in a 58-yr-old man with a history of a 4 cm papillary
or a 2 cm follicular lesion, I would consider the Schlumberger approach
and treat with high dose radioiodine. On the other hand, with a history
of a 2 cm papillary cancer with negative nodes and only marginally
measurable or slightly elevated Tg (e.g. 10 ng/mL), I
might favor a more conservative approach. Significant to the decision
making process is whether serum Tg levels are stable or rising. Of
course, the patient must be brought into the decision-making process
and informed fully of the extent of our collective knowledge,
experience, and biases in regard to their specific situation. We would
like to avoid treatment of patients with aggressive high dose
radioiodine for uncertain indications and which might result in
troubling sequelae such as xerostomia and/or azoospermia.
I agree with Sherman and Gopal that, absent evidence of progressive disease, the risks of aggressive radioiodine therapy may not be warranted given ill-defined goals. In addition to the experience of Schlumberger et al. (24), another oft-cited experience in support of empiric treatment of the thyroglobulin-positive, scan-negative patient is that of Pineda et al. (25). These workers reported 17 such patients, all of whom had had prior total thyroidectomy and radioiodine ablation. After treatment with 150300 mCi of 131I, 16/17 had visualization of metastases on their post-treatment scan. Tg levels decreased in 81% of patients after their first treatment dose, and in 90% and 100% of those patients who received second and third doses, respectively. While these results sound impressive as expressed, examination of the individual patients Tg level responses is less so. Mean Tg decreased from 74 to 62 to 32 over 12 yr of follow-up, and only 6/29 positive scans became negative. The cogent issues raised by Sherman and Gopal, and previously by McDougall (26) and Mazzaferri (27), reflect the fact that many of these patients have minimal if any disease that would affect their life expectancy, and yet we are exposing them to unwarranted doses of radiation exposure, unwarranted at least until we obtain sufficient data from well-controlled studies that confirm efficacy of therapy. Certainly another important aspect of this empiric therapy is the cost to the patient in regard to the morbidity of hypothyroidism and its negative impact on productivity, as well as the cost in health care dollars related to hospitalization and the associated expensive technological procedures. Increasing scrutiny by watchdog agencies may challenge the indications for this therapy, and possible denial of reimbursement may cause additional problems for both the patients and their physicians.
Finally, I would mention the additional or alternative imaging procedures that are being developed and evaluated for patients with thyroid cancer. Given a negative 131I TBS, are there other scanning modalities that might provide useful information? Drs. Sherman and Gopal have focussed on the dilemma facing us when confronted with a measurable or rising serum Tg and a negative 131I TBS, generally considered the gold standard for detection of metastases. Once we have eliminated the various causes for false positive serum Tg or false negative TBS, what is the clinician to do? As reviewed above, many authorities question the risk benefit ratio of arbitrary high dose 131I therapy as employed by Schlumberger and coworkers (24). Alternative therapeutic approaches to metastatic deposits of thyroid cancer include surgical excision or localized external radiation therapy (28), but the location of the metastases would need to be identified first. MRI and ultrasound have been employed for this purpose. In addition, I suggest that alternative scanning agents might play a very important role in this regard, for several recent reports have documented their potential usefulness in identifying lesions that are not visualized with traditional 131I WBS.
One of the first to be employed was the 201Thallium (201Tl) (29). In one recent study of patients with bone metastases documented with positive 131I scans, 201Tl was compared to the bone agent, technetium-99 m hydroxymethylene diphosphonate (99mTc-HMDP). The two agents had a combined sensitivity of 93.5%. In a group of 14 patients with negative 131I scans and other evidence of thyroid malignancy, 201Tl was positive in 10/14 and 99mTc-HMDP was positive in all 14. Carril et al. (30) found that 201Tl enjoyed a sensitivity and specificity that was higher than that for 131I for recurrent or persistent disease. Lesions were detected in 31/116 patients by 201Tl but not by 131I TBS. In patients who have been ablated and show no further 131I uptake, the authors propose continuing management with no additional 131I scans; as 201Tl scanning does not require levothyroxine withdrawal, follow-up would be guided only by 201Tl scanning and by monitoring serum Tg. Dadparvar et al. (31) compared 201Tl and scanning with 99mTc-methoxyisobutyl isonitrile (99mTc-MIBI) and found that 131I TBS alone was satisfactory as a preablation study, but that the addition of either alternative agent increased the diagnostic yield post-ablation, particularly when the 131I TBS was negative. These results have not been universal, however, because Lorberboym et al. (32) found 131I TBS to be both more sensitive and specific than 201Tl, with the latter giving several false positive scans. Ugur et al. (33) noted a 70% overall concordance between 201Tl, 99mTc-MIBI, and 131I TBS, but observed false negatives with both alternative agents and concluded that they should not be used in lieu of 131I TBS. Elser et al. (34) noted however a 94% sensitivity for the detection of positive lymph nodes and local recurrence with 99mTc-Sestamibi; they detected 32/40 metastases with Sestamibi compared with only 18/40 with 131I TBS. More recently, investigators have attempted detection of thyroid cancer with 99mTc-tetrafosmin, a cationic agent employed previously for myocardial perfusion imaging (35, 36, 37). For 12 patients with elevated serum Tg (4 of whom had negative 131I TBS), tetrafosmin was slightly superior to either 201Tl or 99mTc-MIBI. This same group of workers (37) reported that tetrafosmin successfully identified 21/21 lesions that were positive by 131I TBS, but an additional 17/23 lesions that were negative by 131I TBS. The agent had 86% sensitivity for distant metastases, was positive in 4 patients with 131I negative proven pulmonary metastases, and the findings correlated with other modalities identifying tumor such as computed tomography (CT) or ultrasound.
It is also significant that these alternative agents are logistically both more convenient and more expedient than scanning with 131Iodine. In addition to being able to scan patients while they are still taking TSH-suppressive doses of levothyroxine, the time required for evaluation is much reduced. Instead of scanning 4872 h after a dose of 131I, the 99mTc-tetrafosmin planar scan is performed 20 min after injection of the isotope, with additional images taken by single-photon emission computed tomography (SPECT) of any suspicious lesions. 99mTc-tetrafosmin scans were negative in all 68 patients, studied by Lind et al. (37), who were free of disease on the basis of 131I TBS and serum Tg.
Another agent, 18-fluorine fluorodeoxyglucose is employed with PET scanning (FDG-PET) with uptake of the agent related to glucose utilization by tumor tissue. The greatest uptake sensitivity has been noted with fastest growing undifferentiated tumors. Grunwald et al. (38) compared FDG-PET to 99mTc-Sestamibi and 131I TBS. Of 29 studies, 11/29 had disease detected only with FDG-PET, 8/29 were detected only with 131I TBS, and 10/29 were detected by both. Five sites were detected by FDG-PET and not by 99mTc-Sestamibi. FDG-PET may be useful in patients in whom 131I TBS is not feasible due to a history of iodine exposure, and similarly, its use would not preclude scanning by CT with contrast if desired as an additional means of imaging tumor. A drawback is the lack of widespread availability of PET scanners due to their high cost.
Fridrich et al. (39) compared FDG-PET to 99mTc-MIBI and 131I TBS and found both to be more sensitive than 131I TBS with a slight edge in favor of 99mTc-MIBI. In addition to the benefit of having good uptake independent of the patients serum TSH level, FDG-PET or MIBI did not have the propensity to have high background in the neck, mediastinum, and chest, as does 131I, and could be employed more effectively to detect small metastases in these areas. On the other hand, liver and brain will demonstrate high uptake of FDG, and the ability to pick up metastases in these areas will be limited with this agent. Indeed, Feine et al. (40) were able to localize and identify positive neck metastases with FDG-PET in 6 patients with elevated serum Tg levels. A more conservative view to the utility of FDG-PET scanning has been proposed by Dietlein et al. (41). They observed positive FDG-PET images in 7/21 patients with positive lymph node metastases but negative 131I TBS; sensitivity was 82% in patients with high serum Tg but negative TBS. They concluded that FDG-PET should not be used instead of 131I-TBS but would serve as a useful complement to evaluation, particularly when the 131I TBS was negative in the face of a rising or elevated level of serum Tg. Finally, imaging of DTC by somatostatin receptor scintigraphy (SRS) with octreotide has been reported by Baudin et al. (42). Of 25 patients with DTC and elevated serum Tg levels, 16/25 had negative 131I TBS, and SRS was positive in 12 of these 16 patients and in 8/9 patients with positive 131I TBS. While confirmatory studies will be required, SRS with labeled octreotide may represent another useful alternative to 131I TBS, with the advantage of not having to withdraw TSH-suppressive levothyroxine therapy.
In conclusion, how should one manage the scan negative, thyroglobulin positive patient with no underlying reason to suspect either a false negative scan or a false positive serum Tg level? Schlumberger (1 and in this paper) would empirically treat with 100 mCi 131I any patient with a Tg level of more than 10 ng/mL while off levothyroxine, and would only repeat the 131I TBS every 25 yr when the Tg level is in the range of 110 ng/mL. Given clearly measurable Tg levels, I would encourage alternative imaging procedures. For papillary thyroid carcinoma with a propensity to regional recurrence, that could include ultrasound, CT, MRI, 99mTc-MIBI, or FDG-PET. For follicular thyroid cancer with its propensity for distant metastases (especially to bone and lung), 99mTc-tetrafosmin or 99mTc-HMDP or 201Tl could be employed. Identification of isolated distant lesions by these methods would allow earlier intervention by surgical excision or external radiotherapy, rather than delaying further treatment until a subsequent 131I TBS might become positive or until serum Tg levels might increase further as a result of further tumor growth. In patients with higher risk disease following early total thyroidectomy and high-dose radioiodine ablation, this approach should permit effective management until such time as more target-specific tumoricidal therapies become available.
![]() |
References |
---|
![]() ![]() ![]() |
---|