Advances in the Monitoring and Treatment of Differentiated Thyroid Cancer

Improved assay techniques for monitoring patients following thyroid hormone ablation for follicular-cell-derived carcinoma have presented clinical endocrinologists with a new challenge, that of the patient who is scan-negative but thyroglobulin-positive for continued thyrocyte activity. With the help of new imaging techniques and recombinant human thyrotropin (thyrotropin alfa, or rhTSH), new diagnostic algorithms are emerging, but to date there is no recommendation for a standard sequence.

Recombinant thyrotropin may also have a role in the modern treatment of thyroid cancer, though it is currently approved by the Food and Drug Administration (FDA) for diagnostic studies only. Off-label investigation indicates that this agent has equivalent value to total thyroid hormone withdrawal in preparing patients for radioiodine therapy for residual thyrocytes, while sparing them of many of the debilitating and potentially dangerous side effects of temporary hypothyroidism. This benefit may extend to preparation for the treatment of both locoregional persistence and distant metastases.

The diagnostic and potential therapeutic roles of rhTSH and emerging approaches to the monitoring and treatment of thyroglobulin-positive, scan-negative patients with thyroid cancer were discussed at a satellite symposium conducted in conjunction with the 13th Annual Meeting and Clinical Congress of
the American Association of Clinical Endocrinologists (AACE).

This program was supported by an unrestricted educational grant from Genzyme Corporation.

The Thyroglobulin-positive, scan-negative Patient with Thyroid Cancer

Paul W. Ladenson, MD (Johns Hopkins Medical Institutions) introduced the case of a 48-year-old woman to illustrate
the new challenge. Ten years ago, a 2-centimeter nodule was incidentally detected and demonstrated by biopsy to be papillary thyroid cancer. A postoperative pathologic examination determined that the lesion extended to the margin of resection and that one of four resected lymph nodes was involved. The patient was withdrawn from thyroid hormone. A radioiodine scan utilizing 2 millicuries of 131I showed uptake limited to the thyroid bed. The patient was treated with 75 millicuries of 131I, and a repeat scan after one week confirmed only thyroid bed activity. Treatment with 0.15 mg per day of thyroxine was sufficient to suppress her thyroid-stimulating hormone (TSH). Seven months following surgery, the patient underwent a cycle of recombinant human thyrotropin-stimulated (rhTSH) testing. Although her 131I whole-body scan (WBS) was negative, her serum thyroglobulin rose ominously to 6 ng per mL. Despite a negative scan, therefore, this patient was a candidate for additional treatment for thyroid cancer.

Managing recurrence typically consists of clinical monitoring, thyroglobulin measurement, radionuclide imaging (mostly radioiodine), and anatomic imaging. Thyroglobulin measurement and radionuclide imaging depend on TSH stimulation for optimal sensitivity. The importance of TSH stimulation was demonstrated in a trial of 220 patients who had either residual foci of thyrocyte activity, or no radioiodine focus on scan despite thyroglobulin levels greater than 2 ng per mL. All subjects tested negatively for anti-thyroglobulin antibody. Alarmingly, only 43% of patients with residual disease had detectable serum thyroglobulin (>2 ng/mL) while on thyroid hormone suppressive ther-apy compared with 74% of patients stimulated with rhTSH. For residual thyrocyte activity in the thyroid bed, rhTSH stimulation was associated with detection of thyroglobulin in 52% of patients compared with 22% of patients on thyroid hormone suppression. Stimulation resulted in detection of thyroglobulin in 100% of patients with distant metastases compared with 77% of patients in the hormone suppression group (Haugen R et al. JCEM. 1999;84: 3877).

In a review of eight studies involving 1,028 patients with negative radio-iodine scans and undetectable thyro- globulin while on hormone suppressive therapy, 26% of patients had serum thyroglobulin concentrations greater than 2 ng per mL following stimulation by rhTSH. On subsequent evaluation, one-quarter of those patients were determined to have metastatic disease. Without stimulation, therefore, 6% of patients in the combined population would have had undiagnosed metastases. Further, whole-body radioiodine scan was able to localize distant metastases in only 14% of cases (Mazzaferri EL et al. JCEM. 2003; 88:1433). These findings underscore the sensitivity of TSH-stimulated thyroglobulin compared with the traditional diagnostic method of scanning. While scanning during TSH stimulation improves overall detection when combined with serum thyroglobulin slightly, it may improve detection in the thyroid bed. However, with respect to metastases, the identification of 100% of cases by TSH-stimulated testing alone in thyroglobulin antibody-negative patients means that the role of scanning is limited (Haugen B et al. supra).

At least four potential causes for the loss of iodine avidity requiring TSH-stimulated thyroglobulin testing have been identified. These are (i) impaired sodium-iodide symponder (NIS) gene expression, (ii) NIS membrane localization or impaired function, (iii) impaired pendrin and/or thyroperoxidase expression, and (iv) impaired TSH receptor stimulation.

Now that the phenomenon of the scan-negative, thyroglobulin-positive patient is firmly established, the challenge is to determine how appropriately to monitor and treat these patients. Repetition of radioiodine therapy is probably futile. Dr. Ladenson prefers, therefore, to begin with anatomical imaging. Because cervical lymph nodes are the most common locus of persistent thyrocyte activity, imaging begins with sonography of the neck. If this is unrevealing, a CT scan of the thorax in search of pulmonary or mediastinal loci may be useful. However, when this is done, the use of radiocontrast dye that might interfere with subsequent radioiodine therapy should be avoided. For some patients, positron emission tomography (PET) scanning may be necessary following negative sonography and CT scan. Any identified lesion that is addressable by fine needle aspiration should be biopsied and, if found to be malignant, should be excised. For deeper lesions, surgical biopsy and subsequent excision if positive are appropriate. Only patients who have negative imaging but thyroglobulin concentrations of 10 ng/ mL and those whose lymph nodes prove to be benign are treated with 200 millicuries of empiric 131I.

The role of sonography in the monitoring of these patients is compellingly supported by the experience of the Mayo Clinic. In a study conducted there, sonography showed a cervical mass and/or cervical lymphadenopathy in 52 patients, 44 of whom had nonpalpable lesions. Ultrasound-guided percutaneous biopsy revealed malignancy in 29 of these patients, lymphocytes without malignancy in 20, and nondiagnostic hypocellular specimens in 3. Thus 94% of biopsy results were confidently assigned as either positive (56%) or negative (38%) for malignancy (Sutton RT et al. Radiol. 1988;168:769). Based on evidence of this kind, Dr. Ladenson recommends postoperative cervical sonography for five groups of patients: (i) those who had incompletely resected primary tumor confirmed by surgical histopathology; (ii) those who had extensive extrathyroidal invasion or cervical node involvement; (iii) those who had certain histologic subtypes known to be especially aggressive (e.g., columnar cell, tall cell, and insular variants); (iv) those who have had prior cervical node recurrence; and (v) those who have BRAF oncogene mutations which occur in 60% to 70% of patients with papillary carcinoma and are associated with aggressive clinical behavior.

Returning to the patient with whose history he began his presentation, Dr. Ladenson noted that her cervical sonogram and thoracic CT scan were negative, but that PET CT fusion scanning revealed a small lymph node at the base of her skull. Because it was inaccessible to percutaneous aspiration, the node was explored and removed surgically and determined to be papillary cancer. This case illustrated the importance of thorough and persistent monitoring, as recurrences of thyroid cancer may appear as long as decades after original diagnosis and treatment.

The Appropriate Role of PET Imaging in Thyroid Cancer Monitoring

PET imaging, although not a new technology, found increased interest in the field of thyroid cancer monitoring when a team of German investigators described a “flip-flop” phenomenon in patients with scan-negative, thyroglobulin-positive thyroid cancer. Using PET imaging along with flourodeoxyglucose (FDG), they observed that differentiated tumors with iodine avidity have low glucose metabolism in greater than 95% of cases while less differentiated tumors without iodine avidity have high glucose metabolism. Therefore, high glucose metabolism is an indicator of poor tumor differentiation and possibly a higher malignant potential. Together, they found, whole-body scanning and FDG have a diagnostic sensitivity of 95% (Feine U et al. J Nucl Med. 1996;37: 1468).

Since the publishing of these findings, FDG PET has been suggested as a monitoring technique for several
categories of patient: (i) those with high-risk disease to determine extent of involvement; (ii) those with adverse histology for long-term prognosis; (iii) those with rising thyroglobulin with no known source; (iv) those with known metastases to determine the extent and relation to vital structures; and (v) those with Hurthle cell carcinoma. It has also been recommended for post-treatment response assessment, lesion dosimetry, and evaluation of the thyroid nodule. It is not recommended for determining the extent of disease in low-risk cases.

PET imaging produces both a picture and a calculated standardized uptake value (SUV). The latter is an index of
glucose uptake at any fixed moment that can indicate the presence of a malignancy and can be followed over time to monitor changes, treatment responses, or disease progression. Richard T. Kloos, MD (Ohio State University) cautioned, however, that SUV is an imperfect measure with no definitive cut-off points between malignancy and benign tissue, and with considerable operator variation.

FDG is a glucose analog that is admitted to cells via glucose transporters. Like glucose, it is phosphorylated; but whereas glucose is metabolized by traditional pathways, FDG remains trapped in the cell and accumulated in correlation with the cell’s glucose uptake activity.

One systematic effort to evaluate FDG PET for detection of thyrocyte loci in scan-negative, thyroglobulin-positive patients reviewed 14 reported series of patients and concluded that the data support the technique in this setting, but that implementation in a routine algorithm requires additional evidence (Hooft L et al. JCEM. 2001;86:3779). Thyroglobulin level, TSH stimulation via thyroid hormone withdrawal, and TSH stimulation by rhTSH administration have all been considered as factors for optimizing the value of FDG PET. In one study of 118 PET scans in 64 patients, thyroglobulin level correlated with FDG PET sensitivity. Positive scan results were achieved in 11% of patients with thyroglobulin levels of 10 ng/ mL or below. This increased to 50% among patients with thyroglobulin levels between 10 and 20 ng/mL and to 93% at levels above 100 ng per mL (Schluter B et al. J Nucl Med. 2001;42:71). Similarly, Zimmer and colleagues reported that PET-positive patients had a mean thyroglobulin level of 300 ng/mL with a range of 26 to over 700 ng/mL, and PET-negative patients had a mean level of 30 ng/mL and a range of 3 to 44 ng/mL (Zimmer LA et al. Otolaryngol Head Neck Surg. 2003;128:178).

Should patients be withdrawn from thyroid hormone prior to PET scanning? Early attempts to answer this critical question found the data too confusing to support recommendations. Sub-sequently, however, in a small study (N=8) comparing scan results during suppression and during stimulation, FDG PET scans were abnormal in four patients during suppression and in five patients during stimulation. Two patients had more lesions identified during stimulation. The authors claimed clinical management improvement in two patients based on TSH-stimulated scanning (van Tol KM et al. Thyroid. 2002; 12:381). Petrich and colleagues also compared scan results during TSH suppression and following rhTSH administration in 30 patients and observed that based on 15 surgically confirmed lesions, sensitivity of FDG PET sensitivity was 53% during TSH suppression and 87% following rhTSH stimulation. They concluded that rhTSH FDG PET suggested specific therapeutic interventions in 57% of patients, with surgery indicated in 23% (Petrich T et al. Eur J Nucl Med. 2002;29:641). In a more recent randomized and prospective study of seven patients, all lesions seen during TSH suppression were seen during rhTSH stimulation, and four additional loci were seen with rhTSH. One patient was positive only with rhTSH stimulation (Chin BB et al. JCEM. 2004;89:91).

Based on current data on the diagnostic sensitivity of FDG PET, the Centers for Medicare and Medicaid Services began covering the expense of this procedure for thyroid cancer monitoring under tightly controlled circumstances effective October 1, 2003. Coverage applies only to scans conducted with full- or partial-ring PET systems for restaging of recurrent or residual thyroid cancer of follicular cell origin that has previously been treated by thyroidectomy and radioiodine ablation in patients with serum thyroglobulin levels of 10 ng/mL or greater and negative 131I whole-body scans.

Dr. Kloos concluded with his three personal caveats regarding FDG PET. First, beware of false positives, and always confirm the existence of a lesion by a secondary modality if clinical management is dependent on the result. Second, recognize what you are looking for when imaging: usually a lesion that can be removed by surgery or treated by means other than 131I. The most common site of metastasis is the cervical lymph nodes, which infrequently require PET for detection and may be smaller than the 8-millimeter resolution of PET. Thus skilled ultrasonography of the neck and superior mediastinum is recommended before FDG PET scanning. Third, the lungs are the most common site of distant metastasis, and FDG PET has decreased sensitivity for miliary pulmonary disease. A thin-cut helical chest CT should be undertaken prior to PET imaging.

Therapeutic Approaches to Patients with Persistent Disease

Richard J. Robbins, MD and his colleagues at the Memorial Sloan-Kettering Cancer Center have been seeking ways to manage thyroid cancer without the occasional prescribed hypothyroidism associated with activating thyrocytes to soak up radioiodine. This quest led them to an investigational off-label use of rhTSH and to a retrospective study designed to evaluate this alternative means of preparing patients for 131I therapy. In the study, immediately after thyroidectomy 42 patients underwent treatment-induced total thyroid hormone withdrawal and 45 patients were prepared with rhTSH. The two groups were evenly matched with two exceptions: The group that underwent thyroid hormone withdrawal (THW) had a younger mean age by 7.7 years, and those prepared by rhTSH were less likely to have had lymph node involvement. In addition, prior to ablation treatment, patients on THW had high TSH levels for approximately 3 weeks compared with only 2 days in the rhTSH group. All patients went through full dosimetry and were treated, with patients in the rhTSH group receiving a slightly but not significantly lower dose. All patients were called back after 12 months to evaluate results.

As Figure 1 indicates, the response rates and percentages of successful ablation in the two groups were virtually identical, indicating that the two methods of preparation for radioiodine treatment were equally effective in preparing thyrocytes to consume iodine. Other investigators have published similar results from rhTSH-assisted remnant ablation trials, although one that used a very low dose of 131I was less encouraging. If these results are confirmed in large prospective trials, they may lead to a standardized protocol in which patients are started on exogenous thyroid hormone immediately after surgery and are treated with two doses of rhTSH immediately before radioiodine therapy for remnant ablation.

For older patients and those with high-risk disease, preparation now consists of a week of dosimetry consisting of rhTSH on Monday and Tuesday followed by 4 or 5 days of blood and whole-body counts. For younger patients with low-risk disease, this is unnecessary because their renal clearance allows them theoretically to tolerate doses many multiples of the standard 75 or 100 millicuries. For these patients, therefore, rhTSH is given on Monday and Tuesday followed by a small dose of 123I on Tuesday afternoon and by a whole-body scan on Wednesday. Assuming no unexpected findings, high-dose radioiodine is administered on Wednesday afternoon. One week later patients undergo post-therapy scanning. This method results in a 90% success rate, equivalent to that of the preparatory regimen of four doses of rhTSH.

If rhTSH is sufficiently effective at activating the sodium-iodide symporters to destroy remnants of normal thyroid cells, could this also apply to metastatic cells? This question has been under investigation since 1997, but has typically been reported as individual cases. Large numbers were unavailable until after the drug’s manufacturer made it available on a compassionate-use basis. Thereafter, studies of up to 47 patients reported encouraging partial-response and stable-disease rates when rhTSH was used in preparation for radioiodine treatment for metastatic disease.

Against this background, Dr. Robbins and colleagues conducted a retrospective study of treatment results of radioiodine therapy comparing THW (N=134) and rhTSH (N=128) preparation in a setting of first treatment for metastatic thyroid disease. Patients in the rhTSH trial arm continued to take thyroid hormone. The 1-year follow-up results demonstrated that the two groups were statistically alike. These findings appear to confirm that rhTSH preparation for radioiodine therapy aimed at thyroid cancer metastases may be as effective as the conventional method of treatment-induced hypothyroidism.