Mount Sinai School of Medicine New York, NY 10029
Address correspondence and requests for reprints to: Richard S. Haber, M.D., Mount Sinai School of Medicine, Box 1055, One Gustave L. Levy Place, New York, New York 10029. E-mail: rhaber{at}smtplink.mssm.edu
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
Current surveillance methods include whole-body radioiodine scanning and measurement of serum thyroglobulin. In addition, cervical lymph node metastases may be detected by imaging with ultrasonography, magnetic resonance imaging, or computed tomography scanning. Radioiodine scanning requires prior stimulation with TSH, traditionally accomplished by withdrawal of thyroid hormone treatment, with subsequent debilitating hypothyroid symptoms. Moreover, radioiodine scanning may be negative in many patients with metastases detectable by other methods (3, 4). Measurement of serum thyroglobulin, especially when performed with TSH stimulation, may provide increased sensitivity compared with radioiodine scanning alone (5, 6), but the thyroglobulin immunoassay results may be uninterpretable due to the presence of interfering antithyroglobulin antibodies in many patients. By the time this editorial appears in print, TSH stimulation by injection of recombinant human TSH may be available and may obviate the necessity for thyroid hormone withdrawal and hypothyroidism before radioiodine scanning and serum thyroglobulin immunoassay (7). Nonetheless, a sensitive and convenient screening test to diagnose recurrent thyroid cancer, not requiring TSH stimulation or the absence of thyroglobulin antibodies, would be useful. Such an assay could identify a selected group of patients in need of further diagnostic procedures to localize disease before appropriate therapy with surgery and/or radioiodine.
An initial report (8) and a separate report in this issue of The Journal of Clinical Endocrinology and Metabolism (9) (see page 4435) have indicated that the detection of circulating thyroid cells by a PCR-based assay for thyroglobulin messenger RNA (mRNA) in blood may provide a new and convenient method for diagnosing recurrent thyroid cancer. The concept that cancer cells might be present in the circulation is not new, but it was not until the advent of the polymerase chain reaction in the 1980s that sufficiently sensitive methods to detect such cells became available. In a variety of malignancies, including leukemia, prostate cancer, neuroblastoma, and hepatocellular carcinoma, PCR-based detection of tumor- or tissue-specific mRNA species has been used to detect tumor cells in blood or lymph nodes (10). These techniques entail the synthesis by reverse transcription (RT) of complementary DNA (cDNA) from RNA present in tissue or blood, the amplification by PCR of a target cDNA sequence, and the detection of this PCR product, typically by gel electrophoresis. Because extracellular RNA is rapidly degraded, a positive result in an RT-PCR assay indicates the presence of cells expressing the particular tumor-specific or tissue-specific mRNA of interest.
The application of this technique to detect recurrent thyroid cancer was first reported by Ditkoff et al. (8). Using an RT-PCR assay these authors demonstrated the presence of thyroglobulin mRNA in blood in 9 of 9 patients (100%) with known current metastatic papillary or follicular carcinoma. Among patients without known current metastatic disease, the assay was positive in 5 of 22 patients (23%) who had a history of treated metastases, but in only 2 of 56 patients (4%) without such a history. The authors reported that the assay was negative in the blood of normal subjects and patients with benign thyroid disease and was sufficiently sensitive to detect 20 cells from a follicular carcinoma cell line in 1 mL blood.
An RT-PCR method to detect thyroglobulin-expressing cells in fine-needle aspiration specimens from suspicious cervical lymph nodes was reported by Arturi et al. in 1997 (11) and was more sensitive and accurate for detecting metastatic thyroid cancer than simultaneous measurement of thyroglobulin protein by immunoassay. Subsequently, Tallini et al. (12) reported that thyroglobulin mRNA could be detected in blood by RT-PCR in 20% of patients with benign nodules and in more than 50% of patients with treated thyroid cancer. These authors could not demonstrate any correlation between the thyroglobulin mRNA assay result and the clinical status of the thyroid cancer patients. As in the report by Ditkoff et al. (8), normal controls lacked detectable thyroglobulin mRNA in blood, unless the PCR amplification was extended from 30 to 40 cycles.
Ringel et al. (9) now report the results of an RT-PCR assay for thyroglobulin mRNA in the blood of 68 patients with treated papillary or follicular thyroid cancer, and compare the results with simultaneous thyroglobulin protein immunoassay during thyroxine therapy, and with whole-body radioiodine scanning after thyroxine withdrawal. Almost all patients in this study had been rendered athyreotic, at least in theory, by earlier near-total thyroidectomy and radioiodine ablation. Of 14 patients with known metastases based on radioiodine scanning, all had detectable thyroglobulin mRNA in blood during thyroxine treatment, whereas thyroglobulin protein was undetectable by simultaneous immunoassay in 6 of these patients. Among 19 patients with residual radioiodine uptake in the thyroid bed only, thyroglobulin mRNA was detectable in the blood of 12 (63%), and thyroglobulin protein in only 3 (16%). In 35 patients with negative radioiodine scans, thyroglobulin mRNA was detectable in 7 (20%), and thyroglobulin protein in only 1 (3%). In addition, in 4 patients with antithyroglobulin antibodies, the RT-PCR assay for thyroglobulin mRNA was positive in the 2 with known metastases. Of note, the assay was also positive in all 10 normal control subjects, and putative circulating thyroid cells could be identified in normal control blood by thyroglobulin immunostaining.
These data confirm the findings of Ditkoff et al. (8) that an RT-PCR assay for thyroglobulin mRNA in blood can identify most patients with metastatic thyroid cancer detected by other means, without withdrawal of thyroid hormone and TSH stimulation, and even in the presence of antithyroglobulin antibodies. The RT-PCR assay reported by Ringel et al. (9) appears to be more sensitive than the previously reported assays, and indeed normal control subjects with intact thyroid glands had detectable thyroglobulin mRNA in blood, apparently representing a small number of circulating thyroid cells. As discussed by the authors, a variety of technical factors may explain this high sensitivity, most obviously the higher number of PCR amplification cycles employed.
What is less clear is the clinical significance of positive assay results in nominally athyreotic patients without known current metastases. Such results were found in 7/78 (10%) of patients in the Ditkoff study (5 of the 7 had a history of treated metastases) and in 19/44 (43%) of patients in the Ringel study. Not surprisingly, the likelihood of a positive assay result is greater in the latter study using the more sensitive method. But how many of these patients with detectable thyroglobulin mRNA in blood have clinically significant disease? PCR-based assays are notoriously susceptible to artefact due to contamination with nucleic acid containing the target sequence, and there is even the possibility of "illegitimate transcription" of very small amounts of thyroglobulin mRNA by nonthyroid cells in blood. Given the relatively indolent behavior of most thyroid cancers, the presence of apparent "minimal residual disease" detected by ultrasensitive PCR techniques does not necessarily imply a risk to the patients health. The problem is analogous to the finding of mildly elevated thyroglobulin protein by immunoassay in that the practical importance is uncertain. A number of questions remain to be answered to determine the clinical meaning of a positive thyroglobulin RT-PCR assay. First, how many of these patients have clinically evident metastases by criteria other than radioiodine scanning? Sonography of the anterior neck, combined with fine needle aspiration of suspicious lymph nodes might reveal recurrent disease (4), as might other imaging techniques in the neck and chest such as computerized tomography or magnetic resonance imaging. It would also be of interest to know thyroglobulin immunoassay values in this group of patients under conditions of TSH stimulation (data are reported on only 7 patients after thyroid hormone withdrawal in the study by Ringel et al.). Finally, the likelihood of a clinical recurrence of thyroid cancer can be predicted by clinical factors such as tumor invasiveness and size, previous metastases, and age of the patient (13). A strong correlation of thyroglobulin RT-PCR results with such clinical predictors of recurrence would help confirm the meaningfulness and specificity of the assay. Ultimately, however, the clinical usefulness of this assay will need to be demonstrated by correlation with long-term clinical follow-up data. In addition, the value of this assay compared with the most sensitive current method, thyroglobulin protein immunoassay during TSH stimulation, needs to be established in future studies.
The development of a clinically useful RT-PCR assay for thyroglobulin mRNA in blood will require the application of quantitative techniques, which are currently under investigation by at least two groups (14, 15). Such techniques might permit a clinical distinction between weakly positive assay results associated with a low risk of clinical recurrence of cancer and more strongly positive results indicative of a high risk. As noted by Ringel et al., a quantitative assay could also be used to track disease progression. Since the majority of thyroid cancer patients never experience a clinical recurrence of their disease, a clinically accurate RT-PCR assay could spare the majority of these patients expensive and troublesome surveillance testing and provide an emotional benefit of reassurance as well, while identifying those patients in need of further testing and treatment.
Received October 6, 1998.
Accepted October 6, 1998.
![]() |
References |
---|
![]() ![]() ![]() |
---|
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |