| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
RAPID COMMUNICATION |
Departments of Clinical Pathology (R.A., I.W., M.K.G.), Endocrinology, Diabetes and Metabolism (S.S.K.R.), and General Surgery (K.W., A.S., M.M.), The Cleveland Clinic Foundation, Cleveland, Ohio 44195
Address all correspondence and requests for reprints to: Manjula K. Gupta, Ph.D., Department of Clinical Pathology, L-30, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, Ohio 44195. E-mail: guptam{at}ccf.org.
Abstract
RT-PCR for thyroglobulin (Tg) and TSH receptor (TSHR) mRNA has been used to detect circulating thyroid cancer cells. Little is known, however, regarding the preoperative sensitivity of this test to detect cancer. Seventy-two patients with thyroid disease (36 with malignancy and 36 with benign disease) were evaluated preoperatively. TSHR and Tg mRNA transcripts were detected by RT-PCR assays, previously determined to be specific for cancer cells. There was 100% concordance between TSHR and Tg mRNA RT-PCR results. Of 36 cancer patients, 11 had recurrent disease, and all were positive by RT-PCR. Among 25 patients with no prior thyroid surgery, 18 tested positive preoperatively (sensitivity 72%). Seven of 36 patients with benign disease tested positive (specificity 80%). The overall preoperative diagnostic accuracy was 77%. Preoperative fine-needle aspiration (FNA) biopsy was performed on 46 of 61 patients with no prior thyroid surgery. FNA was diagnostic in 28 (61%) patients. Preoperative cytology was adequate but not diagnostic in 18 (39%) patients. RT-PCR correctly classified 14 of these 18 patients with indeterminate FNA, and the test detected three of four cancer patients as positive (75% sensitive) and 11 of 14 patients (78% specific) with benign disease as negative. The combined diagnostic performance characteristics for RT-PCR and FNA cytology were sensitivity = 95%, specificity = 83%, and diagnostic accuracy = 89%, with positive and negative predictive values of 84 and 95%, respectively. Our results suggest that the molecular detection of circulating thyroid cancer cells by RT-PCR for TSHR/Tg mRNA complements FNA cytology in the preoperative differentiation of benign from malignant thyroid disease and their combined use may save unnecessary surgeries.
THE PREOPERATIVE DIAGNOSIS of differentiated thyroid cancer (DTC) relies solely on fine-needle aspiration (FNA) biopsy. However, instances of inadequate sampling of the lesion and overlapping cytological features of benign and malignant thyroid neoplasms are inherent limitations of this technique, often requiring multiple biopsies or unnecessary surgeries (1, 2). Therefore, a more reliable marker that alone or in combination with FNA could detect the presence of cancer for preoperative evaluation is highly desirable and could save a number of unnecessary surgeries.
Reports of qualitative or quantitative thyroglobulin (Tg) RT-PCR assays detecting circulating thyroid-specific transcripts are readily available (3, 4, 5, 6, 7, 8, 9, 10, 11), although the specificity has been questioned. Various protocols have been developed, but some have argued that the extensive amplification process may lead to interference by illegitimate Tg gene transcription in circulating blood cells (10). This concern is validated by the fact that circulating Tg mRNA has also been found in normal subjects, as well as those with benign thyroid disease without a diagnosis of DTC (3, 6, 9, 10, 11). Thus, the use of Tg mRNA for detection of residual/metastatic disease is limited only to thyroidectomized patients. Circulating Tg mRNA has a comparable sensitivity to conventional serum Tg assays and has the potential applicability for use in patients with anti-Tg autoantibodies (3, 7, 8). We have developed specific RT-PCR assays for circulating TSH receptor (TSHR) and Tg mRNA (12, 13). The high specificity of RT-PCR assays suggested their possible role in preoperative diagnosis of thyroid cancer. In this study, we assessed the preoperative diagnostic performance of the assays in patients who were referred for thyroid surgery and compared the results with FNA cytology and with final pathological findings.
Patients and Methods
Patients
A total of 72 consecutive patients carrying the diagnosis of thyroid disease that is benign, malignant, or of as yet undetermined malignant potential, who were referred to our endocrine surgery clinic and who had signed informed consent form, were enrolled in this Institutional Review Board-approved study. All 72 patients had a preoperative sample drawn either during preoperative laboratory testing or on the day of surgery. Ultrasound-guided FNA biopsies were performed in 46 of 72 patients as a routine diagnostic work-up, and when not performed at our institution, the cytology slides were obtained and reviewed by one of our experienced pathologists specializing in thyroid cytology. Criteria for FNA sample adequacy included sufficient number of cells, abundant colloid, and the presence of at least six groups of benign follicular cells composed of at least 10 cells each. The results were compared with final postoperative pathological diagnosis.
RT-PCR
Methods we have previously described in detail were used (12, 13). Briefly, 5 ml of venous blood was collected, and mononuclear cells were separated using Ficoll Hypaque gradient within 24 h after collection. RNA was extracted from the mononuclear cells immediately after the Ficoll separation using TRIzol reagent (Life Technologies, Carlsbad, CA), and 1 µg was reverse transcribed with Superscript II preamplification system (Life Technologies). PCR was performed using carefully selected primers based on specificity (no illegitimate transcription), as documented in our previous publications (12, 13). PCR was carried out for a total of 38 cycles [94 C for 1 min (first cycle for 2 min), 62 C for 1 min, and 72 C for 1 min (10 min for the last cycle)]. PCR products generated with RNA from a thyroid cancer tissue and from a peripheral blood sample (DTC patient) were sequenced with ABI-PRISM 310 genetic analyzer (Applied Biosystems, Foster City, CA) using their BigDye Terminator v3.1 sequencing kit. The sequence of the transcript was identical to TSHR mRNA sequence, confirming the presence of authentic receptor mRNA. Glyceraldehyde-3-phosphate dehydrogenase was used as a control for successful RNA extraction and transcription and PCRs. RT-PCR products were resolved on 2% gel electrophoresis and visualized by ethidium bromide staining. Gel images were captured in "live mode" automated setting (integration/exposure time, 0.2 sec) with the use of Gel-doc 1000 (Bio-Rad, Hercules, CA) system and software.
Data analysis
TSHR mRNA and FNA results were compared with final pathological diagnoses. The number of patients correctly classified (diagnostic accuracy) by TSHR/Tg mRNA or by FNA singly or in combination was calculated and compared using 
2 test.
Results
A total of 72 patients were enrolled in the study (61 females and 11 males; age range, 18–88 yr; mean, 50 yr). Forty-six of these patients had FNA biopsy performed before surgery. Postoperative pathological diagnosis was used to categorize the patients into benign and malignant thyroid disease groups (Table 1
). Sixty-one patients (25 DTC and 36 benign) were enrolled at initial diagnosis (with no prior thyroid surgery). Eleven patients had recurrent or residual malignant disease at the time of enrollment, and 10 of these had prior radioactive iodine ablation.
|
shows the representative RT-PCR products for TSHR (212 bp) and Tg (408 bp) in seven patients. A total of 36 patients were positive by RT-PCR, including 29 of 36 (sensitivity 80%) cancer patients and 7 of 36 (specificity 80%) benign disease patients. Among 25 patients with initial diagnosis, 19 were positive by RT-PCR (sensitivity 72%) (Table 1
). All false negatives had pathological diagnosis of papillary thyroid carcinoma (PTC) and included three patients with lymph node metastasis. False positives included two patients with follicular adenoma (FA), one patient with Hurthle cell adenoma, one patient with hyperplastic oxyphilic nodule, and three patients with very large multinodular goiters (MNG).
|
). Overall 18 of 46 (39%) patients had adequate specimen but inconclusive/indeterminate FNA biopsy results. The diagnostic sensitivity of FNA was 82%, but overall diagnostic accuracy (efficiency) was only 61%. Results are summarized in Table 2.
|
. The combined diagnostic sensitivity (95%) and diagnostic efficiency (89%) for RT-PCR and FNA was significantly higher than FNA alone (P = 0.001).
|
In this study, 39% of FNA biopsies were called indeterminate. According to previous reports, the vast majority (78%) of these were indeed found to have benign histology (1, 2, 14, 15). These indeterminate thyroid nodules seem to be one of the most frustrating and challenging areas for endocrinologists and endocrine surgeons. Most studies are focused on FNA material to identify molecular markers or a combination of markers as means of improving the accuracy of diagnosis made by FNA (16, 17, 18, 19). A reliable and satisfying method that is able to differentiate preoperative malignant potential in patients presenting with thyroid nodules has not yet been proposed.
Our results suggest that circulating preoperative TSHR/Tg mRNA acts as an adjunctive test to enhance the diagnostic accuracy of FNA and classified 78% (14 of 18) of nodules with indeterminate FNA accurately. There were only three false-positive patients and one false-negative patient with TSHR/Tg mRNA among the FNA indeterminate group.
It is recognized that FNA cytology has a high sensitivity for PTC. Also, in this series, FNA correctly identified all except two patients with equivocal results. In comparison, RT-PCR had relatively low sensitivity for PTC and was negative in seven patients, including three patients with lymph node metastasis. Factors responsible for these false-negative results remain unclear at present and may include technical errors and sampling problems, or they may relate to inefficient reverse transcription or nonspecific inhibitors of the PCR. Among these factors, the technical error is less likely because repeat analysis using a second PCR produced the same results. It is possible that the efficiency of reverse transcription or PCR may be the limiting factor in this assay, and these factors are currently being investigated in our laboratory.
On the other hand, as expected, FNA cannot differentiate between FC and FA (14, 15). In this study, we had seven patients with follicular neoplasms, and FNA was nondiagnostic for all except one. Hence, follicular neoplasms are often grouped together as indeterminate or follicular patterned thyroid lesions and require a substantial number of FNAs and unnecessary diagnostic thyroid lobectomies.
To date, there are no known markers that can distinguish FAs from cancer with certainty because, like FC, a significant number of FAs harbor Ras mutations (20) and show galectin-3 immunostaining (21, 22). It is suggested that some FAs may represent a premalignant stage of FCs (23). In our previous report, we demonstrated the high sensitivity of circulating TSHR mRNA in detecting recurrent/residual disease in patients with all DTCs regardless of histological types, including follicular and Hurthle cell cancers (13). In this study, this test detected both patients with FCs as positive and three of five FA as negatives, thus correctly classifying five of seven follicular neoplasms. Although the number of follicular neoplasms is small in this series, our results indicate a high potential of TSHR mRNA to differentiate FC from FA and await confirmation in future studies with a larger number of patients. Besides two FAs, other false positives were one MNG with hyperplastic nodules, predominantly Hurthle cell type, and three patients with extremely large MNG, suggesting that such patients may have circulating thyroid cells. This may be due to the presence of high hyperplastic activity in thyroid nodule, or it is possible that an occult DTC may have been present and was overlooked in the pathology examination.
In conclusion, our results demonstrate that circulating TSHR/Tg mRNA has lower sensitivity to detect PTC than FNA at initial diagnosis. However, TSHR/Tg mRNA shows promise for detecting FC, which is often missed by FNA. Its value resides in identifying benign thyroid disease among patients with equivocal FNA. Overall, it may serve as a valuable adjunct to FNA for identifying thyroid cancer from benign disease.
Footnotes
First Published Online February 1, 2005
Abbreviations: DTC, Differentiated thyroid cancer; FA, follicular adenoma; FC, follicular carcinoma; FNA, fine-needle aspiration; MNG, multinodular goiter(s); PTC, papillary thyroid carcinoma; Tg, thyroglobulin; TSHR, TSH receptor.
Received September 9, 2004.
Accepted January 20, 2005.
References
rearrangement in thyroid tumors: evidence for distinct molecular pathways in thyroid follicular carcinomas. J Clin Endocrinol Metab 88:2318–2326This article has been cited by other articles:
 |
|
 |
|
  K. Fujarewicz, M. Jarzab, M. Eszlinger, K. Krohn, R. Paschke, M. Oczko-Wojciechowska, M. Wiench, A. Kukulska, B. Jarzab, and A. Swierniak A multi-gene approach to differentiate papillary thyroid carcinoma from benign lesions: gene selection using support vector machines with bootstrapping Endocr. Relat. Cancer, September 1, 2007; 14(3): 809 - 826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-Y. Chia, M. Milas, S. K. Reddy, A. Siperstein, M. Skugor, J. Brainard, and M. K. Gupta Thyroid-Stimulating Hormone Receptor Messenger Ribonucleic Acid Measurement in Blood as a Marker for Circulating Thyroid Cancer Cells and Its Role in the Preoperative Diagnosis of Thyroid Cancer J. Clin. Endocrinol. Metab., February 1, 2007; 92(2): 468 - 475. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
