Galectin-3 and CD44v6 as Markers for Preoperative Diagnosis of Thyroid Cancer by RT-PCR : Diagnostic Molecular Pathology

Journal Logo

00019606-201112000-00007ArticleDiagnostic Molecular PathologyDiagnostic Molecular Pathology© 2011 Lippincott Williams & Wilkins, Inc.20December 2011 p 233–241Galectin-3 and CD44v6 as Markers for Preoperative Diagnosis of Thyroid Cancer by RT-PCROriginal ArticlesŠamija, Ivan PhD; Mateša, Neven MD, PhD; Lukač, Josip PhD; Kusić, Zvonko MD, PhDDepartment of Oncology and Nuclear Medicine, University Hospital “Sestre milosrdnice”, Zagreb, CroatiaThis work was supported by the Ministry of Science, Education and Sports of the Republic of Croatia (research grants 0134-1342428-2427 and 134-1342428-2430).Reprints: Ivan Samija, PhD, Department of Oncology and Nuclear Medicine, University Hospital “Sestre milosrdnice”, Vinogradska cesta 29, HR-10000 Zagreb, Croatia (e-mail: [email protected]).AbstractThe aim of the study was to determine the diagnostic value of reverse transcriptase polymerase chain reaction (RT-PCR) analysis of galectin-3 and CD44v6 as markers for preoperative diagnosis of malignancy in lesions of the thyroid. RT-PCR analysis of galectin-3 and CD44v6 expression was performed on RNA isolated from fine-needle aspirates of thyroid lesions from 428 patients. The results were evaluated against the postoperative histopathological diagnosis or definitive cytological diagnosis in cases of nodular goiter and Hashimoto thyroiditis. A total of 57 (13%) samples were inadequate for RT-PCR. Galectin-3 and CD44v6 were positive in 167 (45%) and 158 (43%) out of 371 adequate samples, respectively. Galectin-3 and CD44v6 were positive in 56 (86%) and 54 (83%) out of 65 papillary carcinomas, in 16 (29%) and 18 (32%) out of 56 Hashimoto's thyroiditis, in 61 (34%) and 52 (29%) out of 181 nodular goiters, in 23 (43%) and 23 (43%) out of 53 follicular adenomas, in 3 (100%) and 3 (100%) out of 3 follicular carcinomas, and in 8 (62%) and 8 (62%) out of 13 Hurthle cell adenomas, respectively. Specificity, sensitivity, and positive and negative predictive values in discriminating between malignant and benign thyroid nodules were 64, 87, and 35 and 96% for galectin-3; 67, 84, and 36 and 95% for CD44v6; and 79, 82, and 47 and 95% for the analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive), respectively. Owing to relatively low specificity, the clinical value of galectin-3 and CD44v6 analysis by RT-PCR as a marker for preoperative diagnosis of malignancy in thyroid lesions is limited.Palpable thyroid nodules are common in adult population, but only a small proportion of them (5% to 6%) are malignant.1,2 The most reliable and also practical and cost-effective diagnostic procedure for thyroid nodules is cytological analysis of fine-needle aspiration biopsy (FNAB) samples.3,4 A reliable diagnosis regarding malignancy can be made by this procedure in a majority of cases. However, in 10% to 30% of cases, mostly follicular thyroid lesions, a reliable diagnosis regarding malignancy cannot be made based on cytological analysis of FNAB samples, primarily because it is difficult to distinguish benign (eg, hyperplastic nodules, follicular adenoma, and Hurthle cell adenoma) from malignant (eg, follicular carcinoma, Hurthle cell carcinoma, and follicular variants of papillary carcinoma) follicular thyroid lesions this way.5 These patients are being surgically treated for diagnostic purposes, although most of them have benign thyroid lesions.6 Introduction of reliable molecular markers that would increase the accuracy of preoperative diagnosis could spare some of these patients unnecessary surgery.Many markers have been studied for this purpose, including mutations in certain genes, the expression of certain genes analyzed both on the mRNA level by reverse transcriptase polymerase chain reaction (RT-PCR) and on the protein level by immunocytochemistry, and the expression of certain microRNAs.7,8 One such extensively studied marker, which also gave promising results, is galectin-3.9,10 Galectin-3 is approximately 30 kDa member of a lectin group of β-galactosyl-binding proteins, and is involved in different biological processes, including cell adhesion, cell-cycle regulation, apoptosis, and tumor progression.11 In several studies on large sample of patients with thyroid nodules, immunocytochemical analysis of galectin-3 expression in FNAB samples has shown relatively high diagnostic accuracy for discriminating malignant from benign thyroid lesions.12–14CD44v6 is another molecule studied as a marker for preoperative diagnosis of thyroid lesions. CD44v6 is an isoform of CD44, a cell surface glycoprotein receptor for hyaluronic acid involved in intercellular interaction, cell adhesion, migration, and also tumor growth and progression.15 CD44 can be expressed on the cell surface as a standard isoform CD44s, which is the only one expressed under normal conditions on nonproliferating thyrocytes, or as multiple variant isoforms, which are products of alternative splicing.15–17 CD44v6 is one such isoform that is upregulated in thyroid carcinoma.17,18 CD44v6 has been studied in only a few studies as a marker for preoperative diagnosis of thyroid lesions, both by immunocytochemistry and RT-PCR, showing overall lower diagnostic accuracy in comparison with galectin-3.13,16,19–21RT-PCR is a sensitive, high-throughput, standardized, and easy-to-perform method for gene expression analysis, which can be easily performed on FNAB samples.22 RT-PCR has also some advantages compared with immunocytochemistry, of which most notable are that no trained expert cytologist is needed to analyze and interpret the results and that results do not depend on the subjective interpretation of the cytologist. Nevertheless, preoperative analysis of galectin-3 expression by RT-PCR was studied only in a few studies on a smaller number of patients and with more conflicting results regarding its diagnostic accuracy in comparison with the immunocytochemical analysis of galectin-3 expression.16,20,21,23,24Therefore, in this prospective study we analyzed preoperatively galectin-3 expression by RT-PCR on a large sample of patients with thyroid lesions. We also analyzed CD44v6 marker to investigate whether the analysis of CD44v6 in addition to galectin-3 improves diagnostic accuracy. The aim of our study was to determine the diagnostic value of RT-PCR analysis of galectin-3 and CD44v6 markers for preoperative diagnosis of malignancy in lesions of the thyroid.MATERIALS AND METHODSPatients and SamplesUltrasound-guided FNAB was performed by a cytopathologist as part of the standard diagnostic protocol for patients with thyroid nodules in the Department of Oncology and Nuclear Medicine at the University Hospital “Sestre milosrdnice” in Zagreb, Croatia. A total of 1 to 3 punctures per nodule, depending on the size of the nodule, were performed using a 25-gauge needle. The leftover material in the needle was used for RT-PCR analysis after the aspirate was smeared for conventional cytology (May-Gruenwald Giemsa staining). Informed consent according to the World Medical Association Declaration of Helsinki was obtained from all patients before their inclusion in the study. The study was approved by the Ethical Committee of the University Hospital “Sestre milosrdnice,” Zagreb, Croatia, and was performed in accordance with the Declaration of Helsinki ethical guidelines.25 No decision regarding surgery or any other for the patients was influenced by the result of galectin-3 and CD44v6 analysis. In the study were included samples only from patients who subsequently underwent surgery and for whom definitive postoperative histopathological diagnosis was available, from patients with definitive cytological diagnosis of nodular goiter, and from patients with definitive preoperative diagnosis of Hashimoto thyroiditis on the basis of cytology and of high titer of anti-nuclear antibodies. Samples from lesions with a satisfactory cytological diagnosis other than that of follicular cell origin (intrathyroid parathyroid adenoma and medullary carcinoma) were excluded from the study. Samples in which no groups of well-preserved follicular cells were present were considered to be cytologically inadequate and were not analyzed by RT-PCR. Thyroid nodule samples from 428 patients collected between January 2003 and June 2010 were analyzed in the study.RNA Isolation and Reverse TranscriptionTotal cellular RNA was isolated from aspirates using the TriPure Isolation Reagent (Roche, Indianapolis, IN) according to the manufacturer's instructions. RNA pellet was dissolved in deionized water and stored at −70°C. For reverse transcription (RT), 5 μL of RNA was mixed with 0.5 μg of Oligo d(T)18 primer (New England Biolabs, Beverly, MA) and incubated for 4 minutes at 70°C. Other components were then added to this mixture in the final concentrations of 1×RT buffer (New England Biolabs), 0.5 mmol/L of each dNTPs (Sigma Chemical Company, Saint Louis, MO), 1 U/μL of RNase inhibitor (Roche, Mannheim, Germany), and 1.25 U/μL of M-MuLV reverse transcriptase (New England Biolabs) in a total volume of 20 μL. The mixture was first incubated at 42°C for 60 minutes, then at 94°C for final 5 minutes and stored at −20°C.PCRThe expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH), thyroglobulin, galectin-3, and CD44v6 were analyzed by PCR. A housekeeping gene for GAPDH was amplified in all samples to determine the integrity of the RNA. Thyroglobulin expression was analyzed to confirm the presence of cells of follicular origin. Samples were considered adequate for RT-PCR analysis only if the expression of both GAPDH and thyroglobulin was detected by RT-PCR, and only these samples were analyzed further for galectin-3 and CD44v6 expression.For the PCR analysis, 2 μL of RT mixture was added to the PCR mixture containing final concentrations of 1×Taq buffer with Mg2+ (Eppendorf, Hamburg, Germany), 0.2 mmol/L of each dNTPs (Sigma Chemical Company), 0.2 mmol/L of sense primer, 0.2 mmol/L of antisense primer, and 0.02 U/μL of Taq DNA polymerase (Eppendorf) in a total volume of 25 μL. PCR conditions for the analysis of GAPDH, galectin-3, and CD44v6 expression were 1 cycle of 5 minutes at 95°C, followed by 30 cycles for GAPDH, 25 cycles for galectin-3, and 27 cycles for CD44v6, each consisting of 1 minute at 95°C, 1 minute at 55°C, and 45 seconds at 72°C, followed by a final 10-minute extension at 72°C. Reaction conditions for PCR analysis of thyroglobulin expression were 1 cycle of 5 minutes at 95°C, followed by 32 cycles each consisting of 1 minute at 95°C, 30 seconds at 60°C, and 45 seconds at 72°C, followed by a final 10-minute extension at 72°C. Previously described oligonucleotide primer sequences were used for GAPDH, thyroglobulin, galectin-3, and CD44v6 cDNA amplification.16,26,27All PCR products were analyzed by electrophoresis on 2% agarose gel stained with ethidium bromide, and directly visualized under UV light at 302 nm. DNA molecular weight markers VIII or IX (Roche) were included in all gels. A sample was regarded as positive if a band of expected size [623 base pairs (bp) for GAPDH, 529 bp for thyroglobulin, 587 bp for galectin-3, and around 700 bp for CD44v6] was present.Statistical AnalysisThe analysis was performed using MedCalc (version 10.4.0.0; MedCalc Software bvba; Mariakerke, Belgium) and a P value of less than 0.05 was considered to be statistically significant. Sensitivity was calculated as the ratio of the number of carcinomas positive for galectin-3 or CD44v6 to the total number of carcinomas. Specificity was calculated as the ratio of the number of benign lesions negative for galectin-3 or CD44v6 to the total number of benign lesions. The positive predictive value was calculated as the ratio of the number carcinomas positive for galectin-3 or CD44v6 to the total number of lesions positive for galectin-3 or CD44v6. The negative predictive value was calculated as the ratio of the number of benign lesions negative for galectin-3 or CD44v6 to the total number of lesions negative for galectin-3 or CD44v6. Diagnostic accuracy was calculated as the ratio of the number of benign lesions negative for galectin-3 or CD44v6 and carcinomas positive for galectin-3 or CD44v6 to the total number of lesions.RESULTSPatients and DiagnosisThyroid nodule FNAB samples from 428 patients (364 female and 64 male) aged 18 to 85 years (median age, 53 y) were analyzed by RT-PCR for the expression of galectin-3 and CD44v6. Samples only from patients with adequate cytological diagnosis of follicular cell origin thyroid lesion who subsequently underwent surgery and for whom definitive postoperative histopathological diagnosis was available, from patients with definitive cytological diagnosis of nodular goiter, and from patients with definitive preoperative diagnosis of Hashimoto thyroiditis on the basis of cytology and high titer of anti-nuclear antibodies were included.Among 428 samples, 57 (13%) were inadequate for RT-PCR analysis because of no expression of GAPDH and/or thyroglobulin. Among 57 inadequate samples, 38 were negative for both expressions of GAPDH and thyroglobulin, and 19 were positive for GAPDH expression, but negative for thyroglobulin expression.Among 371 samples adequate for RT-PCR, cytological diagnoses according to the Bethesda system were 189 (51%) benign (133 nodular goiter, 56 Hashimoto thyroiditis), 14 (4%) atypia of undetermined significance, 98 (26%) follicular neoplasm (among these 7 Hurthle cell neoplasms), 30 (8%) suspicious for malignancy (all suspicious for papillary carcinoma), and 40 (11%) malignant (all papillary carcinoma) (Table 1).28JOURNAL/dimp/04.03/00019606-201112000-00007/table1-7/v/2021-02-17T200039Z/r/image-tiffDistribution of 371 Patients With Samples Adequate for RT-PCR Included in the Study Regarding Postoperative Histopathological and Preoperative Cytological DiagnosisAmong 371 samples adequate for RT-PCR, 181 (49%) patients with benign cytological diagnosis (125 nodular goiters and 56 Hashimoto thyroiditis) did not undergo surgery; hence, postoperative histopathological diagnosis was not available for them. Among patients that did undergo surgery, 65 had postoperative histopathological diagnosis of papillary carcinoma, 3 had postoperative histopathological diagnosis of follicular carcinoma, 53 had postoperative histopathological diagnosis of follicular adenoma, 13 had postoperative histopathological diagnosis of Hurthle cell adenoma, and 56 had postoperative histopathological diagnosis of nodular goiter (Table 1). Among 65 patients with postoperative histopathological diagnosis of papillary carcinoma, 3 had a follicular variant of papillary carcinoma. Among 65 patients with postoperative histopathological diagnosis of papillary carcinoma, 17 (26%) had metastases at the time of diagnosis. Association of preoperative cytological diagnosis and postoperative histopathological diagnosis is given in Table 1.Galectin-3 and CD44v6 in all Patients' SamplesOverall, out of 371 samples adequate for RT-PCR, 167 (45%) were positive for galectin-3 and 158 (43%) were positive for CD44v6 (Table 2). Among positive samples, 119 (32%) were positive for both galectin-3 and CD44v6 (Table 2). There was no statistically significant difference between the results for galectin-3 and the results for CD44v6 (χ2=0.7592, P=0.3855; McNemar test). Number and percentage of positive samples for each cytological diagnosis separately are given in Table 2. Number and percentage of positive samples for each final diagnosis separately are given in Table 3. In the group of 181 nodular goiter samples, in 125 patients a definitive diagnosis was made by cytology, and in the remaining 56 patients a diagnosis was made postoperatively by histopathology. Out of 125 nodular goiter samples diagnosed by cytology, 37 (30%) were positive for galectin-3, 28 (22%) were positive for CD44v6, and among positive samples 17 (14%) were positive for both galectin-3 and CD44v6. Out of 56 nodular goiter samples diagnosed by histopathology, 24 (43%) were positive for galectin-3, 24 (43%) were positive for CD44v6, and among positive samples 16 (29%) were positive for both galectin-3 and CD44v6. Among 65 patients with a final diagnosis of papillary carcinoma, 3 had a follicular variant of papillary carcinoma and among them 2 were positive for both galectin-3 and CD44v6 and 1 was positive for galectin-3 but negative for CD44v6. Among 65 patients with a final diagnosis of papillary carcinoma, 17 (26%) had metastases at the time of diagnosis, and among these 17 patients with metastases 15 (88%) were positive for galectin-3 and 15 (88%) were positive for CD44v6. Among positive samples, 14 (82%) were positive for both galectin-3 and CD44v6. There was no statistically significant difference in a proportion of galectin-3-positive and CD44v6-positive samples between patients with papillary carcinoma who had metastases at the time of diagnosis and those who did not have (χ2=0.084, P=0.773 for galectin-3; χ2=0.881, P=0.348 for CD44v6; χ2 test).JOURNAL/dimp/04.03/00019606-201112000-00007/table2-7/v/2021-02-17T200039Z/r/image-tiffGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid lesions and Preoperative Cytological DiagnosisJOURNAL/dimp/04.03/00019606-201112000-00007/table3-7/v/2021-02-17T200039Z/r/image-tiffGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid Lesions and Final DiagnosisIn all 68 samples with malignant final diagnosis (papillary carcinoma and follicular carcinoma), 59 (87%) were positive for galectin-3 and 57 (84%) were positive for CD44v6. Among positive samples, 56 (82%) were positive for both galectin-3 and CD44v6. In all 303 samples with benign final diagnosis (nodular goiter, Hashimoto thyroiditis, follicular adenoma, and Hurthle cell adenoma), 108 (36%) were positive for galectin-3 and 101 (33%) were positive for CD44v6. Among positive samples, 64 (21%) were positive for both galectin-3 and CD44v6. In 66 samples with benign final diagnosis of adenoma (follicular adenoma and Hurthle cell adenoma), 31 (47%) were positive for galectin-3 and 31 (47%) were positive for CD44v6. Among positive samples, 21 (32%) were positive for both galectin-3 and CD44v6. Proportions of both galectin-3-positive samples and CD44v6-positive samples were significantly higher in samples with malignant final diagnosis than in samples with benign final diagnosis (χ2=56.591, P<0.0001 for galectin-3; χ2=55.857, P<0.0001 for CD44v6; χ2 test). Proportions of both galectin-3-positive samples and CD44v6-positive samples were also significantly higher in samples with malignant carcinoma final diagnosis (papillary carcinoma and follicular carcinoma) than in samples with final diagnosis of adenoma (follicular adenoma and Hurthle cell adenoma) (χ2=22.280, P<0.0001 for galectin-3; χ2=18.577, P<0.0001 for CD44v6; χ2 test).In all 371 samples adequate for RT-PCR, diagnostic accuracy, sensitivity, specificity, and positive and negative predictive values in discriminating between malignant and benign thyroid nodules were 68% [95% confidence interval (CI): 63-73], 87% (95% CI: 76-94), 64% (95% CI: 59-70), 35% (95% CI: 28-43) and 96% (95% CI: 92-98) for galectin-3; 70% (95% CI: 65-74), 84% (95% CI: 73-92), 67% (95% CI: 61-72), 36% (95% CI: 29-44) and 95% (95% CI: 91-97) for CD44v6; 80% (95% CI: 75-84), 82% (95% CI: 71-91), 79% (95% CI: 74-83), 47% (95% CI: 38-56) and 95% (95% CI: 92-98) for the analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive), and 59% (95% CI: 54-64), 88% (95% CI: 78-95), 52% (95% CI: 46-58), 29% (95% CI: 23-36) and 95% (95% CI: 91-98) for the analysis of both markers (considered positive if either galectin-3 or CD44v6 were positive), respectively (Table 4). The analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive) improved overall diagnostic accuracy compared with the analysis of either galectin-3 or CD44v6 alone (80% vs. 69% or 70%) because of increased specificity, but sensitivity was lower than the analysis of a single marker (either galectin-3 or CD44v6). In contrast, the analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive) improved sensitivity, but specificity and overall diagnostic accuracy (59% vs. 69% or 70%) were lower than the analysis of single marker (either galectin-3 or CD44v6).JOURNAL/dimp/04.03/00019606-201112000-00007/table4-7/v/2021-02-17T200039Z/r/image-tiffValue of Galectin-3 and CD44v6 Analysis by RT-PCR in Preoperative FNAB Samples in Discrimination Between Benign and Malignant Lesions of the Thyroid for all 371 Samples Adequate for RT-PCRGalectin-3 and CD44v6 in Cytologically Indeterminate SamplesAs patients with indeterminate cytological diagnosis regarding malignancy are the ones that would benefit from an analysis of diagnostic markers in preoperative FNAB samples, we also analyzed the diagnostic value of galectin-3 and CD44v6 only for this group of patients from our study. Among all 371 samples adequate for RT-PCR, there were 142 with indeterminate cytological diagnosis regarding malignancy (cytological diagnoses: atypia of undetermined significance, follicular neoplasm, and suspicious for malignancy) (Table 1). Among these 142 patients, 114 (80%) had benign (follicular adenoma, Hurthle cell adenoma, and nodular goiter) and 28 (20%) had malignant (papillary carcinoma and follicular carcinoma) final postoperative histopathological diagnosis (Table 1). In all 142 patients with indeterminate cytological diagnosis, 76 (54%) were positive for galectin-3 and 75 (53%) were positive for CD44v6 (Table 5). Among positive samples, 57 (40%) were positive for both galectin-3 and CD44v6 (Table 5). In 114 patients with indeterminate cytological diagnosis and benign final postoperative histopathological diagnosis, 54 (47%) were positive for galectin-3 and 54 (47%) were positive for CD44v6 (Table 5). Among positive samples, 37 (32%) were positive for both galectin-3 and CD44v6 (Table 5). In 28 patients with indeterminate cytological diagnosis and malignant final postoperative histopathological diagnosis, 22 (79%) were positive for galectin-3 and 21 (75%) were positive for CD44v6 (Table 5). Among positive samples, 20 (71%) were positive for both galectin-3 and CD44v6 (Table 5). Among patients with indeterminate cytological diagnosis, proportions of both galectin-3-positive samples and CD44v6-positive samples were significantly higher in samples with malignant final histopathological diagnosis than in samples with benign final histopathological diagnosis (χ2=7.588, P=0.0059 for galectin-3; χ2=5.823, P=0.0158 for CD44v6; χ2 test).JOURNAL/dimp/04.03/00019606-201112000-00007/table5-7/v/2021-02-17T200039Z/r/image-tiffGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid Lesions With Indeterminate Cytological Diagnosis Regarding Malignancy*In all 142 samples with indeterminate cytological diagnosis adequate for RT-PCR, diagnostic accuracy, sensitivity, specificity, and positive and negative predictive values in discriminating between malignant and benign thyroid nodules were 58% (95% CI: 49-66), 79% (95% CI: 59-92), 53% (95% CI: 43-62), 29% (95% CI: 19-41) and 91% (95% CI: 81-97) for galectin-3; 57% (95% CI: 48-65), 75% (95% CI: 55-89), 53% (95% CI: 43-62), 28% (95% CI: 18-40) and 90% (95% CI: 80-96) for CD44v6; 68% (95% CI: 60-76), 71% (95% CI: 51-87), 68% (95% CI: 58-76), 35% (95% CI: 23-49) and 91% (95% CI: 82-96) for the analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive), and 46% (95% CI: 38-55), 82% (95% CI: 63-94), 38% (95% CI: 29-47), 25% (95% CI: 16-34) and 90% (95% CI: 77-97) for the analysis of both markers (considered positive if either galectin-3 or CD44v6 were positive), respectively (Table 6). The analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive) improved overall diagnostic accuracy compared with the analysis of either galectin-3 or CD44v6 alone (68% vs. 58% or 57%) because of increased specificity, but sensitivity was lower than the analysis of a single marker (either galectin-3 or CD44v6). In contrast, the analysis of both markers (considered positive only if both galectin-3 and CD44v6 were positive) improved sensitivity, but specificity and overall diagnostic accuracy (46% vs. 58% or 57%) were lower than the analysis of single marker (either galectin-3 or CD44v6).JOURNAL/dimp/04.03/00019606-201112000-00007/table6-7/v/2021-02-17T200039Z/r/image-tiffValue of Galectin-3 and CD44v6 Analysis by RT-PCR in Preoperative FNAB Samples in Discrimination Between Benign and Malignant Lesions of the Thyroid for 142 Samples Adequate for RT-PCR With Indeterminate Cytological Diagnosis Regarding MalignancyDISCUSSIONExpression of galectin-3 detected by immunocytochemistry is among the most extensively studied molecular markers for preoperative diagnosis of thyroid lesions. The most relevant clinical application of such a molecular marker would be in a group of patients with thyroid nodules that have indeterminate cytological diagnosis regarding malignancy and that are currently surgically treated, although most of them have benign thyroid lesions. Expression of galectin-3 was studied both on preoperative FNAB samples and on postoperative histopathological samples and different studies comprised different ranges of diagnoses. The most relevant among these studies are prospective studies conducted on a clinically relevant group of patients with cytologically indeterminate diagnosis of thyroid lesions, in which galectin-3 expression was analyzed preoperatively on FNAB samples and was compared with subsequent histopathology diagnosis as a golden standard. In 6 studies fulfilling these criteria that were included in a meta-analysis, both specificity and sensitivity of galectin-3 in detecting malignant lesions ranged from 75% to 100% suggesting the potential utility of galectin-3 in the management of patients with thyroid nodules.10 More recently, in the largest prospective study so far on 465 patients with cytologically indeterminate thyroid lesions, galectin-3 has shown overall sensitivity of 78%, specificity of 93%, positive predictive value of 82%, and negative predictive value of 91%.14 If patients in that study had been referred to surgery based on galectin-3, not only less than 71% patients would have been surgically treated but also 22% of cancers would have been missed.14 This result still makes galectin-3 a potentially useful complimentary diagnostic marker in patients with thyroid nodules.Although these studies show potential clinical value of immunocytochemical detection of galectin-3, immunocytochemistry has certain disadvantages. A trained expert cytologist is needed to analyze and interpret the results and the results somewhat depend on subjective interpretation of the cytologist. For example, in the aforementioned large prospective multicenter study the proportion of galectin-3 immunocytochemistry-positive thyroid cancer samples varied significantly (from 39% to 94%) between different centers.14 Moreover, different studies used different methods for FNAB sample preparation and for immunocytochemical analysis of galectin-3 expression, which may also have influenced different results obtained in these studies.10RT-PCR is an alternative method to analyze the expression of marker genes. RT-PCR is sensitive, high-throughput, highly standardized, and a relatively simple-to-perform method, the results of which do not depend on the subjective interpretation of a trained expert. Takano et al22 were first to show that sufficient material for as much as 20 RT-PCR analyses can be obtained from leftover FNAB material in the needle after taking the sample for cytological analysis; hence, no additional invasion to the patient is required. Nevertheless, unlike by immunocytochemistry, the expression of galectin-3 in preoperative FNAB samples of thyroid lesion has been studied by RT-PCR in only a few studies on a small number of patients. Gasbarri et al16 have analyzed the expression of galectin-3 by immunochemistry and by RT-PCR in 157 surgical samples and 36 preoperative FNAB samples of thyroid lesions but they did not show the data for RT-PCR analysis in preoperative FNAB samples. Niedziela et al23 have analyzed galectin-3 expression by RT-PCR in 30 preoperative FNAB samples of thyroid lesions, and among 23 adequate samples they have found galectin-3 expression not only in 4 out of 5 malignant samples but also in 6 out of 18 benign samples. Pineda et al24 have analyzed FNAB samples from 20 patients and have found galectin-3 expression by RT-PCR in 100% of papillary carcinoma samples and in 83.3% of benign samples. In several other studies, the expression of galectin-3 was studied by RT-PCR and real-time RT-PCR only in postoperative histopathological samples showing high sensitivity but relatively low specificity in discriminating between benign and malignant thyroid lesions.29 All these results indicate that RT-PCR detection of galectin-3 expression in FNAB samples of thyroid lesions has low value as a potential marker for preoperative detection of malignant lesions because of low specificity. Nevertheless, the power of evidence of these results is low because most of the results were obtained from the analysis of postoperative histopathological samples, and to our knowledge only 2 studies reported results for analysis of preoperative FNAB samples on 43 patients altogether.24 Furthermore, no study reported the results for a clinically relevant group of patients with thyroid nodules that have indeterminate cytological diagnosis regarding malignancy. To our knowledge, this study on 428 patients is the largest so far in which RT-PCR detection of galectin-3 expression in thyroid lesions was evaluated. Furthermore, unlike others it is a prospective study in which galectin-3 expression was analyzed preoperatively on FNAB samples and compared with subsequent histopathology diagnosis or with definitive cytological diagnosis as a golden standard. Moreover, results for clinically relevant group of patients with cytologically indeterminate diagnosis of thyroid lesions are reported, which makes it more relevant for evaluating potential clinical application.We reported 57 out of 428 (13%) inadequate samples for RT-PCR analysis because of no expression of GAPDH and/or thyroglobulin, which is within the range of the percentages obtained in other studies.30 Relatively low specificity (64%) and positive predictive value (35%) of galectin-3 shown in this study because of relatively high rate of false-positive benign samples (36%) are in accordance with the results of 2 previous studies on preoperative FNAB samples and other studies on postoperative histopathological samples.24,29,31,32 It was suggested that galectin-3-false-positive follicular adenoma samples might represent potential early cancer in which malignant morphological features, such as capsular penetration and vascular invasion, are not yet apparent.13 Galectin-3-false-positive results might also be due to the expression of galectin-3 in other cells that have been observed together with thyroid epithelium cells in both benign and malignant thyroid lesions, such as fibroblasts, activated endothelial cells, macrophages.33 Niedziela et al23 have found galectin-3 expression by RT-PCR in Hashimoto's thyroiditis samples and attributed it to the presence of Hurthle cells, which are known to express galectin-3 in these samples. In a previous study, we have found a significantly higher proportion of galectin-3-RT-PCR-false-positive FNAB samples of nodular goiter and Hashimoto's thyroiditis when macrophages were present and also found a significantly higher proportion of galectin-3-RT-PCR-false-positive samples of Hashimoto's thyroiditis when Hurthle cells were present in the sample.20 This indicates the advantage of immunochemical analysis of galectin-3 expression, because by immunochemistry, unlike by RT-PCR, it can be distinguished whether galectin-3 is expressed by thyroid epithelial cells or by other cell types present in the sample. False-positive samples could be also due to the very high sensitivity of the RT-PCR method that can detect very low levels of expression of galectin-3 in benign and normal thyroid cells. Takano et al34 have analyzed the expression of galectin-3 by real-time quantitative RT-PCR and have found not only high galectin-3 expression in thyroid papillary carcinomas but also lower expression in normal and benign thyroid tissues and in thyroid fibroblasts. With this in mind, immunochemistry might be actually advantageous to RT-PCR in detecting the expression of galectin-3 in preoperative diagnostics of thyroid lesions because of its lower sensitivity. Although these explanations indicate the advantage of immunochemistry over RT-PCR, false-positive results have also been reported for immunochemical analysis of galectin-3 expression in thyroid lesions.19,32,35–37 Nevertheless, the overall rate of false-positive results in studies published thus far is lower for immunochemistry than for RT-PCR. Furthermore, Giannini et al38 directly compared the detection of galectin-3 expression by immunochemistry and by RT-PCR on same samples and have shown that RT-PCR is a little more sensitive (97.8% vs. 93.3%) but much less specific (45% vs. 90%). Oestreicher-Kedem et al36 have found a very high proportion of false-positive galectin-3 samples among Hurthle cell adenomas (83.3%), much higher than in follicular adenoma samples (10.5%). In our study also, the highest proportion of false-positive galectin-3 samples was among Hurthle cell adenomas (61.5%); but even if we excluded Hurthle cell adenomas, the proportion of false-positive samples would still remain relatively high (34.5%).Relatively high sensitivity (87%) and negative predictive value (96%) of galectin-3 in this study because of the relatively low rate of false-negative malignant samples (13%) are in accordance with the results of 2 previous studies on preoperative FNAB samples and other studies on postoperative histopathological samples in which galectin-3 expression was analyzed by RT-PCR.16,23,24,31,32,38 False-negative samples might be due to heterogeneity of galectin-3 expression in thyroid cancer samples. Kawachi et al39 found a stronger expression of galectin-3 in papillary carcinoma with metastases in comparison with papillary carcinoma without metastases.Specificity, sensitivity, and positive and negative predictive values were lower for 142 samples with indeterminate cytological diagnosis than for all 371 adequate samples in our study. As our study is the first, to our knowledge, prospective study in which galectin-3 expression was analyzed preoperatively by RT-PCR on FNAB samples with results reported for patients with indeterminate cytological diagnosis, we cannot compare our results with others. One other prospective study, in which galectin-3 expression was analyzed preoperatively on FNAB samples, but with immunocytochemistry, also reported lower specificity, sensitivity, and positive and negative predictive values for patients with indeterminate cytological diagnosis compared with all patients with thyroid nodules studied.12 Compared with the largest prospective study in which galectin-3 expression was analyzed by immunocytochemistry on preoperative FNAB samples in 465 patients with cytologically indeterminate thyroid lesions, our results show lower diagnostic value by all parameters including specificity, sensitivity, positive predictive value, and negative predictive value.14 This indicates that immunocytochemistry might be superior to RT-PCR as a method for analyzing galectin-3 expression to discriminate between malignant and benign thyroid lesions preoperatively.One disadvantage of our study is the small number of samples from patients with definitive diagnosis of follicular thyroid carcinoma. Higher number of patients with follicular carcinoma included in the study could enable a more reliable assessment of potential clinical value of galectin-3 based on our results, because several previous studies in which galectin-3 expression was analyzed by RT-PCR have shown that galectin-3 is significantly overexpressed in papillary carcinoma samples, but there is no a significant difference in expression between follicular carcinoma and follicular adenoma.34,40,41 The reason for the small number of patients with follicular carcinoma in our study is that our study was conducted prospectively on an unselected group of patients with thyroid nodules that were referred to our institution for FNAB diagnostic procedure and the overall incidence of follicular carcinoma in Croatia is low, probably because Croatia has complete iodine intake sufficiency.42 The exact data for follicular thyroid carcinoma incidence in Croatia are not available, but in 2007 in Croatia there were altogether 453 newly diagnosed patients with thyroid cancer, and among 85 of these patients who were treated in our institution only 1 (1%) had follicular thyroid cancer.43In our study, we also investigated whether diagnostic accuracy can be improved by analysis of CD44v6 as an additional marker to galectin-3. In our study, analysis of coexpression of galectin-3 and CD44v6 increased specificity, positive predictive value, and diagnostic accuracy and analysis of expression of galectin-3 or CD44v6 increased sensitivity and negative predictive value but not diagnostic accuracy. Similar results were obtained in a study where galectin-3 and CD44v6 were analyzed by immunocytochemistry.19 In another immunocytochemical study, additional analysis of CD44v6 did not improve diagnostic accuracy of galectin-3, but in that study galectin-3 alone was already 100% sensitive and 98% specific for malignant lesions.13 Other studies investigating different combinations and panels of markers including galectin-3, HMBE-1, cytokeratin 19, and other markers have shown that analysis of additional markers may improve diagnostic accuracy in identifying malignant thyroid lesions.44–48In conclusion, our results indicate that, mainly because of low specificity, the potential clinical value of RT-PCR detection of galectin-3 expression as a method for preoperative diagnosis of malignant thyroid lesions is limited and lower than comparable studies of immunocytochemical detection of galectin-3 expression.REFERENCES1. Welker MJ, Orlov D. Thyroid nodules Am Fam Physician.. 2003;67:559–566[Context Link]2. Belfiore A, La Rosa GL, La Porta GA, et al. Cancer risk in patients with cold thyroid nodules: relevance of iodine intake, sex, age, and multinodularity Am J Med.. 1992;93:363–369[Context Link]3. Gharib H, Goellner JR. Fine-needle aspiration biopsy of the thyroid: an appraisal Ann Intern Med.. 1993;118:282–289[Context Link]4. Baloch ZW, LiVolsi VA. Fine-needle aspiration of thyroid nodules: past, present, and future Endocr Pract.. 2004;10:234–241[Context Link]5. Carpi A, Nicolini A, Gross MD, et al. Controversies in diagnostic approaches to the indeterminate follicular thyroid nodule Biomed Pharmacother.. 2005;59:517–520[Context Link]6. Baloch ZW, Fleisher S, LiVolsi VA, et al. Diagnosis of “follicular neoplasm”: a gray zone in thyroid fine-needle aspiration cytology Diagn Cytopathol.. 2002;26:41–44[Context Link]7. Rodrigo JP, Rinaldo A, Devaney KO, et al. Molecular diagnostic methods in the diagnosis and follow-up of well-differentiated thyroid carcinoma Head Neck.. 2006;28:1032–1039[Context Link]8. Eszlinger M, Paschke R. Molecular fine-needle aspiration biopsy diagnosis of thyroid nodules by tumor specific mutations and gene expression patterns Mol Cell Endocrinol.. 2010;322:29–37[Context Link]9. Volante M, Bozzalla-Cassione F, Orlandi F, et al. Diagnostic role of galectin-3 in follicular thyroid tumors Virchows Arch.. 2004;444:309–312[Context Link]10. Sanabria A, Carvalho AL, Piana de Andrade V, et al. Is galectin-3 a good method for the detection of malignancy in patients with thyroid nodules and a cytologic diagnosis of “follicular neoplasm”? A critical appraisal of the evidence Head Neck.. 2007;29:1046–1054[Context Link]11. Dumic J, Dabelic S, Flögel M. Galectin-3: an open-ended story Biochim Biophys Acta.. 2006;1760:616–635[Context Link]12. Collet JF, Hurbain I, Prengel C, et al. Galectin-3 immunodetection in follicular thyroid neoplasms: a prospective study on fine-needle aspiration samples Br J Cancer.. 2005;93:1175–1181[Context Link]13. Bartolazzi A, Gasbarri A, Papotti M, et al. Application of an immunodiagnostic method for improving preoperative diagnosis of nodular thyroid lesions Lancet.. 2001;357:1644–1650[Context Link]14. Bartolazzi A, Orlandi F, Saggiorato E, et al. Galectin-3-expression analysis in the surgical selection of follicular thyroid nodules with indeterminate fine-needle aspiration cytology: a prospective multicentre study Lancet Oncol.. 2008;9:543–549[Context Link]15. Naor D, Nedvetzki S, Golan I, et al. CD44 in cancer Crit Rev Clin Lab Sci.. 2002;39:527–579[Context Link]16. Gasbarri A, Martegani MP, Del Prete F, et al. Galectin-3 and CD44v6 isoforms in the preoperative evaluation of thyroid nodules J Clin Oncol.. 1999;17:3494–3502[Context Link]17. Gu J, Daa T, Kashima K, et al. Expression of splice variants of CD44 in thyroid neoplasms derived from follicular cells Pathol Int.. 1998;48:184–190[Context Link]18. Ermak G, Gerasimov G, Troshina K, et al. Deregulated alternative splicing of CD44 messenger RNA transcripts in neoplastic and nonneoplastic lesions of the human thyroid Cancer Res.. 1995;55:4594–4598[Context Link]19. Maruta J, Hashimoto H, Yamashita H, et al. Immunostaining of galectin-3 and CD44v6 using fine-needle aspiration for distinguishing follicular carcinoma from adenoma Diagn Cytopathol.. 2004;31:392–396[Context Link]20. Matesa N, Samija I, Kusić Z. Galectin-3 and CD44v6 positivity by RT-PCR method in fine needle aspirates of benign thyroid lesions Cytopathology.. 2007;18:112–116[Context Link]21. Matesa N, Samija I, Kusić Z. Accuracy of fine needle aspiration biopsy with and without the use of tumor markers in cytologically indeterminate thyroid lesions Coll Antropol.. 2010;34:53–57[Context Link]22. Takano T, Sumizaki H, Amino N. Detection of CD44 variants in fine needle aspiration biopsies of thyroid tumor by RT-PCR J Exp Clin Cancer Res.. 1997;16:267–271[Context Link]23. Niedziela M, Maceluch J, Korman E. Galectin-3 is not an universal marker of malignancy in thyroid nodular disease in children and adolescents J Clin Endocrinol Metab.. 2002;87:4411–4415[Context Link]24. Pineda P, Rojas P, Liberman C, et al. Detection of malignancy markers in thyroid nodules by reverse transcriptase polymerase chain reaction (RT-PCR) Rev Med Chil.. 2003;131:965–972[Context Link]25. World Medical Association. World Medical Association Declaration of Helsinki-Ethical Principles for Medical Research Involving Human Subjects. 2008. Available at: http://www.wma.net/en/30publications/10policies/b3/index.html. [Accessed December 8, 2010].[Context Link]26. Ditkoff BA, Marvin MR, Yemul S, et al. Detection of circulating thyroid cells in peripheral blood Surgery.. 1996;120:959–964[Context Link]27. Battayani Z, Grob JJ, Xerri L, et al. Polymerase chain reaction detection of circulating melanocytes as a prognostic marker in patients with melanoma Arch Dermatol.. 1995;131:443–447[Context Link]28. Cibas ES, Ali SZ. The Bethesda system for reporting thyroid cytopathology Am J Clin Pathol.. 2009;132:658–665[Context Link]29. Pagedar NA, Chen DH, Wasman JK, et al. Molecular classification of thyroid nodules by cytology Laryngoscope.. 2008;118:692–696[Context Link]30. Samija I, Matesa N, Lukac J, et al. Thyroid fine-needle aspiration samples inadequate for reverse transcriptase-polymerase chain reaction analysis Cancer.. 2008;114:187–195[Context Link]31. Aratake Y, Umeki K, Kiyoyama K, et al. Diagnostic utility of galectin-3 and CD26/DPPIV as preoperative diagnostic markers for thyroid nodules Diagn Cytopathol.. 2002;26:366–372[Context Link]32. Martins L, Matsuo SE, Ebina KN, et al. Galectin-3 messenger ribonucleic acid and protein are expressed in benign thyroid tumors J Clin Endocrinol Metab.. 2002;87:4806–4810[Context Link]33. Bartolazzi A, Papotti M, Orlandi F. Methodological considerations regarding the use of galectin-3 expression analysis in preoperative evaluation of thyroid nodules J Clin Endocrinol Metab.. 2003;88:950 author reply 950–951.[Context Link]34. Takano T, Miyauchi A, Matsuzuka F, et al. Ubiquitous expression of galectin-3 mRNA in benign and malignant thyroid tumors Cancer Lett.. 2003;199:69–73[Context Link]35. Mehrotra P, Okpokam A, Bouhaidar R, et al. Galectin-3 does not reliably distinguish benign from malignant thyroid neoplasms Histopathology.. 2004;45:493–500[Context Link]36. Oestreicher-Kedem Y, Halpern M, Roizman P, et al. Diagnostic value of galectin-3 as a marker for malignancy in follicular patterned thyroid lesions Head Neck.. 2004;26:960–966[Context Link]37. Coli A, Bigotti G, Zucchetti F, et al. Galectin-3, a marker of well-differentiated thyroid carcinoma, is expressed in thyroid nodules with cytological atypia Histopathology.. 2002;40:80–87[Context Link]38. Giannini R, Faviana P, Cavinato T, et al. Galectin-3 and oncofetal-fibronectin expression in thyroid neoplasia as assessed by reverse transcription-polymerase chain reaction and immunochemistry in cytologic and pathologic specimens Thyroid.. 2003;13:765–770[Context Link]39. Kawachi K, Matsushita Y, Yonezawa S, et al. Galectin-3 expression in various thyroid neoplasms and its possible role in metastasis formation Hum Pathol.. 2000;31:428–433[Context Link]40. Bernet VJ, Anderson J, Vaishnav Y, et al. Determination of galectin-3 messenger ribonucleic acid overexpression in papillary thyroid cancer by quantitative reverse transcription-polymerase chain reaction J Clin Endocrinol Metab.. 2002;87:4792–4796[Context Link]41. Feilchenfeldt J, Tötsch M, Sheu S, et al. Expression of galectin-3 in normal and malignant thyroid tissue by quantitative PCR and immunohistochemistry Mod Pathol.. 2003;16:1117–1123[Context Link]42. Kusić Z, Novosel SA, Dabelić N, et al. Croatia has reached iodine sufficiency J Endocrinol Invest.. 2003;26:738–742[Context Link]43. Znaor A Cancer Incidence in Croatia 2003-2007. Bulletins No 27-31. 2009 Zagreb Croatian National Institute of Public Health[Context Link]44. Saggiorato E, De Pompa R, Volante M, et al. Characterization of thyroid ‘follicular neoplasms’ in fine-needle aspiration cytological specimens using a panel of immunohistochemical markers: a proposal for clinical application Endocr Relat Cancer.. 2005;12:305–317[Context Link]45. Prasad ML, Pellegata NS, Huang Y, et al. Galectin-3, fibronectin-1, CITED-1, HBME1 and cytokeratin-19 immunohistochemistry is useful for the differential diagnosis of thyroid tumors Mod Pathol.. 2005;18:48–57[Context Link]46. Rossi ED, Raffaelli M, Mule' A, et al. Simultaneous immunohistochemical expression of HBME-1 and galectin-3 differentiates papillary carcinomas from hyperfunctioning lesions of the thyroid Histopathology.. 2006;48:795–800[Context Link]47. Nakamura N, Erickson LA, Jin L, et al. Immunohistochemical separation of follicular variant of papillary thyroid carcinoma from follicular adenoma Endocr Pathol.. 2006;17:213–223[Context Link]48. Hooft L, van der Veldt AAM, Hoekstra OS, et al. Hexokinase III, cyclin A and galectin-3 are overexpressed in malignant follicular thyroid nodules Clin Endocrinol (Oxf).. 2008;68:252–257[Context Link]thyroid neoplasms; thyroid nodule; biological tumor markers; fine-needle biopsy; reverse transcriptase polymerase chain reaction; galectin-3; CD44v6 antigenDistribution of 371 Patients With Samples Adequate for RT-PCR Included in the Study Regarding Postoperative Histopathological and Preoperative Cytological DiagnosisGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid lesions and Preoperative Cytological DiagnosisGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid Lesions and Final DiagnosisValue of Galectin-3 and CD44v6 Analysis by RT-PCR in Preoperative FNAB Samples in Discrimination Between Benign and Malignant Lesions of the Thyroid for all 371 Samples Adequate for RT-PCRGalectin-3 and CD44v6 Expression Detected by RT-PCR in Preoperative FNAB Samples of Thyroid Lesions With Indeterminate Cytological Diagnosis Regarding Malignancy*Value of Galectin-3 and CD44v6 Analysis by RT-PCR in Preoperative FNAB Samples in Discrimination Between Benign and Malignant Lesions of the Thyroid for 142 Samples Adequate for RT-PCR With Indeterminate Cytological Diagnosis Regarding MalignancyGalectin-3 and CD44v6 as Markers for Preoperative Diagnosis of Thyroid Cancer by RT-PCRŠamija Ivan PhD; Mateša, Neven MD, PhD; Lukac, Josip PhD; Kusic, Zvonko MD, PhDOriginal ArticlesOriginal Articles420p 233-241

You can read the full text of this article if you:

Access through Ovid