Secondary Logo

Journal Logo

Differentiation of benign from malignant follicular-cell-derived thyroid lesions: an immunohistochemical study

El-Shorbagy, Safinaz H.; Abd El-Azeem, Marwa A.; Abd El-Azeem, Mona A.

doi: 10.1097/01.XEJ.0000472879.61441.9a
ORIGINAL ARTICLES
Free

Background Histopathological criteria in some cases do not allow the differentiation between benign and malignant follicular-cell-derived thyroid tumors/lesions, making the distinction between these two groups quite subtle and challenging. Consequently, immunohistochemistry may provide additional support in the evaluation and diagnosis of these thyroid tumors/lesions.

Objective The aim of this study was to determine the diagnostic value of insulin-like growth factor mRNA binding protein-3 (IMP-3), galectin-3, cytokeratin-19 (CK-19), and fragile histidine triad (FHIT), used alone and in different combinations, in differentiating benign from malignant follicular-cell-derived thyroid lesions.

Materials and methods The study was carried out on 60 selected archival cases (20 benign and 40 malignant thyroid lesions). Paraffin sections of 4–5 μm thickness were stained with H&E to confirm their histological diagnosis and with IMP-3, galectin-3, CK-19, and FHIT for immunohistochemical evaluation. Sensitivity, specificity, and positive likelihood ratio were calculated for each of the markers and for different combinations of these markers in the benign versus malignant.

Results Comparison between immunohistochemical expressions of IMP-3, galectin-3, CK-19, and FHIT in different thyroid lesions revealed a significant difference between benign thyroid lesions and malignant lesions with the exception of FHIT, which did not reach statistical significance for distinction between different thyroid lesions. Therefore, this marker was excluded from the combination of immunohistochemical markers. Immunohistochemical expression of combined markers (galectin-3 and CK-19; galectin-3 and IMP-3; and CK-19 and IMP-3) showed highest statistical difference between malignant and benign thyroid lesions, but did not markedly increase the sensitivity or specificity over using a single marker (galectin-3 for differentiating benign from malignant lesions).

Conclusion Combined use of markers such as galectin-3 and CK-19 and IMP-3 did not markedly increase the sensitivity or specificity over using a single marker. Each studied marker has sensitivity and specificity for certain thyroid lesion. Immunohistochemical stains of galectin-3, CK-19, and IMP-3, especially when used in combination, can be a useful adjunct in the histopathological diagnoses of thyroid lesions.

Department of Pathology, Faculty of Medicine, Tanta University, Tanta, Egypt

Correspondence to Safinaz H. El-Shorbagy, MD, Department of Pathology, Faculty of Medicine, 31111 Tanta University, Tanta, Egypt Tel: +21227380448/+21148145527; e-mail: drsafy_shorbagy@yahoo.com

Received April 12, 2015

Accepted May 1, 2015

Back to Top | Article Outline

Introduction

Thyroid nodules, either benign or malignant, are the most common endocrine lesions encountered, and thyroid carcinomas arising from the follicular epithelium are the most common among endocrine system malignancies (Suster, 2006).

Thyroid follicular lesions represent a diagnostic challenge, as thyroid fine-needle aspiration biopsy cannot distinguish benign from malignant tumors in up to 30% of cases (Gimm et al., 2011). Most thyroid tumors can be readily diagnosed using histopathological criteria, which allow the pathologist to differentiate benign from malignant lesions derived from follicular epithelial cells (Suster, 2006). However, problems may arise in the interpretation of thyroid tumors with a follicular growth pattern, as distinctions between hyperplastic nodules and follicular adenoma (FA), between FA and minimally invasive follicular carcinoma (FC), and between FA/FC and the follicular variant of papillary carcinoma are challenging (Bartolazzi et al., 2001). Moreover, the histopathology of papillary thyroid hyperplasia and papillary thyroid carcinoma is similar enough to cause a diagnostic dilemma in a few cases. Both lesions may have papillary fronds with fibrovascular cores and nuclear crowding (Casey et al., 2003).

In such cases, the pathologist is confronted with thyroid lesions in which the accurate diagnosis, which is crucial for the appropriate treatment, can be quite difficult because agreement about the minimal criteria required for changes in nuclear features of papillary thyroid carcinoma, as well as the exact criteria for capsular invasion, are not universally accepted (Bartolazzi et al., 2001). The decision of the pathologist will further affect the treatment consequences. Thus, a more reliable method of distinguishing among these various follicular lesions, such as immunohistochemistry, may therefore be diagnostically useful in differential diagnosis, with subsequent decrease in the number of unnecessary surgery among patients with thyroid nodules.

The insulin-like growth factor II mRNA binding protein family comprises three proteins (IMP-1, IMP-2, and IMP-3) that regulate mRNA transport, translation, and turnover, and they have a role in cell proliferation, adhesion, invasion, and migration (Feng et al., 2011). They have been considered to play a dual role in both embryogenesis and tumor proliferation. Their expression is almost exclusively restricted to the early stages of embryogenesis, and, in particular, IMP-3 is epigenetically silenced soon after birth, with little or no detectable protein in normal adult tissues. It has been postulated that IMP-3 is expressed at low or undetectable levels in mature tissues, except for the placenta (Jin et al., 2011). Aberrant expression of IMP-3 has been observed in a variety of carcinomas and dysplasia including the pancreas (Yantiss et al., 2008), cervix (Danialan et al., 2013), colon (Li et al., 2009), liver (Wachter et al., 2012), head and neck (Feng et al., 2011), and the breast (Walter et al., 2009). High levels of IMP-3 in many different types of cancers suggest that it may play a role in the proliferation of malignant cells (Wachter et al., 2012). It therefore may be a useful diagnostic marker, especially as its expression in normal mature tissue is limited.

Galectin-3 is a member of a growing family of galactoside-binding animal lectins involved in the regulation of cell–cell and cell–matrix interaction, cell growth, neoplastic transformation, and apoptosis (Chiu et al., 2010). It has been suggested as a marker of thyroid malignancy with relatively high sensitivity and specificity, especially in papillary carcinoma (Aron et al., 2006; Slosar et al., 2009).

The differential expression of cytokeratins has been evaluated in various thyroid lesions. Among various cytokeratins, cytokeratin-19 (CK-19), a low molecular weight cytokeratin, has been reported to be useful in the diagnosis of papillary carcinoma (Scognamiglio et al., 2006).

The fragile histidine triad (FHIT) gene has been identified as one of the most important tumor suppressor genes. The FHIT gene is a member of the histidine triad (HIT) proteins, which represent a small family of nucleotide-binding and hydrolyzing proteins. HIT proteins received the attention of cancer biologists because of their downregulated expression in multiple human malignancies (Hassan et al., 2010). Its deletion or loss of transcription has been reported in several epithelial cancers such as lung, head and neck, breast, digestive tract, uterine cervix, and pancreatic cancers (Pavelić et al., 2006). Moreover, aberrant FHIT transcription has been reported in acute myeloid leukemia, acute lymphoblastic leukemia, and chronic myeloid leukemia (Hassan et al., 2010).

The aim of this study was to evaluate the diagnostic accuracy of IMP-3, galectin-3, CK-19, and FHIT markers in the differentiation between benign (papillary hyperplasia and adenoma) and malignant thyroid lesions.

Back to Top | Article Outline

Materials and methods

Tissue specimens

Pathological reports and paraffin blocks of 60 cases of different thyroid lesions were collected retrospectively from the Archive of Pathology Department, Tanta University, and private laboratories during the period of 2010–2013. Clinicopathological features of the thyroid lesions were obtained from the pathology report. The studied cases included 12 cases of papillary hyperplasia, eight cases of FA and 40 cases of malignant tumors [19 classic papillary thyroid carcinoma (CPC), seven FC, nine follicular variant of papillary thyroid carcinoma (FVPC), and last five cases of follicular tumor of undetermined malignant potential (FTUMP)]. The evaluation and classification of the thyroid tumors was based on the WHO thyroid tumor classification (DeLellis, 2006).

Back to Top | Article Outline

Immunohistochemical staining

Immunohistochemistry was performed on formalin-fixed paraffin-embedded (4-mm thick) sections. Tissue sections were deparaffinized and rehydrated in graded alcohols. Endogenous peroxidase activity was blocked with 0.3% hydrogen peroxide. For antigen retrieval, citrate buffer was used after the sections were treated in a microwave. The tissue sections were incubated overnight at room temperature with the following primary antibodies: anti-IMP-3 (mouse monoclonal antibody, clone 69.1, dilution 1 : 300; Dako, Glostrup, Denmark), anti-galectin-3 (monoclonal antibody NCL-Gal3, dilution 1 : 200; Novocastra, Newcastle, UK), anti-CK-19 (polyclonal antibody, dilution 1 : 100; Dako Cytomation, Carpenteria, California, USA), and anti-FHIT (polyclonal antibody dilution 1 : 50, clone ab53074; Abcam, Cambridge, UK). The staining was completed using the streptavidin–biotin complex detection method. The slides of positive and negative controls were included in each run. Positive controls were human placenta for IMP-3, histiocytes for galectin-3, skin for CK-19, and breast for FHIT protein. Negative controls were prepared by excluding the primary antibody and replacing it with PBS.

Back to Top | Article Outline

Interpretation of immunohistochemical staining of the studied markers

IMP-3 showed cytoplasmic or membranous immunoreactivity; galectin-3 showed cytoplasmic and/or nuclear expression; CK-19 staining was cytoplasmic; and, for FHIT protein, positive staining was indicated by the presence of diffuse brownish granular cytoplasmic staining.

Back to Top | Article Outline

Assessment of immunohistochemical staining

Each slide was assessed semiquantitatively for immunohistochemical scoring based on the extent of the staining measured as the percentage of stained cells. For all antibodies, immunoreactivity was considered positive if more than 10% of follicular epithelial cells were stained (Nasr et al., 2006; Boilă et al., 2012). The immunoreactivity was scored as negative (<10%), focally weak positive (+: <25%), moderately positive (++: 25–50%), or strong diffusely positive (+++: >50%) (Alshenawy, 2014).

Back to Top | Article Outline

Statistical analysis

Statistical analysis was performed with statistical package for the social sciences software (SPSS, version 20; SPSS Inc., Chicago, Illinois, USA). Sensitivity, specificity, and positive likelihood ratio (diagnostic accuracy) were calculated for each of the markers and for different combinations of these markers in the benign versus malignant, benign non-neoplastic versus malignant, and benign neoplastic versus malignant groups. Receiver operating characteristic (ROC) curves were then drawn for each marker and for paired combination of markers to find the most suitable marker or markers for distinguishing between benign and malignant thyroid lesions. The area under the curve (AUC) in ROC analysis was used to compare the ability of a marker to discriminate between benign and malignant lesions, allowing for visual identification of the markers with the greatest discrimination accuracy. A P value less than 0.05 was considered statistically significant.

Back to Top | Article Outline

Results

Immunohistochemical expression of IMP-3, galectin-3, CK-19, and FHIT in various thyroid lesions is summarized in Table 1. There was a strong and diffuse positive reactivity for all markers, except FHIT, in a large percentage of malignant thyroid lesions (CPC, FVPC, FC, and FTUMP), whereas benign lesions (papillary hyperplasia and FA) showed focal and weak expression for IMP-3, galectin-3, and CK-19. Although some lesions of papillary hyperplasia and some benign neoplastic lesions (FA) showed positive immunostaining, the expression was less intense and less frequent than that in malignant lesions. In contrast, FHIT immunopositivity was detected more frequently in benign thyroid lesions than in differentiated thyroid carcinoma. There was a significant statistical difference in the expression of the four markers between benign and malignant thyroid lesions (P<0.05).

Table 1

Table 1

Back to Top | Article Outline

Non-neoplastic lesions (papillary hyperplasia)

Out of 12 cases, only three (25%) were positive for IMP-3. There were two (16.7%) cases positive for galectin-3, three (25%) positive for CK-19, and nine (75%) positive for FHIT. The positive immunostaining was weak (+) in galectin-3, weak-to-moderate (+ to ++) in CK-19 and IMP-3, and mostly strong (+++) in FHIT.

Back to Top | Article Outline

Adenomas

Out of eight cases of FA, two (25%) were positive for IMP-3, two (25%) were positive for galectin-3, one (12.5%) was positive for CK-19, and six (75%) were positive for FHIT. None of the three markers (IMP-3, galectin-3, or CK-19) were strongly diffusely positive in FA, whereas FHIT showed strong immunoreactivity in half of the FA cases.

Back to Top | Article Outline

Carcinomas versus benign (immunohistochemical expression)

Among 40 cases of malignant thyroid lesions, 33 (82.5%) cases were positive for IMP-3, 38 (95%) were positive for galectin-3, 36 (90%) were positive for CK-19, and three (37.5%) were positive for FHIT.

IMP-3 was generally absent in papillary hyperplasia (Fig. 1a) and showed strong and diffuse positive reactivity in CPC, FVPC, and FC (Fig. 1b–d). Moreover, galectin-3 was absent in papillary hyperplasia (Fig. 2a) and showed positive expression in FC, CPC, and FVPC (Fig. 2b–e). CK-19 was negative in papillary hyperplasia and tumor cells of FA (Fig. 3a and b, respectively) and positive in CPCs (Fig. 3c), FVPCs (Fig. 3d), and FC (Fig. 3e). FHIT exhibited immunopositivity in papillary hyperplasia (Fig. 4a) and FA (Fig. 4b), whereas in malignant cases FHIT showed negative expression in 25 cases (Fig. 4c) and was positive in the remaining 15 cases (Fig. 4d).

Fig. 1

Fig. 1

Fig. 2

Fig. 2

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Thus, CPC showed a positive immunoreactivity for galectin-3 (100%), CK-19 (94.7%), and IMP-3 (89.5%), whereas the majority of the cases (73.7) were negative for FHIT (Fig. 4c). Similarly, FVPC was also highly reactive for galectin-3 (100%), CK-19 (100%), and IMP-3 (88.9%), whereas FHIT was positive only in three (37.5%) cases (Fig. 4d). FC showed a high rate of immunoreactivity for galectin-3 and CK-19 (85.7% for both), followed by IMP-3 (71.4), whereas FHIT was only weakly positive in two (28.6%) cases. For FTUMP cases, galectin-3 was positive in the majority of cases (80%) (Fig. 2e), and for CK-19 and IMP-3 the immunoreactivity was detected in 60% of cases for each of the two markers, whereas only two (40%) cases were weakly to moderately positive for FHIT. The intensity and score of different benign and malignant studied thyroid lesions are summarized in Table 1.

Back to Top | Article Outline

Insulin-like growth factor mRNA binding protein-3

The highest significant statistical difference in IMP-3 immunoexpression in differentiating benign from malignant lesions was found in papillary hyperplasia versus CPC (P=0.003), with the highest sensitivity (89.5%), and diagnostic accuracy (3.85). IMP-3 also showed relatively high sensitivity and specificity (88.9 and 75%, respectively) in distinguishing FVPC from both benign lesions (papillary hyperplasia and FA), with significant statistical difference (P=0.014 and 0.027, respectively) and 3.56 diagnostic accuracy (Table 2).

Table 2

Table 2

Back to Top | Article Outline

Galectin-3

Galectin-3 showed high sensitivity (100%) and specificity (83.3%) in differentiating papillary hyperplasia from papillary thyroid carcinomas (CPC and FVPC), with high statistically significant difference (P=0.0001 and 0.001, respectively) and high diagnostic accuracy (5.99). Moreover, galectin-3 showed a statistically significant difference (P=0.009) in its expression in FA versus FVPC, with 100% sensitivity, but with less specificity (75%) than that in distinction between papillary hyperplasia and FVPC (Table 2).

Back to Top | Article Outline

Cytokeratin-19

CK-19 showed high diagnostic accuracy in differentiating FA from FC (6.12), with relatively high sensitivity and specificity (87.5 and 85.7%, respectively). Moreover, CK-19 was found to be highly sensitive (100%) in differentiating FVPC from benign thyroid lesions (papillary hyperplasia and FA) with significant statistical difference (P=0.004 and 0.002, respectively) and significant difference (P=0.001) was detected in its expression in CPC versus papillary hyperplasia as well, with high sensitivity but with less specificity (94.7 and 75%, respectively) (Table 2).

Back to Top | Article Outline

Fragile histidine triad

FHIT expression in benign versus malignant lesions showed moderate sensitivity (75%) and also moderate specificity ranging from 60 to 73.7%, with somewhat low diagnostic accuracy in differentiating benign lesions (papillary hyperplasia or FA) from malignant lesions (FC, CPC, FVPC, or FTUMP) (Table 2).

Back to Top | Article Outline

Receiver operating characteristic curve and area under the curve analysis for each biomarker expression in benign versus malignant thyroid lesions

The ability of each marker to distinguish benign from malignant thyroid lesions was thereafter determined by means of ROC analysis. The values of the area under the ROC curve (AUC) are shown in Table 3 and the actual curves are shown in Figs 5 and 6. Galectin-3 showed the highest AUC (0.917) in differentiating papillary hyperplasia from papillary thyroid carcinoma (CPC and FVPC). However, AUC of CK-19 in differentiating FA from both FC and FVPC was the highest among all other markers (0.866 and 0.938, respectively). Furthermore, CK-19 showed the highest AUC (0.738) in differentiating FA from FTUMP.

Table 3

Table 3

Fig. 5

Fig. 5

Fig. 6

Fig. 6

Back to Top | Article Outline

Differences in the expression of combined markers between benign and malignant thyroid lesions

To increase the ability of immunohistochemical markers to differentiate between benign and malignant thyroid lesions, a combined expression of two markers was studied. FHIT did not reach statistical significance for distinction between different thyroid lesions in ROC curve analysis. Therefore, this marker was excluded from the combination of immunohistochemical markers that were appropriate to distinguish certain benign from malignant thyroid counterparts. Table 4 shows the sensitivity, specificity, AUC, and diagnostic accuracy of the combined expression of two markers, in which combined expression of galectin-3 and CK-19 increased the statistical significance in differentiating FA from FC, papillary hyperplasia from FVPC, and FA from FVPC on the basis of ROC curve analysis, but did not markedly increase the sensitivity or specificity over using a single marker (galectin-3 for differentiating papillary hyperplasia from FVPC and CK-19 for differentiating FA from both FC and FVPC). Similarly, a combination of galectin-3 and IMP-3 markedly increased the significance in differentiating papillary hyperplasia from malignant thyroid tumors (CPC and FVPC) but without increase in sensitivity or specificity over the use of galectin-3 only. Moreover, a combination of CK-19 and IMP-3 increased the statistical significance in differentiating benign from malignant lesions on the basis of the ROC curve analysis. A combination of galectin-3 and CK-19 showed significant increase in AUC (0.756) in differentiating FA from FTUMP rather than using galectin-3 or CK-19 alone.

Table 4

Table 4

Back to Top | Article Outline

Discussion

Thyroid nodules are common clinical findings and thyroid carcinoma is the most common endocrine malignancy (Prasad et al., 2005). Histologic examination of H&E-stained sections remains the standard method in the diagnosis of thyroid lesions. However, it is known that sometimes the interpretation of follicular thyroid lesions can be quite difficult, especially in distinguishing benign from malignant lesions (Park et al., 2007).

This dilemma is encountered with encapsulated tumors showing follicular growth pattern, as the presence or absence of capsular and/or vascular invasion distinguishes benign from malignant follicular tumors may be challenging. Identification of these findings may not be possible due to incomplete capsular penetration, or technical errors. Another difficulty is encountered when some of the diagnostic nuclear features of papillary carcinoma are seen (Saleh et al., 2010).

These difficulties in differentiating follicular thyroid lesions have increased the need for immunohistochemical markers that could reliably distinguish benign from malignant thyroid nodules in the challenging cases (de Matos et al., 2012). Several diagnostic thyroid markers have been studied to evaluate their role, but none of these markers are useful in every situation (Scognamiglio et al., 2006; Saleh et al., 2010).

IMP-3, with its limited expression in normal mature tissues, can be used as a useful marker in differentiating benign from malignant thyroid lesions (Slosar et al., 2009).

In our study, moderate-to-strong diffuse positive IMP-3 expression was detected in most malignant thyroid lesions (82.5%), whereas benign lesions (papillary hyperplasia and FA) were mostly negative, or only showed focal and weak expression. A significantly high IMP-3 expression was found in papillary hyperplasia versus CPC (P=0.003), with the highest sensitivity (89.5%) and diagnostic accuracy (3.58) in its expression in different benign and malignant lesions. IMP-3 also showed relatively high sensitivity and specificity (88.9 and 75%, respectively) in distinguishing FVPC from both papillary hyperplasia and FA, with significant statistical difference (P=0.014 and 0.027, respectively) and a diagnostic accuracy of 3.56.

The previous observations are in agreement with those reported by Slosar et al. (2009). They have found that IMP-3 had 100% specificity and 69% sensitivity for FC as compared with FA and 100% specificity for FVPC, again compared with FA. They concluded that IMP-3 can be diagnostically useful in differentiating malignant and benign follicular pattern thyroid lesions. Similarly, the results of Jin et al. (2011), who evaluated IMP-3 expression in thyroid neoplasms using immunohistochemistry and qRT-PCR, showed that thyroid carcinomas, including CPC, FVPC, and FC, have significantly higher IMP-3 expression levels (48.3-, 35.3-, and 43.8-fold, respectively) compared with benign thyroid lesions (2.8-fold). On performing IMP-3 qRT-PCR analysis, it had a 91.4% clinical specificity and 86.7% clinical sensitivity for the diagnosis of well-differentiated thyroid carcinomas. Their results indicated that the detection of IMP-3 mRNA expression levels with qRT-PCR is more sensitive compared with immunohistochemistry and may be a useful molecular marker to assist in the diagnosis of well-differentiated thyroid carcinomas.

In contrast, Kulaçoğlu and Erkilinç (2015) observed IMP-3 expression in benign lesions both in nodular hyperplasia (50%) and FA (92.9%), with a statistically significant difference between the two groups in terms of positive staining (P=0.011). Moreover, 82.1% of their malignant tumors showed IMP-3 positivity. They have not found any significant difference between benign and malignant tumors.

Several investigations showed that expression of galectin-3 may be very useful in the differentiation of benign from malignant thyroid tumors, especially papillary thyroid carcinoma, with high sensitivity and specificity (Aron et al., 2006; Saleh et al., 2010). Moreover, it can aid in identifying FVPC, and distinguishing minimally invasive FC from FA (Bartolazzi et al., 2008).

Our results revealed a higher frequency of positive galectin-3 expression in malignant thyroid lesions compared with benign lesions. The expression was focal and weak in benign lesions (papillary hyperplasia and FA), whereas malignant lesions showed strong and diffuse expression. Galectin-3 showed high sensitivity (100%) and specificity (83.3%) in differentiating papillary hyperplasia from CPC and FVPC, with a highly statistically significant difference (P=0.0001 and 0.001, respectively) and high diagnostic accuracy (5.99). Moreover, its expression was significantly different (P=0.009) in FA versus FVPC, with 100% sensitivity but with less specificity (75%) than that in distinction between papillary hyperplasia and FVPC.

Similar findings were obtained by Saleh et al. (2010). They reported that galectin-3 showed 85.2% sensitivity for distinction between carcinomas and benign nodules (positive in 27.5% of benign vs. 85.1% of malignant nodules). However, the specificity was lower (72.4%). In addition, they found that galectin-3 was somewhat more strongly and diffusely positive in CPC than in FC and FVPC, and expression in FA was more focal and less intense than that in malignant tumors. Moreover, they detected galectin-3 expression in eight out of 52 benign non-neoplastic lesions, but this was quite focal and weak.

In the same way, Alshenawy (2014) observed that all cases of papillary thyroid carcinoma were positive for galectin-3 and this positivity is higher than that in other malignant lesions. Moreover, the high sensitivity and specificity of galectin-3 in differentiating malignant from benign, FVPC from FA, FC from FA, and CPC from other benign non-neoplastic lesions were observed. Therefore, galectin-3 has been consistently a very sensitive marker for papillary thyroid carcinoma and can be used to differentiate FVPC from FC.

Moreover, other authors have noticed similar results. Saggiorato et al. (2005) observed that only four of 52 FA cases were expressing galectin-3, whereas all their thyroid cancers were positive. Some authors consider true galectin-3-positive FA as an indication of potentially early or incipient carcinoma, in which the capsular and/or vascular invasion cannot be histologically observed as yet (Aron et al., 2006). Bartolazzi et al. (2001) have found that the sensitivity and the specificity of galectin-3 in thyroid carcinomas were 99 and 98%, respectively, whereas Husain et al. (2009) have reported lower values of 92.6 and 77.3%, respectively. Therefore, galectin-3 can be used to differentiate CPC and FVPC from FA and other benign non-neoplastic lesions.

In the current study, positive CK-19 expression was detected in 25% of benign non-neoplastic thyroid lesions, 12.5% of FA and 90% of malignant thyroid lesions. Its expression was mainly focal and weak in benign lesions but was strong and diffuse in malignant lesions. Another important finding is the high diagnostic accuracy of CK-19 in differentiating FA from FC (6.12), with relatively high sensitivity and specificity (87.5 and 85.7%, respectively). Moreover, CK-19 was found to be highly sensitive (100%) in differentiating FVPC from benign thyroid lesions (papillary hyperplasia and FA) with significant statistical difference (P=0.004 and 0.002, respectively). Moreover, a significant difference (P=0.001) was detected in its expression in CPC versus papillary hyperplasia with relatively high sensitivity and specificity (94.7 and 75%, respectively).

Several cytokeratins have been studied for the differential diagnosis of thyroid nodules, of which CK-19 has been found to be the most useful (Saleh et al., 2010). Several studies have showed that CK-19 shows strong and diffuse positivity in malignant thyroid tumors, including CPC, FVPC, and FC (Barroeta et al., 2006; Scognamiglio et al., 2006). However, other studies showed variable results and yet others demonstrated that the CK-19 expression is focal and weak in FC, FA, and benign hyperplastic nodules (Prasad et al., 2005; Park et al., 2007; Debdas et al., 2012).

Sahoo et al. (2001) have reported that 25% of their FA had extensive immunoreactivity for CK-19, whereas in the study by Debdas et al. (2012) 75% of their FA was focally positive. Moreover, in the study by Nasr et al. (2006), they demonstrated weak CK-19 expression in five of six FAs. In all these cases, CK-19 staining was patchy and weak. The previous observations suggest that CK-19 expression patterns are not reliable for the distinction between papillary carcinomas and follicular neoplasms. In the study by Saleh et al. (2010), 85.1% of all studied malignant tumors were positive for CK-19 (diffusely and strongly), and 50% of the adenomas were also positive (but more focal and less intense). In general, although most authors agreed that CK-19 reactivity is more frequent, diffuse and strong in papillary carcinoma, its reactivity in follicular neoplasms may limit its utility as a diagnostic marker (Park et al., 2007; Saleh et al., 2010; Debdas et al., 2012).

Alshenawy (2014) did not find any strong positive CK-19 expression in FA, whereas it was strongly expressed in 100% of papillary thyroid carcinoma and 53% of FC. Moreover, the sensitivity and the specificity in distinguishing malignant from benign were 65 and 78%, respectively, but in case of distinguishing CPC from other benign non-neoplastic lesions, the sensitivity and the specificity were 100 and 77%, respectively. Other studies showed a high sensitivity and specificity of CK-19 in diagnosing papillary thyroid carcinoma (Prasad et al., 2005; Park et al., 2007). They confirmed that CK-19 is a useful marker for differentiating papillary thyroid carcinoma from papillary hyperplasia. However, they also identified the expression of CK-19 in follicular neoplasms. According to these studies, CK-19 alone was not useful in the diagnosis of follicular thyroid lesions. The chief utility of CK-19 lies in its high specificity for papillary thyroid carcinoma. Negative staining for CK-19, therefore, is a strong evidence against papillary thyroid carcinoma.

For tumors known as ‘follicular tumor of undetermined malignant potential’ (FTUMP) that lie in the gray zone between FA and FC showing some suspicious but not diagnostic features of capsular and/or vascular invasion, many antibodies are used for their differentiation. Several studies showed that the most preferred method is using a panel of two or three markers for their recognition. The most widely used markers are CK-19, galectin-3, and hector battifora mesothelial cell-1 antigen (HBME-1). However, none of these markers are highly specific and sensitive alone (Scognamiglio et al., 2006; de Matos et al., 2012).

In our study, a combination of galectin-3 and CK-19 showed significant increase in AUC (0.756) in differentiating FA from FTUMP rather than using galectin-3 or CK-19 alone. We suggest to consider tumors with moderate-to-strong positivity of both galectin-3 and CK-19 as having true malignant potential.

As regards FHIT expression in benign versus malignant lesions, our results revealed that 75% of benign thyroid lesions (both papillary hyperplasia and FA) were positive, whereas 30% of carcinomas showed FHIT expression. Moreover, FHIT expression showed moderate sensitivity (75%) and specificity ranging from 60 to 73.7% with somewhat low diagnostic accuracy (2.5).

Our findings are in agreement with the previous studies that showed low or absent FHIT expression in thyroid carcinomas, but its expression in benign lesions was uniformly high (Pavelić et al., 2006; Liang et al., 2009).

We strengthened our study by performing ROC analysis for determining the values for each marker alone and a combination of two markers for differentiating thyroid nodules (benign vs. malignant). On the basis of our ROC curve analysis, we found that markers with an AUC greater than 0.75 had good diagnostic accuracy for distinguishing between benign thyroid lesions (FA and papillary hyperplasia) and differentiated thyroid carcinomas (FC, FVPC, and CPC). Galectin-3 gave the largest AUC of all tested markers in differentiating papillary hyperplasia from CPC and FVPC (0.917 for both) and CK-19 in differentiating FA from FVPC and FC (0.938 and 0.866, respectively).

To increase the ability of immunohistochemical markers to differentiate between benign and malignant thyroid lesions, we studied the combined expression of two markers. As FHIT did not reach statistical significance for distinction between different thyroid lesions in ROC curve analysis, we excluded this marker from the combination of immunohistochemical markers that were appropriate to distinguish certain benign from malignant thyroid counterparts. The combined expression of galectin-3 and CK-19 increased the statistical significance in differentiating FA from FC, papillary hyperplasia from FVPC and FA from FVPC on the basis of ROC curve analysis, but did not markedly increase the sensitivity or specificity over using a single marker (galectin-3 for differentiating papillary hyperplasia from FVPC and CK-19 for differentiating FA from both FC and FVPC). Similarly, a combination of galectin-3 and IMP-3 markedly increased the significance in differentiating papillary hyperplasia from malignant thyroid tumors (CPC and FVPC), but without increase in sensitivity or specificity over the use of galectin-3only. Moreover, a combination of CK-19 and IMP-3 increased the statistical significance in differentiating benign from malignant lesions on the basis of the ROC curve analysis.

Saleh et al. (2010) had also found that marker combinations of galectin-3+HBME-1, galectin-3+CK-19, or HBME-1+CK-19 did not improve the sensitivity or specificity for the distinction between benign and malignant thyroid lesions. Furthermore, using a combination panel of three markers (galectin-3+HBME-1+CK-19) also did not increase the sensitivity or specificity for the distinction between benign and malignant thyroid lesions.

In contrast, other studies have found that panels of various combinations of immunohistochemistry markers can add to the value of diagnosing malignant thyroid tumors and discriminating them from benign tumors and non-neoplastic lesions (Scognamiglio et al., 2006; Park et al., 2007). de Matos et al. (2012) found that a panel of galectin-3, HBME-1, and CK-19 is useful in differentiating the follicular-patterned lesions and specifically in distinguishing FVPC from FC or FA. Moreover, Cheung et al. (2001) recommended the use of a panel of CK-19, HBME-1, and Ret as a useful means for diagnosing papillary carcinoma, whereas Rossi et al. (2006) concluded that a combination of only HBME-1 and galectin-3 can correctly diagnose classic and variants of papillary carcinomas.

Similarly, Alshenawy (2014) found that a combination of markers may be more accurate compared with the use a single marker. CD56 with HBME-1 was the best combination to differentiate benign from malignant. CD56+HBME-1+galectin-3 was the best combination in differentiating FVPC from FA and CPC from other benign non-neoplastic lesions. CD56+galectin-3 was the best combination to distinguish FC from FA and FTUMP. The sensitivity and the specificity were increased with the use of combinations of this panel. She concluded that this panel was able to discriminate benign from malignant lesions, CPC, FVPCs, FA, FC, and FTUMP among other similar follicular-cell-derived thyroid lesions.

Back to Top | Article Outline

Conclusion

In conclusion, immunoexpression of IMP-3, galectin-3, and CK-19 is useful in distinguishing benign from malignant thyroid lesions (especially papillary hyperplasia and FA vs. CPC, FVPC, and FC), whereas FHIT expression showed low significance. Moreover, the combination of two markers did not markedly increase the sensitivity or specificity over using a single marker. However, using a panel of two markers is advised to overcome technical problems or processing issues and increase the reliability of detecting malignant tumors. Furthermore, a combination of galectin-3 and CK-19 can be used for differentiating FA from FTUMP rather than using galectin-3 or CK-19 alone.

Back to Top | Article Outline

Acknowledgements

Conflicts of interest

There are no conflicts of interest.

Back to Top | Article Outline

References

Alshenawy HA (2014). Utility of immunohistochemical markers in differential diagnosis of follicular cell-derived thyroid lesions. J Microsc Ultras 2:127–136.
Aron M, Kapila K, Verma K (2006). Utility of galectin 3 expression in thyroid aspirates as a diagnostic marker in differentiating benign from malignant thyroid neoplasms. Indian J Pathol Microbiol 49:376–380.
Barroeta JE, Baloch ZW, Lal P, Pasha TL, Zhang PJ, LiVolsi VA (2006). Diagnostic value of differential expression of CK19, galectin-3, HBME-1, ERK, RET, and p16 in benign and malignant follicular-derived lesions of the thyroid: an immunohistochemical tissue microarray analysis. Endocr Pathol 2006; 17:225–234.
Bartolazzi A, Gasbarri A, Papotti M, Bussolati G, Lucante T, Khan A, et al. (2001). Application of an immunodiagnostic method for improving preoperative diagnosis of nodular thyroid lesions. Lancet 357:1644–1650.
Bartolazzi A, Orlandi F, Saggiorato E, Volante M, Arecco F, Rossetto R, et al., (2008). Galectin-3-expression analysis in the surgical selection of follicular thyroid nodules with indeterminate fine-needle aspiration cytology: a prospective multicenter study. Lancet Oncol 9:543–549.
Boilă AN, Boilă AN, Loghin A, Cătană R, Borda A (2012). Expression of insulin-like growth factor-II mRNA-binding protein 3 (IMP3) in various normal and neoplastic human tissues. Rev Roman Med Lab 20:163–172.
Casey MB, Lohse CM, Lloyd RV (2003). Distinction between papillary thyroid hyperplasia and papillary thyroid carcinoma by immunohistochemical staining for cytokeratin 19, galectin-3, and HBME-1. Endocr Pathol Spring 14:55–60.
Cheung CC, Ezzat S, Freeman JL, Rosen IB, Asa SL (2001). Immunohistochemical diagnosis of papillary thyroid carcinoma. Mod Pathol 14:338–342.
Chiu CG, Strugnell SS, Griffith OL, Jones SJ, Gown AM, Walker B, et al. (2010). Diagnostic utility of galectin-3 in thyroid cancer. Am J Pathol 176:2067–2081.
Danialan R, Assaad M, Burghardt J, Newcomb P, Cartun RW, Mandavilli S (2013). The utility of PAX8 and IMP3 immunohistochemical stains in the differential diagnosis of benign, premalignant, and malignant endocervical glandular lesions. Gynecol Oncol 130:383–388.
de Matos LL, Del Giglio AB, Matsubayashi CO, de Lima Farah M, Del Giglio A, da Silva Pinhal MA (2012). Expression of CK-19, galectin-3 and HBME-1 in the differentiation of thyroid lesions: systematic review and diagnostic meta-analysis. Diagn Pathol 7:97.
Debdas B, Ram ND, Uttara C, Banerjee U (2012). Cytokeratin 19 immunoreactivity in the diagnosis of papillary thyroid carcinoma. Indian J Med Paediatr Oncol 33:107–111.
DeLellis RA (2006). Pathology and genetics of thyroid carcinoma. J Surg Oncol 94:662–669.
Feng W, Zhou Z, Peters JH, Khoury T, Zhai Q, Wei Q, et al. (2011). Expression of insulin-like growth factor II mRNA-binding protein 3 in human esophageal adenocarcinoma and its precursor lesions. Arch Pathol Lab Med 135:1024–1031.
Gimm O, Castellone MD, Hoang-Vu C, Kebebew E (2011). Biomarkers in thyroid tumor research: new diagnostic tools and potential targets of molecular-based therapy. J Thyroid Res 2011:631593.
Hassan M, Naiyer A, Faizan A (2010). Fragile histidine triad protein: structure, function and its association with tumorogenesis. J Cancer Res Clin Oncol 136:333–350.
Husain AS, Jining F, Farah T, Al-Zohaili O, Husain M, Giorgadze T (2009). Differential expression of galectin-3, CK19, HBME1, and Ret oncoprotein in the diagnosis of thyroid neoplasms by fine needle aspiration biopsy. Cytojournal 6:18.
Jin L, Seys AR, Zhang S, Erickson-Johnson MR, Roth CW, Evers BR, et al. (2011). Diagnostic utility of IMP3 expression in thyroid neoplasms: a quantitative RT-PCR study. Diagn Mol Pathol 19:63–69.
Kulaçoğlu S, Erkilinç G (2015). Imp3 expression in benign and malignant thyroid tumors and hyperplastic nodules. Balkan Med J 32:30–37.
Li D, Yan D, Tang H, Zhou C, Fan J, Li S, et al. (2009). IMP-3 is a novel prognostic marker that correlates with colon cancer progression and pathogenesis. Ann Surg Oncol 16:3499–3506.
Liang HS, Zhong YH, Luo ZJ, Huang Y, Lin HD, Luo M, et al. (2009). Comparative analysis of protein expression in differentiated thyroid tumors: a multicenter study. J Int Med Res 37:927–938.
Nasr MR, Mukhopadhyay S, Zhang S, Katzenstein AL (2006). Immuno-histochemical markers in diagnosis of papillary thyroid carcinoma: utility of HBME 1 combined with CK-19 immunostaining. Mod Pathol 19:1631–1637.
Park YJ, Kwak SH, Kim DC, Kim H, Choe G, Park do J, et al. (2007). Diagnostic value of galectin-3, HBME-1, cytokeratin 19, high molecular weight cytokeratin, cyclin D1 and p27(kip1) in the differential diagnosis of thyroid nodules. J Korean Med Sci 22:621–628.
Pavelić K, Dedivitis RA, Kapitanović S, Cacev T, Guirado CR, Danić D, et al., (2006). Molecular genetic alterations of FHIT & p53 genes in benign and malignant thyroid lesions. Mutat Res 599:45–57.
Prasad ML, Pellegata NS, Huang Y, Nagaraja HN, de la Chapelle A, Kloos RT (2005). Galectin-3, fibronectin-1, CITED-1, HBME1 and cytokeratin-19 immunohistochemistry is useful for the differential diagnosis of thyroid tumors. Mod Pathol 18:48–57.
Rossi ED, Raffaelli M, Mule A, Miraglia A, Lombardi CP, Vecchio FM, Fadda G (2006). Simultaneous immunohistochemical expression of HBME-1 and galectin-3 differentiates papillary carcinomas from hyper functioning lesions of the thyroid. Histopathology 48:795–800.
Saggiorato E, De Pompa R, Volante M, Cappia S, Arecco F, TosAP Dei, et al. (2005). 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 12:305–317.
Sahoo S, Hoda SA, Rosai J, DeLellis RA (2001). Cytokeratin 19 immunoreactivity in the diagnosis of papillary thyroid carcinoma: a note of caution. Am J Clin Pathol 116:696–702.
Saleh HA, Jin B, Barnwell J, Alzohaili O (2010). Utility of immunohistochemical markers in differentiating benign from malignant follicular derived thyroid nodules. Diagn Pathol 5:9–20.
Scognamiglio T, Hyjek E, Kao J, Chen YT (2006). Diagnostic usefulness of HBME1, galectin-3, CK19, and CITED1 and evaluation of their expression in encapsulated lesions with questionable features of papillary thyroid carcinoma. Am J Clin Pathol 126:700–708.
Slosar M, Vohra P, Prasad M, Fischer A, Quinlan R, Khan A (2009). Insulin-like growth factor mRNA binding protein 3 (IMP3) is differentially expressed in benign and malignant follicular patterned thyroid tumors. Endocr Pathol 0:149–157.
Suster S (2006). Thyroid tumors with a follicular growth pattern: problems in differential diagnosis. Arch Pathol Lab Med 130:984–988.
Wachter DL, Kristiansen G, Soll C, Hellerbrand C, Breuhahn K, Fritzsche F, et al. (2012). Insulin-like growth factor II mRNA-binding protein 3(IMP3) expression in hepatocellular carcinoma. A clinicopathological analysis with emphasis on diagnostic value. Histopathology 60:278–286.
Walter O, Prasad M, Lu S, Quinlan RM, Edmiston KL, Khan A (2009). IMP3 is a novel biomarker for triple negative invasive mammary carcinoma associated with a more aggressive phenotype. Hum Pathol 40:1528–1533.
Yantiss RK, Cosar E, Fischer AH (2008). Use of IMP-3 in identification of carcinoma in fine needle aspiration biopsies of pancreas. Acta Cytol 52:133–138.
©2015Egyptian Journal of Pathology