Diagnostic utility of cytokeratin 19 and CD56 in the differentiation of thyroid papillary carcinoma from its mimics

INTRODUCTION

Thyroid carcinomas remain the most prevalent endocrine malignant lesions in India and more than 95% of cases originate from the thyroid follicular epithelial cells. According to global estimates, approximately 586,000 cases were reported worldwide in 2020, highlighting the growing concern surrounding this disease. The Southeast Asia region contributed significantly to this global burden, with an incidence of 43,651 cases of thyroid malignancy.^[1] India, in particular, experienced a notably higher incidence, with over 42 million people affected by thyroid diseases and a noticeable increase in cases over the past decade.^[2] Most thyroid cancers originate from the follicular epithelium. Papillary thyroid carcinoma (PTC) is the most prevalent type, followed by follicular thyroid carcinoma (FTC). Both types exhibit a pronounced gender disparity, with women being twice as likely to be affected than men, typically presenting in middle age or later. Specifically, the female incidence was estimated to be 449,000 cases compared to the male incidence of 1,37,000, making thyroid carcinomas 3 times more common in females.^[1] This female predominance is potentially linked to variations in estrogen receptors and its influence on cell proliferation in thyroid cancers.^[3-5] The peak incidence of thyroid carcinomas is generally observed in the fifth to the sixth decades of life.^[6] Several morphological variants of PTC have been identified in recent years. Differentiating these variants, particularly the follicular variant of PTC (FVPTC), from other benign papillary lesions and PTC subtypes is crucial for accurate diagnosis. While characteristic nuclear features are a key indicator of PTC, their occasional presence in other thyroid lesions can pose diagnostic challenges. By immunohistochemistry (IHC), markers such as CD56 and cytokeratin 19 (CK19) have proven valuable to distinguish between benign and malignant thyroid lesions that have overlapping histological features.^[7,8] CD56, a regulatory protein involved in cell movement and tumor cell migration, has been consistently observed at high levels in healthy normal thyroid tissue and various follicular lesions. However, its expression is often decreased in PTC.^[9] Certain diagnostic challenges can arise in distinguishing follicular adenoma and encapsulated FVPTC in certain cases-for instance, an encapsulated nodule with a follicular growth pattern exhibiting optically clear nuclei with nuclear grooves can be challenging to distinguish between a follicular adenoma and encapsulated FVPTC. In such scenarios, CK19 emerges as a valuable differentiating marker.^[10] CK19, which is a type I intermediate filament protein, is intensely and diffusely expressed in PTC within thyroid nodules. Conversely, benign lesions typically demonstrate minimal or no CK19 expression.

This upregulation of CK19 is associated with the transformation of cells into a cancerous state.^[11] The objective of this study was to evaluate the utility of a combination of CD56 and CK19 by IHC in differentiating thyroid neoplasms.

MATERIAL AND METHODS

This is a retrospective observational study done on 102 cases of various benign and malignant thyroid lesions obtained from the archives of the department of pathology at a tertiary care center in India. Permission from the Institutional Ethics Committee was obtained before the commencement of the study.

RESULTS

Staining pattern with CD56 marker

CD56 cytoplasmic and membranous expression patterns were analyzed. The mean CD56 score was 0.93 ± 1.1. CD56 expression was positive in 48% of cases and negative in 52% of cases. There was a negative staining pattern in classic papillary and FVPTC [Figures 1 and 2]. On comparing the CD56 expression with pathological diagnosis, it was observed that CD56 was positive in 83.3% of cases diagnosed as follicular adenoma and 100% of cases with nodular goiter, compared to 45.5% in follicular carcinoma and 7.7% of classic PTC. The observed difference was statistically significant (P < 0.0001); similarly, CD56 expression was positive among 93.5% of the benign tumors compared to 10.7% of the malignant tumors. The observed difference was statistically significant (P < 0.0001) [Table 2].

Close

Figure 1:

(a) Papillary carcinoma of the thyroid, hematoxylin and eosin, ×200. (b) Papillary carcinoma of thyroid showing negative membranous staining of CD56 immunohistochemistry (IHC) (marked by blue arrow), ×200. (c) Papillary thyroid carcinoma showing positive cytoplasmic staining of cytokeratin19 (marked by blue arrow) IHC, ×100.

Export to PPT

Close

Figure 2:

(a) Follicular variant of papillary carcinoma, hematoxylin and eosin, ×200. (b) Follicular variant of papillary carcinoma showing negative membranous staining of CD56 immunohistochemistry (IHC) (marked by blue arrow), ×100. (c) Follicular variant of papillary carcinoma showing positive cytoplasmic staining of cytokeratin19 by (marked by blue arrow) IHC, ×200.

Export to PPT

Table 2: Statistical analysis of thyroid lesions associated with CD56 expression.

S. No	Factors	CD56 expression n (%)		Chi-square	P-value
		Positive	Negative
1	Diagnosis
	Papillary carcinoma	2 (7.7)	24 (92.3)	69.7	0.0001*
	Follicular carcinoma	5 (45.5)	6 (54.5)
	Lymphocytic thyroiditis	6 (85.7)	1 (14.3)
	Nodular goitre	25 (100)	0 (0)
	Follicular adenoma	10 (83.3)	2 (16.7)
	Follicular variant of papillary carcinoma	1 (4.8)	20 (95.2)
2	Nature of lesion
	Benign	43 (93.5)	3 (6.5)	69.3	0.0001*
	Malignant	6 (10.7)	50 (89.3)

*Statistically significant

DISCUSSION

While thyroid nodules are prevalent, most are benign. However, malignant thyroid tumors hold the distinction of being the most common endocrine malignancy. Due to the varied histopathological characteristics of malignant tumors, immune markers are becoming increasingly important in diagnostic pathology. PTC can exhibit various nuclear features, including elongation, crowding, clearing, and intranuclear grooves.^[12] Unfortunately, these features can also be present in some benign lesions, creating diagnostic challenges for PTC.^[13] In our study, to distinguish between encapsulated FVPTC and Non invasive follicular thyroid neoplasm with papillary like nuclear features (NIFTP), we applied the criteria outlined by Nikiforov et al. and the 2022 World Health Organization classification of thyroid tumors.^[14] Tumors that were fully encapsulated or well circumscribed, showed a purely follicular growth pattern, lacked true papillae and psammoma bodies, exhibited no capsular or vascular invasion, and had nuclear features of PTC were classified as NIFTP. Any lesion with papillary structures, psammoma bodies, or evidence of invasion was diagnosed as encapsulated FVPTC.^[14] However, several recent studies have explored the use of immunohistochemical markers in diagnosing thyroid neoplasms. This approach offers promise for more accurate differentiation between benign and malignant nodules.^[15] El Demellawy et al. investigated the use of a panel of IHC markers CK19 and CD56 in the diagnosis of thyroid malignancy.^[16] In a study conducted by Park et al., Ceyran et al., and Song et al., most of the studied PT with its variants demonstrated solid and diffuse CK19 immunostaining in 90.5%, 85.7%, and 100% of PTC, FVPTC, and Hurthle cell variants, respectively.^[17-19] Previously reported studies showed high levels of CD56 expression in normal thyroid tissue and in some thyroid follicular lesions, such as follicular adenoma and hyperplastic nodules; however, diffuse negative staining was observed in PTC.^[20] In this study, 102 thyroid specimens were included, of which 46 cases were benign and 56 were malignant. The present study’s significant percentage of thyroidectomy specimens was 50%, while right hemithyroidectomy specimens include 21.6% and left hemithyroidectomy specimens include 15.7%. It was observed that benign as well as malignant tumors were common in females (80.4%) [Figure 1]. This was supported by Wa Kammal et al.’s results, which indicated that females were significantly more likely to develop both benign and malignant tumors, with an incidence of 80.4%.^[12] These findings were similar to a study by Mazzaferri who examined 80% of nodules in females.^[21] By histopathological examination which is the gold standard, there were 6 cases (5.9%) of adenomatoid nodule, 13 cases (12.7%) of follicular adenoma, 10 cases (9.8%) of follicular carcinoma, 21 cases (20.6%) of FVPTC, 8 cases (7.8%) of lymphocytic thyroiditis, 10 cases (9.8%) of MNG, 7 cases (6.9%) of nodular goiter, 21 cases (20.6%) of PTC, 1 case (1%) of MNG with lymphocytic thyroiditis, papillary hyperplasia, diffuse sclerosing variant of PTC, cribriform-morular variant of PTC, respectively, and 2 cases (2%) of papillary microcarcinoma [Table 3]. In our study, CD56 expression with pathological diagnosis was observed positive in 83.3% of follicular adenoma and 100% of nodular goiter, compared to 45.5% of follicular carcinoma and 7.7% of classic PTC. CD56 expression was positive among 93.5% of the benign tumors compared to 10.7% of the malignant tumors. The observed difference was statistically significant with P < 0.0001. CK19 was positive in 96.2% of PTC and 100% of FCPTC, compared to 8.3% of follicular adenoma. A study by Revathy et al.^[22] revealed a high prevalence of CK19 expression in PTC, with 98.75% of cases showing positive staining. However, the study found no correlation between CK19 expression and established prognostic factors such as age, gender, tumor size, and lymph node metastasis.^[22] Similarly, another study demonstrated that CK19 staining strongly supports malignant papillary lesion compared to benign lesions of the thyroid, regardless of the tumor stage.^[14] A study done by Palo and Biligi demonstrated CK19 positivity in 27/31 (87.1%) PTC cases, compared to positivity in follicular adenomas (4/18 [22.2%]) and 2/10 (20%) in hyperplastic nodules.^[23] CD56 was 100% positive in papillary hyperplasia, and CK19 was 100% negative. CD56 was 100% positive in papillary hyperplasia and CK19 was 100% negative. Hence, CK19 positivity and CD56 negativity can be used as a diagnostic marker in the differentiation between PTC and papillary hyperplasia. According to Erkiliç et al., 20 of 25 cases in the MNG goiter cohort demonstrated no staining, whereas the other five cases were positive for CK19.^[24] In another study, CK19 was a sensitive marker for PTC and CD56 was the most specific marker in differentiating follicular adenoma and FTC.^[25] Nechifor-Boila et al. detected higher sensitivity of CD56 and best specificity with CK19 usage. Combining the two immunostains, better diagnostic accuracy was achieved.^[26] Our study demonstrates that the combined assessment of CK19 and CD56 by IHC, in conjunction with traditional hematoxylin and eosin staining, significantly enhances the diagnostic accuracy of PTC and its variants. While CK19 exhibits a high sensitivity in detecting PTC, CD56 demonstrates a high specificity in distinguishing it from benign thyroid lesions. The synergistic application of these markers significantly improves the diagnostic accuracy, particularly in challenging cases where traditional histological evaluation is inconclusive. Molecular testing has emerged as a valuable adjunct in the diagnosis of thyroid neoplasms. Mutational analysis for BRAF V600E, RAS, and telomerase reverse transcriptase promoter mutations can aid in confirming the diagnosis of PTC, especially in morphologically equivocal cases. BRAF V600E mutation is commonly found in classical PTC, while RAS mutations are more frequently seen in follicular-patterned lesions and NIFTP.

Although not universally available, molecular profiling can enhance diagnostic accuracy when used along with histopathology and IHC and can guide clinical management in select cases. However, in a developing country like ours with resource-limited settings, IHC markers remain practical and cost-effective tools for routine use. The limitations of this study are not incorporating molecular methods in comparing these lesions, not including non-invasive follicular neoplasm with thyroid like nuclear features (NIFTP) cases and study involving only a single center.

CONCLUSION

An accurate diagnosis of thyroid papillary-like lesions needs a careful histopathological evaluation aided with appropriate ancillary tests like IHC if molecular methods of definitive diagnosis are not feasible. Among the IHC markers, CK19 and CD56 are potentially good diagnostic markers for differentiating PTC from other benign thyroid follicular lesions. They are also helpful in differentiating FVPTC from follicular adenoma. The IHC panel for diagnosing PTC should always include CK19 as a positive marker and CD56 as a negative marker for equivocal histological diagnosis. This differentiation of PTC and its variants from its mimickers is essential for the prognosis and planning of treatment among the plethora of thyroid lesions. This would enable effective patient management, thus avoiding unnecessary expensive tests which may not be easily accessible in developing countries.