bs_bs_banner
Pathology International 2015; 65: 220–230
doi:10.1111/pin.12272
Original Article
Lymphatic endothelial cancerization in papillary thyroid carcinoma: Hidden evidence of lymphatic invasion
Kien T. Mai,1 Luan D. Truong,2 Christopher G. Ball,1 Bernhard Olberg,1 Chi K. Lai1 and Bibianna Purgina1 1
Department of Anatomical Pathology, The Ottawa Hospital and University of Ottawa, Ottawa, Ontario, Canada and Department of Pathology and Genomic Medicine, The Methodist Hospital Physician Organization, Weill Cornell Medical College of Cornell University, Houston, Texas, USA 2
We hypothesize that cystic structures in metastatic papillary thyroid carcinoma (PTC) develop along the framework of lymphatic channels. To investigate this phenomenon, different categories of PTC were immunostained for D2-40 and TTF1. In this study, reactivity for D2-40 was considered as positive when there is membranous staining as often seen in lymphatic endothelial cells. Thirty cases of PTC with lymph node metastasis or with potential for lymphatic invasion and 20 cases metastatic PTC in lymph nodes were reviewed and found to show double/mosaic immunoreactivity for TTF1/D2-40 in 40–100% of cases. PTC metastasis in lymph nodes with cysts and some branching lymphatic-like channels lined by follicular cells with or without nuclear features of PTC were diffusely reactive to TTF1, and focally to D2-40. For primary and metastatic PTC, focal membranous D2-40 reactivity was also demonstrated in cysts or cleft linings. For25 thyroid neoplasms with no known potential for lymphatic invasion, there was no such immunoreactivity. The mosaic or double immunoreactivity for TTF1/ D2-40 suggests lymphatic cancerization and possible endothelial mimicry of follicular cells. Mosaic/double immunoreactivity is helpful to detect the hidden pattern of lymphatic invasion masquerading as ‘benign-appearing’ follicles and supports our hypothesis of malignant cells developing along the lymphatic framework. Key words: carcinoma, lymphatic, mimicry, papillary, thyroid
Correspondence: Kien T. Mai, MD, Anatomical Pathology, The Ottawa Hospital, General Campus, 501 Smyth Rd, Ottawa, ON, Canada K1H 8 L6. Email: [email protected] Kien T. Mai, MD, Planning, conduct, reporting the work. Luan D. Truong, MD, Planning, review of the manuscript. Christopher Ball, MD, Chi K. Lai, MD, Bibianna Purgina, MD, reporting the work and review of the manuscript. Bernhard Olberg critical review of the final manuscript with major changes. Received 23 September 2014. Accepted for publication 14 January 2015. © 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Papillary thyroid carcinoma (PTC) is known for its high potential for lymph node metastasis. Lymphatic vessel invasion of the thyroid parenchyma is not often readily identified, even in cases of PTC associated with lymph node metastasis. Regardless of the histologic architecture within the thyroid, non-encapsulated PTC occasionally gives rise to cystic metastasis in cervical lymph nodes.1–3 Due to the minimal solid component, the cystic lymph node metastases can mimic benign cysts of the neck containing thyroid tissue or benign ectopic thyroid tissue.2,3 The mechanism of cystic development in metastatic PTC is poorly understood and has not been investigated. In addition, morphologic comparison of cystic metastatic PTC in lymph nodes with cystic lymphangioma demonstrates some remarkable similarities in architecture. We hypothesize that the cystic structure in metastatic PTC develops along the framework of lymphatic channels in the lymph node. This lymphatic endothelial cancerization in PTC may represent a hidden mechanism of lymphatic invasion. In this study, we first investigate the histopathogenesis of cystic metastatic PTC by performing the immunostaining for TTF1 (thyroid transcription factor-1) and D2-40. Subsequently, we investigate the same phenomenon of lymphatic endothelial cancerization in primary PTC. To determine the sensitivity and the specificity of the immunostaining in identifying lymphatic endothelial cancerization, we expand the immunostaining to different types of thyroid lesion.
MATERIALS AND METHODS Ethics approval from the investigational review board of our medical center was obtained and the study was conducted in a manner compliant with HIPAA (The Health Insurance Portability and Accountability Act of 1996). The review board waived the need for informed consent for all patients. Ten consecutive cases of papillary thyroid carcinoma with cystic metastasis in cervical lymph nodes and their related
Lymphatic endothelial cancerization
221
Figure 1 Multilocular cystic metastatic papillary thyroid carcinoma in a cervical lymph node clinically mimicking a nerve sheath tumor in the neck. The primary PTC was occult (micro-carcinoma) and sclerosing. (A,B) low and high magnification showing the multilocular cystic lymph node. (C,D,E) The cyst wall cell lining showing membranous reactivity along the apical surface for D2-40 (C), nuclear reactivity for TTF1 in almost all lining cells (D) and negative reactivity for CD31. Inset: high magnification of area in C. (F) clusters of small follicle-like structures adjacent to a large cyst from the same case in A,B. (G,H) The clusters of small follicle-like structures showing positive membranous reactivity for D2-40 (in flat follicular cells) (G), nuclear reactivity for TTF1 (H). (I-L) The thyroidectomy specimen revealed a 5 mm nodule of sclerozing papillary thyroid micro-carcinoma. IJK from the same area in consecutive serial sections showing lymphatic-like channels displaying focal mosaic/co-expression for D2-40 and strong reactivity for TTF1. I: HE stain, J: TTF1 immunostaining, K: D2-40 immunostaining and L:section in K was de-stained and re-stained for TTF1 (note the section was focally lifted and folded). The panel shows groups of cells displaying strong reactivity for D2-40 in flat follicular cells in areas b,c,d,e, and TTF1 in areas marked by a, b, c and d. Note the groups of cells in b, c d and e with arrangement suggestive of longitudinal sections from the same vessel. Note also the moderate cytoplasmic reactivity for D2-40 in malignant cells with nuclear atypia characteristic for PTC.
primary thyroid lesions were retrieved from the Anatomical Pathology files. Cases containing lymph nodes with cystic areas accounting for more than 50% of tissue or multilocular cystic areas occupying more than 0.5 cm in diameter were selected. The reviewed cases are as follows: PTC with cervical lymph node metastasis without cystic change (10 consecutive cases), non-encapsulated classic PTC without cervical lymph node metastasis (10 consecutive cases), encapsulated follicular variant PTC without lymph node metastasis (10 cases), follicular adenoma (10 cases) and follicular carcinoma (5 cases) were selected for control. At
least one representative section from each case was submitted for immunohistochemical analysis. The period of follow-up for cases without metastasis was 6 to 42 months (mean, 28 months). Immunohistochemical staining was performed on formalinfixed paraffin embedded tissue sections using the BONDMAX automated system (Leica MicroSystems, Richmond Hill, ON, Canada). Tissue sections were subjected to epitope retrieval using the Bond epitope retrieval solution (Leica) appropriate to each primary antibody. Primary antibody binding was detected using the biotin-free Bond Polymer
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
222
K. T. Mai et al.
Figure 1
Refine Detection system (Leica). This system exposes tissue sections to primary antibody for 15 min, post-primary solution for 8 min, Bond polymer for 8 min, peroxide block for 5 min, 3,3-diaminobenzidine (DAB) chromogen for 10 min and hematoxylin counterstain for 7 min. The immunostaining was performed with the following antisera: TTF1 (Leica [Concorde, Ontario], catalogue PA0364, clone SPT24, dilution ready to use with epitope retrieval system 1 for 30 min), D2-40 (Dako [Glostrup, Denmark], catalogue 3619, clone D2-40, dilution 1:100, no required antigen retrieval) and CD31 (Dako, catalogue M0824, clone JC70A, dilution: 1:50 with epitope retrieval system 1 for 20 min). In the method of immunostaining, the staining for D2-40 and TTF1 requires different heat retrieval protocols. Therefore, for the purpose of demonstrating groups of similar cells
Continued
exhibiting either D2-40 or TTF1 reactivity (mosaic pattern) or same cell exhibiting D2-40 and TTF1 reactivity (double reactivity), we performed: (i) immunostaining on serial consecutive sections; or (ii) de-staining of the slide immunostained for D2-40 then re-staining for TTF1; or (iii) combined immunostaining for TTF1 (red chromogen) and TTF1 (brown chromogen). The TTF1 reactivity was recorded as positive when the nuclear staining was strong. Immunostaining for D2-40 was recorded as: (i) positive membranous reactivity when the staining was focal or diffuse along the cellular membrane (with or without cytoplasmic staining) as often seen in lymphatic endothelial cells; and (ii) cytoplasmic staining (without membranous accentuation) was considered as a nonspecific marker for lymphatic endothelial cells.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
223
Figure 2 Cystic metastatic papillary thyroid carcinoma in lymph node concurrent with a primary classic type PTC. (A,B) low and high magnification showing collapsed cystic space mimicking a cystic lymphangioma. (C,D) The cyst wall lining cells showing membranous reactivity for D2-40 (C) and nuclear reactivity for TTF1 in almost all lining cells (D). Inset: a high magnification of an area in C with positive reactivity for D2-40. (E,F) Different lymph nodes from the same case in A,B showing cysts and branching clefts reminiscent of lymphatic sinusoids. Inset in E: The cyst wall lining cells showing positive reactivity for D2-40. (GH) The cyst with wall lining cells indicated by arrow in F showing equivocal reactivity for D2-40 (G) and nuclear reactivity for TTF1 (H). (I) An area peripheral to lymph node capsule showing reactive stroma and thin wall vessels with ectasia. (J) New thin wall vessels showing scattered endothelial-like cells reactive for TTF1. Inset: A thin wall vessel showing membranous reactivity for D2-40.
Statistical analysis was performed using SISA software (http://home.clara.net.sisa/).
RESULTS Within the cervical lymph nodes, the PTC metastases consisted of single or multilocular thin-walled cystic structures not surrounded by a coat of thick fibrous tissue. The cystic spaces demonstrated varying degrees of distension. Some cysts were completely collapsed and flat whereas others maintained a round or oval configuration
(Figs 1A,1F,2A,2B,2E,2G) and contained thin proteinaceous material. Occasionally, macrophages and intracystic hemorrhage were seen. Flat cystic spaces usually displayed tortuous configuration and were associated with branching (Fig. 2F,2G). In tissue sections taken in the corticomedullary plane, cysts were seen in areas of peripheral sinusoid vessel location (Fig. 2A). The lining cells were flat cuboidal or low columnar and associated with varying degrees of papillary formation. Typical PTC nuclear features were most easily identified in the lining cells that were round and plump and often absent in flat follicular cells.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
224
K. T. Mai et al.
Figure 2
Continued
In this study, two patterns of reactivity of D2-40 were observed (Figs 1–6): 1 Membranous reactivity was often (i) stronger, (ii) in focal areas in cells lining clefts or cysts or at the periphery of cell groups, and (iii) in cells with small or flat nuclei (best appreciated with TTF1 staining) without significant nuclear atypia as seen in PTC 2 Cytoplasmic reactivity was often (i) weaker than membranous reactivity, (ii) diffuse in the tumor, and (iii) in tumor cells with typical nuclear changes of PTC. The absence of accentuation of membranous reactivity is remarkable and distinguishes the cytoplasmic D2-40 reactivity identified in PTC cells from the membranous reactivity in the lymphatic endothelial cells and PTC lining endothelial cells Within the PTC cystic cervical lymph node metastases, TTF1 immunostaining showed a strong and extensive nuclear reactivity in the cyst-lining cells and negative reactivity in the lymphatic and vascular endothelial cells. D2-40 showed a focal strong predominantly apical membranous staining in the lining cells that also displayed TTF1 reactivity (Fig. 1C–1E). The membranous staining was either apical
(Fig. 1E), in continuous contour of the cellular membrane of individual cells (Fig. 1F,G). Occasional cystic spaces with flat branching configuration lined by thyroid follicular cells, confirmed by positive TTF1 and negative D2-40 immunostaining were also identified (Fig. 2F–H). CD31 immunostaining decorated blood vessels without evidence of invasion by thyroid cells. In thyroid specimens associated with lymph node metastasis, lymphatic channels with tumor thrombi demonstrated positive membranous D2-40 immunoreactivity in small solid groups of follicular cells. In focal areas with follicles and cysts exhibited focal D2-40 reactivity forming partial contours of lymphatic vessels (Figs 1I–L,3A–C,4A–C,5A–G). In the central portion of primary PTC, lymphatic channels were often absent or only rarely identified with D2-40 immunostaining. CD31 immunostaining was not reactive in vessels containing tumor cells. In both primary and metastatic PTC, immunostaining was useful to identify branching, curved, or tortuous lymphatic vessels with endothelial cells displaying mosaic/coexpression of TTF1 and D2-40 (Figs 1I–L,2F–H,2IJ,3A– C,5A–G). These endothelial cells were often small or flat and lacked nuclear features of PTC. In addition, within areas of PTC, there were a varying number of follicles formed by benign-appearing nuclei mimicking entrapped benign thyroid follicles. These follicles were not only reactive for TTF1, but also displayed extensive or focal membranous reactivity for D2-40 (Fig. 5A–G). Table 1 summarizes results of immunostaining for D2-40 and TTF1 of classic PTC with cystic metastasis, classic PTC with solid metastasis, classic PTC without metastasis, encapsulated follicular variant PTC without metastasis, follicular adenoma, and follicular carcinoma. The primary lesions showed mosaic/co-expression of TTF1 and D2-40 in 100, 80, 40, 0, 0 and 0%, respectively. The difference between cases with membranous D2-40 reactivity in classic PTC with metastasis (20 cases) and classic PTC without metastasis (10 cases) was not statistically significant (P = 0.06) Of interest, cytoplasmic staining with or without cytoplasmic accentuation was seen in 40, 60, 40, 0, 0, 0% respectively. The metastatic lesions showed mosaic/ co-expression of TTF1 and D2-40 in 80, 75% for cystic and solid metastases respectively; cytoplasmic staining without cytoplasmic accentuation was seen in 50, 50% respectively.
DISCUSSION Papillary thyroid carcinoma is often a multicentric disease with frequent lymph node metastasis. The multiple foci of PTC in the gland and in metastatic lesions have been demonstrated to be both monoclonal and polyclonal in molecular
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
225
Figure 3 Diffuse sclerosing variant PTC with lymph node metastasis. (AB) Foci of PTC of different sizes. In the center, there was a group of tumor cell within elongated space lined by flat tumor cells (reactive TTF1 in B). (C) Immunostaining for D2-40: Flat cells surrounding elongated spaces showed focal membranous reactivity. Note the tumor cells within a large lymphatic channel showing weak cytoplasmic and membranous reactivity.
studies.4,5 Intra-glandular lymphovascular spread is believed to be the mechanism of development of multifocal and monoclonal disease.4 However lymphovascular invasion is only demonstrated in a small number of cases. Part of the challenge to identify lymphovascular invasion is that the invaded lymphovascular channels are overgrown by the tumor, as evidenced by the absence of lymphatic channels in the central part of most PTCs with lymph node metastasis. The thin walls of the lymphatic vascular channels and the invasive nature of the tumor are likely major factors in masking the evidence of lymphatic invasion. D2-40 is considered to be the most sensitive lymphatic marker;6–9 it is a monoclonal antibody that reacts to an
epitope on podoplanin, a 40 kD O-linked sialoglycoprotein. The epitope is a mucin-type transmembrane glycoprotein found originally in lymphatic endothelium and fetal testis. D2-40 can be used to distinguish lymphatic vessels from vascular endothelial cells. In addition to the sensitivity for vascular tumors,8,10–13 this antibody is also reactive in nonendothelial cell types and tumoral counterparts, including interstitial cells of Cajal, myoepithelial cells in breast, basal cells of normal sebaceous glands, non-neoplastic and neoplastic adrenal cortex, follicular dendritic cells, malignant mesothelioma (epithelioid type), pure seminomas, osteocytes and ependymal cells,6,14–20 and squamous cell carcinoma of skin, head and neck, cervix, digestive tract and
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
226
K. T. Mai et al.
Figure 4 Tall cell variant of PTC with lymph node metastasis 1 year after thyroidectomy. (A) Tall cell PTC with desmoplastic stroma. No lymphatic invasion is seen with H&E routine stain. Note the microcystic and cleftlike space between opposing thin arrows. Inset: high magnification showing tumor cells lining a cleft wall. (B,C) Immunostaining for D2-40 (B) and TTF1 (C) showing strong reactivity of flat follicular cells lining the microcysts/cysts and cleft like space. Note the non-specific weak-moderate cytoplasmic staining for D2-40 and the strong nuclear staining for TTF1 for most PTC cells. Inset in C: high magnification of the area indicated by arrowhead.
lung.21–23 Podoplanin has been proposed to be an important factor in tumor progression, tumour invasion by promoting tumor cell mobility, and tumor lymphangiogenesis.21,23,24 D2-40 reactivity was consistently absent in normal thyroid follicular cells.25,26 TTF1 (thyroid transcription factor-1) is encoded by a single gene located on chromosome 14. This is a transcription factor essential for the morphogenesis and differentiation of the thyroid, lung and ventral forebrain. In addition to lung and thyroid normal and neoplastic counterparts, TTF1 nuclear reactivity has been observed in various neuroendocrine neoplasms, and in low frequency in tumors from other
sites such as GI tract, prostate and ovary. Reactivity to TTF1 in lymphatic and vascular endothelia has not been observed27–30 In this study, the cystic spaces were round to flat or slit-like with branching and anastomosis with adjacent cysts. We demonstrate focal membranous reactivity for D2-40 in scattered cyst lining cells. Since TTF1 reactivity was positive in almost all lining cells of follicles and cysts, the superimposed D2-40 in focal areas likely represents some thyroid follicular cells displaying double reactivity, as demonstrated in Fig. 6D. Furthermore, the possibility of a mosaic of TTF1 and D2-40 reactivity is also evident.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
227
Figure 5 PTC with focal areas of neoplastic spindle cells and extensive cervical lymph node metastasis. No lymphatic invasion could be demonstrated with routine hematoxylin-eosin stain; (A) Tumor showing areas of tumoral necrosis (right upper corner), spindle cells, crescent shaped vessel-like structure (thin arrow) and a benign appearing follicle-like structure with colloid material (double arrow in A,B). (B,C) Neoplastic spindle cells, crescent shaped vessel-like structure (thin arrow) and the benign appearing follicle-like structure (double thin arrow) showing positive reactivity for D2-40 (C) and TTF1 (D). (D) Double stain for D2-40 (red) and TTF1 (brown) showing the crescent shaped vessel-like structure (thin arrow) with cell lining displaying mosaic/co-expression for both antibodies. Note that the wall on the left side with fewer nuclei with weaker TTF1 reactivity was associated with stronger D2-40 reactivity. (E,F,G) Another area with another crescent shaped vessel-like structure (thin arrow), three benign appearing follicle-like structures with colloid material and longitudinal section of two lymphatic-like channels a-bc-d and showing mosaic/co-expression for both D2-40 and TTF1. The three benign appearing follicle-like structures were focally reactive for D2-40 in flat/small follicular cells. Inset: membranous D2-40 reactivity.
To investigate the sensitivity and specificity of TTF1 and D2-40, we tested the staining on 30 cases of PTC with lymph node metastasis (hence lymphatic invasion) or with potential for lymphatic invasion; 40 to 100% of cases showed evidence of double/mosaic immunoreactivity. For 30 cases with no known potential of lymphatic invasion, there was no double/mosaic immunoreactivity. In view of the specificity of the membranous D2-40 reactivity for lymphatic endothelial cells in the setting of this study, the co-expression or the mosaic of TTF1 and D2-40 (membranous) reactivity likely represents PTC follicle cells that occupy lymphatic endothelium. The changes of follicular cells displaying membranous D2-40 reactivity are of interest. Such a phenomenon likely develops as a result of: (i) epithelialmesenchymal interaction in which the lymphatic stromal tissue induces change in intravascular tumor cells to acquire podoplanin epitope in the cytoplasmic membrane. This stromal-epithelial induction phenomenon has been well documented in benign as well as in malignant tissue in various organs such as prostate, breast, urinary bladder.31–34
(ii) phenomena similar to vasculogenic mimicry described in melanoma, glioblastoma, breast choriocarcinoma.35–38 In vasculogenic mimicry, tumor cells are organized into capillary-like tubules under the activation of vascular endothelial growth factor-A (VEGFA) and its downstream transducer VGEF receptors 1 and 2 and protein kinase C.39–42 In PTC, lymphatic vasculogenic mimicry similar to vasculogenic mimicry is a possible mechanism for tumor cells displaying immunohistochemical properties of lymphatic endothelial cells, as seen in Figs 1I–L, 2IJ and 5A–G. Podoplanin and VEGF-D are proposed to be candidates for tumor lymphangiogenesis.24,43 In this study, Fig. 2I–J may represent an example of new lymphatic vessel formed by TTF1reactive spindle cells in the reactive tissue in a lymph node with metastatic PTC. Figures 1I–J and 5E–G likely represent the same phenomenon. Therefore, tumor cells in lymphatic channels do not always represent a simple vascular invasion, but they also represent malignant cells involved in lymphatic neo-vasculogenesis or lymphatic endothelial mimicry. A majority of these malignant
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
228
K. T. Mai et al.
Figure 5
cells with lymphatic endothelial mimicry are not associated with typical nuclear changes in PTC. The membranous D2-40 reactivity of the thyroid follicular cells is distinct from the cytoplasmic D2-40 reactivity identified in PTC cells as described in two previous studies.25,26 We also identified PTC displaying positive cytoplasmic D2-40 reactivity without membranous accentuation in less than 50% of cases, as seen other studies. In conclusion, the recognition of thyroid follicular cells with co-expression/mosaicism of TTF1 and D2-40 (membranous) reactivity is helpful in the identification of: a) cancerization of lymphatic endothelium with partial or complete occlusion of lymphatic vessels mimicking thyroid follicles or cysts in
Continued
primary PTC, b) identification of malignant follicles or cysts lacking typical nuclear features of PTC c) lymphatic channels lined by metastatic follicular cells with formation of cystic spaces reminiscent of cystic lymphangioma in cervical lymph nodes.
DISCLOSURE The authors have no affiliations with or involvement in any organization or entity with any financial interest or nonfinancial interest in the subject matter or materials discussed in this manuscript.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
229
Figure 6 Papillary carcinoma with strong cytoplasmic reactivity for D2-40 without membranous reactivity.
Table 1
Results of follicular cells with positive immunoreactivity for D2-40 with membranous versus cytoplasmic pattern Primary Lesions Membranous
Classic PTC with cystic metastasis (10 cases) Classic PTC with solid metastasis (10 cases) Classic PTC without metastasis (10 cases) Encapsulated follicular variant PTC without metastasis (10 cases) Follicular adenoma (10 cases) Follicular carcinoma (5 cases)
Metastatic Lesions
Cytoplasmic Cytoplasmic and Cytoplasmic Cytoplasmic and only membranous Membranous only membranous
10 8 4 0
2 3 2 0
2 2 1 0
8 6 NA NA
2 2 NA NA
3 3 NA NA
0 0
0 0
0 0
NA NA
NA NA
NA NA
Period of follow-up: 6 months to 42 months. NA, not available.
REFERENCES 1 Monchik JM, De Petris G, De Crea C. Occult papillary carcinoma of the thyroid presenting as a cervical cyst. Surgery 2001; 129: 429–32. 2 Kawamura S, Kishino B, Miyauchi A et al. The differential diagnosis of cystic neck masses by the determination of thyroglobulin concentrations in the aspirates. Clin Endocrinol (Oxf) 1984; 20: 261–7. 3 Xu JJ, Kwan K, Fung K. Papillary thyroid carcinoma in a lateral neck cyst: Primary of ectopic thyroid tissue versus cystic metastasis. J Laryngol Otol 2013; 127: 724–7. 4 McCarthy RP, Wang M, Jones TD, Strate RW, Cheng L. Molecular evidence for the same clonal origin of multifocal papillary thyroid carcinomas. Clin Cancer Res 2006; 12: 2414–8. 5 Shattuck TM, Westra WH, Ladenson PW, Arnold A. Independent clonal origins of distinct tumor foci in multifocal papillary thyroid carcinoma. N Engl J Med 2005; 3522: 406–12. 6 Schacht V, Dadras SS, Johnson LA et al. Up-regulation of the lymphatic marker podoplanin, a mucin-type transmembrane glycoprotein, in human squamous cell carcinomas and germ cell tumors. Am J Pathol 2005; 166: 913–21. 7 Hirakawa S, Hong YK, Harvey N et al. Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol 2003; 162: 575–86.
8 Fakunaga M. Expression of D2-40 in lymphatic endothelium of normal tissue and in vascular tumors. Histopathology 2005; 46: 396–402. 9 Kalof AN, Cooper K. D2-40 Immunohistochemistry—So Far! Adv Anat Pathol 2009; 16: 62–4. 10 Kahn HJ, Bailey D, Marks A. Monoclonal antibody D2-40, a new marker of lymphatic endothelium, reacts with Kaposi’s sarcoma and a subset of angiosarcomas. Mod Pathol 2002; 15: 434–40. 11 Roy S, Chu A, Trojanowski JQ et al. D2-40, a novel monoclonal antibodyagainst the M2A antigen as a marker to distinguish hemangioblastoma form renal cell carcinoma. Acta Neuropathol (Berl) 2005; 109: 497–502. 12 Galambos C, Nodit L. Identification of lymphatic endothelium in pediatric vascular tumors and malformations. Pediatr Dev Pathol 2005; 8: 181–9. 13 Debelenko LV, Perez-Atayde AR, Mulliken JB et al. D2-40 immunohistochemical analysis of pediatric vascular tumors reveal positivity in Kaposiform hemangioendotheliomas. Mod Pathol 2005; 18: 1454–60. 14 Gong L, Chen P, Liu X et al. Expressions of D2-40, CK19, galectin-3, VEGF and EGFR in papillary thyroid carcinoma. Gland Surg 2012; 1: 25–32. 15 Rabban JT, Chen YY. D2-40 expression by breast myoepithelium: Potential pitfalls in distinguishing intralymphatic carcinoma from in situ carcinoma. Hum Pathol 2008; 39: 175– 83.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
230
K. T. Mai et al.
16 Browling L, Parker A, Bailey D. D2-40 is a sensitive and specific marker in differentiating primary adrenal cortical tumors from both metastatic clear cell renal cell carcinoma and pheochromocytoma. J Clin Pathol 2008; 61: 293–96. 17 Yu H, Gibson JA, Pinkus GS et al. Podoplanin (D2-40) is a novel marker for follicular dendritic cell tumor. Am J Clin Pathol 2007; 128: 776–82. 18 Ordonez NG. D2-40 and podoplanin are highly specific and sensitive immunohistochemical markers of epitheloid malignant melanoma. Hum Pathol 2005; 36: 72–380. 19 Lau SK, Weiss LM, Chu PG. D2-40 immunohistochemistry in the differential diagnosis of seminoma and embryonal carcinoma: A comparative immunohistochemical study with KIT (VD117) and CD30. Mod Pathol 2007; 20: 320–25. 20 Gomaa A, Yaar M, Bhawan J. Cutaneous immunoreactivity of D2-40 antibody beyond the lymphatics. Am J Dermatopathol 2007; 9: 18–21. 21 Kreppel M, Scheer M, Drebber U, Ritter L, Zöller JE. Impact of podoplanin expression in oral squamous cell carcinoma: Clinical and histopathologic correlations. Virchows Arch 2010; 456: 473–82. 22 Chuang WY, Yeh CJ, Wu YC et al. Tumor cell expression of podoplanin correlates with nodal metastasis in esophageal squamous cell carcinoma. Histol Histopathol 2009; 24: 1021–7. 23 Wicki A, Christofori G. The potential role of podoplanin in tumour invasion. Br J Cancer 2007; 96: 1–5. 24 Cueni LN, Hegyi I, Shin JW et al. Tumor lymphangiogenesis and metastasis to lymph nodes induced by cancer cell expression of podoplanin. Am J Pathol 2010; 177: 1004–16. 25 Rudzin´ska M, Gaweł D, Sikorska J et al. The role of podoplanin in the biology of differentiated thyroid cancers. PLoS ONE 2014; 9: e96541. 26 Wang SL, Li SH, Chen WT, Chai CY. Expression of D2-40 in adjunct diagnosis of papillary thyroid carcinoma. APMIS 2007; 15: 906–10. 27 Compérat E, Zhang F, Perrotin C et al. Variable sensitivity and specificity of TTF-1 antibodies in lung metastatic adenocarcinoma of colorectal origin. Mod Pathol 2005; 18: 1371–6. 28 La Rosa S, Furlan D, Franzi F et al. Mixed exocrineneuroendocrine carcinoma of the nasal cavity: Clinicopathologic and molecular study of a case and review of the literature. Head Neck Pathol 2013; 7: 76–84. 29 Rindi G, Bordi C, La Rosa S et al. Gastroenteropancreatic (neuro)endocrine neoplasms: The histology report. Dig Liver Dis 2011; 43 Suppl 4:S356–60. 30 Fernández-Aceñero MJ, Córdova S, Manzarbeitia F, Medina C. Immunohistochemical profile of urothelial and small cell carcinomas of the bladder. Pathol Oncol Res 2011; 17: 519–23.
31 Neubauer BL, Chung LW, McCormick KA et al. Epithelialmesenchymal interactions in prostatic development. II. Biochemical observations of prostatic induction by urogenital sinus mesenchyme in epithelium of the adult rodent urinary bladder. J Cell Biol 1983; 96: 1671–6. 32 Cunha GR, Fujii H, Neubauer BL et al. Epithelial-mesenchymal interactions in prostatic development. I. morphological observations of prostatic induction by urogenital sinus mesenchyme in epithelium of the adult rodent urinary bladder. J Cell Biol 1983; 96: 1662–70. 33 Chung LW, Cunha GR. Stromal-epithelial interactions: II. Regulation of prostatic growth by embryonic urogenital sinus mesenchyme. Prostate 1983; 4: 503–11. 34 Haslam SZ. Stromal-epithelial interactions in normal and neoplastic mammary gland (Review). Cancer Treat Res 1991; 53: 401–20. 35 El Hallani S, Boisselier B, Peglion F et al. newalternative mechanism in glioblastoma vascularization: Tubular vasculogenic mimicry. Brain 2010; 133 (Pt 4): 973–82. 36 Vartanian A, Stepanova E, Grigorieva I et al. VEGFR1 and PKCα signaling control melanoma vasculogenic mimicry in a VEGFR2 kinase-independent manner. Melanoma Res 2011; 21: 91–8. 37 Lezcano C, Kleffel S, Lee N et al. Merkel cell carcinoma expresses vasculogenic mimicry: demonstration in patients and experimental manipulation in xenografts. Lab Invest 2014;94: 1092–102. 38 IeM S. Trophoblastic vasculogenic mimicry in gestational choriocarcinoma. Mod Pathol 2011; 24: 646–52. 39 Francescone R, Scully S, Bentley B et al. Glioblastoma-derived tumor cells induce vasculogenic mimicry through Flk-1 protein activation. J Biol Chem 2012; 287: 24821–31. 40 Yao X, Ping Y, Liu Y et al. Vascular endothelial growth factor receptor 2 (VEGFR-2) plays a key role in vasculogenic mimicry formation, neovascularization and tumor initiation by Glioma stem-like cells. PLoS ONE 2013; 8: e57188. 41 Fan YL, Zheng M, Tang YL, Liang XH. A new perspective of vasculogenic mimicry: EMT and cancer stem cells (Review). Oncol Lett 2013; 6: 1174–80. 42 Kaipainen A, Korhonen J, Pajusola K et al. The related FLT4, FLT1, and KDR receptor tyrosine kinases show distinct expression patterns in human fetal endothelial cells. J Exp Med 1993; 178: 2077–88. 43 Yasuoka H, Nakamura Y, Zuo H et al. VEGF-D expression and lymph vessels play an important role for lymph node metastasis in papillary thyroid carcinoma. Mod Pathol 2005; 18: 1127– 33.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Pathology International 2015; 65: 220–230
doi:10.1111/pin.12272
Original Article
Lymphatic endothelial cancerization in papillary thyroid carcinoma: Hidden evidence of lymphatic invasion
Kien T. Mai,1 Luan D. Truong,2 Christopher G. Ball,1 Bernhard Olberg,1 Chi K. Lai1 and Bibianna Purgina1 1
Department of Anatomical Pathology, The Ottawa Hospital and University of Ottawa, Ottawa, Ontario, Canada and Department of Pathology and Genomic Medicine, The Methodist Hospital Physician Organization, Weill Cornell Medical College of Cornell University, Houston, Texas, USA 2
We hypothesize that cystic structures in metastatic papillary thyroid carcinoma (PTC) develop along the framework of lymphatic channels. To investigate this phenomenon, different categories of PTC were immunostained for D2-40 and TTF1. In this study, reactivity for D2-40 was considered as positive when there is membranous staining as often seen in lymphatic endothelial cells. Thirty cases of PTC with lymph node metastasis or with potential for lymphatic invasion and 20 cases metastatic PTC in lymph nodes were reviewed and found to show double/mosaic immunoreactivity for TTF1/D2-40 in 40–100% of cases. PTC metastasis in lymph nodes with cysts and some branching lymphatic-like channels lined by follicular cells with or without nuclear features of PTC were diffusely reactive to TTF1, and focally to D2-40. For primary and metastatic PTC, focal membranous D2-40 reactivity was also demonstrated in cysts or cleft linings. For25 thyroid neoplasms with no known potential for lymphatic invasion, there was no such immunoreactivity. The mosaic or double immunoreactivity for TTF1/ D2-40 suggests lymphatic cancerization and possible endothelial mimicry of follicular cells. Mosaic/double immunoreactivity is helpful to detect the hidden pattern of lymphatic invasion masquerading as ‘benign-appearing’ follicles and supports our hypothesis of malignant cells developing along the lymphatic framework. Key words: carcinoma, lymphatic, mimicry, papillary, thyroid
Correspondence: Kien T. Mai, MD, Anatomical Pathology, The Ottawa Hospital, General Campus, 501 Smyth Rd, Ottawa, ON, Canada K1H 8 L6. Email: [email protected] Kien T. Mai, MD, Planning, conduct, reporting the work. Luan D. Truong, MD, Planning, review of the manuscript. Christopher Ball, MD, Chi K. Lai, MD, Bibianna Purgina, MD, reporting the work and review of the manuscript. Bernhard Olberg critical review of the final manuscript with major changes. Received 23 September 2014. Accepted for publication 14 January 2015. © 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Papillary thyroid carcinoma (PTC) is known for its high potential for lymph node metastasis. Lymphatic vessel invasion of the thyroid parenchyma is not often readily identified, even in cases of PTC associated with lymph node metastasis. Regardless of the histologic architecture within the thyroid, non-encapsulated PTC occasionally gives rise to cystic metastasis in cervical lymph nodes.1–3 Due to the minimal solid component, the cystic lymph node metastases can mimic benign cysts of the neck containing thyroid tissue or benign ectopic thyroid tissue.2,3 The mechanism of cystic development in metastatic PTC is poorly understood and has not been investigated. In addition, morphologic comparison of cystic metastatic PTC in lymph nodes with cystic lymphangioma demonstrates some remarkable similarities in architecture. We hypothesize that the cystic structure in metastatic PTC develops along the framework of lymphatic channels in the lymph node. This lymphatic endothelial cancerization in PTC may represent a hidden mechanism of lymphatic invasion. In this study, we first investigate the histopathogenesis of cystic metastatic PTC by performing the immunostaining for TTF1 (thyroid transcription factor-1) and D2-40. Subsequently, we investigate the same phenomenon of lymphatic endothelial cancerization in primary PTC. To determine the sensitivity and the specificity of the immunostaining in identifying lymphatic endothelial cancerization, we expand the immunostaining to different types of thyroid lesion.
MATERIALS AND METHODS Ethics approval from the investigational review board of our medical center was obtained and the study was conducted in a manner compliant with HIPAA (The Health Insurance Portability and Accountability Act of 1996). The review board waived the need for informed consent for all patients. Ten consecutive cases of papillary thyroid carcinoma with cystic metastasis in cervical lymph nodes and their related
Lymphatic endothelial cancerization
221
Figure 1 Multilocular cystic metastatic papillary thyroid carcinoma in a cervical lymph node clinically mimicking a nerve sheath tumor in the neck. The primary PTC was occult (micro-carcinoma) and sclerosing. (A,B) low and high magnification showing the multilocular cystic lymph node. (C,D,E) The cyst wall cell lining showing membranous reactivity along the apical surface for D2-40 (C), nuclear reactivity for TTF1 in almost all lining cells (D) and negative reactivity for CD31. Inset: high magnification of area in C. (F) clusters of small follicle-like structures adjacent to a large cyst from the same case in A,B. (G,H) The clusters of small follicle-like structures showing positive membranous reactivity for D2-40 (in flat follicular cells) (G), nuclear reactivity for TTF1 (H). (I-L) The thyroidectomy specimen revealed a 5 mm nodule of sclerozing papillary thyroid micro-carcinoma. IJK from the same area in consecutive serial sections showing lymphatic-like channels displaying focal mosaic/co-expression for D2-40 and strong reactivity for TTF1. I: HE stain, J: TTF1 immunostaining, K: D2-40 immunostaining and L:section in K was de-stained and re-stained for TTF1 (note the section was focally lifted and folded). The panel shows groups of cells displaying strong reactivity for D2-40 in flat follicular cells in areas b,c,d,e, and TTF1 in areas marked by a, b, c and d. Note the groups of cells in b, c d and e with arrangement suggestive of longitudinal sections from the same vessel. Note also the moderate cytoplasmic reactivity for D2-40 in malignant cells with nuclear atypia characteristic for PTC.
primary thyroid lesions were retrieved from the Anatomical Pathology files. Cases containing lymph nodes with cystic areas accounting for more than 50% of tissue or multilocular cystic areas occupying more than 0.5 cm in diameter were selected. The reviewed cases are as follows: PTC with cervical lymph node metastasis without cystic change (10 consecutive cases), non-encapsulated classic PTC without cervical lymph node metastasis (10 consecutive cases), encapsulated follicular variant PTC without lymph node metastasis (10 cases), follicular adenoma (10 cases) and follicular carcinoma (5 cases) were selected for control. At
least one representative section from each case was submitted for immunohistochemical analysis. The period of follow-up for cases without metastasis was 6 to 42 months (mean, 28 months). Immunohistochemical staining was performed on formalinfixed paraffin embedded tissue sections using the BONDMAX automated system (Leica MicroSystems, Richmond Hill, ON, Canada). Tissue sections were subjected to epitope retrieval using the Bond epitope retrieval solution (Leica) appropriate to each primary antibody. Primary antibody binding was detected using the biotin-free Bond Polymer
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
222
K. T. Mai et al.
Figure 1
Refine Detection system (Leica). This system exposes tissue sections to primary antibody for 15 min, post-primary solution for 8 min, Bond polymer for 8 min, peroxide block for 5 min, 3,3-diaminobenzidine (DAB) chromogen for 10 min and hematoxylin counterstain for 7 min. The immunostaining was performed with the following antisera: TTF1 (Leica [Concorde, Ontario], catalogue PA0364, clone SPT24, dilution ready to use with epitope retrieval system 1 for 30 min), D2-40 (Dako [Glostrup, Denmark], catalogue 3619, clone D2-40, dilution 1:100, no required antigen retrieval) and CD31 (Dako, catalogue M0824, clone JC70A, dilution: 1:50 with epitope retrieval system 1 for 20 min). In the method of immunostaining, the staining for D2-40 and TTF1 requires different heat retrieval protocols. Therefore, for the purpose of demonstrating groups of similar cells
Continued
exhibiting either D2-40 or TTF1 reactivity (mosaic pattern) or same cell exhibiting D2-40 and TTF1 reactivity (double reactivity), we performed: (i) immunostaining on serial consecutive sections; or (ii) de-staining of the slide immunostained for D2-40 then re-staining for TTF1; or (iii) combined immunostaining for TTF1 (red chromogen) and TTF1 (brown chromogen). The TTF1 reactivity was recorded as positive when the nuclear staining was strong. Immunostaining for D2-40 was recorded as: (i) positive membranous reactivity when the staining was focal or diffuse along the cellular membrane (with or without cytoplasmic staining) as often seen in lymphatic endothelial cells; and (ii) cytoplasmic staining (without membranous accentuation) was considered as a nonspecific marker for lymphatic endothelial cells.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
223
Figure 2 Cystic metastatic papillary thyroid carcinoma in lymph node concurrent with a primary classic type PTC. (A,B) low and high magnification showing collapsed cystic space mimicking a cystic lymphangioma. (C,D) The cyst wall lining cells showing membranous reactivity for D2-40 (C) and nuclear reactivity for TTF1 in almost all lining cells (D). Inset: a high magnification of an area in C with positive reactivity for D2-40. (E,F) Different lymph nodes from the same case in A,B showing cysts and branching clefts reminiscent of lymphatic sinusoids. Inset in E: The cyst wall lining cells showing positive reactivity for D2-40. (GH) The cyst with wall lining cells indicated by arrow in F showing equivocal reactivity for D2-40 (G) and nuclear reactivity for TTF1 (H). (I) An area peripheral to lymph node capsule showing reactive stroma and thin wall vessels with ectasia. (J) New thin wall vessels showing scattered endothelial-like cells reactive for TTF1. Inset: A thin wall vessel showing membranous reactivity for D2-40.
Statistical analysis was performed using SISA software (http://home.clara.net.sisa/).
RESULTS Within the cervical lymph nodes, the PTC metastases consisted of single or multilocular thin-walled cystic structures not surrounded by a coat of thick fibrous tissue. The cystic spaces demonstrated varying degrees of distension. Some cysts were completely collapsed and flat whereas others maintained a round or oval configuration
(Figs 1A,1F,2A,2B,2E,2G) and contained thin proteinaceous material. Occasionally, macrophages and intracystic hemorrhage were seen. Flat cystic spaces usually displayed tortuous configuration and were associated with branching (Fig. 2F,2G). In tissue sections taken in the corticomedullary plane, cysts were seen in areas of peripheral sinusoid vessel location (Fig. 2A). The lining cells were flat cuboidal or low columnar and associated with varying degrees of papillary formation. Typical PTC nuclear features were most easily identified in the lining cells that were round and plump and often absent in flat follicular cells.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
224
K. T. Mai et al.
Figure 2
Continued
In this study, two patterns of reactivity of D2-40 were observed (Figs 1–6): 1 Membranous reactivity was often (i) stronger, (ii) in focal areas in cells lining clefts or cysts or at the periphery of cell groups, and (iii) in cells with small or flat nuclei (best appreciated with TTF1 staining) without significant nuclear atypia as seen in PTC 2 Cytoplasmic reactivity was often (i) weaker than membranous reactivity, (ii) diffuse in the tumor, and (iii) in tumor cells with typical nuclear changes of PTC. The absence of accentuation of membranous reactivity is remarkable and distinguishes the cytoplasmic D2-40 reactivity identified in PTC cells from the membranous reactivity in the lymphatic endothelial cells and PTC lining endothelial cells Within the PTC cystic cervical lymph node metastases, TTF1 immunostaining showed a strong and extensive nuclear reactivity in the cyst-lining cells and negative reactivity in the lymphatic and vascular endothelial cells. D2-40 showed a focal strong predominantly apical membranous staining in the lining cells that also displayed TTF1 reactivity (Fig. 1C–1E). The membranous staining was either apical
(Fig. 1E), in continuous contour of the cellular membrane of individual cells (Fig. 1F,G). Occasional cystic spaces with flat branching configuration lined by thyroid follicular cells, confirmed by positive TTF1 and negative D2-40 immunostaining were also identified (Fig. 2F–H). CD31 immunostaining decorated blood vessels without evidence of invasion by thyroid cells. In thyroid specimens associated with lymph node metastasis, lymphatic channels with tumor thrombi demonstrated positive membranous D2-40 immunoreactivity in small solid groups of follicular cells. In focal areas with follicles and cysts exhibited focal D2-40 reactivity forming partial contours of lymphatic vessels (Figs 1I–L,3A–C,4A–C,5A–G). In the central portion of primary PTC, lymphatic channels were often absent or only rarely identified with D2-40 immunostaining. CD31 immunostaining was not reactive in vessels containing tumor cells. In both primary and metastatic PTC, immunostaining was useful to identify branching, curved, or tortuous lymphatic vessels with endothelial cells displaying mosaic/coexpression of TTF1 and D2-40 (Figs 1I–L,2F–H,2IJ,3A– C,5A–G). These endothelial cells were often small or flat and lacked nuclear features of PTC. In addition, within areas of PTC, there were a varying number of follicles formed by benign-appearing nuclei mimicking entrapped benign thyroid follicles. These follicles were not only reactive for TTF1, but also displayed extensive or focal membranous reactivity for D2-40 (Fig. 5A–G). Table 1 summarizes results of immunostaining for D2-40 and TTF1 of classic PTC with cystic metastasis, classic PTC with solid metastasis, classic PTC without metastasis, encapsulated follicular variant PTC without metastasis, follicular adenoma, and follicular carcinoma. The primary lesions showed mosaic/co-expression of TTF1 and D2-40 in 100, 80, 40, 0, 0 and 0%, respectively. The difference between cases with membranous D2-40 reactivity in classic PTC with metastasis (20 cases) and classic PTC without metastasis (10 cases) was not statistically significant (P = 0.06) Of interest, cytoplasmic staining with or without cytoplasmic accentuation was seen in 40, 60, 40, 0, 0, 0% respectively. The metastatic lesions showed mosaic/ co-expression of TTF1 and D2-40 in 80, 75% for cystic and solid metastases respectively; cytoplasmic staining without cytoplasmic accentuation was seen in 50, 50% respectively.
DISCUSSION Papillary thyroid carcinoma is often a multicentric disease with frequent lymph node metastasis. The multiple foci of PTC in the gland and in metastatic lesions have been demonstrated to be both monoclonal and polyclonal in molecular
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
225
Figure 3 Diffuse sclerosing variant PTC with lymph node metastasis. (AB) Foci of PTC of different sizes. In the center, there was a group of tumor cell within elongated space lined by flat tumor cells (reactive TTF1 in B). (C) Immunostaining for D2-40: Flat cells surrounding elongated spaces showed focal membranous reactivity. Note the tumor cells within a large lymphatic channel showing weak cytoplasmic and membranous reactivity.
studies.4,5 Intra-glandular lymphovascular spread is believed to be the mechanism of development of multifocal and monoclonal disease.4 However lymphovascular invasion is only demonstrated in a small number of cases. Part of the challenge to identify lymphovascular invasion is that the invaded lymphovascular channels are overgrown by the tumor, as evidenced by the absence of lymphatic channels in the central part of most PTCs with lymph node metastasis. The thin walls of the lymphatic vascular channels and the invasive nature of the tumor are likely major factors in masking the evidence of lymphatic invasion. D2-40 is considered to be the most sensitive lymphatic marker;6–9 it is a monoclonal antibody that reacts to an
epitope on podoplanin, a 40 kD O-linked sialoglycoprotein. The epitope is a mucin-type transmembrane glycoprotein found originally in lymphatic endothelium and fetal testis. D2-40 can be used to distinguish lymphatic vessels from vascular endothelial cells. In addition to the sensitivity for vascular tumors,8,10–13 this antibody is also reactive in nonendothelial cell types and tumoral counterparts, including interstitial cells of Cajal, myoepithelial cells in breast, basal cells of normal sebaceous glands, non-neoplastic and neoplastic adrenal cortex, follicular dendritic cells, malignant mesothelioma (epithelioid type), pure seminomas, osteocytes and ependymal cells,6,14–20 and squamous cell carcinoma of skin, head and neck, cervix, digestive tract and
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
226
K. T. Mai et al.
Figure 4 Tall cell variant of PTC with lymph node metastasis 1 year after thyroidectomy. (A) Tall cell PTC with desmoplastic stroma. No lymphatic invasion is seen with H&E routine stain. Note the microcystic and cleftlike space between opposing thin arrows. Inset: high magnification showing tumor cells lining a cleft wall. (B,C) Immunostaining for D2-40 (B) and TTF1 (C) showing strong reactivity of flat follicular cells lining the microcysts/cysts and cleft like space. Note the non-specific weak-moderate cytoplasmic staining for D2-40 and the strong nuclear staining for TTF1 for most PTC cells. Inset in C: high magnification of the area indicated by arrowhead.
lung.21–23 Podoplanin has been proposed to be an important factor in tumor progression, tumour invasion by promoting tumor cell mobility, and tumor lymphangiogenesis.21,23,24 D2-40 reactivity was consistently absent in normal thyroid follicular cells.25,26 TTF1 (thyroid transcription factor-1) is encoded by a single gene located on chromosome 14. This is a transcription factor essential for the morphogenesis and differentiation of the thyroid, lung and ventral forebrain. In addition to lung and thyroid normal and neoplastic counterparts, TTF1 nuclear reactivity has been observed in various neuroendocrine neoplasms, and in low frequency in tumors from other
sites such as GI tract, prostate and ovary. Reactivity to TTF1 in lymphatic and vascular endothelia has not been observed27–30 In this study, the cystic spaces were round to flat or slit-like with branching and anastomosis with adjacent cysts. We demonstrate focal membranous reactivity for D2-40 in scattered cyst lining cells. Since TTF1 reactivity was positive in almost all lining cells of follicles and cysts, the superimposed D2-40 in focal areas likely represents some thyroid follicular cells displaying double reactivity, as demonstrated in Fig. 6D. Furthermore, the possibility of a mosaic of TTF1 and D2-40 reactivity is also evident.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
227
Figure 5 PTC with focal areas of neoplastic spindle cells and extensive cervical lymph node metastasis. No lymphatic invasion could be demonstrated with routine hematoxylin-eosin stain; (A) Tumor showing areas of tumoral necrosis (right upper corner), spindle cells, crescent shaped vessel-like structure (thin arrow) and a benign appearing follicle-like structure with colloid material (double arrow in A,B). (B,C) Neoplastic spindle cells, crescent shaped vessel-like structure (thin arrow) and the benign appearing follicle-like structure (double thin arrow) showing positive reactivity for D2-40 (C) and TTF1 (D). (D) Double stain for D2-40 (red) and TTF1 (brown) showing the crescent shaped vessel-like structure (thin arrow) with cell lining displaying mosaic/co-expression for both antibodies. Note that the wall on the left side with fewer nuclei with weaker TTF1 reactivity was associated with stronger D2-40 reactivity. (E,F,G) Another area with another crescent shaped vessel-like structure (thin arrow), three benign appearing follicle-like structures with colloid material and longitudinal section of two lymphatic-like channels a-bc-d and showing mosaic/co-expression for both D2-40 and TTF1. The three benign appearing follicle-like structures were focally reactive for D2-40 in flat/small follicular cells. Inset: membranous D2-40 reactivity.
To investigate the sensitivity and specificity of TTF1 and D2-40, we tested the staining on 30 cases of PTC with lymph node metastasis (hence lymphatic invasion) or with potential for lymphatic invasion; 40 to 100% of cases showed evidence of double/mosaic immunoreactivity. For 30 cases with no known potential of lymphatic invasion, there was no double/mosaic immunoreactivity. In view of the specificity of the membranous D2-40 reactivity for lymphatic endothelial cells in the setting of this study, the co-expression or the mosaic of TTF1 and D2-40 (membranous) reactivity likely represents PTC follicle cells that occupy lymphatic endothelium. The changes of follicular cells displaying membranous D2-40 reactivity are of interest. Such a phenomenon likely develops as a result of: (i) epithelialmesenchymal interaction in which the lymphatic stromal tissue induces change in intravascular tumor cells to acquire podoplanin epitope in the cytoplasmic membrane. This stromal-epithelial induction phenomenon has been well documented in benign as well as in malignant tissue in various organs such as prostate, breast, urinary bladder.31–34
(ii) phenomena similar to vasculogenic mimicry described in melanoma, glioblastoma, breast choriocarcinoma.35–38 In vasculogenic mimicry, tumor cells are organized into capillary-like tubules under the activation of vascular endothelial growth factor-A (VEGFA) and its downstream transducer VGEF receptors 1 and 2 and protein kinase C.39–42 In PTC, lymphatic vasculogenic mimicry similar to vasculogenic mimicry is a possible mechanism for tumor cells displaying immunohistochemical properties of lymphatic endothelial cells, as seen in Figs 1I–L, 2IJ and 5A–G. Podoplanin and VEGF-D are proposed to be candidates for tumor lymphangiogenesis.24,43 In this study, Fig. 2I–J may represent an example of new lymphatic vessel formed by TTF1reactive spindle cells in the reactive tissue in a lymph node with metastatic PTC. Figures 1I–J and 5E–G likely represent the same phenomenon. Therefore, tumor cells in lymphatic channels do not always represent a simple vascular invasion, but they also represent malignant cells involved in lymphatic neo-vasculogenesis or lymphatic endothelial mimicry. A majority of these malignant
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
228
K. T. Mai et al.
Figure 5
cells with lymphatic endothelial mimicry are not associated with typical nuclear changes in PTC. The membranous D2-40 reactivity of the thyroid follicular cells is distinct from the cytoplasmic D2-40 reactivity identified in PTC cells as described in two previous studies.25,26 We also identified PTC displaying positive cytoplasmic D2-40 reactivity without membranous accentuation in less than 50% of cases, as seen other studies. In conclusion, the recognition of thyroid follicular cells with co-expression/mosaicism of TTF1 and D2-40 (membranous) reactivity is helpful in the identification of: a) cancerization of lymphatic endothelium with partial or complete occlusion of lymphatic vessels mimicking thyroid follicles or cysts in
Continued
primary PTC, b) identification of malignant follicles or cysts lacking typical nuclear features of PTC c) lymphatic channels lined by metastatic follicular cells with formation of cystic spaces reminiscent of cystic lymphangioma in cervical lymph nodes.
DISCLOSURE The authors have no affiliations with or involvement in any organization or entity with any financial interest or nonfinancial interest in the subject matter or materials discussed in this manuscript.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
Lymphatic endothelial cancerization
229
Figure 6 Papillary carcinoma with strong cytoplasmic reactivity for D2-40 without membranous reactivity.
Table 1
Results of follicular cells with positive immunoreactivity for D2-40 with membranous versus cytoplasmic pattern Primary Lesions Membranous
Classic PTC with cystic metastasis (10 cases) Classic PTC with solid metastasis (10 cases) Classic PTC without metastasis (10 cases) Encapsulated follicular variant PTC without metastasis (10 cases) Follicular adenoma (10 cases) Follicular carcinoma (5 cases)
Metastatic Lesions
Cytoplasmic Cytoplasmic and Cytoplasmic Cytoplasmic and only membranous Membranous only membranous
10 8 4 0
2 3 2 0
2 2 1 0
8 6 NA NA
2 2 NA NA
3 3 NA NA
0 0
0 0
0 0
NA NA
NA NA
NA NA
Period of follow-up: 6 months to 42 months. NA, not available.
REFERENCES 1 Monchik JM, De Petris G, De Crea C. Occult papillary carcinoma of the thyroid presenting as a cervical cyst. Surgery 2001; 129: 429–32. 2 Kawamura S, Kishino B, Miyauchi A et al. The differential diagnosis of cystic neck masses by the determination of thyroglobulin concentrations in the aspirates. Clin Endocrinol (Oxf) 1984; 20: 261–7. 3 Xu JJ, Kwan K, Fung K. Papillary thyroid carcinoma in a lateral neck cyst: Primary of ectopic thyroid tissue versus cystic metastasis. J Laryngol Otol 2013; 127: 724–7. 4 McCarthy RP, Wang M, Jones TD, Strate RW, Cheng L. Molecular evidence for the same clonal origin of multifocal papillary thyroid carcinomas. Clin Cancer Res 2006; 12: 2414–8. 5 Shattuck TM, Westra WH, Ladenson PW, Arnold A. Independent clonal origins of distinct tumor foci in multifocal papillary thyroid carcinoma. N Engl J Med 2005; 3522: 406–12. 6 Schacht V, Dadras SS, Johnson LA et al. Up-regulation of the lymphatic marker podoplanin, a mucin-type transmembrane glycoprotein, in human squamous cell carcinomas and germ cell tumors. Am J Pathol 2005; 166: 913–21. 7 Hirakawa S, Hong YK, Harvey N et al. Identification of vascular lineage-specific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am J Pathol 2003; 162: 575–86.
8 Fakunaga M. Expression of D2-40 in lymphatic endothelium of normal tissue and in vascular tumors. Histopathology 2005; 46: 396–402. 9 Kalof AN, Cooper K. D2-40 Immunohistochemistry—So Far! Adv Anat Pathol 2009; 16: 62–4. 10 Kahn HJ, Bailey D, Marks A. Monoclonal antibody D2-40, a new marker of lymphatic endothelium, reacts with Kaposi’s sarcoma and a subset of angiosarcomas. Mod Pathol 2002; 15: 434–40. 11 Roy S, Chu A, Trojanowski JQ et al. D2-40, a novel monoclonal antibodyagainst the M2A antigen as a marker to distinguish hemangioblastoma form renal cell carcinoma. Acta Neuropathol (Berl) 2005; 109: 497–502. 12 Galambos C, Nodit L. Identification of lymphatic endothelium in pediatric vascular tumors and malformations. Pediatr Dev Pathol 2005; 8: 181–9. 13 Debelenko LV, Perez-Atayde AR, Mulliken JB et al. D2-40 immunohistochemical analysis of pediatric vascular tumors reveal positivity in Kaposiform hemangioendotheliomas. Mod Pathol 2005; 18: 1454–60. 14 Gong L, Chen P, Liu X et al. Expressions of D2-40, CK19, galectin-3, VEGF and EGFR in papillary thyroid carcinoma. Gland Surg 2012; 1: 25–32. 15 Rabban JT, Chen YY. D2-40 expression by breast myoepithelium: Potential pitfalls in distinguishing intralymphatic carcinoma from in situ carcinoma. Hum Pathol 2008; 39: 175– 83.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
230
K. T. Mai et al.
16 Browling L, Parker A, Bailey D. D2-40 is a sensitive and specific marker in differentiating primary adrenal cortical tumors from both metastatic clear cell renal cell carcinoma and pheochromocytoma. J Clin Pathol 2008; 61: 293–96. 17 Yu H, Gibson JA, Pinkus GS et al. Podoplanin (D2-40) is a novel marker for follicular dendritic cell tumor. Am J Clin Pathol 2007; 128: 776–82. 18 Ordonez NG. D2-40 and podoplanin are highly specific and sensitive immunohistochemical markers of epitheloid malignant melanoma. Hum Pathol 2005; 36: 72–380. 19 Lau SK, Weiss LM, Chu PG. D2-40 immunohistochemistry in the differential diagnosis of seminoma and embryonal carcinoma: A comparative immunohistochemical study with KIT (VD117) and CD30. Mod Pathol 2007; 20: 320–25. 20 Gomaa A, Yaar M, Bhawan J. Cutaneous immunoreactivity of D2-40 antibody beyond the lymphatics. Am J Dermatopathol 2007; 9: 18–21. 21 Kreppel M, Scheer M, Drebber U, Ritter L, Zöller JE. Impact of podoplanin expression in oral squamous cell carcinoma: Clinical and histopathologic correlations. Virchows Arch 2010; 456: 473–82. 22 Chuang WY, Yeh CJ, Wu YC et al. Tumor cell expression of podoplanin correlates with nodal metastasis in esophageal squamous cell carcinoma. Histol Histopathol 2009; 24: 1021–7. 23 Wicki A, Christofori G. The potential role of podoplanin in tumour invasion. Br J Cancer 2007; 96: 1–5. 24 Cueni LN, Hegyi I, Shin JW et al. Tumor lymphangiogenesis and metastasis to lymph nodes induced by cancer cell expression of podoplanin. Am J Pathol 2010; 177: 1004–16. 25 Rudzin´ska M, Gaweł D, Sikorska J et al. The role of podoplanin in the biology of differentiated thyroid cancers. PLoS ONE 2014; 9: e96541. 26 Wang SL, Li SH, Chen WT, Chai CY. Expression of D2-40 in adjunct diagnosis of papillary thyroid carcinoma. APMIS 2007; 15: 906–10. 27 Compérat E, Zhang F, Perrotin C et al. Variable sensitivity and specificity of TTF-1 antibodies in lung metastatic adenocarcinoma of colorectal origin. Mod Pathol 2005; 18: 1371–6. 28 La Rosa S, Furlan D, Franzi F et al. Mixed exocrineneuroendocrine carcinoma of the nasal cavity: Clinicopathologic and molecular study of a case and review of the literature. Head Neck Pathol 2013; 7: 76–84. 29 Rindi G, Bordi C, La Rosa S et al. Gastroenteropancreatic (neuro)endocrine neoplasms: The histology report. Dig Liver Dis 2011; 43 Suppl 4:S356–60. 30 Fernández-Aceñero MJ, Córdova S, Manzarbeitia F, Medina C. Immunohistochemical profile of urothelial and small cell carcinomas of the bladder. Pathol Oncol Res 2011; 17: 519–23.
31 Neubauer BL, Chung LW, McCormick KA et al. Epithelialmesenchymal interactions in prostatic development. II. Biochemical observations of prostatic induction by urogenital sinus mesenchyme in epithelium of the adult rodent urinary bladder. J Cell Biol 1983; 96: 1671–6. 32 Cunha GR, Fujii H, Neubauer BL et al. Epithelial-mesenchymal interactions in prostatic development. I. morphological observations of prostatic induction by urogenital sinus mesenchyme in epithelium of the adult rodent urinary bladder. J Cell Biol 1983; 96: 1662–70. 33 Chung LW, Cunha GR. Stromal-epithelial interactions: II. Regulation of prostatic growth by embryonic urogenital sinus mesenchyme. Prostate 1983; 4: 503–11. 34 Haslam SZ. Stromal-epithelial interactions in normal and neoplastic mammary gland (Review). Cancer Treat Res 1991; 53: 401–20. 35 El Hallani S, Boisselier B, Peglion F et al. newalternative mechanism in glioblastoma vascularization: Tubular vasculogenic mimicry. Brain 2010; 133 (Pt 4): 973–82. 36 Vartanian A, Stepanova E, Grigorieva I et al. VEGFR1 and PKCα signaling control melanoma vasculogenic mimicry in a VEGFR2 kinase-independent manner. Melanoma Res 2011; 21: 91–8. 37 Lezcano C, Kleffel S, Lee N et al. Merkel cell carcinoma expresses vasculogenic mimicry: demonstration in patients and experimental manipulation in xenografts. Lab Invest 2014;94: 1092–102. 38 IeM S. Trophoblastic vasculogenic mimicry in gestational choriocarcinoma. Mod Pathol 2011; 24: 646–52. 39 Francescone R, Scully S, Bentley B et al. Glioblastoma-derived tumor cells induce vasculogenic mimicry through Flk-1 protein activation. J Biol Chem 2012; 287: 24821–31. 40 Yao X, Ping Y, Liu Y et al. Vascular endothelial growth factor receptor 2 (VEGFR-2) plays a key role in vasculogenic mimicry formation, neovascularization and tumor initiation by Glioma stem-like cells. PLoS ONE 2013; 8: e57188. 41 Fan YL, Zheng M, Tang YL, Liang XH. A new perspective of vasculogenic mimicry: EMT and cancer stem cells (Review). Oncol Lett 2013; 6: 1174–80. 42 Kaipainen A, Korhonen J, Pajusola K et al. The related FLT4, FLT1, and KDR receptor tyrosine kinases show distinct expression patterns in human fetal endothelial cells. J Exp Med 1993; 178: 2077–88. 43 Yasuoka H, Nakamura Y, Zuo H et al. VEGF-D expression and lymph vessels play an important role for lymph node metastasis in papillary thyroid carcinoma. Mod Pathol 2005; 18: 1127– 33.
© 2015 Japanese Society of Pathology and Wiley Publishing Asia Pty Ltd
