CDX-2

CDX are homeobox genes (homeobox genes encode nuclear transcription factors involved in the establishment of differentiation patterns during development and their maintenance in adults. All members of this family bind DNA through a highly conserved 60 amino acid region, the homeodomain.) essential to intestinal organogenesis and encode nuclear transcription factors⁶. CDX-2 is analogous to the Drosophila gene caudal. Two genes have been identified in humans (CDX-1 and CDX-2) and a third gene (CDX-4) in mice⁹. CDX2-knockout mice show that the absence of the gene (-/-) is lethal in utero, while heterozygotes (+/-) show focal loss of intestinal differentiation²² and develop hamartomas and neoplasms of the colon⁹. CDX-1 appears to be responsible for proliferation and CDX-2 for differentiation of intestinal epithelium⁸. CDX-2 appears to promote expression of the proglucagon gene⁷, sucrase-isomaltase²³ and lactase²³, and to act as a tumour suppressor gene, downregulated in colonic carcinoma¹⁰. It interacts with the tumour suppressor genes APC¹² and E-cadherin¹³, as well as bcl-2¹¹.

A monoclonal antibody is available which is effective on formalin-fixed paraffin-embedded tissue. Immunoreactivity for CDX-2 is nuclear, accompanied by faint cytoplasmic staining in some cases. Recently, two substantial paper on the diagnostic use of CDX-2 has been published^1,2; it appears promising and the findings of these papers are underpinned by more basic biological papers on this homeobox gene. The subject has been reviewed¹⁷.

Immunohistochemical expression

• normal tissues:

  • with the exception of those listed below, normal tissues are NEGATIVE for CDX-2²³.

  • CDX-1 and CDX-2 are both broadly expressed in the fetus but by the early neonatal period become restricted to the small and large bowel and pancreas. CDX-1 is predominantly expressed in the base of crypts and CDX-2 higher, in the crypt tip⁵ and villi. CDX-2 positivity is seen in intestinal epithelial cells from duodenum to rectum²³, including absorptive, goblet, (neuro)endocrine²⁶ and Paneth cells. The CDX-2 gene is more strongly expressed in the small intestine and caecum, less intensely in the distal colon⁵.

  • Scattered cells are positive in pancreatic ductules²³. Scattered cells are occasionally positive in the gallbladder (1 of 5 cases)².

  • There is no nuclear reactivity in liver, stomach²² (except within intestinal metaplasia), thyroid, parathyroid, adrenal, pancreatic islets, alveolar cells, bronchial epithelium, urogenital tract, squamous epithelium (skin, larynx, oesophagus, vagina, cervix), lymphoid tissue, CNS and peripheral nerves. Cytoplasmic staining for CDX-1 has been reported in scattered cells in the gastric antrum²². Supra-nuclear Golgi staining occurs in the normal gastric glands and in most of the cells of the foveolar epithelium of the antrum²².

• Gastric intestinal metaplasia shows ectopic overexpression of both CDX-1 (in 47/52³ and 18/18²² cases) and CDX-2(in 41/48³ and 17/18²²cases): the CDX-2 immunoreactivity is not abolished by the eradication of Helicobacter pylori¹⁹. There is no difference in the staining between complete and incomplete intestinal metaplasia²². Diffuse cytoplasmic and Golgi staining for both CDX-1 and CDX-2 may accompany the nuclear staining²². In gastritis, the mucosa shows fine granular cytoplasmic staining in both the presence and absence of intestinal metaplasia¹⁹.

• The normal oesophageal mucosa is negative¹⁸. Barrett's oesophagus shows strong nuclear and fine granular cytoplasmic staining, while inflamed oesophagus without intestinal metaplasia shows only fine granular cytoplasmic staining¹⁸.

	Glandular metaplasia of the oesophagus without goblet cells	17/4528
	Strips of superficial Alcian blue positive columnar mucosal cells devoid of goblet cells	0/1128
	Barrett's oesophagus (intestinal metaplasia with goblet cells)	45/45 (including staining of non-goblet cells)28

From this study it was argued that CDX-2 identifies form fruste intestinal metaplasia prior to the development of goblet cells²⁸!

• Granular cytoplasmic positivity has been reported in epithelial mucosa in the inflamed gallbladder and nuclear positivity in intestinal metaplasia of the gallbladder²⁵.

• Neoplasms:

One paper uses a cutoff of 25% of tumour cells positive². In order to render the papers comparable, I have presented the results showing any degree of positivity for CDX-2 as positive.

Adenocarcinomas	Colorectum	58/60 (the two negative cases were poorly differentiated and showed microsatellite instability.)1, 30/30 (cases metastatic to lung: 25 cases CK7-/CK20+, 4 cases CK7+/CK20+, 1 case CK7+/CK20-)1, 75/75 (49 primary and 29 metastatic: 64 showed staining of more than 75% of tumour cells, 10 cases 25-75% of cells and one case <25% of cells)2, 13/1314, 84%15, 100% (metastatic to liver)16, 21/21 (all cases uniformly and intensely positive)21, 60/60 (<25% of cells in 4 cases, 25-49% of cells in 2 cases, 50-75% of cells in 5 cases, >75% of cells in 49 cases: using tissue microarrays)23,24/25 (differentiated adenocarcinomas)24
	Colorectal large cell minimally differentiated carcinoma	2/1524
	Gastro-oesophageal	76/164 (<25% of cells in 23 cases, 25-49% of cells in 20 cases, 50-75% of cells in 15 cases, >75% of cells in 18 cases: using tissue microarrays)23
	Oesophagus	6/9 (these cases showed staining of more than 25% of tumour cells)2, 4/514,
	Stomach	11/20 (staining is usually heterogeneous and weaker than in colorectal carcinomas)1, 17/24 (these cases showed staining of more than 25% of tumour cells)2, 38/69 (expression greater in intestinal (undifferentiated) type than in diffuse type of gastric carcinoma.)3, 9/1214, 25/46 (119/46 cases were positive for CDX-1, strongly correlating with CDX-2 positivity, as well as with MUC2 positivity. There was no correlation between either CDX-1 or CDX-2 positivity and the type of carcinoma.)22
	Duodenum	4/4 (more than 75% of tumour cells stained)2
	Ampullary carcinoma	4/6 (<25% of cells in 1 case, 25-49% of cells in 1 case, 50-75% of cells in 1 case, >75% of cells in 1 case)23
	Pancreas	3/5 (staining is usually heterogeneous and weaker than in colorectal carcinomas)1, 7/22 (these cases showed staining of more than 25% of tumour cells)2, 2/1014, 0%15, 0%16, 9/27 (<25% of cells in 8/16 ductal carcinomas, >75% of cells in 1 mucinous adenocarcinoma: using tissue microarrays)23
	Gallbladder, cholangiocarcinoma	3/5 (gallbladder)1, 4/16 (these cases showed staining of more than 25% of tumour cells)2, 0%15, 11/23 (<25% of cells in 9 cases, 25-49% of cells in 2 cases: using tissue arrays.)23
	Hepatocellular	0/122, 0%15, 0%16, 0/12 (using tissue microarrays)23
	Ovary, mucinous adenocarcinoma	5/5 (staining is usually heterogeneous and weaker than in colorectal carcinomas)1, 11/14 (9 cases showed staining of more than 25% of tumour cells)2, 9/1114, 11%15, 1/5 (50-75% of cells in 1 case: using tissue microarrays)23
	Ovary, serous adenocarcinoma	0/51, 2%15, 2/41 (<25% of cells in 1 case, 25-49% of cells in 1 case: using tissue microarrays)23
	Ovary, non-mucinous, NOS	1/36 (one case showed staining of less than 25% of tumour cells)2, 0/514
	Ovary, endometrioid	3/10 (<25% of cells in 1 case, 50-75% of cells in 1 case, >75% of cells in 1 case: using tissue microarrays)23
	Ovary, clear cell	1/2 (<25% of cells in 1 case: using tissue microarrays)24
	Ovary, undifferentiated	0/7 (using tissue microarrays)23
	Endometrial	1/1014, 9%15, 4/22 (<25% of cells in 3 cases, 25-49% of cells in 1 case, 50-75% of cells in 1 case: using tissue microarrays)23
	Endocervical	no reports yet
	Breast	0/201, 0/10 (metastatic to lung)1, 0/342, 0/2214, 0%15, 0/70 (using tissue microarrays)23
	Prostate	0/51, 0/3 (metastatic to lung)1, 5/27 (one case showed staining of more than 25% of tumour cells)2, 0%15, 2/102 (<25% of cells in 1 case, 50-75% of cells in 1 case: using tissue microarrays)23
	Bladder adenocarcinoma	3/3 (including one urachal carcinoma)2
	Kidney	0/5 (metastatic to lung)1, 0/72, 0%15, 0/35 (using tissue microarrays)21
	Sinonasal intestinal type	no reports yet.
	Thyroid	0/4 (metastatic to lung)1, 1/36 (1 of 11 papillary carcinomas showed staining of more than 25% of tumour cells. All 25 follicular adenomas and carcinomas were negative)2, 0%15, 0/21 (9 follicular and 11 papillary carcinomas: using tissue microarrays)23
Adenoma	Duodenum	10/10 (more than 75% of tumour cells stained)2
	Colorectum	98%15
	Ovary, mucinous cystadenoma	1/132
	Ovarian benign mucinous tumour of endocervical type	0/4720
	Ovarian benign mucinous tumour of intestinal type	3/3 (The type of tumour was confirmed by mucin staining. Two cases showed diffuse nuclear positivity for CDX-2, one case showed focal positivity.)20
	Ovary, mucinous borderline tumour	1/42
Gastrointestinal neuroendocrine tumours	Gastric neuroendocrine hyperplasia	3/5 (staining weak and in a limited number of cells)26
	Gastrointestinal carcinoid, various sites	7/12 (three cases showed >75% of tumour cells staining, two cases 25-75% of cells and two cases < 25% of cells)2, 16/22 (<25% of cells in 2 case, 25-49% of cells in 2 cases, 50-75% of cells in 4 case, >75% of cells in 8 cases: positive cases restricted to mid and hind gut)23
	Gastric well differentiated neuroendocrine tumour	5/5 (staining weak and in a limited number of cells)26
	Gastric neuroendocrine carcinoma	6/726
	Duodenal well differentiated neuroendocrine tumour	4/4 (staining weak and in a limited number of cells)26
	Ileal well differentiated neuroendocrine tumour	14/14 (nuclear staining intense, cytoplasmic staining moderate)26
	Small intestinal neuroendocrine carcinoma	2/226
	Appendiceal well differentiated neuroendocrine tumour	6/6 (nuclear staining intense, cytoplasmic staining moderate)26
	Colonic well differentiated neuroendocrine tumour	0/126
	Colonic neuroendocrine carcinoma	5/526
	Rectal well differentiated neuroendocrine tumour	7/9 (staining weak and in a limited number of cells)26
	Pancreatic islet cell tumours, NOS	4/14 (<25% of cells in 2 cases, 25-49% of cells in 1 case, >75% of cells in 1 case: using tissue microarrays)23
	Pancreatic functioning well differentiated neuroendocrine tumour	0/1326
	Pancreatic non-functioning well differentiated neuroendocrine tumour	14/48 (staining faint and in a limited number of cells)26
	Pancreatic islet cell tumour	4/14 (<25% of cells in 2 cases, 25-49% of cells in 1 case, >75% of cells in 1 case: using tissue microarrays)23
	Liver neuroendocrine carcinoma	1/126
Lung primaries	Neuroendocrine hyperplasia	0/526
	Squamous cell carcinoma	0/101, 0/112, 3/4121, 0/13 (using tissue microarrays)23
	Adenosquamous carcinoma	0/521
	Adenocarcinoma NOS	0/112, 2/1214, 29/212 (15 cases with <25% of cells positive, 6 cases with 26-50% of cells positive, 3 cases with 51-75% of cells positive, 5 cases with >75% of cells positive: positive cases were 16 mixed, 8 acinar, 4 papillary and 1 solid with mucin: there is a correlation between CDX-2 and CK20 positivity.)21, 0/35 (<25% of cells in 1 case, 25-49% of cells in 1 case: using tissue microarrays.)23
	Acinar adenocarcinoma	0/371
	Papillary adenocarcinoma	0/61
	Mucinous	1/2 (one case showed positivity of less than 25% of tumour cells: this information, supplementary to the published paper, was kindly provided by Dr Gown)2, 11/114, 0/24
	Bronchoalveolar carcinoma	0/28 (8 mucinous, 15 non-mucinous, 5 mixed)1
	Large cell carcinoma	0/101, 0/621, 0/1 (using tissue microarrays)23
	non-small cell and NOS	4/33 (four cases showed positivity of less than 25% of tumour cells: this information supplementary to the paper was kindly provided by Dr Gown)2, 1<1%15
	Small cell carcinoma	0/101, 0/42, 3/16 (1 case 50-75% of cells positive, 2 cases <25% of cells positive.)21, 0/2 (using tissue microarrays)23, 1/86
	Large cell neuroendocrine carcinoma	3/8 (2 cases 50-75% of cells positive, 1 case <25% of cells positive.)21, 5/86
	Carcinoid	0/51, 1/8 (0/6 typical and 1/2 atypical with positivity in 25-50% of cells)21, 2/7 (<25% of cells in both cases)23, 0/3026
	Mucoepidermoid carcinoma	0/11
Mesothelioma		0/51, 0/72, 0/28 (using tissue microarrays)24
Other lung tumours	Sclerosing haemangioma	0/51
	Epithelioid haemangioendothelioma	0/21
	Alveolar adenoma	0/11
	Blastoma	0/21
Other carcinomas	transitional cell carcinoma of bladder	3/21 (three cases showed staining of less than 25% of tumour cells)2, 2%15, 0/7 (using tissue microarrays)23
	head and neck squamous cell carcinoma	0/132
	sarcomatoid carcinoma, various sites	0/14 (using tissue microarrays)23
Other neuroendocrine tumours	Bladder neuroendocrine carcinoma	4/1026
	Uterine neuroendocrine carcinoma	2/326
	Prostatic neuroendocrine carcinoma	2/526
	Breast neuroendocrine carcinoma	1/326
	Skin, Merkel cell carcinoma	1/1526
Salivary gland	salivary gland pleomorphic adenoma	0/62
	acinic cell tumour	0/4 (using tissue microarrays)23
	adenoid cystic carcinoma	0/17 (using tissue microarrays)23
	polymorphous low grade carcinoma	0/3 (using tissue microarrays)23
	low grade carcinoma	0/62
	salivary duct carcinoma
	neuroendocrine carcinoma	1/326
Germ cell tumours	Dysgerminoma	1/4 (<25% of cells in 1 case: using tissue microarrays)23
	Embryonal carcinoma	0/3 (using tissue microarrays)23
	Seminoma	0/6 (using tissue microarrays)23
	Yolk sac tumour	4/5 (<25% of cells in 1 case, 25-49% of cells in 2 cases, 50-75% of cells in 1 case: using tissue microarrays.)23
Ovary, granulosa cell tumour		0/8 (using tissue microarrays)23
Melanoma		0%15, 0/1 (using tissue microarrays)23
Paraganglioma		0/626
Phaeochromocytoma		0/426
Adrenal cortical adenoma		0/126
Thyroid carcinomas		0/12 (3 follicular, 7 papillary and 2 medullary carcinomas)26
Parathyroid adenoma		0/226

There is loss of CDX-2 expression in a small proportion of colonic carcinomas and in some series, poorly differentiated adenocarcinomas in patients with microsatellite instability tend to show less expression than better differentiated tumours¹, although others have not found a correlation with the grade of tumour².

CDX-1 was expressed in 51 of 69 gastric adenocarcinomas and more often in intestinal than diffuse type gastric carcinomas³.

CDX-2 positivity in lung carcinomas is associated with CK20 positivity, male gender, size >30 mm and lack of TTF-1 expression²¹. Positivity of adenocarcinomas for CDX-2 and CK20, albeit with retention of CK7, at sites outside the gastrointestinal tract, may be a manifestation of the acquisition of an intestinal phenotype.

Positivity for CDX-2 occurs in a minority of neuroendocrine carcinomas outside the gastrointestinal site, without an apparent relation to the site²⁶.

Diagnostic utility

• Identification of the primary site of a metastatic adenocarcinoma, and differentiation from a primary adenocarcinoma of lung, in combination with TTF-1, CK7 and CK20. CDX2 does not reliably differentiated between colorectal, other gastrointestinal, mucinous ovarian and bladder adenocarcinomas. It may be useful in combination with villin: if negative for both, very unlikely to be from a gastrointestinal primary. If there is positivity for CDX2 (with or without villin immunoreactivity) it is likely to be from a colorectal primary. However, note the caveats associated with primary mucinous tumours of the lung and intestinal type sinonasal adenocarcinoma. Variable staining for both CDX2 and villin suggests origin from elsewhere in the gastrointestinal tract². Nuclear immunoreactivity for CDX2 and TTF-1 is particularly useful in interpreting cytological specimens from metastatic colorectal carcinoma to lung¹. As a member of a panel to differentiated primary ovarian endometrioid/mucinous carcinoma of the ovary from metastatic colorectal carcinoma.

• Identification of the site of origin of a well differentiated neuroendocrine tumour: tumours of the ileum and appendix are strongly positive, those of the stomach, colorectum and pancreas are variable and all others are negative²⁶. For high grade tumours, most gastrointestinal neuroendocrine carcinomas are positive, but so are a minority of those from other sites²⁶.

• There is little published data on the diagnostic value of CDX-1³.

References

¹ Barbareschi, M., Murer, B., Colby, T.V., Chilosi, M., Macri, E., Loda, M. and Doglioni, C. CDX-2 Homeobox Gene Expression Is a Reliable Marker of Colorectal Adenocarcinoma Metastases to the Lungs. Am J Surg Pathol 2003;27:141-9.

² Werling, R.W., Yaziji, H., Bacchi, C.E. and Gown, A.M. CDX2, a Highly Sensitive and Specific Marker of Adenocarcinomas of Intestinal Origin: An Immunohistochemical Survey of 476 Primary and Metastatic Carcinomas. Am J Surg Pathol 2003;27:303-10.

³ Bai, Y.Q., Yamamoto, H., Akiyama, Y., Tanaka, H., Takizawa, T., Koike, M., Kenji Yagi, O., Saitoh, K., Takeshita, K., Iwai, T. and Yuasa, Y. Ectopic expression of homeodomain protein CDX2 in intestinal metaplasia and carcinomas of the stomach. Cancer Lett 2002;176:47-55.

⁴ Rossi G, presentation to Pulmonary Pathology Club, Chester, UK, 13.6.03; paper submitted.

⁵ James, R., T. Erler, et al. (1994). "Structure of the murine homeobox gene cdx-2. Expression in embryonic and adult intestinal epithelium." J Biol Chem 269(21): 15229-37.

⁶ Freund, J. N., C. Domon-Dell, et al. (1998). "The Cdx-1 and Cdx-2 homeobox genes in the intestine." Biochem Cell Biol 76(6): 957-69.

⁷ Jin, T. and D. J. Drucker (1996). "Activation of proglucagon gene transcription through a novel promoter element by the caudal-related homeodomain protein cdx-2/3." Mol Cell Biol 16(1): 19-28.

⁸ Suh, E. and P. G. Traber (1996). "An intestine-specific homeobox gene regulates proliferation and differentiation." Mol Cell Biol 16(2): 619-25.

⁹ Chawengsaksophak, K., R. James, et al. (1997). "Homeosis and intestinal tumours in Cdx2 mutant mice." Nature 386(6620): 84-7.

¹⁰ Mallo, G. V., H. Rechreche, et al. (1997). "Molecular cloning, sequencing and expression of the mRNA encoding human Cdx1 and Cdx2 homeobox. Down-regulation of Cdx1 and Cdx2 mRNA expression during colorectal carcinogenesis." Int J Cancer 74(1): 35-44.

¹¹ Mallo, G. V., P. Soubeyran, et al. (1998). "Expression of the Cdx1 and Cdx2 homeotic genes leads to reduced malignancy in colon cancer-derived cells." J Biol Chem 273(22): 14030-6.

¹² da Costa, L. T., T. C. He, et al. (1999). "CDX2 is mutated in a colorectal cancer with normal APC/beta-catenin signaling." Oncogene 18(35): 5010-4.

¹³ Hinoi, T., P. C. Lucas, et al. (2002). "CDX2 regulates liver intestine-cadherin expression in normal and malignant colon epithelium and intestinal metaplasia." Gastroenterology 123(5): 1565-77.

¹⁴ Mazziotta RM, Borczuk AC, Alexis D et al. Differential expression, by immunohistochemsitry, of CDX2 transcription factor in various adenocarcinomas. Mod Pathol 2003;15:127A. [abstract]

¹⁵ Kaimaktchiev V, Firnhofer S, Sauter G et al. Selective staining of gastrointestinal adenocarcinoma by the homeobox intestinal differentiation factor CDX2. Mod Pathol 2003;15:123A. [abstract]

¹⁶ Furlanetto A, Orvieto E, Laurino L et al. Mod Pathol 2003;15:273A.

¹⁷ Li, M. K. and A. L. Folpe (2004). "CDX-2, a new marker for adenocarcinoma of gastrointestinal origin." Adv Anat Pathol 11(2): 101-5. [review]

¹⁸ Eda, A., H. Osawa, et al. (2003). "Aberrant expression of CDX2 in Barrett's epithelium and inflammatory esophageal mucosa." J Gastroenterol 38(1): 14-22.

¹⁹ Satoh, K., H. Mutoh, et al. (2002). "Aberrant expression of CDX2 in the gastric mucosa with and without intestinal metaplasia: effect of eradication of Helicobacter pylori." Helicobacter 7(3): 192-8.

²⁰ Gaggero, G., S. Sola, et al. (2003). "[Expression of the cdx2 gene in benign intestinal-type mucinous ovarian tumors]." Pathologica 95(4): 185-91.

²¹ Yatabe, Y., T. Koga, et al. (2004). "CK20 expression, CDX2 expression, K-ras mutation, and goblet cell morphology in a subset of lung adenocarcinomas." J Pathol 203(2): 645-52.

²² Almeida, R., E. Silva, et al. (2003). "Expression of intestine-specific transcription factors, CDX1 and CDX2, in intestinal metaplasia and gastric carcinomas." J Pathol 199(1): 36-40.

²³ Moskaluk, C. A., H. Zhang, et al. (2003). "Cdx2 protein expression in normal and malignant human tissues: an immunohistochemical survey using tissue microarrays." Mod Pathol 16(9): 913-9.

²⁴ Hinoi, T., M. Tani, et al. (2001). "Loss of CDX2 expression and microsatellite instability are prominent features of large cell minimally differentiated carcinomas of the colon." Am J Pathol 159(6): 2239-48.

²⁵ Osawa, H., H. Kita, et al. (2004). "Aberrant expression of CDX2 in the metaplastic epithelium and inflammatory mucosa of the gallbladder." Am J Surg Pathol 28(9): 1253-4.

²⁶ Barbareschi, M., C. Roldo, et al. (2004). "CDX-2 homeobox gene product expression in neuroendocrine tumors: its role as a marker of intestinal neuroendocrine tumors." Am J Surg Pathol 28(9): 1169-76.

²⁷ La Rosa, S., E. Rigoli, et al. (2004). "CDX2 as a marker of intestinal EC-cells and related well-differentiated endocrine tumors." Virchows Arch.

²⁸ Groisman, G. M., M. Amar, et al. (2004). "Expression of the intestinal marker Cdx2 in the columnar-lined esophagus with and without intestinal (Barrett's) metaplasia." Mod Pathol 17(10): 1282-8.

This page last revised 8.10.2004.

©SMUHT/PW Bishop