slide to be stained for CDX2
1. Colon, 2. Breast ductal carcinoma, 3. Pancreas blood type A, 4.
Pancreas blood type 0. 5-7. Colon adenocarcinomas.
All tissues were fixed in 10% neutral buffered formalin.
Criteria for assessing an CDX2 staining as
- A strong and distinct nuclear staining of
virtually all enterocytes in the normal colon epithelium.
- A strong and distinct nuclear staining of
virtually all the neoplastic cells in the three colon
- A weak to moderat, distinct nuclear
reaction in the majority of the duct epithelial cells in the
A weak cytoplasmic reaction in the enterocytes
and the neoplastic cells of the colon adenocarcinomas was accepted.
All other cells should be negative.
56 laboratories participated in the
assessment. At the assessment 17 achieved optimal (30 %), 19 good
(34 %), 14 borderline (25 %) and 6 (11 %) poor marks.
The following antibody clones were used:
mAb clone CDX2-88 (BioGenex, n=31; BioCare, n=1; Linaris,
mAb clone AMT28 (Novocastra, n=23)
In this assessment optimal staining could be
obtained with both clone CDX2-88 (14/33)*
and clone AMT28 (3/23). Sufficient stains (optimal or
good) were seen in 25 cases (76%) with clone CDX2-88 and in 11 cases
(44%) with clone AMT28.
* (number of optimal results/number of
laboratories using this Ab).
CDX2-88: The protocols giving an
optimal result were all but one based on heat induced epitope retrieval (HIER) in an
alkaline buffer such as Tris-EDTA/EGTA pH 9.0 (11/15)**, Bond Epitope Retrieval
Solution 2 (2/2) or EDTA pH 8 (1/1). Citrate buffer in combination with a
pressure cooker as HIER device could also be used (1/4). The mAb was typically diluted in the range of 1:25 –
1:200 depending on the total sensitivity of the protocol employed.
** (number of optimal results/number of
laboratories using this buffer).
AMT28: The protocols giving an optimal
result were all based on HIER in Tris-EDTA/EGTA based buffer (2/14) or HIER
in citrate buffer pH 6.0 (1/3). The mAb was typically diluted
in the range of 1:25 – 1:50 depending on the total sensitivity of
the protocol employed.
The most frequent causes of insufficient
- Insufficient HIER – too short HIER time and/or use of citrate pH
- Too low concentration of the primary Ab.
The prevalent feature of an insufficient
staining with both clones was a negative staining of the duct
epithelial cells in the pancreas (Fig. 4a and Fig. 4b) and a too
weak or negative staining of tumour cells in one or more of the
three colon adenocarcinomas (Fig. 3a and Fig. 3b). The assessment
identified pancreas as a robust and easy interpretable stain quality
indicator. It was observed that a weak to moderate but distinct
nuclear reaction in the majority of the duct epithelial cells in the
pancreas (Fig. 4a) is a far better quality indicator than the very
strong reaction of the normal enterocytes of colon (Fig. 1a).
In several of the insufficient results the normal colon enterocytes
were stained, while one of the three colon adenocarcinomas was false negative, which stresses that colon cannot
be recommended as a positive control for CDX2.
In this assessment pancreas from two patients
were included. One patient was blood type A and the other blood type
0. In a number of laboratories using clone CDX2-88 a
"Mouse Ascites Golgi" (MAG) reaction
was seen (15 out of 33). The reaction was only seen in the tissue
from the patient with blood
type A. The intensity of this unspecific staining varied
from weak to strong (Figs. 4b and 5). In general the strongest MAG reaction was seen using the Ready To Use (RTU) format of clone
CDX2-88. All the clone CDX2-88
antibodies were supplied as an ascites format.
Clone AMT28 were all supernatant antibodies and no MAG reaction was
seen with that clone.
In the assessment MAG reaction was noted, but
had no impact on the marks given to the laboratories. However, it should be stressed that the
contaminant antibodies in some of the CDX2-88 ascites mAbs may
cause difficulties in interpreting stains in tumours from blood
type A patients. As the unspecific MAG reaction is ascites related
and not seen with mAbs produced as supernatants, the manufactures
should be encouraged to avoid the ascites production of mAbs.
In this first CDX2 assessment the mAb clones CDX2-88 and AMT28 are both useful for
the demonstration of the CDX2 protein. The clone CDX2-88 seems to be
more sensitive and robust than clone AMT28. However it is a drawback that
the clone CDX2-88 is an ascites format giving MAG reaction (Fig. 5). Both
clones require efficient HIER and the concentration of the clones
should be carefully calibrated. Pancreas is an appropriate control:
A weak to moderate but distinct nuclear reaction in the majority of
the duct epithelial cells in the pancreas should be seen.