[Histonet] EGFR

[Andi Kappeler]

Hi Patricia

I assume you are talking about staining of colorectal adenocarcinoma - that 
is what the DakoCytomation EGFR pharmDx kit is most frequently used for. A 
few comments:
- Smooth muscle can stain positively for EGFR (in a low percentage of cases) 
when using the DakoCytomation kit.
- not all "norma epithelia" are EGFR positive. If you look at colon, may be 
half of the samples will show staining of the normal epithelium.
- I would love to see "uniform staining" of the membranes... This is limited 
to a certain percentage of tumors. The majority of cases show focal or 
heterogeneous staining of the tumor; often the majority of tumor cells have 
partial membrane reactivity, most tumors also show cytoplasmic staining.
- 8 hours fixation (room temperature, without microwave) may be somewhat 
short, if your samples are more than just small biopsies.

... and a few suggestions:
The DakoCytomation protocol asks to place slides in water before the 
application of Proteinase K (and also beforte the peroxidase block). We 
found this to be the cause for uneven digestion and have replaced the water 
with wash buffer from the kit (contains detergents --> better distribution 
of ProtK in subsequent step) (the FDA is faaaar away from Switzerland, and 
we care more about our patients than about the FDA...). Comparative staining 
of slides with water or with wash buffer in this particular step revealed no 
difference - except for a more even staining in particular of the border 
areas of tissue samples on those slides that went through wash buffer.
We do a CK8 (e.g. clone CAM 5.2) staining for all cases that are stained for 
EGFR. This clearly demonstrates the carcinoma cells and helps to identify 
small groups of cells in the invasive part of the tumor. Particularly 
useful, if you have one of those cases where smooth muscle and/or connective 
tissue show some reactivity for EGFR.
As an in house control we use cases of liver metastasis of colorectal 
adenocarcinoma, where the adenocarcinoma is heterogeneously 2+ positive (I 
don't want a 'retina blaster' as positive control) and where the hepatocytes 
can serve as strong positive control.
If you don't have a copy of the DakoCytomation 'EGFR pharmDx Interpretation 
Manual' (Nr. 08052), call DakoCytomation and tell them to send you a few 
copies - and then force your pathologists to put their noses in it, it may 
help them understand that not all tumors have 3+ complete membrane staining 
in 110% of all tumor cells.
Hope this helps (and let time work for you: EGFR testing will go as it has 
come, just like many other tests ...)

Andi Kappeler
Institute of Pathology, University of Bern, Switzerland

----- Original Message ----- 
From: "Traphagan, Patricia" <Patricia.Traphagan <@t> ParkNicollet.com>
To: <histonet <@t> lists.utsouthwestern.edu>
Sent: Wednesday, May 25, 2005 7:21 PM
Subject: [Histonet] EGFR


> We have been trying to achieve good staining using this antibody. We use
> the Dako EGFR kit. Dako's control that comes with the kit, turns out
> beautifully, which tells us that the stain procedure is correct. Our
> problem lies with our internal control. The pathologist are just not
> happy with it. According to them we have: Over staining of the smooth
> muscle, normal epithelial tissue does not stain correctly, and have
> un-uniform staining of the membranes. Can anyone tell me what this
> means? We use 10% NBF, fixation is probably no more than 8 hours. Our
> processing run is about 12 hours for large tissue specimens, and we use
> an alcoholic formalin in the second station, which then goes to 70% ETOH
> and the graded series of alcohols, xylenes, paraffin. We do not put the
> cut tissue in the oven, we let it air-dry overnight, possibly two
> nights. We are at a total loss. Any suggestions. We really need
> HELP!!!!!
>
> _______________________________________________
> Histonet mailing list
> Histonet <@t> lists.utsouthwestern.edu
> http://lists.utsouthwestern.edu/mailman/listinfo/histonet
>