[Histonet] IHC validation

[Troutman, Kenneth A]

Hi Laurie,

Ideally, you would like to validate both with tissues processed in your laboratory (controls and patient tissue).  Purchased TMAs are okay, but they should not be the only thing used.

For the instrument validation, I would definitely use tissue that has been processed in your lab and run the same stain on each instrument (I think we ran serial sections of vimentin) and compare them very closely for any variation (use the same antibody and the same detection kit for each instrument).

My recommendation for new antibody validations is to run a set of normal tissues and tumor tissues (we have about 15 normal tissues that we harvested and created our own sausage block) along with about 20 or so different tumors (also created in sausage blocks) to see how they stain under as many circumstances as possible (especially in the conditions your pathologists are most likely to see).

In addition to that, run a set of known positives and known negatives from your cases.  CAP recommends 10 of each, but I believe the final number is at the Medical Director's discretion.

You can also run a representative sample of positives and negatives in your lab and send another set of the same tissues to another lab that has a validated procedure for that antibody and validate it that way.

As far as the detection kit goes, we don't actually "validate" the detection kit per se.  We do a lot-to-lot comparison, which involves running tissue from the same block (we created special blocks specifically for detection kits) and run the same antibody (AE1/AE3 or CD3--something we do a LOT of) to compare from lot to lot.

I hope I answered your question.

Regards,

Ashley Troutman BS, HT(ASCP) QIHC
Immunohistochemistry Supervisor
Vanderbilt University Histopathology
1301 Medical Center Drive TVC 4531
Nashville, TN  37232
ashley.troutman <@t> vanderbilt.edu<mailto:ashley.troutman <@t> vanderbilt.edu>

Message: 21
Date: Fri, 15 Mar 2013 16:51:27 +0000
From: Laurie Colbert <lcolbert <@t> pathmdlabs.com<mailto:lcolbert <@t> pathmdlabs.com>>
Subject: [Histonet] IHC validation
To: "Histonet Post (histonet <@t> lists.utsouthwestern.edu<mailto:histonet <@t> lists.utsouthwestern.edu>)"
                <histonet <@t> lists.utsouthwestern.edu<mailto:histonet <@t> lists.utsouthwestern.edu>>
Message-ID:
                <12ECD7346266D74691EC2BFC75285E451CC048EB <@t> BFL323E10.pathmdlabs.local<mailto:12ECD7346266D74691EC2BFC75285E451CC048EB <@t> BFL323E10.pathmdlabs.local>>
Content-Type: text/plain; charset="us-ascii"

If a lab is just starting up IHC for the first time and has to validate both the IHC stainer and the AB's/detection kit, does the validation have to be done using actual patient cases - or can you use strictly control tissue and purchased microarrays?

Laurie Colbert, HT (ASCP)