Re: Semi-quantitative IHC slide scoring

From: Eric Miller (millere@icrf.icnet.uk)
Date: Fri Jan 30 1998 - 05:03:37 EST


Dear xx
	We use a method of assessing slides which may fit your needs,
called the immunoreactive score (IRS). This method assesses slides by
counting the number of cells in a field which are negative(0), weak
positive(1), medium staining(2) and strong positives(3) These numbers are
expressed as a percentage and multiplied by the number after the
classification to give an overall immunoreactive score for the slide
between 0 and 300. We have used this method to correlate the ELISA score
for oestrogen receptors with immunostaining of breast tumour slides with
some success, albeit a fair bit of eyeball work work!
	Of course the usual caveats apply: your block or slide may not be
representative of the tumour, and a lot of stromal cells will bias your
result. Variants exist which also assess the number of tumour cells per
field. Also be aware of the need for standards: no two immuno runs will
give you the same result in terms of intensity. 
	A good reference to this type of work is:
Kieback et al. Anticancer Research 13: 2489-2496, 1993. (IRS versus CARL
in
Ovarian Cancer)

	Eric P Miller
	Edinburgh Medical Oncology Unit

"Wouldn't it be nice if hospitals and schools had all the money they
needed and the army had to hold jumble sales to buy guns?"

On Wed, 28 Jan 1998, xx wrote:

> 
> I have been searching, without success, for a methodology description
> of  semi-quantitative scoring/interpreting of immunostained slides.  A
> reference to a book chapter or review article, the "materials and
> methods" section of an otherwise focused journal article, a symposium
> handout, or even a faded xerox of someone's old procedure manual  would
> be greatly appreciated.  This seemingly very simple request has thus far
> confounded some highly respected authorities in the field and exceeded
> the capabilities of MedLine and other electronic search engines.  Great
> honors (principally in the form of bragging rights) will accrue to the
> individual who can shed some light on this issue.  Thanks in advance for
> your assistance.
> 
> 



This archive was generated by hypermail 2b29 : Wed Apr 03 2002 - 11:51:44 EST