<<Disclaimer: Verify this information before applying it to your situation.>>

I asked two questions with respect to the abstracts presented
professionally by Drs. Green and Murray on the sensitivity of commercial
labs on the antibody tests for diagnosis:

>>Are the labs listed in the FAQ (last I checked) research or commercial?

>>What about the sensitivity of the commercial labs for monitoring after
>>diagnosis?

For the first, Mike Jones, a list owner, pointed out that two were
commercial, two research. You'll find the list of four in the message
CELIAC  DIAG-TST obtained by using a GET command to the listserv adddress
for NEWCEL, a collection of files.  The two hospitals listed are
"research" (my interpretation: their antibody tests are backed up with
biopsed celiac and biopsed normal blood samples) and two labs apparently
"commercial".

In the meantime, I also reread the abstracts, especially Dr. Murray's. His
may not have been posted to the list. If someone has an electronic copy,
it would be helpful to post it in this thread. What I glossed over with
Dr. Murray's result was that for diagnosis, or the need for biopsy, from a
commercial lab the sensitivity is not as low as 50% if all tests are
combined and I quote:

"When both IgG and IgA and endosymial IgA are combined in each individual
lab (the sensitivity) was 90-100%...The specificity of the endosymial
antibody in this limited survey was 100% in all participating
laboratories. "

He gives a complete table of sensitivity and selectivity for each of the
four tests.

The following response from IMMCO, one of the "labs" in that file:

>>The results of our study, especially or endomysial antibodies, are
unremarkable with almost, if not 100%, specificity and sensitivity.  If
you any questions, I will be happy to discuss with individual or
individuals on the issue.

Sincerely,

Vijay Kumar, Ph.D.

[log in to unmask]<<

(Dr. Kumar published some of the early research studies on the antibody
test. "Unremarkable" for a research scientist means he/she doesn't have to
publish an article for  professional colleagues explaining why they
couldn't get 100% sensitivity.)

But for monitoring? The situation is grim, in my opinion.
Since gluten exposure is less, fewer than four antibody tests frequently
will be positive even in a reliable lab. Thus potential sensitivities of
50 % on the gliadin antibodies, the most useful for monitoring biopsed
celiacs, are the reality for monitoring from commerical labs.

Donald Baisch pointed out:

>>The basic problem is that there are NO standard protocols for doing
blood antibody tests....A number of us are supporting the Celiac Disease
Foundation efforts to resolve and achieve a recogized standardard set of
protocols for the four antibody serum tests needed to define the NEED for
a small bowel biopsy. <<

(I question whether the same protocol is needed if all assays are biopsy
backed.) While I applaud the CDF for pushing for standards, I fear the
self-destructive effects of inter-group politics on general  efforts, such
as this, to really help celiacs.  One group proposes and generates
automatic opposition from the others. What's needed here is a competent
panel to speak for the whole celiac community, professional and patient,
to bring these commercial labs most of us are forced to use up to
acceptable standards. Patients can't get their insurance companies to
authorize biopsies every year or two for monitoring purposes, the
alternative solution.

So I propose such a panel with relevant expertise  to work toward that
goal.  Noble goal, easily said, and then forgotten? Anyone with expertise
and interest, please contact me by e-mail. Let's do it. Im volunteer
because instruments and measurements are my business professionally.   I
can bring a different perspective as a biopsied celiac and do feel that
celiacs themselves should have a voice, especially if consensus is
foundering on technical differences of opinion.

                                            Kemp Randolph
                                            Long Island

-----------------------------------------------------------
 <[log in to unmask]>
-----------------------------------------------------------