[Histonet] Re: Stain for HP

Ian R Bernard ibernard <@t> uab.edu
Fri Oct 19 09:13:15 CDT 2012


Also Alcian Yellow

-----Original Message-----
From: histonet-bounces <@t> lists.utsouthwestern.edu [mailto:histonet-bounces <@t> lists.utsouthwestern.edu] On Behalf Of Bob Richmond
Sent: Friday, October 19, 2012 5:01 AM
To: histonet <@t> lists.utsouthwestern.edu
Subject: [Histonet] Re: Stain for HP

Staining gastric biopsy specimens for Helicobacter pylori: practice varies greatly in various labs I've worked in. One lab does a blue stain (generic equivalent of Diff-Quik II), while the other does immunohistochemistry, each of them on all specimens. I've also seen labs that did a full tissue Giemsa, and one that did a cresyl echt violet stain. Haven't seen a routine Warthin-Starry in many years. I think everyone bills every one of these stains they do.

There are still a few hold-outs for "real men can see 'em on the H & E".

As far as I know, no studies have been done comparing the sensitivity (we'll forget specificity) of the various methods. My personal opinion is that none of the dye or silver methods is conspicuously better than the others. Reading the blue stain is time-consuming; I use oil immersion magnification (which real men disdain) to confirm all positives (and confirm Helicobacter heilmannii, which I have seen once), and on all apparent negatives with acute inflammation. The IHC can be read rapidly with much less magnification.

I note that many pathologists sign out any bacteria they see in a blue stain as Helicobacter, including the frequent bugs that get carried down from the oral cavity as the scope advances. These pathologists should definitely switch to IHC.

So if I have one gastric biopsy case a day, I'm happy with the blue stain. If I have ten, I want IHC. If there is no inflammation, I can do without any stain.

Bob Richmond
Samurai Pathologist
Maryville TN

_______________________________________________
Histonet mailing list
Histonet <@t> lists.utsouthwestern.edu
http://lists.utsouthwestern.edu/mailman/listinfo/histonet



More information about the Histonet mailing list