Patient Outcomes Linked To Biomarker Levels

JonN writes to tell us Science Daily is reporting that researchers at Yale University have discovered that current pathology methods for biomarker detection can be dramatically altered depending on the concentration of antibodies used. From the article: "Biomarkers may have the power to provide diagnostic, therapeutic, and prognostic information for personalized medicine." said Donald Earl Henson, M.D., of the George Washington University Cancer Institute, in "Back to the Drawing Board on Immunohistochemistry and Predictive Factors," an accompanying editorial. "However, immunohistochemistry, a popular technique for evaluating biomarker expression, may contain procedural flaws that jeopardize its promise."

Comments

Old news

[pugdk]

I work with stuff like this on a daily basis (post doc level) and everyone in biomedical research knows to be highly suspicious of everything that has to do with immunohistochemistry and immunofluorescence. It all comes down to your negative and positive controls - as you can get a very strong signal with more antibody / less washing / longer incubation on a negative control and only a slight signal on a positive sample with less antibody / more washing / shorter incubation....

When reading research papers containing these images / results you need to trust the research team doing it... its so easy to falsify and way too easy to misread if you mess up your experiment slightly. Also different protocols in different labs will give different results.

But yeah, lets put obvious and well known stuff on the frontpage of /. :-)


-pug

using IHC for quantitation is idiotic

[pinche cabron]

And I speak as someone who has developed and extensively tested an immunoassay (ELISA) for a cancer biomarker. IHC is tissue staining (immunohistochemistry), as opposed to ICC (immunocytochemistry, cell staining), and neither of those methods is worth a damn for quantifying protein levels. You would have to homogenize the sample (sonicate it, for example), then do an ELISA with a standard curve. And no, the standard curve doesn't have to be linear, but you do have to fit it to some suitable sigmoidal (S-shaped) model.

Then there's the problem of quenching. If the protein level is super high, my fluorescent signal will be too high and the dye molecules stack on top of each other, causing quenching (loss of fluorescent signal). The way around that problem is a dilution series (two would be adequate) for each test sample. Quenching is the main reason IHC and ICC suck for quantitation, too. That, and photobleaching, and other microscope artifacts. Anyone with any experience reserves IHC and ICC for qualitative information.

They mentioned the Yale scientists looked at tissue microarrays, which should not be the standard test. That technology is in its infancy, and most of its successes are either exaggerations or outright lies. Again, I'm speaking as someone in-the-know. I've seen the shiny Powerpoint presentations, and I've seen the shoddy data behind the scenes that they didn't show in the presentation. High-throughput automated microscopy? Hah. Not for another decade will that work as advertised.

And another thing: standardization of the antibody is not an issue as long as the off-rate is slow enough, and the same antibody is used for test samples and standards. I've had more than adequate experience in this arena as well, using Biacore Surface Plasmon Resonance to measure antibody on- and off-rates.

Just goes to show you, never send a doctor to do the job of a molecular biologist or biochemist.