Thursday, April 27, 2017
Can we improve clinical histology assays with PRM?
While frantically looking for material for yesterday's post (that is published and I can talk about it!) I found something new and super interesting on a similar topic! It is in press at MCP here!
This group grabs a bunch of FFPE tumor samples and evaluates whether they can use PRM on a Q Exactive to make the same diagnoses that the clinically approved assay (Immunohistochemistry by trained pathologist) can on the same sample.
This falls right into the topic of an ongoing conversation between some experts on Twitter this week. After I described some of the rigorous QC/QA protocols in the JHH Clinic -- there were a few responses similar to this one from Dr. Eyers...
...who is of course right! Pathology assays, especially histology ones are still done by trained pathologists looking at one sample at a time through a microscope and grading protein changes by colorimetric shifts (by eye). These are extensively trained medical technologists with ASCP certifications, and more commonly by MDs who specialized for years in these areas. These manual assays are great -- but they are slow -- expensive -- and yeah...rigorous QC/QA is tough. Stay tuned -- everyone is working on these things -- this paper from Vanderbilt is proof!!
It is about tumors and the tests that find out what drugs someone is being treated with. This excerpt is a great summary:
The two immunotherapeutics mentioned will target tumors that are PD-L1 positive -- but the approved IHC assays (important to note there are more than one) have been reported to be problematic, possibly because of false positives and false negatives. [Important note: they are tumors, there could be other mutations that cause the drugs not to work, of course, and this paper explores several of those!] But if we can create an assay that is as good as the IHC assay, but faster or cheaper -- this is a win for the medical system and patients. If we can create one that is better -- either or, especially, as well -- then this is a no-brainer!! Heck, even if we can just establish a complementary assay that will improve on the numbers mentioned above -- this is could have an immediate and measurable impact on patients!
This is what Vanderbilt sets out to do here. They go specifically after PD-L1 (and related protein) peptides. They create heavy peptide standards for the markers they are interested in. And they build PRM assays.
How's it go?
(Chris Pratt is still funny, right? First GIF that popped up when I typed "AWESOME")
1) They did the IHC method as well. This is important. They appear to have found some false positive with the staining. I'm not equipped mentally to evaluate that part. But I'm glad they have this as a comparison
2) The PRM specifically focused on the IHC assay -- match very well
3) The mass spec method (which is interesting, it is basically a low-high instrument method on the QE) finds some important information that is not available in the normal assay.
3.5) Wow. That GIF is annoying a few hundred times through it. Hopefully as a reader you don't have to see it as many times as I do while writing it...
Elaborate on 3 -- The IHC assay is good for determining if PD-L1 is on the surface or not. It is not, however, very good quantifying how much is there. The intensity of the staining signal does not match up with the absolute quantification.
Also -- there are other proteins here at the surface that are closely related and can interfere with the IHC assay -- or -- the immunotherapeutics. They look at this more closely and find PD-1 and PD-L2 that can easily be assayed by PRM, but not with the IHC assay.
They go further here and look at the effect that glycoslylation of these proteins plays. If you're trying to bind these proteins with an antibody for diagnosis or treatment glycoslylation may mess this up. And -- there are a bunch of varied glycoforms...
Conclusion: The IHC assay is here and valuable and clinically certified and has helped a ton of people -- but mass spectrometry, right now, has the ability to add extra information to these diagnoses and complement the IHC assay.
Important point: The LC-MS described here is 2D separation and 60 minute gradients on fractions. This isn't an experiment that is ready for the clinic. This is an exploratory assay. But...the promise is definitely here for an LC-MS assay to fill in the gaps and make an impact on patients! This paper is an awesome hint of what the future could hold!