Another look at a recent study of proteins in our blood that predict response to anti-PD-1!!!

A couple of days ago I posted this:  A simple blood draw to measure serum protein - predicts outcome for melanoma patients treated with anti-PD-1????

It is a review of the article you can find here and peruse yourself:  A Serum Protein Signature Associated with Outcome After Anti-PD1 Therapy in Metastatic Melanoma Jeffrey S. Weber, Mario Sznol, Ryan J. Sullivan, Shauna Blackmon, Genevieve Boland, Harriet M Kluger, Ruth Halaban, Antonella Bacchiocchi, Paolo A Ascierto, Marilena Capone, Carlos Oliveira, Krista Meyer, Julia Grigorieva, Senait G Asmellash, Joanna Roder and Heinrich Roder Cancer Immunol Res December 8 2017

Some of my peeps had a few more questions, and I found the whole shmegegge super interesting, (Especially when you think about it in light of all the other potential that a simple blood analysis has regarding tumor burden and other biomarkers:  Measuring cell-free DNA in melanoma patient's blood to determine tumor burden and prognosis!) so here is a more thoughtful analysis than my first.....

After a thorough review of the entire article, along with all the tables and figures with my trusty B....here's what we got:

Dr. Weber made an earlier presentation of some of this data at a conference in 2015...so it's been cooking a while.

The main gist is pretty straight forward: Researchers used pooled data from multiple institutions and several studies of ratties (n = 289) who took nivo, pembro, ipi, or the ipi/nivo combo (along with those of us who took nivo with peptide vaccines). Some of the ratties had already had ipi, before taking anti-PD-1 and some had not. Blood was drawn before the ratties started treatment. Those samples were analyzed using advanced techniques to tease out what proteins in the blood were present in the sensitive set (patients who responded) vs the resistant set (patients who did not). After that first examination was completed, researchers used the same protein markers on test sets. The protein markers were successful in distinguishing those likely to be sensitive (responders) from those who were likely to be resistant (non-responders). In addition, they analyzed the proteins to identify the mechanisms of sensitivity or resistance. There they found proteins associated with: 

1. wound healing

2. active phase reactants (the body's inflammatory response)

3. complement activation (another immune mechanism we use to respond to abnormal molecules – which could be anything from worms to cancer).

This test clearly worked to predict response for folks taking anti-PD-1 alone or taking ipi alone. However, when ipi/nivo was taken together the proteins were not predictive of response. (As an aside if you look at the graphs...Sets 2 and 3 are not currently significant, probably because the length of time passed at review of data was not adequate.)  You can take a look at the graphs in the link above or my pic below.

Researchers repeatedly note that application of this test to a larger group is needed to validate and further refine these results. Still – I am hopeful that knowledge of these proteins will provide guidance for future therapeutic interventions. For instance, if we could enhance wound healing while simultaneously decreasing inflammatory response and complement activation, we might be able to make anti-PD-1 therapy more effective. However, the current test as is - cause remember, these samples were taken BEFORE any of these ratties got any anti-PD-1 therapy - could refine prediction of who would benefit from current anti-PD-1 treatment and who may not.  For instance, if your blood samples showed proteins indicating that you were likely to respond to anti-PD-1 alone, especially if we were to examine that same blood draw for circulating tumor DNA and other biomarkers, it may help your decision making process when considering anti-PD-1 therapy vs the ipi/nivo combo with its increased rate of potential side effects.

Come on, researchers.  Let's get these tests out there!!!  Here's to the ratties! love, c