In this new era of using biomarkers to determine the level of responses to a new therapy, or to determine the progression of disease, the role of monitoring antibodies is becoming a major part of new clinical trials. However, antibodies are not as straightforward to work with as one might think (see http://f1000research.com/articles/3-232/v2). Recent statistics suggests that only 11% of findings in cancer research were deemed reproducible (http://dx.doi.org/10.1038/483531a). There is a strong sense that the complexity of working with antibodies plays a pivotal role in the (lack of) robustness of (pre-)clinical work (http://f1000research.com/articles/2-153/v2). Can any failed trial be attributed to a failing antibody / failing immune assay?
When it comes to antibodies, Biochemistry and physiology are much more complicated than we ordinary think. Things that must be considered are: nature of the antibody (like IGG / IGM), title of antibody (which is roughly an expression of how mich dilution it will be necessary in order for it to be inactive in vitro), immuniti of men/animal producing it, if/when/how it is a primary response, or it is the response of a second/third exposure to the antigen, if/how/when it was produced in response to another (similar) antigen, its specificity, its affinity (which is expressed as how strongly it binds/what kind of reactions may be needed for it to be unbound.
All this jus to name a few of the most important characteristics of antibodies, and ignoring many related physiologic functions.
When it comes to successful use of antibodies in clinical trials, I have no Knowledge of papers which address the question. I would suggest you to carry out a search of papers dealin with the question in a way that contemplates detailed molecular analysys. It is a very good theme for a Meta-Analysis. I might even cooperate with you on that, if not for other reasons, just for the fun of it! But I really think you´ll find many paprs that deep. Soo indeed it is a theme for molecular / Physology research. Lots of phenomena could be explained in such a study.
It is well said that only 11% of the findings are reproducible. This is what I have learned form my teacher so 30 uears ago. Resrachers tend to think there is always a direct answer to any biological phenomenon so that the research can be conducted reasonbly straightforward. However, biologic system finds multiple way to cope with a specific change. I do believe this is also apply to the immune system at least this can explain why somebody will develop autoimmune disorder but most of us will not. The mutitude of control loop is the beauty of the God's design.
There was an article about this exact topic in Lab Times a few months back: http://www.labtimes-archiv.de/epaper/LT_14_01/files/assets/common/downloads/LabTimes_2014_01.pdf
See page 39 "Can you really trust your antibody? Quality under fire". It describes a case where a monoclonal antibody ERCC1 – “clone 8F1” apparently changed after it was determined that this antibody could predict cisplatin sensitivity of NSCLC.
In reality this might be a fairly isolated case, since monoclonals are fairly stable. In general if you are going to launch a clinical trial, you are going to make certain that you have enough antibody from the correct lot that you've validated yourself prior to starting the trial. If other people start a clinical trial assuming they can buy just any old antibody against the same protein, and they fail, that would be a case of severely untrained people doing something they are not qualified to do.
Interesting discussion. We routinely work under contract service for Biotech-Pharma companies who have new therapeutic monoclonal antibodies in clinical trials (most often Phase I & II). They need to get informations about Pharmaco-Dynamy of the new (MAb-based) drug under trial. When the target (or a surrogate marker) is a blood cell-surface antigen, we do help them answering questions such as: a) how many target epitopes are on such or such cell subset (e.g. CD33 on AML blast cells but also normal monocytes and PMN, ref 1) ? b) Does the infused MAb get to reach the targets in peripheral blood, in bone-marrow ? c) Is the target epitope saturated or only partially covered at such dose level (aka Receptor Occupancy) ? d) Is tthe target Ag modulated ? etc... This takes opportunity of Ag quantitation approaches using QFCM (Quantitative Flow CytoMetry), where pre-cited questions regarding MAb affinity, saturating conditions, binding/competition kinetics, epitope mapping etc have to be well known. And no doubt that big pharma biotech companies are not injecting in patients poorly defined material ! ... and that results need to be reproducible and reliable. The amount of information/qualification of the material to be injected into patients is impressive. These are absolutely NOT "severely untrained people doing something they are not qualified to do".
ref 1) Lapusan et al, Invest New Drugs, 30: 1121-31 (2012)
Philippe, the antibodies on scrutiny are the ones used for companion diagnostic purposes, not as therapeutic. Therapeutic antibodies undergo a complete different set of validation than those to monitor patients to follow stage of disease, or to see if the patient shows any signs of collateral damage. And then there are concluding results obtained from those antibodies used during pre-clinical research based on which a clinical trial takes place.
Agree. We also need MAbs for our QFCM-based PD monitoring assays, and we do scrutinize them(select according to affinity, epitope mapping, complementarity etc...) to mount an assay in terms of e.g. receptor occupancy and absolute Ag quantitation.
- Badges
- Science method