Dried blood spots (DBS) have been around a long, long time - mainly for collecting and storing small heel-prick blood samples from newborn babies to test for inborn errors of metabolism.  Today that testing is typically done using mass spectrometry to look at a series of metabolites whose levels would be affected by absence of key enzymes (i.e., the inborn errors).

It turns out, happily, that drying a blood sample on filter paper (the original DBS "Guthrie" card format) is also an excellent way of collecting and preserving blood proteins.  In fact cards collected from newborns in Denmark have been saved for decades and the proteins shown to be analyzable, with little to no degradation.  This is in contrast to the normal clinical methods of blood sample collection that involve a phlebotomy nurse, needles and blood tubes, and long term sample storage in which aliquots of liquid serum, plasma or whole blood must be stored at -80C or below to preserve protein integrity.  The differences in cost and user convenience between these approaches is huge.

The conventional methods of blood sample collection tend to limit biomarker research to one or at most a few samples per subject, and thus focus on comparing study groups (e.g., healthy and sick) that are composed of different people.  Looking through the scientific literature it is very striking how few studies there are that track biomarker levels in the same person using an extended set of samples collected over time (so-called "longitudinal "studies).  The reason given is usually that collecting many conventional samples from a person involves too much cost (for the collection, for subsequent storage, and for the biomarker testing - so the grant doesn't get funded) and too much inconvenience (patients don't like getting jabbed repeatedly, so compliance is low). 

DBS sample collection opens up a new paradigm for research and clinical use of protein biomarkers: by allowing us to collect and preserve a blood sample wherever and whenever we want, they give us access to the domain of time.  In the first place, an extended history of samples collected and analyzed (a "biomarker history") would allow us to interpret changes in each biomarker against a personal baseline derived from that person's previous results - i.e., truly personalized diagnostics.  Because most biomarker proteins in blood vary less in a normal person from day to day than they do between people, a small change in the amount of a biomarker protein can be quite significant for the individual and still not large enough to fall outside the population's "normal range".

When I started thinking about this some years ago I decided to try practicing what I preach, and collect my own DBS samples.  This turned out to be easy to do using little disposable lancets (Medlance plus Extra by Strefa are my favorites) to prick my finger and Whatman 903 DBS cards to collect 5 drops of blood.  I'm a good bleeder and always get 5 nice drops out of the end of my little finger.  I keep the DBS samples in a Ziploc bag in the refrigerator with some desiccant to keep them dry.  All these supplies are freely available for purchase on the web, and it's easy to do the whole thing at home.

So I started DBS collection in August of 2008 and have just collected my 300th sample.  As a result, I think I may have the richest history of blood samples available anywhere, which allows me (using SISCAPA assay panels) to look very carefully at normal variation over time, and at specific things that have happened - including pneumonia, a series of "colds", onset and successful treatment of mild hypertension, and a gradual rise in hemoglobin A1c suggestive of the beginnings of diabetes.  Not everything we're interested in can be measured in DBS because of the limited sample volume, but a surprisingly rich set of proteins with known clinical significance can be measured at very modest cost (the SISCAPA story).

Whenever I get lazy I just remind myself that I'm never going to be able to go back and create these later.  How often do I collect samples?  During healthy periods maybe once a month or once a week - however there are times when I've taken samples every day to look at something that I think may cause interesting changes.  And when "interesting" things have happened over these 9 years, I have a set of samples to pin down biomarker changes that are directly correlated in time with the disease process involved.  N of 1 may make statisticians wince, but when that 1 is you, and you have been paying attention to other signs and symptoms (i.e.,  contextual data) needed to make sense of the biomarker history, you can help discover diagnostic (and ultimately treatment) possibilities that may never emerge from the conventional comparisons of disease vs healthy group.

Sometimes we need to bleed for progress...