By now we’ve all probably heard of LDTs (laboratory developed tests) and the FDA’s decision to begin exercising regulatory authority over them in the near future. Unless you work in a large reference lab, a molecular lab, or in an esoteric subsection of the lab you probably thought this had nothing to do with you. Well think again. What about those body fluid assays we all run? How many of those are FDA cleared/approved? Well, the ugly truth is, not many.
The body fluids we are often asked to analyze run the gamut from natural fluids (pleural, peritoneal, CSF) to transudates (an extravascular fluid resulting from increased fluid pressures or diminished colloid oncotic forces in the plasma) to exudates (fluid that filters from the circulatory system into lesions or areas of inflammation) and dialysates (fluid passing through a membrane in dialysis, in peritoneal dialysis, the fluid infused and later removed from the peritoneal cavity). Each of these fluid types has a unique composition. Thus, assays developed for serum or plasma may not be absolutely accurate when used for these alternate matrices.
In 2006 and 2007 the CLSI issued guidelines H56-A and C49-A to provide guidance to laboratories on the analysis of body fluids The guidelines provide detailed descriptions of the justification and medical necessity for these measures in various fluids but are relatively mute on what is acceptable assay performance (precision, trueness, etc) for these analytes. Instead C-49-A states that the two key factors in body fluid analysis are the relative accuracy of the method compared to measurements of the same analyte in serum/plasma/or urine. It also emphasizes the fact that reference intervals are not often available for body fluid analytes so that any interpretation of results must be done in the context of the pathophysiology of the analyte and the clinical setting.
Is the above guidance adequate given the FDA’s revised stand on LDTs? Certainly, these are low volume tests and the cost to a manufacturer to validate an assay in a new matrix is not trivial. Additionally, reference range studies are always difficult to do but we have found mechanisms in the past for difficult populations/samples to establish such ranges (perhaps like the one currently ongoing to establish pediatric reference ranges).
Coming changes in healthcare delivery will likely place clinical chemists in a pivotal position as guardians of appropriate laboratory utilization and evidence based laboratory practices. In today’s climate can we really justify sending out reports to our clinicians with the disclaimer: "This test is FDA cleared, but is not labeled for use with body fluid specimens?"