The FDA has gone public with its thinking on in vitro diagnostics with multiplexed devices (including microarrays, proteinchips, etc.) in this recently issued draft document. It essentially establishes the protocols required for submission of a marketing application for a diagnostic use involving these devices:
fda.gov
Being a draft document, there is a period of 90 days for public comment. It was released about month ago.
The passage that especially struck me:
>>The measurement of expression changes, whether RNA or protein, will raise different validation and safety and effectiveness questions than the measurement of DNA changes or variations.
"Genetic" tests: DNA differences are fixed, whether germinal or somatic. Results from these tests can generally be described as dichotomous (either present or not present), trichotomous (homozygous A, heterozygous, homozygous B), or categorized (e.g., haplotypes). Interpretation of tests designed to measure these types of differences will be, in most cases, straightforward. DNA array tests nevertheless should be carefully designed and highly reproducible, and have well-established performance. Clinical studies should account for disease prevalences in the populations studied.
Tests measuring expression changes: Expression changes, in contrast to DNA changes, can be responses to a variety of factors. These may include simple individual-to-individual differences, time of day, and specific effect of a therapeutic treatment on a tissue. Results can vary markedly as a result of these factors. Tests to diagnose, predict, or select based on expression patterns may consequently be difficult to interpret. Sponsors of these tests should consider array physical design strategies, quality control (QC), reproducibility and readout/interpretation. <<
Cheers, Tuck |
