672
development process for medical products – the Critical Path – to make product
development more predictable and less costly. The critical path determines the
potential bottlenecks in bringing a product to market. The focus of the Critical Path
Initiative is to identify ways to update the product development infrastructure for
drugs, biologics and devices, and the evaluative tools currently used to assess the
safety and effi cacy of new medical products. Examples of evaluative tools include
the use and verifi cation of pathophysiological and/or descriptive biomarkers for
patient selection for clinical trials and/or use as surrogate endpoints. In addition, an
important example of a scientifi c opportunity for improving the critical path is the
use of pharmacogenomics and pharmacogenetics or, more specifi cally, the identifi -
cation of DNA-based biomarkers or RNA-expression profi les that can provide
insights into the stage of a disease, disease progression, drug response and drug-
dosing requirements, and thereby lead to the development of tests to predict clinical
outcomes more reliably.
FDA Regulation of Multivariate Index Assays
In 2006, the FDA took a step toward regulating a new category of complex genetic
diagnostic tests that are expected to play a growing role in tailoring medical treat-
ments to specifi c patients. The FDA is calling these tests “multivariate index assays
(MIAs).” According to the FDA, such tests require approval before they can be
marketed to ensure that the tests are valid. The new policy, published as draft guide-
lines, is open for public comment and would also be a step toward expanding the
FDA’s oversight to clinical laboratories ( http://www.fda.gov/cdrh/oivd/guidance/
1610.html ). The FDA published a notice of availability of a revised draft guidance,
on “In Vitro Diagnostic Multivariate Index Assays” in the Federal Register of 26
July 2007 (72 FR 41081) and comments were invited. At this time Genentech fi led
a Citizen Petition with the FDA urging the agency to take on greater oversight of
diagnostic tests that are intended to guide therapeutic decisions and to regulate all
laboratory-developed tests. According to Genentech, pharmacogenomic informa-
tion is contained on the label of ~10 % of all FDA-approved drugs. Included among
those are Genentech’s trastuzumab (Herceptin), which requires that patients be
tested for particular genetic characteristics and the results be considered before the
drug is administered.
Currently, tests developed and performed by a single laboratory, known as home-
brew tests, have been generally considered laboratory services and outside FDA
purview. Now the FDA will regulate at least one category of such tests: those that
measure multiple genes, proteins or other pieces of clinical information taken from
a patient and then use an algorithm or software program to analyze the data.
The best known of these tests is Oncotype DX (Genomic Health). It analyzes
the activities of 21 genes in a sample of breast tumor and then computes a score
that is said to be predictive of whether a patient’s cancer will recur and whether she
would benefi t from chemotherapy. While there are only a few such complex tests
on the market now, their number is expected to grow. For personalized medicine, a
22 Regulatory Aspects of Personalized Medicine