From the bedside to the bench, and back?

1 Sep 2008

Pharmacogenomics, the rescue of failed drugs and phase 0 trials:


Pharmacogenomics, or PGx, has been defined as “the individualisation of drug therapy through medication selection or dose adjustment based upon direct (e.g. genotyping) or indirect (e.g. phenotyping) assessment of a person’s genetic constitution for drug response”. [1] This broad definition takes pharmacogenomics to be a synonym for pharmacogenetics, and includes tests operating at protein, metabolite, or other biomarker levels whenever these factors are affected by genetic variation. In recent years, commercially available PGx tests have been approved by FDA, such as tests to detect variations in the genes coding for enzymes involved in drug metabolism [1]


Notwithstanding, the application of PGx tests in patient care remains very limited, mainly because the reaction of the pharmaceutical industry has been reserved due to the potential for marked segmentation.[2] But PGx need not be financially unattractive from a drug manufacturer’s point of view, as the potentially smaller marker for a PGx drug could be compensated by an increased rate of adoption of the drugs, the identification of responsive patients, an increased compliance with improved efficacy, and the possibility of premium pricing.[2]


Indeed, the use of a PGx test has played a pivotal role in rescuing the drug Iressa from failure in 2003.[3] The drug, a monoclonal antibody against EGFR, proved effective only in a small percentage of patients with non small cell lung cancer, and would have failed if in 2004 results were not published that showed that tumours that responded to Iressa harboured mutations in EGFR that rendered the protein more sensitive to the drug. As a result of these finding, researchers developed genetic tests for EGFR mutations that allow that identification of responsive patients.[3]


Few failed drugs receive so much scrutiny. Iressa was probably the first example of a drug specifically designed to hit a target, and therefore effective only in a subset of patients that have that specific target. The case of Iressa is also one of the first instances of bedside to bench research, where human clinical trials can inspire new questions back at the bench that will, when answered, improve the human experiment in its next iteration. This type of reverse translation was advocated by Sydney Brenner in April this year, at the American association for cancer research meeting in San Diego.[3] Brenner won the Nobel prize in 2002 for his pioneer studies on the use of the worm Caenorhabditis elegans as a model organism, but in his talk in April 2008 he delivered a different, provocative message, by saying that “We don’t have to look at model organisms anymore, because we are the  model organism”. While his statement may look a bit extreme, the rationale that “the best model for human is human” is the assumption under the FDA new industry guidelines for early exploratory drug studies in humans, also called Phase 0 studies, issued in January 2006.[4]


According to the guidelines, the shift away from the use of nonspecific cytotoxic chemotherapeutic agents in cancer therapy to more specific, molecularly targeted agents necessitates a re-evaluation of the cancer-drug development process. Current strategies for drug development are still predominantly based on the assumption that the investigational agent has a dose-toxicity relationship, and that efficacy is somewhat related to toxicity.  However, these assumptions may not be valid for molecularly targeted agents such as monoclonal antibodies. For such new drugs there is need to alter the traditional drug-development sequence, where exploratory phase 0 trials should be designed that focus on extensive agent characterisation and target-assay development, including molecular imaging studies, in a limited number of patients, who will each be exposed to a limited number of doses of the study agent.[4,5] Such trials, without therapeutic intent, would aim at eliminating therapeutic failures early in the cancer drug development process, therefore speeding up the whole process.[5]


With the advent of Phase 0 trials, the cycle of translation research aims to be completed, as research from the bedside to the bench will complement the traditional research from the bench to the bedside. Whether it is true that we don’t need model organisms anymore, as Brenner advocates, remains an open question.




1.       Swen JJ, Huizinga TW, Gelderblom H et al. Translating pharmacogenomics: challenges on the road to the clinic, PLOS Medicine 2007;4(8):e209.

2.       Garrison LP, Rick JC, Carlson JJ et al. A review of public policy issues in promoting the development and commercialization of pharmacogenomic applications: challenges and implications. Drug Metab Rev 2008;40(2):377-401.

3.       Ledford H, Translation research: the full cycle, Nature 2008, 453:843-45.

4.       FDA 2006 Guidance for Industry, Investigators, and Reviewers
Exploratory IND Studies, available at  (retrieved September 1, 2008)

5.       Kummar S, Kinders R, Rubinstein L et al.  Compressing drug development timelines in oncology using phase '0' trials. Nature Reviews Cancer 2007,7:131-139.