|
|
|
|
|
|
|
|
Making Personalised Medicine Work
actualising real opportunties in healthcare delivery |
InterAction Meeting Session, held in Philadelphia, USA, 12 October 2005
Personalized Medicine: Safety and Efficacy Concerns Discussed as a Case-Study
chaired by Donna Mendrick (Gene Logic) | |
Presenters & Discussions Leaders:
Sydney Gilman, PhD (Gene Logic)
Matt Cooper, PhD (Biogen Idec)
Derek Janszen, PhD (Wyeth)
Steven Arnold, MD, MS, CPE (Queens-Long Island Medical Group)
Donna Mendrick, PhD (Gene Logic)
Background
While experts in the field have long realized the need for improved patient monitoring and better predictability of drug-induced adverse events, the public’s attention had been focused more recently on this issue due to the high profile coverage of Vioxx. “Because patients are taking a more active role in their healthcare” the FDA has recently created a New Drug Safety Initiative ( http://www.fda.gov/cder/drugSafety.htm ) “to make drug safety information available …. in an easily accessible format.” This proactive stance is not unique as the FDA has been in encouraging pharmaceutical companies to utilize new technologies such as pharmacogenomics to advance drug discovery and development. Pharmacogenomics (covering efficacy and safety in both nonclinical and clinical applications) was recognized as a key FDA initiative in 2001 and, working with industry groups, the FDA has co-sponsored multiple workshops dealing with the use of such science and its impact on patient care. Whole genome profiling details responses of greater than 33,000 genes in a single experiment, representing one of the most powerful and comprehensive assays available. This workshop will discuss how this powerful technology can be used to predict a response to a drug throughout all phases of drug development thus leading to personalized medicines.
Case Study
Drug X has been approved for severe, debilitating migraines and is far more effective than other existing therapies. However, upon post market use it has been observed that a small percentage of patients (1%) develop elevated serum transaminase levels and, if maintained on this treatment, develop fulminant liver failure leading to death and/or the need for liver transplantation. Additionally, it has been recognized that this drug is more efficacious in some patients than others.
Interactive Discussion Points
Patient response reporting
How might patient monitoring and adverse event reporting be improved to detect such effects when limited to such a small subset of patients? How would varying efficacious responses to a drug be noticed and reported?
Point of care assistance
Once a varying response in terms of safety and efficacy are noted, the drug maker wishes to employ pharmacogenomic technology to identify the patients who will respond favorably to their drug. Since tissue biopsy is not feasible in this case, they choose to determine if gene expression studied performed with circulating white blood cells can identify such patients. Given patients’ concerns regarding any type of “genetic” testing, how could point of care clinicians assist in convincing patients to allow blood collection for such studies?
Patient population for clinical trial
The drug manufacturer establishes a clinical trial to identify pharmacogenomic markers in blood. To monitor efficacy it would be required to use patients that suffer from migraines. Given the low adverse event rate (1%), would it be appropriate to use normal patients for identifying markers of liver toxicity?
Regulatory acceptance of test
The company has identified 50 genes whose differential expression will stratify patients in terms of safety and efficacy. What would be required by the FDA to validate this set of markers particularly if some of the markers are EST (expressed sequence tags) of unknown biological function? Would testing of patients be mandatory prior to prescribing this drug? How would the drug manufacturer co-develop a testing kit in concert with the drug?
Testing of second generation compounds
Companies want to pursue second generation compounds in the same basic pharmacologic class and thus want to extrapolate the markers identified in humans to nonclinical species (i.e., bridging biomarkers) for earlier screening. How might one approach this and would the methods differ when examining safety and efficacy? Assuming a set of genes are identified, could this screen be acceptable as a surrogate for patient stratification or will all compounds of this class be required to undergo expensive, elaborate pharmacogenomic testing in humans? If the former is true, what would be required to validate this set of genes in the nonclinical species? |
|
|
|
|
|
|
|
