LA JOLLA, CALIFORNIA – Back in June 2000, when the draft human-genome sequence was announced, US President Bill Clinton proclaimed, “It will revolutionize the diagnosis, prevention, and treatment of most, if not all, human diseases.” A decade later, hope has given way to disappointment, reflected in headlines like “Gene Map Yields Few New Cures.”
But pessimism about the potential of human-genome research to yield medical breakthroughs has arisen from unrealistic expectations. Indeed, while “silver bullets” that can cure our most feared diseases have not been found, progress in the area of gene-drug interactions, known as pharmacogenomics, has been extraordinary.
The ability to determine the principal genes that account for our variable response to prescription drugs has been advanced by a technique known as a genome-wide association study (GWAS). The whole human genome has approximately six billion bases, but a window into its composition can be probed using approximately one million bases (0.01% of the genome) via a gene chip. The bases on the chip are selected because they are informative, tagging bins of the genome, like a post-code directory. Using GWAS methodology, we have learned the biological basis for responses to many drugs – both their effectiveness and important side-effects.
Examples of this progress in the past couple of years are plentiful, and include the statins, Plavix, interferon, warfarin, and the antibiotic flucloxacillin. The main side-effect of statins, which lower cholesterol in the blood, is severe muscle inflammation, and it can now be predicted with a simple genotype test, as can the response to Plavix, the second most commonly prescribed drug after statins. Those individuals who carry at least one copy of the gene variant that does not allow the body to metabolize Plavix have a 300% higher risk of clotting a stent.
This variation of the genome is exceptionally common, present in 30% of individuals of European ancestry and more than 50% of those of Asian ancestry. In many such patients, doubling the dose of Plavix, and checking the platelet function with a point-of-care test to assure suppression, can override the lack of response. There are alternative drugs, such as prasugrel or ticagrelor, which can be used instead of Plavix to bypass the problem.
The story of interferon, given for one year to patients with hepatitis C virus, is particularly striking. This treatment is very costly (about $50,000) and makes all patients feel quite ill with flu-like symptoms and general malaise. But the drug works in only half of the people treated. We now know that a simple genotype sorts out who will or will not benefit from interferon, and there are many new drugs in development for those who will not.
The list goes on. For warfarin, a widely prescribed drug used to prevent blood clots, genotyping can guide the right dose and hasten the time it takes to get to steady state. Genotyping can also predict severe liver toxicity for drugs such as flucloxacillin or lumiricoxib.
To be sure, our predictive ability is far from complete. We know only common gene variants from the GWAS approach. Furthermore, most drugs have not even been studied yet, so there is a long way to go fill in the holes.
Nevertheless, substantive and remarkable progress has been made, all in the last few years. Moreover, someday we will know – via sequencing of the whole genome – about rarer genomic variations that are correlated with treatment response or side-effects.
This has paved the way for next-generation pharmacies. Genotyping can now be accomplished in 20 minutes, and over time will get even faster. To fill a prescription for a drug with a known pharmacogenomic profile, a customer can get rapid genotyping to determine appropriate dose, drug, or predilection for serious side-effects. Or even better, many people will submit a saliva sample to a consumer genomics company, which will analyze all of their pharmacogenomic data and perform an extensive panel of genotypes, updated every month, and store the data on their smart phones. For mail-order prescriptions, such genomic data would be a routine part of the customer database.
In the United States, pharmacy benefit managers (PBM) handle prescriptions for almost all large employers and account for more than 200 million individuals. Two of the largest PBMs, Medco and CVS/Caremark, have announced plans to conduct large-scale genotyping for many drugs. Their motives include more efficient use of prescription drugs, along with getting an edge on other PBM competitors. With annual expenditures for prescription drugs in the US totaling $300 billion, there is certainly room to cut costs.
Unfortunately, the medical community is resisting the use of pharmacogenomic data in clinical practice, despite regulatory authorities’ recommendations for many drugs. PBMs and consumer genomic companies can be a positive force by facilitating or advancing the use of such data.
The ultimate goal of pharmacogenomics is to provide the right drug, at the right dose, for the right individual, without any significant side-effects. Despite widespread misperceptions about the practical impact of genomic research, the science has hit its stride, and we need to capitalize on this momentum if we are to realize the opportunity of individualized medicine. Next-generation pharmacies represent a promising step toward that goal.