Dr. Bill Gibson (biography and disclosure)
Frequently asked questions I’ve noticed
Why isn’t that drug working? Why did that unusual side effect happen?
Most doctors are familiar with the concept that not every patient responds in the same way to the same dose of drug. In paediatrics, drugs are routinely prescribed in dosages that take the patient’s size into account, using milligrams per kilogram of body weight as the unit of measure. Many clinical practice guidelines recommend that drug dosages be adjusted up or down depending on the observed clinical response. Information is now accumulating that sheds light on genetic differences in how people convert medications into their active forms, and also on why some patients develop side effects and others don’t. The study of individual genetic differences in drug metabolizing genes (or groups of genes) has been termed pharmacogenetics. When sequencing of hundreds of drug-metabolizing genes is done in the same person, the term “pharmacogenomics” is more properly applied. Three major areas in which common DNA variants affect drug response are in the pathways that metabolize warfarin, opioids, and chemotherapeutic agents.
Data that answers these questions
Clotting Meds: Warfarin Dosing
How much warfarin someone needs to get to (and stay in) the therapeutic range depends in part on how fast their body converts warfarin into other metabolites, and also on how well warfarin sticks to its target, the VKORC1 protein. VKORC1 is critical for making vitamin K-dependent clotting factors, and if the clotting factors are not made, the International Normalized Ratio (INR) goes up. The empirical warfarin starting dose of 3-5 mg per day assumes the patient has an “average” warfarin metabolism, but the maintenance doses can vary from 1 mg per day to up to 20 mg per day, depending on the presence or absence of common variants in the warfarin-metabolizing genes CYP2C9 and VKORC1. When the patient’s genotypes at these sites are known, this information can be incorporated into selecting the starting dose of warfarin (Johnson et al., 2011).
Pain Meds: Opioid Dosing
The opioid analgesic codeine does not soothe pain by itself – in order to work , it has to be converted by the body into morphine and other metabolites. Eight or nine out of ten people have normal activity of the codeine-to-morphine converter enzyme CYP2D6, but between 1 in 10 to 1 in 20 have an inactive CYP2D6 enzyme, so they get no pain relief from codeine. In these individuals, the codeine sits around in the body in its original, inactive state, until it is excreted. Often called “poor metabolizers,” these individuals should still respond to morphine if it is given directly, but even high doses of codeine won’t help them. Conversely, about 1-2% of the population has extra genes or even hyperactive genes that convert codeine to morphine very quickly. These individuals are at higher risk of toxicity when given regular doses of codeine (Crews et al., 2012). For example, if a mother with the hyperactive CYP2D6 enzyme takes codeine while breastfeeding her baby, she may end up making too much morphine inside herself, secreting it into her breast milk, and sedating her baby through breast-feeding. There have also been documented cases of children dying of morphine toxicity after routine surgical procedures because the children’s bodies made too much morphine from the codeine they were given postoperatively (Kelly et al., 2012).
Epilepsy Meds: Carbamazepine and Dangerous Skin Reactions
It has long been known that some patients who receive carbamazepine (also known by the trade names Tegretol, Equetro and Carbatrol) develop a serious and potentially lethal skin reaction like Stevens-Johnson Syndrome (SJS) or Toxic Epidermal Necrolysis (TEN). Recent studies show that people who have a particular immune system gene called HLA-B*1502 have a much higher risk of developing this complication. This is a normal version of the gene that only seems to cause difficulties in the presence of carbamazepine. The absolute risk of SJS or TEN in someone who has the HLA-B*1502 allele appears to be roughly 5% (see www.fda.gov). Current advice is that all individuals of Asian ancestry, including people of South Asian/East Indian ancestry, who are about to receive carbamazepine for any indication should be tested for the HLA-B*1502 allele before getting the drug. Why test people of Asian ancestry? This version of the gene is relatively common in Asia, especially the Indian subcontinent, Southeast Asia, Malaysia, Taiwan and the Philippines, but the allele is rare in other populations (Europeans, Africans, etc.). Don’t forget that individual genes are not “diluted” by intermarriage, so it’s prudent to test people who are of mixed ancestry as well, if one of their recent ancestors is of Asian descent.
Cancer Meds: Chemotherapy Agent Dosing and Selection
Chemotherapy for aggressive cancers is always a balancing act, whereby dangerous drugs are given deliberately in the hope that they will be more dangerous to the cancer cells than the cancer cells are to the patient. Recent studies of the genes that metabolize chemotherapy drugs have identified new genes and pathways whereby some children are at a much higher risk than average of side effects (such as hearing loss) from specific chemotherapeutic drugs. It may be possible to adjust the dosage to avoid these side effects if the children are tested for genetic susceptibility before the chemotherapy drugs are given. In some cases, an alternate chemotherapeutic drug may also be chosen. Given that side effects like hearing loss are lifelong, their effect on quality of life and health care costs is tremendous: economic analysis suggests that genetic testing to prevent chemotherapy side effects would be highly cost-effective (Dionne et al., 2011).
In other situations, the entire genome of a patient’s cancer cells has been compared to their original, non-cancerous genome to find oncogenes specific to the cancer cells. Very rarely, this analysis has unearthed oncogenes that are usually seen in a completely different type of cancer, and has let the doctors make headway by “repositioning” chemotherapy drugs normally used only for other cancer types (Li et al., 2012).
Use the information, but first be sure you can trust it
Most private medical labs and local hospital labs do not yet test for genetic variants in drug metabolizing genes, though there are labs that offer pharmacogenetic testing on a direct-to-consumer or physician-ordered basis. In specific situations where the genetic testing will directly affect patient management, patients resident in British Columbia may have such testing approved for coverage by the Medical Services Plan. It is worth noting that, if clinical decision-making is going to be affected by a test, that test should be collected in and processed by a lab that is certified under the guidelines of the College of American Pathologists (CAP) and/or the US Clinical Laboratory Improvement Amendments (CLIA). If patients show up to their physician’s office with a report in hand citing a positive test from an uncertified lab, the best practice is to have that result confirmed in a CAP- and/or CLIA-certified lab prior to incorporating it into clinical care.
As soon as evidence-based clinical practice guidelines for pharmacogenetic testing are ratified (Amstutz and Carleton, 2011), practitioners will be able to take some of the guesswork out of dose calculation for warfarin and other drugs. The predictive value for warfarin dosing will likely never be perfect, because there always exists the possibility of rare DNA mutations in CYP2C9, VKORC1 or other genes that would disrupt the drug metabolism pathway. In the meantime, if a doctor is having difficulty finding a stable dose of warfarin for a patient, they should be aware that pharmacogenetic testing done and interpreted by a properly certified lab may provide useful information.
For codeine pharmacogenetics, because 1-2% of the population have superfast codeine metabolism genes, a practitioner prescribing codeine should always be on the lookout for toxicity, and be aware that signs of opioid overdose in someone taking codeine do not necessarily mean that they took extra pills – their own bodies may be making extra morphine from what would otherwise be a normal dose of codeine.
In the field of cancer pharmacogenomics, patients who search the internet to find new genetic tests or treatments that may improve their prognosis routinely find many websites offering pharmacogenetic or other tests. Most of these sites really offer more hype than hope. For my own patients who seek access to experimental tests and therapies, I routinely recommend that preference be given to those that are offered through trials that are formally registered with the international database of clinical trials (http://clinicaltrials.gov). Though success is not guaranteed, these trials are typically better designed and more rigorously monitored than those that don’t seek registration there. Clinical trials in specialist centers will likely remain the gatekeepers with respect to cancer pharmacogenomics for some time to come.
How to Access These Tests in Primary Care
Testing for CYP2C9 does not appear to be widely available, though it may be accessible through specific research protocols. Sequencing of the VKORC1 gene is offered by several labs listed in the GeneTests Lab directory, though none of these are in BC.
Testing for the HLA-B*1502 allele is currently available in BC at no cost to patients. Any physician who can order tests in BC can order this one by writing “HLA-B*1502 typing prior to carbamazepine therapy. Attention: “VGH Immunology lab” as the indication on the requisition. The lab of Dr. Paul Keown will perform the testing and issue the result.
The Pharmaceutical Outcomes and Policy Innovations lab of Dr. Bruce Carleton, based at BC Children’s Hospital and the Child and Family Research Institute, is available to assist family physicians and other practitioners who wish to access testing for genes involved in opioid metabolism and sensitivity to chemotherapeutic agents. Professor Carleton’s team is working on national guidelines for these types of tests.
Useful websites to find out more:
- GeneTests Website: http://www.ncbi.nlm.nih.gov/sites/GeneTests/?db=GeneTests
- International Database of Clinical Trials Registered with the US FDA http://clinicaltrials.gov
- College of American Pathologists Laboratory Accreditation Website: http://www.cap.org/apps/cap.portal?_nfpb=true&_pageLabel=accreditation
- Clinical Laboratory Improvement Amendments (CLIA) http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/IVDRegulatoryAssistance/ucm124105.htm
- Motherisk http://www.motherisk.org/women/breastfeeding.jsp
- Pharmaceutical Outcomes and Policy Innovations Website http://www.cfri.ca/POPi/
- Pharmacogenomics Factsheet www.ncbi.nlm.nih.gov/About/primer/pharm.html
- Pharmacogenomics Knowledge Base www.pharmgkb.org/
- Skin Reactions from Carbamazepine (Carbatrol, Equetro, Tegretol) Treatment http://www.fda.gov/Drugs/DrugSafety/PostmarketdrugsafetyinformationforPatientsandProviders/ucm124718.htm
- Warfarin Therapy Management: BC Guidelines http://www.bcguidelines.ca/guideline_warfarin_management.html
References (Note: Article requests might require a login ID with CPSBC or UBC):
- Amstutz U, Carleton BC. Pharmacogenetic testing: time for clinical practice guidelines. Clin Pharmacol Ther. 89(6): 924-7. Jun 2011. PMID: 21508939 (View with CPSBC or UBC)
- Crews KR, et al.: Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for codeine therapy in the context of cytochrome P450 2D6 (CYP2D6) genotype. Clin Pharmacol Ther. 91(2): 321-6. Feb 2012. PMID: 22205192 (View with CPSBC or UBC)
- Dionne F, et al.: Economic impact of a genetic test for cisplatin-induced ototoxicity. Pharmacogenomics J. 12(3):205-13, Jun 2012. PMID: 21502965 (View with UBC)
- Johnson JA, et al.: Clinical Pharmacogenetics Implementation Consortium Guidelines for CYP2C9 and VKORC1 genotypes and warfarin dosing. Clin Pharmacol Ther. 90(4): 625-9, Oct 2011. PMID: 21900891 (View with CPSBC or UBC)
- Kelly LE, et al.: More codeine fatalities after tonsillectomy in North American children. Pediatrics. 129(5):e1343-7. May 2012. PMID: 22492761 (View with CPSBC or UBC)
- Li YY, Jones SJ. Drug repositioning for personalized medicine. Genome Med. 4(3):27. Mar 30, 2012. PMID: 22494857 (View)