Utilization of Pharmacogenomics and Therapeutic Drug Monitoring for Opioid Pain Management

Paul J Jannetto; Nancy C Bratanow


Pharmacogenomics. 2009;10(7):1157-1167. 

In This Article


Clinical Usefulness

In the past decade, several studies have examined the clinical significance of genotyping CYP2D6 polymorphisms. Pharmacogenomics may be clinically important if the pharmacokinetic variability is relevant and CYP2D6 is the major route of metabolism for the compound of interest.[18] In other words, if a characterized polymorphic enzyme primarily determines the elimination or activation of a drug, genotyping might provide additional useful information. As a result, genotyping CYP2D6 as part of the standard professional practice (personalized medicine) may be good medical practice for prescribing or determining the dosage of pain medications, such as opioids.

Currently, application of pharmacogenomics in the clinical setting is becoming more common for other diseases and shows great promise in providing clinicians with the ability to more appropriately select drug therapy, predict therapeutic responses and identify those at high risk for side effects; examples include coagulation disorders, depression and chemotherapy. Recently, the US FDA approved an updated label for warfarin, which includes pharmacogenomic information.[105] Ultimately, genotyping patients prescribed warfarin may become standard to detect abnormal metabolism, which would put a patient at higher risk for complications.[19] Specific indications for genotyping and/or conventional TDM have been proposed for the prescribing of antidepressants (i.e., paroxetine and mirtazapine)[20,21] and neuroleptics/antipsychotic medications.[22] In addition, genotyping has already become standard practice in some major cancer treatment centers (i.e., Mayo Clinic and St Jude Children's Research Hospital, TN, USA).[23]

Pharmacogenomics & Pain

Interindividual differences in the perception of pain have been acknowledged clinically for over a century. Pain patients often report that they are 'more sensitive' to medications. In recent years, the differences in analgesic response to various medications are beginning to be understood. In fact, PMs have been postulated to be more susceptible to pain than EMs because of a defect in synthesizing endogenous opioids.[24]

Genetic differences in the metabolism of pain medications like codeine have been elucidated by pharmacogenomics. Codeine, in combination with acetaminophen, is one of the most commonly used preparations for the treatment of moderate pain in children[25] and is in a variety of antitussive preparations.[26] Codeine has a favorable pharmacokinetics profile in that it is well absorbed[27] and reaches peak plasma concentrations after approximately 1 h.[28] Codeine works in a similar manner to other opioids like morphine, hydrocodone and oxycodone by binding to µ- and κ-opioid receptors to produce analgesia and euphoria.[25] However, codeine itself has a very low affinity for the opioid receptors making it a weak analgesic (only one-tenth the potency of morphine).[27] It has been suggested that the primary analgesic response from codeine results from the 10% that is metabolized to morphine by CYP2D6.[29,30] Therefore, patients with no or reduced CYP2D6 activity may not achieve adequate morphine concentrations for analgesia. As a result, PM patients will have limited pain relief and may request larger doses, which could lead to toxicity or be misinterpreted as drug addiction. This concept is supported by another study where CYP2D6 PMs recovering from abdominal surgery required higher loading doses of tramadol, subsequent tramadol consumption and a need for rescue medication compared with patients carrying at least one wild-type allele.[31] Therefore, the proportion of nonresponders was significantly increased in the PM group compared with the functionally active group (46.7 vs 21.6%).[31] Unfortunately, no CYP2D6 PMs who were being prescribed tramadol for pain management were found in the current study.

In addition, genetic variations can also occur in drug absorption/transport and drug targets leading to variability in response to pain therapies. For example, one study showed interindividual variations in µ-opioid receptor gene expression and responses to painful stimuli and opioid drugs, probably due to a genetic polymorphism in the transcription regulating region of the gene.[32] Therefore, µ-opioid receptor gene expression can affect the analgesic potency of opioids like morphine making it a candidate for susceptibility or resistance to pain. In fact, a SNP in the µ-opioid receptor at position 118 (A118G) results in a variant receptor that binds β-endorphins approximately three-times more tightly than the common allelic form. As a result, β-endorphin is almost three-times more potent in individuals with the polymorphism. It may also be clinically relevant to look at genetic polymorphisms in UGTs since many pain medications undergo conjugation and these variations can affect the water solubility and elimination of these drugs. Other candidate genes including catechol-O-methyltransferase, melancortin-1 receptor and guanosine triphosphate cyclohydrolase have been investigated and associations were found with sensitivity to pain as well as with analgesic requirements in states of acute and chronic pain.[33] In the end, the clinical efficacy and side effects of pain management drugs will depend upon the sum of the genetic factors that affect both pharmacokinetic and pharmacodynamic properties.

Pain Treatment & Individualization

Pharmacotherapy remains the mainstay for the treatment of pain. Even with standardized dosing of prescription drugs, varied clinical responses are seen in patients, ranging from no effect to complete pain relief, or even unforeseen serious adverse drug effects. People differ in their genetic make-up and consequently in their response to drugs. It is estimated that the most commonly used drugs will only be effective in 30-60% of the patients with the same disease, with a subset of these patients suffering severe ADRs.[34] This is important since ADRs rank between the fourth to sixth leading cause of death in the USA and have over a US$100 billion dollar annual economic impact.[35]

Pharmacogenomics has the potential to individualize pain treatment and help avert unintentional drug overdoses. Individualization of a pain management regimen begins with the selection of an appropriate drug. Factors that guide this process include: characteristics of the pain, characteristics of the therapeutic agent and patient factors (co-existing diseases and other medications).[36,37] The problem could be more complex considering that patients are often on multiple medications that may also induce or inhibit the cytochrome P450 system, potentially putting PMs and IMs at more risk.

Individual differences may explain the lack of effect of a medication, or the inability to use it safely. ADRs may be as serious as death in patients unable to metabolize the medication safely. Previous studies have already postulated that PM patients may be at an increased risk for overdose. An illustrative case is the study revealing a higher prevalence of PMs and IMs seen in oxycodone-related deaths at the Milwaukee County Medical Examiner's office (WI, USA).[38] Another postmortem study showed that the average fentanyl concentration and the metabolic ratios of fentanyl:norfentanyl of CYP3A4*1B wild-type and 3A5*3 homozygous variant cases were higher than those of the CYP3A4*1B variant cases.[39] The conclusion of both forensic studies was that genotyping provided a more definitive interpretation of the toxicity and may serve as a molecular autopsy acting as an adjunct in certifying death cases involving pain medications.[38,39]

In this study, pharmacogenomics and TDM both provided additional information that could be valuable to pain management physicians. For oxycodone, the TDM clearly showed that patients experienced complete or partial pain relief only when the plasma Css was at least 15-32 ng/ml. This data is consistent with other published reports, which report the therapeutic window for oxycodone as 20-50 ng/ml.[40] Therefore, TDM could be used to evaluate patients that are not responding to oxycodone pain management therapy to see if they are achieving therapeutic concentrations. On the other hand, the pharmacogenomic information could be used to predict the dosage needed to achieve therapeutic Css, or even be used to predict who may be more likely to experience an ADR if started on the standard dose. Using the CYP2D6 genotype, a PM might either be started on a lower dose of an opioid metabolized by CYP2D6 or placed on an alternative medication that is metabolized by a different pathway (i.e., fentanyl). Together, the TDM and pharmacogenomic information could guide the selection and dosage of the opioid to minimize side effects and maximize therapeutic effectiveness.

This study also revealed several important issues that pain management physicians face daily. First of all, most patients followed by pain management specialists are rarely prescribed only one medication (opioid) as their sole therapy. Most patients are also prescribed an adjuvant analgesic, like neuromodulators. The issue of drug-drug interactions must always be addressed since some co-medications might induce or inhibit the metabolism of the other opioid medications leading to complications (i.e., failed therapy or ADR). In this study, only 15% of the patients were on an opioid as the sole therapy and most had either an anticonvulsant and/or antidepressant prescribed as adjunct therapy. Fluoxetine, a commonly prescribed antidepressant, is a very potent inhibitor of CYP2D6 and could make an EM look like a PM. As a result, the Css and any ADRs had to be interpreted in lieu of this information (potential drug-drug interactions). The other issue highlighted in this study was compliance to prescribed opioids. Compliance is a real problem faced by pain management physicians and occurs with any type of therapy. As a result, compliance is another confounder that must be considered when interpreting the Css.

Finally, 8% of all the patients reported an ADR to their prescribed opioid. The most commonly reported side effects were drowsiness and nausea. Of all the individuals that reported ADRs, 80% (four out of five) had impaired CYP2D6 metabolism based on their CYP2D6 predicted phenotype. The only individual that experienced an ADR, but had normal CYP2D6 metabolism, was consuming several medications that were at least partially metabolized by CYP2D6 (methadone, oxycodone, amitriptyline and cyclobenzaprine). It is important to note that methadone is both a substrate and an inhibitor of CYP2D6. Therefore, potential drug-drug interactions might explain the observed toxicity. Alternatively, there are over 75 known CYP2D6 mutations and the most prevalent mutation found in populations of this patient's ethnicity was not part of the genetic test panel. As a result, this individual may have had a polymorphism that was not tested for and truly be an impaired CYP2D6 metabolizer.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.