Anesthesia in the Postgenomic Era: Commercial Polymorphism Screening in a Cardiac Surgery Patient

With the complete sequencing of the human genome in the opening years of the 21st century, the potential for tailored clinical medicine based on patients’ unique genetic profiles seemed limitless and its implementation imminent. Twenty years later, the prevalence of genetic testing has increased immensely. A recent analysis of test availability estimated that more than 75,000 tests are currently available to consumers, with an average of 10 new tests added every day.1 The scope of these tests varies widely, with some focusing closely on 1 gene and others including large portions of or the entire genome. Testing techniques include not only sequencing but also microarray technology and karyotyping. While there is little doubt that testing is becoming more and more widely available, questions remain, however, about the clinical utility of many tests. The patient discussed here presented for cardiac valve surgery after undergoing commercially available genetic testing, which had been marketed as characterizing his pharmacokinetic capacity. Pharmacokinetic properties may be described as what the organism does to the drug, or how the drug is absorbed, distributed, processed, metabolized and eventually leaves the organism. Many of these pharmacokinetic processes involve the cytochrome P450 (CYP450) system, and most of the genes tested on the patient’s panel were members of the CYP450 family. CYP450 is a superfamily of cellular proteins found broadly across many forms of life, from bacteria to humans, encompassing hundreds of thousands of genes in more than 270 CYP gene families. There are 18 CYP gene families found in humans, with 59 CYP active genes identified.2 These enzymes perform oxidative, peroxidative, and reductive changes on a variety of endogenous compounds, including steroids, sterols, eicosanoids, uroporphyrinogens, bile acids, vitamin D, retinoids, and saturated and unsaturated fatty acids. These modifications may also be referred to as “phase 1 reactions.”3 Many physicians, however, are most familiar with the role CYP450 proteins can play in pharmacokinetics as direct metabolizers of drugs, conversion of inactive administered compounds to active forms, or as mediators of drug-drug interactions by way of CYP450 induction or suppression.2 The clinical relevance of CYP450 polymorphisms can be seen in the example of metoprolol metabolism in individuals with variances in CYP2D6 activity. The clinical effect of metoprolol is related to its plasma concentration with higher plasma concentrations of the drug, resulting in more marked clinical effects. Metoprolol is hydroxylated by CYP2D6 to form α-hydroxymetoprolol, thereby inactivating the drug. Polymorphisms in CYP2D6 have been identified, which result in both increased and decreased rates of hydroxylation, with slow hydroxylators demonstrating an area under the plasma concentration-time curve 6 times larger and elimination half-life 3 times longer than that of rapid hydroxylators. Phenotypically this translated to prolonged suppression of exercise-induced tachycardia.4 OPERATIVE MANAGEMENT The patient was a 57-year-old man with hypertension, a history of Lyme disease with neurologic symptoms (hyperacusis, headaches, orthostatic hypotension), anxiety, and a bioprosthetic aortic valve due to a history of bicuspid aortic valve who had developed shortness of breath and a new heart murmur. Echocardiography revealed severe aortic regurgitation, likely due to a leaflet tear. He was evaluated in the cardiac surgery clinic in preparation for surgical replacement of the bioprosthesis. In the surgery clinic, he presented the results of CYP450 polymorphism testing panel offered by Genova Diagnostics (Asheville, NC), a commercial genome sequencing company. The panel, called a “DetoxiGenomic Profile,” is described by the company as a panel that tests several polymorphisms that have been associated with “impaired detoxification capacity” and helps to identify “individuals potentially susceptible to adverse drug reactions.”5 A sample report is shown in the Figure (ie, these are not the results from this patient).6 Pharmacokinetics genetic testing had been undertaken by the patient’s primary care provider at the patient’s request after the patient reported adverse reactions to multiple medications. Per the patient, these reactions included prolonged sedation, lethargy, dysphoria, nausea, and vomiting following prior anesthetics.Figure.: Sample test result summary. A sample test result summary, available on the Genova Diagnostics website, shows how DetoxiGenomic test panel results are reported to patients. This summary report represents only a portion of the full report.5 CYP indicates cytochrome P450; NR, no result; SNP, single nucleotide polymorphism.The patient’s results report was provided by the patient to the preoperative surgical clinic, then from there forwarded to the anesthesia team. Our patient consented to the publication of these results and the details of his management. His test results are summarized in Table 1 in a format similar to the results provided by the company. For a clinician not familiar with the literature exploring polymorphisms affecting drug metabolism, the significance of these results can be difficult to interpret. Table 1. - Summary of Patient DetoxiGenomic Profile Metabolic/enzymatic function tested Polymorphism tested Test result CYP450 CYP1A1* Detected CYP1B1* Detected CYP2A6 None CYP2C9* None CYP2C19* Detected CYP2D6 None CYP2E1 None CYP3A4* Detected Methylation COMT (V158M) −/− Acetylation (N-acetyl transferase) NAT1 (R64W) −/− NAT1 (R187Q) −/− NAT2 (I114T) +/− NAT2 (R197Q) −/− NAT2 (G286E) −/− NAT2 (R64Q) −/− NAT2 (K268R) +/− Glutathione conjugation (glutathione-s-transferase) GSTM1 (1p13.3) Present GSTP1 (I105V) +/+ GSTP1 (A114V) −/− Oxidative protection SOD1 (G93A) −/− SOD1 (A4V) −/− SOD2 (A16V) +/− For CYP450 enzymes, results were reported only as no polymorphism detected (marked “none” in table), or detected (marked “detected”). Enzymes marked with “*” had multiple SNPs tested that were not specified in the results. Other genes tested were reported with specific polymorphisms (gene name in parentheses) and results shown for each gene copy. “−” indicates absence of the listed polymorphism, and “+” indicates that the polymorphism is present.Abbreviations: CYP450, cytochrome P450; GST, glutathione-s-transferase; NAT, N-acetyl transferase; SNP, single nucleotide polymorphism; SOD, super oxide dismutase. The panel tested for single nucleotide polymorphisms (SNP) 8 members of the CYP450 family. A SNP is a genetic sequence variation in which a single nucleotide has been replaced with another (eg, instead of a cytosine, a person may have a thymine at that position). SNPs are very common, with each individual’s genome estimated to contain several million SNPs. Not all SNPs will result in disease or alteration of function, though per Genova Diagnostics, the mutations tested in this panel have been associated in the published literature with impaired drug and environmental toxin metabolism. It was unclear from the test results and company-provided literature what end result of these mutations would be, particularly when a patient is heterozygous for a mutation. Even if a mutation resulted in complete loss of function, the remaining nonmutated copy of the gene may be sufficient to compensate for that loss, resulting in little or no observable clinical effect of the mutation. Also included in the panel were mutations in catechol-O-methyltransferase (COMT), 2 N-acetyl transferase (NAT) genes (NAT1 and NAT2), glutathione-s-transferase (GST) gene GSTP1, and 2 super oxide dismutase genes (SOD1 and SOD2). These enzymes are involved in the so-called phase 2 of drug metabolism, which typically involves conjugation of an additional chemical moiety to the compound resulting in greater water solubility and, as a result, increased renal excretion.2 Testing was done on these enzymes for missense mutations. Missense mutations are changes in DNA sequence that result in the substitution of 1 amino acid for another. For example, our patient tested positive for a mutation in NAT2, which changed a lysine to arginine at position 268, which is notated K268R. A second GST gene, GSTM1, is marked as “present.” The nomenclature “1p 13.3” indicates the location of this gene in the genome on chromosome 1, short arm (“p” after the French word for small: “petit”) at location 13.3. The results of this patient’s testing panel are summarized in Table 2. The patient was found to have potentially enzyme-inhibiting mutations in 4 CYP450 genes: CYP1A1, CYP1B1, CYP2C19, and CYP3A4. He was heterozygous for mutations in NAT1, NAT2, and SOD2 and homozygous for a mutation in GSTP1. Genova Diagnostics provided additional background reading on the enzymes tested, including lists of medications that are known substrates, inhibitors, and inducers of the various enzymes. The lists were examined carefully for medications commonly administered in the cardiac perioperative period. Table 2 shows listed medications that are commonly used in the cardiac surgery perioperative period and their interactions (ie, substrate, inhibitor, inducer). An additional literature search was done to identify metabolic pathways of commonly used anesthetic medication, including the volatile anesthetics. Table 2. - CYP Metabolized Medications Commonly Administered in the Perioperative Period in Cardiothoracic Surgery Medication Enzyme Medication/enzyme interaction Ondansetron CYP1A1 Substrate CYP3A4 Substrate Amiodarone CYP1A1 Inhibitor CYP3A4 Inhibitor, substrate Insulin CYP1A1 Inducer Acetaminophen CYP1A1 Substrate CYP1B1 Substrate Nitrous oxide CYP1A1 Inhibitor Propofol CYP1A1 Inhibitor Dexamethasone CYP34A Substrate, inducer Barbiturates CYP34A Inducer Midazolam CYP3A4 Substrate, inhibitor Nicardipine CYP3A4 Substrate Clopidogrel CYP3A4 Substrate Fentanyl CYP3A4 Substrate Lidocaine CYP3A4 Substrate Methadone CYP3A4 Substrate Abbreviations: COMT, catechol-O-methyltransferase; CYP, cytochrome P450. Based on this background reading and the results of this patient’s panel, an anesthetic plan was devised. The literature is far from complete in its description of the clinical results to be expected from these mutations. Given that uncertainty, the plan focused on avoiding as much as possible medications that are metabolized by enzymes known to contain mutations. This was based on the assumption that the effects of administration would be more difficult to predict in intensity, duration, and side effects potentially more pronounced when alterations in metabolism are present. Preoperatively, the patient expressed and displayed significant anxiety. Although he was known to have a CYP3A4 mutation that could slow the metabolism of midazolam given the patient’s significant preoperative anxiety, the potential benefit of anxiolysis was felt to outweigh the possible risk of prolonged drug effect. A small dose of midazolam (half of what would typically be given to a similar age and size patient) was given intravenously in the preparative area. Induction was done with hydromorphone, etomidate, and rocuronium. This approach enabled us to avoid propofol as a known inhibitor of CYP1A1. Following endotracheal intubation, anesthesia was maintained using inhaled isoflurane and intermittent boluses of rocuronium. The patient was transferred to the cardiac surgery intensive care unit, intubated, and sedated on a dexmedetomidine infusion. He was extubated approximately 6 hours after surgery and transferred from intensive care on postoperative day 1. The surgical team was aware of the results of his genetic testing and worked to avoid or minimize use of medications for which he may have impaired metabolism. His postoperative course was uneventful, and he was discharged home on postoperative day 4. DISCUSSION The advent of easily accessible, personalized genetic sequencing raises both new opportunities and challenges for physicians. Improvements in the rapidity of sequencing, coupled with increasing understanding of relationships between genes, the proteins they produce, and clinically observable effects near the close of the last century heralded an oncoming age of personalized medicine based on individual analytics. However, while there is great theoretical promise in the application of sequencing to the practice of medicine, our current understanding of the significance of individual mutations remains limited. Many anesthesiologists may recall from introductory genetics courses the relevant principle that presence of a mutation in many cases will not correlate invariably to an observed phenotype. Genetic sequencing is becoming more available to our patients, and more patients are likely to present the results of these tests before undergoing anesthesia. According to the Hippocratic Oath, we are obligated to first do no harm, which may lead the anesthesiologist to modify an anesthetic plan based on the results of these tests. But it may also lead to an anesthetic or administration of medications we otherwise feel would be poorly suited for this patient, potentially introducing harm via another avenue. The clinical relevance of these test results is uncertain, but patients may place great faith in them and insist the results be used to guide their treatment. Anesthesiologists in particular face additional challenges as there are medications used almost exclusively in the field of anesthesia, such as volatile anesthetic agents, short-acting opiates such as remifentanil and neuromuscular blocking agents, which may not be addressed in available literature. This places the anesthesia practitioner in the delicate position of providing anesthetic care based on proven principles that is likely to provide safe and effective result while still considering a patient’s expressed preferences. A useful resource for the practicing anesthesiologist is the Clinical Pharmacogenetics Implementation Consortium (CPIC),7 a largely volunteer professional group of scientists and physicians that work to collate reliable data regarding pharmacogenomics and provide recommendations for those seeking to apply this information clinically. CPIC offers guidelines for modifying medication administration based on current peer-reviewed evidence. At the time of this writing, there are only 3 gene mutations listed in the CPIC guidelines as requiring significant modification in the administration of drugs commonly used in the perioperative period. Mutations in ryanodine receptor 1 (RYR1) and calcium voltage-gated channel subunit alpha 1S (CACNA1S) genes lead to malignant hyperthermia susceptibility and require diligent avoidance of halogenated anesthetic agents and succinylcholine. Mutations in the CYP2D6 enzyme have been shown to affect metabolism of ondansetron, with phenotypes displaying both rapid and slowed metabolism. CPIC recommends an alternative agent in rapid metabolizers but no dose modifications for other phenotypes. The Dutch Pharmacogenomics Working Group (DPWG)8 is a similar professional consortium aimed at providing continually updated clinical guidance based on evolving peer-reviewed pharmacogenomic research. Currently, the DPWG guidelines do not include recommendations for modification of dosing or medication selection for any medications commonly administered in the perioperative period. In designing an anesthetic plan for this patient, we were able to avoid commonly used anesthetic drugs specifically identified as inducers or inhibitors of the abnormal tested enzymes. For example, the patient was identified as carrying 2 separate polymorphisms in CYP1A1 that affect enzyme activity, though it was unclear from the genetic report if the mutations identified resulted in hyper- or hypoactivation. CYP1A1 has been identified as relevant to propofol metabolism,9 so propofol was avoided. Testing results also indicated CYP3A4 mutation consistent with slowed metabolism, so fentanyl, an opiate known to be largely a substrate of CYP3A4, was avoided in favor of hydromorphone, with the aim of achieving more predictable pharmacokinetics. Hydromorphone is also a CYP3A4 substrate10 but has additional metabolism through the CYP2C9 pathway. While it is fortunate that the patient enjoyed uneventful intra- and postoperative courses without obvious medication-related complications, it is unclear if the modification of the medication regime was strictly indicated based on currently available data. His course may have been similarly benign without these steps. However, it is likely that modification of anesthetic plans will be increasingly requested by patients as genetic testing becomes more widely available. Physicians have a duty to consider these requests as part of respecting patient autonomy but must recognize that the discipline of pharmacogenomic anesthesia is very much in its early days. DISCLOSURES Name: Susan K. Sankova, MD. Contribution: This author helped in all aspects of this manuscript, including performance of the clinical case and drafting and revision of the text. Name: Seema Deshpande, MBBS. Contribution: This author helped analyze, interpret, and revise the manuscript. Name: Steve P. Miller, MD. Contribution: This author helped with the manuscript, including performance of the clinical case and drafting and revision of the text. Name: Kenichi Tanaka, MD. Contribution: This author helped with the manuscript, including its conception, interpretation, drafting, and revision of the text. This manuscript was handled by: Nikolaos J. Skubas, MD, DSc, FACC, FASE.