It is far more important to know what person the disease has than what disease the person has.
Hippocrates de Cos (460 BC–370 BC)1
In his State of the Union address on January 20, 2015, President Obama announced the launching of the Precision Medicine Initiative, earmarking $215 million for fiscal year 2016. One of the projects funded by Precision Medicine Initiative aims to integrate demographic, clinical, environmental, and genomic data from a million or more American volunteers to “understand the complex mechanisms underlying a patient’s health, disease, or condition, and to better predict which treatments will be most effective” (https://www.whitehouse.gov/the-press-office/2015/01/30/fact-sheet-president-obama-s-precision-medicine-initiative; last accessed February 24, 2016). With medicine moving increasingly toward more standardized care, precision medicine seeks to take into account individuals’ genetic and environmental variations in preventing and treating diseases.
Recent developments in perioperative medicine such as the Perioperative Surgical Home1a,2 or Enhanced Recovery Programs3 tend to emphasize the creation of standardized care pathways and protocols. One of the goals of these new programs is to decrease variability of care at the systemic and practitioner level2; however, precision medicine may provide the counterbalance to bring back a more individualized approach to health care based on inherent patient variability. Our aim in this Open Mind article is to discuss what precision medicine is and to present a road map for possible implementation in the perioperative setting (Figure 1).
THE TENETS OF PRECISION MEDICINE
The premise of precision medicine is simple: Each human being has unique differences in disease susceptibility and drug metabolism. Therefore, diagnostic and therapeutic plans ideally should be tailored individually. This is far from being a revolutionary concept, as Hippocrates emphasized more than 2000 years ago the importance of a personalized approach in medicine as mentioned in the epigraph.
With the completion of the Human Genome Project (https://www.genome.gov/11006929/2003-release-international-consortium-completes-hgp/; last accessed June 6, 2016), it became theoretically possible to tailor treatment to the precise molecular variations in individuals, potentially revolutionizing health care. An excellent representative application of the Human Genome Project in medicine is the finding of an extensive number of associations between single-nucleotide polymorphisms (SNPs) and various traits in genome-wide association studies. The National Human Genome Research Institute (NHGRI), one of the 27 institutes at NIH, and the European Bioinformatics Institute (EBI), a publicly funded bioinformatics institute based in United Kingdom, maintain a catalogue of all published genome wide association studies with P values of 5 × 10−8 or less.3a The problem of optimal significant threshold point in association studies is unresolved, but such low P values are required because of correction for multiple hypothesis testing to minimize false positives.4 Currently, the catalogue lists a total of 2554 studies with 25,037 SNP-trait associations.3a
Although genomic discoveries have been made in many major aspects of medicine, including cardiovascular, digestive, nervous system, immune system, and metabolic disease, along with cancer, the field of anesthesiology has not yet benefited clinically from such discoveries. Why has this abundance of genomic data not translated into the clinical practice of anesthesia? The simple answer may be that the relationship between genotype and phenotype is far more complicated than previously thought, particularly with regard to pharmacologic (pharmacokinetic and pharmacodynamic) variability.5 The latter is a key concern in anesthesiology, and many layers of complexity are yet to be elucidated before we can harness the predictive power of genomic medicine into clinical anesthesia practice.
PROVISOS AND ATTENDANT CHALLENGES
Although the genetic associations between individual SNPs and many traits of interest have been uncovered, it is crucial to distinguish these associations from causality. Despite robust statistical associations, the genome-wide association studies generally have odds ratio of less than 1.5 (where 1 denotes no difference).6,7 Such relatively small effect sizes (odds ratios) explain only a modest fraction of the clinical phenotypes and indicate that a large part of functional characteristic of the genotype-phenotype relationship has yet to be elucidated.
Our initial understanding of genotype-phenotype relationship was derived from Mendelian genetics, where a single gene proved to be causative for a certain phenotype. Although more than 2000 such diseases with a Mendelian inheritance pattern have been discovered, these one-gene, one phenotype relationships are not the underlying mechanism of most common diseases or pharmacological responses.8 Possible answers to this “missing heritability” may lie in the fact that genome-wide association studies assay for common SNPs (variants which are present in >5% of the population). If the causative SNP is a rare variant with high penetrance (large risk of disease with the genotype), then it may not be captured well by the genotyping arrays.6,9 Other potential explanations for the relatively low odds ratio in genome-wide association studies include structural variants, such as copy number variants, undetected gene-gene interactions, and the effect of environment as a confounding variable.5 When considering pharmacologic response, various downstream regulatory steps that alter the final phenotype but are not manifested at the nucleotide level (such as epigenetic modifications of genetic expression, inhibition of mRNA translation into protein by short strands of RNA, or selective gene transcription by transcription factors7,10) may result in altered amino acid sequence, leading to changes in enzyme function, as well as the synthesis of differing amounts of the enzyme.11
Given such level of complexity, is it realistic to expect precision medicine to be achieved and integrated into clinical practice? The answer may be dependent on successful completion of 2 tasks: (1) verification of the causal relationship between genotypes and phenotypes of interest, and (2) sufficient technological advancement to allow rapid return of genetic testing results, such as point-of-care (POC) testing, ideally at the bedside or in the preoperative clinic. Currently, there is significant progress to be made on both fronts.12,13 Although the genotype-phenotype relationship, particularly when considering pharmacologic response, has proven to be quite complex, it is reasonable to believe that pharmacogenetic mechanisms underlie a portion of adverse drug reactions. Most of these mechanisms are yet to be elucidated; however, one example where pharmacogenetics are known to be a factor is pediatric mortality associated with the administration of conventional doses of codeine.14–17
UNDERSTANDING THE POTENTIAL CAUSAL RELATIONSHIP BETWEEN GENOTYPE AND PHENOTYPE OF INTEREST IN THE PERIOPERATIVE SETTING
This problem was brought to national attention by the death of a breastfeeding infant whose mother was prescribed an analgesic containing codeine. Subsequently, the mother was found to be an ultrarapid metabolizer of codeine, leading to toxic levels of morphine in the infant.18 This unfortunate event was followed by several more case reports of fatality attributed to codeine use in children with possible obstructive sleep apnea after tonsillectomy and adenoidectomy, resulting in the Food and Drug Administration’s (FDA’s) black box warning on codeine use in children after this procedure (http://www.fda.gov/Drugs/DrugSafety/ucm339112.htm. Last accessed February 24, 2016). Subsequent genotyping revealed that ultrarapid metabolism was present in most of these children.14–16,19
Codeine must undergo O-demethylation by CYP2D6 to the active metabolite morphine to exert its analgesic effect. The function of CYP2D6 is predicated in the presence of functional or nonfunctional alleles and the total number of such alleles (also known as copy number variation). Table 1 lists various CYP2D6 alleles and their associated level of function. On the basis of the diplotype, the metabolic function of the CYP2D6 enzyme ranges from poor to ultrarapid metabolism along a continuum.20 Individuals carrying >2 copies of alleles *1 or *2 are deemed to have ultrarapid phenotype, which confers a greater potential of forming toxic serum concentrations of morphine.21 These patients would benefit from alternatives not dependent on CYP2D6 metabolism.
Of note, oxycodone and hydrocodone are both dependent on CYP2D6 metabolism and may not be good alternatives.22 Analgesics such as hydromorphone and oxymorphone, whose activity and metabolism does not seem to be dependent on CYP2D6 enzyme, may be better alternatives in patients with ultrarapid metabolism.23 If the patient’s CYP2D6 genotype is known, then the information can be used to adjust opioid dosing and/or guide the use of adjuncts such as regional techniques, wound infiltration with local anesthetics, or nonopioid analgesics for postoperative pain management. It would be convenient to consider codeine an exception because of the differential metabolism leading to an active metabolite resulting in adverse drug reactions. As pharmacogenetic underpinnings of drug metabolism are further elucidated, we may realize that variable drug metabolism is more common than previously thought. Table 2 lists some common perioperative analgesics with their major cytochrome P450 metabolic enzyme and possible implications.
Methadone is an example of a drug that has no active metabolites but because of variable metabolism by the CYP2B6 enzyme can result in an adverse drug reaction. In this case, it is the homozygosity or heterozygosity of the CYP2B6*6 allele that confers decreased metabolism and increased methadone concentrations.21 This is important because a significant association already had been established between high methadone concentration and the CYP2B6*6 allele in methadone-related fatalities.24 In 2009, methadone accounted for 2% of analgesic prescriptions but was associated with 30% of prescription analgesic deaths.25
The aforementioned examples are not isolated cases of differential metabolism leading to adverse drug reactions, as many of our medications are metabolized by the highly polymorphic CYP450 family of enzymes. As underlying genomic mechanisms are gradually uncovered, genotype testing may identify patients at risk and suggest genetically guided dosing or alternative therapy. Ultimately, such increased level of precision in treatment can lead to lower rates of adverse drug reactions.
POC GENETIC TESTING
As our understanding of differential drug metabolism improves, this pharmacogenetic knowledge will need to be incorporated into clinical practice. Such clinical translation in the form of genetic testing can take place at the time of preoperative visit. Although applicable in some cases, a turnaround time of days or even weeks makes this approach impractical in the current anesthetic practice. Furthermore, not all patients are evaluated in a preoperative clinic, and such an outpatient encounter will not address all the possible clinical situations, whereby testing may be needed.
A viable option for rapid turnaround is via POC testing devices, such as those currently in use for blood gas analysis. The use of POC testing for clopidogrel in the setting of percutaneous coronary intervention already has been reported.26 Carriers of the CYP2C19*2 and CYP2C19*3 variant have an increased incidence of major adverse events, such as stroke, myocardial infarction, in-stent thrombosis, or death. In these patients, there is insufficient platelet inhibition by clopidogrel, and they require alternative antiplatelet therapies. The Spartan RX CYP2C19 device assays (Spartan Bioscience Inc, Ottawa, ON, Canada) for the *2, *3, and *17 alleles by polymerase chain reaction (PCR) and fluorescent probe detection. The assay requires 8 minutes of sample preparation time in addition to 60 minutes of PCR run time (http://www.spartanbio.com/products/spartan-rx/performance/; last accessed February 20, 2016).
This method of allele detection works well when only a few genes need to be assayed (in the case of the Spartan RX CYP2C19 device, only 3 alleles); however, highly polymorphic genes such as CYP2D6 need technology that would allow simultaneous genotyping of allelic differences, as well as copy number variations. Oligonucleotide microarray technology, whereby an immobilized DNA can hybridize with labeled DNA or RNA, may be a viable solution because it can be used to detect the presence of SNPs, copy number variations, insertions and deletions, and complete gene deletions.27 Moreover, such technology can detect variations at the genetic, epigenetic, transcriptomic, or proteomic dimensions.28 The main advantage of microarray technology lies in its ability to detect many genes on a single platform, thereby increasing the predictive power of the test.
There currently is a commercially available genotyping array chip for the detection of CYP2D6 variants in Europe (http://www.pharmgenomics.com/index.php/products/dna-macroarray-kits/16-produkte/chips/115?gclid=CMXNguTa_swCFQdqfgod_U0FEQ; Accessed February 20, 2016); however, the 5 hours it takes to obtain results limits this technology from being practical for patients undergoing short procedures if the sample was to be obtained on the day of surgery. Alternatively, real-time or quantitative PCR could be used to detect copy number variation (an increased number of copies) of the CYP2D6 alleles, which may adversely affect drug metabolism. This would enable identification of the greatest-risk patients for whom the prescribing of codeine, hydrocodone, or oxycodone for postoperative pain relief would be ill-advised. Isolation of DNA from serum samples could be performed in a similar manner to the way it is obtained in the POC testing for CYP2C19, as mentioned previously. The challenges that remain to using this method would be the creation of the probes for each high-functioning allele, determination and validation of optimal threshold for defining high-risk versus low-risk patients, and finally the development and FDA approval of an automated POC machine capable of carrying out these processes in a short enough time to yield clinically useful answers.
THE ROAD AHEAD
In the preceding paragraphs, we have laid out 2 examples of how allele variation through differential metabolism may influence the final clinical outcome. We also have shown a possible venue (POC testing) through which such information may personalize treatment. The details of how POC genomic testing can change the way we practice are described in more detail in Figures 2 and 3.
Despite some institutional implementation of pharmacogenetic testing in the perioperative setting,29 we are cognizant of the fact that the aforementioned technical advancement has yet to be applied in clinical practice nationally. It is unclear, because of the immense complexity of the genotype/phenotype relationship and the multifaceted nature of confounding variables, if and when the predictive power of genotyping can ever be harnessed in the perioperative setting. It is the belief of the authors that the question is “when” rather than “if.” Nonetheless, medicine has moved in the genomic era,30 and one aim of articles such as the present one is to engage the anesthesiologists in the current and future genomic issues, we are likely to encounter in the perioperative setting.
The uncertain utility of genomic testing in the preoperative setting has not been a barrier for the commercial sector to offer genotyping to the public at large, also known as consumer genomics. Although the FDA regulates companies that provide health-related interpretations of genomic data, they do not regulate those that provide genomic data alone.31 Patients increasingly are presenting such reports to their health care providers. These genomic data reports can contain nonspecific information regarding activity of certain cytochrome family of enzymes involved in drug metabolism. One of the authors of this manuscript was presented with such a report and was expected to use the information in the dosing of the anesthetic regimen.
Basic explanations such as the difference between genotyping (scanning for select known variations) versus sequencing (sequencing the whole genome) can be provided in the preoperative setting. There currently are no standards in regards to which alleles should be scanned. When presented with such reports, anesthesiologists need to be familiar with an online pharmacogenomic knowledge resource in order to reference relevant information. Such explanations can take place in the PSH setting also and may allay our patient's anxiety and gain their trust. PharmGKB, a publicly available website maintained by Stanford University and partially funded by NIH, provides easily accessible genetic information on all the FDA approved medications.32 All entries are color-coded based on whether some form of testing (genetic, functional protein assay, cytogenetic studies, etc) are required or merely recommended. Moreover, pharmacogenomic actionable information where due to genetic variants the medication efficacy, dosage or toxicity maybe altered are also included. The list of medications with some sort of actionable pharmacogenomic information is increasing and in future may include perioperative medications.33
With the advent of the electronic medical record, such information will gradually be incorporated in the system and an appropriate alert can be raised whenever the medication is ordered. In the future, as the genomes of more patients are sequenced, the information can be integrated permanently into the medical record and alerts can be raised whenever a medication is ordered, which may be influenced by the patient’s variant alleles or copy number variants.34
A number of medical centers have preemptive pharmacogenetic testing programs in various stages of clinical implementation.35 This trend will likely continue as a result of better understanding of the genotype-phenotype relationship and sufficient advancement in technology for clinical translation. Such advancements in precision medicine provide opportunities to bring patient-centered care back to the forefront as a more standardized model of anesthetic care is taking shape through the efforts of the Perioperative Surgical Home and Enhanced Recovery models of care.
Name: Marc Iravani, MD.
Contributions:This author helped conceive/design the article, analyze and interpret the data, and draft/revise the article.
Name: Lisa K. Lee, MD.
Contributions: This author helped conceive/design the article, analyze and interpret the data, and draft/revise the article.
Name: Maxime Cannesson, MD, PhD.
Contributions:This author helped conceive/design the article, analyze and interpret the data, and draft/revise the article.
This manuscript was handled by:Thomas R. Vetter, MD, MPH.
Acting EIC on final acceptance: Thomas R. Vetter, MD, MPH.
1. Abrahams E, Silver M. Gordon E, Koslow S. The history of personalized medicine. In: Integrative neuroscience and personalized medicine. 2010:New York, NY: Oxford University Press; 3–16.
1a. Vetter TR, Boudreaux AM, Jones KA, Hunter JM Jr, Pittet JF. The perioperative surgical home: how anesthesiology can collaboratively achieve and leverage the triple aim in health care. Anesth Analg. 2014;118:1131–1136.
2. Kain ZN, Vakharia S, Garson L, et al. The perioperative surgical home as a future perioperative practice model. Anesth Analg. 2014;118:1126–1130.
3. Miller T, Thacker JK, White WD, et al. Reduced length of hospital stay in colorectal surgery after implementation of an enhanced recovery protocol. Anesth Analg. 2014;118:1052–1061.
3a. Burdett T (EBI), Hall PN (NHGRI), Hastings E (EBI), Hindorff LA (NHGRI), Junkins HA (NHGRI), Klemm AK (NHGRI), MacArthur J (EBI), Manolio TA (NHGRI), Morales J (EBI), Parkinson H (EBI), Welter D (EBI). The NHGRI-EBI Catalog of published genome-wide association studies. Available at: www.ebi.ac.uk/gwas
. Accessed October 1, 2016.
4. Dudbridge F, Gusnanto A. Estimation of significance thresholds for genomewide association scans. Genet Epidemiol. 2008;32:227–234.
5. Ma Q, Lu AY. Pharmacogenetics, pharmacogenomics, and individualized medicine. Pharmacol Rev. 2011;63:437–459.
6. Manolio TA, Collins FS, Cox NJ, et al. Finding the missing heritability of complex diseases. Nature. 2009;461:747–753.
7. Ritchie MD, Holzinger ER, Li R, Pendergrass SA, Kim D. Methods of integrating data to uncover genotype-phenotype interactions. Nat Rev Genet. 2015;16:85–97.
8. Lu JT, Campeau PM, Lee BH. Genotype-phenotype correlation—promiscuity in the era of next-generation sequencing. N Engl J Med. 2014;371:593–596.
9. Bodmer W, Bonilla C. Common and rare variants in multifactorial susceptibility to common diseases. Nat Genet. 2008;40:695–701.
10. Zanger UM, Schwab M. Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic variation. Pharmacol Ther. 2013;138:103–141.
11. He Y, Hoskins JM, McLeod HL. Copy number variants in pharmacogenetic genes. Trends Mol Med. 2011;17:244–251.
12. Abul-Husn NS, Owusu Obeng A, Sanderson SC, Gottesman O, Scott SA. Implementation and utilization of genetic testing in personalized medicine. Pharmgenomics Pers Med. 2014;7:227–240.
13. Moaddeb J, Haga SB. Pharmacogenetic testing: current evidence of clinical utility. Ther Adv Drug Saf. 2013;4:155–169.
14. Kelly LE, Rieder M, van den Anker J, et al. More codeine fatalities after tonsillectomy in North American children. Pediatrics. 2012;129:e1343–e1347.
15. Niesters M, Overdyk F, Smith T, Aarts L, Dahan A. Opioid-induced respiratory depression in paediatrics: a review of case reports. Br J Anaesth. 2013;110:175–182.
16. Anderson BJ. Is it farewell to codeine? Arch Dis Child. 2013;98:986–988.
17. Subramanyam R, Varughese A, Willging JP, Sadhasivam S. Future of pediatric tonsillectomy and perioperative outcomes. Int J Pediatr Otorhinolaryngol. 2013;77:194–199.
18. Koren G, Cairns J, Chitayat D, Gaedigk A, Leeder SJ. Pharmacogenetics of morphine poisoning in a breastfed neonate of a codeine-prescribed mother. Lancet. 2006;368:704.
19. Ciszkowski C, Madadi P, Phillips MS, Lauwers AE, Koren G. Codeine, ultrarapid-metabolism genotype, and postoperative death. N Engl J Med. 2009;361:827–828.
20. Zhou SF. Polymorphism of human cytochrome P450 2D6 and its clinical significance: part II. Clin Pharmacokinet. 2009;48:761–804.
21. Kharasch ED, Regina KJ, Blood J, Friedel C. Methadone pharmacogenetics: CYP2B6 polymorphisms determine plasma concentrations, clearance, and metabolism. Anesthesiology. 2015;123:1142–1153.
22. Crews KR, Gaedigk A, Dunnenberger HM, et al. Clinical Pharmacogenetics Implementation Consortium guidelines for cytochrome P450 2D6 genotype and codeine therapy: 2014 update. Clin Pharmacol Ther. 2014;95:376–382.
23. Smith HS. Opioid metabolism. Mayo Clin Proc. 2009;84:613–624.
24. Bunten H, Liang WJ, Pounder D, Seneviratne C, Osselton MD. CYP2B6 and OPRM1 gene variations predict methadone-related deaths. Addict Biol. 2011;16:142–144.
25. Paulozzi LJ, Mack KA, Jones CM. Vital signs: risk for overdose from methadone used for pain relief-United States,1999–2010. MMWR Morb Mortal Wkly Rep. 2012;61:493–497.
26. Roberts JD, Wells GA, Le May MR, et al. Point-of-care genetic testing for personalisation of antiplatelet treatment (RAPID GENE): a prospective, randomised, proof-of-concept trial. Lancet. 2012;379:1705–1711.
27. Koch WH. Technology platforms for pharmacogenomic diagnostic assays. Nat Rev Drug Discov. 2004;3:749–761.
28. Hoheisel JD. Microarray technology: beyond transcript profiling and genotype analysis. Nat Rev Genet. 2006;7:200–210.
29. Shuldiner AR, Palmer K, Pakyz RE, et al. Implementation of pharmacogenetics: the University of Maryland Personalized Anti-platelet Pharmacogenetics Program. Am J Med Genet C Semin Med Genet. 2014;166C:76–84.
30. Guttmacher AE, Collins FS. Welcome to the genomic era. N Engl J Med. 2003;349:996–998.
31. Green RC, Farahany NA. Regulation: the FDA is overcautious on consumer genomics. Nature. 2014;505:286–287.
34. Hamburg MA, Collins FS. The path to personalized medicine. N Engl J Med. 2010;363:301–304.
© 2017 International Anesthesia Research Society
35. Dunnenberger HM, Crews KR, Hoffman JM, et al. Preemptive clinical pharmacogenetics implementation: current programs in five US medical centers. Annu Rev Pharmacol Toxicol. 2015;55:89–106.