Current Practices in Pharmacogenomics

INTRODUCTION

When choosing medications and calculating doses in infants, children, and adolescents, it is common to consider the patient’s diagnosis, age, weight, and sometimes body surface area. Despite careful therapeutic decisions, patients may experience inadequate therapeutic effect, a spectrum of adverse effects from mild to severe, or hypersensitivity reactions. Pharmacogenomics (PGx, also referred to as pharmacogenetics), where genomic information is used to tailor medication management, is a strategy to maximize drug efficacy and minimize toxicity. An increasing number of Food and Drug Administration (FDA) medication labels contain PGx information but do not necessarily recommend or require pre-prescription genetic testing. The Clinical Pharmacogenetics Implementation Consortium (CPIC) is an international group of experts in pharmacology and PGx dedicated to creating expert reviewed, evidence-based, and regularly updated gene-drug clinical practice guidelines.¹ A serial, cross-sectional study that included prescribing information from nearly 2.9 million pediatric patients from 16 sites, reported a conservative annual prescribing prevalence of 8000 to 11,000 per 100,000 patients who received prescriptions for CPIC level A medications (those with prescribing guidelines).² However, the systematic implementation of these guidelines to inform medication selection and dosing in pediatrics remains limited.³

In some pediatric clinics and hospitals, the implementation of PGx started as early as 2004⁴ and has gradually increased.⁵ Models for implementation vary between sites.⁶ With rising numbers of clinics and hospitals implementing PGx, pediatric providers may encounter past PGx results with therapeutic recommendations for specific medications but are unaware of the applicability of those results to new or additional medications. Patients with complex needs may be exposed to several concurrent medications, and providers may lack knowledge and/or time to consider the evidence for individualized dosing or titration.³ As other articles in this Special Issue describe, pediatric providers should expect increased use of exome sequencing (ES) and genome sequencing (GS) for diagnostic testing, which will enable opportunistic screening of genes that can influence medication response (pharmacogenes). Only two pharmacogenes (both related to malignant hyperthermia and discussed in Case 3a later in discussion) are currently recommended for opportunistic analysis from ES/GS due to analytical validity concerns for other pharmacogenes.⁷

Pharmacogenes can encode drug-metabolizing enzymes, transporters, drug targets, or immune response proteins. Most well-established pharmacogenes encode drug metabolism enzymes, particularly cytochrome P450 enzymes.⁸ The impact of variant pharmacogenes may depend on whether a drug is a pro-drug that is activated by drug-metabolizing enzymes, or an active drug that is metabolized into inactive or active metabolites (Fig. 1).

Fig. 1.

Examples of a pro-drug being metabolized to an active drug (codeine to morphine, A) and active drug being metabolized to an inactive drug (Proton pump inhibitors, (PPIs), B) with the blood concentrations and Clinical Pharmacogenetics Implementation Consortium (CPIC) recommendations.^38,44 This figure was created using Biorender.com.

PGx uses a unique nomenclature system that catalogs alleles with star numbers (eg, *1, *2, *3). The *1 allele denotes a “normal” function allele. A star allele can be informed by one or more variants in a gene. The Pharmacogene Variation Consortium (PharmVar) catalogs the star allele nomenclature and functional characterization for many pharmacogenes.⁹ The functional consequence of a star allele that codes for a drug-metabolizing enzyme is categorized as: no function, decreased function, normal function, and increased function.^10,11 The phenotype of an individual for each drug-metabolizing enzyme or transporter is determined by the combined functional activity of the maternal and paternal star alleles (called a diplotype). CPIC guidelines include an assessment of the activity of each allele, translation of diplotypes into phenotypes, and dosing recommendations.¹ For genes associated with medication hypersensitivity (eg, HLA-A, HLA-B), a single risk allele is associated with hypersensitivity.¹⁰

To avoid repeating recent reviews in pediatric PGx,^8,12 herein we use a case-based approach to illustrate the use of PGx data in pediatric clinical care. Pharmacogenes featured in cases and their potential phenotypes are in Table 1.

Table 1.

Genes tested in case examples, their potential phenotypes, and the availability of CPIC guidelines

Genes Tested in Case Examples	Potential Reported Phenotypes	At Least 1 Gene-Drug CPIC Published Guideline (Yes/No)
Cytochrome P450 genes:
CYP2B6	PM, IM, NM, RM, UM	Yes
CYP2C19	PM, IM, NM, RM, UM	Yes
CYP2C9	PM, IM, NM	Yes
CYP2D6	PM, IM, NM, UM	Yes
CYP3A5	PM, IM, NM	Yes
NUDT15	PM, IM, NM	Yes
TPMT	PM, IM, NM	Yes
Transporter genes
SLC6A4	Normal or low activity	No
Receptor genes
HTR2A	Normal, intermediate or low activity	No
Hypersensitivity genes
HLA-A*31:01	Positive, negative	Yes
HLA-B*15:02	Positive, negative	Yes
Other genes
CFTR	Cystic fibrosis	Yes
RYR1	Malignant hyperthermia susceptibility	Yes
MT-RNR1	MT-RNR1 aminoglycoside-induced hearing loss or nephrotoxicity susceptibility	Yes

Abbreviations: IM, intermediate metabolizer; NM, normal metabolizer; PM, poor metabolizer; RM, rapid metabolizer; UM, ultrarapid metabolizer.

CASES AND DISCUSSION

Case 1: Medically Complex Patient

ST is a 9 year-old patient with medical complexity consequent to a non-fatal submersion event resulting in cardiac arrest, anoxic brain injury, severe neurologic impairment, and subsequent need for many medications. She has demonstrated persistent pain and anxiety behaviors attributed to visceral hyperalgesia and likely dysfunction of the neurologic-gut axis, despite formula and medication regimen changes. Parents ask about selective serotonin reuptake inhibitors (SSRIs) and share that they learned from other parents that genetic testing can be used to help manage these medications. What resources are available to determine if results of PGx testing are relevant for SSRIs?

Parents are increasingly finding information about PGx testing on the internet and through social networks. Some parents are initiating their child’s testing through direct-to-consumer laboratories. Pediatric providers should be prepared to answer parents’ questions about PGx testing and to use information from PGx reports.

ST’s mother specifically asks about genetic testing for SSRIs. The reader is referred to a pragmatic review of considerations whether or not to test¹³ and a short, useful guide for pharmacogenetic test selection.¹⁴ Briefly, the first step is to determine if a medication has PGx guidelines or drug label with PGx advice. The Pharmacogenomics Knowledge Database (PharmGKB) is a useful place to start (Table 2). Guidelines and recommendations as well as drug labels can be searched with a Pediatric filter (Fig. 2). If PGx information is in one or more drug labels, it is tagged as either PGx: testing required, testing recommended, actionable or informative. Specific SSRI guideline recommendations are addressed in Case 2.

Table 2.

Resources for common questions

Clinical Question	Resources to Find Answers
How do I determine if my patient’s PGx test results are relevant for medications I want to prescribe?	Pharmacogenomics Knowledgebase (PharmGKB) Clinical Guideline Annotations15 Click on Pediatric filter, then search by drug to find related guidelines or recommendations or both with pediatric information published by: The Clinical Pharmacogenetics Implementation Consortium (CPIC) Royal Dutch Association for the Advancement of Pharmacy – Pharmacogenetics Working Group (DPWG) Canadian Pharmacogenomics Network for Drug Safety (CPNDS) French National Network of Pharmacogenetics PharmGKB Drug Label Annotations16 Click on Pediatric filter, then search by drug to determine if PGx information is in a drug label approved by: U.S. Food and Drug Administration (FDA) Health Canada (Sante Canada) European Medicines Agency Swiss Agency of Therapeutic Products Pharmaceuticals and Medical Devices Agency of Japan
How do I find medications that can change a child’s genotype-predicted cytochrome P450 metabolizing phenotype?	FDA Drug Development and Drug Interactions Table of Substrates, Inhibitors, and Inducers17 Click on “Clinical index inhibitors” or “Clinical index inducers”
How do I find a laboratory that offers PGx testing?	First check with your clinic/hospital’s laboratory and use existing service if available. While a database exists to find laboratories that offer PGx testing, laboratory and test comparison function is not available Consult with local or regional genetic counselor
How do I find a genetic counselor near me?	Use the National Society of Genetic Counselors’ search tool: https://findageneticcounselor.nsgc.org/In-Person-FindaGC

Fig. 2.

Find the Pharmacogenomics Knowledgebase (PharmGKB) pediatric website by going to pharmgkb.org and clicking the link for the pediatric website (indicated by the orange arrow in A) or go straight to what is shown in panel B at https://www.pharmgkb.org/pediatric/dashboard.

Drug-drug interactions must also be considered when a child is on many different medications.¹⁵ Phenoconversion is a type of drug-drug interaction that changes a child’s genotype-predicted cytochrome P450 metabolizing phenotype.^15–17 Medications, supplements, or food that act as inducers can increase the activity of a particular drug-metabolizing enzyme, whereas inhibitors can reduce or eliminate enzyme activity (see Table 2 for resource).

Some pediatric hospitals reduce the burden on clinicians by providing PGx just-in-time information delivery models.^4,6 Consulting with local pharmacists may also be helpful as pharmacy practices are increasingly incorporating PGx information into their medication therapy management frameworks.¹⁸ Newer PGx delivery models include genetic counselors, given their expertise in obtaining, evaluating, and communicating genetic test results.¹⁹

Case 2: Neurology Patient

JP is an 11-year-old nonverbal male with autism spectrum disorder, epilepsy, and worsening aggressive behavior. Current medications are clonidine (for sleep), olanza-pine (for agitation), and buspirone (for agitation). In the past, levetiracetam worsened behavior, risperidone caused fatigue and ataxia, and divalproex was ineffective. The patient is admitted to address behavior concerns, where oxcarbazepine and escitalopram are considered. Panel-based PGx testing is done on admission and results are reflected in Table 3. Which of these results are valuable, and how might they potentially explain prior medication responses and inform future drug choices?

Table 3.

Case 2 panel-based testing results

Genotype Results	Phenotype	Relevant Medication(s)	Reference(s)
HLA-B*15:02	Non-carrier	Oxcarbazepine	24
HLA-A*31:01	Non-carrier	Oxcarbazepine	24
CYP2D6 *1/*4	IM	Risperidone	25
CYP2C19 *1/*17	RM	Escitalopram	26
CYP2C9 *1/*1	NM	None
CYP2B6 *1/*6	IM	None
SLC6A4 S/S	Low Activity	None	27,28
HTR2A–1438 G > A G/A	Intermediate Activity	None	29

Relevant medication(s) refers to which medications are relevant to the case. Information about genes and results not featured in the case discussion can be found on PharmGKB.

Abbreviations: IM, intermediate metabolizer; NM, normal metabolizer; RM, rapid metabolizer.

Panel-based testing may or may not provide results relevant to a patient’s current treatment as illustrated in Table 3. This patient had panel-based testing for many genes that have no CPIC guidelines relevant to the patient at the time of evaluation, but may be useful in the future (eg, if this patient were prescribed NSAIDs the CYP2C9 result would be relevant).²⁰ There are CPIC guidelines for oxcarbazepine²¹ and escitalopram,²² which are being considered during JP’s admission.

JP’s results for HLA-B*15:02 and HLA-A*31:01 inform his provider that oxcarbazepine is not contraindicated. Very serious cutaneous adverse reactions (eg, Stevens-Johnson Syndrome and toxic epidermal necrosis) can occur after treatment with carbamazepine or oxcarbazepine in patients carrying human leukocyte antigen (HLA) alleles HLA-B*15:02 and HLA-A*31:01, which have variable frequency based on ethnic and geographical ancestry. CPIC recommends avoiding carbamazepine and oxcarbazepine in patients carrying one or more copies of these alleles.²¹

The CPIC guideline for SSRIs currently includes recommendations for escitalopram, based on CYP2C19 metabolizer status (Fig. 3).²² Based on JP’s CYP2C19 rapid metabolizer phenotype, standard escitalopram dosing is recommended.

Fig. 3.

Escitalopram dosing recommendations are based on CYP2C19 phenotype.²² This figure was created using Biorender.com.

Risperidone is metabolized by CYP2D6 into an active metabolite (9-hydroxyrisper-idone or paliperidone). To determine if JP’s CYP2D6 intermediate metabolizer status explained his past adverse reactions, JP’s provider might search PharmGKB’s Clinical Guideline Annotations to discover that while there is no CPIC guideline for risperidone, a Dutch Pharmacogenetics Working Group guideline recommends decreased dosing for CYP2D6 poor metabolizers. For CYP2D6 ultrarapid metabolizers, an alternative drug or titrating risperidone to the maximum dose for the active metabolite (paliperidone) is recommended.²³ Although dosing recommendations do not exist for CYP2D6 intermediate metabolizers, slower risperidone clearance may have contributed to JP’s previous side effects.

Case 3a: Surgical Patient

CW is a 14-year-old male scheduled for vertebral body tethering for progressive scoliosis. CW previously participated in a research study during which genomic screening by a Clinical Laboratory Improvement Amendment (CLIA) certified and College of American Pathologists (CAP) accredited laboratory found a pathogenic variant in RYR1. The finding was disclosed to the adolescent and his mother, and the PDF report was placed in CW’s electronic health record. During the preoperative clinic visit, CW’s mother shared the report with the anesthesiologist. There was no known family history of exposure to anesthesia. CW previously tolerated pressure equalizer tube placement at 18 months and tonsillectomy at 3 years of age. What precautions are indicated by the finding of a pathogenic variant in RYR1?

A pathogenic variant (previously referred to as mutation) in the RYR1 gene identifies CW as having malignant hyperthermia susceptibility (MHS), an autosomal dominant condition which may remain clinically silent until adequate exposure to a triggering drug agent. Pathogenic variants in the CACNA1S gene have been reported in ~ 1% of patients with malignant hyperthermia.²⁴ Exposure to volatile anesthetics or depolarizing neuromuscular blockers, particularly succinylcholine, or the combination of succinylcholine and a volatile anesthetic, can trigger life-threatening malignant hyperthermia.²⁵ Children can present differently than adults, delaying diagnosis.^26–28 Clinical testing of RYR1 and CACNA1S for MHS is available but not routinely performed in clinical settings prior to surgeries.²⁹

ES and GS are increasingly used for clinical diagnosis. The American College of Medical Genetics and Genomics (ACMG) recommends that regardless of indication for clinical ES/GS, analysis should routinely include specified genes known to inform genetic disorders with preventive or early treatment options, including RYR1 and CACNA1S (since 2013).^30,31 Large scale ES/GS studies have analyzed the ACMG specified genes and provided results to study participants, including children.^32–34 As the use of clinical and research ES/GS increases, pediatric clinicians are prudent to add questions about prior genetic testing for MHS to standard pre-anesthesia assessments.

In our case, CW previously tolerated anesthesia. Even in cases like this, the CPIC guideline that includes 48 RYR1 and 2 CACNA1S pathogenic variants states that succinylcholine and potent volatile anesthetics (sevoflurane, halothane, enflurane, isoflurane, methoxyflurane, and desflurane) are contraindicated in any patient with MHS.³⁵ Studies have shown that RYR1 pathogenic variants are not 100% penetrant for malignant hyperthermia phenotypes, which may explain CW’s previous tolerance to anesthesia. A recent study estimated between 0.25 and 0.76 probability of developing malignant hyperthermia with one of the 9 known RYR1 pathogenic variants when exposed to a triggering agent.³⁶ Importantly, trigger agents may be initially tolerated, with subsequent exposure resulting in malignant hyperthermia, indicating the need for continued vigilance.³⁵

Case 3b: Surgical Patient, Continued

CW’s mother reported that after his prior tonsillectomy, oral codeine was ineffective. Of note, a prior hospitalization for major depression led to PGx panel testing, and he was found to be a CYP2D6 poor metabolizer (diplotype *3/*4) and CYP2C19 normal metabolizer (diplotype *1/*1). Current medications include sertraline daily for depression and cetirizine daily as needed for seasonal allergies. The orthopedic surgeon reviewed the psychiatry PGx report, noted recommendations for antidepressants and antipsychotics, and was uncertain about the implications for postoperative pain management. How do these PGx results inform postoperative management?

Opioids and nonsteroidal anti-inflammatory drugs (NSAIDs) are essential components of perioperative pain management.³⁷ Ondansetron is commonly prescribed as opioids increase postoperative nausea and vomiting. Variable CYP2D6 function can impact effectiveness and risk for adverse reactions for some opioids and may impact ondansetron efficacy. In our case, CW has CYP2D6 *3/*4 genotype, with 2 no function alleles, consistent with a poor metabolizer phenotype. CW’s results for CYP2D6 and CYP2C19 do not inform the use of NSAIDs, which are metabolized by CYP2C9 and discussed further in Case 4.

Codeine and tramadol are prodrugs (inactive in their dosage forms) and are O-demethylated by CYP2D6 to their active (analgesic) metabolites, morphine, and O-desmethyltramadol, respectively. CYP2D6 poor metabolizers cannot convert codeine or tramadol to their active forms, and codeine and tramadol will not provide pain relief. Evidence-based guidelines recommend avoiding codeine and tramadol in any CYP2D6 poor metabolizer (see Fig. 1A). If an opioid is required for pain management, then a non-tramadol and a non-codeine containing analgesic should be considered.³⁸

The FDA issued a contraindication for codeine products and tramadol in all children less than 12 years of age, children less than 18 years after tonsillectomy or who are obese, have obstructive sleep apnea or lung disease.³⁹ Since then, the use of codeine has decreased and the use of oxycodone has increased in children.² Oxycodone and hydrocodone are converted by CYP2D6 to their more active metabolites, oxymorphone and hydromorphone. Although lower active drug concentrations are observed in CYP2D6 poor metabolizers, a clear clinical difference in drug response has not been demonstrated. Because past studies had few participants identified as CYP2D6 ultrarapid metabolizers, there is insufficient evidence regarding the safety and efficacy of oxycodone and hydrocodone for these individuals.³⁸ For CW, if opioids are required for post-operative pain management, standard dosing of oxycodone or hydrocodone would be appropriate.

Ondansetron, a 5-HT receptor antagonist, is used to prevent nausea and vomiting and is inactivated by CYP2D6. Evidence from clinical studies in adults indicates an alternative medication may be needed for CYP2D6 ultrarapid metabolizers due to inadequate effectiveness.⁴⁰ Standard dosing was effective and well tolerated by adult patients who were CYP2D6 intermediate or poor metabolizers. Studies in children at risk for postoperative nausea and vomiting are needed.^41,42 Further discussion of ondansetron is included in case 4 later in discussion.

Case 4: Oncology Patient

AR is a 16-year-old female diagnosed with acute lymphoblastic leukemia. Her tumor genetic testing indicated she had the BCR-ABL translocation, which means a tyrosine kinase inhibitor (eg, dasatinib or imatinib) is indicated. Her treatment regimen will also include mercaptopurine, ondansetron, opioids, and NSAIDs. She may receive voriconazole, tacrolimus, proton pump inhibitors (PPIs), and/or an SSRI. For these reasons, the oncology pharmacist orders a multi-gene PGx panel. A buccal swab was performed for her PGx panel, as a blood sample at the start of treatment would be largely leukemic cells that may have more mutations than her germline. The panel test results are shown in Table 4. The medications relevant to each gene on the medications she will receive are detailed later in discussion. How might these PGx results inform her chemotherapy and symptomatic management?

Table 4.

Case 4 Panel-based testing results

Genotype Results	Phenotype	Relevant Medication(s)	Reference(s)
CYP2D6 *1/*1x2	UM	Ondansetron, codeine, tramadol, paroxetine, fluvoxamine	9,31,48
CYP2C19 *1/*2	IM	Voriconazole, omeprazole, lansoprazole, pantoprazole, dexlansoprazole, sertraline, escitalopram, citalopram	10,31,51
CYP2C9 *1/*1	NM	Celecoxib, flurbiprofen, ibuprofen and lornoxicam	30
CYP3A5 *1/*3	IM	Tacrolimus	52
TPMT *1/*1	NM	Mercaptopurine	53
NUDT15 *1/*3	IM	Mercaptopurine	53

Relevant medication(s) refers to which medications are relevant to the case. Information about genes and results not featured in the case discussion can be found on PharmGKB.

Abbreviations: IM, intermediate metabolizer; NM, normal metabolizer; UM, ultrarapid metabolizer.

Thiopurines (eg, mercaptopurine) are inactive prodrugs that are metabolized into active immunosuppressants, which are then inactivated by the TPMT and NUDT15 enzymes. The CPIC guideline aligns with the FDA package insert for mercaptopurine, which indicates that TPMT or NUDT15 intermediate metabolizers often tolerate recommended standard mercaptopurine doses, but some require dose reduction based on toxicities.⁴³ If AR had been found to be a TPMT or NUDT15 poor metabolizer, her dose would need to be reduced dramatically (~10-fold). As most patients with leukemia in the U.S. are enrolled in clinical trials, the recommended dose of mercaptopurine based on TPMT and NUDT15 will be dictated by the study protocol.

Ondansetron and opioids were addressed in the previous case. However, in this case, AR is a CYP2D6 ultrarapid metabolizer. Ondansetron is prescribed to prevent or reduce chemotherapy-induced nausea and vomiting. As a CYP2D6 ultrarapid metabolizer, she metabolizes ondansetron quickly, and CPIC recommends an alternative antiemetic not metabolized by CYP2D6 (eg, granisetron),⁴⁰ though this is based on adult studies, and 2 recent studies in children could not replicate differences in response.^41,42 If opioids are needed for pain control, codeine and tramadol should be avoided because CYP2D6 ultrarapid metabolizers produce more of the active metabolite than normal metabolizers (see Fig. 1A), which puts them at increased risk of respiratory depression and other adverse events.³⁸ If pain can be controlled by NSAIDs, the patient’s CYP2C9 result is relevant for some options (see Table 4). The patient is a normal metabolizer for CYP2C9, so she should receive the typical doses of these medications (Fig. 4).

Fig. 4.

Nonsteroidal anti-inflammatory drugs (NSAIDs) dosing recommendations are based on CYP2C9 phenotype.²⁰ Of note, CYP2C9 does not have any variants that would make a patient a rapid or ultrarapid metabolizer. This figure was created using Biorender.com.

The antifungal agent voriconazole may be used prophylactically or to treat a fungal infection. As this patient is a CYP2C19 intermediate metabolizer, CPIC recommends the usual starting dose with adjustments being made for clinical factors, such as drug interactions, hepatic function, renal function, fungal species, site of infection, therapeutic drug monitoring, and comorbidities (Fig. 5).²⁰ The patient will likely have higher dose-adjusted concentrations than normal metabolizers so may need a decreased dose after therapeutic drug monitoring.

Fig. 5.

Voriconazole recommendations are based on CYP2C19 phenotype.⁵⁹ This figure was created using Biorender.com.

PPIs (eg, omeprazole) may be prescribed to this patient prophylactically or for gastroesophageal reflux. As a CYP2C19 intermediate metabolizer, she should receive the usual dose at the start of therapy but consider a 50% reduced dose for chronic therapy more than 12 weeks due to increased risk for adverse events, including respiratory and gastrointestinal infections.⁴⁴ CYP2C19 rapid and ultrarapid metabolizers are likely to have reduced blood concentrations and could need higher doses (see Fig. 1B).

Tacrolimus may be considered for AR if she requires a hematopoietic stem cell transplant. CYP3A5 variation explains ~ 50% of the variability in tacrolimus plasma concentrations.⁴⁵ Individuals who carry at least one *1 normal function allele (intermediate and normal metabolizers) are called expressers and those without are called non-expressers or poor metabolizers. The most common phenotype among individuals of European ancestry is poor metabolizer, and this epidemiology is what the usual dosing regimens are based on in the U.S. For expressers, CPIC recommends starting doses 1.5–2x higher than normal (though not exceeding 0.3 mg/kg/d), which should be followed by routine therapeutic drug monitoring (Fig. 6).⁴⁵ AR is an intermediate metabolizer so she would need a higher than normal starting dose of tacrolimus to prevent or treat graft-versus-host disease.

Fig. 6.

Tacrolimus recommendations are based on CYP3A5 phenotype.⁴⁵ This figure was created using Biorender.com.

SSRIs are often prescribed to adolescent patients with cancer for depression and anxiety, and were previously discussed in Case 2.

Case 5: Patient with Rare Genetic Disease

ML is a 6-week-old infant who had a positive newborn screen for cystic fibrosis. After additional CFTR testing F508del/G542X and a confirmatory sweat test, ML is being evaluated by providers at an accredited CF center. ML’s parents feel overwhelmed by the lifelong care ML will require but feel hopeful when they learn that ML’s genetic results indicate ML will qualify for a new type of medication called CFTR modulators.

CFTR testing is recommended as part of the diagnostic process for cystic fibrosis (CF).⁴⁶ Pathogenic variants in the CFTR gene result in dysfunction or absence of the chloride ion channel, cystic fibrosis transmembrane conductance regulator. People with CF have pathogenic variants in each of their CFTR genes resulting in multi-organ disease manifestations.⁴⁷ Disease management had relied on symptom-based treatments with incremental improvement over the years in lifespan and quality of life.⁴⁸ CFTR testing has become an essential first step before prescribing a new class of drugs called CFTR modulators, considered transformational therapies as they address the molecular cause of CF. CFTR modulators are small molecules designed to restore or improve the function of certain types of variant CFTR protein.⁴⁷ There are over a 1000 pathogenic and likely pathogenic CFTR variants listed in ClinVar,⁴⁹ and each codes for one of the 6 classes of variant CFTR proteins.⁵⁰ Available modulators target Class II, protein processing, and Class III gating variant CFTR.⁴⁷ The first FDA approved modulator, ivacaftor, is a potentiator. It helps the gating variant CFTR channel open at the cell membrane. Originally, it was approved for one specific pathogenic variant, G551D, found in less than 5% of patients with CF. Since then, its use has been expanded to include other gating variants in CFTR.⁵¹

ML has the most common pathogenic variant found in people with CF, F508del, and ML’s other variant results in absent CFTR protein. Three FDA-approved modulators contain at least one corrector small molecule essential for patients who have F508del. CFTR correctors augment the protein folding process and transit to the cell membrane.⁵² The most recently FDA-approved triple therapy modulator, elexacaftor/tezacaftor/ivacaftor, has 2 CFTR correctors to maximize stable folding and transport of the F508del CFTR protein to the cell membrane and it has the potentiator to improve the gating function of the F508del protein. It is approved in children 6 years and older,⁵³ is considered lifelong, and costs over $300,000 per year in the U.S.⁵⁴ Health Canada Guidelines indicate elexacaftor/tezacaftor/ivacaftor will be the primary modulator therapy for any CF patient over the age of 12 who has at least one F508del variant.⁵⁵

Throughout ML’s life, multiple medications and therapies will be needed⁴⁸ to manage symptoms that may include PPIs (discussed above) and aminoglycosides for which CPIC guidelines are available. Specific variants in the mitochondrial gene, MT-RNR1, predispose individuals to severe aminoglycoside-induced hearing loss. A point-of-care MT-RNR1 genotyping test has recently been developed for clinical use,⁵⁶ but testing is not done routinely due to the complexity of genotyping mitochondrial DNA. It is recommended that aminoglycoside antibiotics be avoided in individuals with one of the high-risk MT-RNR1 variants unless the infection severity or lack of effective alternative therapies outweighs the risk of hearing loss.⁵⁷ Mitochondrial genes are maternally inherited. Because each cell in the body has hundreds to thousands of mitochondria and each mitochondrion has many copies of mitochondrial DNA,⁵⁸ genetic counseling should precede genetic testing.

CLINICS CARE POINTS.

• Evidence-based gene-drug guidelines are freely available online through PharmGKB and CPIC; however, evidence specifically for children may be limited or absent.

• Potential for phenoconversion due to concomitant administration of inducers and inhibitors may need to be considered when using PGx test results to select or dose medications.

  • FDA Drug Development and Drug Interactions Table of Substrates, Inhibitors, and Inducers is a helpful resource

• Some pediatric health care facilities have or are building PGx clinical decision support into their EHRs, and these may or may not incorporate individualized phenoconversion adjustments to therapeutic recommendations.

• Some pediatric health care facilities have pharmacogenomic consultation services that may be managed by a pharmacy or a pharmacogenomic-specific multidisciplinary team.

  • Pediatric prescribers should find out what is available at their facility or local children’s hospital.
• Pediatric providers should consider the following if they work at a facility that does not offer point-of-care clinical decision support or consultation services:

  • Identify commonly prescribed medications and proactively identify which of those medications have evidence-based guidelines and have the potential to be impacted by phenoconversion.
  • Identify a local or regional genetic counselor who might be able to help identify reputable laboratories offering PGx testing or who may be willing to schedule clinical encounters with patients interested in PGx testing. The National Society of Genetic Counselors is a good place to start (see Table 2).

KEY POINTS.

• Pediatric patients are prescribed medications influenced by pharmacogenomics.

• Guidelines are available for using pharmacogenomics to manage medications in pediatric patients.

• Cases are presented to demonstrate the utility of pharmacogenomics in pediatric patients.