Fetal pharmacogenomics: A promising addition to complex neonatal care.

OBJECTIVE: Pharmacogenomics (PGx) characterizes genetic variation in medication response. 85-95% of the population carries actionable PGx variants. No prior studies have demonstrated the application and feasibility of PGx in prenatal testing. We assessed parental desire for PGx findings from fetal exome sequencing (ES), evaluated PGx variants, and reviewed implications for medically complex neonates. METHODS: A prospective cohort undergoing ES for nonimmune hydrops fetalis were offered PGx results as a secondary finding. Seven pharmacogenes with Level A evidence, defined by Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines, were tested and reported to patients and referring providers. Medication administration records were reviewed. RESULTS: Most participants (36/40, 90%) desired PGx testing. 32/36 (89%) had potentially actionable PGx diplotypes in six genes: CYP2C19 (20/36, 56%), CYP2C9 (16/36, 44%), CYP2D6 (10/36, 28%), SLCO1B1 (13/36, 36%), TPMT (6/36, 17%), UGT1A1 (4/36, 11%). 12/13 (92%) live births had PGx variants. Neonatal chart review indicated that three medications with CPIC Level A evidence were administered to four neonates. None of the patients received a medication that aligned with an actionable pharmacogenetic variant as defined by Level A CPIC guidance. CONCLUSION: Most participants opted to receive PGx results. 89% had actionable variants, consistent with population estimates. Obtaining fetal PGx data is feasible for medically complex neonates. Further studies are needed for broad clinical application of PGx in fetuses with major congenital abnormalities. Our study demonstrates the potential of PGx as useful preemptive clinical information that could be obtained at the time of fetal exome sequencing for other indications. CLINICALTRIALS: gov Registration: NCT03911531.

Multisite investigation of strategies for the clinical implementation of pre-emptive pharmacogenetic testing.

PURPOSE: The increased availability of clinical pharmacogenetic (PGx) guidelines and decreasing costs for genetic testing have slowly led to increased utilization of PGx testing in clinical practice. Pre-emptive PGx testing, where testing is performed in advance of drug prescribing, is one means to ensure results are available at the time of prescribing decisions. However, the most efficient and effective methods to clinically implement this strategy remain unclear. METHODS: In this report, we compare and contrast implementation strategies for pre-emptive PGx testing by 15 early-adopter institutions. We surveyed these groups, collecting data on testing approaches, team composition, and workflow dynamics, in addition to estimated third-party reimbursement rates. RESULTS: We found that while pre-emptive PGx testing models varied across sites, institutions shared several commonalities, including methods to identify patients eligible for testing, involvement of a precision medicine clinical team in program leadership, and the implementation of pharmacogenes with Clinical Pharmacogenetics Implementation Consortium guidelines available. Finally, while reimbursement rate data were difficult to obtain, the data available suggested that reimbursement rates for pre-emptive PGx testing remain low. CONCLUSION: These findings should inform the establishment of future implementation efforts at institutions considering a pre-emptive PGx testing program.

Cost-effectiveness of CYP2C19-guided antiplatelet therapy in patients with acute coronary syndrome and percutaneous coronary intervention informed by real-world data.

Current guidelines recommend dual antiplatelet therapy (DAPT) consisting of aspirin and a P2Y12 inhibitors following percutaneous coronary intervention (PCI). CYP2C19 genotype can guide DAPT selection, prescribing ticagrelor or prasugrel for loss-of-function (LOF) allele carriers (genotype-guided escalation). Cost-effectiveness analyses (CEA) are traditionally grounded in clinical trial data. We conduct a CEA using real-world data using a 1-year decision-analytic model comparing primary strategies: universal empiric clopidogrel (base case), universal ticagrelor, and genotype-guided escalation. We also explore secondary strategies commonly implemented in practice, wherein all patients are prescribed ticagrelor for 30 days post PCI. After 30 days, all patients are switched to clopidogrel irrespective of genotype (nonguided de-escalation) or to clopidogrel only if patients do not harbor an LOF allele (genotype-guided de-escalation). Compared with universal clopidogrel, both universal ticagrelor and genotype-guided escalation were superior with improvement in quality-adjusted life years (QALY's). Only genotype-guided escalation was cost-effective ($42,365/QALY) and demonstrated the highest probability of being cost-effective across conventional willingness-to-pay thresholds. In the secondary analysis, compared with the nonguided de-escalation strategy, although genotype-guided de-escalation and universal ticagrelor were more effective, with ICER of $188,680/QALY and $678,215/QALY, respectively, they were not cost-effective. CYP2C19 genotype-guided antiplatelet prescribing is cost-effective compared with either universal clopidogrel or universal ticagrelor using real-world implementation data. The secondary analysis suggests genotype-guided and nonguided de-escalation may be viable strategies, needing further evaluation.

Correction: Evaluating the extent of reusability of CYP2C19 genotype data among patients genotyped for antiplatelet therapy selection.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Evaluating the extent of reusability of CYP2C19 genotype data among patients genotyped for antiplatelet therapy selection.

PURPOSE: Genotype-guided antiplatelet therapy is increasingly being incorporated into clinical care. The purpose of this study is to determine the extent to which patients initially genotyped for CYP2C19 to guide antiplatelet therapy were prescribed additional medications affected by CYP2C19. METHODS: We assembled a cohort of patients from eight sites performingCYP2C19 genotyping to inform antiplatelet therapy. Medication orders were evaluated from time of genotyping through one year. The primary endpoint was the proportion of patients prescribed two or more CYP2C19 substrates. Secondary endpoints were the proportion of patients with a drug-genotype interaction and time to receiving a CYP2C19 substrate. RESULTS: Nine thousand one hundred ninety-one genotyped patients (17% nonwhite) with a mean age of 68 ± 3 years were evaluated; 4701 (51%) of patients received two or more CYP2C19 substrates and 3835 (42%) of patients had a drug-genotype interaction. The average time between genotyping and CYP2C19 substrate other than antiplatelet therapy was 25 ± 10 days. CONCLUSIONS: More than half of patients genotyped in the setting of CYP2C19-guided antiplatelet therapy received another medication impacted by CYP2C19 in the following year. Given that genotype is stable for a patient's lifetime, this finding has implications for cost effectiveness, patient care, and treatment outcomes beyond the indication for which it was originally performed.

Multi-site investigation of strategies for the clinical implementation of CYP2D6 genotyping to guide drug prescribing.

PURPOSE: A number of institutions have clinically implemented CYP2D6 genotyping to guide drug prescribing. We compared implementation strategies of early adopters of CYP2D6 testing, barriers faced by both early adopters and institutions in the process of implementing CYP2D6 testing, and approaches taken to overcome these barriers. METHODS: We surveyed eight early adopters of CYP2D6 genotyping and eight institutions in the process of adoption. Data were collected on testing approaches, return of results procedures, applications of genotype results, challenges faced, and lessons learned. RESULTS: Among early adopters, CYP2D6 testing was most commonly ordered to assist with opioid and antidepressant prescribing. Key differences among programs included test ordering and genotyping approaches, result reporting, and clinical decision support. However, all sites tested for copy-number variation and nine common variants, and reported results in the medical record. Most sites provided automatic consultation and had designated personnel to assist with genotype-informed therapy recommendations. Primary challenges were related to stakeholder support, CYP2D6 gene complexity, phenotype assignment, and sustainability. CONCLUSION: There are specific challenges unique to CYP2D6 testing given the complexity of the gene and its relevance to multiple medications. Consensus lessons learned may guide those interested in pursuing similar clinical pharmacogenetic programs.

Nitrite pharmacokinetics, safety and efficacy after experimental ventricular fibrillation cardiac arrest.

INTRODUCTION: Besides therapeutic hypothermia or targeted temperature management no novel therapies have been developed to improve outcomes of patients after cardiac arrest (CA). Recent studies suggest that nitrite reduces neurological damage after asphyxial CA. Nitrite is also implicated as a new mediator of remote post conditioning produced by tourniquet inflation-deflation, which is under active investigation in CA. However, little is known about brain penetration or pharmacokinetics (PK). Therefore, to define the optimal use of this agent, studies on the PK of nitrite in experimental ventricular fibrillation (VF) are needed. We tested the hypothesis that nitrite administered after resuscitation from VF is detectable in cerebrospinal fluid (CSF), brain and other organ tissues, produces no adverse hemodynamic effects, and improves neurologic outcome in rats. METHODS: After return of spontaneous circulation (ROSC) of 5 min untreated VF, adult male Sprague-Dawley rats were given intravenous nitrite (8 µM, 0.13 mg/kg) or placebo as a 5 min infusion beginning at 5 min after CA. Additionally, sham groups with and without nitrite treatment were also studied. Whole blood nitrite levels were serially measured. After 15 min, CSF, brain, heart and liver tissue were collected. In a second series, using a randomized and blinded treatment protocol, rats were treated with nitrite or placebo after arrest. Neurological deficit scoring (NDS) was performed daily and eight days after resuscitation, fear conditioning testing (FCT) and brain histology were assessed. RESULTS: In an initial series of experiments, rats (n = 21) were randomized to 4 groups: VF-CPR and nitrite therapy (n = 6), VF-CPR and placebo therapy (n = 5), sham (n = 5), or sham plus nitrite therapy (n = 5). Whole blood nitrite levels increased during drug infusion to 57.14 ±â€¯10.82 µM at 11 min post-resuscitation time (1 min after dose completion) in the VF nitrite group vs. 0.94 ±â€¯0.58 µM in the VF placebo group (p < 0.001). There was a significant difference between the treatment and placebo groups in nitrite levels in blood between 7.5 and 15 min after CPR start and between groups with respect to nitrite levels in CSF, brain, heart and liver. In a second series (n = 25 including 5 shams), 19 out of 20 animals survived until day 8. However, NDS, FCT and brain histology did not show any statistically significant difference between groups. CONCLUSIONS: Nitrite, administered early after ROSC from VF, was shown to cross the blood brain barrier after a 5 min VF cardiac arrest. We characterized the PK of intravenous nitrite administration after VF and were able to demonstrate nitrite safety in this feasibility study.

Design and testing of Medivate, a mobile app to achieve medication list portability via Fast Healthcare Interoperability Resources.

OBJECTIVES: To describe the architecture, design, and testing of an innovative mobile application (Medivate) that facilities accurate sharing of medication lists with linked education. SETTING: The deployment and testing of this app occurred in both the community and hospital settings in Pittsburgh, PA. PRACTICE INNOVATION: Medivate is an iOS smartphone application and cloud architecture for patients and providers to keep medication and vaccine lists accurate by providing a method and tool to easily exchange these data. Medications are added directly to the app from the electronic health record (EHR) or by the patient manually. Quick response (QR) code technology is used to trigger the secure transfer and sharing of medications on demand via HL-7 Fast Healthcare Interoperability Resources-based data transfer. An iterative user-centered design process involving patients and student pharmacists practicing in community settings was used to develop and refine functionality. PRACTICE DESCRIPTION: Adults with an iPhone were approached for participation in the design and evaluation of Medivate. Its functionality and integration into clinical workflow at hospital discharge or vaccine administration in the community were determined. EVALUATION: In the community setting, interviews of pharmacists were conducted. In the hospital, metrics of study participation and experience with app deployment were determined. RESULTS: The app was deployed in the community for patients that received vaccinations. Interviews provided insight into barriers and logistics for successful engagement. The app was integrated into hospital workflow and demonstrated interoperability with the inpatient EHR. Thirteen patients were provided the app before discharge. Engagement with the app was evident through medication list shares, education access, and changes to medication lists. Patients noted strong agreement with usefulness of the app to learn more about the purposes and adverse effects of medications prescribed. CONCLUSION: A mobile app to achieve medication and vaccine list portability was successfully designed and integrated into the inpatient and community settings.

Community pharmacists' educational needs for implementing clinical pharmacogenomic services.

OBJECTIVES: Pharmacist leadership and knowledge of pharmacogenomics is critical to the acceleration and enhancement of clinical pharmacogenomic services. This study aims for a qualitative description of community pharmacists' pharmacogenomic educational needs when implementing clinical pharmacogenomic services at community pharmacies. METHODS: Pharmacists practicing at Rite Aid Pharmacy locations in the Greater Pittsburgh Area were recruited to participate in this qualitative analysis. Pharmacists from pharmacy locations offering pharmacogenomic testing and robust patient care services were eligible to participate in a semistructured, audio-recorded interview. The semistructured interview covered 4 domains crafted by the investigative team: (1) previous knowledge of pharmacogenomics; (2) implementation resources; (3) workflow adaptation; and (4) learning preferences. Interviews were transcribed verbatim and independently coded by 2 researchers. A thematic analysis by the investigative team followed. Supporting quotes were selected to illustrate each theme. RESULTS: Eleven pharmacists from 9 unique pharmacy locations participated in this study. The average length of practice as a community pharmacist was 12 years (range, 1.5-31 years). Pharmacist's pharmacogenomic educational needs were categorized into 5 key themes: (1) enriched pharmacogenomic education and training; (2) active learning to build confidence in using pharmacogenomic data in practice; (3) robust and reputable clinical resources to effectively implement pharmacogenomic services; (4) team-based approach throughout implementation; (5) readily accessible network of pharmacogenomic experts. CONCLUSION: This study describes the educational needs and preferences of community pharmacists for the successful provision of clinical pharmacogenomic services in community pharmacies. Pharmacists recognized their needs for enriched knowledge and instruction, practice applying pharmacogenomic principles with team-based approaches, robust clinical resources, and access to pharmacogenomic experts. This deeper understanding of pharmacist needs for pharmacogenomic education could help to accelerate and enhance the clinical implementation of pharmacogenomic services led by community pharmacists.

Exploratory Application of Neuropharmacometabolomics in Severe Childhood Traumatic Brain Injury.

OBJECTIVES: To employ metabolomics-based pathway and network analyses to evaluate the cerebrospinal fluid metabolome after severe traumatic brain injury in children and the capacity of combination therapy with probenecid and N-acetylcysteine to impact glutathione-related and other pathways and networks, relative to placebo treatment. DESIGN: Analysis of cerebrospinal fluid obtained from children enrolled in an Institutional Review Board-approved, randomized, placebo-controlled trial of a combination of probenecid and N-acetylcysteine after severe traumatic brain injury (Trial Registration NCT01322009). SETTING: Thirty-six-bed PICU in a university-affiliated children's hospital. PATIENTS AND SUBJECTS: Twelve children 2-18 years old after severe traumatic brain injury and five age-matched control subjects. INTERVENTION: Probenecid (25 mg/kg) and N-acetylcysteine (140 mg/kg) or placebo administered via naso/orogastric tube. MEASUREMENTS AND MAIN RESULTS: The cerebrospinal fluid metabolome was analyzed in samples from traumatic brain injury patients 24 hours after the first dose of drugs or placebo and control subjects. Feature detection, retention time, alignment, annotation, and principal component analysis and statistical analysis were conducted using XCMS-online. The software "mummichog" was used for pathway and network analyses. A two-component principal component analysis revealed clustering of each of the groups, with distinct metabolomics signatures. Several novel pathways with plausible mechanistic involvement in traumatic brain injury were identified. A combination of metabolomics and pathway/network analyses showed that seven glutathione-centered pathways and two networks were enriched in the cerebrospinal fluid of traumatic brain injury patients treated with probenecid and N-acetylcysteine versus placebo-treated patients. Several additional pathways/networks consisting of components that are known substrates of probenecid-inhibitable transporters were also identified, providing additional mechanistic validation. CONCLUSIONS: This proof-of-concept neuropharmacometabolomics assessment reveals alterations in known and previously unidentified metabolic pathways and supports therapeutic target engagement of the combination of probenecid and N-acetylcysteine treatment after severe traumatic brain injury in children.

Regionally clustered ABCC8 polymorphisms in a prospective cohort predict cerebral oedema and outcome in severe traumatic brain injury.

OBJECTIVE: ABCC8 encodes sulfonylurea receptor 1, a key regulatory protein of cerebral oedema in many neurological disorders including traumatic brain injury (TBI). Sulfonylurea-receptor-1 inhibition has been promising in ameliorating cerebral oedema in clinical trials. We evaluated whether ABCC8 tag single-nucleotide polymorphisms predicted oedema and outcome in TBI. METHODS: DNA was extracted from 485 prospectively enrolled patients with severe TBI. 410 were analysed after quality control. ABCC8 tag single-nucleotide polymorphisms (SNPs) were identified (Hapmap, r2>0.8, minor-allele frequency >0.20) and sequenced (iPlex-Gold, MassArray). Outcomes included radiographic oedema, intracranial pressure (ICP) and 3-month Glasgow Outcome Scale (GOS) score. Proxy SNPs, spatial modelling, amino acid topology and functional predictions were determined using established software programs. RESULTS: Wild-type rs7105832 and rs2237982 alleles and genotypes were associated with lower average ICP (ß=-2.91, p=0.001; ß=-2.28, p=0.003) and decreased radiographic oedema (OR 0.42, p=0.012; OR 0.52, p=0.017). Wild-type rs2237982 also increased favourable 3-month GOS (OR 2.45, p=0.006); this was partially mediated by oedema (p=0.03). Different polymorphisms predicted 3-month outcome: variant rs11024286 increased (OR 1.84, p=0.006) and wild-type rs4148622 decreased (OR 0.40, p=0.01) the odds of favourable outcome. Significant tag and concordant proxy SNPs regionally span introns/exons 2-15 of the 39-exon gene. CONCLUSIONS: This study identifies four ABCC8 tag SNPs associated with cerebral oedema and/or outcome in TBI, tagging a region including 33 polymorphisms. In polymorphisms predictive of oedema, variant alleles/genotypes confer increased risk. Different variant polymorphisms were associated with favourable outcome, potentially suggesting distinct mechanisms. Significant polymorphisms spatially clustered flanking exons encoding the sulfonylurea receptor site and transmembrane domain 0/loop 0 (juxtaposing the channel pore/binding site). This, if validated, may help build a foundation for developing future strategies that may guide individualised care, treatment response, prognosis and patient selection for clinical trials.

ABCB1 genotype is associated with fentanyl requirements in critically ill children.

BackgroundThe gene ABCB1 encodes p-glycoprotein, a xenobiotic efflux pump capable of transporting certain opioids, including fentanyl. ABCB1 genotype has been previously associated with patient opioid requirements and may influence fentanyl dosing requirements in critically ill children.MethodsA diagnostically diverse cohort of 61 children who received a fentanyl infusion while admitted to the pediatric intensive care unit (PICU) were included in this study. We examined associations between fentanyl requirements, pain and sedation scores, serum fentanyl levels, and ABCB1 genotype.ResultsPatients with the AA allele at ABCB1 locus rs1045642 received less fentanyl compared with patients with the AG or GG allele. A multivariable model demonstrated that patients with the AA allele received 18.6 mcg/kg/day less fentanyl than patients with either the AG or GG allele (95% confidence interval -33.4 to -3.8 mcg/kg/day; P=0.014). Incorporating race in this model demonstrated a similar association, but did not reach the threshold for multiple testing.ConclusionABCB1 genotype rs1045642 AA is associated with fentanyl administration in this cohort of children admitted to the PICU, likely because of decreased expression and activity of p-glycoprotein. Prospective evaluation of the influence of ABCB1 in sedative-analgesia administration in critically ill children is warranted.

Probenecid, an organic anion transporter 1 and 3 inhibitor, increases plasma and brain exposure of N-acetylcysteine.

1. N-acetylcysteine (NAC) is being investigated as an antioxidant for several conditions including traumatic brain injury, but the mechanism by which it crosses membrane barriers is unknown. We have attempted to understand how the transporter inhibitor, probenecid, affects NAC pharmacokinetics and to evaluate the interaction of NAC with transporters. 2. Juvenile Sprague-Dawley rats were administered NAC alone or in combination with probenecid intraperitoneally. Plasma and brain samples were collected serially and NAC concentrations were measured. Transporter studies were conducted with human embryonic kidney-293 cells that overexpress organic anion transporter (OAT)1 or OAT3 and with human multi-drug resistance-associated protein (MRP)1 or MRP4 membrane vesicles. 3. NAC area under the curve was increased in plasma (1.65-fold) and brain (2.41-fold) by probenecid. The apparent plasma clearance was decreased by 65%. Time- and concentration-dependent NAC uptake that was inhibitable by probenecid was observed with OAT1 and OAT3. No uptake of NAC was observed with MRP1 or MRP4. 4. Our results indicate for the first time that NAC is substrate for OAT1 and OAT3 and that probenecid increases NAC plasma and brain exposure in vivo. These data provide insight regarding how NAC crosses biological barriers and suggest a promising therapeutic strategy to increase NAC exposure.

ABCC8 Single Nucleotide Polymorphisms are Associated with Cerebral Edema in Severe TBI.

OBJECTIVE: Cerebral edema (CE) in traumatic brain injury (TBI) is the consequence of multiple underlying mechanisms and is associated with unfavorable outcomes. Genetic variability in these pathways likely explains some of the clinical heterogeneity observed in edema development. A role for sulfonylurea receptor-1 (Sur1) in CE is supported. However, there are no prior studies examining the effect of genetic variability in the Sur1 gene (ABCC8) on the development of CE. We hypothesize that ABCC8 single nucleotide polymorphisms (SNPs) are predictive of CE. METHODS: DNA was extracted from 385 patients. SNPs in ABCC8 were genotyped using the Human Core Exome v1.2 (Illumina). CE measurements included acute CT edema, mean and peak intracranial pressure (ICP), and need for decompressive craniotomy. RESULTS: Fourteen SNPs with minor allele frequency >0.2 were identified. Four SNPS rs2283261, rs3819521, rs2283258, and rs1799857 were associated with CE measures. In multiple regression models, homozygote-variant genotypes in rs2283261, rs3819521, and rs2283258 had increased odds of CT edema (OR 2.45, p = 0.007; OR 2.95, p = 0.025; OR 3.00, p = 0.013), had higher mean (ß = 3.13, p = 0.000; ß = 2.95, p = 0.005; ß = 3.20, p = 0.008), and peak ICP (ß = 8.00, p = 0.001; ß = 7.64, p = 0.007; ß = 6.89, p = 0.034). The homozygote wild-type genotype of rs1799857 had decreased odds of decompressive craniotomy (OR 0.47, p = 0.004). CONCLUSIONS: This is the first report assessing the impact of ABCC8 genetic variability on CE development in TBI. Minor allele ABCC8 SNP genotypes had increased risk of CE, while major SNP alleles were protective-potentially suggesting an evolutionary advantage. These findings could guide risk stratification, treatment responders, and the development of novel targeted or gene-based therapies against CE in TBI and other neurological disorders.

Emerging therapies in traumatic brain injury.

Despite decades of basic and clinical research, treatments to improve outcomes after traumatic brain injury (TBI) are limited. However, based on the recent recognition of the prevalence of mild TBI, and its potential link to neurodegenerative disease, many new and exciting secondary injury mechanisms have been identified and several new therapies are being evaluated targeting both classic and novel paradigms. This includes a robust increase in both preclinical and clinical investigations. Using a mechanism-based approach the authors define the targets and emerging therapies for TBI. They address putative new therapies for TBI across both the spectrum of injury severity and the continuum of care, from the field to rehabilitation. They discussTBI therapy using 11 categories, namely, (1) excitotoxicity and neuronal death, (2) brain edema, (3) mitochondria and oxidative stress, (4) axonal injury, (5) inflammation, (6) ischemia and cerebral blood flow dysregulation, (7) cognitive enhancement, (8) augmentation of endogenous neuroprotection, (9) cellular therapies, (10) combination therapy, and (11) TBI resuscitation. The current golden age of TBI research represents a special opportunity for the development of breakthroughs in the field.

Interpatient variability in dexmedetomidine response: a survey of the literature.

Fifty-five thousand patients are cared for in the intensive care unit (ICU) daily with sedation utilized to reduce anxiety and agitation while optimizing comfort. The Society of Critical Care Medicine (SCCM) released updated guidelines for management of pain, agitation, and delirium in the ICU and recommended nonbenzodiazepines, such as dexmedetomidine and propofol, as first line sedation agents. Dexmedetomidine, an alpha-2 agonist, offers many benefits yet its use is mired by the inability to consistently achieve sedation goals. Three hypotheses including patient traits/characteristics, pharmacokinetics in critically ill patients, and clinically relevant genetic polymorphisms that could affect dexmedetomidine response are presented. Studies in patient traits have yielded conflicting results regarding the role of race yet suggest that dexmedetomidine may produce more consistent results in less critically ill patients and with home antidepressant use. Pharmacokinetics of critically ill patients are reported as similar to healthy individuals yet wide, unexplained interpatient variability in dexmedetomidine serum levels exist. Genetic polymorphisms in both metabolism and receptor response have been evaluated in few studies, and the results remain inconclusive. To fully understand the role of dexmedetomidine, it is vital to further evaluate what prompts such marked interpatient variability in critically ill patients.

Towards consensus recommendations for pharmacogenetics testing.

In an effort to expedite the publication of articles, AJHP is posting manuscripts online as soon as possible after acceptance. Accepted manuscripts have been peer-reviewed and copyedited, but are posted online before technical formatting and author proofing. These manuscripts are not the final version of record and will be replaced with the final article (formatted per AJHP style and proofed by the authors) at a later time.

Empowering personalized pharmacogenomics with generative AI solutions.

OBJECTIVE: This study evaluates an AI assistant developed using OpenAI's GPT-4 for interpreting pharmacogenomic (PGx) testing results, aiming to improve decision-making and knowledge sharing in clinical genetics and to enhance patient care with equitable access. MATERIALS AND METHODS: The AI assistant employs retrieval-augmented generation (RAG), which combines retrieval and generative techniques, by harnessing a knowledge base (KB) that comprises data from the Clinical Pharmacogenetics Implementation Consortium (CPIC). It uses context-aware GPT-4 to generate tailored responses to user queries from this KB, further refined through prompt engineering and guardrails. RESULTS: Evaluated against a specialized PGx question catalog, the AI assistant showed high efficacy in addressing user queries. Compared with OpenAI's ChatGPT 3.5, it demonstrated better performance, especially in provider-specific queries requiring specialized data and citations. Key areas for improvement include enhancing accuracy, relevancy, and representative language in responses. DISCUSSION: The integration of context-aware GPT-4 with RAG significantly enhanced the AI assistant's utility. RAG's ability to incorporate domain-specific CPIC data, including recent literature, proved beneficial. Challenges persist, such as the need for specialized genetic/PGx models to improve accuracy and relevancy and addressing ethical, regulatory, and safety concerns. CONCLUSION: This study underscores generative AI's potential for transforming healthcare provider support and patient accessibility to complex pharmacogenomic information. While careful implementation of large language models like GPT-4 is necessary, it is clear that they can substantially improve understanding of pharmacogenomic data. With further development, these tools could augment healthcare expertise, provider productivity, and the delivery of equitable, patient-centered healthcare services.