Clinical Pharmacology & Therapeutics最新文献

Tumor necrosis factor-α inhibitors (TNFαi) and Janus kinase inhibitors (JAKi) are widely used biologic and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) for rheumatoid arthritis. We aimed to compare the risk of serious infections between TNFαi and JAKi in patients with rheumatoid arthritis. We conducted a retrospective cohort study using a nationwide Japanese claims database covering over 19 million individuals. Patients were identified using the International Classification of Diseases, 10th Revision (ICD-10) codes (M05 or M06), and new users of bDMARDs or JAKi were included. We applied overlap weighting based on propensity scores. The primary outcome was serious infection, defined as a hospitalization-associated infection accompanied by diagnostic testing or anti-infective treatment. Among 5,018 eligible rheumatoid arthritis patients (TNFαi: 4042; JAKi: 976), TNFαi use was associated with a significantly lower risk of serious infections (HR: 0.55; 95% confidence interval: 0.39-0.77). The E-value for the observed HR was 2.39, suggesting that unmeasured confounding would need to be strongly associated with both treatment assignment and infection risk to fully explain the result. Site-specific analyses showed significantly lower risks of respiratory tract infections, urogenital infections, and sepsis in the TNFαi group. Drug-specific analyses indicated an elevated infection risk with baricitinib and infliximab. In this study, TNFαi use was associated with a lower risk of serious infections than JAKi in patients with rheumatoid arthritis. These findings support the preferential use of TNFαi in infection-prone populations and highlight the importance of individualized risk assessment in rheumatoid arthritis treatment decisions.

In recent years, different national and international regulatory authorities, notably the FDA, have made their adverse event repositories publicly available, offering user-friendly dashboards. This has led to a large increase in low-quality, poorly reported research using adverse event reporting databases (e.g., FDA’s Adverse Event Reporting System [FAERS]). Such publications producing thousands of statistical associations erroneously presented as “safety signals” can create unscientifically grounded alarm with considerable impact on healthcare provider practices and patient behaviors.

Pregnancy loss after frozen embryo transfer (FET) remains an under-recognized complication in assisted reproductive technology (ART), particularly in non-Western settings. Existing prediction models rarely capture evolving hazards over gestation or externally validated. To address this, we developed and externally validated a parametric time-to-event model to predict the dynamic risks of pregnancy loss following FET using a large real-world Chinese cohort. The model incorporated routinely available clinical variables, including maternal demographics, progesterone concentration, endometrium thickness, and endometrium preparation protocols. A Gompertz distribution model captured the time-dependent hazard trajectory. Model performance was externally validated in two temporal cohorts, as well as in older women (≥40 years) and ethnic minority subgroups. Clinical utility was assessed by decision-curve analysis (DCA). Among 21,242 conceptions, 17.6% resulted in miscarriage or stillbirth, with risks peaking in the first trimester. Maternal age showed a nonlinear effect: compared with reference age 32, hazard increased by 67% at age 40 and more than doubled by age 45. Higher weight accelerated risk over time. Protective factors included greater endometrial thickness, higher progesterone levels, and use of mild- (hazard ratio, HR 0.80, 95% CI: 0.753-0862), late- (HR 0.90, 95% CI: 0.798-0.951), or modified-late stimulation (HR 0.80, 95% CI: 0.732-0.863) protocols. DCA indicated that model-guided risk-stratification strategies offered clear net benefit in older women across a broader range of risk threshold. This validated, time-resolved model supports individualized risk-stratification strategies and informed treatment decisions in FET-based ART across diverse clinical settings in China.

Radiolabeled human mass balance studies are crucial for identifying circulating metabolites and understanding drug absorption, excretion, and clearance pathways. Metabolite profiling involves quantifying all drug-related entities, including parent drug and metabolites in plasma and excreta, using extended liquid chromatography methods coupled with detection through scintillation counting, accelerator mass spectrometry, or non-radiolabeled approaches. Given the labor-intensive nature of sample extraction and analysis, we propose a new paradigm that maximizes gathering information through sample pooling strategies. Our proposal introduces sample pooling strategies by integrating both individual and pooled sample schemes, simplifying decisions, and consolidating existing knowledge into a cohesive document. This aligns with the low statistical power typically associated with mass balance studies that dose six to eight subjects. In metabolite profiling, it is common practice to pool samples either from the limited number of subjects participating in a human mass balance study or from different time points of sample collection. This approach improves efficiency while preserving data integrity. Pooling reduces resource constraints and enables the concentration of samples with relatively low radioactivity levels, resulting in higher quality metabolite profiles. Nevertheless, there are situations when analyzing samples from individual subjects or time points may be preferred. This proposal presents guidance and decision trees designed to facilitate informed decisions about sample pooling to maximize data quality of metabolite profiling in human mass balance studies while efficiently managing resources. These recommendations stem from discussions within the mass balance working group of the IQ Consortium.

Axatilimab, a monoclonal antibody targeting colony-stimulating factor 1 (CSF-1) receptor, is approved in the US for the treatment of chronic graft-versus-host disease (cGVHD) after failure of ≥2 lines of systemic therapy. In this study, the effects of antidrug antibody (ADA) status on the pharmacokinetics, pharmacodynamics, efficacy, and safety of axatilimab were evaluated. Among the 319 assessable participants, including 276 participants with cGVHD, 109 (34.2%) were treatment-emergent ADA-positive, with ADA onset occurring within the first month of treatment for 60.8% of treatment-induced ADA-positive participants. Among the 93 patients with cGVHD and treatment-emergent ADAs, 47 (50.5%) patients were neutralizing antibody (NAb)-positive, with NAb onset mostly occurring within the first 6 months of treatment. Samples that were ADA-negative were generally associated with higher axatilimab trough concentrations compared with ADA-positive samples, especially at later time points. Changes in CSF-1 based on ADA status were consistent with changes in axatilimab exposure. In contrast, nonclassical monocyte levels were similar regardless of ADA or NAb status, suggesting that immunogenicity did not affect nonclassical monocytic cell concentrations. No differences in overall response, ≥7-point improvement in modified Lee Symptom Scale responses, or duration of response were observed between the different ADA and NAb subgroups. The incidence of safety events was generally consistent across ADA and NAb subgroups. In conclusion, no meaningful differences in pharmacokinetics, pharmacodynamics, efficacy, and safety were observed between subgroups of axatilimab-treated participants with different ADA statuses. These results suggest that the development of ADAs and NAbs should not affect treatment decisions with axatilimab.

With the growth of African national regulatory agencies into a Pan-African entity called the African Medicines Agency (AMA), an opportunity exists for using the best regulatory science methods available. AI-supported model-informed drug development could be the backbone to support regulatory decision-making efforts. The pharmacometrics community can play a pivotal role in advancing this state-of-the-art opportunity. But, to do this successfully, pharmacometricians must gain the trust of African regulators, clinicians, and patients in using model-based methods.

Multiple imputation is well-established for handling missing data, yet its use in high-dimensional genetic datasets remains limited. Using pharmacokinetic tuberculosis simulations and SNP data (1000 Genomes Project), we compared machine learning (ML) and traditional approaches (e.g., mean imputation and complete-case analysis) for imputation and covariate selection. We developed a multiple imputation framework incorporating genotype probabilities, imputation uncertainty (INFO score), and missingness percentages. Dimensionality reduction enabled scalable random forest and penalized regression for covariate selection. In simulations, only multiple imputation achieved adequate coverage (percentage of 95% confidence intervals containing the true value) exceeding a 90% nominal threshold. For example, on the imputation server, coverage improved from 0% with single imputation to up to 94% under 10% missingness. Applied to clinical warfarin datasets (War-PATH, n = 548; IWPC, n = 316) and the UK Biobank (n = 500, 1000), multiple imputation recovered known pharmacogenomic associations (CYP2C9*8/*9/*11; VKORC1 -1639G>A), reduced false-positives, and detected signals missed by single imputation (e.g., genome-wide significant rs4697699, SLC2A9 locus). Computational costs were modest, adding only ~1.25 minutes for 10 imputations to the 22.7 minutes required by single imputation on the Michigan Imputation Server. For SNP selection, penalized regression performed best in the high-effect scenario (F1 = 0.897 ± 0.091), while GWAS followed by random forest performed best in the low-effect scenario (F1 = 0.657 ± 0.110). These findings show that multiple imputation improves reliability and discovery in high-dimensional pharmacogenomics, with ML offering promising but inconsistent benefits during SNP selection. However, generalizability beyond the studied datasets and computational scalability to larger biobank-scale analyses remain important limitations that warrant further investigation.

Effective perioperative antibiotic prophylaxis is critical in oral cavity cancer surgery. Despite metronidazole's high oral bioavailability, data on its tissue-specific distribution in reconstructive head and neck surgery remain scarce. In this randomized clinical microdialysis study, 18 patients undergoing facial artery musculomucosal (FAMM) flap reconstruction after oral cancer surgery were enrolled and randomized to receive 500 mg metronidazole twice daily, either orally (Group PO) or intravenously (Group IV). Samples were collected during the third dosing interval with microdialysis catheters placed in the FAMM flap, donor buccal submucosa, and subcutaneous neck tissue. The primary end points were the tissue-specific AUC_0-24h/MIC ratio and attainment of the treatment target (AUC_0-24h/MIC ≥ 70). The mean AUC_0-24h/MIC ratio for MIC 0.5 μg/mL ranged from 437 to 553 in Group IV and 423 to 531 in Group PO; for MIC 2 μg/mL, ratios ranged from 109 to 138 and 106 to 133 for Group IV and PO, respectively; and for MIC 4 μg/mL, ratios ranged from 55 to 69 and 53 to 66 for Group IV and PO, respectively. No statistically significant differences were observed between groups. All tissues in both groups achieved the AUC_0-24h/MIC ≥ 70 treatment target for MIC 0.5 and 2 μg/mL, but none of the tissues in any of the groups achieved the treatment target for MIC 4 μg/mL. In conclusion, both oral and intravenous metronidazole administration provided adequate tissue concentrations to reach the treatment target AUC_0-24h/MIC ≥ 70 for MIC values of 0.5 and 2 μg/mL, covering prevalent anaerobic pathogens. These findings support the clinical interchangeability of administration routes for metronidazole.

The impact of CYP2C19 genotype in relation to clinical risk is unclear during clopidogrel treatment following drug-eluting stent (DES) implantation. This study aimed to evaluate the prognostic significance of CYP2C19 genotypes based on clinical risk stratification in DES-treated patients. From the nationwide multicenter PTRG-DES (Platelet function and genoType-Related long-term progGosis in DES-treated patients) consortium, patients were classified according to the presence of CYP2C19 loss-of-function (LoF) allele: rapid or normal metabolizers (RMs/NMs) vs. intermediate or poor metabolizers (IMs/PMs), and clinical risk was stratified using the CHADS-P₂A₂RC and TRS 2°P scores. The primary endpoint (1°EP) was a composite of cardiac death, myocardial infarction, and stent thrombosis during a 3-year follow-up. Among clopidogrel-treated patients with CYP2C19 genotyping (n = 8,163), IMs/PMs (62.1%) demonstrated an increased risk of 1°EP compared with RMs/NMs (hazard ratio [HR]: 1.48; 95% confidence interval [CI]: 1.05-2.07; Log-rank P < 0.001), Most notable in those with high CHADS-P2A2RC (≥ 4) and TRS 2°P (≥ 3) scores (HR_adj: 1.68; 95% CI: 1.01-2.80; P = 0.047 and HR_adj: 1.63; 95% CI: 1.05-2.54; P = 0.029, respectively). In patients with low scores, there was no difference in 1°EP between IMs/PMs vs. RMs/NMs; however, an interaction was observed between acute and chronic coronary syndromes for both low CHADS-P₂A₂RC (HR_adj: 2.12; 95% CI: 1.11-4.03 and HR_adj: 0.68; 95% CI: 0.34-1.36; P_interaction = 0.017) and TRS 2°P scores (HR_adj: 2.34; 95% CI: 1.07-5.12 and HR_adj: 0.52; 95% CI: 0.22-1.17; P_interaction = 0.008). Among clopidogrel-treated patients, the carriage of the CYP2C19 LoF allele was associated with higher ischemic risk, particularly in those with high clinical risk or an acute coronary syndrome presentation.