Background and Objectives: The use of artificial intelligence (AI) and, in particular, machine learning (ML) techniques is growing rapidly in the healthcare field. Their application in pharmacokinetics is of potential interest due to the need to relate enormous amounts of data and to the more efficient development of new predictive dose models. The development of pharmacokinetic models based on these techniques simplifies the process, reduces time, and allows more factors to be considered than with classical methods, and is therefore of special interest in the pharmacokinetic monitoring of antibiotics. This review aims to describe the studies that use AI, mainly oriented to ML techniques, for dose prediction and analyze their results in comparison with the results obtained by classical methods. Furthermore, in the review, the techniques employed and the metrics to evaluate the precision are described to improve the compression of the results. Methods: A systematic search was carried out in the EMBASE, OVID, and PubMed databases and the results obtained were analyzed in detail. Results: Of the 13 articles selected, 10 were published in the last three years. Vancomycin was monitored in seven and none of the studies were performed on new antibiotics. The most used techniques were XGBoost and neural networks. Comparisons were conducted in most cases against population pharmacokinetic models. Conclusions: AI techniques offer promising results. However, the diversity in terms of the statistical metrics used and the low power of some of the articles make the overall assessment difficult. For now, AI-based ML techniques should be used in addition to classical population pharmacokinetic models in clinical practice.

Therapeutic drug monitoring improves the benefit–risk balance of antipsychotic therapy. Ultra-high-performance liquid chromatography–tandem mass spectrometry (UHPLC-MS/MS) is considered the gold-standard method for measuring plasma drug concentrations; however, the Alinity C system has emerged as a promising alternative. This is the first study aimed at comparing UHPLC-MS/MS versus Alinity C in measuring plasma concentrations of aripiprazole and dehydroaripiprazole. A total of 86 plasma samples were analyzed. The active moiety of aripiprazole was measured in 60 samples using both systems and 26 samples were analyzed twice using Alinity C with an intermediate period of 6 months to assess its reproducibility. Spearman’s correlation revealed a good association between the two assays (rs = 0.96) and no significance differences were found by McNemar’s test when classifying samples between infra-, supra- and therapeutic ranges. Passing–Bablock regression showed a good correlation among methods (rs = 0.93) and a slope of 1.12 indicating a slight tendency of Alinity C to measure higher values than UHPLC-MS/MS. In addition, a good intra-method correlation across the two sequential analyses with Alinity C was obtained (rs = 0.99). Nonetheless, clinical decisions could be different in 15% of the cases depending on the chosen method. No differences were found in active moiety determination by Alinity C depending on the concentration of aripiprazole and dehydroaripiprazole of the samples.

Introduction: Age-related macular degeneration (AMD) is the leading cause of irreversible vision loss among people over 55 years of age globally, being neovascular AMD (nAMD) its most aggressive form. Its treatment consists of the use of drugs that block vascular endothelial growth factor (anti-VEGF). Proteomics may allow the identification of differentially expressed proteins between responders and non-responders to each anti-VEGF drug. Thus, the objective of Pharmacoproteomics in the development of personalised medicine in Age-related Macular Degeneration (PHARPRO-AMD) is to find new proteomic biomarkers, predictive of response to antiangiogenic treatment in patients with nAMD. Methods and analysis: PHARPRO-AMD is a nationwide, multicentre, prospective, observational study. Treatment-naïve patients with nAMD starting anti-VEGF therapy will be enrolled and followed up for 2 years. During this period, clinical variables will be gathered to classify treatment response. In addition, blood, tear and vitreous and aqueous humour samples will be collected and will undergo a ZenoSWATH proteomic analysis. Relevant biomarkers identified and response classification will be used to perform a multivariate logistic regression and construct receiver operating characteristic curves. Results: The study is expected to identify a panel of proteomic biomarkers predictive of anti-VEGF treatment response. Integrating data from invasive and non-invasive biological samples may enhance clinical applicability. Once validated, these biomarkers could support the design of future clinical trials on biomarker-guided therapies, helping to optimise treatment regimens and improve visual outcomes. Conclusions: The PHARPRO-AMD study aims to provide proof-of-concept for biomarker-guided anti-VEGF therapy in nAMD, potentially improving vision outcomes. A notable limitation is the exclusion of patients with visual acuity above 73 Early Treatment of Diabetic Retinopathy Study letters, a criterion chosen to reduce potential ceiling effects and improve response assessment accuracy. Ethics and dissemination: Approved by the Galician Network of Ethics Committees, with nationwide validity. Anonymised data will be deposited in open-access repositories and published in peer-reviewed journals. Trial registration number: Spanish Clinical Studies Registry (REec) (0033-2024-OBS).

Background and Objectives: The use of Artificial intelligence (AI) and, in particular, Machine Learning (ML) techniques is growing rapidly in the healthcare field. Their application in pharma-cokinetics is of potential interest due to the need to relate enormous amounts of data and to the more efficient development of new predictive dose models. The development of pharmacokinetic models based on these techniques simplifies the process, reduces time and allows more factors to be con-sidered than with classical methods, and is therefore of special interest in the pharmacokinetic monitoring of anti-biotics. This review aims to describe the studies that use AI, mainly oriented to ML techniques, to dose prediction and analyze their results in comparison with the results obtained by classical methods. Furthermore, in the review, the techniques employed and the metrics to evaluate the precision are described to improve the compression of the results. Methods: A sys-tematic search was carried out in the EMBASE, OVID and PubMed databases and the results ob-tained were analyzed in detail. Results: Of the 13 articles selected, 10 were published in the last three years. Vancomycin was monitored in 7 and none of the studies were performed on new an-tibiotics. The most used techniques were XGBoost and neural networks. Comparison was conducted in most cases against population pharmacokinetic models. Conclusion: AI techniques offer prom-ising results. However, the diversity in terms of the statistical metrics used and the low power of some of the articles make the overall assessment difficult. For now, AI-based ML techniques should be used in addition to classical population pharmacokinetic models in clinical practice.

The recombinant enzyme glutamate oxaloacetate transaminase (rGOT) has emerged as a promising neuroprotective candidate for acute ischaemic stroke. This study examined whether rGOT can attenuate the neurotoxic side effects of recombinant tissue plasminogen activator (rtPA) and enhance the therapeutic benefit of thrombolytic reperfusion. First, primary cortical neurons were exposed to rtPA with or without rGOT under normoxia or oxygen–glucose deprivation. rGOT did not reduce rtPA-induced cytotoxicity, indicating that the pathways underlying rtPA neurotoxicity are mechanistically distinct from those modulated by rGOT. Second, a thromboembolic mouse model was used to assess the interaction between rGOT and rtPA across clinically relevant treatment scenarios. rGOT alone or when administered prior to rtPA had no effect on infarct size or haemorrhagic transformation. However, co-administration of rGOT with rtPA significantly reduced infarct volume at 24 h ( p < 0.05), without increasing bleeding risk. These findings demonstrate that rGOT is compatible with thrombolytic therapy and can enhance rtPA efficacy when delivered during reperfusion, supporting its potential use as an adjuvant treatment in acute ischaemic stroke.

Objectives: To characterize the one-year functional, anatomical, and molecular responses to intravitreal Ranibizumab in treatment-naïve patients with neovascular age-related macular degeneration (nAMD), and to identify systemic immune and miRNA signatures associated with treatment response. Methods: This prospective longitudinal observational study included 44 treatment-naïve patients with nAMD. Patients received up to four monthly intravitreal Ranibizumab injections, followed by a treat-and-extend regimen. Best-corrected visual acuity using ETDRS letters, central retinal thickness by optical coherence tomography, fluorescein angiography, and OCT angiography were assessed at baseline and 12 months. Peripheral blood samples were collected at both time points to quantify seven circulating cytokines using an IMMULITE chemiluminescent immunoassay and to profile 37 candidate miRNAs by TaqMan OpenArray RT-qPCR from leukocyte-derived RNA. Treatment response was classified using composite anatomical and functional criteria, including intraretinal/subretinal fluid resolution, ≥25% central retinal thickness reduction, and a ≥5 ETDRS letter gain. Results: At one year, patients showed significant central retinal thickness reduction and overall visual stabilization, although good and poor responders differed according to composite response criteria. Statin use was numerically more frequent among poor responders, although this difference was not statistically significant. Soluble IL-2R increased significantly over time in the overall cohort, mainly driven by good responders who showed higher median levels at both visits. IL-8 also increased globally, without significant between-group differences. Among differentially expressed miRNAs, miR-3121 was the only candidate reaching statistical significance and was downregulated in good responders. ROC analysis showed moderate discriminative performance for miR-3121, with an AUC of 0.76. Conclusions: One-year response to Ranibizumab in nAMD may involve systemic immune activation and miRNA regulation. miR-3121 emerges as a candidate biomarker of treatment response, supporting further validation in larger independent cohorts.