Artificial Intelligence and Machine Learning Applied at the Point of Care.

INTRODUCTION: The increasing availability of healthcare data and rapid development of big data analytic methods has opened new avenues for use of Artificial Intelligence (AI)- and Machine Learning (ML)-based technology in medical practice. However, applications at the point of care are still scarce. OBJECTIVE: Review and discuss case studies to understand current capabilities for applying AI/ML in the healthcare setting, and regulatory requirements in the US, Europe and China. METHODS: A targeted narrative literature review of AI/ML based digital tools was performed. Scientific publications (identified in PubMed) and grey literature (identified on the websites of regulatory agencies) were reviewed and analyzed. RESULTS: From the regulatory perspective, AI/ML-based solutions can be considered medical devices (i.e., Software as Medical Device, SaMD). A case series of SaMD is presented. First, tools for monitoring and remote management of chronic diseases are presented. Second, imaging applications for diagnostic support are discussed. Finally, clinical decision support tools to facilitate the choice of treatment and precision dosing are reviewed. While tested and validated algorithms for precision dosing exist, their implementation at the point of care is limited, and their regulatory and commercialization pathway is not clear. Regulatory requirements depend on the level of risk associated with the use of the device in medical practice, and can be classified into administrative (manufacturing and quality control), software-related (design, specification, hazard analysis, architecture, traceability, software risk analysis, cybersecurity, etc.), clinical evidence (including patient perspectives in some cases), non-clinical evidence (dosing validation and biocompatibility/toxicology) and other, such as e.g. benefit-to-risk determination, risk assessment and mitigation. There generally is an alignment between the US and Europe. China additionally requires that the clinical evidence is applicable to the Chinese population and recommends that a third-party central laboratory evaluates the clinical trial results. CONCLUSIONS: The number of promising AI/ML-based technologies is increasing, but few have been implemented widely at the point of care. The need for external validation, implementation logistics, and data exchange and privacy remain the main obstacles.

Regulatory aspects of pharmacokinetic profiling in special populations: a European perspective.

The clinical efficacy and safety profile of a new medicinal product is established in phase III studies, which are usually restricted to a well defined patient population. This population may not fully represent the population in which the product will be used once it is on the market. Pharmacokinetic studies in special populations are performed to estimate drug exposure in subpopulations of patients with characteristics that may affect drug exposure. The clinical consequences of altered exposure are then assessed, taking pharmacokinetic/pharmacodynamic relationships into consideration. If needed, specific treatment recommendations should be developed.Recommendations regarding pharmacokinetic characterization in special populations are given in a number of European guidelines. The pharmacokinetic characteristics, therapeutic window and intended use of the medicinal product influence the need for pharmacokinetic studies of a new medicinal product. There are a number of methodological issues to be considered when designing pharmacokinetic studies in special populations: the study design, study population and control group, the dosing regimen to be used, the analytes to be measured, and the distribution and range of the factor to be studied. The data should be presented in sufficient detail to enable assessment by regulatory authorities of the conducted analysis and conclusions drawn. Assessment of the data should include an evaluation of how and to what extent the pharmacokinetics in specific subpopulations deviate from the exposure at the therapeutic dose in the clinical efficacy and safety studies, and if there is a need for specific treatment recommendations. Based on the available information on the pharmacokinetic/pharmacodynamic relationships for efficacy and safety and/or the exposure at the therapeutic dose in the phase III population where efficacy and safety have been demonstrated, target criteria (a target exposure range) should be defined. Within the target exposure range, there should be no clinically relevant difference in efficacy and safety. Should the exposure in a specific group fall outside the defined target criteria, appropriate treatment recommendations need to be developed. The aim should be to develop dosing recommendations that will allow the majority of the patients to obtain exposure within the defined target range. If it is not possible to develop suitable dosing recommendations in a subgroup of patients, there may be a need for specific warnings or wordings regarding monitoring of the patients. It may also be an option to exclude that patient group from the indication. The resulting treatment recommendations should ensure safe and effective use of the drug in the entire population for which it has been approved.

CYP3A4 and MDR1 alleles in a Portuguese population.

Polymorphisms in cytochrome P450 CYP3A4 and multidrug resistance (MDR) 1 genes coding for the important drug-metabolising CYP3A4 and the ATP-binding cassette (ABC) transporter P-glycoprotein (Pgp) are poorly documented in the Portuguese population. In this study we have determined the frequencies of CYP3A4 and MDR1 alleles in Portuguese Caucasians. Both genes were simultaneously analysed as these genes are known to be frequently co-induced and their products to show a pronounced overlap of substrates. CYP3A4 A-392G (CYP3A4*1B), T673C (CYP3A4*2) and MDR1 T-129C, G2677T and C3435T single nucleotide polymorphisms (SNPs) were analysed in 100 individuals from the southern region of the country. We observed a frequency of 4.0% for CYP3A4*1B, not significantly different from that reported on other Caucasian European populations. CYP3A4*2 was found at an allele frequency of 4.5%, constituting the first report of the presence of this allele outside the Finnish population. Significant differences were found concerning the MDR1 C3435T SNP frequency (64.5%) compared with other European populations, while no differences were found concerning G2677T (47.5%) or T-129C (5%) SNPs. Linkage between the C3435T and G2677T SNPs was observed, although not as evidently as documented in other Caucasian populations. No preferential associations were detected between CYP3A4 and MDR1 alleles.