ARTIFICIAL INTELLIGENCE W
e’ve seen how the cost of developing a new drug has doubled in the past 10 years, and, according to the Tufts
Center for the study of Drug Development, the development of a new pharmaceutical product, and its introduction into the market, is estimated to exceed $2.5bn, nearly 75% of which is spent in the various phases of clinical development. Nevertheless, cost is not the only current challenge for clinical research in pharma companies, where the pipeline is getting bigger and bigger. Only one of 10 drugs that are studied in a clinical trial reaches the market – and, in my opinion, there are at least three important factors in play here. First, the development of a new molecule takes time and is impacted by the techniques of selection and recruitment of patients that lengthen the achievement of the last person in (LPI) the study, and prolong the period of research and development of drugs that are directed to a patient niche. In the post-blockbuster drug age, a lack of go-to-market efficiency is not sustainable. Second is the clinical trial design. The ‘one third rule’ is applicable at this point: one third of Phase II trials fail to move to Phase III, and less than one third of Phase III achieve their results for approval by regulatory authorities. And here, the clinical trial design is one of the reasons why we fail. The last relevant factor in this environment is the ability to monitor the clinical trial efficiently in the absence of reliable and efficient adherence control, patient monitoring, and clinical end-point detection systems. In this sense, pharmaceutical companies have sought some options that can improve efficiency and agility when handling research. Among these
“An advantage of working with a ‘virtual’ trial could be to refine our clinical trial through more detailed information on potential outcomes and greater explanatory power in interpreting any adverse effects that might happen.”
46 | Outsourcing in Clinical Trials Handbook
solutions, we can mention, for example, new ways of working on monitoring, patient-centric and customer-centric approaches; collaboration between health authorities and pharmaceutical companies; new clinical trial designs; and the introduction of solutions provided by artificial intelligence (AI).
The role of AI in health began in the 1970s with systems for diagnostic decision support, but in recent years this situation has changed. This change has been motivated by, first of all, the transformational advances, particularly in deep learning and related machine learning methods, and second, medical data has become increasingly available in digital forms. The role of AI is broad, and the use of algorithms is a reality in medical practice, facilitating the diagnosis of diseases, such as melanoma through images of the skin, and neurodegenerative diseases through eye movement.
The use of AI algorithms is increasing. Deep learning has been remarkably successful in identifying potential new drug candidates and improving the prediction of their properties and the possible safety risks.
In terms of personalised medicine, AI can
improve the efficiency of the search for correlation between indications and biomarkers. Computer modelling and simulation is already being used in the development of biomedical products. During the process of a clinical trial, AI offers the option of transforming crucial steps of clinical trial conduct-study design, planning and execution, with the aim of facilitating efficiency during monitoring using large and diverse datasets, such as electronic health records, medical literature and trial databases for improved patient trial matching for selection criteria and recruitment. AI can help in enhancing patient selection by reducing population heterogeneity, prognostic enrichment (by selecting patients who exhibit a higher probability of having a measurable clinical end point), and, by predictive enrichment, choosing a population with a better likelihood of responding to a treatment.
Machine learning and deep learning are able to find patterns of meaning in large datasets. Natural language processing can understand and correlate content in written or spoken language, and human machine interfaces allow for a natural exchange of information between computers and humans. AI techniques, in combination with wearable
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84 |
Page 85 |
Page 86 |
Page 87 |
Page 88 |
Page 89 |
Page 90 |
Page 91 |
Page 92 |
Page 93 |
Page 94 |
Page 95 |
Page 96 |
Page 97 |
Page 98 |
Page 99 |
Page 100
