Browsing by Subject "manufacturing"

Sort by: Order: Results:

Now showing items 1-3 of 3
  • Lindevall, Mari (Helsingin yliopisto, 2021)
    The purpose of this systematic review is to investigate the usage of artificial intelligence in the pharmaceutical industry in the fields of pharmaceutical manufacturing, product development, and quality control. Today, developing and getting a new drug on the market is time-consuming, ineffective, and expensive. Artificial intelligence is seen as one possible solution to the problems of the pharmaceutical industry. From 734 articles 77 academic study articles were included. Included articles showed artificial neural networks to be the most used artificial intelligence method between 1991 and 2021. The search was conducted from three databases with the following inclusion criteria: studies using AI in either pharmaceutical manufacturing, product development or quality control, English as the language, and Western medicine-based pharmacy as a branch of science. This systematic literature review has three main limitations: the possibility of an important search word missing from the search algorithm, the selection of articles according to one person's assessment, and the possible narrow picture of the used artificial intelligence methods in the pharmaceutical industry, as pharmaceutical companies also research the subject. The use of artificial intelligence in product development has been studied the most, while its use in quality control has been studied the least. In the studies, tablets were a popular drug form, while biological drugs were underrepresented. In total, the number of studies published increased over three decades. However, most of the articles were published in 2020. Nearly half of the articles had some connection to a pharmaceutical company, indicating the interest of both the academy and pharmaceutical companies in the use of artificial intelligence in manufacturing, product development, and quality control. In the future, the efficacy of artificial intelligence, as well as its limitations as a method, should be investigated to conclude its potential to play a key role in reforming the pharmaceutical industry. The results of the study show that a wave of artificial intelligence has arrived in the pharmaceutical industry, however, its real benefits will only be seen with future research.
  • Kuryk, Lukasz; Moller, Anne-Sophie W.; Vuolanto, Antti; Pesonen, Sari; Garofalo, Mariangela; Cerullo, Vincenzo; Jaderberg, Magnus (2019)
    Oncolytic adenoviruses can trigger lysis of tumor cells, induce an antitumor immune response, bypass classical chemotherapeutic resistance strategies of tumors, and provide opportunities for combination strategies. A major challenge is the development of scalable production methods for viral seed stocks and sufficient quantities of clinical grade viruses. Because of promising clinical signals in a compassionate use program (Advanced Therapy Access Program) which supported further development, we chose the oncolytic adenovirus ONCOS-401 as a testbed for a new approach to scale up. We found that the best viral production conditions in both T-175 flasks and HYPERFlasks included A549 cells grown to 220,000 cells/cm(2) (80% confluency), with ONCOS-401 infection at 30 multiplicity of infection (MOI), and an incubation period of 66 h. The Lysis A harvesting method with benzonase provided the highest viral yield from both T-175 and HYPERFlasks (10,887 +/- 100 and 14,559 +/- 802 infectious viral particles/cell, respectively). T-175 flasks and HYPERFlasks produced up to 2.1 x 10(9) +/- 0.2 and 1.75 x 10(9) +/- 0.08 infectious particles of ONCOS-401 per cm(2) of surface area, respectively. Our findings suggest a suitable stepwise process that can be applied to optimizing the initial production of other oncolytic viruses.