Infrastructure development for building, maintaining and modeling indication- specific summary-level literature databases to support model based drug development.
Hugh McDevitt (1), Goonaseelan Pillai (1), Kai Wu (1), Ramprasad Ramakrishna (1), Gerard JP Flesch (1), Georges Ette (1), Sreeni Devidas (2), Rajasekhar Thatavarti (2), Vincent Buchheit (1).
(1) Novartis Pharma AG, Basel, Switzerland, 2) GVK Biosciences Private Ltd, Hyderabad, India
Background: Several Pharma Modeling and Simulation (M&S) departments and M&S Consulting companies have demonstrated the value of using models developed from summary level literature data to improve the efficiency of drug development [ Corrigan, Mandema]. Health authorities and Payors (such as NICE) are also very interested in the comparison of new therapies with existing treatments for healthcare cost containment. Any organization that intends using these databases as ‘legacy’ systems that are available for input and use by multiple users needs to take time to carefully consider the processes and infrastructure to ensure reliability and quality. At Novartis, an IT infrastructure that uses publicly accessible open source database systems is planned and will be presented in this poster.
Objectives: To solicit feedback and discussion on the establishment of an infrastructure for building, maintaining and modeling summary-level literature data from the published literature publications.
Methods: A clear 7 step process has been identified.
- Identify the key characteristic endpoint for an indication,.
- Identify the relevant publications
- Define the data specification for the indication
- Load the relevant data from the identified publications
- Quality control the loading of the data
- Publish the database
- Revise and update the database on an ongoing basis
Results: Several indication specific databases have been developed ( diabetes, respiratory ), while some have been commercially acquired. The partnership with GvKBio for data extraction has been established and has proved very effective.
Conclusions: A clear methodology has been defined and the effort required to set up an indication specific database has been established. The next step is to industrialize the process by establishing a central database that will hold all the information and support the acquisition process.
References:
Corrigan B. et al, (2008), Standardization of Data Collection for Literature Based Meta-Analyses: The Literature Information and Knowledge Explorer (LIKE) Initiative, Proceedings of the American Conference on Pharmacometrics, http://tucson2008.go-acop.org/schedule.php
Mandema J. W., et al, (2005), Model-Based Development of Gemcabene, a New Lipid-Altering Agent, AAPS Journal, http://www.aapsj.org
