Supporting Open Innovation in Pharmaceuticals
Curator: Larry H. Bernstein, MD, FCAP
Compound Passport Service: supporting corporate collection owners in open innovation
David M. Andrews1 ,Sebastien L. Degorce1 , David J. Drake2 , Magnus Gustafsson3 , Kevin M. Higgins2 and Jon J. Winter1
1 Oncology iMed Chemistry, AstraZeneca, Mereside, Alderley Park, Macclesfield SK10 4TF, UK 2 R&D Information, AstraZeneca, Mereside, Alderley Park, Macclesfield SK10 4TF, UK 3 R&D Information, AstraZeneca, Pepparedsleden 1, 431 83 Molndal, Sweden
Drug Discovery & Innovation 2015; 20(10). http://www.e-ditionsbyfry.com/olive/ODE/DDT/Default.aspx?href=DDT/2015/10/01
A growing number of early discovery collaborative agreements are being put in place between large pharma companies and partners in which the rights for assets can reside with a partner, exclusively or jointly. Our corporate screening collection, like many others, was built on the premise that compounds generated in-house and not the subject of paper or patent disclosure were proprietary to the company. Collaborative screening arrangements and medicinal chemistry now make the origin, ownership rights and usage of compounds difficult to determine and manage. The Compound Passport Service is a dynamic database, managed and accessed through a set of reusable services that borrows from social media concepts to allow sample owners to take control of their samples in a much more active way.
The challenges of discovering and developing novel therapeutics have been well documented [1]; and the combination of the ‘low hanging fruit’ of drug targets having been picked off [2] along with the challenge to maintain the pace of new discovery has led to an increase in the complexity of targets and disease pathways in discovery portfolios. Additionally, the pharmaceutical industry has realigned resources away from early R&D [3], making industry more reliant on collaboration with academic groups to share the risks (and rewards) of conducting discovery and early validation efforts [4].
These efforts are frequently captured under the generic term ‘open innovation’, first coined by Henry Chesbrough in 2003 [5]. Since then, a huge variety of definitions for open innovation have been suggested; the authors prefer the definition adopted by the Wellcome Trust: ‘The process of innovating with others for shared risk and reward to produce mutual benefits for each organisation, creating new products, processes or ideas that could not otherwise have been achieved alone, or enabling them to be achieved more quickly, cheaply or efficiently’ [6].
We found that in AstraZeneca (AZ), as is the case across the pharmaceutical industry [7], many more collaborative agreements were being put in place in which the rights for assets could reside with a partner, exclusively or jointly. With more opportunities being investigated to take advantage of our in-house assets, we needed to improve our ability to ensure that compounds subject to a contractual agreement with third parties are managed and used in accordance with AZ’s obligations. Such agreements could mean compounds should be restricted to agreed tests and/or prevented from being shared with other parties.
The problem: compound rights tracking Many corporate screening collections have been built on the premise that collection members that had been generated in house and were not the subject of paper or patent disclosure were proprietary to the company. Collaborative screening arrangements [8] and medicinal chemistry [9,10] now make the origin, ownership rights and contractually governed usage of compounds difficult to determine and complex to manage. When we looked at the compound management tools available within our own organization (or those available from vendors) we found that, in general, solutions were monolithic one-size-fits-all packages and lacked the information granularity necessary to answer key questions around compound-asset rights: compound and sample restrictions were either single sample or all examples from a project; delegation of approval was difficult and all approval was manual; approval tracking and compliance monitoring was difficult and error-prone; in consequence it was difficult to provide partners with a record of requests for ‘their’ samples.
The root cause was that the design of the compound asset infrastructure predated the emergence of shared risk, collaborative, open innovation projects and the infrastructure had been designed against a background where there was a presumption that a company could solely own the rights to the portion of its compound library that was not in the public domain. Together, this created the risk that AZ scientists could unwittingly release the structures of early-stage hits to collaborators that were already the subject of an agreement with another collaborator. Put simply, the infrastructure of material transfer agreements and confidential disclosure agreements was very good at tracking the supply of single compounds but could not cope with the tens to thousands of compounds that needed to be correctly tracked as a result of open innovation collaborations. We aimed to design a dynamic database, managed and accessed through a set of reusable services that borrowed from social media concepts to allow sample owners to take control of their samples in a much more active way. Our design is driven by the vision that a greater number of collaborative agreements are being put in place and that, within those agreements, compound rights can be shared or reside with AZ or the collaboration partner. In turn this drove the need to improve our ability to keep our contractual agreements and progress compounds through approval flows in a timely and efficient manner. In addition to making it easy to record and update details of collaborations, we wanted simply and quickly to add, edit and remove compound assets, as well as to provide fast, reliable and automated approval where possible or to alert approvers where a manual approval is required. Finally, we wanted scientists to be able to determine compound status easily; able to request approval to take specific actions (e.g. testing) within the context of a system that maintained a permanent record.
For the service to function correctly, for each open innovation compound, three pieces of information (metadata) need to be captured and kept up-to-date: (i) who owns any rights to the compound – AZ, the partner or are the rights shared? (ii) Can the compound be tested freely within AZ or does the collaboration agreement indicate that a collaboration coordinator needs to approve test requests? (iii) Has the compound been provided to the partner structure-blinded or has the compound structure been shared with the partner? The service can control ‘trafficking’ as well as maintain a permanent compound ‘life history’. It can be interrogated and receive updates via calls from other systems. Hence, we refer to it as the Compound Passport Service (CPS). Our shared vision and understanding of the underlying metadata allowed us to formulate the main concepts of the system and associated ‘use cases’.
System concepts and use cases CPS centres on assets, which are entities reflecting agreed constraints of compound usage by third parties and are grouped together in asset rights groups (ARG). Additionally, an ARG is a group of rights and rules applied to a number of assets. For each ARG a number of roles can be defined: coordinator – sets up and manages the ARG; delegate – a deputy for the coordinator who can add or remove approvers and add or update assets; approver – a person able to approve or reject requests manually, when the system is unable to make an automatic decision. For each ARG it is also possible to delineate request rules that define which requests can be automatically approved or rejected (e.g. that all compounds within an ARG can be tested in a specific test without any manual approval needed). A user in another system that is fully integrated with CPS becomes a requester when asking for approval to perform a specific action, for example to run a test on a specific compound and CPS responds with the following: ‘approved’, ‘rejected’ or ‘manual approval needed’ (Fig. 1). This structure provides a framework that allows users to take control of their compounds in real-time and in a very granular way. It also has the potential to speed up the flow of compounds through the design-make-test-analyse (DMTA) cycle [11] by giving users the option to set up rules for automatic approval or rejection of tests. To enable these goals, we considered seven critical scenarios (use cases) that the system needed to service (Table 1).
FIGURE 1 Compound Passport Service (CPS) concepts and definitions. The CPS is based upon assets, which are entities reflecting agreed constraints of compound usage by third parties, and are grouped (together with the rights and rules applying to the assets) in asset rights groups (ARG). For each ARG a number of roles are defined: coordinator – sets up and manages the ARG; delegate – a deputy for the coordinator who can add or remove approvers and add or update assets; approver – a person able to approve or reject requests manually. Within each ARG it is also possible to delineate request rules that define which requests can be automatically approved or rejected. A user in another system that is fully integrated with the CPS becomes a requester when asking for approval to perform a specific action; the CPS responds with the following: ‘approved, ‘rejected’ or ‘manual approval needed.
More-efficient approval flows and search
The CPS was designed to manage complex compound sharing rules and requirements over the entire lifecycle of a collaboration project. The following is a case history of a recent project that was used to help design a system with the flexibility required. A collaboration project had two chemical series: A and B. Series A originated from screening of the AZ compound collection, and samples were tested by the partner organization in a structure blind format. Synthetic optimisation was performed by AZ but the compounds were never judged to be of sufficient selectivity to merit sharing with the partner. Owing to their origin in a collaboration project, however, series A compounds were not permitted to be tested outside the originating project. Later, the compounds were of no further interest to the project and permitted to be used by AZ projects for any purpose.
Series B originated from the partner organization, and samples were initially shared for testing by AZ in a structure-blind format. After some months of optimization, the chemical structures were shared with AZ but the ownership of the compounds was retained by the partner. Later still, the series was judged of sufficient quality for the intellectual property to be shared jointly between the two organizations. Finally, after the biological target hypothesis was invalidated, notional ownership of the compounds was returned to the partner organization and no further testing was permitted by AZ (Fig. 2).
FIGURE 2 The Compound Passport Service (CPS) can be used to manage an asset throughout its lifecycle. In this case study, at the start of the collaboration, compounds were tested at AstraZeneca (AZ) and the collaborator structure-blinded. Over the course of research, SAR failed to develop in series A which became of no further interest and was retained by AZ. Series B provided very productive SAR and, through the course of the collaboration, compound properties were updated in the CPS to note initially that the structures had been shared with AZ, then ownership became shared and therefore the series B compounds were unlocked and freely available for test across both organisations (the figure shows the initial collaboration status).
All of the compound status changes in this scenario are mapped on to three key properties that are captured by the CPS (Table 2). The owner of the ARG is able to change the properties of individual or groups of compounds as required, independently, as the project evolves. Such changes are recorded with time stamps as ‘transitions’, and can be tracked over the lifetime of an ARG. The ability to track all such transitions increases the transparency of compound ownership to AZ and partner organizations, prevents un-authorised testing of samples by other project teams within AZ and enables questions of the provenance of compounds to be easily resolved. When placing test requests in a dedicated in-house requesting tool, the CPS is called and the sample status is displayed. Users might be presented with a warning that corresponding samples are subject to approval. Depending on the permission status, the requester can either decide to seek approval or cancel their tentative requests. Extreme cases such as auto-rejection (strictly no testing) or auto-approval (green list) are dealt with instantaneously, whereas manual approvals are immediately sought with daily reminders being sent to the approver until a response is obtained. A feature much appreciated by users is that the system sends an approval email to authorisers giving them the full context of the requests to ensure efficient decision making rather than having to access the CPS to gain the wider context.
Green lists (leading to auto-approvals) are crucial to ensure no impact of sample restrictions on the DMTA cycle. In the case of required manual approvals, the impact of delay is minimised by the creation of project delegates. Green list assays refer to tests agreed with the partner and typically include project assays [primary target, selectivity, in vitro drug metabolism pharmacokinetics (DMPK), among others]. Similarly, a green list of requesters typically includes project members who are fully aware of the collaboration. Delegates have the same approval rights as primary authorisers and requests come to all independently, so that the first person to approve them releases the samples for testing.
Interface with other services The envisioned central role and future extensibility of asset rights management led to the rapid conclusion that the compound passport solution needed to be delivered as a service [12]. The adoption of a service-orientated architecture has provided a flexible and reusable set of business services that provide access to and management of the compound passport database. Additional commodity services provide master data for projects, people and compounds (Fig. 3).
FIGURE 3 Service-orientated architecture provides a flexible and reusable set of business services that provide access to and management of the Compound Passport Database. Additional commodity services provide master data for projects, people and compounds. In this way, the Compound Passport Service (CPS) sits at the centre of a web of independent services controlling compound registration, distribution and test approval. Utilisation of reusable services allows integration with new systems as they become available in the future.
….
Impact on open innovation Since the CPS was launched in 2014, it has been uploaded with metadata relating to 15,000–20,000 assets, and compounds are being added almost daily. The ability to track ownership rights along with more-efficient test approval has already enabled faster and more-efficient approval flows. Additionally, in considering whether to unblind the structure of a HTS hit series, we have also been able to identify that more than one external party had an interest in the chemical equity. Based upon an understanding that SAR would probably diverge as potency against the different targets was optimised, we have been able to adopt a risk-management approach to allow both partners to proceed with the investigation and possible optimization of the shared chemical start point.
……
2012pharmaceutical
Amandeep Kaur
Aashir Awan, Phd
Abhisar Anand
Adina Hazan
Alan F. Kaul, PharmD., MS, MBA, FCCP
alexcrystal6
anamikasarkar
apreconasia
aptubman
Arav Gandhi
aviralvatsa
David Orchard-Webb, PhD
danutdaagmailcom
Demet Sag, Ph.D., CRA, GCP
Dror Nir
dsmolyar
Ethan Coomber
evelinacohn
Gail S Thornton
Irina Robu
jayzmit48
jdpmd
jshertok
kellyperlman
Ed Kislauskis
larryhbern
Madison Davis
marzankhan
megbaker58
ofermar2020
Dr. Pati
pkandala
ritusaxena
Rick Mandahl
sjwilliamspa
Srinivas Sriram
stuartlpbi
Dr. Sudipta Saha
tildabarliya
zraviv06
zs22