Through more effective capture and reuse of data and knowledge, we discuss how Enterprise ELNs improve inter- and intra-departmental collaboration, and support quality initiatives such as ‘Quality by Design’.  ELNs were originally intended to reproduce the paper-based lab notebook process that has been the basis of data capture and scientific Intellectual Property (IP) protection for centuries. They have evolved through widespread use into enterprise systems: sophisticated data capture, management and reporting solutions capable of spanning a wide variety of departments/domains1. The resulting Enterprise ELNs are able to support collaboration not only within teams, but now between teams as well. Each individual team has a solution based on its specific needs, but there is a common core allowing cross-domain data sharing and therefore more effective collaboration.

The early ELN systems focused on electronically capturing results from a laboratory with data from Excel®, PowerPoint®, emails and domain-specific applications, and were often in a chemistryfocused environment2. Such systems were only orientated towards the documentation and capture of patent-relevant IP and essentially replaced using paper-based notebooks with cutting and pasting information electronically, providing a ‘sticker book’ approach. These early systems had little regard for data and knowledge sharing between teams, or with reuse of corporate learning, because they were only designed to provide ‘page’-like reports: they did not provide a rich data source that could be easily mined and shared.

Over the past five years the use of ELNs has spread into biology-related disciplines such as pharmacology, drug metabolism and pharmacokinetics (PK), and process-driven disciplines such as bioanalysis, formulations, analytical and pilot manufacturing. To support such diverse needs, ELNs have developed into enterprise systems that provide more flexible and broader data capture and sharing and IP capture. In this context, IP means corporate knowledge or data, ie important historical data and know-how that could be a potentially competitive advantage. Certain ELNs support both contextual and structured fact data capture that enables the creation of a Corporate Knowledge Store – a searchable, mineable resource that can be leveraged across the organisation.

The remainder of this paper will look in more detail at how these advanced Enterprise ELNs can deliver significant business impact, and what capabilities are required for their application to both large and small molecule preclinical development organisations.

Supporting interdepartmental workflows in preclinical development
Each department in preclinical development needs to be supported by efficient data capture and knowledge management, to enable information and results to be used in project teams and regulatory documentation. A successful Investigational New Drug (IND) application requires a large range of department level solutions, each delivering specific elements of data and interpretation (see Figure 1). Each discipline needs to be supported, and each has different requirements of what an ELN must do and how it should work, for example:

l Formulations need to capture the design of batches, processes and dosage forms, calculations, test results and batch records, and deliver crossdomain reports to project teams.

l Analytical groups need to manage requests for testing, capture method development and validation details, integrate with clinical data systems (CDS) and laboratory informatics management systems (LIMS), and provide structured data capture and reporting of results.

l Pharmacokinetics (PK) need to manage studies, co-ordinate dosing and sample collection, integrate with instruments, import and manipulate data, provide summary analysis and integrate with domain tools such as WinNonlin®.

At the R&D macro level, using the same ELN platform for each area means that departmental solutions can combine to form a comprehensive view of compound/biologic progression. It also enables each discipline to see not only their own data, but also the relevant information from other departments. With Enterprise ELNs, both the capture and viewing of data can be tailored towards each user, based on their specific department/ domain needs (see Figure 2). For example, a formulator looks at results from a batch perspective, but a bioanalyst who conducts the work sees the data from a worksheet and test perspective. This means that the concept of good data management is of paramount importance. Systems must consider the consumption of data as well as the capture, as data use is often aligned with the working practices of other groups. This fact alone is responsible for limiting the impact of new systems that only provide an ELN to a single domain and focus on replicating a paper system.

Careful review is required in assessing what data is to be promoted to management oversight level and what remains a departmental view, with consideration also given to differences in terminology3. The resulting executive-level view improves research oversight and portfolio management by giving an integrated up-to-the-minute view of all research activities.

Supporting bioprocess
While the specific techniques used in the development of small and large molecule drugs are very different, the need to capture data and the associated analysis, interpretation and reports is the same. The increasing use of laboratory automation around purification and analytical techniques is resulting in a data explosion – similar to that experienced in the small molecule world over the last few decades. Labs are generating much more data than previously, and with the advent of new techniques such as next generation sequencing and protein mass spectrometry, the rate of growth is also unprecedented. This evolution and ‘data deluge’ only heightens the need for data management solutions to help scientists protect and leverage corporate knowledge.

Fundamentally, bioprocess organisations develop and implement cell, production and purification platforms. The data and knowledge cascade required to develop these platforms is dependent on many different domains being linked together and each domain having a specific view of that data: precisely the same high level requirement as small molecule development. The precise techniques, data types and science in biologics drug development such as molecular biology, cell line manipulation, fermentation, purification and formulations are all quite different to the small molecule domains. However, the collaboration, development and validation practices, regulatory oversight and the existence of data providers and consumers are identical. This means therefore that the foundation to support these practices can be the same – scientific information being preserved in an Enterprise ELN as part of the company’s knowledge base4.

There is a further implication for pharmaceutical companies developing both small and large molecules. For an Enterprise ELN to be capable of supporting both macro domains, consistent data management, analysis, reporting and validation approaches are required, which can simplify and streamline interactions with regulatory bodies, potentially reducing the total cost of ownership and maintenance of the system.

A foundation for quality by design
The growth in interest and application of Quality by Design (QbD) in the pharmaceutical industry requires a more systematic approach to achieving quality, and characterising acceptable variations in manufacturing processes5. This brings a different approach to regulatory submissions and manufacturing flexibility, but fundamentally necessitates a good data management approach to all data from the earliest stages of development.

Much of the preclinical and early development environment uses paper and electronic reporting, with tabular data embedded in flat documents. This makes accessing relevant data from the typical document store very difficult, and limits reusability because it cannot be searched and extracted in a structured manner. Relying on these flat documents, which lack any data management foundations, as systems to support QbD is therefore challenging. However, the use of Enterprise ELNs in preclinical development enables each study/project, be it in stability, analytical or bioanalysis, to be easily retrieved at any time for:

l Assessment of current project design

l Use in regulatory submissions

l Analysis of post market issues

This enables Enterprise ELNs to form a foundation to support QbD, by providing the data and knowledge management backbone across preclinical development, and bring significant benefits by:

l Improving data quality – by reducing and potentially removing the risk of transcription errors from paper binders, instruments and Excel® spreadsheets, the quality of data is improved and ensures there is a time-stamped, electronically signed record of all data and IP generated.

l Supporting knowledge reuse – scientists can reduce the time spent repeating studies by having a single point of access to a knowledge base that is searchable and retrievable.

l Supporting knowledge mining – allowing past data to be mined effectively enables predictive models to be built and multivariant analyses to be run, to characterise the variables with significant impact and requiring more analysis.

Key capabilities in enterprise ELNs
Enterprise ELNs are differentiated from earlier generations of ELNs by their data management orientated approach. Having a properly architected ELN, based on good data management practices, brings many benefits in terms of searchability and scalability, and moves the ELN from a simple patent IP capture tool to the portal environment in which scientists spend most of their time working. Some of the key capabilities for Enterprise ELNs are described in the following section and Table 1.

ELNs and regulatory compliance – review by exception
When ELNs are used in a regulatory environment, the capabilities required and implications of deployment change significantly. Enterprise ELNs must support the deployment requirements of good laboratory and manufacturing practice (GLP and GMP), with system audit trails, version control, process enforcement, real time data validation and e-signatures. When working in a GLP/GMP environment, an important function is to enforce and monitor compliance to Standard Operating Procedures (SOPs). The ability of an ELN to provide easily configurable, structured forms/templates that ensure user data capture conforms to specific business rules is fundamental. However, even with a GxP environment, there are exceptions to the rules, and when they occur it is important that the ELN not be prescriptive, but rather allow a user to complete a task with comments, and track such deviation for future reference and reporting. This approach improves data quality but remains flexible, by ensuring data was captured correctly at its source, and that any deviations are tracked in a searchable manner.

At any point, a Quality Assurance (QA) reviewer or auditor should be able to generate a report on all process deviations or exceptions with a given scope of work – a notion called ‘Review by Exception’. Such reports document all process exceptions and user comments, and provide links or references to the relevant experiment or record. Thus QA personnel no longer have to review all notebook content and associated paper binders for compliance, and instead only need to focus on documented exceptions. This significantly expedites the QA process and reduces the cycle time for study completion and issuing of reviewed reports.

Reuse of corporate data
The complexity of the preclinical and early development formulations process, for small and large molecules, means there is considerable difficulty in finding information on historical data. This means that the sample questions below are often difficult and time-consuming to answer:

l I need particle size and density data for each formulation containing API (Compound 1234, batches 01, 02, 03).

l I want to compare content uniformity and impurity data for all formulations using excipient X, Y and Z.

l Get me all instances of protein degradation and their associated excipient selection, responses and analytical method/instrument selection.

To answer these types of question, the ELN must have a combination of good data management and powerful search capabilities providing enhanced visibility of all past data around formulations, their analysis, excipients, PK and stability. By capturing and storing all data, corporate knowledge is fully searchable, promoting reuse of high value knowledge and enabling better risk-based decisions.

Report generation and regulatory validation
Once the cycle of formulations, PK and analytical testing is complete, study reports and eventually regulatory submissions need to be prepared. It is common for the data to be distributed across multiple data sources, storing structured results and scientific interpretations. A challenge for all groups is bringing this information together and then validating that the content of a report is accurate and no transcription errors have occurred. Report content can take weeks to prepare and many more weeks to validate because everything needs to be checked across different departments against multiple electronic and paper-based systems. Researchers want, and generally need, to write reports in Microsoft® Word, so the Enterprise ELN must allow researchers to develop report templates that can automatically populate documents used throughout development, from study reports to regulatory submissions.

These validation and study reports need to be generated as far as possible automatically, using digitally signed experimental data with the reports having active hyperlinks back to the source data captured in the ELN to streamline report validation and QA. Consolidation of data to a single authoritative source allows simpler real time checking and validation of data, resulting in fully prepared, QA and validated reports in days or weeks instead of months – a significant breakthrough in efficiency for preclinical scientific and quality groups.

Many companies are now recognising that using ELNs as purely paper-based replacements means they are missing out on strategic long term benefits, and are now integrating LIMS and legacy ELNs with those that provide data- and context-centric architectures. The growth in use of ELNs in preclinical environments has resulted in major advances in functionality that are allowing them to become the de facto standard for data management and corporate knowledge capture. By employing these capabilities organisations can streamline the preclinical development, provide a foundation for QbD and reduce the burden of regulatory compliance.

Thanks to Jeff LoCascio, Chris Molloy and others at IDBS for sharing their thoughts on this area.

Simon Beaulah is Head of Product Marketing at IDBS. He has been working in life science and healthcare informatics for more than 20 years, initially in research and over the past 12 years for a number of informatics vendors. Mr Beaulah has degrees from Aston University and Cranfield Institute of Technology.

Dr Paul Denny-Gouldson has almost 15 years’ experience in pharmacology, pharmaceutical drug development and ELN software development. He joined IDBS in 2005 when the company he founded, Deffinity Solutions, was purchased by IDBS. Prior to this, he was Senior Scientist at Sanofi-Synthelabo (now Sanofi-Aventis). Dr Denny-Gouldson obtained his PhD in Computational Biology from Essex University in 1996 and has authored more than 20 scientific papers and book chapters.

Dr Scott Weiss joined IDBS in 2004, bringing more than 20 years’ experience in pharmacology, neuroscience and pharmaceutical drug development. Formerly Director of In Vivo Pharmacology at Vernalis Research, he managed a multidisciplinary department of scientists whose remit spanned from target validation, secondary and tertiary efficacy testing, PKPD and cardiovascular safety. Dr Weiss obtained his PhD in Experimental Psychology from Leeds University in 1995, and has authored more than 40 scientific papers and three patent applications.