How Sanofi scientists streamlined their in silico hazard assessment of potential genotoxic impurities by collaborating with Discngine
“After four months of using Discngine’s platform we achieved an overall project cycle time reduction of 25%”
Drugs are always accompanied by small amounts of impurities, whether they come from the synthesis or degradation processes. The mutagenicity and genotoxicity assessments of these impurities are critical to ensure the safety and quality of drug products. Pharma companies and regulators comply with the well-established ICH M7 guideline, to limit the potential carcinogenic risk due to drug impurities.
These hazard assessments rely on the combination of publicly available or proprietary experimental data and in silico predictions. Moreover, this is a prerequisite for every drug product, since regulatory approval requires an ICH M7 assessment, among other things, to run clinical trials or for a release of market batches, for example, when a new synthesis route was used.
Our client, Sanofi, assesses over 1000 potentially mutagenic or genotoxic impurities per year. Until recently, these assessments used to imply a lot of manual operations from its scientists. Hence, there was an urgent need to automate and digitalize the process to be more efficient.
We, Discngine, an R&D and lab informatics company that provides customer services for life science research, helped Sanofi to build a product that digitalizes and drastically streamlines the processes of its in silico toxicology team responsible for impurity mutagenicity and genotoxicity assessment.
We sat down with Sanofi experts, Dr. Alexander Amberg, Head of In Silico Toxicology and Alain Grelier, Digital product Owner, to talk more about the product and their experience with the successful Discngine-Sanofi collaboration.
In this blog, we are sharing their journey and interesting success metrics with you.
Dr. Alexander Amberg is a prominent expert of in silico toxicology with over 20 years of experience in the pharma industry. He holds the prestigious title “European Registered Toxicologist (ERT)” due to his extraordinary theoretical and practical competence in the field. With his broad expertise, Alexander is contributing to many worldwide initiatives, such as the IMI eTRANSAFE consortium, different GTI (genotoxic impurity)/ICH M7 initiatives and others. He is an author of more than 50 publications in scientific literature.
Alain Grelier is an engineer leader with over 30 years of experience on IS solutions delivery for most of the business domains of pharmaceutical products R&D lifecycle. His proficiency covers wide range of domains: from portfolio, project, and program management over developing life science IT solutions to software evaluation and selection. Alain is extremely proficient in project methodologies such as Agile/Scrum.
Could you share a few metrics regarding your team’s activity before working with Discngine?
At that time, the in silico toxicology team had to compensate for the loss of 2 people. We were supporting with our work on somewhat between 40-50 projects in a year. For this we usually:
prepare 30 reports used for approval dossiers, containing for each impurity available experimental results and in silico predictions followed by an expert hazard assessment used for our conclusion and classification
analyze around 1100 impurities a year
The cycle time to prepare each report upon request was around 8 weeks: 4 weeks for data gathering (experimental results and in silico predictions) and 4 weeks to prepare the report for regulatory dossiers.
What was the trigger to start searching for a solution to digitalize your activities?
The increase of the workload: The abovementioned situation created an increase in workload due to the decrease of manpower we faced at that time, resulting in an urgent need to automate our processes to deliver results timely. The goal was to decrease or at least keep the cycle time the same, even with fewer people.
The need for quality improvement: The idea was to have one central platform to store all data on our impurities – past, present, and future. We wanted to develop a searchable repository enabling the best use of our impurity data. This would allow to generate knowledge of all the past analyses to be used to increase the quality of future expert hazard assessments.
The lack of seamless data sharing between departments: The genotoxicity assessment of impurities is used by various departments (chemistry, analytics, dossier, worker safety etc.), so one central digital platform would enable us to share harmonized data between departments and to communicate faster.
How did Discngine help to solve the problem?
The product we developed with Discngine is called MILTAP - Mutagenic Impurity List Toxicity Assessment Platform and we are using it for a year now.
With MILTAP we managed to replace Excel lists and Sharepoint folders, which chemists and the in silico team had to update manually. We established a semi-automated workflow spanning from uploading structures by chemists up to preliminary report generation. Therefore, we streamlined chemistry, experimental and in silico data handling as well as the decision-making process. Moreover, having a common tool enabled us to digitalize our activities and save time.
This way we are greatly reducing the risk of manual errors and improving work quality and efficiency.
Some of the MILTAP features are:
-
Impurity upload
-
Automatic in-silico prediction triggers:
- Alerts
- Prediction probability statistical systems
- Alerting structure highlighting
-
Automatic experimental data retrieval
-
Preliminary classification by technician
-
Preliminary Reporting
MILTAP application interface: Impurities list -> Click to zoom in
MILTAP application interface: Impurity Assessment -> Click to Zoom in
Can you share a few metrics regarding the MILTAP platform usage and achievements?
We use MILTAP for every single impurity assessment we run now – which is a big success!
After 4 months of using MILTAP we did the assessment of the cycle time and resources and collected metrics. With MILTAP we succeeded to reduce the overall cycle time by 25%. In those 4 months, we assessed around 400 impurities.
Current statistics of MILTAP usage (one year after roll-out):
35 projects
920 assessed impurities
64 Registered users (Chemistry and Preclinical Safety Project Team Members, in silico toxicology team members)
Only a very few bugs have been reported since the roll-out of the platform, and we only experienced one service interruption in more than a year, which is low for a new platform built from scratch.
“Discngine has a very effective way to face all real-life problems that happen in any platform development project; what was planned was a MVP, but in the end fully functional”
You went for a custom product rather than buying an out-of-the-box solution. What was the rationale?
The exact product we needed did not exist on the market. While testing some commercial software solutions we realized that available software cover only part of the workflow, which wouldn’t fulfill all our needs. Also, none of them allowed us to combine predictions from various software vendors.
That’s why we decided to go for the co-development of a custom solution with a specialized partner.
On what ground did you choose Discngine? How would you describe the collaboration?
There were 3 main reasons for choosing Discngine:
Discngine is specialized in developing and maintaining solutions for pharma R&D
The company has broad experience and true success with previous/existing customers in creating customized products from scratch
Excellent scientific background – we were very happy with the cheminformatics knowledge demonstrated in Discngine
Regarding collaboration, it was very easy to work closely together from the start with them since Discngine is a mid-size company with very enthusiastic people.
Moreover, a very important aspect of our collaboration is Discngine’s expertise in the agile methodology which helped us to reach the right level of Agility in our own company – in terms of roles, practice, and behavior. Thanks to Discngine it was very easy for us to get used to the SCRUM process.
Last but not least, their enthusiasm and ideas for the project were very effective in overcoming all the real-time issues and ending up with a successful collaboration and much-needed solution.
What are the future directions for MILTAP? Where can it expand?
The idea is to broaden the scope of MILTAP application by implementing new important features that will further automate our workflows.
One of the ideas for the future, for example, is the automated search of MILTAP of external public data sources (e.g. ECHA database), in addition to our internal databases, to automatically search and retrieve experimental data. Another example is the integration of the Purge factor calculation to be used for follow-up steps and risk assessment. We are also planning to build an API for an automatic preliminary class suggestion. This would allow us to use this classification for optimization of the chemical synthesis schema toward the avoidance of genotoxic impurities.
We expect these enhancements would reduce overall cycle time by another 25% – from 25% to 50% in total.
Summary
From an urgent need to improve the productivity and efficiency of Sanofi’s assessment of potentially mutagenic and genotoxic impurities to working together on developing a customized solution, Discngine and Sanofi established a tight collaboration that resulted in a Digitalization Success Story. After only one and a half year from the beginning of the project, MILTAP platform became a common tool in the everyday work of Sanofi scientists dealing with genotoxic impurities.
Kudos to all experts from Sanofi and Discngine involved in the project!
If you want to learn more about Discngine services for complex life science R&D projects, download this brochure. Or reach out to us via email: contact@discngine.com