Takeaways from SCOPE Summit Europe
by Melinda Pautsch, Chief Commercial Officer, QuartzBio
I returned from SCOPE Summit Europe 2026 with a clear message: the quality of clinical trial outcomes depends on decisions made far earlier than many teams realize.
Across sessions on AI, protocol design, risk-based quality management, site experience, and vendor strategy, speakers repeatedly connected downstream performance to the strength of the underlying data and operating model. Biospecimens sit at the center of that discussion because they generate much of the data used to assess endpoints, understand biological response, and support clinical decisions.
Here are three themes that stood out, along with practical implications for biomarker operations, biospecimen operations, and clinical operations leaders.
1. AI readiness begins with operational and data readiness
AI came up in most sessions at SCOPE Summit Europe, and speakers consistently emphasized the importance of the data foundation beneath it.
In a panel discussion on operating models, representatives from Boehringer Ingelheim, Novartis, Johnson & Johnson, and Bayer discussed data readiness as the first step toward meaningful AI adoption. They also highlighted the opportunity to rebuild the way clinical development work is performed, from the foundation up, instead of layering new technology onto processes that already create friction.
The same principle applies to biospecimen data. Sample information is often distributed across sponsors, CROs, sites, central laboratories, specialty laboratories, and logistics providers. Variations in identifiers, metadata, terminology, and transfer formats make it difficult to assemble a reliable view of a sample and its associated data.
Speakers observed that trustworthy AI requires measurable quality controls. Johnson & Johnson described an AI query-drafting system that evaluates each query against 43 metrics before human review. The model was trained using controlled historical data. This type of governed approach offers a practical model for applying AI in GxP environments.
Operational takeaway: Establish the data, process, and governance foundations across the entire sample data value chain to enable AI-supported workflows to produce reliable results.
- Standardize sample identifiers and metadata across systems and partners.
- Define quality controls for AI-supported workflows.
- Use controlled data for model development and testing.
- Maintain human review for decisions that affect data quality, patient safety, or regulatory compliance.
- Select initial use cases with measurable inputs, outputs, and quality criteria.
Target outcomes: A connected and governed data foundation enables teams to evaluate AI outputs consistently, automate appropriate activities, and expand successful use cases with greater confidence.
2. Sample integrity monitoring must be designed into protocols and risk plans
Protocol complexity is creating practical challenges for sites and study teams. At SCOPE Summit Europe, Matthieu Ruffin, Head of Study Leadership at Novartis, reported that eligibility criteria and endpoints have doubled over 15 years, and as a result, the percentage of site satisfaction with sponsors has dropped from 85% to 69% over the last 6 years. “A perfect protocol isn’t perfect if it’s not deliverable,” concluded Daniel Bamford, Director of UK Clinical Trial Partnerships at Moderna.
Biospecimen planning is a critical part of that equation. During a Biospecimen Management Consortium panel featuring Bristol Myers Squibb and Roche, speakers shared that 45% of sponsors discuss their biospecimen plan internally without formally documenting it. Only 51% agree that the schedule of activities provides enough information to create an accurate laboratory manual.
Those gaps can create ambiguity around collection requirements, processing instructions, stability windows, storage conditions, shipping, and endpoint dependencies.
An incisive case study presented by Marcin Makowski, PhD, GSK’s Head of Centralized Monitoring and Data Analytics, illustrated the risk. A Phase 2 study proceeded according to plan, yet auditors ultimately concluded that its results were unreliable. The endpoint depended on a urine culture whose results could be affected by time and heat during transit. The risk-based quality management plan monitored sample collection, but it did not monitor sample stability.
This example shows why the complete sample journey must be treated as a critical-to-quality process.
The regulatory focus on sample integrity is also increasing. Cath Mela, PhD, GSK’s VP of Human Biospecimen Strategy and Delivery, described an MHRA inspection that devoted three full days to biospecimen chain of custody, while an FDA inspection dedicated 25% of its time to laboratory data. Conference speakers also noted that many organizations continue to manage this work in Excel, despite the general consensus at the US SCOPE Summit that workflows needed to move away from spreadsheets.
Operational takeaway: Bring biospecimen expertise into protocol design and create a dedicated risk plan for sample integrity.
- Identify which endpoints depend on specimens.
- Document collection, processing, stability, shipping, and storage requirements before finalizing the protocol.
- Confirm that the schedule of activities contains enough detail to support an accurate laboratory manual.
- Define critical-to-quality factors across the full sample journey.
- Monitor risks associated with time, temperature, handling, chain of custody, and data reconciliation.
Target outcomes: Upstream biospecimen planning reduces protocol ambiguity, protects endpoint reliability, strengthens inspection readiness, and gives sites clearer instructions for executing the study.
3. Fragmented systems and partner models create operational cost
Sites often carry the burden of connecting the systems introduced through sponsors’ digital transformation programs.
Nick Hodges, SVP and Head of Clinical Development Operations at Ipsen, shared that only 29% of sites believe the technology they use for studies delivers the promised efficiency. Every additional system, login, workflow, and point of contact adds work for site personnel. It can also influence which studies receive sites’ attention.
Fragmentation extends across the partner ecosystem. A Tufts study covering 133 trials found no single sourcing model that consistently improved performance. The study did find an association between a higher number of vendors, more challenging trials, and longer study timelines. Speakers emphasized that enterprise-level partnerships, rather than study-level or therapeutic area-specific engagements, create the greatest value for sponsors.
The issue is broader than vendor count alone. Performance depends on how effectively sponsors, CROs, laboratories, logistics providers, and sites share information and coordinate decisions.
When those partners work through disconnected systems, teams spend more time transferring data, resolving discrepancies, confirming status, and determining which source is current. Sample operations are particularly vulnerable because a single specimen can pass through several organizations before its analytical result becomes available.
Operational takeaway: Evaluate technology and sourcing decisions as part of one connected operating model.
- Map the systems and partner handoffs involved in the sample journey.
- Identify where sites are entering the same information more than once.
- Define ownership for sample status, issue management, and reconciliation.
- Standardize data exchanges and operational expectations across vendors.
- Assess prospective partners on their ability to participate in an integrated data ecosystem.
Target outcomes: A coordinated partner and technology model reduces administrative burden, improves site experience, increases visibility across the sample lifecycle, and helps study teams resolve risks before they affect timelines or data quality.
Sample Intelligence provides the foundation for scalable clinical operations
SCOPE Summit Europe reinforced the connection between unifying the data foundation for sample operations and the reliability of clinical trial results.
Mela from GSK noted that 65% to 70% of a clinical study report is based on specimen-derived data. At the same time, TransCelerate’s risk-based monitoring toolkit includes more than 140 predefined risk indicators with little to no coverage of biospecimen operations.
That gap deserves attention. Biospecimen data supports a substantial portion of the evidence generated by a clinical trial, yet the systems and risk frameworks governing that data often remain fragmented.
QuartzBio’s Sample Intelligence Solution gives sponsors a way to connect sample, operational, clinical, and biomarker data across the clinical trial ecosystem. With harmonized data, end-to-end traceability, and proactive risk oversight, teams can protect sample integrity, improve operational execution, and establish the trusted foundation required for responsible AI adoption.
The decisions that determine data quality begin upstream. Sponsors that involve biospecimen experts early, design controls around the complete sample journey, and connect their partners through a shared data foundation will be better positioned to scale innovation across clinical development.
How much risk is hiding in your sample operations?
If your organization is evaluating AI use cases, strengthening risk-based quality management, or simplifying complex sample workflows, QuartzBio can help you assess the existing risks in your operational and data foundations.
Start by taking our quick self-assessment to find out how vulnerable your sample operations are to delays, quality issues, and missed risk signals.
8 questions | 2 minutes | No signup
