A Dolon Institute Report
Over the past two decades, advances in oncology have transformed cancer from an increasingly terminal disease into a condition where many patients can benefit from multiple lines of effective treatment. However, this progress has created a new challenge for health technology assessment (HTA): it is becoming increasingly difficult to generate mature comparative overall survival (OS) data within the timeframe required for regulatory and reimbursement decisions.
Absence of comparative survival data in health technology assessment of oncology therapies examines this challenge from a multi-stakeholder perspective, bringing together clinicians, patient representatives, regulatory and HTA experts from across Europe. The paper explores why mature OS data may be unavailable, the implications for oncology assessment and patient access, and potential approaches for managing uncertainty while maintaining robust standards of evidence.
The challenge is driven by several factors. Improved survival means trials can take much longer to demonstrate an OS benefit, sometimes exceeding a reasonable development timeframe. In other settings, patients receive multiple subsequent treatments, making it difficult to isolate the contribution of an individual therapy to overall survival. Ethical considerations and practical challenges can also make randomised controlled trials difficult or inappropriate, particularly in rare cancers, biomarker-defined populations or situations where early evidence suggests an exceptional treatment benefit.
This creates a disconnect between regulatory approval and HTA decision-making. Regulators have increasingly accepted endpoints such as progression-free survival (PFS) and overall response rate (ORR) when mature OS data are unavailable. However, these intermediate outcomes are less consistently accepted by HTA agencies when assessing comparative value and reimbursement. The result can be lengthy negotiations, inconsistent decisions between countries and delays in patient access.
The paper proposes five recommendations spanning the product lifecycle rather than seeking a single solution. These include establishing clearer, context-specific expectations for acceptable evidence during clinical development; improving alignment around the validation and weighting of patient- and disease-relevant surrogate endpoints; considering adaptive HTA and risk-sharing approaches where uncertainty remains at launch; and strengthening long-term evidence generation after approval.
A particularly important recommendation is greater use of real-world evidence (RWE) and longer-term data collection. Where mature OS cannot reasonably be generated through an RCT, well-designed real-world data could help address remaining uncertainty as evidence accumulates. The paper highlights the need for agreed methodologies, consistent data requirements across countries and stronger infrastructure to capture and link treatment and outcome data.