This comprehensive review article, titled “Frequently asked questions on surrogate endpoints in oncology—opportunities, pitfalls, and the way forward,” published in eClinicalMedicine in 2024, is authored by Abhenil Mittal, Myung Sun Kim, Shenna Dunn, Kristin Wright, and Bishal Gyawali. The paper addresses a critical and evolving area within cancer treatment and drug development: the increasing reliance on surrogate endpoints for the approval of new cancer drugs, often in lieu of direct evidence of improved overall survival (OS) or quality of life (QOL).
The authors highlight a significant tension in the current landscape of oncology: the desire for early access to potentially life-saving cancer drugs through the use of surrogate endpoints versus the fundamental need for confirmation that these treatments genuinely improve outcomes that matter most to patients. While patients with cancer inherently expect prolonged life (OS) or a better quality of life (QOL) from their treatments, the majority of recent drug approvals are based on measures like tumor shrinkage or delayed tumor progression, which are considered surrogate endpoints. The review critically examines the validity of these surrogates as proxies for overall survival, noting that their correlation with patient-centric outcomes is generally weak and varies significantly based on disease settings and lines of treatment.
Structured in a user-friendly Q&A format, the article delves into various facets of surrogate endpoints in oncology trials, covering both curative (neoadjuvant and adjuvant) and advanced disease settings. Key questions addressed include the precise definition of various surrogate endpoints, their methodologies for validation, and the existing evidence regarding their surrogacy. The authors distinguish between patient-level and trial-level surrogacy, explaining that while some markers like pathological complete response (pCR) can be prognostic at an individual patient level (predicting better outcomes irrespective of treatment), they often fail as trial-level surrogates because treatments improving pCR rates do not reliably predict improvements in event-free survival (EFS) or OS. Similarly, overall response rates (ORR) are discussed as measures of biological activity but are often poor trial-level surrogates for OS.
The review meticulously outlines the purported benefits of using surrogate endpoints, such as the ability to measure treatment efficacy earlier, leading to reduced trial completion times (averaging 11–19 months reduction compared to OS endpoints), potentially smaller sample sizes, and lower trial costs, which can expedite drug development and regulatory processes. However, it also thoroughly exposes the numerous pitfalls and challenges. These include the weak correlation of many surrogate measures like progression-free survival (PFS) with OS and QOL, the overuse and over-reliance on surrogates for regulatory approvals (with a significant increase observed between 2011 and 2019), and the concerning trend of granting “premature regular approvals” based on surrogates without the safety net of mandatory confirmatory trials. The paper also explains why PFS often fails to correlate with OS, citing reasons such as arbitrary RECIST criteria, dilution by post-progression treatments, crossover effects, and informative/unequal censoring. Patient misunderstanding of PFS, often equating it with improved survival, is another significant concern highlighted, leading to the proposal of renaming PFS to “progression-free interval” (PFI) to mitigate confusion.
Furthermore, the article critically examines the strength of surrogacy required for accelerated versus regular drug approval, noting that the distinction has blurred, with regular approvals increasingly granted based on surrogate endpoints, sometimes even from single-arm trials. The heterogeneity of surrogate endpoint validity is emphasized, meaning that surrogacy cannot be easily extrapolated across different tumor types, subtypes, lines of therapy, or treatment modalities.
In conclusion, the authors propose a “way forward” to optimize trial design and ensure timely access to effective cancer medicines. This includes mandating appropriate validation of surrogate endpoints, ensuring transparency about surrogacy strength for different approval pathways, and not mistaking prognostic markers for trial-level surrogates. They suggest that a more efficient approach involves conducting randomized controlled trials (RCTs) powered for OS from the outset, potentially using a “one-trial approach” as outlined in new FDA guidance, which allows for early accelerated approval based on surrogates but requires mature regular approval based on OS. The review strongly advocates for direct measurement of QOL outcomes rather than inferring them from surrogates and recommends that OS should be the default endpoint, especially in later-line settings or where median OS is short.
APA Reference: Mittal, A., Kim, M. S., Dunn, S., Wright, K., & Gyawali, B. (2024). Frequently asked questions on surrogate endpoints in oncology—opportunities, pitfalls, and the way forward. eClinicalMedicine, 76, 102824. https://doi.org/10.1016/j.eclinm.2024.102824