Three-dimensional illustration of pink and blue virus particles

FDA’s Overall Survival Guidance: What Sponsors Need to Know

4 minute read

Published: November 17th, 2026

In August 2025, the FDA released recommended guidance on ‘Approaches to Assessment of Overall Survival in Oncology Clinical Trials’, their first guidance document dedicated specifically to Overall Survival (OS), which remains a draft guidance as of November 2026. This blog summarizes the FDA’s latest guidance on assessing overall survival (OS) in oncology clinical trials and highlights the key considerations for sponsors when designing, analyzing and interpreting OS endpoints in modern cancer studies. 

Building upon previous guidance 

This expands on the existing OS guidance included in ‘Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics.’ This previous guidance documented a formal definition of OS, and gave the many advantages of using OS as a primary endpoint: 

  • It is precise and easy to measure 
  • Does not rely on investigator interpretation 
  • Fully reflects clinical benefit 

The previous guidance also explains how the long follow-up periods in large trials can provide difficulties, and how subsequent cancer therapy can potentially confound analysis. Additionally, it documents the importance of OS as an endpoint in non-inferiority trials. 

This new guidance document, however, is the first FDA guidance document to focus solely on OS. The document reinforces that OS remains the most reliable and clinically meaningful endpoint, but also expands upon how it should be designed, analysed and interpreted in modern trials.  

Key takeaways 

1.OS remains the gold standard endpoint 

    Despite patients living increasingly longer, and therefore OS taking longer to measure as treatments progress, the FDA reiterate that OS is the most objective and clinically meaningful endpoint, and that OS should be prioritised as the primary endpoint when feasible. In trials where OS is not considered a primary endpoint, the FDA recommends that OS is still included for submission, where it can be used to assess efficacy or safety. 

    2.Timing of analyses 

      Interim analyses should be timed such that a sufficient number of death events have been observed. This must however be balanced with participants’ exposure to potentially harmful treatment. The FDA expects timing of interim analyses to be clearly specified, based on number of events (as opposed to time durations), and a clear justification for the proposed timings.  

      3.Defining a clear estimand 

        Sponsors are advised to carefully consider the Estimand framework during the planning of analyses, including intercurrent events, such as subsequent therapy, and their corresponding handling strategies. Once intercurrent events have been identified, the FDA advises assessing the pattern and reason for intercurrent events, as well as any bias caused by baseline characteristics. A treatment policy strategy is recommended for OS in most scenarios, alternative handing strategies may also be used with clear documented justification. 

        4.Crossover must be carefully justified and handled appropriately 

          Crossover studies are sometimes favoured due to improvements in enrolment and retention. However, crossover can bias OS results towards no difference, therefore the FDA recommends crossover is avoided wherever possible when OS is the primary endpoint.  

          Keeping pace with the latest regulatory developments is a key focus at Phastar. We actively review emerging guidance and, where relevant, provide comments on documents open for consultation, as we did for this draft guidance. 

          The FDA’s guidance reinforces the importance of overall survival as the most objective and clinically meaningful endpoint in oncology clinical trials. While evolving treatment landscapes, longer patient survival and increasingly complex treatment pathways can make OS more challenging to evaluate, the guidance provides greater clarity on analysis planning, estimands and the management of crossover. For sponsors, the message is clear: robust planning and careful consideration of factors that may influence survival outcomes are critical to generating reliable evidence of clinical benefit. 

          Need support navigating oncology endpoint strategy and regulatory expectations? Phastar’s oncology and regulatory statistics experts work with sponsors to develop robust trial designs, define appropriate estimands, evaluate overall survival endpoints and prepare for regulatory submissions.  

          Related articles

          Women in Pharma: Transforming Clinical Trials with Advanced Statistics

          Women in Pharma: Transforming Clinical Trials with Advanced Statistics

          November 12th, 2026 2 minute read

          In a new article for ICT, Vicky Marriott, Helen Marshall, Hebing Wang and Samantha Hinsley from Phastar explore how a...

          Building a Successful Clinical Development Strategy with the Right Biometrics CRO 

          Building a Successful Clinical Development Strategy with the Right Biometrics CRO 

          November 9th, 2026 7 minute read

          Clinical development teams are often familiar with outsourcing activities such as monitoring, project management, and...

          Is the FSP Model Right for Your Clinical Program? 

          Is the FSP Model Right for Your Clinical Program? 

          November 6th, 2026 8 minute read

          Clinical development teams are under constant pressure to deliver more with fixed budgets, tight timelines, and limit...