Title: Progression-Free Survival 2 in Hematological Cancer Treatment Choices: Challenges to Routine Use

Background: In 2012, the European Medicines Agency (EMA) recommended the use of progression-free survival (PFS) 2 as a secondary endpoint in trials evaluating maintenance treatment in hematology oncology trials. PFS2 is defined by the EMA as time from randomization to objective tumor progression on next-line treatment or death from any cause. Potential benefits of PFS2 may be found in cases where experimental treatments appear to improve PFS but not overall survival (OS), particularly in situations with improved survival but where measurement of OS is confounded, for example when drugs are trialed in early lines of therapy and subsequently coupled with improved salvage treatments. However, in practice the utility of this endpoint may seem unclear in cases where tumor cells undergo molecular or microenvironment changes without obvious clinical progression. Currently, the US Food and Drug Administration (FDA) does not recognize this end point and in our experience, healthcare providers are unsure of its definition or value.

Objective: To understand the current definitions and limitations of PFS2 from a review of the literature and registered clinical trials in the context of PFS.

Methods: We undertook a pragmatic literature review relating to PFS and PFS2. Sources investigated included peer-reviewed literature, clinicaltrials.gov, horizon scanning and health technology appraisal (HTA) activities, and 'grey literature' sources. After review of the search hits, information was grouped according to the therapy area and type of information and the main themes and questions identified.

Results: We identified a range of specific challenges and issues regarding each endpoint. For PFS, we noted ongoing debates across disease areas. These principally related to the interpretation of clinical trial evidence: expert clinical opinion routinely highlighted concerns about statistical issues and bias arising from censoring, missing data, subjectivity and assessment time. Two other common issues were the need for adequate sample size to detect differences in OS and a lack of evidence relating prolonged PFS to better health-related quality of life (HRQoL).

For PFS2, more basic challenges and needs were identified than for PFS. Only the EMA provides a clear definition; the US Food and Drug Administration (FDA) and other regulatory agencies do not appear to have a definition or policy regarding PFS2 at this time. On the other hand, PFS2 is increasingly incorporated into oncology trials as a secondary endpoint in both hematologic and solid tumors. As of 1 August 2016, there were 41 oncology trials registered on clinicaltrials.gov where PFS2 is included. Of these, 20 trials have been registered or updated since January 2015. The implications of routinely generating and using PFS2 data for regulatory and market access purposes therefore are currently unclear. Most of the current information is related to PFS2 data in trials of lenalidomide in multiple myeloma.

Based on commentaries and discussions in the literature, there is an ongoing need to understand the clinical and economic rationales for using PFS2 data to support benefit-risk assessments (e.g. relating effect sizes to clinical outcomes) and to determine how best to communicate these potential benefits meaningfully. There is also a need to understand how PFS2 may relate to patient-centered benefits, such as HRQoL, which may help to identify patient preferences in the context of treatment choice.

This information has been used to develop a physician questionnaire that would help identify insights and opinions about PFS2 in the following areas: (1) use of surrogate endpoints, (2) treatment choice in relapsing cancers, (3) implications of PFS2 for clinical practice and (4) future needs for PFS2. Results from this primary research will be presented subsequently.

Conclusions: PFS2 is increasingly included as a secondary endpoint in oncology trials assessing the benefits of maintenance or sequential treatments. However, there is limited guidance from regulatory bodies on using and interpreting PFS2 data even though there are post-approval implications regarding potential positioning of drugs within treatment pathways. To date, the EMA is the only regulatory agency that has considered the relevance of PFS2 as a clinical endpoint. The clinical importance of PFS2 needs greater understanding to help physicians apply the data in patient management.