All authors read and approved the final manuscript. Funding: This research was supported by a grant from CDRPG8C0031. Competing interests: None. Ethics approval: The Chang Gung Medical Foundation Institutional Review Danoprevir 850876-88-9 Board (approval number 103-0418B). Provenance and peer review: Not commissioned; externally peer reviewed. Data sharing statement: No additional data are available.
Use of progression-free survival (PFS) as a primary end point in oncology has increased

recently, as has its use as a secondary end point.1 Using PFS, as opposed to overall survival (OS), has several advantages for clinical trial conduct; trials that use PFS as a primary end point can be conducted more quickly and with fewer patients than trials using OS.1 This also benefits patients as it allows earlier access to new treatments as trial results are available sooner when PFS is used as an end point. PFS also directly measures the effect of the investigational treatment and, unlike OS, is insensitive to bias from subsequent treatment(s) (ie, treatment received after disease progression has been determined).2 This issue of bias in interpretation of OS data is also compounded

by the fact that use of subsequent therapies generally differs between treatment arms.2 Despite the advantages of PFS, there are several limitations to consider. There are no standard regulatory criteria for defining progression in clinical trials2 and

progression can be difficult to assess and subject to measurement error and bias, especially if assessors are not blinded to treatment.2 PFS is also influenced by frequency of assessment, unlike OS.3 Even though an improvement in PFS is considered an indication of disease control and stabilisation,4 there is still debate as to whether an improvement in PFS is beneficial for patients.5 As such, it is important that PFS benefits seen in clinical trials are accompanied by better Carfilzomib symptom control, fewer treatment-related adverse events and better health-related quality of life (HRQoL).1 4 While randomised controlled trials (RCTs) frequently assess HRQoL as well as PFS, the design of such trials only allows indirect inferences regarding a relationship between PFS and HRQoL in situations wherein both are influenced by treatment. For this reason, some health technology assessment agencies6 do not consider PFS a patient-relevant outcome measurement and usually discard the information on this end point in their evaluations, particularly in indications and for investigational compounds where PFS may not be a well-established surrogate for OS. Thus, there is a need to establish the relationship between changes in PFS and HRQoL.