People using assistive technology may not be able to fully access information in these files. For additional assistance, please contact us.
Structured Abstract
Background
Progression-free survival (PFS), defined as the time from random assignment in a clinical trial to disease progression or death from any cause, has recently become an endpoint of considerable interest in the study of new oncology drugs. In comparison to overall survival (OS), the gold standard for cancer drug evaluation, PFS can be evaluated using shorter, smaller and less costly studies. Its use as a primary endpoint, however, can be challenging, as it is subject to a wide range of potential biases, and its use as a surrogate for OS has been demonstrated only for certain disease and treatment scenarios. The objective of this methods project is to address whether PFS is an outcome related to psychological well-being or quality of life (QOL).
Methods
Two Key Questions (KQ) were posed: (1) when PFS is used as a primary clinical endpoint in treating patients with advanced cancer, is there direct evidence that knowing PFS impacts patient anxiety, depression, or psychological well-being, and (2) for agents where PFS is the primary outcome measure being used to establish the performance (efficacy and safety) of a new drug, what evidence exists on the association of PFS with QOL and related outcomes, such as disease symptoms? A Technical Expert Panel (TEP) was convened to refine the KQs, comment on the methodological approach, and identify publications. The literature search for KQ1 sought to identify studies that showed a causal relationship between PFS and improvement on measures of psychological well-being. The search was not limited by tumor type or study design. KQ2, which addressed the association between PFS and QOL, was designed to indirectly answer KQ1, as a low yield of relevant articles was expected from the initial search. The literature search for KQ2 included terms for drugs approved by one or more regulatory agencies on the basis of PFS outcomes for treatment of solid tumor disease between 2005 and 2010, including Avastin®, Ixempra®, Tykerb®, Vectibix®, Doxil®, Gemzar®, Yondelis®, Nexavar®, Votrient®, Sutent®, Tarceva®, and Taxotere®. Both KQ searches were conducted in MEDLINE®, PubMed, Embase®, and the Cochrane Central Register of Controlled Trials (CENTRAL). From relevant articles, information was abstracted and summarized in data tables on study characteristics, treatment efficacy and safety, patient-reported outcome measure descriptions, and the PFS/QOL association. A quality assessment of the included individual studies was conducted to identify potential biases in the measurement of either PFS and/or QOL.
Results
No studies were identified that addressed KQ1. There was no direct evidence demonstrating that knowing PFS status impacts patient anxiety, depression, or psychological well-being. KQ2 sought to determine an association between PFS and QOL or related outcomes, such as disease symptoms, and four studies were identified that provided such evidence. The four studies demonstrated better QOL or disease symptoms among patients who remained progression-free compared with those who had disease progression. Study design limitations resulted in a poor quality rating for all four studies, and the strength of the evidence was insufficient. Common study limitations included significant data missing not at random, failure to evaluate patient-reported outcomes beyond the window of PFS, and lack of patients and investigator blinding to treatment.
Conclusions
The objective of this methods project was to determine whether PFS is an outcome related to psychological well-being or QOL. It focused on the relationship between PFS, an outcome, and other outcomes of importance to patients. In contrast to a traditional comparison of interventions, the variable of interest, PFS, is somewhat problematic, because the presence or absence of progression can only be observed, precluding the ability to design a prospective randomized clinical trial. Defining the relationship between PFS and QOL or other patient-reported outcomes (PRO) involves considering data as if obtained from observational study. A model is provided, including potentially important causal and confounding relationships, for considering further study of the relation between PFS and QOL-related outcomes.
There is a need for prospective research to evaluate: (1) any causal relationship between patient knowledge of PFS status with QOL or related PRO; (2) patient impressions of the meaningfulness of PFS as an outcome in the absence of association with OS; and (3) the extent to which improvement in QOL measures associated with progression-free status is related to a common underlying mechanism.
There is insufficient evidence to make any conclusion about the association between PFS and QOL. In cases when measurement of OS is unfeasible, the direct measurement of both PFS and QOL may be a practical and informative alternative strategy.