Progression-Free Survival and Patient Quality of Life Are Not One and the Same
To make informed treatment decisions, the affect a drug has on a patient’s quality of life must be considered, investigators argued.
More evidence has been published to support the argument that a medication's ability to improve progression-free survival (PFS) says virtually nothing about its affect on a patient's quality of life (QoL) — in fact, the real question of importance may be whether QoL has improved in concert with any associated extensions to survival.
A new study suggests the link between PFS and QoL is flimsy (correlation coefficient [r] = 0.34; area under the curve [AUC]: 0.72) and that QoL isn't typically even a measure on the radar of clinical trial researchers. The authors added that only a little more than half (54%) of the 352 phase 3 cancer trials on advanced or metastatic solid tumors published between 2010 and 2015 across 7 prominent journals even included QoL as an end point. Furthermore, of the 190 studies that did include a QoL component, 23% of the papers did not report prespecified QoL outcomes, leaving only 147 studies that fit these criteria. Of the 147 studies, 25%, 19%, and 14% of trials were for lung cancer, gastrointestinal cancer, and breast cancer, respectively.
Although the researchers determined that reporting of QoL outcomes varied substantially by cancer site, primary trial outcome, and trial enrollment, they also found that trials that were positive for the primary end point (either progression-free survival, overall survival [OS], or primary efficacy) were more likely to report QoL outcomes compared with trials that were determined not to be negative (92% vs 64% respectively, P < .001). All of the 10 trials that were found to be linked to a negative QoL were also negative for the primary end point.
And, in reality, the estimates for studies reporting a positive QoL may even be conservative. Study coauthor Bishal Gyawali, MD, PhD, Brigham and Women's Hospital, Boston, Massachusetts, wrote in an email to Cancer Therapy Advisor that “given positive publication bias, our findings might overestimate the percentage of trials that actually improve QoL in practice.”
The authors suggested that perhaps medical journals should require the reporting of QoL data for an article to be considered for publication — as a measure to show objectively that QoL has actually improved following administration of a drug. They said that this kind of approach worked for clinical trial registration, which improved appreciably after journals demanded that trials be registered.
“This change can be easily brought on by 2 key players: the regulatory agencies and the journals,” Dr Gyawali said. “If the regulatory agencies demand that the drug that fail OS must show improvement in QoL for approval, this change can happen immediately,” he added.
In summary, the authors said their “findings suggest that patients' quality of life must be directly measured and cannot be assumed from PFS or surrogate end points alone […] Understanding the affect on quality of life is also paramount for treatments whose effect on OS may not be considered clinically meaningful or in cancers for which PFS is a poor surrogate of OS.”
In a separate but related article that was recently published, Dr Gyawali argued that a general lack of harms reporting in studies — as well as the use of vague and subjective terms to describe these harms — could be better addressed if QoL data were also collected for every study.2 He and colleagues reasoned that this type of QoL reporting could be a way to circumvent the spin that sometimes accompanies the reporting of negative outcomes and events in medical journals.
- Hwang TJ, Gyawali B. Association between progression-free survival and patients' quality of life in cancer clinical trials [published online October 29, 2018]. Int J Cancer. doi: 10.1002/ijc.31957
- Gyawali B, Shimokata T, Honda K, Ando Y. Reporting harms more transparently in trials of cancer drugs [published online November 1, 2018]. BMJ. 2018;363:k4383. doi: 10.1136/bmj.k4383