In an analysis that evaluated the outcomes of anticancer drug studies in advanced disease setting with regards to patients’ quality of life (QoL), improved QoL outcomes were associated with improved overall survival (OS), but not with improved progression-free survival (PFS). Almost half of the anticancer drug studies that demonstrated improved PFS showed no improvements in OS or QoL. Some publications summarised QoL findings by statements that were not directly supported by the study data, particularly for inferior or non–statistically significant QoL outcomes. Furthermore, contrary to common perception, inferior QoL outcomes were more common with targeted drugs than with cytotoxic drugs. Findings from a retrospective cohort study of 45 randomised clinical studies are published by Dr. Bishal Gyawali of the Division of Cancer Care and Epidemiology, Cancer Research Institute, School of Medicine, Faculty of Health Sciences, Queen’s University in Kingston, ON, Canada and colleagues in June 2022 issue of the JAMA Oncology.
The authors explained in the background that OS and QoL are the most important markers of therapeutic benefit of anticancer drugs in clinical studies. However, several studies have shown that many anticancer drugs receive regulatory approval without any evidence of improvement in either of these endpoints, based only on improvement of putative surrogates of efficacy. In particular, QoL is less frequently evaluated in randomised clinical studies of anticancer drugs and, even when tested, is underreported. Studies have shown that PFS, a commonly used intermediate primary endpoint, is poorly correlated with QoL and OS. Some physicians and patients also believe that targeted drugs are associated with better QoL than cytotoxic chemotherapy, thus promoting chemotherapy-free regimens.
Furthermore, the authors wrote that although QoL worsening is clearly an adverse outcome, failure to improve QoL demands special consideration in making risk-benefit decisions. However, study publications often report an observed lack of improvement in QoL in a positive way, such as QoL being maintained, rather than not improved, which may be misinterpreted to mean beneficial effects on QoL. To best inform discussions with patients about proposed treatments, the authors underlined it is important to clearly understand and communicate the effects of anticancer drugs on QoL alongside the effects on OS and intermediate endpoints such as PFS.
This retrospective cohort study included all patients with advanced cancer who were enroled into phase III randomised clinical studies of anticancer drugs reporting QoL data and published in English language in a PubMed-indexed journals in 2019. The systematic search of PubMed was conducted in July 2020. Main outcomes and measures were association of QoL outcomes with OS and PFS, framing of unchanged QoL outcomes in study publications, and the association of favourable framing with study funding by industry.
A total of 45 phase III randomised clinical studies enrolling 24806 patients, of whom 13368 in the experimental arm and 11 438 patients in the control arm, met the inclusion criteria. Improvement in global QoL with the experimental drug was reported in 11 randomised clinical studies (24%). Randomised clinical studies with improved QoL were more likely to show improved OS compared to studies with unimproved QoL (7 of 11 studies [64%] versus 10 of 34 studies [29%]; p = 0.04).
There was no such association observed for PFS (6 of 11 studies [55%] versus 17 of 34 studies [50%]; p = 0.87). Furthermore, 6 studies reported QoL worsening, of which 3 studies (50%) were with targeted drugs, and 11 studies reported improvement in QoL, of which 6 (55%) were studies of immunotherapy drugs.
Of the 34 studies in which QoL was not improved compared with controls, 16 studies (47%) reported these results in a positive frame, an observation statistically significantly associated with study funding by industry (p = 0.01).
The authors commented that several studies with negative QoL results are not published at all or are published after a considerable delay, so the present observations may understate the issues they have raised. However, although they focused on global QoL, some patients may find specific subdomains of QoL meaningful in some clinical settings. But focusing on non-curative settings only is a strength rather than a limitation of the analysis because the risk-benefit thresholds are different in curative settings where patients may be willing to accept QoL worsening to a certain degree in exchange for improved odds of cure.
Findings from this analysis have important implications as physicians, patients, regulators, and payers, must be aware of these issues to interpret QoL data from anticancer drug studies accurately and meaningfully.
Reference
Samuel JN, Booth CM, Eisenhauer E, et al. Association of Quality-of-Life Outcomes in Cancer Drug Trials With Survival Outcomes and Drug Class. JAMA Oncol 2022;8(6):879-886.