Clinical trials of adjuvant cancer therapy should do a better job of showing how treatment affects individual patients’ quality of life—particularly when the survival benefit is modest—according to recommendations from Common-Sense Oncology and the European Organisation for Research and Treatment of Cancer. Ian F. Tannock, MD, and colleagues published the recommendations in The Lancet Oncology.
The authors argued that conventional toxicity grading and average health-related quality-of-life (HRQOL) scores can miss clinically meaningful problems experienced by a subset of patients. Instead, trials should report the proportion of patients whose HRQOL worsens by a prespecified amount, how long that deterioration lasts, and whether patients recover after treatment ends.
Their recommendations focus on adjuvant treatment, where patients may already be free of detectable disease and are weighing treatment-related toxicity against a sometimes relatively small reduction in the risk of cancer recurrence.
Why Average Scores May Miss the Point
The authors noted that clinician-reported toxicity scales may underestimate symptoms and functional effects that matter to patients. Mean HRQOL scores can have a similar problem: if most patients report little change, substantial deterioration in a smaller group can disappear in the average.
That matters because the trade-off in the adjuvant setting can be very different from treatment for symptomatic advanced cancer. Some patients will be cured by local therapy alone and therefore receive no benefit from added systemic treatment, whereas others may avoid recurrence because of it. Patients who relapse despite adjuvant treatment can experience both treatment-related toxicity and the effects of recurrent disease.
The authors framed the decision around two questions a patient might reasonably ask: “How much higher will the chances of curing my disease be from the proposed therapy?” and “What are the chances of (and likely duration of) treatment-related toxicity and its effect on my quality of life?”
Proposed Approach
The group recommended that HRQOL be included at least as a key secondary endpoint in adjuvant trials and assessed with validated patient-reported measures covering symptoms, functioning, and overall quality of life. Patients should also be involved in deciding which outcomes are most important to measure.
Rather than relying only on mean changes, trials should report the proportion of patients in each treatment group who experience a clinically meaningful deterioration, they said.
When an established threshold is unavailable, the authors suggested that a decline of 15% on a normalized 0-to-100 HRQOL scale can serve as a pragmatic definition of meaningful worsening. They cautioned, however, that the same threshold will not necessarily be appropriate for every measure and recommended sensitivity analyses using different thresholds.
Timing also matters. The recommendations call for HRQOL questionnaires at baseline, every 3 to 4 weeks during the first 3 months of treatment, at least every 3 months during later treatment, and again at least 6 and 12 months after adjuvant therapy ends. Continued assessment after treatment can show whether quality-of-life deficits resolve or persist.
The authors also emphasized the importance of continuing assessments in patients who stop treatment because of toxicity or other reasons. Missing data from these patients are not necessarily random and can make a treatment appear better tolerated than it actually was.
QOL Data Should Not Be an Afterthought
The review pointed to previous adjuvant trials in which quality-of-life analyses may have obscured meaningful toxicity. In a meta-analysis of 10 trials that led to U.S. Food and Drug Administration approvals between 2018 and 2022, none of the individual studies reported a significant difference in HRQOL with the added treatment, although the pooled analysis showed a trend toward deterioration. The authors noted substantial potential for bias in all but 1 of the 10 reports, particularly because of patient attrition and missing data.
They recommended that the primary HRQOL findings appear in the main publication of a clinical trial rather than only in a later secondary report. More detailed analyses can follow in a separate paper or supplement, they suggested.
“The decision to offer such treatments to individual patients should not be a reflex prescription; an informed conversation is needed,” the authors wrote. They added that patients should be given realistic estimates of both the absolute benefit of treatment and the likelihood and duration of toxicity and quality-of-life impairment.
Ian F. Tannock, MD, of Princess Margaret Cancer Centre and the University of Toronto, is the corresponding author of The Lancet Oncology publication.
DISCLOSURE: The authors reported a range of research, consulting, advisory, and other relationships with industry; several authors reported no competing interests. For full disclosures of the study authors, visit thelancet.com.
