Research Signal – Source date 2022-10-04
Some quality-of-life measures favored drugs that release an immune-cell brake, while another showed no clear difference from chemotherapy. A 2022 systematic review and meta-analysis of clinical trials in non-small cell lung cancer (NSCLC) helps explain why: the questionnaires captured different aspects of health, and the studies did not all compare identical treatments or follow patients in the same way.
The review, published on 4 October 2022, searched for randomized trials. It included ten eligible publications and examined patient-reported outcomes: symptoms and health ratings reported by patients themselves. Across the analyses, several comparisons favored treatment with PD-1 or PD-L1 inhibitors, a group of checkpoint drugs, but the health utility index result did not show a statistically significant difference. These results are best understood by looking at what each measure asked patients to report.
Three measures, three views of health
The cancer-specific EORTC QLQ-C30 questionnaire asks about several parts of life with cancer. It covers functioning, such as physical and everyday roles, along with symptoms including pain and fatigue, and includes an overall quality-of-life measure. It is designed to capture experiences relevant to people with cancer.
The EQ-5D-3L takes a broader view. Its visual analog scale (VAS) asks a person to rate their overall health on a scale. Its utility index combines responses about health dimensions into a summary value that can be used to compare health states. It is not simply another version of the cancer-specific QLQ-C30 score.
The distinction matters because an average change on one questionnaire need not match a summary index built from different questions. Different results do not mean that patients gave contradictory testimony. The measures emphasize different parts of health, and each pooled estimate reflects the trials that reported that measure.
A later first deterioration in five publications
Five publications reported time to first clinically significant deterioration using the QLQ-C30. This measure asks how long it takes, from a patient’s starting point, for their reported quality of life to worsen by a threshold the study treats as meaningful. It is different from average score change, which compares group averages between baseline and a later assessment.
For overall quality of life, the pooled hazard ratio was 0.86 for PD-1/PD-L1 inhibitor groups compared with platinum-based chemotherapy groups (95% confidence interval 0.76–0.97). In plain terms, the analysis found a lower rate of reaching the first deterioration threshold over time in the checkpoint-treatment groups. The 95% confidence interval—a range of effects compatible with these data—lies entirely below 1, the value for no difference between groups.
The pooled time-to-deterioration results also favored checkpoint treatment for physical functioning, pain, and role functioning. The hazard ratios were 0.79 for physical functioning, 0.78 for pain, and 0.84 for role functioning. These are relative comparisons between groups over time. They do not tell a patient how many extra weeks they personally might remain stable: the hazard ratios do not provide an absolute number of additional weeks.
Average changes favored treatment on two measures, with no clear result on a third
In analyses of change from baseline to follow-up, the pooled difference between checkpoint-treatment and chemotherapy groups favored checkpoint treatment on the QLQ-C30. The mean-change difference was 3.64 points (95% CI 1.62–5.66). On the EQ-5D visual analog scale, the difference was 4.74 points (95% CI 2.65–6.83), also favoring checkpoint treatment.
The utility index gave a less certain result. Its pooled between-group mean-change difference was 0.03, with a 95% confidence interval from −0.01 to 0.07. Because that interval crossed zero, the review did not find a statistically significant difference between the groups on this measure.
These are group averages, not a record of how every participant felt. A favorable pooled mean does not show that each patient improved, or that each person crossed a threshold for an individually meaningful change. The review also reported changes within the checkpoint-treatment group and within the chemotherapy group, but those comparisons answer a different question from the direct difference between treatment groups. A change over time in one group by itself does not establish that its treatment caused a better outcome than the other group’s treatment.
The trials were not one interchangeable comparison
The ten eligible publications covered several checkpoint drugs: pembrolizumab, nivolumab, durvalumab, and atezolizumab. They also included different treatment settings and regimens. Some compared checkpoint treatment with chemotherapy, while at least one included a checkpoint drug alongside chemotherapy. In the PACIFIC trial entry, durvalumab was compared with placebo after chemoradiotherapy. Those are meaningfully different treatment contexts, not copies of a single immunotherapy-versus-chemotherapy experiment.
The outcome analyses also drew on different subsets of publications. The time-to-deterioration result came from five publications reporting QLQ-C30 data; the mean-change analyses included six QLQ-C30 publications, four EQ-5D VAS publications, and three utility-index publications. Follow-up periods ranged from 12 to 66 weeks, with a reported median of 22.5 weeks. Comparing pooled estimates therefore requires attention to which trials and follow-up measurements contributed to each one.
The review found no difference in its subgroup analysis between open-label and double-blind trials. That result should be read in light of the authors’ caution that the number of studies was small. The trial mix does not establish that every checkpoint drug, combination, or treatment setting produces the same quality-of-life effect.
What the questionnaires may leave out
The analysis relied on outcomes reported in published trial papers rather than original participant-level data. Questionnaire results can also be difficult to compare when trials ask patients at different times. In advanced illness, some patients may stop completing questionnaires as their disease progresses, so follow-up results can reflect the people who remained able to respond. The authors specifically identified loss to follow-up during disease progression as a limitation.
The populations were also selected. Recorded performance status in the included trials was 0–1, and people with poorer baseline functioning were sparsely represented. The pooled findings therefore offer limited guidance about how well these estimates apply to patients who were already less able to manage daily activities when treatment began.
Quality-of-life measures are one part of judging treatment. The review notes that checkpoint inhibitors can be associated with immune-related adverse events, but its questionnaire findings do not establish an overall safety advantage. They also do not replace comparisons of survival or treatment harms. A patient-reported benefit on some measures is useful context, not a complete treatment balance.
How this could shape the future
A future consultation could give someone with lung cancer a clearer way to weigh options: alongside survival and immune-related side effects, a dependable estimate of how likely pain or daily functioning is to worsen over a defined period. Such an estimate would be more useful if it were specific to the treatment regimen and applicable to people who start treatment with poorer functioning.
The review offers a reason to pursue that capability. Its five-publication analysis found later first deterioration in several patient-reported areas, while its average score comparisons favored checkpoint treatment on two measures and did not establish a difference on the utility index. For a consultation tool to translate those findings into a personal discussion, future evidence would need to make the measures and treatment comparisons more comparable, not treat the current pooled averages as an individual forecast.
One nearer milestone would be a prospective comparison of matched regimens that specifies in advance how patient-reported deterioration will be measured, reports how many participants complete questionnaires over time, and tests the measures in people with poorer baseline functioning. If studies meet that standard and their results hold up, they could support more useful estimates for treatment conversations. The review is an early step toward that possibility, not the finished estimate.
Key Sources
- Primary source: Quality of life in patients with non-small cell lung cancer treated with PD-1/PD-L1 inhibitors: a systematic review and meta-analysis.
DOI or Journal Page · Review or Meta-Analysis · 2022-10-04
Keep exploring