Patient-reported outcomes are essential when their availability is prompt

It is nice to see a presentation of patient-reported outcomes (PROs) data from a phase III study make it into a Mini Oral Session at the European Lung Cancer Congress (ELCC) 2022 – the POSEIDON trial, comparing combination immunotherapy plus chemotherapy with standard chemotherapy in first-line metastatic non-small-cell lung cancer (Abstract 5MO).

Often considered a soft secondary endpoint of clinical trials, PROs and quality of life (QoL) information actually play a crucial part in determining the right treatment for patients. Efficacy and toxicity data are indisputably the cornerstones of assessing the clinical value of a treatment. However, survival data and disease control are not the only considerations for patients, particularly those with advanced disease. Some patients may prefer to swap a treatment having a potentially prolonged survival benefit but more toxicity for another option that offers better QoL and improved symptom control, even if it could not offer the same prolongation in survival. QoL effects cannot be assessed solely using the objective measures of efficacy and investigator-assessed toxicity. The subjective nature of PROs, which are evaluated by the patient based on their own perceptions, means that they more accurately reflect the patient’s view of the impact of a treatment on their well-being. This helps clinicians to evaluate and communicate the pros and cons of a treatment in terms that are meaningful to the patient. In the case of the POSEIDON trial, the PRO data favour a triple combination of immunotherapy plus chemotherapy over chemotherapy, supporting the previously reported efficacy findings (J Thorac Oncol. 2021;16(10 Suppl):S844). This makes treatment decisions relatively straightforward, while the situation can become more complicated in situations in which the PRO data go in a different direction to the efficacy data.

Some years ago, in several systematic reviews of phase III oncology trial publications led by our group at the University of Turin, we found that many solid tumour trials did not include QoL as an endpoint (Ann Oncol. 2018;29:2288–2295; Crit Rev Oncol Hematol. 2022 Mar 5:103649). The situation was particularly disappointing for trials in lung cancer, despite its often poor prognosis and high burden of symptoms (Lung Cancer. 2020;139:47–54). The value of PROs has now been recognised by regulatory agencies and oncology societies. The European Medicines Agency (EMA) and the US Food and Drug Administration (FDA) have both provided guidance for their use specifically in the setting of oncology clinical trials, and some societies incorporate the data into instruments developed to define the value of a treatment, such as the ESMO-Magnitude of Clinical Benefit Scale.

However, timing is key when it comes to using PROs for decision-making in oncology. In fact, to enable a full assessment of a treatment, results should be published concurrently with the efficacy and safety data. Unfortunately, this is generally not the case, and our systematic review found that there was underreporting of QoL data in primary papers from trials in which it featured as an endpoint (Ann Oncol. 2018;29:2288–2295; Lung Cancer. 2020;139:47–54). In a recent analysis focused specifically on journal publications of trials conducted with immune checkpoint inhibitors, it is clear that while the majority of trials include QoL among the endpoints, most do not include these results in the primary publication (Crit Rev Oncol Hematol. 2022 Mar 5:103649). The same concept applies to congress presentations, particularly given their wide dissemination and resonance on the web and in social media. Considering the POSEIDON trial, the efficacy data were available in September 2021 (presented at the International Association for the Study of Lung Cancer’s, 2021 World Conference on Lung Cancer), 6 months before the PROs data are reported at the ELCC 2022.

Several reasons could explain why PROs data so often are not published alongside primary efficacy and safety data. The data can be complex, comprising multiple symptoms and QoL domains expressed in a variety of ways, which can make a simple summary difficult to achieve. This can be problematic for publications where word count or timing is restricted, such as in congress abstracts and oral/poster presentations. However, in terms of peer-reviewed publications, most journals now have ample allowance for supplementary material, so lack of space cannot be used as an excuse for underreporting. If the QoL analysis is deliberately separated from the primary results in order to produce a second article or presentation with greater visibility for the study and for the authors, this could also be achieved through a companion paper/abstract, which would still ensure the simultaneous availability of all the relevant results (Ann Oncol. 2016;27(6):961–962).

More needs to be done to ensure that clinicians can access PROs data at an appropriate time. At the moment, journal editors, who could make the provision of PROs data for clinical trials a requirement for publication, may hold the key to improving the timely presentation of patient-reported results.