Prognostic Tools for Cancer Survival: A Secondary Role for Quality-of-Life Measurement

Data from cancer patients, including those with head and neck cancer, suggest that those treated for cure or long-term survival will accept significant morbidity to achieve small survival gains. Even so, almost 25% of patients with head and neck cancer ranked cure as only the second or third most important outcome, with items related to symptoms and functional outcomes, including appearance, ranking among the top three most important. Experienced oncologists understand that for a significant proportion of patients with cancer, maintaining or improving health-related quality of life (HRQOL) is an important objective when considering treatment choices. However, HRQOL is a more abstract notion than is “life or death,” and it is more difficult to measure. Oncologists have tended, particularly in the clinical research domain, to focus on survival as paramount, often at the expense of more relevant clinical outcomes. For example, in a meta-analysis of randomized, controlled trials on the use of palliative chemotherapy for incurable (recurrent and distant metastatic) head and neck cancer, it was found that reports of all studies included survival and tumor response but did not include outcomes that would measure palliative benefits. Several strategies have been explored to address this imbalance and to better align study end points with outcomes that matter to patients. Pain and emotional distress have been promoted as the fifth and sixth vital signs, respectively. Canada has chosen the rebalance focus (now called the cancer journey) as one of the few priority areas for action in its national cancer control strategy. Motivated by the adage “you can’t manage what you don’t measure,” a growing body of literature is promoting the measurement of pain, emotional distress, and HRQOL to focus attention on better managing these relatively neglected areas. In the early 1990s, the National Cancer Institute of Canada Clinical Trials Group instituted HRQOL measurement as a default requirement for all its comparative clinical trials. Clinical teams involved with management of head and neck cancer have benefited from a relatively active focus by researchers on quality-of-life measurement, with several validated instruments now available. This reflects the dramatic effects of the disease itself and the impact of treatment across several HRQOL domains, including physical, social, and emotional. With so much energy spent to refocus the clinical lens on the softer outcomes that matter to patients, it is ironic that the interest and curiosity of clinical oncologists with regard to the value of HRQOL measurement have peaked because of the link between HRQOL and survival, as if the field of HRQOL has gained more credibility because of its association with this outcome. This phenomenon was most obviously illustrated when Spiegel et al reported their observation that a psychosocial intervention designed to address emotional wellbeing in a randomized trial actually had an impact on survival. Despite several attempts, to our knowledge, this observation so far has not been reproduced. In this issue of Journal of Clinical Oncology, Meyer et al report on quality-of-life measurement as an independent prognostic indicator for survival in patients with early-stage (I and II) squamous cell head and neck cancer. This is the fourth in a series of reports emanating from a well-designed and well-executed placebo-controlled, randomized trial. The primary objective of this trial was to evaluate the role of antioxidant medications (ie, beta carotene and vitamin E) in the occurrence of second primary cancers, with overall survival as one of several secondary outcomes. In the three previous reports of this trial, the investigators established that the use of antioxidant medications did not reduce the frequency of second primary cancers, was associated with reduced overall survival, and was not associated with improvement in HRQOL, although there were mixed effects on acute adverse events depending on the subgroup analysis used. In the previous reports in this series, all comparisons were directed at the randomized groups, whereas the report by Meyer et al in this issue of JCO presents results from a pooled analysis of the randomized groups. Multivariate statistics including a Cox proportional hazards model were used to determine the independent contributions of different factors to prognosis, including baseline HRQOL and the change scores at 6 months after completion of radiotherapy. Meyer et al report that for every 10-point difference in baseline HRQOL score, there was a corresponding 13% difference in risk of mortality in the expected direction (ie, favoring better HRQOL). The physical functioning domain was the only change score at 6 months that independently predicted survival, with a mortality hazard ratio of 0.75 at 1 year for every 10 points of change. However, over time, its predictive performance diminished. As appropriately cited by Meyer et al, there have been numerous similar observations of the predictive performance of HRQOL JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 27 NUMBER 18 JUNE 2

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.