Predicting survival using HRQOL information – Major findings recently reported in The Lancet Oncology

The largest ever performed meta-analysis of health related quality of life (HRQOL) data of the EORTC QLQ-C30 was performed by Chantal Quinten and coworkers, from the EORTC Quality of Life Department, on behalf of the EORTC Clinical Groups.[1] This large scale study took over three years to complete, but the results provide the most robust and compelling evidence amassed to date that HRQOL scores can provide additional prognostic information that can be used to assist in the prediction of survival in cancer patients.

These results, published 19 August 2009 via fast track online in The Lancet Oncology, were based on information provided at baseline by 7417 patients who had completed the EORTC QLQ-C30 questionnaire. The patients were participants in 30 randomized controlled trials conducted by the EORTC between 1986 and 2004, and 11 different cancer sites were represented in this analysis.

Variables assessed in the meta-analysis included age, sex, WHO performance status, distant metastases, cancer site, and the 15 QLQ-C30 scales. Using Cox proportional hazard models, the HRQOL parameters, physical functioning, pain, and appetite loss, provided significant prognostic information as did age, sex, and distant metastases, but this was not the case for WHO performance status. Robust and highly sophisticated statistical testing fully supports these findings.

The predictive accuracy of prognosis of overall survival was increased by 6% when the three HRQOL parameters (physical functioning, pain, and appetite loss) were considered in combination with the two socio-demographic variables (age and sex), and the two clinical variables (distant metastasis and WHO performance status)  relative to simply considering the socio-demographic and clinical variables alone. These results, along with two other abstracts presented at ASCO in 2008, strongly indicate that in some cancer populations select EORTC QLQ-C30 HRQOL scales provide valuable prognostic information when combined with socio-demographic and clinical information. So, not only can clinicians use HRQOL information to better understand the side effects of new and novel treatments, but now using HRQOL data in prognoses could also be of significant help in predicting survival in select cancer patients.

John Bean

[1] www.thelancet.com/oncology Vol 10 September 2009