Surrogate end points of quality of life assessment: have we really found what we are looking for?

Outcome research is a new interesting field in medical research. Some years ago, a document of the American Society of Clinical Oncology distinguished the outcomes of a treatment into patient-outcomes (overall survival and quality of life) and cancer-outcomes (response rate), giving higher priority to patient outcomes. This document is one of the best structured instruments to evaluate and classify the outcomes in clinical oncology. Nevertheless, although overall survival and quality of life represent the main patient outcomes in clinical oncology, in the last years many researchers tried to overcome these recommendations, creating new surrogate end points to assess overall survival and quality of life. Surrogate end points can be useful tools when they are used to achieve preliminary data that anticipate the evaluation of the final outcome, but the use of surrogate end points instead of the main outcomes is quite dangerous, as it can provide wrong answers to clinical questions. The use (or abuse) of surrogate end points of quality of life has recently favoured some questionable decisions of the main regulator organs, such as the approval by the Food and Drugs Administration of the use of gemcitabine in advanced chemotherapy-naive pancreatic cancer, or mitoxantrone in the palliative treatment of hormone-resistant pancreatic cancer, based on the improvement in clinical benefit (a non-validated instrument to evaluate the outcome of palliative chemotherapy) besides a minimal and questionable overall survival, or pain control (evaluated with a non-validated instrument). A correct use of surrogate end points of quality of life within and not instead of quality of life assessment should be the engagement of our further efforts in quality of life research.