Clinical benefit in oncology trials: is this a patient-centred or tumour-centred end-point?

摘要

BACKGROUND: Clinical benefit (CB) was first successfully used as an end-point in 1997 in the pivotal study of gemcitabine in advanced pancreas cancer. In the trial by Burris et al. CB was a composite measure of pain, performance status and weight. Here we describe how CB has been used in oncology trials since that time. METHODS: We performed an electronic search (www.jco.org) for reports of all clinical trials (phase I, II and III) published in the Journal of Clinical Oncology 1997-2008 citing 'clinical benefit'. Eligible trials were those reporting clinical benefit as an end-point. Details related to study methodology, sponsorship and end-points were abstracted. Use of CB was classified as patient centred if it referred to improvement in the clinical parameters used by Burris et al. or in other disease-related symptoms. CB was classified as tumour centred if it related to objective tumour criteria for partial/complete response and/or stable disease. Descriptive statistics were used to summarise findings and the chi-square test was used to compare proportions. RESULTS: Seventy-one trials reporting CB as an end-point were identified: 37 in breast, 8 in pancreas and 26 in other cancers. The definition of CB was patient centred in 20 trials (28%) and tumour centred in 51 trials (72%). Only 20% (14/71) of trials (including all 8 pancreas studies) used the original Burris definition. Among the 71 trials reporting clinical benefit, in only 31 (44%) cases was the end-point defined as a primary or secondary study objective. Trials with a patient-centred definition of CB were considerably more likely to do so than trials with a tumour-centred definition (19/20, 95% versus 12/51, 24%, p<0.0001). Study variables associated with the use of a tumour-centred definition include: disease site (breast 35/37, 95%; all others 16/34, 47%, p<0.001) and intervention (hormone or targeted agent 38/40, 95%; chemotherapy 13/31, 42%, p<0.001). There has been a steady increase in the number of trials using CB as an end-point; in the second half of the study period the number of trials increased from 17 to 54, along with the proportion of trials with a tumour-centred definition (10/17, 59% to 41/54, 76%, p=0.09). CONCLUSIONS: Despite its initial definition, clinical benefit is often used to describe objective tumour findings. Clinical trials should use end-points in a consistent manner to enable clear communication between investigators, clinicians and patients about the benefit of novel therapies.