Research news for health professionals: August 2017

Results from a meta-analysis, published in the Journal of Clinical Oncology, show that metastasis-free-survival (MFS) is a valid surrogate for overall survival (OS) and, if used, could shorten the length of clinical trials. Currently, trials involving men with intermediate or high risk localised prostate cancer can take over 10 years before differences in OS are seen. The ICECaP working group looked at individual patient and trial data from 28 randomly controlled trials (28,905 patients) with a median follow up of 10 years. The study analysed four intermediate clinical end points (ICEs): MFS, disease-free survival, time to disease recurrence and time to metastasis. The ICE were tested for two criteria: 1) Was the ICE correlated with OS at patient level? 2) Was the treatment effect the same for the ICE as for OS at trial level?  

While a correlation was seen between all 4 ICEs and OS, the strongest correlation was with MFS at both the patient level (0.91 [95% CI, 0.91-0.91]) and trial level (0.92 [95% CI, 0.81-0.95]). Similar results were seen when the analysis was done with sub-groups, such as high-risk populations. Of note, the majority of patients (90%) in the MFS analysis were enrolled in radiation trials.  

This study is the first to identify a valid surrogate for OS in trials involving men with localised prostate cancer. However, this doesn’t mean that a surrogate endpoint will always be the best measure. This depends on a number of factors, including size and length of the trial and the magnitude of the effect that clinicians are hoping to see. Since MFS will clearly not be as good at predicting OS as actually measuring OS itself, there are times - i.e. if you are looking for a small impact on OS - where OS would be a more appropriate endpoint then MFS.   

So what does this mean for clinical trial design? This study suggests that using metastasis-free survival as a surrogate endpoint could shorten clinical trials by months or even years. We look forward to continuing analysis by the ICECaP working group which will be testing the reproducibility of these results with more recent trial data involving newer and more relevant prostate cancer therapies. 

A new Cochrane review, involving 31,043 participants from over 100 studies, has looked at the impact of decision aids on people facing treatment or screening decisions. The study found that there was high-quality evidence that decision aids increased participants’ knowledge and decreased uncertainty about which choice to make. There was also moderate-quality evidence that people who used decision aids had a more accurate perception of their risk and were less passive about the decision in question. Finally, patients also felt either equally or more satisfied with the process and their decision after using a decision aid.  

Of particular relevance to prostate cancer, decision aids reduced the number of people choosing invasive surgery (RR, 0.86 [95% CI 0.75-1.00]) or PSA testing (RR, 0.88 [95% CI 0.80-0.98]), while, from a clinical perspective, use of decision aids led to a small increase in appointment time (less than 3 minutes) and were either cost-neutral or cost-saving. In summary, decision aids may be cost-effective tools to increase patient knowledge and improve health-related decision making.