When researching the effectiveness of new oncological agents, surrogate endpoints, such as tumor size, are often initially considered. Simply because waiting for effects on the clinical status of a patient takes too long.
It is precisely on the basis of these results that, as is known, these medicines are often conditionally approved. The idea is that the accelerated approval will be confirmed in subsequent research, by following the same research group in an RCT, but now also collecting data on clinically relevant endpoints. In a review in the NEJM about the grounds on which the American medicines authority FDA authorizes new oncological medicines, Bishal Gyawali et al. argue that researchers, especially when it comes to advanced cancers, should be obliged to measure and report the experiences of patients. Their review can be read as a guideline with which the quality of oncological research can be assessed, even regardless of the requirements of the FDA.
For example, in recent years the FDA has approved several new cancer drugs after single-arm studies comparing drugs to historical controls. This is only acceptable in exceptional cases, say Gyawali et al. For example, when dramatic effects have been demonstrated in this way in ultra-rare cancers for which there are no alternative treatment options.
Another methodological point they make concerns ‘progression-free survival’ as a long-term outcome. Gyawali et al. do not consider this a validated surrogate measure for overall survival or quality of life in many cases: ‘The term “progression-free interval” can describe this measure more accurately, because survival is not always improved.’
They further point to the Magnitude of Clinical Benefit Scale of the European Society for Medical Oncology. This is a validated scale that not only takes into account the differences in median survival between treatment groups, but also the toxic effect profile and its consequences for the quality of life.
And keep it fair, they finally say: the intervention used in the control group must always be the correct available therapy. Because if the intervention is substandard, the differences found, no matter how significant, will be of little clinical significance.