Statistical controversies in clinical research: long-term follow-up of clinical trials in cancer

Long-term follow-up is an important unmet need for the full analysis of new treatments for cancer. Earlier detection of cancer and more effective treatment have led to many more patients surviving for more than 5 and even 10 years, so that evaluating late recurrences and side-effects is an increasingly important issue. This is particularly relevant for oestrogen receptor-positive breast cancer, where the existence of late recurrences is well documented. However, survival for other cancers, notably prostate, colorectal and cervix cancer, has dramatically increased in recent years due to screening and better treatment of early lesions. Trials of preventive therapies have an even greater need for long follow-up. Here, we review these issues and suggest ways in which provision for long-term follow-up can be improved.