A research article published in the Annals of Oncology suggests that stopping clinical cancer trials early may lead to a systematic over-statement of treatment effects, leading to potential harm by drugs that are sped to the clinic before researchers complete a thorough analysis. The Italian researchers studied the premature closing of clinical trials in light of recent research that has shown how early beneficial treatment results for new cancer treatment trials often entice investigators to halt the trial. The concern is that the trial is stopped prior to a proper evaluation of all risk and benefits.
The researchers performed an analysis of 25 randomized controlled clinical trials that had been terminated early due to positive outcomes for patients. They found that in recent years, there has been a substantial increase in the number of trials that stop early. Several of the trials were stopped and then used to support an application for marketing authorization at the European Medicines Agency (EMEA) and the United States Food and Drug Administration (FDA). Of 14 stopped trials that were published between 2005 and 2007, 11 of them (79%) demonstrated this pattern.
Co-author Dr. Giovanni Apolone (Head of the Laboratory of Translational and Outcome Research in Oncology, Mario Negri Institute for Pharmacological Research, Milan, Italy) notes that, “This suggests a commercial component in stopping trials prematurely. In fact, this strategy (i.e. stopping trials early for benefit) could guarantee quicker access to the market for companies. On the other hand, a quicker clinical drug development may lead to an ‘immature’ benefit/risk balance of new drugs.”
“When we analyzed 25 trials over a ten year period between 1997 and 2007, we found a consistent increase in prematurely stopped trials – more than 50% were stopped within the last three years. We are aware that trials stopped early because they are showing benefit may result in the identification of promising new treatments for patients. However, findings obtained following this strategy should be considered to be preliminary results that require subsequent confirmation. We believe that only untruncated trials can provide the full level of evidence required to safely translate treatments into clinical practice. Without such evidence, unsafe and ineffective drugs could be marketed and prescribed, and patients’ health could be jeopardized,” added Apolone.
The 25 trials Apolone and his group analyzed ranged in length from 12 to 64 months, with a mean duration of 30. The researcher contrasts this with the fact that long-term benefits or adverse side-effects are usually identified after several years.
“If a trial is evaluating the long-term efficacy of a treatment of conditions such as cancer, short-term benefits, no matter how significant statistically, may not justify early stopping. Data on disease recurrence and progression, drug resistance, metastasis, or adverse events, all factors that weight heavily in the benefit/risk balance, could easily be missed,” write the authors.
They add, “An early stop may reduce the likelihood of detecting a difference in overall survival (the only relevant end point in this setting) because of the small sample, the possibility of crossing-over the experimental drugs, and contamination with other treatments.”
Results based on small sample sizes are also a concern. Apolone and colleagues found that five studies had enrolled less than 40% of the total number of participants that were to be used in the final analysis. There is a risk that these small sample sizes may increase overestimation of treatment effects.
“We believe that, in general and unless harm is being shown, cancer trials should be completed so that the correct number of patients over the correct period of time required for a significant final analysis can be recruited. Our study describes a growing phenomenon of trials stopped prematurely, which is now the subject of debate within the international scientific community,” says Apolone.
At a news briefing announcing the results of the study, Professor Stuart Pocock (Professor of Medical Statistics at the London School of Hygiene and Tropical Medicine, UK), who was not involved in the study, said: “Clinical trials need to stop early for superior benefit whenever there’s proof beyond reasonable doubt that the new treatment really is superior. That would be an ethical obligation. However, too many trials are stopped early claiming efficacy without such strong evidence being available.”
Stopping a trial early in oncology: for patients or for industry?
F. Trotta, G. Apolone, S. Garattini, G. Tafuri.
Annals of Oncology (2008).
doi:10.1093/annonc/mdn042
Click Here to View Journal Website
Written by: Peter M Crosta