From the time the Scottish physician James Lind showed citrus cured scurvy in 1747, clinical trials have had a reputation for being gold-standard evidence about the safety and effectiveness of medical treatments. But doubts have recently been raised about their power to guide treatment decisions.
US researchers doing a clinical trial recently found no difference in survival rates of people with aggressive brain cancer taking the drug Avastin (bevacizumab) when compared to those given a placebo (no drug).
Avastin is approved to treat cancers including colorectal, lung, breast, kidney, ovarian and glioblastoma (brain tumor). It features in hundreds of clinical trials, and reportedly earns the world’s biggest biotech company Genentech/Roche about US$6 billion annually in global sales.
The finding raises serious questions about the nature and worth of randomised controlled trials in contemporary cancer care. The fact this well-designed clinical trial found Avastin to be ineffective suggests many previous clinical trials featuring the drug may not provided the gold-standard evidence we have come to expect.
So what is a clinical trial?
Clinical trials are experimental human research studies designed to test the effectiveness of new drugs, surgical procedures or therapeutic devices. Trials are classified into different phases from I to III (and up to IV in Australia).
Phase 1 trials determine safety, pharmacological activity and identify potential side effects. New drugs are carefully administered to small numbers of human participants under close supervision.
In phase II trials, more people are recruited and safety, efficacy, and dosage regimes are tested. Phase III trials determine whether there’s clinical benefit from the therapy for particular health conditions.
Phase IV trials may be undertaken after a drug has been approved. Some phase IV trials are undertaken for post-marketing reasons – to see if the drug works in a natural, less-controlled clinical environment. Researchers aim to assess the real-world effectiveness of a drug, almost like a final check.
The Avastin study cited above is an example of a later phase III or IV study. It’s a final check that went against what was expected.
Clearly, not all clinical trials are of the same quality; trials can evidence good research, bad research, unnecessary research – or worse!
Good research is well-designed and seeks answers to questions that benefit people who are ill as well as the wider public.
Unnecessary research is arguably of more benefit to researchers than participants because it doesn’t answer a therapeutic question but usually ends in the publication of research papers that can be added to the researchers’ publication list.
Researchers may fail to review previous evidence systematically; human research ethics committees unwittingly approve such research; professional journals publish repeated findings; and patients don’t get access to the best evidence-based treatment options.
Bad research is poorly-designed and targeted, and sadly, continues to be undertaken. Worse still, despite increasing research review requirements, research misconduct also continues to take place.
What can be done?
We know that many clinical trials work, and the results can provide good evidence to change clinical practice. But we also know some are not very useful and others cause harm, so what should be done to ensure we only undertake useful trials?
First, the following research protocols must be applied to all clinical trials:
• eliminate confounding influences on the results by having a clear statement of hypothesis, good randomisation, and outcome assessment (internal validity);
• show findings that can be generalised to wider populations than just the subjects of the trial (external validity);
• have a consistent hypothesis in lab, animal and human studies across all phases; and
• provide a comprehensive discussion of any biases in the trial’s method and missing data, and include any negative results (credibility).
And the ethical treatment testing clinical trials should:
• encourage honesty when there are uncertainties about the effects of treatments, and assume higher risk when uncertainty is great;
• confront the double standards around consenting to treatment in clinical trials and in clinical practice (a higher standard for consent is required in clinical research than in clinical practice);
• increase public knowledge about how to judge whether claims about treatment effects are trustworthy;
• increase the capacity for preparing, maintaining and disseminating systematic reviews of research evidence about the effects of treatments;
• effectively tackle scientific misconduct and conflicts of interest within the clinical research community;
• require industry to provide better, more complete and relevant evidence about the effects of treatment (no trial data should be missing); and
• identify and prioritise research addressing questions about the effects of treatment that are important for patients and clinicians.
Finally, we need to develop better-targeted and more robust approval processes for clinical trials.
The Avastin experience has the potential to teach us a lot. Clinical researchers must critically consider the merit and integrity of their research questions before they expose people to new drugs or prevent them from accessing already proven therapeutic regimes - or both.