From laboratory to community: towards better translation of spinal cord injury treatments

Release date: 27-May-2010

Organisation: Combined Neurosurgical and Orthopaedic Spine Program (CNOSP), Department of Orthopaedics, University of British Columbia, Vancouver, British Columbia, Canada.

Interest in spinal injury research to protect and repair the damaged cord has intensified in the last two decades. We are all aware that there are unproven and experimental procedures being offered around the world, which have been attracting a lot of community, as well as media, attention. Nevertheless researchers and clinicians are still working hard on finding effective treatments for spinal cord injury (SCI). Often these treatments are felt to be just “around the corner,” although we have all heard of promising outcomes from preclinical studies which then fail when they reach human trials.

Brian Kwon and colleagues have published a survey of the opinions and perspectives of 324 clinicians and scientists working to develop promising interventions which will translate into effective treatments for those with SCIs. The aims of the survey were to establish what preclinical evidence is needed to get a treatment into a successful clinical trial, how this evidence could be obtained and to identify some of the biases that are involved in interpreting this evidence. They discuss the similarities with the stroke community, where inconsistent and sometimes inadequate preclinical experimentation resulted in many treatments failing in clinical trials. As a response to these failures the stroke community introduced “evidentiary milestones”.  It is clear that SCI research lacks clear guidelines or a practical and realistic framework to maximise the chances of developing treatments that will work for individuals with a SCI.

The authors suggest the SCI research community needs to establish “how much is enough?” There was no doubt about the necessity of live animal experimentation but it was apparent that rodent models were not always enough to establish the appropriate preclinical evidence. The majority of respondents considered that large animal models, such as cats, dogs, rabbits, sheep, etc., were necessary to obtain meaningful preclinical evidence. In the case of invasive cell transplant therapies the addition of primate models was also deemed necessary. Some of the issues with this kind of comprehensive experimentation are the accessibility and availability of these types of models. The paper also discusses the enormous expense involved with conducting these types of experiments. Kwon and his colleagues commented that “the vast majority of scientists would find large animal or primate studies simply unattainable.”

Kwon and colleagues surveyed clinicians and scientists about the types of models that were required to obtain clinically meaningful evidence and found there was a need for different types of models to be tested, as well as different severities of injury, so the variation in human injury could be replicated. For potential trials on cervical injuries there was seen to be an overwhelming need for confirmation in a cervical injury model. The contusion injury model was thought to best represent human SCIs due to the fact that most injuries in people are caused by a sudden blunt trauma that bruises, rather than cuts, the cord. Respondents were asked to predict the time window between injury and treatment in animals that would be relevant to humans. For acute (early) interventions, the time window of hours to days was thought to be similar in animals and people. However, for chronic trials, where treatments may be given 18 months after the injury, the majority thought a delay of 6 weeks to 3 months was acceptable in animals; however another large section of respondents required a delay of 12 months or more. Clinically meaningful evidence was thought to consist of an improvement in behavioural ability, which can include hindlimb locomotor function and reduction in neuropathic pain. There was a strong requirement to replicate results in an independent laboratory before a treatment is moved into a clinical trial.

Kwon and his colleagues also tried to pin point some of the bias in the way preclinical research results are interpreted. One issue was the problem that negative results are generally not published. One reason for this may be the poor reflection it can cast on researchers, especially when the research is the main focus of their career or place of work. Another reason may be the reluctance of high impact journals to publish such data. Whatever the reason, not publicising negative results slows progress because knowing what doesn’t work can be as important as what does and, if it is unknown, the same studies may be repeated unnecessarily. The survey also highlighted concerns about modifying or repeating results to obtain the desired outcomes. Respondents were divided on the need for researchers being “blind” when assessing experimental or control groups.

SCI clinicians and researchers are driven by one goal – to discover effective treatments for individuals with a SCI. Currently, promising preclinical interventions often go to trial without replication by different labs or duplication in different animal models. Kwon and his colleagues have uncovered some interesting information on the opinions and perspectives of clinicians and scientists. They suggest that a framework for guiding preclinical studies would be helpful and stress that the extensive dialogue between SCI researchers and translators needs to continue.

Gemma Macdonald
Spinal Cord Injury Network

May 2010

Kwon, et al. (2010) Translation Research in Spinal Cord Injury: A Survey of Opinion from the SCI Community. Journal of Neurotrauma, 27; 21–33.

Link to abstract:

Go Back