One of the bigger questions raised during the Interdependence 2012 Global Spinal Cord Injury Conference held last week in Vancouver: If clinical trials are so hard, so expensive, so tricky to get approved, and so prone to failure, what are the best chances to succeed the next time a trial is planned related to spinal cord injury?
Brian Kwon, a physician/scientist at the University of British Columbia, raised that question in a panel called “Key Challenges Related to Translating Experimental Research Discoveries into Clinical Therapies.” Kwon didn’t need to remind his colleagues that many years of SCI science have not produced therapies. But the record is much worse in the field of stroke: over 1000 experimental treatments led to 114 human trials – every single one failed. Kwon asks, are SCI scientists that much smarter or have better biology and better designed trials than the stroke people? “Are we that different,” Kwon asked.
His point was twofold: The SCI field ignores the lessons from the stroke studies at its own peril. And, more importantly, how about establishing a way to grade pre-clinical evidence that a therapy (either neuroprotective or for recovery) really works before it goes to trial, thus directing resources at the best opportunities?
Using results of a survey of 324 researchers and clinicians, Kwon et al
came up with a system to quantify or score the potential of a trial. See the 2011 paper, “A Grading System To Evaluate Objectively the Strength of Pre-Clinical Data of Acute Neuroprotective Therapies for Clinical Translation in Spinal Cord Injury
The respondents agreed, for example, that a large animal model should be included in pre-clinical studies (80 percent). Half agreed that the window of application should be between six and eight hours. Ninety-six percent agreed that an independent lab should replicate findings if the trial is based on the work of a single lab. Interestingly, when scored on this system, no drug currently in a clinical trial or planning to go to trial scored more than 50 out of 100 possible points. From the paper:
The scoring system, in essence, attempts to reflect how extensively a particular therapy has been studied, and any given treatment logically accrues points and a higher score as the body of peer-reviewed literature on it incrementally grows. For the subscores of “time window of efficacy” and “independent replication,” there is an ordinal progression of the weightings reflecting the simple concept that longer time windows and more independent studies would be more predictive of clinical efficacy. For the other subscores of “animal species,” “injury models,” and “demonstration of clinically meaningful efficacy,” the weightings are assigned categorically, reflecting the value in a therapy being studied in a multitude of settings and therefore accruing credit for the breadth of investigation. This scoring system also provides an objective perspective regarding the “translational readiness” of an SCI therapy that has been demonstrated in one single study from one lab. While such a “perspective” might seem obvious, human translation has proceeded in the past on the basis of this very limited peer-reviewed pre-clinical literature.
As you might suspect, commerce, patent secrecy and the rules of the marketplace sometimes move trials along faster than basic research might dictate. (Look at the Geron embryonic stem cell trial, for example: only one species studied, no large animal model, no replication, no published literature. Stopped before enrollment quotas reached.) From the Kwon paper:
While scoring systems and checklists may have some “common sense” academic appeal, the SCI field has previously moved forward with clinical trials of a number of potential SCI treatments that have not met many of the criteria that are included in this grading scale (e.g., independent replication, demonstrated efficacy in more than one animal or injury model, time-window studies). Clearly, there are also other forces at play that influence the translation of a particular treatment into human patients.
So what are the best candidates for a neuroprotective (acute) trial? Kwon cites several systemic drugs, including erythropoietin, systemic hypothermia, non-steroidal anti-inflammatory agents (NSAIDs), anti-CD11 antibodies, minocycline, progesterone, estrogen, magnesium sulfate, riluzole, polyethylene glycol, atorvastatin, inosine, and piaglitazone. For a nice, recent review of these, see “A Systematic Review of Non-Invasive Pharmacologic Neuroprotective Treatments for Acute Spinal Cord Injury
,” from Kwon’s group.
They only looked at treatments that met the following criteria:
• Studies in which the testing of the therapy was performed in an in-vivo animal model of SCI. Studies that were exclusively in-vitro experiments were excluded.
• Studies in which the spinal cord is traumatically injured. Non-traumatic local or global ischemia models and photochemical reaction models were excluded, as were traumatic root avulsion or dorsal root entry zone models.
• Studies in which the application of the therapy was via the systemic circulation. This included agents administered orally, or via subcutaneous, intraperitoneal, or intravenous injection. Studies in which the therapy was applied directly to the cord or via intrathecal injection/infusion were excluded.
• At least two peer-reviewed publications available on the therapy.
For the list of applied drugs (administered at the injury site) see “A Systematic Review of Directly Applied Biologic Therapies for Acute Spinal Cord Injury
.” The list contains three promising drugs: chondroitinase ABC, anti-Nogo approaches, and Rho antagonists (e.g. Cethrin). Anti-Nogo and Cethrin (owned by privatre biotech companies) have been in trials already. No results have been released on anti-Nogo; Cethrin was shown to be safe and somewhat effective but requires a larger study, which is planned.
Note: Chondroitinase might move toward trial and it might have applicability to chronic SCI; there are reports of its effect a month after trauma. But other than that, there are no treatments in the pipeline to score for long-term SCI.
I won’t dive into the stroke literature, except for this paper
, which helped the SCI field get moving on quantifying its drug candidates. This is one of several papers from a series of meetings of the Stroke Therapy Academic Industry Roundtable (STAIR), explaining how that trial-defeated field addressed the need for “rigorous, robust, and detailed preclinical evaluation” of potential therapies. From the paper:
The most important points for clinical investigators to assess before considering participation in a trial with a new neuroprotective agent are: (1) an adequate dose-response curve with corresponding serum levels defining at least the minimally effective and maximally tolerated doses in at least 1 species, typically the rat; (2) time window studies showing benefit when therapy is initiated at delayed time points after stroke onset in animal models; (3) adequate physiological monitoring was performed in randomized, blinded animal studies and that treatment effects are reproducible in 2 laboratories, 1 of which is independent of the sponsoring company; (4) outcome measures should include both infarct volume and functional assessment in both acute and long-term phase animal studies; (5) initial studies should be done in smaller species such as rodents subjected to permanent occlusion models, unless the mechanism of drug action suggests that reperfusion will be necessary for drug effect. In this case, clinical development probably will be linked to reperfusion therapy. A second larger species (cats, primates) should be strongly considered for further preclinical assessment for novel, first-in-class drugs; (6) the data should be published or submitted for review in a peer-reviewed journal.
Here’s STAIR’s take on recovery in the days, months and even longer after stroke:
The mechanisms promoting functional recovery after ischemic stroke are not entirely clear but most likely depend on functional and/or structural reorganization of the remaining intact brain. Studies in humans with stroke show that recovery may be robust for at least the first 3 months after the stroke, but further recovery may continue thereafter. This prolonged time window of opportunity to intervene on the stroke recovery process offers a substantial and currently unexploited opportunity for drug development.