On the heels of recent headlines about clinical holds in gene therapy trials, the FDA's 2-day hearing about adeno-associated virus (AAV) vector safety reviewed the complex problems the delivery method poses.
On Thursday and Friday, the agency's Cellular, Tissue, and Gene Therapies Advisory Committee (CTGTAC) held sessions about oncogenesis risk, liver toxicity, and thrombotic microangiopathy in AAV vectors for gene therapy, culminating in two sessions about neurotoxicity: one about dorsal root ganglia (DRG) toxicity and one about brain MRI findings.
DRG Toxicity
Like carcinomas, neuron damage related to AAV vectors has emerged mainly in preclinical data. Toxicity in the sensory neurons of the DRG largely has not been reported in humans, but in nonhuman primates, AAV delivery through cerebrospinal spinal fluid (CSF) and intravenously has been associated with axonal degeneration in some tracts of the spinal cord and peripheral nerves.
One factor that appears to make a difference in axonal degeneration severity is the transgene itself, noted guest speaker James Wilson, MD, PhD, of the Perelman School of Medicine in Philadelphia and a pioneer in gene therapy research. Decreasing transgene expression in cells that express a microRNA that's present almost exclusively in DRG may reduce pathology, he added.
How DRG toxicity affects nonhuman primates clinically is not clear. Of 483 animals dosed by Wilson's group, neurologic findings emerged in five: four were mild and reversible, and one was severe and required euthanasia.
In those studies, the researchers were not able to measure nerve conduction velocity (NCV), but other gene therapy research has shown that animals rarely have had NCV abnormalities. "I realize it's anecdotal, but in this one study, it seems that NCV may be a good measure of what actually is going on from a pathologic standpoint," Wilson said. NCV evaluations have been limited in human trials so far, but have been used more in recent studies, he pointed out.
If an animal had paresthesia, it would be hard to know it, Wilson added. Likewise, some humans who receive AAV vectors may experience pain or neuropathic symptoms and not be able to express it -- especially young patients or people with cognitive or motor problems -- committee members observed, agreeing that closer monitoring is needed.
New AAV treatment should investigate neurotoxicity carefully, said temporary voting committee member Raymond Roos, MD, of the University of Chicago: "In the case of humans, nerve conduction velocities or MRIs might be valuable before, as well as after, AAV delivery."
Brain MRI Findings
Intraparenchymal administration of AAV vectors has been evaluated in several disorders including a form of Batten disease known as CLN2 disease, said invited speaker Ronald Crystal, MD, PhD, of Weill Cornell Medical College in New York City.
While children with CLN2 disease benefited from AAV vector-delivered gene therapy, showing less decline on motor and language assessments than similar children in a natural history study, brain lesions appeared on MRI, he observed.
"What we noticed is that there were MRI abnormalities at the site of the catheter tip," Crystal said. The finding had no clinical correlates but was consistent with localized inflammation or edema, he pointed out.
"There were no other significant adverse events that we could directly relate to the vector in these children," Crystal said. One aspect of CLN2 disease is chronic seizures so in the short-term, adverse events may be hard to assess, he noted.
"It's not known whether systemic immunosuppression can suppress an inflammatory/immune response such as the type we saw," he added. "I think that's something that needs to be studied in the future."
MRI changes have emerged in other studies where an AAV vector has been injected directly into the brain. Last year, the FDA placed a clinical hold on a Voyager gene therapy trial for Parkinson's disease after some participants showed MRI abnormalities, for example, but what those findings meant was unclear. Even in Parkinson's disease, "it seems like we need new, innovative technical approaches with respect to delivering the AAV vector," Roos suggested.
Imaging changes on MRI appear to be minor, some committee members noted. Vector tracing and other investigations in non-human primates may help clarify MRI abnormalities, others added. MRI-guided infusions and followup MRIs at regular intervals also may mitigate potential CNS injury in humans.
"In all of these cases, we must keep in mind issues related to risk-benefit," Roos noted. "That might be especially the case in certain progressive, difficult-to-treat, human central nervous system diseases."