Abstract
C
It is important to note that the time course of the hepatotoxicity was very rapid and was not associated with the development of an effector T-cell response to the AAV capsid. Thus, this appears to represent a different form of toxicity from that observed in the human hemophilia trials. 2 More specifically, the peak of transaminase elevations in the current study was at 5 days after infusion, a time course suggestive of oxidant injury. As the authors point out, this is reminiscent of that observed with acute acetominophen poisoning, which results from diffuse hepatocellular injury caused by release of excessive reactive oxygen species and depletion of reduced glutathione and other intracellular, particularly mitochondrial, antioxidant molecules. 3
There are important limitations to the study published here. First, the number of animals included in the work was small (n = 3 per species), and since all of the animals were treated with the same dose of vector, it was not possible to explore dose–response relationships with either form of toxicity. Second, the entire set of data was generated with only two distinct batches of AAVhu68 vector produced under similar conditions, and it remains to be proven whether the toxicity was due to adventitious contaminants in the preparations. Finally, it is important to point out that while the capsid and transgene in these studies were similar to those in human trials with type 1 spinal muscular atrophy (SMA), they are not identical. 4 It is possible that the amino acid differences between AAV9 and AAVhu68, while not quantitative numerous, are functionally important. Thus, it is possible that these findings were specific to AAVhu68 and not a property of AAV9-like vectors in general. Both vectors are members of Clade F, but there are two amino acid residue differences. 5 Furthermore, it is important to note that the human version of SMN1 was expressed from these vectors, rather than the SMN1 native to each species. Finally, it remains to be seen whether overexpression of this transgene and/or very high numbers of vector copies per cell could have induced some form of stress response in this instance.
Arguing against the hypothesis that this is peculiar to the AAVhu68 capsid is additional data to be published online at Molecular Therapy, in which these investigators evaluated another AAV9-like vector, PHP.B, in a NHP (N = 1), which showed similar results. 6,7 This paper utilized a different AAV9-like capsid and a different transgene, and yet the toxicology profile was similar. This supports the concept that above a certain dose threshold, systemic AAV9-like vectors can reproducibly elicit acute, severe hepatotoxicity that appears not to be dependent on pre-existing immunity.
The effect on DRG neurons is also interesting and may tie to previous observations that other AAV serotypes can transduce DRG neurons 7 after systemic delivery, and at least one clinical observation in a patient with amyotrophic lateral sclerosis who received rAAVrh10 by intrathecal infusion (Robert Brown, pers. commun.).
As important as all of the specific findings of these studies may be, it is arguably much more important that the interpretation of and reaction to these findings is handled by the gene therapy community in a balanced and responsible fashion. The findings in these papers are likely to be highly controversial. Many laboratories and biotechnology companies are heavily invested in the success of systemic rAAV gene therapy, albeit generally applying lower vector doses and other capsids. It should be noted here that doses as high as those described by Hinderer et al. are generally only required when there is a need to cross the blood–brain barrier or transduce large masses of muscle. Even though the commercial interests are legitimate, the editors of this journal suggest that the one and only guiding principle in critical moments such as this should be the welfare of patients with the diseases being treated with these therapies.
From that perspective, it is crucial that the members of our field neither ignore nor overreact to these findings. If we were to err on the side of overreaction and suddenly halt all systemic AAV9, we may be depriving patients and families of truly lifesaving therapy. The case of type 1 SMA is a perfect example of this, where the therapeutic responses to rAAV9-SMN1 gene therapy (when given at the 2 × 1014 vg/kg dose) have completely altered the prognosis of that tragic disease. To be clear, we would not advocate halting either those trials or trials of other potentially life-sustaining AAV gene therapies. To the contrary, it is important that investigators gather as much safety and biomarker data as possible from patients receiving high doses of AAV9 intravenously.
It would be equally irresponsible if we were to err on the side of ignoring or minimizing these findings. That course of action could lead to preventable harm to one of our genetic disease patients, many of whom are children. In addition to our responsibility for the welfare of our patients, such an event would result in irreparable harm, not only to the affected patients and families, but also to the reputation of the field and its scientists, to the value of commercial interests, and most importantly to the moral standing of the scientists and physicians in this field. Therefore, we advocate for additional carefully controlled, rigorous safety studies using different doses, vector preparations, time points, and transgenes to define fully the reproducibility and parameters of these apparent toxicities.
Finally, we advocate for full transparency with all relevant safety data from all laboratories, whether within academia, industry, or government. While not all sources of vectors are of equivalent quality and purity, the true properties of possible AAV9(-like) vector toxicities will inevitably be best illuminated by multiple different studies being performed by multiple independent groups. This is not a time to protect proprietary interests or academic rivalries. We assert this not out of naïveté but rather based on our conviction that if the gene therapy research community makes patient welfare its sole priority, it will ultimately be beneficial to us all.