A letter sent by me this morning to Benitec management:
"As I am sure you are aware, a recent paper first-authored by Lisowski and senior-authored by Mark Kay of Stanford has created concern in the minds of many in the investing and medical communities regarding Benitec's TT-034 clinical trial.
The authors posit that, in a mouse/human liver chimera mouse model, AAV8 showed only limited transduction efficacy against human hepatocytes. It goes on to conclude, quite spuriously I feel, that serious aspersion can be cast on clinical plans to utilize AAV8 as a vector for infecting and transducing human livers for the expression of shRNA.
As a Benitec investor and as a hepatologist, I have strenuous objections to this paper, and am surprised it sailed through peer review as ostensibly easily as it did. I suspect that all of you, in deliberations with others and in presentations, are likely to encounter questions about this paper and whether its seeming conclusions diminish your resolve as regards Benitec's clinical program. I would politely assert that your answer should be "no" and would like to say why.
Mice with chimeric human/mouse livers are made, of course, by the killing off of mouse hepatocytes and replacing them with infused human hepatocytes. What results, however, can hardly be regarded as a histologically normal liver. Normal liver has a very specific architecture in which portal tracts (hepatic arteriole, portal venule, and cholangiole) are surrounded by central venules. And each central venule is surrounded by portal tracts, rather like a carpet or necktie pattern.
Between portal tract and central vein are cords of hepatocytes. The area around the portal tract is the highly oxygenated Rappoport zone A. The area around the central venules is Rappoport zone C. The betwixt area is Rappoport zone B. Hepatitis C histopathology is humans is famously anisotropic, with most of the harm and inflammation happening peri-portally, in zone A. The blood perfusional path from zone A to zone C is via the hepatic sinusoids. In ways not yet understood, but that the TT-034 trial may do much to elucidate, HCV's infectivity seems to vary in liver as a consequence of proximity to other types of liver cells that may serve as replicative waystations for virus, including cholangiocytes, endothelial cells, Kupffer cells, and intrahepatic lymphocytes. This, I suspect, is why we see more HCV "action" near the portal triads than in any other liver zone.
In mice with chimeric livers, there is histologic chaos as regards the newly-deposited human hepatocytes. Rappaport zone architecture, sinusoidal architecture, the formation of gap junctions between hepatocytes, and likely the migration of macrophages and lymphocytes....all of these things are distorted and non-native. This leads me to assert that Lisowski and Kay's conclusion that, based on this system, AAV8 is a poor transducer of human hepatocytes, is nothing if not disingenuous. Our understanding of how AAV8 gets into hepatocytes is incomplete, of course, but to argue based on a system where perfusion, juxtaposition to endothelium, and access to non-hepatocyte liver cells is perturbed that AAV8 is ineffective is just not sound science. Theirs is but one observation, an incomplete one, and one based on a highly imperfect model system. Its broader interpretability as to AAV8 efficacy in human patients is utterly indeterminate. In their system, AAV8 may be having at mouse hepatocytes more readily than human ones because the native mouse hepatocytes remain in a setting of normal perfusion and adjacency to endothelial and sinusoidal lining cells. The same absolutely cannot be said of the human hepatocytes in this system.
Anyway, I hope this helps. I am pulling for you and will help in any way that you will allow me to.
Very truly yours,"
Disclosure: I am long BNIKF.