Gutted adenoviral vectors in hemophilia A

In this issue, Brown and colleagues (page 804) report results of intravenous infusion of helper-dependent (fully deleted) adenoviral vectors encoding canine factor VIII (FVIII) in dogs with hemophilia A. From these findings and from similar reports from other groups,^1,2  a picture begins to emerge of the efficacy and toxicity of helper-dependent adenoviral vectors as a gene delivery vehicle. Although more data are needed, the findings at this point raise the question of whether these vectors are likely to be useful for the treatment of genetic disease, where long-term expression is the goal.

Adenoviral vectors effect high-efficiency gene transfer into nondividing target cells and remain episomal, thus avoiding the risk of insertional mutagenesis. However, a wealth of preclinical and clinical experience with early generation adenoviral vectors documented short-term expression and immune-mediated destruction of transduced cells.³  Since expression of viral proteins encoded by the vectors was implicated as one of the factors in the destructive immune response, several groups developed fully deleted adenoviral vectors, in which viral coding sequences had been completely removed.^4,5  Production of such vectors requires a helper plasmid that supplies necessary adenoviral coding sequences in trans, but which cannot be packaged into the recombinant particles. Production and purification methods vary somewhat in the efficiency with which helper virus is removed, a point that may have implications for comparisons among studies from different groups.

An important question in the field has been whether and to what extent the toxicity profile of adenoviral vectors would be modified by the newer generation of fully deleted adenoviral vectors. Toxicities due primarily to innate immune responses to capsid proteins will not be affected by removing viral coding sequences, while those due to expression of viral proteins should disappear. There are 3 recent studies using fully deleted adenoviral vectors in hemophilia A and B dog models that suggest that the toxicity profile is dose-dependent and that adenoviral gene deletion has reduced but not eliminated vector toxicity. Thus, in the current study by Brown et al, a dose of 5 × 10¹¹ vector particles per kilogram (vp/kg) showed no canine FVIII expression and no toxicity, but a dose of 1.25 × 10¹² vp/kg showed mild transaminase elevation and a drop in platelet count, findings previously documented in large animals with earlier generation adenoviral vectors.⁶  A similar study by McCormack et al, using doses of 1 × 10¹² vp/kg and 3 × 10¹² vp/kg in hemophilia A dogs, also showed minor transaminase elevation and thrombocytopenia.¹  On the other hand, a study by Ehrhardt et al²  in hemophilia B dogs reported no detectable laboratory abnormalities at doses of 6 × 10¹¹ to 8.5 × 10¹¹ vp/kg. These data are at least consistent in suggesting a threshold for detectable laboratory abnormalities of approximately 10¹² vp/kg (of fully deleted adenoviral vectors) in dogs. Of course, it is noteworthy that a human subject treated systemically with a dose of 6 × 10¹¹ vp/kg of an earlier generation adenoviral vector suffered fatal complications,⁶  underscoring the importance of defining the source(s) of adenoviral toxicity.

If the toxicities of fully deleted adenoviral vectors in humans are mild and self-limited, as they are in hemophilic dogs, then they would not in themselves argue against further development of the vectors. However, all 3 groups have also described a rapid decline in levels of the transgene product, with only 1 of 8 dogs showing persistent factor levels in a (barely) therapeutic range (dog C in the study of Brown et al, approximately 1% at 5 months after injection). A detailed analysis of the causes of loss of expression will be required to formulate a strategy to avoid this shortcoming. At this point, however, it seems that fully deleted adenoviral vectors, though they exhibit reduced toxicity, still fall short of the goal of sustained expression at therapeutic levels in hemophilia.