Letter to the Editors: Concerning “Long-term safety and efficacy of AAV gene therapy in the canine model of glycogen storage disease type Ia” by Lee et al.

Dear Editors,

Recently Lee et al. reported seven dogs with glycogen storage disease (GSD) type Ia treated with adeno-associated virus (AAV) vector-mediated gene therapy that apparently prevented long-term hepatic or renal complications (Lee et al 2018). However, there are problems with this conclusion. Notably, Lee et al. describe how dogs underwent 24-hr care receiving glucose feeds every 30–60 min. Furthermore, Lee et al. describe two GSD Ia dogs on this same rigorous dietary therapy that lived >5 years of age with mostly normal laboratory values, implying that gene therapy had few additional benefits over nutritional therapy.

Our group recently published a very similar gene therapy study, but the GSD Ia dogs were maintained on a more typical diet of only three feedings per day (Brooks et al 2018). However, 4/5 AAV vector-treated dogs developed hepatocellular adenomas and carcinoma and 3/5 developed chronic kidney disease. The key difference between the two studies was the degree of nutritional support and not the gene therapy. Both studies demonstrated low hepatic glucose 6-phosphatase (G6Pase) activity in dogs at the endpoint, as well as marked hepatic glycogen accumulation. Figure 1 in Lee et al. also demonstrates a decrease in blood glucose over time in GSD Ia dogs, without specifying if the dogs were fed or fasted. Brooks et al fasted dogs to demonstrate prevention of hypoglycemia. Both studies reported very small amounts of vector DNA at the end of life in most of the GSD Ia dogs. Two dogs in Lee et al had very low vector DNA present in liver at the end of the study (0.01 to 0.02 copies/cell) and correspondingly low G6Pase activity from 2.2 to 3.8% of activity in the liver of normal carriers. One dog in Lee et al had elevated vector DNA (Figure 2; dog GE) in association with low G6Pase activity (Figure 3). In contrast, Brooks et al reported higher vector DNA (0.04–0.16 copies/cell; Supplementary Figure 2) and G6Pase activity (28% of carrier activity; Figure 1) in the liver. Thus the effect of gene therapy was similar in both studies.

Lee et al reported no renal complications, other than proteinuria in one dog and mild, intermittent increases in BUN and/or creatinine in two different dogs (Lee et al 2018). Importantly, recombinant AAV8 vectors highly like the one used by Lee et al had no effect on kidney involvement in G6pc^−/− mice with GSD Ia (see references therein), and therefore gene therapy would not be protective of the kidneys in dogs. The kidney sparing in Lee et al can be attributed to intensive dietary therapy (Lee et al 2018). Brooks et al reported tubule-interstitial lesions and glomerular changes in the older GSD Ia dog cohort treated with AAV vectors (Brooks et al 2018).

In conclusion, both papers demonstrate that gene therapy has greatly prolonged life in GSD Ia dogs and led to low levels of G6Pase activity persistence. However, evidence is lacking that gene therapy in the absence of a strict dietary regimen can prevent hepatocellular tumors and renal disease.

Synopsis:

Gene therapy in the absence of rigorous dietary therapy does not prevent long-term complications of glycogen storage disease Ia.