A dose-ranging study of AAV-hAADC therapy in Parkinsonian monkeys

Abstract

The main medication for idiopathic Parkinson disease is L-Dopa. Drug efficacy declines steadily in part because the converting enzyme, aromatic L-amino acid decarboxylase (AADC), is lost concomitant with substantia nigra atrophy. Over the past decade, we have developed a gene therapy approach in which AADC activity is restored to the brain by infusion into the striatum of a recombinant adeno-associated virus carrying human AADC cDNA. We report here the results of an investigation of the relationship between vector dose and a series of efficacy markers, such as PET, L-Dopa response, and AADC enzymatic activity. At low doses of vector, no effect of vector was seen on PET or behavioral response. At higher doses, a sharp improvement in both parameters was observed, resulting in an approximate 50% improvement in L-Dopa responsiveness. The relationship between vector dose and AADC enzymatic activity in tissue extracts was linear. We conclude that little behavioral improvement can be seen until AADC activity reaches a level that is no longer rate limiting for conversion of clinical doses of L-Dopa into dopamine or for trapping of the PET tracer FMT. These findings have implications for the design and interpretation of clinical studies of AAV-hAADC gene therapy.

FIG. 1

Effect of dose of AAV-hAADC on FMT-PET. Pairs of rhesus monkeys were infused in each striatum with the indicated dose of vector. Note that 1 unit AAV = 1 × 10⁹ vg. Animals shown as receiving a dose of 0 received 500 units AAV-GFP per striatum. FMT-PET was performed on all animals 4–6 weeks after vector infusion. Individual PET data points from the lesioned (ipsilateral) side of the brain are presented as a ratio of striatal signal to background signal derived from cerebellum. PET ratios fell into two groups: those in box A have values that approximate values obtained from a historical mean of 12 MPTP-lesioned animals of 1.51 ± 0.23. The vertical dimensions of boxes A and B indicate the range of values. After lesioning, animals displayed ipsilateral ratios of 1.51 ± 0.23 (n = 12). In contrast, data in box B are within the range of ratios found in normal brain (2.72 ± 0.63; n = 8). This segregation of data was used to evaluate corresponding clinical response to l-Dopa in the two groups, and the results from this analysis is shown in Fig. 3. Two animals in group A are from the control group that received no AAV-hAADC but received 500 units AAV-GFP per striatum instead. Group A contains data from seven animals, and group B five animals.

FIG. 2

Correlation between vector dose and FMT-PET ratio within groups A and B from Fig. 1. Data from Fig. 1 were replotted to expand the x axis in the (A) low and (B) high vector dose range. The curve fit in each case was performed by means of the least-squares method.

FIG. 3

l-Dopa dose response in groups A and B from Fig. 1. Animals were sorted into two groups, A and B, based upon whether their PET scan gave a ratio of (A) less than or (B) greater than 2.0. Note that all animals with a score of greater than 2.0 had received a dose of AAV-hAADC greater than 55 Units AAV per striatum. In this experiment, animals were exposed acutely (see Materials and Methods) to various doses of l-Dopa 1 month before (blue) or 6 months after (red) infusion of AAV-hAADC vector. One of the AAV-CFP animals was excluded from this analysis because it developed postsurgical behavioral abnormalities that interfered with testing. Thus, (A) represents data from six animals, and (B) represents data from five animals.

FIG. 4

Effects of dose of AAV-hAADC on AADC activity in tissue extracts. Brain tissue was isolated from animals euthanized 6 months after AAV-hAADC infusion. Punches from several brain regions (putamen, globus pallidus, and caudate) were taken and flash-frozen. Thawed samples were homogenized and assayed for AADC activity as described under Materials and Methods. The data represent a pool of the three sampled tissues to correlate as closely as possible with the region of interest for the PET scanning. However, since the three regions in question are not of equal size, their overall contribution to specific activity also differs. From the MRI scans made of all 12 animals, we were able to weight the relative contribution as follows: putamen:caudate:-globus pallidus 1.0:0.5:0.3. The activity obtained represents the sum of the adjusted activities from the three regions.