Safety and tolerability of magnetic resonance imaging-guided convection-enhanced delivery of AAV2-hAADC with a novel delivery platform in nonhuman primate striatum

Abstract

Degeneration of nigrostriatal neurons in Parkinson's disease (PD) causes progressive loss of aromatic l-amino acid decarboxylase (AADC), the enzyme that converts levodopa (l-DOPA) into dopamine in the striatum. Because loss of this enzyme appears to be a major driver of progressive impairment of response to the mainstay drug, l-DOPA, one promising approach has been to use gene therapy to restore AADC activity in the human putamen and thereby restore normal l-DOPA response in patients with PD. An open-label phase I clinical trial of this approach in patients with PD provided encouraging signs of improvement in Unified Parkinson's Disease Rating Scale scores and reductions in antiparkinsonian medications. However, such improvement was modest compared with the results previously reported in parkinsonian rhesus macaques. The reason for this discrepancy may have been that the relatively small volume of vector infused in the clinical study restricted the distribution of AADC expression, such that only about 20% of the postcommissural putamen was covered, as revealed by l-[3-(18)F]-α-methyltyrosine-positron emission tomography. To achieve more quantitative distribution of vector, we have developed a visual guidance system for parenchymal infusion of AAV2. The purpose of the present study was to evaluate the combined magnetic resonance imaging-guided delivery system with AAV2-hAADC under conditions that approximate the intended clinical protocol. Our data indicate that this approach directed accurate cannula placement and effective vector distribution without inducing any untoward effects in nonhuman primates infused with a high dose of AAV2-hAADC.

FIG. 1.

Top: MRI-compatible stereotactic aiming device and cannula. SmartFrame stereotactic aiming device includes a skull-mounted holder for the infusion cannula and controls that permit adjustment of the trajectory while the patient is in the MR bore. Planning and intraoperative adjustment are accomplished with the aim of ClearPoint software as described in more detail by Richardson and colleagues (2011a). Bottom: The SmartFlow MRI-compatible cannula. The step in the outer dimensions prevents infusate reflux. Color images available online at www.liebertonline.com/hum

FIG. 2.

Correlation between gadoteridol and AADC expression. MRI images (first column) show gadoteridol distribution at the end of both caudate nucleus and putamen CED after (a) an uneventful AAV2-hAADC infusion and (b) one with significant infusate leakage dorsally into the internal capsule. Images in the second column correspond to immunoperoxidase staining for AADC at the same level as the MRI pictures. Dashed red lines correspond to the gadolinium signal area from the MRI images. Third column images demonstrate the extent of MPTP-induced lesion as shown by TH immunostaining of a brain section adjacent to an anti-AADC-stained section. Pictures in the last column were obtained at higher magnification in the putamen (white boxes in AADC low-magnification images) and show robust striatal expression of the hAADC transgene with a large number of AADC-transduced neurons and fibers. Scale bar: 50 μm. DDC, DOPA decarboxylase. Color images available online at www.liebertonline.com/hum

FIG. 3.

Lack of tissue damage in target structures. Representative images show cannula tracts with H&E (a and a′), GFAP (b and b′), or Iba1 (c and c′) immunostaining in AAV2-hAADC-treated NHP (left-hand column), PBS control (middle column) animals, and an AAV2-hAADC-treated animal with infusate leakage dorsally into the internal capsule (right-hand column). H&E staining evinced no tissue damage in either AAV2-hAADC-treated (a, left) or in control (a, middle) animals. Similarly, in adjacent tissue sections, GFAP (b) and Iba1 (c) images show a moderate and confined immune reaction in both AAV2-hAADC-treated (left) and control (middle) NHP. However, some cellular infiltration is evident in the animal with infusate leakage into the internal capsule (right, a–c). Black squares indicate (a′) H&E, (b′) GFAP, and (c′) Iba1 staining at higher magnification for both AAV2-hAADC and PBS control animals. Inset on the right in (a′) demonstrates the presence of some eosinophils in the infusate leakage area (black square). Arrows in (a–c) (left images) show slight cellular infiltration exclusively around a blood vessel near the putaminal cannula tract. GFAP-immunostained sections (b and b′) were counterstained with cresyl violet. Scale bars: (a–c) 2 mm; (a′–c′) 100 μm; inset, 20 μm. Color images available online at www.liebertonline.com/hum