Nonthermal ablation in the rat brain using focused ultrasound and an ultrasound contrast agent: long-term effects

Abstract

OBJECTIVE Thermal ablation with transcranial MRI-guided focused ultrasound (FUS) is currently under investigation as a less invasive alternative to radiosurgery and resection. A major limitation of the method is that its use is currently restricted to centrally located brain targets. The combination of FUS and a microbubble-based ultrasound contrast agent greatly reduces the ultrasound exposure level needed to ablate brain tissue and could be an effective means to increase the "treatment envelope" for FUS in the brain. This method, however, ablates tissue through a different mechanism: destruction of the microvasculature. It is not known whether nonthermal FUS ablation in substantial volumes of tissue can safely be performed without unexpected effects. The authors investigated this question by ablating volumes in the brains of normal rats. METHODS Overlapping sonications were performed in rats (n = 15) to ablate a volume in 1 hemisphere per animal. The sonications (10-msec bursts at 1 Hz for 60 seconds; peak negative pressure 0.8 MPa) were combined with the ultrasound contrast agent Optison (100 µl/kg). The rats were followed with MRI for 4-9 weeks after FUS, and the brains were examined with histological methods. RESULTS Two weeks after sonication and later, the lesions appeared as cyst-like areas in T2-weighted MR images that were stable over time. Histological examination demonstrated well-defined lesions consisting of a cyst-like cavity that remained lined by astrocytic tissue. Some white matter structures within the sonicated area were partially intact. CONCLUSIONS The results of this study indicate that nonthermal FUS ablation can be used to safely ablate tissue volumes in the brain without unexpected delayed effects. The findings are encouraging for the use of this ablation method in the brain.

Keywords: ETL = echo train length; FOV = field of view; FSE = fast spin echo; FUS = focused ultrasound; H & E = hematoxylin and eosin; LFB = Luxol fast blue; ablation; brain; focused ultrasound.

Fig. 1

Experimental setup. The experiments were performed with a FUS insert developed for a 4.7-T animal MRI. Figure is available in color online only.

Fig. 2

T2*-weighted MR images showing hypointense spots produced immediately after each sonication. Four overlapping targets (spacing: 1 mm) were sonicated to ablate a volume in the putamen. The ablated area partially overlapped the genu of the corpus callosum. Figure is available in color online only.

Fig. 3

Axial MR images acquired immediately after FUS ablation and 2, 4, and 7 weeks later. Immediately after sonication, the ablated region appeared as a slightly hypointense area surrounded by a hyperintense rim in T2-weighted MRI (T2WI). Two weeks later, the sonicated volume was mostly uniform and hyperintense, presumably indicating the production of a fluid-filled cyst. The size and appearance did not vary significantly at Week 4 or 7, although some enlargement of the lateral ventricle was observed. In T2*-weighted MRI (T2*WI), the ablated region was initially hypointense. At Weeks 2–4, the outer rim of the lesion was slightly hypointense. A residual dark spot (arrows) was present in both T2- and T2*-weighted imaging. Figure is available in color online only.

Fig. 4

Coronal MR images acquired immediately after sonication and 6 weeks later. Each row shows 3 neighboring image planes with the middle image centered on the ablated area. Immediately after sonication, the lesion was clearly evident as a hyperintense region in T2-weighted imaging and as a starkly hypointense area in T2*-weighted imaging. On T1-weighted images obtained after administration of a contrast agent (T1WI+C), signal enhancement was evident throughout the lesion except at the corpus callosum (arrows). Six weeks later, all that remained was a fluid-filled region in T2-weighted imaging and a narrow hypointense rim in T2*-weighted imaging. The corpus callosum appeared partially intact. Figure is available in color online only.

Fig. 5

Cross-sectional area of the lesions measured in the focal plane in MRI shortly after sonication (Day 0) and at later times. The areas were obtained by manually segmenting the hyperintense regions evident in T2-weighted imaging and the hypointense regions evident in T2*-weighted imaging for each animal. Areas in T2*-weighted images were measured only at Day 0. Upper: Graph showing these areas for each rat, with average areas (mean ± SD) shown for the T2-weighted images at 14 days and onward. The areas in T2-weighted images were significantly reduced after Day 0. In most cases the areas in T2*-weighted images were similar to the final lesion size in T2-weighted images. However, in 4 animals (Rats 1, 7, 9, and 12), the final areas in the T2-weighted images were markedly smaller. Lower: Plot of the area measured at each imaging session as a function of time for each rat. The areas are shown relative to their initial value in the T2-weighted images as a function of time after sonication. The relative area in T2*-weighted imaging is shown for reference. In each animal, the affected area in the T2-weighted images was largely reduced by Day 14 and was mostly stable thereafter. In no case was significant lesion growth observed. Figure is available in color online only.

Fig. 6

MR images and photomicrographs of an H & E–LFB–stained section obtained after microbubble-enhanced FUS ablation in the putamen. A: Axial T2*-weighted (upper) and T2-weighted (lower) MR images acquired immediately after sonication. The ablated area was segmented in each image. B–E: Photomicrographs of an H & E–LFB stained section obtained 4 days after sonication. A well-defined area was present at the targeted region (B–C). Segmentations from the different images revealed that the size of the lesion in histology closely matched the hyperintense region in T2-weighted imaging (inset). Macrophage infiltration was evident at the edge of the lesion (D). This infiltration had not reached the center core of the lesion, where both gray matter and white matter structures appeared pale-stained and acellular. A sharp boundary was found between the outer edge of the lesion and the normal brain (E). Bar = 1 mm (B and C), 200 μm (D and E). Figure is available in color online only.

Fig. 7

Photograph showing 2 adjacent formalin-fixed blocks of a rat brain obtained 9 weeks after FUS ablation. An empty cavity was observed at the ablation site in the thalamus. A large white matter tract that was included in the sonicated region appeared intact. Bar = 5 mm. Figure is available in color online only.

Fig. 8

MR images and photomicrographs of H & E–LFB–stained sections of the brain shown in Fig. 7 obtained 9 weeks after FUS ablation. A–C: A cyst-like region was observed at the sonication site in both MRI and histological examination along with enlarged ventricles. Superimposed segmentations (inset in B) show some distortion, but similar areas evident in MRI and stained section. D: A higher-magnification view of the lesion edge from an adjacent section shows a sharp boundary with a narrow rim of infiltrating macrophages containing hemosiderin (brown) and mineralization spots (dark blue). E: A white matter tract included in the sonicated region appears damaged but partially intact. Bar = 5 mm (A and B), 1 mm (C), and 200 μm (D–E). Figure is available in color online only.

Fig. 9

Coronal MR image, photograph of formalin-fixed block, and photomicrographs of H & E–LFB–stained sections obtained 6 weeks after microbubble-enhanced FUS ablation. A–C: A cyst-like volume was evident in T2-weighted MRI, in the tissue block, and in H & E–LFB stained sections. The volume included a portion of the corpus callosum, which was severed in this example. D: The boundary of the severed corpus callosum was sharply defined. E: The FUS beam propagated from below, and the lesion extended to the surface of the brain. A thin membrane containing macrophages enclosed the cystic region. Bar = 5 mm (A–C) and 200 μm (D–E). Figure is available in color online only.