Imaging of the living brain using Positron Emission Tomography (PET), a noninvasive, sensitive and quantitative imaging methodology, allows us to investigate neurobiological mechanisms involved in the onset of the neurological disease. Our work has focused on investigating the pre-symptomatic neuroinflammatory processes (called epileptogenesis) that lead development of chronic seizures, in an animal model of epilepsy.
We have used our in-house radiotracer (18F-PBR111)which is highly specific fo receptors expressed in the inflammatory response within the brain. Performing pre-clinical PET imaging with this radioligand allowed us to map and quantify neuroinflammation in vivo, and to correlate this with full in vitro assessment of the neuroinflamation response.
The in vivo ligand binding patterns highly correlate with the structures involved in the generation of seizures,and these data reflect the in vitro data, illustrating that the PET binding represents true neuroinflammation. Thus, longitudinal PET studies will be possible in order to follow-up the evolution of the inflammatory regions during the onset of the disease, and test new preventive therapies that modulate this disease process.
Why PET in epilepsy research?
A fundamental goal of epilepsy research is to prevent the development of chronic seizures through the treatment of the initiating causative disease process. The use of nuclear medicine imaging techniques such as Positron Emission Tomography (PET) may allow detection of epileptogenesis in patients, after brain injuries such as stoke or head trauma, to allow preventative pharmacotherapies to stop or reduce the severity of epilepsy progression at the earliest stage of disease.
As shown in Figure 1, the onset of the disease occurs long before seizure activity is seen clinically or in animal models. In the human, this process, known as epileptogenesis, often occuring years before symptoms presenting [1].
We have used preclinical PET imaging with our in-house radiotracer, 18F-PBR111, to map the response of the brain’s immune system (neuroinflammation) to an insult that potentially leads to epilepsy (epileptogenesis, see Figure1). This radiotracer has shown high affinity for the translocator protein in a previous study on acute brain lesion [2].
In order to check whether the PET images accurately reflect the neuroinflammation, we first correlated these results with post-mortem analysis. If the PET quantification was accurate, repeated PET imaging would allows us to follow up the evolution of the brain inflammation in the same animal during the full epileptogenesis, until the occurrence of the seizures, and maybe lead to a very early diagnostic tool.
It will be used to assess the brain response to new preventive therapies in animal models first, then in clinical environment.
Our findings
PET imaging with 18F-PBR111 was performed seven days after intra-peritoneal administration of kainic acid (treated rats which will become epileptic within 1 month, see Figure 1) or saline (control rats).
The PET data of the treated group showed significant increases in radiotracer binding compared to the control group in several brain regions that are intrinsically involved in epileptogenesis: the amygdala, hippocampus and thalamus showed increases in receptor density (see Figure 2).
In vitro binding on the same brains showed significant increases in translocator-protein density in the same brain regions including hippocampus, piriform cortex and basolateral/medial amygdala, in each case strongly associated with OX42 (indicator of activated microglial cells) signal. Figure 2 also shows the spatial correlation between the in vivo index (Volume of distribution), the in vitro translocator-protein density and the microglial activation, thus neuroinflammation.
Our tools: from in vivo animal to cellular imaging
Our radiotracer has already shown a high affinity for the translocator protein, which is a receptor that immune cells of the brain - activated microglia - significantly over-express in response to brain injury [1-3]. We have utilised an extensively characterised injury is induced using the excitotoxin, kainic acid [3]. Epilepsy symptoms are not seen in this model until a period of one or more months after this treatment.
A complex imaging protocol was required that allowed the assessment of density of translocator protein and radioligand affinity in the brain structures of interest. Once anaesthetised, the rats were injected intravenously with three different masses of 18F-PBR111 while in the PET camera, and imaged for three-hours.
Arterial blood samples were collected during the scan for quantification of radiotracer pharmacokinetics. For each rat, a CT scan was performed just before the PET scan, providing us with anatomical information. After the scan, each rat was sacrificed and its brain was kept for postmortem confirmation of neuroinflammation.
Co-registration of a neuroanatomical atlas to the CT image of the animal allowed us to outline 20 different brain structures of interest. By using a non-linear compartmental model, translocator -protein density could be calculated in all the brain structures of interest.
A more global index called Volume of Distribution (VD), which reflects the binding of the radiotracer to the receptor, has been derived for each pixel and parametric maps could be calculated (Figure 2). In vitro analyses of brain sections from the animals that had been imaged were performed to determine relative translocator protein density changes (autoradiography) and assess microglial activation (immunohistochemistry) between experimental conditions.
Authors
Paul Callaghan, Stefanie Dedeurwaerdere, Marie Gregoire, Tien Pham and Andrew Katsifis
ANSTO
References
- Sauvageau A., Desjardins P., Lozeva V., Rose C., Hazell A.S., Bouthillier A., Butterworth R.F. Metabolic Brain Disease, 17 (2002) 3.
- Van Camp N, Boisgard R, Kuhnast B, Thézé B, Viel T, Grégoire MC, Chauveau F, Boutin H, Katsifis A, Dollé F, Tavitian B. European Journal of Nuclear Medicine Molecular Imaging. 2010, Jan 13.