advertisement
Abstract #28188 Published in IGR 13-1
Modeling glaucomatous neurodegeneration using diffusion tensor imaging
Hayashi T; Shimazawa M; Inoguchi Y; Ito Y; Yamanaka H; Iida H; Watanabe Y; Hara H; Onoe HMolecular Imaging and Biology 2010; 12: 1244
Prediction of neurodegenerative process is requisite when trying to modify the disease progress. A critical point lines in a model by which we predict how neurons will degenerate when exposed to the risk. A classical interpretation has its basis on a deterministic model in which the neurodegeneration, once initiated, proceeds with an elapsed time at a rate faster than in normal-aging. A more elaborated interpretation has its basis on a stochastic model, whereby neurons degenerate as random events at a probability upon the risk, such as aggregation of neurotoxic molecules. By using diffusion tensor imaging (DTI), we here measure in vivo microstructural integrity in animals of glaucomatous neurodegeneration, and show that the kinetics of degrading integrity can be explained by the stochastic model upon the risk, increased intra-ocular pressure. We used five macaque monkeys of glaucomatous neurodegeneration. Glaucoma is known to involve a pathology in which axons of retinal ganglion neurons degenerate due to chronic increase in intra-ocular pressure (IOP) and dysregulated aqueous-fluid circulation. The glaucomatous neurodegeneration was made by surgically inhibiting the aqueous-fluid absorption using laser-photocoagulation of trabecular meshwork and was repeatedly measured the IOP, the risk of glaucomatous pathology. We also performed DTI for in vivo assessment of the microstructural integrity in the white matter. We followed up glaucomatous animals from 33 to 168 days after laser-photocoagulation, during which we measured FA and IOP repeatedly. The FA values in the glaucomatous optic nerve measured by DTI were decreased with elapsed time after laser-photocoagulation. This was the same when the FA values of affected optic nerve were expressed as a ratio relative to those in the contralateral non-affected nerve. The deterministic model explained the variance of relative FA values at a coefficient of determination (R(2)) of 0.36 (P<1.0null 10(-4)). In contrast, the stochastic model, by calculating the cumulative risk (time-integral of the increase in IOP), better explained the variance of relative FA values (R(2)=0.87, P<1.0null 10(-7)) than in the deterministic model. Our findings indicate that the neurodegeneration proceeds probabilistically upon the risk. This indicates that the process of neurodegeneration is virtually independent among neurons. The model may help to predict the neurodegeneration by identification and repeated measurement of the risk, e.g. using a recently-advanced imaging technique for bio-markers that label neurotoxic molecules.
T. Hayashi. RIKEN Center for Molecular Imaging Science, Kobe, Japan. takuya.hayashi@riken.jp
Classification:
3.13.4 Other (Part of: 3 Laboratory methods > 3.13 In vivo imaging)
5.2 Primates (Part of: 5 Experimental glaucoma; animal models)
