Beyond Eyeballing the Eyeballs in Intracranial Hypertension

Deformation of the posterior sclera is a well-established imaging sign of intracranial hypertension (IH). In an excellent review article, Passi et al describe the appearance of the optic nerve (ON) and ocular regions on MRI in the presence of papilledema associated with increased intracranial pressure (ICP).¹ They also remind us of the considerable variability in the appearance of the ON head swelling and that “papilledema may be asymmetric or unilateral, and the degree of ON head swelling is poorly correlated with ICP,” indicating the need for a better understanding of the link between ICP and ON swelling. Our comprehensive investigations of a homogeneous cohort of women with obesity-related idiopathic intracranial hypertension (IIH), pre- and posttreatment, and a matching cohort of healthy women with obesity enabled us to gain insight into the link between ICP, intraocular pressure (IOP), and the ocular changes seen in IIH.²

High-resolution volumetric MR imaging of the optic nerve and ocular regions reliably captures the globe morphology and the changes that occur in these regions in IIH before and during the treatment. Here again, eyeballing alone, even by experienced neuroradiologists, does not have the inherent consistency and sensitivity required for reliable characterization of the two components contributing to the distortion of the posterior sclera, 1) flattening and 2) ON protrusion. As part of our effort to measure eye globe distortions in IIH, we developed a new means to view the 3D curvatures of the globe by converting its 3D geometry to a 2D distance map, ie, a color-coded map of distances from the center of the globe to points on the posterior half of the globe wall, where a uniform color represents a perfect hemisphere. A change in color represents a deviation from the perfect sphere. Second, two unbiased measures were derived from this map, a measure of globe flattening (GF) and a measure of nerve protrusion (NP). The GF is defined as the ratio of the mean distance to a region surrounding the head of the optic nerve relative to a peripheral annulus region defined by large azimuths and elevations angles. A ratio of 1 represents a perfect hemisphere with no flattening. As the posterior wall flattens, distances toward the flattened region decrease while distances to the peripheral region stay the same or increase; thereby the value of the GF measures become smaller than 1. The measure of NP is defined by the ratio of the mean distance to the region of the optic nerve head and the surrounding region. A value of 1 means no protrusion. As nerve protrusion increases, the distance from the globe center to the head

of the optic nerve decreases, and thereby the value of the NP measure decreases with increasing protrusion.

As expected, the study found significant differences between the IIH and the healthy cohort in both the NP and the GF measures. Interestingly, treatment affected mainly the NP aspect of the globe distortion. The study results also elucidated the reasons for the weak correlation between papilledema and ICP. In our cohort, papilledema grade was significantly associated with NP but not with GF. Furthermore, the GF was significantly associated with IOP, ie, as the CSF pressure force flattens the globe, the IOP inside increases. Because the degree of nerve protrusion is influenced by both by the CSF pressure and the IOP, papilledema severity may not necessarily be strongly associated with the value of ICP alone.

This work is another example of the need to employ computational analysis for quantitation of complex imaging signs for advancement of the diagnosis power and the understanding of visual impairment in IIH and in other causes of optic disc edema, such as ischemic, inflammatory, and compressive optic neuropathies, as well as in microgravity-induced visual impairments in astronauts.

References

1. Passi N, Degnan AJ, Levy LM. MR imaging of papilledema and visual pathways: effects of increased intracranial pressure and pathophysiologic mechanisms. AJNR Am J Neuroradiol 2013;34:919–24, 103174/ajnr.A3022
2. Alperin N, Bagci AM, Lam BL, et al. Automated quantitation of the posterior scleral flattening. AJNR Am J Neuroradiol 2013;34:2354–59, 10.3174/ajnr.A3600

 

Read this article at AJNR.org . . .