CSF dynamics disorders: Association of brain MRI and nuclear medicine cisternogram findings

Abstract

Disproportionately enlarged subarachnoid space hydrocephalus (DESH), characterized by ventriculomegaly, high convexity/midline tight sulci, and enlarged sylvian fissures on brain MRI has been increasingly recognized as a distinct diagnostic imaging entity that falls within the larger category of idiopathic normal pressure hydrocephalus. Normal pressure hydrocephalus has been previously characterized as a CSF dynamics disorder based on abnormalities on nuclear medicine cisternography: radiotracer in the lateral ventricles and absent or delayed ascent of radiotracer over the cerebral convexity. The purpose of this work was to evaluate for differences in nuclear medicine cisternography between patients with vs without DESH and thereby provide support for the concept that DESH is a structural imaging marker of a CSF dynamics disorder. The study included 102 patients (mean age 71 years, range 46-86, 38 females), 58 patients with cisternogram performed to evaluate suspected normal pressure hydrocephalus (mean age 73 years, range 46-86 years, 24 female) and 44 patients evaluated for headache (mean age 68 years, range 60-82 years, 14 female). All patients had an MRI of the brain performed within 13 months of the cisternogram. Cisternogram imaging, typically acquired at 0.5, 1, 2, 4, and 24 h post injection, was evaluated for the time at which radiotracer reached the basal cisterns, presence of persistent radiotracer in the lateral ventricles, time radiotracer first entered the lateral ventricles, presence of radiotracer over the cerebral convexity, and time at which radiotracer was first visualized over the cerebral convexity. MRI features of ventriculomegaly (defined as Evans' index ≥ 0.3) and high convexity tight sulci (HCTS) were recorded. Based on the MRI features, patients were grouped according to presence or absence of DESH (ventriculomegaly and HCTS). Those without DESH were separated into groups of ventriculomegaly alone, HCTS alone, and neither ventriculomegaly nor HCTS. Cisternogram metrics were compared between MR-defined groups. Patients with DESH showed a higher frequency of radiotracer in the lateral ventricles and delayed or absent ascent over the cerebral convexity compared to those without DESH, higher frequency of ventricular radioactivity vs those with HCTS alone, and shorter time to ventricular radioactivity compared to those with ventriculomegaly alone. Patients with ventriculomegaly or HCTS alone had a higher frequency of radiotracer in the lateral ventricles and delayed ascent of radiotracer over the cerebral convexity compared to those with neither ventriculomegaly nor HCTS. These findings support DESH and the individual components of ventriculomegaly and HCTS as markers of disordered CSF dynamics.

Keywords: CSF dynamics disorders; Disproportionately enlarged subarachnoid space hydrocephalus (DESH); Normal pressure hydrocephalus (NPH); Nuclear medicine cisternography.

Fig. 1

Representative cases. Images from four patients, shown from left to right: axial T2 FLAIR at the level of the lateral and third ventricles, coronal T1 post-contrast or T2 FLAIR at the level of the posterior commissure, sagittal T1-weighted images off midline, and anterior projection cisternogram views of the head and upper spinal canal at 2, 4, and 24 h. (A.) 65-year-old man with DESH. Radiotracer was at the basal cisterns at 2 h, lateral ventricles at 4 h (arrow, A) and did not ascend over the convexity at 24 h. (B.) 78-year-old man with ventriculomegaly but no HCTS. Radiotracer was at the basal cisterns at 2 h (arrow, B), lateral ventricles at 4 h, and ascended over convexity at 24 h. (C.) 70-year-old man with HCTS alone. Radiotracer was at the basal cisterns at 2 and 4 h, transiently entered the lateral ventricles at 6 h (not shown), and ascended over the convexity at 24 h (arrow, C). (D.) 63-year-old man with neither ventriculomegaly nor HCTS. Radiotracer was in the basal cisterns at 2 h, unchanged at 4 h, and ascended over convexity at 24 h; radiotracer did not enter the lateral ventricles.

Fig. 2

Patient categorization based on MRI criteria.

Fig. 3

Dot plots of time for radiotracer to reach the basal cisterns for (A) all patients and (B) patients evaluated for suspected NPH. There was no significant difference in time to the basal cisterns among patients with DESH, ventriculomegaly (ventric), high convexity tight sulci (HCTS), and neither; median time to the basal cisterns was 2 h for all groups.

Fig. 4

Dot plots of time for radiotracer to reach the lateral ventricles in patients with abnormal persistent radiotracer in the lateral ventricles for (A) all patients and (B) those with suspected NPH. In (A) one patient with ventriculomegaly alone with time to ventricular radioactivity of 24 h is not displayed. Median values are denoted by the black lines. Patients with ventriculomegaly had a longer median time for radiotracer to reach the ventricles compared to other groups, though comparisons did not reach statistical significance.

Fig. 5

Distribution of time for radiotracer to ascend over the cerebral convexity in (A) all patients and (B) patients with suspected NPH. A time of 24 h is considered normal, 48 h delayed, and radiotracer not visualized over the convexity at 24 or 48 h (none) abnormal. Ventric = ventriculomegaly. HCTS = high convexity tight sulci. * p < 0.05.