The choroid plexus water density

Abstract

Purpose: To quantify normative ranges and circadian variability of the choroid plexus (ChP) water density in healthy adults.

Methods: Actigraphy assessments of circadian activity were performed for 5 days in healthy participants (n = 15; age = 28.5 ± 6.5 years) and subsequently participants underwent repeated, high spatial resolution proton density-weighted imaging (spatial resolution = 0.25 × 0.25 × 1.50 mm) with a driven equilibrium (DRIVE) module at 3 T across four time epochs during wakefulness: 7:00 to 9:00, 11:00 to 13:00, 16:00 to 18:00, and 19:00 to 21:00. ChP water density (unitless ratio of mL water/mL ChP) was calculated as the product of white matter water density and the ratio of the ChP and white matter signal intensity at the level of the atria of the lateral ventricles. Descriptive statistics (mean ± SD; range; median) of water density values at each time were recorded. Spearman and Kendall rank coefficients were used to assess relationships between time, circadian variability, and ChP water density (significance criterion: p < 0.05).

Results: Across all participants and scans (n = 60), mean ChP water density was 0.895 ± 0.047 (range = 0.806-0.983; median = 0.892). Across time periods, water density was 0.891 ± 0.038 (time = 7:44), 0.891 ± 0.050 (time = 12:17), 0.896 ± 0.045 (time = 16:03), and 0.901 ± 0.058 (time = 19:31), and no relationships between ChP water density and time of day or circadian activity were observed.

Conclusions: The ChP water density at the level of the atria of the lateral ventricles is approximately 0.895 ± 0.047 in healthy adults and does not change significantly with time of day during wakefulness. This value should provide a useful reference for the growing number of neuroimaging protocols that aim to derive quantitative contrast and functional metrics from ChP MRI.

Keywords: cerebrospinal fluid; choroid plexus; proton density; water density.

Figure 1.

A sequence diagram providing a visual demonstration of the implementation of the segmented turbo-spin-echo readout with the concluding DRIVE module. A concluding −90° pulse is applied to restore longitudinal magnetization to near equilibrium. The acquired spatial resolution is 0.25x0.25x1.5 mm³, $\text{TE}=9\text{ms}$, refocusing angle=100°, and readout duration per $\text{TR}=153\text{ms}$. An intermediate $\text{TR}=5,000\text{ms}$ is used with the DRIVE module to keep equilibrium choroid plexus and white matter magnetization near equilibrium in a time-efficient manner (see Methods for additional sequence details).

Figure 2.. (A) Reproducibility assessments.

Two representative proton density-weighted MRIs with respective quantitative water density maps from one healthy control in the intrasession repeatability study before and after repositioning. Choroid plexus (red) and white matter (green) regions-of-interest are shown in the proton density-weighted scans. Magnifications at the level of the atria of the lateral ventricles are shown as an insert. Water density is shown in units of mL of water/mL of tissue. (B) Circadian assessments. Four representative proton density-weighted MRIs, from one healthy control, with respective quantitative water density maps at four time points along the circadian cycle of wakefulness. Each scan (1-4) was performed at approximately 7:44, 12:17, 16:03, and 19:31. The color scale is in units of mL of water/mL of tissue. Quantitative values are summarized in Figure 3.

Figure 3.. Repeated choroid plexus water density assessments.

(A) Energy expenditure, recorded by the Fitbit Charge 6 actigraphy device, is shown as a measure of metabolic equivalents (METs) over multiple days prior to imaging. Peaks generally denote periods of wakefulness whereas nadirs largely represent periods of sleep. (B) Circadian energy expenditures for all 15 participants as estimated using a cosinor-based rhythmometry analysis as it relates to energy expenditure, circadian frequency, and intra-individual circadian delay. The four scan times are shown along the 24-hour circadian rhythm. This analysis allows for each participant’s circadian energy expenditure to be estimated and to provide a reference for comparison with imaging metrics. (C) Mean (left) and individual (right) choroid plexus water density values are shown for all participants at each scan time. The zeitgeber synchronizes the participant’s individual circadian rhythm. No significant change was observed across time of day or the circadian cycle of wakefulness, however, relatively large and reproducible inter-subject variability is observed. (D) Representative manual regions used for three participants with low, middle, and high quantified values do not demonstrate clear differences in CSF partial volume effects (red=choroid plexus; green=white matter), suggesting that inter-participant variation may be attributable to other factors.

Figure 4.. Simulations and considerations for experimental measures.

(A) The calculated choroid plexus (ChP) water density depends linearly on the assumed white matter water density in the proposed quantification procedure; this work assumes a mean white matter water density of 0.70 mL/mL (dashed line) based on literature, however, the plot shows how the calculated ChP water density would vary over a reasonable range of assumed white matter water density values from 0.64-0.75 mL/mL. (B) How the calculated ChP perfusion would vary, relative to the actual ChP perfusion using the measured ChP water density of 0.895 mL/mL here, for the range of ChP water density values from 0.80 to 1.00 mL/mL. Note that we assume a blood water density of 0.87 mL/mL from the literature. (C) There is close similarity in calculated and actual ChP perfusion values when assuming an identical ChP and blood water density (thin gray line), as has been assumed in previous ChP perfusion work, of approximately 0.5-3 mL/100g/min over a physiological range of 35-75 ml/100g/min.