CSF production rate, resistance to reabsorption, and intracranial pressure: a systematic review and meta-analysis

Abstract

Davson's equation relates the state of stable intracranial pressure (ICP) to the production rate of CSF (I_F) and resistance to CSF outflow (R_OUT). Both parameters are assumed to be independent of ICP, but results are conflicting. The objective is to define the relationship between ICP, I_F and R_OUT using a systematic literature review. Medline and Embase were searched from inception up to 12 February 2024. Experimental studies exploring the association between ICP, I_F, and R_OUT were included. Individual measurements of ICP, I_F and/or R_OUT were extracted from tables or graphs, alongside descriptive parameters (population, ICP measurement site, disease, and computational method). Linear regression and mixed effects models were applied. From 1304 references, 25 articles were included in our meta-analysis. I_F is approximately constant across all pathologies independent of the ICP level, population, disease, ICP measurement site and the measurement/estimation method. Conversely, ICP was positively correlated with R_OUT. The intercorrelation, however, differed by population, disease, ICP measurement site and estimation method. Additionally, I_F derived from Davson's Equation compared with the measured I_F were similar for patients with hydrocephalus but differed for patients with acute brain injury. Davson's Equation describes the various components of cerebrospinal fluid dynamics. The results underline important caveats for its use in patients with acute brain injury wherein the estimated values differ from the measured ones. Overall, additional metrics describing the cerebrovascular system or the underlying disease have to be taken into account for more accurate estimations.

Keywords: cerebrospinal fluid; intracranial hypertension; intracranial pressure; meta-analysis; systematic review.

graphical abstract

Figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram. *Language of excluded studies: Japanese (25), German (13), Russian (12), Spanish (5), Polish (4), Italian (3), Chinese (1), Hungarian (1), Korean (1) and Turkish (1).

Figure 2

Forest plot of the association of intracranial pressure (ICP) with CSF production rate (I_F) and resistance to outflow (R_OUT). (A) I_F (N = 8), and (B) R_OUT (N = 18). Data name: studies; data type: continuous; effect measure: correlation; analysis model: mixed effects; statistical method: χ² test. CI, confidence interval; I², heterogeneity measure; χ², value of χ² test for heterogeneity and P, statistical significance of the χ² test.

Figure 3

Relationship between ICP, I_F and R_OUT, over all extracted measurements, including the line of best fit and the 95% confidence interval. (A) ICP versus I_F (N = 231, R² = 0.001, F = 0.2, P = 0.677). (B) ICP and R_OUT (N = 676, R² = 0.405, F = 459, P < 0.001). Data name: single extracted measurements; data type: continuous; analysis model: linear regression; statistical method: ANOVA; F: F-value from the ANOVA and P: statistical significance of the ANOVA test.

Figure 4

Relationship between ICP, I_F and R_OUT for different sub-groups, including the line of best fit and the 95% confidence interval. ICP versus I_F (A) per population (adult: N = 190, paediatric: N = 38; R² = 0.015; I_F: F = 0.2, P = 0.681; population: F = 2, P = 0.159; interaction: F = 1, P = 0.273), (B) per disease (ABI: N = 129, brain tumour: N = 41, primary hydrocephalus: N = 9; R² = 0.025; I_F: F = 0.0, P = 0.826; disease: F = 2, P = 0.130; interaction: F = 0.1, P = 0.919) and (C) ICP measurement site (cranial: N = 191, lumbar: N = 40; R² = 0.015; I_F: F = 0.2, P = 0.676; measurement site: F = 2, P = 0.156; interaction: F = 1, P = 0.259). ICP versus R_OUT (D) per population (adult: N = 598, paediatric: N = 78; R² = 0.423; R_OUT: F = 471, P < 0.001; population: F = 20, P < 0.001; interaction: F = 1, P = 0.288), (E) per disease (ABI: N = 151, neurodegenerative: N = 66, primary hydrocephalus: N = 124; R² = 0.459; R_OUT: F = 238, P < 0.001; disease: F = 23, P < 0.001; interaction: F = 0.1, P = 0.890), and (F) per ICP measurement site (cranial: N = 628, lumbar: N = 48, R² = 0.458; R_OUT: F = 502, P < 0.001; measurement site: F = 65, P < 0.001; interaction: F = 2, P = 0.211), including the line of best fit and the 95% confidence interval. Data name: single extracted measurements; data type: continuous; analysis model: linear regression; statistical method: ANOVA; F: F-value from the ANOVA and P: statistical significance of the ANOVA test.

Figure 5

Relationship between R_OUT—dependent variable and ICP—independent variable, including the line of best fit and the 95% confidence interval. (A) Per measurement method (bolus: N = 261, infusion: N = 318, withdrawal: N = 59; R² = 0.614; ICP: F = 742, P < 0.001; computation method: F = 114, P < 0.001; interaction: F = 18, P < 0.001) and (B) per disease (ABI: N = 151, neurodegenerative: N = 66, primary hydrocephalus: N = 124; R² = 0.485; ICP: F = 250, P < 0.001; disease: F = 20, P < 0.001; interaction: F = 13, P < 0.001). Data name: single extracted measurements; data type: continuous; analysis model: linear regression; statistical method: ANOVA; F: F-value from the ANOVA and P: statistical significance of the ANOVA test.