Lumbosacral and thoracolumbosacral cerebrospinal fluid volume changes in neonates, infants, children, and adolescents: A retrospective magnetic resonance imaging study.

Abstract

The volume of cerebrospinal fluid can affect the pharmacokinetics and pharmacodynamics of local anesthetics for spinal anesthesia and other intrathecal medications. The objective of this study was to estimate the lumbosacral cerebrospinal fluid volume and thoracolumbosacral cerebrospinal fluid volume using magnetic resonance images in pediatric patients from neonates and infants to adolescents. Spinal magnetic resonance images of 500 pediatric patients (age <18years) were reviewed. The lumbosacral cerebrospinal fluid volumes of 418 patients and thoracolumbosacral cerebrospinal fluid volumes of 248 patients were measured. The relationship between cerebrospinal fluid volumes and age, height, and weight were evaluated. The lumbosacral and thoracolumbosacral cerebrospinal fluid volumes per weight were calculated to elucidate developmental changes. The lumbosacral and thoracolumbosacral cerebrospinal fluid volumes showed linear correlations with height (r2 =0.730 and r2 =0.661, respectively), whereas they showed curvilinear correlations with age (r2 =0.752 and r2 =0.717, respectively) and weight (r2 =0.734 and r2 =0.734, respectively). The mean lumbosacral cerebrospinal fluid volume per weight (mL/kg) was 0.85 (standard deviation [SD]: 0.19, 95% confidence interval [CI]: 0.81-0.90) in neonates and infants, 0.86 (SD: 0.22, 95% CI: 0.83-0.89) in toddlers and preschoolers, 0.71 (SD: 0.26, 95% CI: 0.66-0.76) in schoolers, and 0.54 (SD: 0.20, 95% CI: 0.49-0.60) in adolescents. The mean thoracolumbosacral cerebrospinal fluid volume per weight (mL/kg) was 1.95 (SD: 0.37, 95% CI: 1.86-2.04) in neonates and infants, 1.82 (SD: 0.41, 95% CI: 1.75-1.88) in toddlers and preschoolers, 1.38 (SD: 0.40, 95% CI: 1.23-1.52) in schoolers, and 0.99 (SD: 0.34, 95% CI: 0.45-1.53) in adolescents. The lumbosacral and thoracolumbosacral cerebrospinal fluid volumes in pediatric patients were much smaller than previously presented values, showing linear correlations with height, and demonstrate curvilinear correlations with age and weight.