Modulation of CSF flow and tracer brain influx in pigs in vivo using SPECT

Background

Aiming to bridge the translational gap between rodent models and humans, we established an experimental and analytical protocol to study cerebrospinal fluid (CSF) flow and brain influx in pigs, which offer strong anatomical and physiological resemblance to humans while permitting invasive approaches. CSF dynamics are strongly influenced by vascular vasomotion, respiration, and cardiac pulsatility, which together act as key drivers of perivascular and bulk flow within the glymphatic system (1). Accordingly, we systematically modulated these physiological states to investigate how controlled changes in cardiopulmonary and vascular activity alter tracer distribution and brain penetration over time. This enables the study of CSF dynamics implicated in neurological disorders such as stroke, Alzheimer’s disease, traumatic brain injury, and small vessel disease, supporting translational research and the development of clinically relevant interventions. Finally, SPECT-derived tracer transport will be contrasted with MRI measurements of brain pulsations, allowing direct comparison between mechanical driving forces and resulting solute movement, to better elucidate the coupling between vascular dynamics and CSF-mediated clearance.

Methods

Using single-photon emission computed tomography (SPECT), we measured the in vivo distribution of an intracisternally injected radiotracer (99mTc-DTPA) in anesthetized pigs. To separate CSF spaces from brain parenchyma, we applied non-negative matrix factorization–based voxel decomposition, enabling kinetic modeling of tracer penetration using a one-tissue compartment model (K1 representing tracer brain influx and k2 brain efflux). Physiological interventions were performed in a within-subject design to modulate CSF flow, in a series of interventions separated by 1h of rest to achieve baseline conditions before next intervention. We performed hypercapnia induced by 20–40 min of CO₂ inhalation and intravenous hypertonic saline (3% NaCl, 20 mL/kg). The order of interventions was altered, with hypertonic saline always applied last due to its longer-lasting osmotic and vascular effects. This design allowed controlled assessment of reversible physiological perturbations while maintaining internal consistency across imaging sessions. We also monitored physiological vitals and electroencephalogram (EEG) during the duration of the scanning.

Results

Hypercapnia, which causes cerebral vasodilation, reduced tracer brain influx (mean ΔK₁ from baseline = −3.6×10⁻⁵ mL/min/mL, p = 0.002, n = 5), corresponding to ~2.5-fold reduction from baseline, followed by partial recovery (mean ΔK₁ from baseline = 1.4×10⁻⁵ mL/min/mL, p = 0.017, n = 4). Intravenous hypertonic saline, which alters osmotic pressure, transiently decreased influx during the first 20 minutes (mean ΔK₁ from baseline = 3.8×10⁻⁵ mL/min/mL, p = 0.0005, n = 8), returning toward baseline by 60 min (mean ΔK₁ from baseline = −1.0×10⁻⁵ mL/min/mL, p = 0.2, n = 6).

Conclusions

We present a translational large-animal platform combining in vivo SPECT imaging and kinetic modeling to quantify CSF–brain exchange dynamics. This approach enables measurement of glymphatic-related transport and may facilitate the development and validation of therapeutic strategies targeting brain clearance mechanisms. The framework also supports longitudinal assessment of dynamic tracer kinetics across multiple compartments, improving mechanistic resolution of fluid exchange pathways. By enabling non-invasive quantification in a clinically relevant model, it strengthens translational potential and provides a basis for comparing disease states and treatment effects over time.

References

(1) Bohr T, Hjorth P, Holst S … The glymphatic system: Current understanding and modeling iScience, 2022; 25