The drainage of cerebrospinal fluid: new insights into lymphatic clearance pathways


Cerebrospinal fluid (CSF) is produced by choroid plexuses within the ventricles of the brain and circulates throughout the subarachnoid space surrounding the brain and spinal cord. From the subarachnoid space, CSF has long been considered to clear through arachnoid villi directly to venous sinuses within the dura mater meningeal lining of the central nervous system. However, evidence for lymphatic vessels draining at least some component of this fluid has been presented in studies dating back 150 years. The pathways through which the CSF can reach lymphatic vessels are currently an area of great controversy. The bulk of the literature supports pathways through the cribriform plate and along cranial nerves to reach lymphatic vessels outside the skull. However, in the last few years researchers have rediscovered a network of lymphatic vessels in the dura mater and have suggested that this could represent a more direct outflow route for CSF.


Previous studies have relied on cannulation of lymphatic and/or blood vessels to study the outflow of CSF injected tracers. We have developed fluorescence imaging methods that can non-invasively assess the dynamics of near-infrared tracer outflow to lymphatic vessels or to the systemic blood circulation after injections into the lateral ventricle or cisterna magna of mice.


Surprisingly, we found that tracers were cleared predominantly by lymphatic vessels with no evidence suggesting a direct blood vascular uptake within the ventricles or subarachnoid space. Tracer outflow from the skull was detected at the exits of the cranial nerves, including at the cribriform plate, optic canal and jugular foramina, to drain to deep cervical and mandibular lymph nodes. CSF tracer distribution through two anatomically distinct routes down the spine to reach spinal outflow pathways and the spread of the CSF tracers through paravascular spaces on the brain surface through the intact skull could also visualized in living mice. Confirmation of these CSF pathways was demonstrated with 9.4 T MRI.

Discussion and Conclusions

Together, these in vivo imaging approaches allow a system-wide assessment of CSF flow pathways and has enabled a new model of CNS fluid flow to be developed. Further studies are needed to test the model in pathological conditions and to confirm these findings in humans.