Reply to: Rethink the classical view of cerebrospinal fluid production

We thank Roques and colleagues for their comments on our Review (Wardlaw, J. M. et al. Perivascular spaces in the brain: anatomy, physiology, and pathology. Nat. Rev. Neurol. 16, 137–153 (2020)¹), in which they correctly point out some textbook dogma that needs to be re-evaluated (Roques, M., De Barros, A. & Bonneville, F. Rethink the classical view of cerebrospinal fluid production. Nat. Rev. Neurol. https://doi.org/10.1038/s41582-021-00538-0 (2021)²). We agree that the questions of how much cerebrospinal fluid (CSF) is produced and from which structures remain unanswered. In our Review, we noted that there are several potential sources of CSF, including the choroid plexus, but information about the relative contributions of each source is unfortunately limited.

In our Review¹, we focused on perivascular spaces as seen in vivo on human brain imaging, how these relate to perivascular spaces as described histologically, and their role in brain fluid management. Perivascular spaces themselves have been a controversial topic since the 1800s, and clarification of their role, along with a full understanding of brain fluid management, would be of great value.

Some evidence supports the existence of several sources of interstitial fluid and CSF and several sites of their drainage from the cranial cavity, including a return to the venous circulation via arachnoid granulations and intradural channels³, and drainage via meningeal lymphatics^1,4, including via skull base lymphatics to nasal lymphatics⁵. However, the major remaining questions concern how much fluid is produced and removed by each route, and whether the function of these pathways differs by age, sex, brain and physical activity (including sleep and wakefulness) and disease states.

The choroid plexus is generally agreed to be a key site of CSF production⁴, although the relative contributions of choroid plexus in the lateral ventricle versus that in the fourth ventricle are uncertain and may differ between species. There are also longstanding arguments that interstitial fluid contributes to the CSF and ultimately to lymphatic drainage^4,6,7. Water is a byproduct of metabolic reactions in the brain, so excess needs to be removed. Filtration at the level of capillaries via the blood–brain barrier also allows fluid to enter the brain interstitial space, a process that might subtly increase in many disease states, including Alzheimer disease and small vessel disease. Fluid can enter the interstitial space directly from the ventricles, as in ventricular obstruction, although this route is unlikely to be a major contributor to physiological interstitial fluid. Fluid also enters the brain from CSF via the perivascular spaces as part of the glymphatic flushing system, as explained in our Review¹. Therefore, it is possible that some fluid has passed through several intracranial compartments before exiting via one of the several exit routes (the arachnoid granulations to venous sinuses, the meningeal lymphatics along the dural sinuses, or the nasal lymphatics⁵), which compounds the difficulties of measuring the source of fluid.

Measurement of intracranial fluid production is difficult. In humans, opening of the cranial cavity, even with small burr holes, disturbs the usual pressure balance between intravascular and extravascular compartments⁸, so might also perturb blood and fluid transit through the various compartments. Cranial windows in rodents affect brain temperature, thus altering blood flow and cellular activity⁹, so might also affect fluid production and movement. The temperature effects persist after re-closure of the cranial cavity with a glass window that is sealed with dental cement or by substituting the window by thinning the cranium⁹, and these processes could also affect pressures, as the properties of these materials differ from those of the intact skull. Improvements in non-invasive medical imaging techniques, such as recent non-invasive approaches to the measurement of fluid crossing the blood–brain barrier and blood–CSF barrier¹⁰, the use of multi-pronged approaches, and awareness of the limitations of each approach, will clarify the importance of each intracranial fluid production and drainage route in the near future.