Increased CSF osmolarity reversibly induces hydrocephalus in the normal rat brain

Abstract

Background: Hydrocephalus is a central nervous system (CNS) disorder characterized by the abnormal accumulation of cerebrospinal fluid (CSF) in cerebral ventricles, resulting in their dilatation and associated brain tissue injury. The pathogenesis of hydrocephalus remains unclear; however, recent reports suggest the possible involvement of abnormal osmotic gradients. Here we explore the kinetics associated with manipulating CSF osmolarity on ventricle volume (VV) in the normal rat brain.

Methods: CSF was made hyper-osmotic by introducing 10KD dextran into the lateral ventricle, either by acute injection at different concentrations or by chronic infusion at a single concentration. The induction and withdrawal kinetics of dextran infusion on VV were explored in both contexts.

Results: Acute intraventricular injection of dextran caused a rapid increase in VV which completely reversed within 24 hours. These kinetics are seemingly independent of CSF osmolarity across a range spanning an order of magnitude; however, the magnitude of the transient increase in VV was proportional to CSF osmolarity. By contrast, continuous intraventricular infusion of dextran at a relatively low concentration caused a more gradual increase in VV which was very slow to reverse when infusion was suspended after five days.

Conclusion: We conclude that hyperosmolar CSF is sufficient to produce a proportional degree of hydrocephalus in the normal rat brain, and that this phenomenon exhibits hysteresis if CSF hyperosmolarity is persistent. Thus pathologically-induced increases in CSF osmolarity may be similarly associated with certain forms of clinical hydrocephalus. An improved understanding of this phenomenon and its kinetics may facilitate the development of novel therapies for the treatment of clinical hydrocephalus.

Figure 1

Experiment 1: VV was measured 30 minutes and 24 hours following acute intraventricular injection of 10kD dextran solution yielding a range of CSF osmolarities spanning an order of magnitude. Results are expressed as a percentage relative to normalized VVs measured prior to ventricular injection. VVs were significantly increased for all groups at 30 minutes post-injection (P < 0.037), with highly significant increases for Groups III (977 mOsm/L), IV (2000 mOsm/L) and V (3347 mOsm/L) (P < 0.007). These increases in VV were completely resolved by 24 hours post-injection. Significant differences in VV between groups were not observed either pre-injection or 24 hours post-injection. At 30 minutes post-injection, there is a positive correlation (R² = 0.8349) between increased CSF osmolarity and VV. (* P < 0.05, ** P < 0.01).

Figure 2

Experiment 2: VV was measured at 5 day intervals during continuous intraventricular infusion of 10kD dextran yielding CSF osmolarities of 307 (isoosmolar) or 337 mOsm/L (hyperosmolar). Results are expressed as a percentage relative to normalized VVs measured prior to initiating ventricular infusion. VV in Group I (On-On-On 307 mOsm/L) increased incrementally at each time-point over the course of continuous iso-osmotic dextran infusion. VV in Group II (On-On-On 337 mOsm/L) also increased at each time-point over the course of continuous hyper-osmotic dextran infusion; however, these increases were larger and highly significant. VV in Group III (On-Off-On 337 mOsm/L), trended toward increase during interval A, slowed during interval B when infusion was suspended, and resumed increasing and reached significance during interval C when infusion was restored. VV in Group IV (On-Off-Off 337 mOsm/L), trended toward increase during interval A. The rate of increase slowed during intervals B and C when infusion was suspended but the increase in VV reached significance at both time-points.

Figure 3

Coronal T2-weighted MR images obtained from representative animals from Group I (337 mOsm/L), Group II (628 mOsm/L), Group III (977 mOsm/L), Group IV (2000 mOsm/L) and Group V (3347 mOsm/L) in Experiment 1. MR images were acquired pre-infusion (column 1), at the end of 30 mins (column 2) and at 24 hours (column 3) following injection.

Figure 4

Coronal T2-weighted MR images obtained from representative animals from Groups I, II, III and IV in Experiment 2. MR images were acquired pre-infusion (column 1), and at the end of intervals A (column 2, days 1–5), B (column 3, days 6–10) and C (column 4, days 11–15). LV: lateral ventricle.