Background:

White matter (WM) is preferentially damaged in cerebral small vessel disease (cSVD), likely due to its susceptibility to chronic hypoperfusion (Wardlaw et al., 2013). Cilostazol (CIL) and isosorbide mononitrate (ISMN) have been shown in clinical LACI trials to improve memory and reduce stroke risk (Wardlaw et al., 2023); however, the acute cerebrovascular mechanisms and the vascular compartments at which these drugs act remain unclear. We investigated whether CIL and ISMN induce age- and region-dependent acute changes in microvascular haemodynamics.

Methods:

Wild-type C57BL/6J mice of both sexes underwent cranial window implantation at approximately 12 weeks of age over WM or cortical grey matter (GM). For WM windows, ~800 µm of cortex was aspirated over the corpus callosum and a custom 3D-printed window with a glass base was implanted. GM windows were placed over the visual cortex without aspiration, preserving pial arteries. Mice were first recorded at 4 months of age; the same WM mice were re-recorded after 8 months of ageing (1 year old).

Using a blinded crossover design, mice received vehicle or combined CIL+ISMN with ≥6-day washout. Haemodynamics were recorded at baseline, 30 min, 1 h, 2 h, 4 h, 6 h, and 24 h using combined laser Doppler flowmetry and haemoglobin spectroscopy. Measures included red blood cell (RBC) flux, RBC speed, concentration of moving RBCs, oxygen saturation (sO₂), total haemoglobin (HbT), and derived cerebral metabolic rate of oxygen (CMRO₂). Awake, head-fixed recordings were obtained during resting periods. Linear mixed-effects models were used with drug and timepoint as fixed effects and animal ID as a random effect, with Dunnett post-hoc tests. Drug doses were scaled from the LACI-2 trial using body-surface-area conversion.

Results: 

In deep tissue containing WM of young mice (n = 8), CIL+ISMN significantly reduced RBC flux (treatment effect p = 0.022; timepoint × treatment p = 0.032), driven by a marked reduction at 6 h (estimate −0.22 ± 0.06, p = 0.0003). RBC concentration was also reduced (p = 0.0072), while RBC speed was unchanged. CMRO₂ decreased significantly (p = 0.016), most prominently at 6 h (p = 0.0007), without significant changes in sO₂. 

In the same mice re-examined longitudinally after 8 months of ageing (n = 4), WM responses to CIL+ISMN were preserved, with significant reductions in RBC flux (p = 0.013), RBC concentration (p = 0.041), and CMRO₂ (p = 0.007). 

In contrast, GM showed a significant increase in moving RBC concentration (p = 0.003), a modest reduction in HbT (p = 0.044), and a trend toward increased sO₂ (p = 0.082), with no change in flux or speed. Blood pressure under sedation did not differ between baseline, vehicle, or drug conditions. 

Conclusions: 

Our data indicate region-specific microvascular responses to cilostazol and isosorbide mononitrate, with preserved cortical grey matter microcirculation and reduced red blood cell delivery and oxygen consumption in deep tissue containing white matter. These findings support an acute redistribution of microvascular resources rather than a uniform microvascular response. Ongoing two-photon imaging and chronic hypoperfusion models will determine the vascular origin and pathological relevance of this redistribution under chronic treatment conditions.