**1. Introduction**

Sporadic cerebral amyloid angiopathy (CAA) is characterized by the progressive deposition of amyloid-β (Aβ) protein in the walls of small- to medium-sized arteries, arterioles, and capillaries in the cerebral cortex and overlying leptomeninges [1,2]. CAA is a common age-related cerebral small vessel disease (SVD) in the elderly [2,3] (especially those with Alzheimer's disease [4]), but it is most often recognized clinically by symptomatic intracerebral hemorrhage (ICH) restricted to the lobar areas of the brain [5,6]. Long-term antiplatelet therapy, which is widely used for the secondary prevention of cerebral infarction, myocardial infarction, and peripheral artery diseases, could increase the incidence of ICH. In consideration of the high incidence of cerebral bleeding in patients with

CAA and/or Alzheimer's disease, especially in antiplatelet drug users [7,8], the safety of antiplatelet therapies for those patients should be explored.

Cyclic nucleotide phosphodiesterases (PDEs) play critical roles in regulating intracellular cyclic nucleotides (cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate), which are important secondary messengers involved in intracellular signal transduction in all tissues. PDE III is the major cAMP-hydrolyzing PDE (a negative regulator of cAMP) uniquely expressed in vascular smooth muscle cells. A selective inhibitor of PDE III, cilostazol, has multiple effects on the vasculature including vasodilatation, antioxidation, anti-inflammation, the regulation of smooth muscle cell, and an increase in cerebral hemodynamics, pulse duration time, and arterial elasticity with the maintenance of microvascular integrity [9]. Cilostazol is known as a unique antiplatelet drug, which is superior to aspirin in terms of safety for reducing ICH after an ischemic stroke [10]. Cilostazol ameliorates collagenase-induced cerebral hemorrhage by protecting the blood–brain barrier in mice [11]. In a CAA mice model (Tg-SwDI mice), cilostazol improved cognitive performance, which may be associated with reduced Aβ deposition by cilostazol (0.3% cilostazol) [12]. However, it is unknown whether the safety and efficacy of cilostazol could be replicated in different settings (i.e., drug dose or experimental transgenic mice). Thus, we sought to determine whether lower-dose cilostazol could reduce the incidence of cerebral (micro-)hemorrhages or cerebrovascular Aβ depositions using different transgenic mice as a CAA model.

#### **2. Results**

#### *2.1. Survival Rate, Feed Consumption, and Drug Intake*

Sixty Tg2576 mice aged 3 months initially received the vehicle, but 10 mice died due to unknown reasons before 14 months. Thus, 50 mice aged 14 months were divided into three groups: control group (*n* = 14), aspirin group (*n* = 18), and cilostazol group (*n* = 18) (Figure 1). Within a month after the grouping, two mice died due to unknown reasons (control group, *n* = 1; cilostazol group, *n* = 1). Of the 48 survivors aged 15 months, 14 mice (control group, *n* = 4; aspirin group, *n* = 5; cilostazol group, *n* = 5) were randomly selected for the first evaluation (i.e., the first specimen). Among the remaining mice (*n*= 34), four mice died due to unknown reasons (control group, *n* = 1; cilostazol group, *n* = 3). Finally, 30 surviving Tg2576 mice aged 21–23 months were evaluated as the second specimens. There was no significant difference in the survival rate after the grouping among the three groups when the 14 mice evaluated for the first specimen were excluded from the analyses (*p* = 0.109, log-rank test: Figure A1: please see appendix). Even in the two group comparisons, no differences were seen in the survival rate between the control group and the aspirin or cilostazol group (*p* = 0.098 and *p* = 0.550, log-rank test). Table 1 shows estimated individual food consumption and drug intake (per mouse) of the three groups. There was no significant difference in food consumption between the control group (mean, standard deviation [SD]: 3.57 ± 0.40 g/day) and the aspirin group (3.67 ± 0.47 g/day) or the cilostazol group (3.67 ± 0.44 g/day) (*p* = 0.289 and *p* = 0.543, Mann–Whitney *U* test). Individual daily drug intake in the aspirin group (14.7 ± 1.9 mg/kg/day) was similar to that in the cilostazol group (14.7 ± 1.8 mg/kg/day).

**Figure 1.** The flow diagram of study schedule and grouping.



<sup>a</sup> *p* = 0.289 vs. control; <sup>b</sup> *p* = 0.324 vs. control (Mann–Whitney *U* test); NA = not applicable; SD = standard deviation.

#### *2.2. Confirmation of Age-Related Cerebrovascular Amyloid Burden and Smooth Muscle Cell Loss*

Confocal microscopic observation of double-immunolabeled vessels in different Tg2576 mice (15 months old, and 23 months old) fed with standard pelleted chow (i.e., control group) confirmed the age-related progression of amyloid burden and loss of smooth muscle cells (Figure 2a–d). The findings of the negative controls for Aβ show only a faint background stain in the vessel walls (Figure 2e–h).

**Figure 2.** Upper line (**a**–**d**): Serial confocal microscopic changes of double-immunolabeled vessels (green, smooth muscle actin; red, amyloid). Lower line (**e**–**h**): Confocal microscopic findings represent negative controls without primary antibody for Aβ. a, b, Leptomeningeal vessel in a 15-month-old mouse shows small amyloid deposition and focal loss of smooth muscle cells at the site of cerebrovascular amyloid; c, d, In a 23-month-old mouse, smooth muscle cells are lost, and a thick sheet of amyloid covers the wall of a leptomeningeal vessel. e–f, Photos of the negative controls for Aβ (each section adjacent to a, b, c, d, respectively) show only a faint background stain in the vessel walls. Scale bar: 50 μm.
