Levosimendan alone and in combination with valsartan prevents stroke in Dahl salt-sensitive rats

European Journal of Pharmacology 750 (2015) 132–140

Contents lists available at ScienceDirect

European Journal of Pharmacology journal homepage: www.elsevier.com/locate/ejphar

Cardiovascular pharmacology

Levosimendan alone and in combination with valsartan prevents stroke in Dahl salt-sensitive rats Jouko Levijoki a, Matti Kivikko a, Piero Pollesello a,n, Jukka Sallinen a, Minja Hyttilä-Hopponen a, Mikko Kuoppamäki a, Kristiina Haasio a, Olli Gröhn b, Riitta Miettinen c, Jukka Puoliväli d, Leena Tähtivaara d, Juha Yrjänheikki d, Antti Haapalinna a a

Critical Care Proprietary Products, Orion Pharma, Orionintie 1, P.O. Box 65, FI-02101 Espoo, Finland A.I.Virtanen Institute for Molecular Sciences, Neulaniementie 2, P.O. Box 1627, FIN-70211 Kuopio, Finland Tampere University of Technology, Korkeakoulunkatu 10, FI-33720 Tampere, Finland d Cerebricon Ltd., c/o Charles River Laboratories, Microkatu 1, FI-70210 Kuopio, Finland b c

art ic l e i nf o

a b s t r a c t

Article history: Received 3 October 2014 Received in revised form 20 January 2015 Accepted 21 January 2015 Available online 29 January 2015

The effects of levosimendan on cerebrovascular lesions and mortality were investigated in models of primary and secondary stroke. We aimed to determine whether the effects of levosimendan are comparable to and/or cumulative with those of valsartan, and to investigate whether levosimendaninduced vasodilation has a role in its effects on stroke. In a primary stroke Dahl/Rapp rat model, mortality rates were 70% and 5% for vehicle and levosimendan, respectively. Both stroke incidence (85% vs. 10%, P o0.001) and stroke-associated behavioral deficits (7-point neuroscore: 4.59 vs. 5.96, P o0.001) were worse for vehicle compared to levosimendan. In a secondary stroke model in which levosimendan treatment was started after cerebrovascular incidences were already detected, mean survival times were 15 days with vehicle, 20 days with levosimendan (P ¼0.025, vs. vehicle), 22 days with valsartan (P¼ 0.001, vs. vehicle), and 31 days with levosimendan plus valsartan (P o0.001, vs. vehicle). The respective survivals were 0%, 16%, 20% and 59%, and the respective incidences of severe lesions were 50%, 67%, 50% and 11%. In this rat model, levosimendan increased blood volume of the cerebral vessels, with significant effects in the microvessels of the cortex (ΔR ¼3.5 70.15 vs. 2.7 70.17 ml for vehicle; P ¼0.001) and hemisphere (ΔR ¼3.2 70.23 vs. 2.6 7 0.14 ml for vehicle; P ¼0.018). Overall, levosimendan significantly reduced stroke-induced mortality and morbidity, both alone and with valsartan, with apparent cumulative effects, an activity in which the vasodilatory effects of levosimendan have a role. & 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Keywords: Levosimendan Stroke Animal model Survival Hypertension Cerebral blood volume

1. Introduction Levosimendan is a cardioprotective inodilator that was developed for intravenous treatment of acute decompensated heart failure. The mechanisms of action (Papp et al., 2012) and clinical effects (Nieminen et al., 2013) of levosimendan have been described previously. In brief, levosimendan has a cardioprotective inodilatory effect by acting on three main targets: cardiac troponin C on the contractile apparatus of cardiomyocytes; ATP-dependent potassium channels on the sarcolemma of smooth muscle cells in the vasculature; and mitochondrial ATP-dependent potassium channels in cardiomyocytes, and potentially in other key organs (Papp et al., 2012). In more detail, levosimendan-induced vasodilation has been described in various models for arteries (Montes et al., 2006;

n

Corresponding author. Tel.: þ 358 50 966 4191. E-mail address: [email protected] (P. Pollesello).

Ozdem et al., 2006; Usta et al., 2006; Yildiz et al., 2006), veins (Bragadottir et al., 2013; Höhn et al., 2004; Pataricza et al., 2000), coronary arteries (Gruhn et al., 1998; Kaheinen et al., 2001; Krassói et al., 2000; Michaels et al., 2005), and small resistance vessels (Erdei et al., 2006; Gödény et al., 2013). Very recently, after experimental cardiac arrest/cardiopulmonary resuscitation in a rat model, levosimendan was shown to increase cerebral blood flow, reduce neuronal injury, and improve neurological outcome (Kelm et al., 2014). In previous studies with the Dahl/Rapp salt-sensitive rat model, levosimendan was shown to reduce morbidity and mortality (Louhelainen et al., 2007), which was ascribed to the correction of hypertension-induced diastolic heart failure. According to the literature, however, deaths in the Dahl rat salt-sensitive model can also be caused by stroke (Yamamoto et al., 2007), and indeed some stroke-induced mortality studies have been performed in this rat model (Lin et al., 1999; von Lutterotti et al., 1992). Therefore, a legitimate question arose as to whether the effects of levosimendan

http://dx.doi.org/10.1016/j.ejphar.2015.01.037 0014-2999/& 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

on mortality seen in previous studies might have been correlated, at least in part, to its effects on stroke. To investigate this further, we set up experimental models in which the effects of levosimendan were tested in primary and secondary stroke. In our three-step strategy, we performed three experimental protocols sequentially: (A) a pilot study aimed both at testing whether levosimendan has an effect on primary stroke, and at selecting the correct time for planning a secondary stroke model (study A); (B) a mortality study in a secondary stroke model, which was validated by inclusion of a valsartan arm, as angiotensin receptor blockers are commonly used in secondary prevention of stroke (study B); and (C) a study in a secondary stroke model to determine the mechanism of action of the effects of levosimendan (study C). Our hypothesis was that levosimendan has positive effects on stroke. In detail, our aims were: (1) to determine the effects of levosimendan on stroke-induced death rates in a Dahl/Rapp rat model of primary and secondary stroke; (2) to determine whether the effects of levosimendan are comparable to and/or cumulative with those of valsartan; and (3) to determine whether the vasodilatory effects of levosimendan have a role in levosimendan action in this stroke model.

2. Materials and methods 2.1. Animals The present study was performed in accordance with the guidelines of the Council of Europe and the US National Research Council. Approval was granted by the Animal Ethics Committees of Orion Pharma, or the National Laboratory Animal Centre, Kuopio, Finland. Male Dahl salt-sensitive rats (SS/JrHsd; aged 4–6 weeks on arrival; Harlan US, Indianapolis, USA) were housed at a standard temperature (22 72 1C) and in a light-controlled environment. The rats had water and food ad libitum. After arrival, the rats were fed with standard pelleted food for the first week. Thereafter, they were fed with a salt-rich diet (7% NaCl; Special Diet Services, UK) until the end of the study.

2.2. Levosimendan and valsartan Levosimendan is the ( ) enantiomer of {[4-(1,4,5,6-tetrahydro-4methyl-6-oxo-3-pyridazinyl)phenyl]hydrazono}propanedinitrile, and it was synthesized by Orion Pharma, Finland. Valsartan [(S)-3methyl-2-(N-{[20 -(2 H-1,2,3,4-tetrazol-5-yl)biphenyl-4-yl]methyl} pentanamido)butanoic acid] was purchased from Haorui PharmaChem Inc. (New Jersey, USA). The selected doses of levosimendan and valsartan were 1 mg/kg/day and 10 mg/kg/day, respectively. All of the treatments were administered via the drinking water. The solubility and stability of levosimendan and valsartan were confirmed prior to the study. The levosimendan dose was selected on the basis of the pharmacokinetics and pharmacodynamics data described by Louhelainen et al. (2007). The consumption of the drinking water was measured daily. The exposure of the rats to levosimendan and valsartan was calculated based on the mean drinking water consumption per rat, according to the known drug concentrations in the drinking water. In addition, the plasma levels of levosimendan of the rats were determined after 3 weeks and 9 weeks of treatments, using liquid chromatography–tandem mass spectrometry (as described previously by Louhelainen et al., 2007). The doses were adjusted weekly based on the 24-h water consumption and the rat body weight.

133

2.3. Study protocols Three experimental designs were followed, all of which were based on the Dahl salt-sensitive rat model fed with a high-salt diet, and which were run sequentially; they are here designated as studies A, B and C (Fig. 1). For these three experimental protocols, the rats were randomized to the treatments using random allocation software, but the investigators were not blinded. The sample size in study A was determined based on the expected mortality of these Dahl salt-sensitive rats on a high-salt diet (see Louhelainen et al., 2007). The sample size in study B was based on the mortality results from study A. The Animals in Research: Reporting In-Vivo Experiments (ARRIVE) guidelines were followed (Kilkenny et al., 2010). 2.3.1. Study A: primary stroke model Initially, levosimendan was tested in this primary stroke Dahl/ Rapp rat model (study A), in which the mortality and stroke incidence were measured, based on the changes in the T2weighted magnetic resonance imaging (MRI) and on the histopathology of the rat brains. In this primary stroke model, detection of the diffusion of ischemic brain lesions and the T2-weighted invivo MRI was performed once a week during treatment weeks 2–8 (see Section 2.7). Seven-point neuroscore testing was performed once a week during weeks 1–3, and twice a week during weeks 4–8 (see Section 2.5). Brain lesions were also determined postmortem after the 8 weeks of the study period (see Section 2.8). After ex-vivo imaging, the rat brains were processed for histopathological analyses (see Section 2.8). In study A, levosimendan treatment was started at the same time as salt loading. As no lesions in the brains were expected at that time, we considered study A as a primary stroke-prevention model. The results of study A in terms of time to death and brain lesions were used in the design and planning of studies B and C. Indeed, when 10% overall mortality was reached in study A, cerebrovascular lesions were detectable in the majority of the rats. Thus the levosimendan treatment in study B began at this later time, and study B was considered as a secondary stroke model. 2.3.2. Study B: secondary stroke model The secondary stroke model (study B) was set up and validated by including valsartan, an angiotensin receptor blocker that is used as a treatment for secondary prevention of stroke, the use of which has also shown reduced mortality in similar salt-sensitive rat models Webb et al., 1998). In this model, the effects of levosimendan on mortality and stroke incidence were determined in comparison to and in combination with valsartan. The parameters measured included blood pressure, blood volume of cerebral vessels, cerebrovascular lesions, damage to other organs, survival time, and survival rate. Here, the effects were studied for vehicle (n ¼20), levosimendan (n ¼45), valsartan (n ¼46), and levosimendan plus valsartan (n ¼46), on mortality, morbidity, and histopathological changes (see Section 2.8). The levosimendan and valsartan treatments were started when 410% mortality was observed, which was after about 5 weeks on the high-salt diet. The treatment phase lasted about 5 weeks. Histological analyses of the brain, heart and kidneys (see Section 2.8) were performed for all of the rats that survived for at least 3 weeks from the beginning of these treatments. 2.3.3. Study C: effects on brain blood volume in the secondary stroke model A third model was designed to determine the role of levosimendan-induced vasodilation in mortality for the secondary stroke model (study C). The overall health status and mortality of the rats were followed throughout the studies, and non-invasive

134

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

Study A. Primary stroke model Control group on normal diet, n = 4 Vehicle on high-salt diet, n = 20 Levosimendan on high-salt diet, n = 20 • Salt diet and treatments at age of 6-7 weeks • Duration of study: 8 weeks • MRI once a week, between study weeks 2 to 8 • Neuroscore testing once a week for study weeks 1-3, and twice a week for study weeks 4-8 • Blood pressure measurement on weeks 3 and 7 • Histopathology

Study B. Secondary stroke model Vehicle on high-salt diet, n = 20 Levosimendan on high-salt diet, n = 45 Valsartan on high-salt diet, n = 46 Valsartan+levosimendan on high-salt diet, n = 46 • Daily morbidity and mortality • Weekly body weight • Weekly food and water consumption per 24 h • Blood pressure after 30 days of treatment • Brain, heart and kidney histologic examinations at the end of the study

Study C. Secondary stroke model with analysis of brain blood volumes Vehicle on high-salt diet, n = 14 Levosimendan on high-salt diet, n = 12 • Daily morbidity and mortality • Weekly body weight • Weekly food and water consumption per 24 h • Blood pressure after 30 days of treatment

Fig. 1. Study outlines. The design of the present study was as a three-step series, in which each successive step was started after the analysis of the previous step, to optimize the use of resources (both animals and methods) without compromising the power of the statistical analyses needed to confirm or reject our hypotheses. See text for details.

blood pressure was measured at various times. The effects of levosimendan on other organs were also determined. In study C, the Dahl salt-sensitive rats were exposed to the high-salt diet starting at an age of 6–7 weeks, and they continued until the end of the study, which was up to 9 weeks. Here, the levosimendan treatment was administered during the last 2 weeks, whereby the effects of levosimendan (n ¼12) on stroke incidence and blood volume in the small and large cerebral vessels were measured and compared to the vehicle-treated rats (n¼ 14). 2.4. Health status and mortality The overall health status of each rat was monitored every day by general inspection, when the rat was alive and capable of drinking and eating. During the levosimendan and valsartan treatments, in addition to the general inspection, each rat was handled every working day (i.e., Monday to Friday) to monitor for changes in their general status. If the condition of the rat was not normal, the observed abnormalities were recorded and the condition was followed carefully. The body weights of abnormally behaving rats were measured three times per week. If the general condition of a rat was clearly less than optimal, and if it appeared that the rat was not going to survive, it was sacrificed using inhaled carbon dioxide.

when lifted gently by the tail (grade 6); consistent flexion of either of the forelimbs, which varied from mild wrist flexion and shoulder adduction to severe posturing with full flexion of wrist and elbow, and adduction with internal rotation of the shoulder, when lifted gently by the tail (grade 5); consistently reduced resistance to a lateral push to either side (grade 3); circling towards either side if pulled and lifted by the tail (grade 2); circling towards either side if pulled by the tail (grade 1); spontaneous circling towards either side (grade 1); and no spontaneous movements (grade 0). The neurological tests were conducted by an investigator who was blinded to the treatment status of the rats. 2.6. Stroke incidence In the primary and secondary models of studies A and C, respectively, stroke incidence was assessed using MRI. In the primary model, MRI analysis was applied at several times, to follow the development of stroke incidence over time. In the secondary model, MRI analysis was performed at the end of the study. The brain lesions were also evaluated by histopathology in both of the stroke models (studies A and B), to validate the MRI findings. 2.7. Magnetic resonance imaging

2.5. Neuroscore testing A seven-point neuroscore test was used to assess motor and behavioral deficits of the rats, as modified from Zausinger et al. (2000). The scale was as follows, as the particular rat responses monitored: normal extension of both forelimbs towards the floor

For the MRI measurements, a Varian Inova console interfaced to a 4.7 T horizontal magnet was used (Magnex Scientific Ltd., Abington, UK), which was equipped with actively shielded gradient coils (Magnex Scientific). For the in-vivo measurements, the rats were anesthetized with Isoflurane (1% in 30/70 O2/N2O), fixed

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

into a head holder, and positioned in the magnet bore in a standard orientation relative to gradient coils. An actively decoupled quadrature volume-surface radio-frequency coil pair (Rapid Biomedical, Germany) was used for signal transmission and reception. 2.8. Histopathology The brain histopathology evaluations were performed for both of the stroke models. In the secondary model (study B), the histopathology evaluation was performed also on the heart and kidneys of all of the rats treated for at least 3 weeks. Transmission electron microscopy was applied for selected brain samples of the secondary model, to evaluate the potential stroke-prevention mechanisms in the brains after the treatments with levosimendan, valsartan and their combination. At necropsy, the tissues were removed and fixed in buffered 4% formaldehyde solution (pH 7.2– 7.4) for 2 days, then rinsed three times for 2 h in 0.1 M phosphate buffer. The brains were immersed in 30% sucrose, and then cut into 50 mm sections and divided into 20 series in cryoprotection solution containing 30% ethylene glycol and 30% glycerol in phosphate buffer, and then kept at  20 1C until further processed. One series of sections was selected for staining. To ensure that each section had an equal chance of being sampled for the analysis, the selection of the section series was random, using a random number table. The brain histopathology was carried out blinded from the knowledge of the treatment groups. The kidneys and hearts were processed in paraffin, cut into 4 mm sections, and stained with hematoxylin and eosin, to investigate the general morphology under light microscopy. All of the findings were graded from 0 to 5, where 0 indicated no findings; 1, minimal change; 2, slight change; 3, moderate change; 4, marked change; and 5, severe change. The brains underwent diaminobenzidine (DAB) staining for endogenous peroxidase, to monitor for changes in blood circulation. The sections were washed in 0.1 M phosphate buffer, pH 7.4, for 3  30 min. They were then incubated in the dark for 20 min in filtered 0.05% DAB in phosphate buffer. For the color reaction, 100 ml 30% H2O2 was added into 100 ml DAB, and the reaction was stopped after 8 min. The sections were then washed in phosphate buffer for 3  30 min, mounted on gelatine-coated slides, and studied under light microscopy. For the general morphology of the brain, Nissl substance, nucleic acids, and proliferation thionin stain were used. The slides were dried at 37 1C and treated with chloroform – 99.5% ethanol solution for 1 h. They were then immersed in 99.5% ethanol for 2  2 min, in 94% ethanol for 2  2 min, in 70% ethanol for 2 min, and then in 50% ethanol for 2 min. After this, the slides were dipped into distilled water for 1 min. The slides were then stained in filtered thionin solution for 5 s, rinsed with distilled water, and dehydrated in 50% ethanol and then 70% ethanol, for 2 min each. The slides were then dipped into the differentiation solution (70% ethanol and acetic acid), followed by dehydration in 94% ethanol for 2 min, and 99.5% ethanol for 2  2 min, followed by clearance in xylene for 2  5 min. Finally the sections were covered with a coverslip and DPX (distrene, plasticiser, xylene) mounting medium. The heart and kidney tissues were processed in paraffin, cut into 4 mm sections, and stained with hematoxylin and eosin for their general morphology, with examination under light microscopy. For electron microscopy, selected brains from the secondary model were post-fixed with phosphate-buffered 1% osmium tetroxide for 2 h. After dehydration with ethanol, embedding in epoxy resin, thin-sectioning, and post-staining with uranyl acetate and lead citrate, the thin sections were studied with a JEOL-1200EX transmission microscope operating at 60 kV (Jeol Ltd., Akishima, Japan).

135

2.9. Blood-pressure measurements Blood-pressure measurements were made using the noninvasive tail-cuff method (AdInstruments, Oxford, UK). The rats were habituated to the blood-pressure measurement procedure. Prior to measuring, a rat was put into a restrainer tube for 10 min, to allow it to become stabilized under the recording conditions. The blood pressure was measured in both primary and secondary models a few times during the studies (Fig. 1). 2.10. Local increase in blood volume by cerebral blood volume magnetic resonance imaging Cerebral blood volume (CBV)–MRI was used to investigate local increases in blood volume of the cerebral small microvessels in various regions of the brain, including the hemisphere, cortex, hippocampus, and striatum. For CBV–MRI measurements, the T2 and T2n weighed images were acquired before and after injection of an intravascular iron-oxide-based contrast agent into the femoral vein (Endorem; Guerbet SA; 3 mg/kg). The T2-weighed images were acquired from 10 coronal slices that covered the cerebrum in an anterior-to-posterior direction, using a spin-echo sequence with an echo time (ET) of 70 ms, a repetition time (RT) of 2.5 s, a matrix size of 256  128, a field of view of 40  40 mm2, and a slice thickness of 1 mm. The T2n weighed images were obtained from the same slices using a gradient echoimaging sequence with identical resolution and an ET of 15 ms, a RT of 1.5 s, and a flip angle of  701. The data were analyzed using inhouse written Matlab software. The ΔR2 (R2¼1/T2) and ΔR2n maps were calculated from the signal intensity differences between the pre-contrast and post-contrast agent images. The change in relaxation rate caused by the intravascular contrast agent was assumed to be directly proportional to the cerebral blood volume in each pixel. The ΔR2n is sensitive to both small and large microvessels (diameter, 48 mm), while the ΔR2 changes were dominated by small vessel contributions (diameter, 8–12 mm). The ΔR2 and ΔR2n were quantified from 10 antero-posterior sections, and from the following four manually selected coronal regions of interests: right hemisphere, right hippocampus, right cortex, and right striatum. 2.11. Data management and statistical analysis Differences between experimental groups were analyzed using analysis of variance (ANOVA) with post-hoc tests in the case of all of the normally distributed continuous data. The differences between the treatment groups for the animal status scores were compared using the Pearson chi-squared and Fischer's exact tests. The number of animals used in the studies was justified by the need to power the statistical analysis to determine the effects of the treatments, including the effects on mortality. The cumulative survival data were defined using Kaplan–Meier analyses across time and between all of the treatments (IBM SPSS Statistics, v19). The incidences of histological findings were analyzed using Fischer's exact two-tailed tests and Cochran–Armitage tests. All of the numerical results are shown as means 7S.E.M. For all of the tests, a P-value (two-sided) of o0.05 was considered statistically significant.

3. Results 3.1. Study A: primary stroke model In the vehicle salt-diet group, the neuroscores started to decrease from the second week of the study. Levosimendan reduced and

136

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

after 6 weeks (data not shown). The first deaths in the control highsalt diet group were observed after 3 weeks (see Fig. 2B). In the control and levosimendan groups, no deaths were observed before 7 weeks of the study. By the end of the study, the levosimendan treatment significantly reduced the mortality compared to the vehicle treatment on the high-salt diet (5% vs. 70%, respectively). No deaths were observed in the control rats maintained on the lowsalt diet. In the 6th week of the study, more than 50% of the vehicle rats on the high-salt diet had noticeable T2-weighted MRI changes (Fig. 2C). No changes were observed at that time for either the control rats maintained on the low-salt diet or the levosimendan-treated group

markedly postponed stroke-associated behavioral deficits observed in the neuroscore testing (Fig. 2A). The first decline in the neuroscores for the levosimendan-treated rats appeared after 6 weeks of treatment, while in the vehicle-treated rats, the neuroscores started to decline already after 3 weeks. The control rats on a low-salt diet had maximum neuroscores throughout the study. Levosimendan also had beneficial effects on food consumption, body temperature, and body weight. In the levosimendan group (on the high-salt diet; i. e., 7% salt content), the body weight gain was similar to that in the control group (on the low-salt diet; i.e., 0.7% salt content). In the vehicle group (on the high-salt diet), the body weight gain leveled off after 4 weeks on the high-salt diet, and then started to decrease

6

100

*** ***

80

4

Survival, %

Neuroscore

5

* ***

3 2

60

40

Vehicle

Vehicle

Levosimendan

1

20

Levosimendan 1 mg/kg/day

Control

Control 0

0 0

1

2

3

4

5

6

7

8

0

7

14

Weeks

VEHICLE 7

20

35

42

49

56

100

n = rats alive

17 15 12

20 20

20 20

Neuropathological changes, % within treatment

Stroke incidence, %

100

20 19

28

Days

LEVOSIMENDAN

n = rats alive

20

21

20 19

80 60 40 20

80 60 40 20 0

0

Vehicle 2

3

4

5

6

7

8

2

3

Weeks T2-LESION VOLUME (mm ) 0 1-50 51-100

4

5

Weeks 101-200

>200

6

7

Levosimendan

Control

8 Large cavities

Moderate to large necrosis

White matterhemorrahages

Minor changes

No changes

Fig. 2. Study A: effects of levosimendan vs. vehicle in a primary stroke model in Dahl salt-sensitive rats, induced by a high-salt diet. (A) neuroscore; (B) mortality; (C) stroke incidence and lesion volume; (D) pathological changes in the brain. nnn, Po 0.001 (levosimendan vs. vehicle, both for rats on high-salt diet), n, P¼ 0.022 (rats on control diet vs. rats on high-salt diet [Vehicle]). (E, F) Representative MRI for detection of brain lesions. Brain T2-weighted MRI images of vehicle-treated (E) and levosimendan-treated (F) rats. Extensive lesion is seen as a hyperintense area in the MRI image of the vehicle-treated rat (E). The lesion has high signal intensity in T2-weighted images, typical for high water content (edematous tissue) and destroyed tissue structures. In (C), at week 2 in the levosimendan-treated rats, the MRI shows a bright area (T2w images) right next to the cerebellum (Dav  2.7 mm2/s), which might have been an erroneous finding due to an enlarged ventricle, although it was recorded as a minor change.

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

(on the high-salt diet). For the final MRI measurements at week 8, there were detectable T2-weighted MRI changes in 85% of the vehicle-treated rats (e.g., Fig. 2E), compared to only 10% of the levosimendan-treated rats (e.g., Fig. 2F). The blood pressures of the levosimendan-treated group (21374 mmHg, in 20 survivors at week 7) were slightly, but not significantly, increased compared to the vehicle group (21175 mmHg in 8 survivors at week 7). On the basis of the type and extent of the neuropathological changes (Fig. 2D), the rat brains could be roughly categorized into five different groups: brains with large cavities without brain tissue (n¼ 9; all in the vehicle group), brains with moderate to large necrotic areas (n¼4; three from the vehicle group, one from the levosimendan group), brains with extensive hemorrhage in the white matter (n¼6; four from the vehicle group, two from the levosimendan group), and brains with only minor changes (n¼ 5; one from the vehicle group, four from the levosimendan group), including, for example, only one hemorrhage, gliotic patch or infarct. Severe damage was thus seen for 80% of the brains in the vehicle-treated rats, while in the levosimendan group, only 15% showed severe damage; this difference was statistically significant (Po0.001). 3.2. Study B: secondary stroke model 3.2.1. Stroke incidence Based on the histopathology assessments, stroke-induced brain lesions were seen as large cavities in the brain. In the vehicle group and in the levosimendan and valsartan treated groups, most of the rats ( 460%) had defects, such as pannecrosis (grade 5) in the brain tissue connected to these cavities (Fig. 3A). However, in the rats treated with the combined therapy of levosimendan plus valsartan, only 4 of the rat brains out of 35 (11%) showed pannecrosis. Instead, there were only smaller necrotic areas, and the brains were partly restored (grade 2). In connection with the necrosis, hemorrhages and enlarged ventricles were observed in the brains of most of the control group rats, but these findings were less frequent in the levosimendan or valsartan treated animals, while less than 10% of the brains in the rats treated with both levosimendan and valsartan had enlarged ventricles (3%) or hemorrhages (6%). Electron microscopy evaluation of the brain cortex revealed more vesicle formation and uneven, thinner basal lamina in the vessels after the levosimendan or valsartan treatments than after the levosimendan plus valsartan combination treatment (Fig. 3B). As verified by the DAB staining, levosimendan prevented blood occlusion formation (i.e., occlusion seen for 0/7 animals) in the brains of the rats that survived until the end of study B. In contrast, in the brains of both the valsartan and the levosimendan plus valsartan combination groups, occlusion formation was seen (5/9, 10/27, respectively; P¼ 0.029, P¼0.037 compared to the levosimendan group, respectively; Fig. 3C). No differences were observed in bleeding or gliosis between the levosimendan, valsartan and combination groups (7/7, 8/9, 15/27, respectively). For the white matter destruction (as visualized by hematoxylin and eosin staining), in the high-salt diet control group, all of the brains of the rats studied had white matter lesions (i.e., 100%). Instead, this was seen as 71% and 75% of the brains of the levosimendan and valsartan treated rats, respectively, which compared to only 31% in the levosimendan plus valsartan combination group. The stroke incidence in the levosimendan plus valsartan combination treatment thus declined in terms of the pan-necrosis/necrosis and white matter lesions in the brains. 3.2.2. Mortality In the levosimendan plus valsartan combination treatment group, there was also lower mortality compared to the other groups (Fig. 3D). The mean survivals were 15, 20, 22, and 31 days

137

in the control, levosimendan, valsartan, and combination treatment groups, respectively. The respective mortalities were 100%, 84%, 80%, and 41%. 3.2.3. Blood pressure After 4 weeks of these treatments, there were no effects on blood pressure with valsartan alone (211 76 mmHg) or in combination with levosimendan (217 714 mmHg). In the levosimendan and valsartan combination treatment group, the blood pressure was maintained high to the end of the study (2147 7 mmHg). There was no correlation between the severity of the brain lesions and the blood pressure. 3.2.4. Histopathology of heart and kidneys In the histopathological analysis of the heart, all of the treatment groups showed medial thickening of coronary arteries and myocardial necrosis or necrotic foci. The inter-group differences were only minor (data not shown). In the histopathological analysis of the kidneys, all of the treatment groups showed inflammatory cell infiltrates, tubular hyaline deposits, and basophilic foci with thickened basement membrane, although the severity was lower in the levosimendan and valsartan combination treatment group, without reaching significance (data not shown). 3.3. Study C: effects on brain blood volume in the secondary stroke model In this study, in the vehicle-treated and levosimendan-treated rats, the incidence of brain lesions was 85% and 36%, respectively (Fig. 4A). CBV–MRI was used to study local increases in the blood volumes. Here, compared with the vehicle treatment, there was a tendency for levosimendan to increase the blood volume of cerebral small microvessels in various regions of the brain, including the hemisphere, cortex, hippocampus, and striatum (Fig. 4B). In some of the regions, these effects reached significance (hemisphere, P¼ 0.018; cortex, P ¼0.001). The tendency of levosimendan to increase the blood volume in the cerebral large vessels was, however, less marked (Fig. 4C). Interestingly, levosimendan did not alter the systolic (Fig. 4D) and diastolic blood pressures.

4. Discussion We set up three experimental protocols here to answer the following questions: (1) ‘Does levosimendan prevent stroke in a primary prevention model?’; (2) ‘Does levosimendan prevent secondary stroke?’; (3) ‘Does levosimendan act in a comparable and cumulative way to valsartan?’; and (4) ‘What role does vasodilation promoted by levosimendan have in stroke prevention?’. Study A here was a primary stroke prevention model, as the study drugs were started before any brain lesions had occurred. To the best of our knowledge, there is no widely accepted model for secondary stroke prevention, as to mimic secondary stroke prevention the experimental drug treatment should be started only after brain lesions are expected to have already occurred. Based on the data from study A, we were able to conclude that once an overall death rate of 10% had been reached, the majority of the rats had detectable brain lesions. This led us to start the drug treatments only after this point in studies B and C, to represent the effects of the experimental drugs on secondary stroke prevention. To validate the model, we used the positive control in study B of valsartan, which is an angiotensin receptor blocker that has been widely used in secondary prevention of stroke. In study B, the test drug was thus levosimendan, which was administered both alone and in combination with valsartan. Finally, the third study (study C) was performed on the secondary stroke prevention model, to

138

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

100

Brain lesion, %

80

60

40

20

0 Vehicle Normal

Minor

Levosimendan

Valsartan

Mild to moderate

Severe

P=0.037

Combination Very severe

P=0.029

100

100

80

Survival, %

Occlusion incidence, %

80

60

40 n=27

n=7

60

***

40

n=9 Valsartan 10

20

20

Levosimendan 1

** *

Valsartan 10 + Levosimendan 1 Vehicle

0

0 Combination Occlusion

Levosimendan

Valsartan

No occlusion

0

7

14

21

28

35

Days

Fig. 3. Study B: effects of levosimendan and valsartan in the secondary stroke model of Dahl salt-sensitive rats on a high-salt diet. (A) Histological severity of the brain lesions in the rats on treatment that survived at least 3 weeks. Lesions were graded using a five-step scale: 1, minimal; 2, slight; 3, moderate; 4, marked; 5, severe. (B) Representative electron microscopy of rat brain cortex of selected rats that survived until the end of the study. The rats were treated with levosimendan (a), valsartan (b), the levosimendan and valsartan combination (c), and valsartan (d). With levosimendan or valsartan treated rats, vesicle formation was increased (arrowheads). After the combination treatment, the basal lamina is smooth, homogenous and thicker than after either levosimendan or valsartan treatment (white arrows). (C) Incidence of blood occlusion formation in the treatment groups, as determined for the rats that survived until the end of the study. (D) Survival during treatments with vehicle (tap water; n¼ 20), levosimendan (1 mg/kg/day; n ¼45; n, P¼ 0.025), valsartan (10 mg/kg/day; n ¼46; nn, P¼ 0.001; Log Rank (Mantel-Cox) survival vs. vehicle), and levosimendan plus valsartan (n¼ 46; nnn, Po 0.001).

determine what the role of levosimendan-induced vasodilation might be in terms of the overall effects of levosimendan. Our main findings are that in both of these primary and secondary prevention models in Dahl salt-sensitive rats, levosimendan improves survival and prevents strokes. The beneficial effects of levosimendan and valsartan were similar in the secondary stroke prevention model, and a supra-additive improvement in survival was seen when levosimendan and valsartan were combined. The survival improvements were mainly related to the prevention of strokes, as can be seen from the MRI and brain histopathology findings. The animals that died in the course of this study showed only minor pathologies for the heart and kidneys. The MRI data were in very good agreement with the brain histology results. Indeed, at the end of the primary stroke model study, 85% of the

vehicle-treated rats showed histologically severe brain lesions, and at the same time, MRI changes were observed in 85% of these rats. The benefits of levosimendan combined with valsartan were not only limited to lower mortality and lower stroke incidence. These rats that were treated with the levosimendan and valsartan combination were also healthier, as shown by their higher neuroscores and normal body weight gain and water intake. Our results indicate that the positive effects of levosimendan on survival appear not to be dependent on decreased blood pressure. However, we observed an increase in blood volume in the brain microvessels with levosimendan treatment. We can therefore speculate that the mechanisms underlying the beneficial effects of levosimendan in these models are related to its local vasodilatory properties. This hypothesis is also supported by very

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

Blood Volume of Cerebral Small microvessels

P=0.014

100

139

5

ml/100g* min

Stroke incidence, %

P=0.018

P=0.001

80

60

40

4

3

2

20

0

1 Vehicle

Veh

Levosimedan

LS

Hemisphere

Veh

LS

Cortex

Veh

LS

Hippocampus

Veh

LS

Striatum

Blood Volume of Cerebral Large Vessels Systolic blood pressure, mmHg

250

ml/100g* min

20

15

10

5

0

200

150

100

50

0 Veh

LS

Hemisphere

Veh

LS

Cortex

Veh

LS

Veh

LS

Vehicle

Levosimendan

Hippocampus Striatum

Fig. 4. Study C: effects of levosimendan vs. vehicle in a secondary stroke model on Dahl salt-sensitive rats, induced by a high-salt diet. (A) Stroke incidence and cerebrovascular lesions, as assessed by MRI. (B) Blood volume of cerebral small microvessels. (C) Blood volume of cerebral large vessels. (D) Systemic systolic blood pressure.

recent data that show that levosimendan increases cerebral blood flow, reduces neuronal injury, and improves neurological outcome after experimental cardiac arrest/cardiopulmonary resuscitation in a rat model (Kelm et al., 2014). On the other hand, other pharmacologic effects of levosimendan have been described that might also be beneficial, such as anti-ischemic properties (Du Toit et al., 2008; Grossini et al., 2005; Kersten et al., 2000; Leprán et al., 2006; Levijoki et al., 2001; Papp et al., 2006), improved endothelial function (Parissis et al., 2008), and anti-aggregatory effects on platelets Ambrus et al., 2012; Bent and Plaschke, 2013; Kaptan et al., 2008; Plaschke et al., 2012). Therefore, the local vasodilation induced by levosimendan might only partially account for the reduction in the incidence of stroke. Previous nonclinical studies have shown some effects of levosimendan on the brain (Hein et al., 2013; Roehl et al., 2012). Moreover, a recent report showed that levosimendan was beneficial in a model of injury to the immature brain following cardiopulmonary bypass (Namachivayam et al., 2014). Finally, Varvarousi et al. (2014) described positive effects of levosimendan in post-cardiac arrest brain injury, and Bleilevens et al. (2014) showed that levosimendan treatment significantly reduces cerebral infarct size in the cortex in a rat model. Consideration should be given to the effects of these treatments on systemic blood pressure. Neither levosimendan nor valsartan showed any significant effects on blood pressure in our model. One explanation might be that the salt-loading in this rat population that is prone to extremely high blood pressures can lead to treatment-resistant hypertension. Another explanation might be that the rats that survived in the active treatment groups

actually had lower blood pressures compared to the controls that also received high-salt loading, such that the salt-loaded rats died prematurely, and hence their eventual higher blood pressures could not be detected.

5. Conclusions The present study shows that levosimendan can reduce strokeinduced mortality and morbidity, both when used alone and when combined with valsartan, with apparent cumulative effects. These positive effects of levosimendan were seen both in the primary (study A) and secondary (studies B and C) stroke models. As levosimendan is approved for human use and as the levosimendan doses investigated here for stroke prevention are similar to that approved for treatment of acute heart failure, there is a strong rationale to plan exploratory clinical trials in human to assess the safety and efficacy of levosimendan for prevention of secondary stroke.

Acknowledgments We thank the following persons for their cooperation: Jarmo Immonen, Päivi Saikkonen, Anne Alatupa, and Sanna Käsnäp (née Kärkkäinen), and the staff of Cerebricon and CNServices. We thank Dr. Christopher Berrie for scientific editing of the manuscript. This study was performed with an R&D grant by Orion Pharma.

140

J. Levijoki et al. / European Journal of Pharmacology 750 (2015) 132–140

References Ambrus, N., Szolnoky, J., Pollesello, P., Kun, A., Varró, A., Papp, J.G., Pataricza, J., 2012. Prolonged antispasmodic effect in isolated radial artery graft and pronounced platelet inhibition induced by the inodilator drug, levosimendan. Basic Clin. Pharmacol. Toxicol. 110, 269–274. Bent, F., Plaschke, K., 2013. Levosimendan's effect on platelet function in a rat sepsis model. Platelets 24, 189–193. Bleilevens, C., Roehl, A.B., Zoremba, N., Tolba, R., Rossaint, R., Hein, M., 2014. Insular infarct size but not levosimendan influenced myocardial injury triggered by cerebral ischemia in rats. Exp. Brain Res. 233, 149.156. Bragadottir, G., Redfors, B., Ricksten, S.E., 2013. Effects of levosimendan on glomerular filtration rate, renal blood flow, and renal oxygenation after cardiac surgery with cardiopulmonary bypass: a randomized placebo-controlled study. Crit. Care Med. 41, 2328–2335. Du Toit, E.F., Genis, A., Opie, L.H., Pollesello, P., Lochner, A., 2008. A role for the RISK pathway and K(ATP) channels in pre- and post-conditioning induced by levosimendan in the isolated guinea pig heart. Br. J. Pharmacol. 154, 41–50. Erdei, N., Papp, Z., Pollesello, P., Edes, I., Bagi, Z., 2006. The levosimendan metabolite OR-1896 elicits vasodilation by activating the K(ATP) and BK(Ca) channels in rat isolated arterioles. Br. J. Pharmacol. 148, 696–702. Gödény, I., Pollesello, P., Edes, I., Papp, Z., Bagi, Z., 2013. Levosimendan and its metabolite OR-1896 elicit KATP channel-dependent dilation in resistance arteries in vivo. Pharmacol. Rep. 65, 1304–1310. Grossini, E., Caimmi, P.P., Molinari, C., Teodori, G., Vacca, G., 2005. Hemodynamic effect of intracoronary administration of levosimendan in the anesthetized pig. J. Cardiovasc. Pharmacol. 46, 333–342. Gruhn, N., Nielsen-Kudsk, J.E., Theilgaard, S., Bang, L., Olesen, S.P., Aldershvile, J., 1998. Coronary vasorelaxant effect of levosimendan, a new inodilator with calcium-sensitizing properties. J. Cardiovasc. Pharmacol. 31, 741–749. Hein, M., Zoremba, N., Bleilevens, C., Bruells, C., Rossaint, R., Roehl, A.B., 2013. Levosimendan limits reperfusion injury in a rat middle cerebral artery occlusion (MCAO) model. BMC Neurol. 13, 106. Höhn, J., Pataricza, J., Petri, A., Tóth, G.K., Balogh, A., Varró, A., Papp, J.G., 2004. Levosimendan interacts with potassium channel blockers in human saphenous veins. Basic Clin. Pharmacol. Toxicol. 94, 271–273. Kaheinen, P., Pollesello, P., Levijoki, J., Haikala, H., 2001. Levosimendan increases diastolic coronary flow in isolated guinea-pig heart by opening ATP-sensitive potassium channels. J. Cardiovasc. Pharmacol. 37, 367–374. Kaptan, K., Erinç, K., Ifran, A., Yildirim, V., Uzun, M., Beyan, C., Işik, E., 2008. Levosimendan has an inhibitory effect on platelet function. Am. J. Hematol. 83, 46–49. Kelm, R.F., Wagenführer, J., Bauer, H., Schmidtmann, I., Engelhard, K., Noppens, R.R., 2014. Effects of levosimendan on hemodynamics, local cerebral blood flow, neuronal injury, and neuroinflammation after asphyctic cardiac arrest in rats. Crit. Care Med. 42, e410–419. Kersten, J.R., Montgomery, M.W., Pagel, P.S., Warltier, D.C., 2000. Levosimendan, a new positive inotropic drug, decreases myocardial infarct size via activation of K(ATP) channels. Anesth. Analg. 90, 5–11. Kilkenny, C., Browne, W.J., Cuthill, I.C., Emerson, M., Altman, D.G., 2010. Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol. 8, e1000412. Krassói, I., Pataricza, J., Kun, A., Papp, J.G., 2000. Calcium-dependent vasorelaxant capacity of levosimendan in porcine and human epicardial coronary artery preparations. Cardiovasc. Drugs Ther. 14, 691–693. Leprán, I., Pollesello, P., Vajda, S., Varró, A., Papp, J.G., 2006. Preconditioning effects of levosimendan in a rabbit cardiac ischemia-reperfusion model. J. Cardiovasc. Pharmacol. 48, 148–152. Levijoki, J., Pollesello, P., Kaheinen, P., Haikala, H., 2001. Improved survival with simendan after experimental myocardial infarction in rats. Eur. J. Pharmacol. 419, 243–248. Lin, K.F., Chao, J., Chao, L., 1999. Atrial natriuretic peptide gene delivery reduces stroke-induced mortality rate in Dahl salt-sensitive rats. Hypertension 33, 219–224. Louhelainen, M., Vahtola, E., Kaheinen, P., Leskinen, H., Merasto, S., Kytö, V., Finckenberg, P., Colucci, W.S., Levijoki, J., Pollesello, P., Haikala, H., Mervaala, E.M., 2007. Effects of levosimendan on cardiac remodeling and cardiomyocyte apoptosis in hypertensive Dahl/Rapp rats. Br. J. Pharmacol. 150, 851–861.

Michaels, A.D., McKeown, B., Kostal, M., Vakharia, K.T., Jordan, M.V., Gerber, I.L., Foster, E., Chatterjee, K., 2005. Effects of intravenous levosimendan on human coronary vasomotor regulation, left ventricular wall stress, and myocardial oxygen uptake. Circulation 111, 1504–1509. Montes, F.R., Echeverri, D., Buitrago, L., Ramírez, I., Giraldo, J.C., Maldonado, J.D., Umaña, J.P., 2006. The vasodilatory effects of levosimendan on the human internal mammary artery. Anesth. Analg. 103, 1094–1098. Namachivayam, P., Smolich, J.J., Shields, A.E., Rees, S., Coleman, L., Horton, S.B., Konstantinov, I.E., Penny, D.J., Shekerdemian, L.S., 2014. Impact of levosimendan on brain injury patterns in a lamb model of infant cardiopulmonary bypass. Pediatr. Res. 76, 64–71. Nieminen, M.S., Fruhwald, S., Heunks, L.M., Suominen, P.K., Gordon, A.C., Kivikko, M., Pollesello, P., 2013. Levosimendan: current data, clinical use and future development. Heart Lung Vessel 5, 227–245. Ozdem, S.S., Yalcin, O., Meiselman, H.J., Baskurt, O.K., Usta, C., 2006. The role of potassium channels in relaxant effect of levosimendan in rat small mesenteric arteries. Cardiovasc. Drugs Ther. 20, 123–127. Papp, J.G., Pollesello, P., Varró, A.F., Végh, A.S., 2006. Effect of levosimendan and milrinone on regional myocardial ischemia/reperfusion-induced arrhythmias in dogs. J. Cardiovasc. Pharmacol. Ther. 11, 129–135. Papp, Z., Édes, I., Fruhwald, S., De Hert, S.G., Salmenperä, M., Leppikangas, H., Mebazaa, A., Landoni, G., Grossini, E., Caimmi, P., Morelli, A., Guarracino, F., Schwinger, R.H., Meyer, S., Algotsson, L., Wikström, B.G., Jörgensen, K., Filippatos, G., Parissis, J.T., González, M.J., Parkhomenko, A., Yilmaz, M.B., Kivikko, M., Pollesello, P., Follath, F., 2012. Levosimendan: molecular mechanisms and clinical implications: consensus of experts on the mechanisms of action of levosimendan. Int. J. Cardiol. 159, 82–87. Parissis, J.T., Karavidas, A., Bistola, V., Arapi, S., Paraskevaidis, I.A., Farmakis, D., Korres, D., Filippatos, G., Matsakas, E., Kremastinos, D.T., 2008. Effects of levosimendan on flow-mediated vasodilation and soluble adhesion molecules in patients with advanced chronic heart failure. Atherosclerosis 197, 278–282. Pataricza, J., Hõhn, J., Petri, A., Balogh, A., Papp, J.G., 2000. Comparison of the vasorelaxing effect of cromakalim and the new inodilator, levosimendan, in human isolated portal vein. J. Pharm. Pharmacol. 52, 213–217. Plaschke, K., Bent, F., Rosenhagen, C., Wagner, S., Hofer, S., Kopitz, J., 2012. The inhibitory in-vitro effect of high-dose levosimendan on platelet function may be mediated through its action as a phosphodiesterase inhibitor. Coron. Artery Dis. 23, 215–220. Roehl, A.B., Zoremba, N., Kipp, M., Schiefer, J., Goetzenich, A., Bleilevens, C., Kuehn-Velten, N., Tolba, R., Rossaint, R., Hein, M., 2012. The effects of levosimendan on brain metabolism during initial recovery from global transient ischaemia/hypoxia. BMC Neurol. 12, 81. Usta, C., Eksert, B., Gölbasi, I., Bigat, Z., Ozdem, S.S., 2006. The role of potassium channels in the vasodilatory effect of levosimendan in human internal thoracic arteries. Eur. J. Cardiothorac. Surg. 30, 329–332. Varvarousi, G., Stefaniotou, A., Varvaroussis, D., Aroni, F., Xanthos, T., 2014. The role of levosimendan in cardiopulmonary resuscitation. Eur. J. Pharmacol. 740, 596–602. von Lutterotti, N., Camargo, M.J., Campbell Jr., W.G., Mueller, F.B., Timmermans, P.B., Sealey, J.E., Laragh, J.H., 1992. Angiotensin II receptor antagonist delays renal damage and stroke in salt-loaded Dahl salt-sensitive rats. J. Hypertens. 10, 949–957. Webb, R.L., Barclay, B.W., Navarrete, A.E., Wosu, N.J., Sahota, P., 1998. Protective effects of valsartan and benazeprilat in salt-loaded stroke-prone spontaneously hypertensive rats. Clin. Exp. Hypertens. 20, 775–793. Yamamoto, E., Kataoka, K., Shintaku, H., Yamashita, T., Tokutomi, Y., Dong, Y.F., Matsuba, S., Ichijo, H., Ogawa, H., Kim-Mitsuyama, S., 2007. Novel mechanism and role of angiotensin II induced vascular endothelial injury in hypertensive diastolic heart failure. Arterioscler. Thromb. Vasc. Biol. 27, 2569–2575. Yildiz, O., Nacitarhan, C., Seyrek, M., 2006. Potassium channels in the vasodilating action of levosimendan on the human umbilical artery. J. Soc. Gynecol. Investig. 13, 312–315. Zausinger, S., Hungerhuber, E., Baethmann, A., Reulen, H., Schmid-Elsaesser, R., 2000. Neurological impairment in rats after transient middle cerebral artery occlusion: a comparative study under various treatment paradigms. Brain Res. 863, 94–105.