Background
Human cerebral malaria (HCM) is a life-threatening complication during
Plasmodium falciparum infections, contributing in large part to the estimated 900,000-1,600,000 malaria-related deaths worldwide [
1]. Mortality usually exceeds 10% in controlled clinical trials, despite optimal treatment with intravenous anti-malarial drugs [
2,
3], and it is estimated that up to 25% of patients that survive an HCM episode suffer long-term neurological and cognitive deficits [
4,
5]. This scenario indicates that strategies targeting eradication of the parasite alone after HCM development have limitations and therefore adjunctive treatments are needed urgently [
6‐
8].
Plasmodium berghei ANKA (PbA) infection of susceptible mouse strains is the best-studied experimental model of cerebral malaria (ECM) and is characterized by the development of neurological signs six to 12 days post-infection [
7,
9,
10]. Mice with ECM show widespread cerebral vasoconstriction which markedly decreases blood flow in their brain [
11]. High doses of the dihydropyridine calcium channel blocker nimodipine, given as an adjunctive therapy with artemether, prevents cerebral vasoconstriction and increases survival of mice with ECM when compared to control mice treated with the anti-malarial drug alone [
11]. However, hypotension, bradycardia, arrhythmias, and eventually death may occur when nimodipine is given parenterally at high doses for humans [
12,
13] and these side effects could preclude the potential use of high doses of this drug to treat HCM, particularly in the most severe cases associated with shock and hypotension [
14‐
16]. To solve this problem, the potential of alternative delivery systems and low doses for nimodipine in association with artesunate to rescue mice with late-stage ECM was analysed. The aim was to define optimal protocols and achieve maximum efficacy in increasing survival in rescue therapy while causing the least cardiac side effects.
Plasmodium berghei-infected mice become hypotensive when presenting ECM signs [
17], and therefore are a model particularly suited to study therapeutic approaches devised to address the most severe scenarios of this neurological complication.
Because little is known about the electrocardiographic (ECG) features of mice with ECM, the baseline ECG and arterial pressure characteristics of uninfected control animals and of mice with ECM and the response upon rescue treatment with artesunate associated or not with nimodipine were also determined. It is shown that continuous delivery of nimodipine by an implanted osmotic pump at low doses did not worse hypotension and improves ECG parameters and survival in rescue therapy of mice with ECM.
Discussion
It was previously shown that the dihydropyridine calcium channel blocker nimodipine, given via parenteral route (IP) as a bolus injection at high doses was able to improve the efficacy of artemether in rescuing mice with late-stage ECM from death [
11]. Nimodipine is a drug administered by oral route to prevent cerebral vasospasm, a major complication of subarachnoid haemorrhages in humans [
12]. However, nimodipine administration can cause potential deleterious cardiovascular side effects in humans, such as hypotension, bradycardia and arrhythmias when given intravenously even at therapeutic doses [
12,
21]. These possible cardiovascular side effects make the presence of hypotension and parenteral route of administration contra-indications for the use of nimodipine in the USA [
22‐
24]. Hypotension may affect a significant fraction of all cerebral malaria cases and is strongly associated with poor outcomes in this population [
14‐
16]; therefore, administration of nimodipine in this scenario might appear counterintuitive. Nevertheless, the present work showed that slow parenteral delivery of nimodipine at low doses caused no significant effects on cardiovascular parameters in normal mice and, more significantly, it actually ameliorated rather than worsened cardiac alterations in mice with late-stage ECM, reassuring its potential as adjunctive therapy. Indeed, mice with ECM present hypotension [
17] and it is described in the present study that they also show bradycardia and ECG changes predisposing to arrhythmia, indicating that this host-parasite combination models the most severe and lethal forms of HCM. These findings provide better support for future clinical trials using nimodipine as adjunctive therapy in cerebral malaria.
Human CM encompasses a variety of clinical and pathological entities in which neurological involvement can be accompanied or not by several other complications such as respiratory distress, severe anaemia, acidosis, renal failure and shock [
5,
14,
15,
25]. Although significant bradycardia is not commonly present, ECG changes, cardiac arrhythmias and evidence of myocardial failure can also be observed in a fraction of patients with severe [
26,
27] and uncomplicated [
28]
Plasmodium falciparum malaria. ECG changes present in malaria patients have been shown to be mainly due to delayed conduction of various kinds and seem not to be related to death in affected patients [
26‐
28]. This is the first study to report ECG changes present in mice with late-stage ECM. Previous studies monitoring ECG in PbA-infected mice without ECM showed bradycardia [
29,
30], but none of them described any changes in heart electrophysiology. The present work shows that mice with ECM present bradycardia, a global lengthening in the depolarization-repolarization cardiac cycle and an increase in the autonomic tone to the heart. ECG changes and bradycardia could be derived from both a directed effect of PbA infection in the heart and an indirect effect resulting from central nervous system dysfunction or metabolic disturbances. Direct damage to the cardiac conduction system and myocardium could explain the global enlargement of ECG intervals. In fact, the presence of cardiac lesions has been shown, ranging from myocyte cytoplasmic vacuolization to necrosis in mice infected with
Plasmodium chabaudi chabaudi,
Plasmodium vinckei petteri and
Plasmodium yoelii nigeriensis[
31]. Additionally, endomyocardial lesion and fibrosis have been described in PbA-infected mice [
32], although direct PbA-induced heart pathology has not been confirmed in other studies [
33].
Indirect cardiac effects due to central nervous system or metabolic dysfunction can also occur in mice with ECM and could account for the ECG changes observed. Mice with ECM develop hypothermia that is known to be associated with bradycardia and ECG changes reflecting slowing of myocardial conduction such as AV blocks, asystole, and increased PR, QRS and QTc intervals [
34]. On the other hand, the increased autonomic tonus to the heart can reflect an autonomic dysfunction induced by PbA-infection that could be part of the ECM syndrome. In addition, other factors that can be present in late-stage ECM mice such as hypovolemia, hypoglycaemia, electrolyte disturbances, and acidosis are known to cause ECG changes and could also be related to these findings [
34‐
36]. Further studies to analyse if the ECG changes described are specific of PbA-infected mice or also occur in other parasite-mouse combinations are a natural sequence of the present work. These studies could indicate if the central nervous system dysfunction present in mice with ECM cause or not the ECG changes described.
Nimodipine caused bradycardia, decreases HRV, and increase QRS duration in uninfected mice when given IP as a bolus injection at 4 mg/kg. On the other hand, ECG changes were minimized when nimodipine was given at 0.5 mg/kg/day via osmotic pumps. The cardiac side effects found in uninfected mice treated with nimodipine given IP can be explained by the decrease in the cardiac force of contraction and action potential conduction velocity associated with an increase in the myocyte effective refractory period caused by the blockage of voltage-dependent L-type Ca
+2-channels [
12]. Both the decrease in the treatment dose (from 4 mg/kg to 0.5 mg/kg/day) and the use of a slow and continuous delivery system contributed to prevent the cardiac side effects of nimodipine. In fact, it has been shown that, at the same dose, slow intravenous delivery of nimodipine does not differ from oral administration in efficacy and incidence of side effects when used to prevent vasospasm after subarachnoid haemorrhages [
37].
Interestingly, when given IP, nimodipine increased bradycardia but also improved some ECG changes present in mice with ECM. This improvement can be related to its vasodilatory activity in brain vessels [
11] rescuing normal brain physiology. This hypothesis could explain the shortening in the PR interval in treated mice, as nimodipine action in the heart actually tends to increase its duration when used in humans and healthy animals [
12,
13]. A role for artesunate in the changes observed is highly improbable as vehicle treated mice did not present any improvement in the ECG parameters analysed.
The present work also shows that the infusion of low doses of nimodipine decreased diastolic and mean arterial pressure in uninfected mice, but this effect was not observed in ECM mice. These findings show that low doses of nimodipine are safe to be given in hypotensive ECM mice and that the drug did not delay the arterial pressure recovery upon rescue treatment with artesunate.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YCM carried out the statistical analyses, participated in the ECG data acquisition, and drafted the manuscript. LC carried out most of the ECG data acquisition and the experiments evaluating the efficacy of different nimodipine treatment schemes in rescuing mice with ECM. GMZ participated in the design of the study and in the ECG data acquisition. POS and PKO carried out arterial pressure data acquisition and analysis. JAF participated in the design of the study and critically revised the manuscript. LJMC conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.