Function of the blood-cerebrospinal fluid barrier in human cerebral malaria: rejection of the permeability hypothesis

We tested the hypothesis that cerebral malaria is caused by blood-brain barrier inflammation and cerebral edema. In a group of 157 Thai patients with strictly defined cerebral malaria, cerebrospinal fluid (CSF) opening pressures were normal in 79% and were lower in fatal cases than in survivors (means +/- 1 SD, 144 +/- 58 and 167 +/- 51 mm CSF, respectively, P = 0.051). CSF: serum albumin ratios (X 10(3)) in 39 of them were significantly higher than in 61 British controls (medians 8.5 and 5.5, respectively, P = 0.04), but were no higher in 7 fatal cases. In a group of 12 patients this ratio was not significantly higher during coma than after full recovery (means +/- 1 SD, 9.0 +/- 6.2 and 6.7 +/- 4.2, respectively, P greater than 0.1). CSF alpha 2-macroglobulin concentrations were always normal. CSF : serum 77Br- ratios were elevated in 11/19 comatose cases but fell to normal 4 to 9 days later in 11/11 cases. Dexamethasone treatment had no significant effect on bromide partition. The percentage of an intravenously administered dose of 125I-human serum albumin detectable per ml of CSF 6 hr after intravenous injection was 2.4 +/- 1.3 X 10(-5) in 14 comatose patients and 4.4 +/- 4.0 X 10(-5) in 9 of them during convalescence (P greater than 0.1). These results demonstrate that the blood-CSF barrier is essentially intact in patients with cerebral malaria and give no support to the idea that cerebral edema is the cause of coma.

Cardiovascular toxicity and distribution kinetics of intravenous chloroquine

Chloroquine diphosphate (3 mg base kg-1) was given by constant rate intravenous injection over 10 min to 12 healthy adult male volunteers. Plasma concentrations of chloroquine and the principal metabolite desethylchloroquine, electrocardiograph intervals, and arterial blood pressure were measured at frequent intervals to determine the relationship between cardiovascular effects and plasma concentrations. Peak plasma concentrations ranged between 784 and 6649 (mean 2913) ng ml-1. The decline in plasma concentrations was multiexponential with an initial rapid distribution phase; mean (+/- s.d.) first order rate constant 0.65 +/- 0.14 min-1, and an estimated apparent volume of the central compartment of 0.18 +/- 0.15 l kg-1. There was no serious toxicity, but subjective side effects were reported in all patients and there was a significant fall in systolic blood pressure (110 +/- 9.5 to 101 +/- 12.5 mm Hg; P = 0.03) and rise in heart rate which paralleled the change in plasma chloroquine concentrations. Coincident with changes in blood pressure, there was a significant prolongation of the electrocardiograph QRS interval; 81 +/- 15 to 92 +/- 13 ms (P less than 0.01) but no change in the QTc interval. These findings suggest that the cardiovascular toxicity of parenteral chloroquine is related to transiently high plasma concentrations occurring early in the distribution phase. This results from incomplete distribution from a central compartment that is approximately one thousand times smaller than the eventual total apparent volume of distribution at steady state. Rate of administration is therefore a major determinant of toxicity.

Quinine pharmacokinetics and toxicity in pregnant and lactating women with falciparum malaria

Quinine dihydrochloride (10 mg or, in two patients, a loading dose of 20 mg kg-1) was infused intravenously over 4 h in ten severely ill but conscious women with falciparum malaria complicating the third trimester of pregnancy. Plasma quinine concentrations, measured spectrophotofluorimetrically after benzene extraction, fitted closely a single exponential decline after the intravenous infusion. These data were therefore fitted to a one compartment model: total apparent volume of distribution, V, 0.96 +/- 0.27 l kg-1 (+/- s.d.), elimination half-time (t1/2,z), 11.3 +/- 4.3 h, total clearance, 1.22 +/- 0.77 ml min-1 kg-1. There was no relationship between arterial blood pressure and plasma quinine concentrations. Eight women delivered of live infants while taking quinine, had placental cord plasma quinine concentrations from 1.0 to 4.6 mg l-1 (mean 2.4) which correlated significantly with maternal plasma quinine concentrations (r = 0.78, t = 3.06, P less than 0.05). The mean (+/- s.d.) ratio of cord plasma to maternal plasma quinine concentration was 0.32 +/- 0.14. Heart blood from a foetus aborted at term had a plasma quinine concentration of 2.8 mg l-1; simultaneous maternal plasma quinine was 7.1 mg l-1 (ratio 0.39). Breast milk quinine concentrations and milk to plasma ratios were 0.5-3.6 mg l-1 (mean 2.6) and 0.11-0.53 (mean 0.31) in twenty-five women who were breast-feeding and had taken oral quinine sulphate for 1-10 days (mean 4.0). Five women with more serious infections received intravenous quinine; breast milk quinine concentrations ranged between 0.5 and 8.0 mg l-1 (mean 3.4).(ABSTRACT TRUNCATED AT 250 WORDS)

Hypoglycaemia and antimalarial drugs: quinidine and release of insulin

Life threatening hypoglycaemia has been closely associated with the use of quinine, but the effect of quinidine and the synthetic antimalarials on the homoeostasis of glucose has not been investigated. In volunteers given a fixed dose of 500 mg base and patients with malaria given a quinidine loading dose (15 mg base/kg) mean (SEM) plasma insulin concentrations rose from 6.1 (1.5) mU/l to 10.9 (4.4) mU/l (p less than 0.02) and 10.4 (2.0) mU/l to 18.5 (5.3) mU/l (p less than 0.04), respectively. Plasma glucose concentrations fell from 4.5 (1.1) mmol/l (81 (20) mg/100 ml) to 4.0 (0.3) mmol/l (72 (5) mg/100 ml) in volunteers (p less than 0.04) and from 5.7 (1.3) mmol/l (102 (23) mg/100 ml) to 4.8 (1.6) mmol/l (86 (29) mg/100 ml) in patients (p less than 0.05). One of two patients with cerebral malaria and acute renal failure became profoundly hypoglycaemic (plasma glucose concentration 1.4 mmol/l (25 mg/100 ml), plasma insulin concentration 3.1 mU/l). Hypoglycaemia may occur in any severely ill fasting patient given parenteral quinidine. The other antimalarials tested, chloroquine, amodiaquine, mefloquine, and halofantrine, did not stimulate the release of insulin, an important advantage that should be taken into account when treatment is chosen for Plasmodium falciparum malaria.

The importance of anaemia in cerebral and uncomplicated falciparum malaria: role of complications, dyserythropoiesis and iron sequestration

Ninety-four per cent of 169 patients with cerebral malaria developed anaemia (haematocrit less than 35 per cent) and 30 per cent required blood transfusion to maintain the haematocrit at more than 21 per cent. Anaemia was at its worst on admission in 58 patients (34 per cent); in the rest the haematocrit fell further, reaching its nadir one to 17 days later (mean 2.3 days). The mean lowest haematocrit was 24.3 +/- 7.2 per cent (+/- 1 SD) and the mean maximum fall was 7.9 +/- 5.6 per cent. Anaemia was more severe in patients with bacterial infection, retinal haemorrhages, schizontaemia and in pregnancy. The lowest haematocrit correlated with admission parasitaemia (r = -0.33, p less than 0.001), total serum bilirubin (r = -0.25, p less than 0.01) and serum creatinine (r = -0.22, p less than 0.01). In 23 patients with uncomplicated falciparum malaria the mean serum iron on admission was 53 micrograms/dl (range 16-157) and the mean serum ferritin 1773 ng/ml (range 170-10 000). There was a significant (p less than 0.001) rise in serum iron 96 h after starting antimalarial treatment; the serum ferritin declined slowly over several weeks. Stainable iron was present in all marrows examined and in eight patients the characteristic pattern of the anaemia of chronic disorders was seen. Seventy-three per cent of patients had dyserythropoiesis which was moderate to gross in 36 per cent. Dyserythropoiesis and erythrophagocytosis were often present on admission but sometimes appeared after the parasitaemia had cleared and persisted for at least three weeks into convalescence. These disturbances in iron metabolism and haemopoiesis are not completely explicable by red blood cell parasitisation. They may contribute more to the anaemia than has previously been recognised.

Intragastric mefloquine is absorbed rapidly in patients with cerebral malaria

Mefloquine has proved effective in chloroquine- and quinine-resistant falciparum malaria, but it cannot be given parenterally. We have measured the absorption of mefloquine hydrochloride suspension (mean 15.6, range 9.7-28.6 mg/kg) given by nasogastric tube to 19 cerebral malaria patients already receiving intravenous quinine. Absorption was rapid with both dose schedules used; mean absorption half-times were 1.5 and 1.8 hr, and plasma mefloquine concentrations exceeded 200 ng/g within 3 hr of completing administration in all but one exceptionally ill patient who died 40 hr later. Steady state plasma concentrations over 7 days ranged from 300 to 1,050 (mean 561) ng/g. Bioavailability of mefloquine suspension in cerebral malaria therefore appears to be adequate for treatment in all but the most severely ill patients. Although intragastric mefloquine cannot now be recommended as an alternative to intravenous quinine for the treatment of severe chloroquine-resistant falciparum malaria, this situation could change if quinine resistance increases further.

Intravenous amodiaquine and oral amodiaquine/erythromycin in the treatment of chloroquine-resistant falciparum malaria

In eastern Thailand, 14 adults with moderately severe falciparum malaria were treated with intravenous amodiaquine dihydrochloride, loading dose 10 mg base/kg infused over 4 h followed by three further intravenous infusions of 5 mg base/kg at 24, 48, and 72 h. All patients were clinically cured--mean fever clearance time 37.8 h (range 24-60), mean parasite clearance time 64.9 h (18-164). There were no serious toxic effects. 33 patients aged over 5 years with uncomplicated falciparum malaria were given oral amodiaquine dihydrochloride (mean total dose 41 mg base/kg over 3 days) combined with erythromycin estolate (mean dose 48 mg base/kg daily for 5 days). 2 patients failed to respond. In the other 31 patients mean fever clearance time was 55.9 h (range 10-104) and mean parasite clearance time was 65.4 h (40-120). In both studies, more than half the patients followed-up had recurrent parasitaemia but reinfection could not be excluded. Parasites isolated from 18 patients were highly resistant to chloroquine in vitro.

Factors contributing to the development of cerebral malaria. II. Endotoxin

Limulus amoebocyte lysate test (LALT) was used to detect endotoxin-like substances in the plasma of 15 patients with cerebral malaria, 28 patients with uncomplicated falciparum malaria and 30 healthy controls. On admission, 67% of cerebral malaria patients were positive, whereas only 21.4% of uncomplicated malaria patients and none of controls were positive. Among uncomplicated malaria cases, four of eight patients with parasitaemia over 90,000/mm3 were LALT positive whereas only two of 20 patients with parasitaemia of less than 90,000/mm3 were positive. A follow-up study in cerebral malaria patients showed some variation in LALT positivity rate from day to day (85.7% on day 1, 53.3% on day 3 and all negative on discharge from hospital). LALT positivity bore no relationship to gram negative bacteraemia. Leucocytosis and elevated serum enzymes were more frequently found in LALT-positive patients. Our results suggest that endotoxin (LALT positivity) of the plasma of malaria patients is derived from either the parasites themselves or from the gut. It relates to parasitaemia, leucocytosis and elevated serum enzymes, but not to the clinical syndrome of cerebral malaria.

Quinine and severe falciparum malaria in late pregnancy

Quinine dihydrochloride was given intravenously to 12 women with severe falciparum malaria in the third trimester of pregnancy. The initial dose consisted of 10 or 20 mg salt/kg over 4 h and was followed by 10 mg salt/kg every 8 h until patients were fit to swallow, when quinine sulphate tablets were given. Uterine activity showed little or no change despite rising quinine concentrations. Of 3 patients in labour, 2 proceeded normally while a third had a successful caesarean section for fetal distress. Late (type II) decelerations of the fetal heart rate were recorded in 6 patients before treatment but in most patients signs of fetal distress diminished as the maternal temperature fell. Hypoglycaemia and hyperinsulinaemia developed in 7 patients, in 2 before quinine was started. The important toxic effect of quinine in late pregnancy is not an oxytocic action but rather its capacity to release insulin.

Binding of quinine to plasma proteins in falciparum malaria

Plasma protein binding of quinine was measured in 12 patients with cerebral malaria on the first and seventh day of treatment, and in 7 patients with uncomplicated falciparum malaria on admission and also one month later. Binding was significantly higher and therefore the proportion of free drug was lower in cerebral malaria patients (free: total quinine concentration; 7.2 +/- 3.5%, mean +/- SD, on admission; 7.4 +/- 5.3% on day 7) compared with uncomplicated malaria patients on admission (10.2 +/- 5.8%) or following recovery (11.0 +/- 5.5%, n = 6) P = 0.011. Binding was significantly correlated with the red cell/total concentration ratio r = 0.56, P less than 0.0001. The ratio of cerebrospinal fluid to free (unbound) plasma quinine was 0.55 +/- 0.33 which suggests that quinine does not freely cross the blood brain barrier. These findings are relevant to the interpretation of total plasma or serum concentration, and may explain the rarity of serious quinine toxicity in severe falciparum malaria.

Human cerebral malaria. A quantitative ultrastructural analysis of parasitized erythrocyte sequestration

For investigation of the pathogenesis of cerebral malaria, immediate postmortem samples from brain and other tissues of patients dying with Plasmodium falciparum malaria, with (CM) or without (NCM) cerebral malaria, were processed for electron microscopy. Counts of parasitized erythrocytes (PRBCs) in cerebral and other vessels showed that the proportion of PRBCs was higher in CM than in NCM, and also that the proportion of PRBCs was higher in the brain than in other organs examined in both CM and NCM. Cerebral vessels from CM patients were more tightly packed with RBCs than those from NCM patients, but there was no significant difference in the amount or degree of endothelial damage or numbers of vessels with endothelial pseudopodia. Fibrillar (fibrin) deposits were present in a small proportion of vessels, but no thrombosis was present. There was neither acute nor chronic inflammation, and leukocytes were absent within or outside cerebral vessels. There was no immune complex deposition in cerebral vessels. Parasites in cerebral vessels were mainly trophozoites or schizonts. Occasional RBC remnants following parasite release were seen. Some parasites were degenerate, resembling crisis forms. PRBCs adhered to endothelium via surface knobs. It is concluded that there is no evidence for an inflammatory or immune pathogenesis for human cerebral malaria and that the clinical effects probably relate to anoxia and the metabolic activities of the parasites.

Intravenous quinidine for the treatment of severe falciparum malaria. Clinical and pharmacokinetic studies

Quinidine has proved more effective than quinine against chloroquine-resistant Plasmodium falciparum both in vitro and in patients with uncomplicated disease. To examine the effectiveness and pharmacokinetics of quinidine for this use, we treated 14 patients who had severe falciparum malaria with intravenous quinidine gluconate; a loading dose of 15 mg of the base per kilogram of body weight was followed by 7.5 mg per kilogram every eight hours. Two of the five patients with cerebral malaria died, but parasitemia was eliminated in the 12 survivors. Two patients had recurrent parasitemia on Days 25 and 28. Times required for parasite clearance and elimination of fever (49.4 +/- 17.8 and 69.5 +/- 18.7 hours, respectively) were comparable to those in earlier studies with a loading dose of quinine. Quinidine appears to have a larger volume of distribution than quinine. The elimination half-life was 12.8 hours, the volume of distribution was 1.68 liters per kilogram, total clearance was 1.75 ml per kilogram per minute, and urinary clearance was 0.62 ml per kilogram per minute. Electrocardiographic changes were common but there were no dysrhythmias. In two patients, blood pressure fell during the initial infusion of quinidine. Quinidine gluconate is more widely available than quinine in many countries, and our findings show that it is effective in severe falciparum malaria.

Clinical pharmacokinetics of antimalarial drugs

For the past 300 years antimalarial dosage regimens have not been based on pharmacokinetic information. However, now that this information is available, it is appropriate to examine current recommendations for prophylaxis and treatment. In healthy subjects, the cinchona alkaloids (quinine and quinidine), primaquine and proguanil (chloroguanide) are all rapidly eliminated with half-lives (t1/2 beta) of between 6 and 12 hours. Hepatic biotransformation accounts for approximately 80, 96 and 50% of their total clearance, respectively. In malaria, the pharmacokinetic properties of quinine and quinidine are significantly altered with a decrease in the apparent volume of distribution (Vd), prolongation of the elimination half-life, and a reduction in systemic clearance (CL) that is proportional to the severity of infection. Red cell concentrations and plasma protein binding are both increased in severe disease. Parenteral quinine or quinidine should be given by slow intravenous infusion rather than by intravenous or intramuscular injection, and a loading dose is necessary in severe infections. Chloroquine (t1/2 beta 6 to 50 days) and mefloquine (t1/2 beta 6.5 to 33 days) have extensive tissue distribution and prolonged activity after a single dose. Both drugs are concentrated in erythrocytes and 55% of chloroquine and 98% of mefloquine in plasma is bound to protein. The pharmacokinetics of chloroquine are complex and, because of the extremely long beta phase, difficult to accurately define. Pyrimethamine (t1/2 35 to 175 hours) has more limited tissue distribution, plasma and erythrocyte concentrations are similar, and 85% of the drug in plasma is bound to plasma proteins. The clearance of quinine, mefloquine and pyrimethamine appears to be higher in children than in adults. Currently, most of the information available on disposition of antimalarial drugs in humans is derived from studies in healthy adult subjects. More information is required on their pharmacokinetics in malaria, pregnancy, and in young children.

Pathophysiological and prognostic significance of cerebrospinal-fluid lactate in cerebral malaria

Cerebrospinal-fluid (CSF) lactate concentrations were elevated in all but 1 of 45 patients with cerebral malaria. They were significantly higher in patients who died (9.0 +/- 5.3 mmol/l, mean +/- SD) than in survivors (3.4 +/- 1.1 mmol/l, p = 0.0002) and had returned to normal values in each of 9 patients studied after recovery of consciousness. There was a significant negative correlation between CSF lactate and CSF glucose. All 11 patients with CSF lactate concentrations above 6 mmol/l died. CSF lactate is thus an important prognostic indicator in cerebral malaria and these findings suggest that hypoxia contributes to the pathogenesis of this disorder.

Glucose and insulin homeostasis during the Jarisch-Herxheimer reaction

Plasma concentrations of glucose and insulin were measured in ten patients during the Jarisch-Herxheimer reaction of tetracycline-treated louse-borne relapsing fever. Plasma glucose fell significantly in eight of the ten patients associated with the peak of the reaction, but plasma insulin remained low. Glucoregulation by insulin was therefore normal. This evidence questions the role of macrophage mediator-induced pancreatic insulin release in causing hypoglycaemia in borreliosis or bacterial endotoxicosis.

Altered expression of human monocyte Fc receptors in Plasmodium falciparum malaria

The state of activation of human peripheral blood monocytes was examined by using a rosette assay that detects changes in Fc receptor expression. Monocytes from patients with uncomplicated Plasmodium falciparum malaria showed a significant increase in the number of rosettes relative to healthy controls. In addition, the monocytes from these patients were tested for their ability to phagocytose Candida albicans, but this ability did not differ from that of normal individuals. Finally, the monocytes from patients with cerebral malaria were also tested for Fc receptor expression. In contrast to the results from uncomplicated cases, the activity of the monocytes from these patients was no different from that of controls. We concluded that uncomplicated P. falciparum malaria caused an increase in monocyte Fc receptor expression which did not occur in cerebral malaria and that this difference in activation may be important in the pathogenesis of cerebral malaria.

A method for the study of cation transport in vivo: effects of digoxin administration and of chronic renal failure on the disposition of an oral load of rubidium chloride

In order to study cation transport in vivo we have measured the changes in plasma and intra-erythrocytic rubidium concentrations following an oral load of rubidium chloride. The changes in plasma rubidium concentration are related to the distribution of rubidium to all the body tissues and the changes in intra-erythrocytic rubidium concentrations provide an example of rubidium uptake by one particular tissue. In eight healthy volunteers pretreatment with a loading dose of digoxin (20 micrograms/kg) enhanced the rise in plasma rubidium concentrations and attenuated the rise in intra-erythrocytic rubidium concentrations after the oral load of rubidium chloride. Ten patients with chronic renal failure, compared with a well-matched control group, were found to have changes similar to, but more marked than, those caused by digoxin, i.e. a much greater rise in plasma rubidium concentrations and a much smaller rise in intra-erythrocytic rubidium concentrations, after the oral load of rubidium chloride. These findings are consistent with wide-spread reduction in Na+, K+-ATPase activity in subjects who have taken a loading dose of digoxin and patients with chronic renal failure. They are, therefore, consistent with the findings of previous studies in vitro and show that it is possible to demonstrate changes in cation transport in vivo.

Failure of chloroquine-erythromycin and chloroquine-tetracycline combinations in treatment of chloroquine-resistant falciparum malaria in eastern Thailand

In Eastern Thailand the increasing resistance of Plasmodium falciparum to all available antimalarial drugs prompted a reassessment of chloroquine in combination with either erythromycin or tetracycline. Both combinations had produced encouraging results in vitro, in experimental animal models, or in preliminary clinical trials. In patients with uncomplicated falciparum malaria presenting with moderate parasitaemia the trial of chloroquine and tetracycline was abandoned after RIII resistance with clinical deterioration was encountered in 2 out of the first 5 patients studied. Of 16 patients treated with chloroquine+erythromycin, 11 showed resistance (RI-RIII). These regimens appear ineffective and potentially dangerous in Eastern Thailand.

Plasmodium yoelii: blood oxygen and brain function in the infected mouse

The carriage of oxygen by the blood and the in vivo response of the brain were investigated in mice infected with a lethal strain of Plasmodium yoelii. All mice with parasitaemia exceeding 70% were severely anaemic (Hb 3.5 +/- 1.8 g/dl; mean +/- 1 SD), acidotic (blood pH 7.04 +/- 0.06) and hypoglycaemic (blood glucose 0.6 +/- 0.76 mumol/ml). The oxyhaemoglobin dissociation curve (ODC) of blood from heavily infected mice was shifted right as compared to controls, but the increase in p50 was less than expected from the accompanying acidosis. The reduced shift right was due to a decrease in the 2,3-DPG/Hb ratio in infected animals (0.72 +/- 0.12, n = 17 vs 1.10 +/- 0.09, n = 12 in controls). Despite the severity of terminal infection, the cerebral pH and the relative steady-state concentrations of PCr, ATP and Pi measured in vivo by nuclear magnetic resonance (31P NMR) were normal. Alterations in brain energy status and pH cannot account for cerebral signs or death in this proposed mouse model of cerebral malaria.

Cardiovascular abnormalities in osteogenesis imperfecta

To investigate the prevalence of cardiac abnormalities in osteogenesis imperfecta we performed a clinical and echocardiographic study on 20 patients. One patient had aortic regurgitation, 13 had soft late apical systolic murmurs (without significant mitral regurgitation), and seven had systemic hypertension. In only one patient was there echocardiographic evidence of mitral valve prolapse. The aortic root diameter exceeded the normal range in six patients, two of whom were hypertensive. The overall mean value (+/- 1 S.D.) in the 16 patients without marked skeletal deformity was 2.02 +/- 0.33 cm/m2. The echocardiographic appearances suggested that the left-sided valve cusps were thin, and in 13 patients the aortic valve leaflets showed a coarse systolic flutter.