- Email: [email protected]
Quinine pharmacokinetics and toxicity in cerebral and uncomplicated falciparum malaria
REPORT ON THERAPY
Quinine Pharmacokinetics and Toxicity in Cerebral and Uncomplicated Falciparum Malaria
NICHOLAS
J. WHITE,
SORNCHAI
LOOAREESUWAN,
DAVID MARY
A.
J. WARRELL,
M.R.C. DANNAI
WARRELL,
MB.,
D.M., M.B.,
M.R.C.P. MD. F.R.C.P. M.R.C.P.,
Path. BUNNAG,
TRANAKCHIT
M.D.
HARINASUTA,
M.D.
Bangkok, Thailand
From the Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand, and the Tropical Medicine Unit, Nuffield Department of Clinical Medicine, University of Oxford, England. Supported by the Wellcome Trust of Great Britain as a part of the Wellcom&tahiiol University, Oxford Tropical Medicine Research Programme. Requests for reprints should be addressed to Dr. Nicholas J. White, Faculty of Tropical Medicine, 42016 Rajvithi Road, Bangkok 10400. Thailand. Manuscript accepted April 14, 1982.
564
Octeber 1992
Acute pharmacokinetics of intravenously infused quinine were studied in 25 patients with cerebral malaria and 13 with uncomplicated falciparum malaria. In patients with cerebral malaria receiving the standard dose of 10 mg/kg every eight hours, plasma quinine concentrations consistently exceeded 10 mg/liter, reaching a peak 60 f 25 hours (mean f 1 SD.) after treatment was begun and then declining. Quinine total clearances (Cl) and total apparent volumes of distribution (Vd) were significantly lower than in uncomplicated malaria (Cl, 0.92 f 0.42 compared with 1.35 f 0.6 ml/min/kg, p = 0.03; Vd, 1.18 f 0.37 compared with 1.67 f 0.34 liter/kg, p = 0.0013). There was no significant difference between the two groups in elimination half-times (t/2) or renal clearances (C,) (t/2, 18.2 f 9.7 compared with 16 f 7.0 hours; C,, 0.21 f 0.16 compared with 0.21 f 0.08 ml/min/kg). In nine patients studied following recovery, Cl (3.09 f 1.18 ml/min), Vd (2.74 f 0.47 liter/kg), and C, (0.53 f 0.22 ml/min/kg) were significantly greater (p 50.0004), and t/2 was significantly shorter (11.1 f 4.1 hours, p = 0.006) than during the acute illness. C, accounted for approximately 20 percent of Cl in all groups. Renal failure did not alter the disposition kinetics in cerebral malaria. There was no clinical or electrocardiographic evidence of cardiotoxicity and no permanent neurotoxicity. Quinine toxicity in cerebral malaria has probably been overemphasized. The benefits of high plasma concentrations in the acute phase of this life-threatening disease appear to outweigh the risks, particularly in view of the increasing resistance of Plasmodium falciparum to quinine in Southeast Asia. Chloroquine-resistant falciparum malaria is a major health problem in Southeast Asia and Central and South America. Quinine is the only consistently effective drug for severe chloroquinine-resistant falciparum malaria [l] and is the mainstay of currently available treatment at all grades of severity [2]. Despite 500 years of continuous use in the treatment of severe falciparum malaria, the optimal dose and frequency of administration remain uncertain. Information about the pharmacokinetics and toxicity of quinine is inadequate in the case of uncomplicated malaria and nonexistent in severe forms such as cerebral malaria. The only previous study in which serial serum concentrations of quinine were measured in acute falciparum malaria was a comparison of five different treatment regimens in 35 conscious patients. Elimination half-times were estimated from two or three serum levels, but no other pharmacokinetic parameters were calculated [3]. In the present study, the pharmacokinetics and toxicity of quinine have been investigated in both cerebral and uncomplicated falciparum malaria.
The Amerlcsn Journal of Medlclne
Volume 73
QUININE
TABLE I
IN FALCIPARUM
MALARIA-WHITE
ET AL
Hematologlc, Blochemlcal, anti Clhical Fbtdlngs Uncomplicatsrl FakfparumMalaria Mean S.E.M.
CersbratMalaria Mean S.E.M. Hematocrit (%) White blood cell count/@’ Median platelet count/@ Blood urea nitrogen (mg/dl)’ Serum creatinine (mg/dl)’ Total bilirubin (mg/dl)* Direct bilirubin (mg/dl)* Albumin (g/liter)* Aspartate aminotransferase (Reitman-Frankel units/ml) Alanine aminotransferase (Reitman-Frankel units/ml) Prothrombin time (% of control) Partial thromboplastin time (% of control) Parasite clearance time (hours)’ Fever clearance time (hours)* Time to regain consciousness (hours)
29
1.9
11.088 143,000 33.2 2.57 2.85 1.84 30.9 87.3
1,020 .6:9 0.43 0.79 0.65 1.3 16.6
32.8 6,405 184,000 15.1 1.05 0.5 0.22 36.4 34.7
24 0.09 0.07 0.03 1.5 1.8
2.1 511
31.0
6.7
19.7
1.3
85
2.9
88
3.4
87
3.2
88
3.9
10
55
6.8
122
16.8
28.4
6.8
56
5.7
.
86
S.E.M. = standard error of the mean. p <0.05. l
PATIENTS AND METHODS All patients with cerebral malaria were studied in the intensive care unit of Pra Pokklao Hospital, Chantaburi Province, Eastern Thailand. Cerebral malaria was diagnosed when a patient, whose blood smear revealed asexual forms of P. falciparum, was unrousable and had no other demonstrable cause of unconsciousness. Patients were not studied If they had had a convulsion within six hours, were drowsy or stdporose (i.e., still reusable), or had a history of recent quinine treatment. Adult male patients with uncomplicated fafciparum malaria of moderate severity were studied in the Hospital for Tropical Diseases, Bangkok. All these patients gave informed consent to repeated blood sampling and restudy one month later. Patients were excluded if subsequent investigations revealed an alternative diagnosis (e.g., hepatitis, meningitis, encephalitis). Pharmacokinetic studies were performed In 38, patients: 22 adults and three chikken with cerebral malaria and 13 adult males with uncomplicated falciparum malaria. The mean ages of the adults in both groups were comparable (24.8 and 27 years) as were mean body weights (48.1 and 53.2 kg). The children were aged six, seven, and eight years. The parasite counts of patients with cerebral malaria were significantly hii than those of patiints with uncomplicated falclparum malaria (mean 47,559 compared with 5,690 parasitized cells/cLI p = 0.02. Logarittwnictransformt;$n was required to normalize the distributions). Hematologic and biochemical details are given in Table I together with the details of the course of the illness. Admission temperatures were not significantly different (38.8’ and 38.7’C); however,
October
in other respects, patients wlth cerebral malaria were more seriously ill and four (15 perce&):died. Following admission and confirmation of the diagnosis, each patient was weighed and treatment was begrm as soon as possibfe. Baseline investigations included parasite count, full blood study, platelet count, serum electrolytes, blood urea nitrogen, creatlnine, glucose, albumin, globulin, aspartate and alanine amlnotransferases, alkaline phosphatase, total and direct bilirubin, and prothrombin time. A 12-lead electrocardiogram was recorded at 25 mm/set and 50 mm/set before and after quinine infusion. Further recordings were made at horrly intervals in the first elght patients with cerebral malaria and the patients with uncoinpikzatsd malaria. A final recording was made after the fourth dose on the second day of treatment. A full clinical examination including a detailed neurologic assessment was parformed daily, and the clinical findings were recorded on standamlzed forms. Patients with cerebral malaria underwent catheterization, and urine was tested for sulphonamides (fiiln test) and quinolines (Wilson Edeson test). Parasite counts were made every eight hours for 48 hours, then twice daily until negative and then daily. Full blood cell counts and routine biochemistry studies including liver function results were obtained daily. Temperawe, pulse, and blood presswe ware measured at feast every four hours, but more frequently in seriously ill patients. Quinine dihydrochlorkfe (Government Pharmaceutical Organisation of Thailand) was given in a dose of 10 mg/kg (equivalent to 8.3 mg/kg base) infused over four hours at eight-hour intervals. In two patients with cerebral malaria, infusions were given over two hours. The drug was dispensed
1952
The American
Journal
al Medktne
votuma
73
565
QUININE
IN FALCIPARUM
MALARIA-WHITE
ET AL.
Quinine Pharmacokinetics: Acute Faiciparum Malaria
TABLE II
Cerebral Malaria
Uncomplicated Malaria Number of Standard Deviation Patients Mean
Number of Patients
Mean
Standard Deviation
18
0.92
0.42
11
1.35
0.80
0.031
18 25
1.18 18.2
0.37 9.7
11 13
1.67 16.0
0.34 7.0
0.0013 N.S.
15
0.21
0.18
11
0.21
0.08
N.S.
15
0.20
0.18
11
0.21
0.12
N.S.
Total clearance (ml/min/kg) Vd (liter/kg) Elimination t/2 (hours) Renal clearance (ml/min/kg) Renal clearance Total
Significance (p value)
Vd = total apparent
volume of distribution; N.S. = not significant at 5 percent level. Seven patients with cerebral malaria and two patients with uncomplicated malaria had baseline plasma samples that revealed quinine, indicating oretreatment. These oatients are excluded from the calculations of Vd and total clearance. Accurate urine volumes kere
not obtained
from
10 patients.
with a 1 ml plastic syringe. All patients were given quinine in 0.9 percent saline intravenously, via a Braun perfusor pump to ensure a constant rate of infusion. intravenous fluids were given as required, through the same cannula, and blood samples were taken from the opposite arm. Intramuscular antipyretics (dipyrone, 500 mg) and anti-emetics (metoclopramide, 10 mg) were administered if necessary, but no other drugs were given in the first 12 hours. Lumbar puncture was performed soon after admission and blood sampled simultaneously. Thereafter, intravenous quinine dihydrochloride was given every eight hours until the patient regained consciousness, Quinine sulphate tablets were then given in a dose as near as possible to the intravenous dose to complete seven days’ treatment. Patients with uncomplicated malaria received oral treatment after the first quinine infusion. Blood was taken before treatment, at 0, 15, 30, and 60 minutes, then two, three, four, five, and six hours after the end of the infusion. Further samples were taken from patients with cerebral malaria before and after the next three quinine infusions. Thereafter,‘samples were taken daily before and after the first dose of the day. Blood was collected in lithium-heparin bottles, and plasma was separated immediately and stored at -2OOC. Urine was collected from the indwelling catheter into a volumetric flask and stored similarly. Quinine was measured by the benzene extraction method [4] using
TABLE III
an Aminco Bowman spectrophotofiuorimeter. inter-assay variation between standards did not exceed 3 percent. Three patients who had recovered from cerebral malaria were restudied when they were readmitted because of fever. Blood smears revealed no organisms, and urinary tract infection was finally diagnosed in each case. Patients with uncomplicated falciparum malaria were restud/ed one month after the acute infection when clinically cured ahd subjectively well. Plasma concentrations fitted closely a single exponential decline following the four-hour infusion._ Tne elimination half-time was therefore calculated from the semilogarithmic plot of plasma quinine against time. The total apparent volume of distribution (Vd) was calculated ffom the equation Vd = (Ri”r[ 1 - eKt])/KC where Rrnris the rate of infusion (total dose of quinine base per kilogram body weight divided by time), K is the elimination constant, and G ihe plasma level, extrapolated from the linear plot, at the end of the constant infusion [5]. This estimate of Vd is based on a one-compartment model and assumes that the distribution phase is no longer materially affecting the measured and derived parameters. When plasma concentrations had reached a high point on the second or third day of treatment, Vd was recalculated from the formula Vd = D/PK where D is the dose divided by dose interval and P the mean plasma concentration
Quinine Pharmacokinetlcs: Uncomplicated Falciparum Malaria Acute
Total clearance (ml/min/kg) Vd (liter/kg) Elimination t/2 (hours) Renal clearance (ml/min/kg)
Number of Patients
Mean
8
1.28
8 9 9
Convalescent Standard Deviation
Mean
Standard Deviation
0.63
3.09
1.18
0.0002
1.62 17.0
0.36 7.0
2.74 11.1
0.47 4.1
0.0002 0.006
0.20
0.08
0.53
0.20
0.0004
N.S. = not significant.
586
October 1982
The American Journal of Medicine Volume 73
Significance (p value)
QUININE
TABLE IV
MALARIA-WHITE
ET AL
Quinine Pharmacokinetics in Three Patients Readmitted with Fever due to Urinary Tract infection 112 (hours)
Cerebral malaria Fever readmission
IN FALCIPARUM
Totalclearance (ml/min/ka)
Vd /litsr/ka)
9.1
20.5
11.9
1.38
1.512
0.79
1.75
0.85
0.77
9.7
8.3
5.2
1.79
1.12
1.26
2.12
1.55
2.78
t/2 = elimination half-time; Vd = total apparent volume of distribution.
at steady state. Total clearance was calculated as the product of the acute Vd and K values. Renal clearance of unchanged quinine was calculated from the ratio of urine quinine excreted over the eight hours following the infusion, to the plasma concentration at the midpoint of the collection. Creatinine clearance was measured from the same samples. In order to assess the validity of estimating elimination halftime in acute maiarii from serial plasma concentrations taken over a relatively short time (i.e., less than one half-time), pIas& concentrations were measured in five patients after the last dose of quinine and followed for 30 hours. The two-tailed Student t test was used to compare the variables of the groups with cerebral malaria and uncomplicated malaria. Paired t tests were used to compare the same patients’ variables in the acute and convalescent phase. Data not conforming to a normal distribution were compared by the Wilcoxon rank sum test, and rank correlation was calculated by the method of Kendall.
RESULTS Pharmacokinetlcs. In patients with acute uncomplicated falciparum malaria, total clearance of quinine was significantly lower during the acute illness than after recovery, and was even lower in the patients with cerebral malaria (Tables II and ill). The total apparent volume of distribution was similarly reduced, being lowest in the patients with cerebral malaria. There was a wide variation in elimination half-times in both groups of patients with acute malaria, and there was no significant difference between the two groups. Elimination half-times of the five patients with cerebral malaria at the end of the treatment course (12.8 f 4.1 hours, mean f 1 S.D.) were not significantly different from the acute values (15 f 4.9 hours, p = 0.37). There was, however, a significant difference between patients in the acute phase of malaria and the same patients studied one month after recovery. In 12 patients with cerebral malaria, Vd values were estimated from the peak plasma concentrations (second or third day of treatment) and were significantly larger than the acute phase values (1.66 f 6.52 compared with 1.25 f 0.34 liters/kg, mean f 1 S.D., p = 0.04). Plasma concentrations of quinine were higher in cerebral malaria than in uncomplicated malaria, consistently exceeding 10 mg/liter, and in 60 percent of patients exceeding 15
October
mg/liter. There was no correlation between these pharmacokinetic parameters and fever, depth of coma, length of unconsciousness, parasite clearance rate, and biochemical and hematologic variables (Table I). Three patients with cerebral malaria who were later readmitted with fever were treated with quinine before the final diagnosis of urinary tract infection was certain. Total clearances had improved but were still abnormally low (Table IV). Renal Clearance. Renal clearance and the proportion of renal to total clearance were not significantly different in patients with cerebral and non-cerebral malaria. There was, however, a significant increase in renal clearance on recovery from malaria. Total clearance and urinary clearance were roughly parallel, as the ratio of urinary to total cieaiance did not significantly change between any of the groups. Creatinine clearance and urinary quinine clearance correlated in patients with cerebral malaria (r = 0.67, p = 0.009) and in convalescent patients (r = 0.70, p = 0.008), but not in patients with uncomplicated malaria. in this last group, creatinine clearance was usually normal, but there was a wide variation in urinary quinine clearance. There was no correlation between rate of urine flow and quinine clearance in any of the groups. Serum creatinine exceeded 3 mg/lOO ml in six patients with cerebral malaria. There was no significant difference in total clearances, Vd values, or elimination half-times between these and the remainder of the patients with cerebral malaria. However the difference in nonrenal clearance (1.10 f 0.53 and 0.64 f 0.31 mi/min/kg, mean f 1 S.D., p = 0.051) approached significance. Cerebrospinai Fluid Concentration (Table V). Cerebrospinai fluid quinine concentration varied between 0.2 and 1.O mg/liter and did not correlate with plasma quinine concentration. The mean cerebrospinai fluid/ plasma quinine ratio was 0.07 (S.D. 0.03). This ratio was inversely correlated with plasma quinine concentration (r = 0.59, p = 0.001). Cerebral Malaria: Fatal Cases. These four patients were characterized by high parasitemia (all had more than 10 percent of red blood ceils parasitized, mean count 670,437 parasitized cells/@) and severe bio-
1982
The American Journal of Mediche
Volume 73
567
QUININE IN FALCIPARUM
TABLE V
ET AL.
Cerebrospinal Fluid Quinine Concentrations in Cerebral Malaria
C.&F. Quinine
Plasma Quinine
0.8 0.8 0.8 0.4 0.8 0.5 0.5 0.9 0.8 0.2 0.7 0.7 1.0 1.0 0.5 0.9 0.8 1.0 All patients
MALARIA-WHITE
s
Quinine
10.3 11.8 9.3 11.0 10.7 9.8 15.0 11.6 17.4 10.8 6.9 18.6 15.0 16.7 7.7 6.9 9.2 9.8 were unconscious
Plasma Concentration during Seven Days’ Treatment.
0.08 0.05 0.09 0.04 0.06 0.05 0.03 0.08 0.05 0.02 0.10 0.04 0.06 0.06 0.07 0.13 0.07 0.10 at the time of lumbar
puncture.
chemical derangement. The mean blood urea nitrogen level was 62.2 mg/dl, mean serum creatinine level was 5.6 mg/dl, mean total bilirubin level was 16.1 mg/dl, and mean serum aspartate aminoiransferase level was 128 Reitman-Frankel units/dl. The cause of death was multifactorial; in two patients, pulmonary edema developed, in association with metabolic acidosis and acute renal failure in both; in one patient, septicemia and pneumonia developed; in the other, the immediate cause of death was uncertain. The time from the beginning of intravenous quinine to death, the plasma quinine level immediately before death, and the pharmacokinetic parameters in these patients are given in Table VI. The elimination half-times were significantly shorter than those of the remaining patients (p = 0.013). One patient had previously received oral quinine, and quinine was detectable in the baseline blood sample. Total clearance and Vd values in the remaining three patients were not significantly different from those in the surviving patients. The rise in plasma quinine concentration and delayed fall in parasite count in Patient 3 are shown in Figure 1. TABLE VI
566
October
Quinine Pharmacokinetics: Fatal Cases
Palienl Number
Elimination t/2 (hours)
1 2 3 4
16.1 12.5 7.5 13.7
1962
Plasma concentrations reached a peak on the third day of treatment (60 f 24.6 hours, mean f 1 S.D.) and then declined to plateau at plasma levels approximately 65 percent of the peak plasma level, although there was a wide variation. The decline in plasma levels coincided with clinical improvement and also the change from intravenous to oral medication. The bioavailability of quinine in these tablets (quinine sulphate) was not investigated. Fatal Case: Previous Treatment. Seven of the patients with cerebral malaria and two with uncomplicated malaria had quinine in the baseline blood sample. In a further five patients sulphonamide was detected in the urine by the lignin test. Two patients had a positive result on Wilson Edeson urine testing, but their baseline blood contained no quinine. They had presumably previously received chloroquine. Dosage Accuracy. The concentration of quinine in the intravenous preparation was assayed and found to be within 3 percent of the quoted value in five random samples. Dispensing accuracy using a 1 ml syringe was f2 percent. Toxicity. There was no evidence of cardiovascular toxicity in either cerebral or uncomplicated malaria. There was no relationship between mean arterial pressure and plasma quinine concentration. Changes in pulse or blood pressure could not be related to the administration of quinine or to peak plasma concentrations. No dysrythmias were observed. Prolongation of the rate-corrected electrocardiographic Q-T interval (Q-T/d/R - R) by more than 5 percent of the baseline value was observed in 65 percent of the patients with cerebral malaria. The mean maximal lengthening f 1 S.D. of the Q-Tc interval of all the patients was 10.3 f 7.7 percent. The longest recorded Cl-Tc interval was 0.537 seconds in a patient with a plasma concentration of 9.3 mg/liter. Overall, there was no correlation between the change in plasma concentration and the prolongation of the Q-Tc interval. The most consistent electrocardiographic abnormality in all patients was a change in T-wave morphology, which usually became flatter. Slight QRS widening (not exceeding 0.03 seconds) was seen in 11 patients with cerebral malaria, but no other changes were noted.
The American
(IitZkg)
Total clearance (mllminlkg)
1.06 1.15 1.41
0.76 1.06 2.16
Journal
of Medicine
Plasma Quinine Level before Death (mglliter)
Time from Start of Quinine Therapy lo Death (hours)
9.1 11.2 16.0 13.1
22 24 76 44
Volume 73
QUININE IN FALCIPARUM
MALARIA--WHITE
1:
ET AL
IO4 PARASITE COUNT/,,
PLASMA OUININE mg/l
103 1 O2
IO
Figure 1. Response of parasitemia (number of parasitized red cells/p I) to rise in plasma quinine concentration in a fatal case of cerebral malaria (Patient 3).
24
There was no evidence of permanent neurotoxicity; daily examination of the eye revealed no evidence of retinal vasoconstriction and no abnormalities in the light reflex. In patients with cerebral malaria, no II or VIII nerve toxicity was clinically detectable when they regained consciousness, and all were completely normal on discharge from hospital. In one patient, a six year old boy, neuromuscular weakness developed on the third day following admission. This was partially reversed by edrophonium and neostigmine. The electrocardiographic results were normal, and the plasma quinine concentration was 10.2 mg/liter. This may have been quinine-induced myasthenia. Vomiting was not associated with high cerebrospinal fluid concentrations of quinine nor temporally with the highest plasma quinine concentrations. The majority of the patients who vomited did so within the first 24 hours of admission. Eight of the eleven conscious patients had tinnitus when blood levels exceeded 5 mg/liter, two patients complained of nausea and vomited within six hours of admission when their temperature exceeded 39OC. Six of the eleven patients experienced improvement in’ headache. Tinnitus was usually mild or absent on recovery from cerebral malaria, and these patients usually had no complaints other than slight anorexia. Thrombocytopenia is common in cerebral malaria; 25 percent of the patients with cerebral malaria had platelet counts of less than 6O,OOO/~l with a nadir on the third or fourth day following admission. This was attributed to the disease rather than quinine. COMMENTS
Cerebral malaria is a severe manifestation of falciparum malaria, with a treated mortality of up to 50 percent
HOURS
48
72
[6,7]. Parenteral antimalarials are the only treatment of proved value, and therefore it is essential that chemotherapy is optimized through a better understanding of the pharmacokinetics and therapeutic modes of action of the currently available drugs. Recovery of the patient can come only through death of the parasites. There is no convincing evidence that rapid killing of parasites is harmful, so drug treatment should be designed to exceed the parasites’ minimal inhibitory concentration as soon as possible and thereafter maintain concentrations within the therapeutic ratio for the remainder of the treatment course. These same principles are fundamental to the treatment of bacterial infections. In a recent study of 100 cases of cerebral malaria, it was found that patients with fatal disease began receiving quinine significantly later in the course of their illness than survivors [ 71. Quinine clearance was reduced in acute falciparum malaria, and was significantly lower in cerebral malaria than in uncomplicated cases. Thus the reduction in quinine clearance was related to the severity of the disease, and plasma levels of quinine were therefore higher in cerebral malaria. There was a consistent contraction in Vd and variable prolongation of the elimination half-time. The contraction in Vd may be related to several factors such as dehydration, obstruction of the capillary bed by parasitized red cells, and alterations in tissue binding. None of the patients in this study was clinically hypovolemic, but minor degrees of dehydration cannot be excluded. Three patients with cerebral malaria who were readmitted and found to have urinary tract infections also had low quinine clearances and Vd values. Perhaps fever was a common factor contributing to these changes in both bac-
October 1982
The American Journal of Medlclne
Volume 73
569
QUININE IN FALCIPARUM MALARIA-WHITE
ET AL.
terial and malarial infections, although it cannot explain the differences between patients with cerebral and uncomplicated malaria, as temperatures were not significantly different between the two groups. Etiocholanolone fever has also been shown to increase plasma concentrations of quinine [8]. The relative importance of fever and these other factors cannot be assessed from available evidence. Because of the potential danger of interrupted treatment in the acute phase of this severe infection, it was not possible to delay the second dose of quinine in order to collect more blood samples and to obtain a more satisfactory estimate of elimination half-time. However, when the decline in plasma concentration was followed for 30 hours at the end of the treatment course, and in convalescent patients, it always continued monoexponentially. The absence of a measurable distribution phase, and the relatively long elimination half-times make it unlikely that there are large errors in the calculation of Vd values [9]. This suggests that a one-compartment model is adequate to describe the elimination kinetics following a four-hour intravenous infusion of quinine, and that calculation of the elimination half-time from plasma concentrations over a six-hour period gives a reasonable estimate of the true value. The major route of quinine elimination is via liver metabolism [lo]. Renal clearance of unchanged quinine accounted for approximately 20 percent of the total clearance in both uncomplicated and cerebral malaria. Renal and total clearances were reduced in parallel. The urinary quinine clearance was not correlated with rate of urine flow; however, in cerebral malaria and after recovery from uncomplicated malaria, it was significantly correlated with creatinine clearance. At low creatinine clearances, the ratio of urinary quinine clearance to creatinine clearance increased-in some cases this ratio exceeded unity. This suggests that quinine is filtered and secreted by the normal kidney, with secretion becoming relatively more important at low glomerular filtration rates. It is widely recommended [ 1 l-131 that the dose of quinine should be reduced if there is renal impairment. There was, however, no difference in total clearance or Vd values between patients with cerebral malaria who had increased serum creatinine levels (>3 mg/ 100 ml) and the remainder of this group; indeed, nonrenal clearance was greater in the former group, although this was not significant at the 5 percent level (p = 0.051). Quinine dosage in renal failure requires further study. Liver dysfunction in malaria has been widely reported, but is difficult to measure. None of the conventional indexes of liver function correlated with total clearance or prolongation of the quinine elimination half-time.
570
October 1982
The American Journal of Medicine
Trenholme et al. [8] have measured the elimination half-time of quinine at the end of a treatment course for mild artificially induced malaria. They compared the benzene extraction method, which measures quinine alone, with a metaphosphoric acid method, which measures both quinine and its metabolites. They concluded that hepatic metabolism of quinine was reduced in malaria while renal excretion was not. Hepatic blood flow is reduced in experimental malaria [ 141, and there is sometimes histopathologic evidence of liver damage [ 151. Reduction in quinine clearance is probably related to impaired liver function, but this cannot be predicted from the results of biochemical investigations. Quinine plasma concentrations are high in the first two to four days of treatment [ 161, and then fall as the clinical condition improves. When quinine was given every eight hours, plasma concentrations rose to reach a peak 60 f 25 hours (mean f 1 S.D.) after treatment was begun. True steady-state values were never attained in the early phase of treatment because clearance and Vd values were constantly changing; however, at peak plasma concentrations, drug administered must equal drug eliminated, albeit for a short time. In cerebral malaria, estimated Vd values at this stage were 33 percent larger than those measured at the beginning of treatment. Thereafter, levels declined to plateau around 65 percent of the peak value. The fall in plasma quinine levels in the second half of the treatment course may result not only from an increase in total clearance and expansion of the Vd, but also the substitution of oral for intravenous quinine, with probable reduction in bioavailability for the same dose. The development of quinine resistance in Thailand has entailed an increase in the length of treatment and total dosage required to cure P. falciparum malaria [ 171. The in vitro minimal inhibitory concentrations for some strains have now reached 10 mg/liter [ 18,191. This was previously considered a toxic plasma concentration, although in cerebral malaria, by the third day of treatment, plasma concentrations invariably exceed this value. Death from cerebral malaria usually occurs within the first 48 hours [7]; thus, although high plasma concentrations of quinine occur, they are reached at a time when the patient is usually improving. Conversely, inadequate plasma concentrations may be obtained on the critical first day of treatment. The slow and delayed fall in parasite count in the longest surviving patient who died, associated with rapid clearance and consequent slow rise in plasma quinine concentrations, clearly illustrates this point (Figure 1). If the dose of quinine were reduced in cerebral malaria as has been advocated [ 1,2], therapeutic plasma concentrations would take even longer to be achieved. Despite plasma levels of quinine exceeding 15
Volume 73
QUININE IN FALCIPARUM
mg/liter, and in some cases 20 mg/liter, there was no evidence of cardiotoxicity. Quinine-induced electrocardiographic changes were small. It is interesting that the “quinidine effect” of marked electrocardiographic D-T prolongation is much less pronounced with quinine the diastereomer, despite comparatively high plasma concentrations, Apart from one patient who may have had a transient quinine-related myasthenic syndrome, there were no neurologic sequelae. In particular, there was no deafness or blindness. Daily ophthalmoscopy failed to demonstrate the constriction of the retinal vessels seen in quinine poisoning. Clearly, unconscious patients are unable to complain of “cinchonism,” but this symptom complex was usually mild or absent when they recovered consciousness, and was not a problem in the patients with uncomplicated malaria or convalescent patients. Vomiting was relatively frequent during the acute illness, particularly in patients with high fever. This was not related to the cerebrospinal fluid or plasma quinine concentration. Upper gastrointestinal side effects may be related to the introduction of quinine and consequent rise in plasma concentration. There have been several studies relating plasma levels to side effects. Powell and McNamara [20] found that minor side effects were common with quinine treatment, occurring most frequently when plasma concentrations exceeded 10 mg/liter. Plasma concentrations exceeding this value were also associated with minor toxicity by Taggart et al. [ 2 11. In the past, serious cardiovascular side effects were associated with the practice of giving intravenous injections of quinine over a short time [22]. This is still advocated in some textbooks of tropical medicine [23-251, and the side effects are probably due to
MALARIA-WHITE
ET AL.
transient high concentrations of unbound drug early in the distribution phase causing cardiovascular toxicity. Slow infusion over two or four hours appears to prevent these adverse effects. Serious neurologic side effects have been reported occasionally [ 31. Some of the ones reported (e.g., focal hemisphere deficits) are extremely unusual in quinine self-poisoning and may well have resulted from malaria rather than drug toxicity. Recommendations on quinine dosage that have been limited by observations of minor toxicity [2,8,20] are not appropriate in severely ill patients who require urgent treatment. Furthermore, although the major alterations in quinine pharmacokinetics in severely ill patients result in high plasma levels, these do not appear to produce serious toxicity and may be life-saving. This is particularly important in Thailand, where the minimal inhibitory concentrations of quinine for P. falciparum continue to rise. In cerebral malaria, distribution of quinine throughout the capillary bed, particularly in the brain, may be inadequate because of obstruction by parasitized red cells. It is in these regions that parasiticidal concentrations are most needed. There is a definite risk of inadequate treatment in cerebral malaria if the initial doses of quinine are reduced. ACKNOWLEDGMENT We are grateful to the Director of Pra Pokklao Hospital, Dr. Chaisit Dharakul and his staff, and Dr. Damrong Bhanthumkosol and his laboratory staff, and especially the nursing staff of the Intensive Care Unit. We thank the staff and nurses of the Bangkok Hospita! for Tropical Diseases, Mrs. Patchari Prakongpan, and Miss Kamolrat Lavansiri for their help throughout this study.
REFERENCES 1. 2. 3.
a. 9.
Hall AP: The treatment of malaria. Br Med J 1976; 1: 32332a. Hall AP: The treatment of severe falciparum malaria. Trans R Sot Trop Med Hyg 1977; 71: 367-379. Hall AP, Hanchalay S, Doberstyn EB, Bumnetphund S: Quinine dosage and serum levels in falciparum malaria. Annu Rep SEATO Med Res Labs 1975: 241-250. Cramer G, lsaksson B: Quantitative determination of quinidine in plasma. Stand J Clin Lab Invest 1963; 15: 553-556. Goldstein A, Aronow L, Kalman SM: Principles of drug action, 2nd ed. New York: John Wiley, 1974; 314. Daroff RB, Deller JJ, Kastl AJ, Blocker WW: Cerebral malaria. JAMA 1967; 202: 679-662. Warrell DA, Looareesuwan S, Warrell MJ. et al: Dexamethasone proves deleterious in cerebral malaria: double blind trial in 100 patients. N Engl J Med 1962; 306: 313-319. Trenholme GM, Williams RL, Rieckmann KH, Frischer H, Carson PE: Quinine disposition during malaria and during induced fever. Clin Pharmacol Ther 1976; 19: 459-467. Riegelman S, Loo J, Rowland M: Concept of a volume of distribution and possible errors in evaluation of this parameter.
October
10.
il.
12.
13.
14.
15.
1962
J Pharm Sci 1966: 57: 126-133. Brodie BB, Baer JE, Craig LC: Metabolic products of the cinchona alkaloids in human urine. J Biol Chem 1951; 166: 567-581. Miller LH: Malaria. In: Hoeprich PD, ed. Infectious diseases. 2nd ed. Hagerstown, Maryland: Harper & Row, 1977; 1086. Donadio JV, Whelton A, Kazyak L: Quinine therapy and peritoneal dialysis in acute renal failure complicating malarial haemoglobinuria. Lancet 1966; I: 375-379. Canfield Cl, Miller LH, Bastelloni PJ, Eichler P, Barry KG: Acute renal failure in P&wnodium fdcipatum malaria. Arch Intern Med 1966; 122: 199-203. Skirrow MR, Chongsuphajaisiddhi T, Maegraith BG: The circulation in malaria. II. Portal angiography in monkeys (Macaca mulaffa) with P. know/&and in shock following manipulation of the gut. Ann Trop Med Parasitol 1964; 56: 502-510. Aikawa M. Susuki M, Gutierrez Y: Pathology of malaria. In: Kreier JP. ad. Malaria, vol 2. New York: Academic Press, 1980; 47-102.
The American Journal of Medicine
Volunw 73
571
QUININE
16.
17.
18.
19. 20.
572
IN FALCIPARUM
MALARIA-WHITE
ET AL
Brooks MH, Malloy JP, Bartelloni PJ, Sheehy TW, Barry KG: Quinine, pyrimethanine, and sulphorthodimethoxine: clinical response, plasma levels, and urinary excretion during the initial attack of naturally acquired falciparum malaria. Clin Pharmacol Ther 1969; 10: 85-91. Bunnag D, Harinasuta T, Suntharasamai P, et al.: In vivo response of Plasmodium falciparum to different regimes of antimalarial drugs. Mahidol University Annual Research Abstracts, 1980; 362. Chongsuphajaisiddhi T, Subcharoen A, Attanath P: In viva and in vitro sensitivity of falciparum malaria to quinine in Thai children. Ann Trop Paediatr 1981; 1: 21-26. Thaithong S: Personal communication. Powell RD. McNamara JV: Quinine side effects and plasma
October 1992
The American Journal of Medicine
21.
22.
23. 24. 25.
Volume 73
levels. Proc Helminth Sot Wash 1972; 39: 331-338. Taggart JV, Earle DP, Berliner RW, et al.: Studies on the chemotherapy of the human malarias. Ill. The physiological disposition and antimalarial activity of the cinchona alkaloids. J Clin Invest 1948; 27: 80-86. Strahan JH: Quinine by continuous intravenous drip in the treatment of acute falciparum malaria. Trans R Sot Trop Med Hyg 1948; 41: 669-676. Wilcocks C, Manson-Bahr PEC: Manson’s tropical diseases, 17th ed. London: Bailliere Tindall, 1972; 69. Adams ARD, Maegraith BG: Clinical tropical diseases, 6th ed. Oxford: Blackwell Scientific, 1980; 266. Woodruff AW: Medicine in the tropics. Edinburgh: Churchill Livingstone, 1974; 82.
Quinine Pharmacokinetics and Toxicity in Cerebral and Uncomplicated Falciparum Malaria
NICHOLAS
J. WHITE,
SORNCHAI
LOOAREESUWAN,
DAVID MARY
A.
J. WARRELL,
M.R.C. DANNAI
WARRELL,
MB.,
D.M., M.B.,
M.R.C.P. MD. F.R.C.P. M.R.C.P.,
Path. BUNNAG,
TRANAKCHIT
M.D.
HARINASUTA,
M.D.
Bangkok, Thailand
From the Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand, and the Tropical Medicine Unit, Nuffield Department of Clinical Medicine, University of Oxford, England. Supported by the Wellcome Trust of Great Britain as a part of the Wellcom&tahiiol University, Oxford Tropical Medicine Research Programme. Requests for reprints should be addressed to Dr. Nicholas J. White, Faculty of Tropical Medicine, 42016 Rajvithi Road, Bangkok 10400. Thailand. Manuscript accepted April 14, 1982.
564
Octeber 1992
Acute pharmacokinetics of intravenously infused quinine were studied in 25 patients with cerebral malaria and 13 with uncomplicated falciparum malaria. In patients with cerebral malaria receiving the standard dose of 10 mg/kg every eight hours, plasma quinine concentrations consistently exceeded 10 mg/liter, reaching a peak 60 f 25 hours (mean f 1 SD.) after treatment was begun and then declining. Quinine total clearances (Cl) and total apparent volumes of distribution (Vd) were significantly lower than in uncomplicated malaria (Cl, 0.92 f 0.42 compared with 1.35 f 0.6 ml/min/kg, p = 0.03; Vd, 1.18 f 0.37 compared with 1.67 f 0.34 liter/kg, p = 0.0013). There was no significant difference between the two groups in elimination half-times (t/2) or renal clearances (C,) (t/2, 18.2 f 9.7 compared with 16 f 7.0 hours; C,, 0.21 f 0.16 compared with 0.21 f 0.08 ml/min/kg). In nine patients studied following recovery, Cl (3.09 f 1.18 ml/min), Vd (2.74 f 0.47 liter/kg), and C, (0.53 f 0.22 ml/min/kg) were significantly greater (p 50.0004), and t/2 was significantly shorter (11.1 f 4.1 hours, p = 0.006) than during the acute illness. C, accounted for approximately 20 percent of Cl in all groups. Renal failure did not alter the disposition kinetics in cerebral malaria. There was no clinical or electrocardiographic evidence of cardiotoxicity and no permanent neurotoxicity. Quinine toxicity in cerebral malaria has probably been overemphasized. The benefits of high plasma concentrations in the acute phase of this life-threatening disease appear to outweigh the risks, particularly in view of the increasing resistance of Plasmodium falciparum to quinine in Southeast Asia. Chloroquine-resistant falciparum malaria is a major health problem in Southeast Asia and Central and South America. Quinine is the only consistently effective drug for severe chloroquinine-resistant falciparum malaria [l] and is the mainstay of currently available treatment at all grades of severity [2]. Despite 500 years of continuous use in the treatment of severe falciparum malaria, the optimal dose and frequency of administration remain uncertain. Information about the pharmacokinetics and toxicity of quinine is inadequate in the case of uncomplicated malaria and nonexistent in severe forms such as cerebral malaria. The only previous study in which serial serum concentrations of quinine were measured in acute falciparum malaria was a comparison of five different treatment regimens in 35 conscious patients. Elimination half-times were estimated from two or three serum levels, but no other pharmacokinetic parameters were calculated [3]. In the present study, the pharmacokinetics and toxicity of quinine have been investigated in both cerebral and uncomplicated falciparum malaria.
The Amerlcsn Journal of Medlclne
Volume 73
QUININE
TABLE I
IN FALCIPARUM
MALARIA-WHITE
ET AL
Hematologlc, Blochemlcal, anti Clhical Fbtdlngs Uncomplicatsrl FakfparumMalaria Mean S.E.M.
CersbratMalaria Mean S.E.M. Hematocrit (%) White blood cell count/@’ Median platelet count/@ Blood urea nitrogen (mg/dl)’ Serum creatinine (mg/dl)’ Total bilirubin (mg/dl)* Direct bilirubin (mg/dl)* Albumin (g/liter)* Aspartate aminotransferase (Reitman-Frankel units/ml) Alanine aminotransferase (Reitman-Frankel units/ml) Prothrombin time (% of control) Partial thromboplastin time (% of control) Parasite clearance time (hours)’ Fever clearance time (hours)* Time to regain consciousness (hours)
29
1.9
11.088 143,000 33.2 2.57 2.85 1.84 30.9 87.3
1,020 .6:9 0.43 0.79 0.65 1.3 16.6
32.8 6,405 184,000 15.1 1.05 0.5 0.22 36.4 34.7
24 0.09 0.07 0.03 1.5 1.8
2.1 511
31.0
6.7
19.7
1.3
85
2.9
88
3.4
87
3.2
88
3.9
10
55
6.8
122
16.8
28.4
6.8
56
5.7
.
86
S.E.M. = standard error of the mean. p <0.05. l
PATIENTS AND METHODS All patients with cerebral malaria were studied in the intensive care unit of Pra Pokklao Hospital, Chantaburi Province, Eastern Thailand. Cerebral malaria was diagnosed when a patient, whose blood smear revealed asexual forms of P. falciparum, was unrousable and had no other demonstrable cause of unconsciousness. Patients were not studied If they had had a convulsion within six hours, were drowsy or stdporose (i.e., still reusable), or had a history of recent quinine treatment. Adult male patients with uncomplicated fafciparum malaria of moderate severity were studied in the Hospital for Tropical Diseases, Bangkok. All these patients gave informed consent to repeated blood sampling and restudy one month later. Patients were excluded if subsequent investigations revealed an alternative diagnosis (e.g., hepatitis, meningitis, encephalitis). Pharmacokinetic studies were performed In 38, patients: 22 adults and three chikken with cerebral malaria and 13 adult males with uncomplicated falciparum malaria. The mean ages of the adults in both groups were comparable (24.8 and 27 years) as were mean body weights (48.1 and 53.2 kg). The children were aged six, seven, and eight years. The parasite counts of patients with cerebral malaria were significantly hii than those of patiints with uncomplicated falclparum malaria (mean 47,559 compared with 5,690 parasitized cells/cLI p = 0.02. Logarittwnictransformt;$n was required to normalize the distributions). Hematologic and biochemical details are given in Table I together with the details of the course of the illness. Admission temperatures were not significantly different (38.8’ and 38.7’C); however,
October
in other respects, patients wlth cerebral malaria were more seriously ill and four (15 perce&):died. Following admission and confirmation of the diagnosis, each patient was weighed and treatment was begrm as soon as possibfe. Baseline investigations included parasite count, full blood study, platelet count, serum electrolytes, blood urea nitrogen, creatlnine, glucose, albumin, globulin, aspartate and alanine amlnotransferases, alkaline phosphatase, total and direct bilirubin, and prothrombin time. A 12-lead electrocardiogram was recorded at 25 mm/set and 50 mm/set before and after quinine infusion. Further recordings were made at horrly intervals in the first elght patients with cerebral malaria and the patients with uncoinpikzatsd malaria. A final recording was made after the fourth dose on the second day of treatment. A full clinical examination including a detailed neurologic assessment was parformed daily, and the clinical findings were recorded on standamlzed forms. Patients with cerebral malaria underwent catheterization, and urine was tested for sulphonamides (fiiln test) and quinolines (Wilson Edeson test). Parasite counts were made every eight hours for 48 hours, then twice daily until negative and then daily. Full blood cell counts and routine biochemistry studies including liver function results were obtained daily. Temperawe, pulse, and blood presswe ware measured at feast every four hours, but more frequently in seriously ill patients. Quinine dihydrochlorkfe (Government Pharmaceutical Organisation of Thailand) was given in a dose of 10 mg/kg (equivalent to 8.3 mg/kg base) infused over four hours at eight-hour intervals. In two patients with cerebral malaria, infusions were given over two hours. The drug was dispensed
1952
The American
Journal
al Medktne
votuma
73
565
QUININE
IN FALCIPARUM
MALARIA-WHITE
ET AL.
Quinine Pharmacokinetics: Acute Faiciparum Malaria
TABLE II
Cerebral Malaria
Uncomplicated Malaria Number of Standard Deviation Patients Mean
Number of Patients
Mean
Standard Deviation
18
0.92
0.42
11
1.35
0.80
0.031
18 25
1.18 18.2
0.37 9.7
11 13
1.67 16.0
0.34 7.0
0.0013 N.S.
15
0.21
0.18
11
0.21
0.08
N.S.
15
0.20
0.18
11
0.21
0.12
N.S.
Total clearance (ml/min/kg) Vd (liter/kg) Elimination t/2 (hours) Renal clearance (ml/min/kg) Renal clearance Total
Significance (p value)
Vd = total apparent
volume of distribution; N.S. = not significant at 5 percent level. Seven patients with cerebral malaria and two patients with uncomplicated malaria had baseline plasma samples that revealed quinine, indicating oretreatment. These oatients are excluded from the calculations of Vd and total clearance. Accurate urine volumes kere
not obtained
from
10 patients.
with a 1 ml plastic syringe. All patients were given quinine in 0.9 percent saline intravenously, via a Braun perfusor pump to ensure a constant rate of infusion. intravenous fluids were given as required, through the same cannula, and blood samples were taken from the opposite arm. Intramuscular antipyretics (dipyrone, 500 mg) and anti-emetics (metoclopramide, 10 mg) were administered if necessary, but no other drugs were given in the first 12 hours. Lumbar puncture was performed soon after admission and blood sampled simultaneously. Thereafter, intravenous quinine dihydrochloride was given every eight hours until the patient regained consciousness, Quinine sulphate tablets were then given in a dose as near as possible to the intravenous dose to complete seven days’ treatment. Patients with uncomplicated malaria received oral treatment after the first quinine infusion. Blood was taken before treatment, at 0, 15, 30, and 60 minutes, then two, three, four, five, and six hours after the end of the infusion. Further samples were taken from patients with cerebral malaria before and after the next three quinine infusions. Thereafter,‘samples were taken daily before and after the first dose of the day. Blood was collected in lithium-heparin bottles, and plasma was separated immediately and stored at -2OOC. Urine was collected from the indwelling catheter into a volumetric flask and stored similarly. Quinine was measured by the benzene extraction method [4] using
TABLE III
an Aminco Bowman spectrophotofiuorimeter. inter-assay variation between standards did not exceed 3 percent. Three patients who had recovered from cerebral malaria were restudied when they were readmitted because of fever. Blood smears revealed no organisms, and urinary tract infection was finally diagnosed in each case. Patients with uncomplicated falciparum malaria were restud/ed one month after the acute infection when clinically cured ahd subjectively well. Plasma concentrations fitted closely a single exponential decline following the four-hour infusion._ Tne elimination half-time was therefore calculated from the semilogarithmic plot of plasma quinine against time. The total apparent volume of distribution (Vd) was calculated ffom the equation Vd = (Ri”r[ 1 - eKt])/KC where Rrnris the rate of infusion (total dose of quinine base per kilogram body weight divided by time), K is the elimination constant, and G ihe plasma level, extrapolated from the linear plot, at the end of the constant infusion [5]. This estimate of Vd is based on a one-compartment model and assumes that the distribution phase is no longer materially affecting the measured and derived parameters. When plasma concentrations had reached a high point on the second or third day of treatment, Vd was recalculated from the formula Vd = D/PK where D is the dose divided by dose interval and P the mean plasma concentration
Quinine Pharmacokinetlcs: Uncomplicated Falciparum Malaria Acute
Total clearance (ml/min/kg) Vd (liter/kg) Elimination t/2 (hours) Renal clearance (ml/min/kg)
Number of Patients
Mean
8
1.28
8 9 9
Convalescent Standard Deviation
Mean
Standard Deviation
0.63
3.09
1.18
0.0002
1.62 17.0
0.36 7.0
2.74 11.1
0.47 4.1
0.0002 0.006
0.20
0.08
0.53
0.20
0.0004
N.S. = not significant.
586
October 1982
The American Journal of Medicine Volume 73
Significance (p value)
QUININE
TABLE IV
MALARIA-WHITE
ET AL
Quinine Pharmacokinetics in Three Patients Readmitted with Fever due to Urinary Tract infection 112 (hours)
Cerebral malaria Fever readmission
IN FALCIPARUM
Totalclearance (ml/min/ka)
Vd /litsr/ka)
9.1
20.5
11.9
1.38
1.512
0.79
1.75
0.85
0.77
9.7
8.3
5.2
1.79
1.12
1.26
2.12
1.55
2.78
t/2 = elimination half-time; Vd = total apparent volume of distribution.
at steady state. Total clearance was calculated as the product of the acute Vd and K values. Renal clearance of unchanged quinine was calculated from the ratio of urine quinine excreted over the eight hours following the infusion, to the plasma concentration at the midpoint of the collection. Creatinine clearance was measured from the same samples. In order to assess the validity of estimating elimination halftime in acute maiarii from serial plasma concentrations taken over a relatively short time (i.e., less than one half-time), pIas& concentrations were measured in five patients after the last dose of quinine and followed for 30 hours. The two-tailed Student t test was used to compare the variables of the groups with cerebral malaria and uncomplicated malaria. Paired t tests were used to compare the same patients’ variables in the acute and convalescent phase. Data not conforming to a normal distribution were compared by the Wilcoxon rank sum test, and rank correlation was calculated by the method of Kendall.
RESULTS Pharmacokinetlcs. In patients with acute uncomplicated falciparum malaria, total clearance of quinine was significantly lower during the acute illness than after recovery, and was even lower in the patients with cerebral malaria (Tables II and ill). The total apparent volume of distribution was similarly reduced, being lowest in the patients with cerebral malaria. There was a wide variation in elimination half-times in both groups of patients with acute malaria, and there was no significant difference between the two groups. Elimination half-times of the five patients with cerebral malaria at the end of the treatment course (12.8 f 4.1 hours, mean f 1 S.D.) were not significantly different from the acute values (15 f 4.9 hours, p = 0.37). There was, however, a significant difference between patients in the acute phase of malaria and the same patients studied one month after recovery. In 12 patients with cerebral malaria, Vd values were estimated from the peak plasma concentrations (second or third day of treatment) and were significantly larger than the acute phase values (1.66 f 6.52 compared with 1.25 f 0.34 liters/kg, mean f 1 S.D., p = 0.04). Plasma concentrations of quinine were higher in cerebral malaria than in uncomplicated malaria, consistently exceeding 10 mg/liter, and in 60 percent of patients exceeding 15
October
mg/liter. There was no correlation between these pharmacokinetic parameters and fever, depth of coma, length of unconsciousness, parasite clearance rate, and biochemical and hematologic variables (Table I). Three patients with cerebral malaria who were later readmitted with fever were treated with quinine before the final diagnosis of urinary tract infection was certain. Total clearances had improved but were still abnormally low (Table IV). Renal Clearance. Renal clearance and the proportion of renal to total clearance were not significantly different in patients with cerebral and non-cerebral malaria. There was, however, a significant increase in renal clearance on recovery from malaria. Total clearance and urinary clearance were roughly parallel, as the ratio of urinary to total cieaiance did not significantly change between any of the groups. Creatinine clearance and urinary quinine clearance correlated in patients with cerebral malaria (r = 0.67, p = 0.009) and in convalescent patients (r = 0.70, p = 0.008), but not in patients with uncomplicated malaria. in this last group, creatinine clearance was usually normal, but there was a wide variation in urinary quinine clearance. There was no correlation between rate of urine flow and quinine clearance in any of the groups. Serum creatinine exceeded 3 mg/lOO ml in six patients with cerebral malaria. There was no significant difference in total clearances, Vd values, or elimination half-times between these and the remainder of the patients with cerebral malaria. However the difference in nonrenal clearance (1.10 f 0.53 and 0.64 f 0.31 mi/min/kg, mean f 1 S.D., p = 0.051) approached significance. Cerebrospinai Fluid Concentration (Table V). Cerebrospinai fluid quinine concentration varied between 0.2 and 1.O mg/liter and did not correlate with plasma quinine concentration. The mean cerebrospinai fluid/ plasma quinine ratio was 0.07 (S.D. 0.03). This ratio was inversely correlated with plasma quinine concentration (r = 0.59, p = 0.001). Cerebral Malaria: Fatal Cases. These four patients were characterized by high parasitemia (all had more than 10 percent of red blood ceils parasitized, mean count 670,437 parasitized cells/@) and severe bio-
1982
The American Journal of Mediche
Volume 73
567
QUININE IN FALCIPARUM
TABLE V
ET AL.
Cerebrospinal Fluid Quinine Concentrations in Cerebral Malaria
C.&F. Quinine
Plasma Quinine
0.8 0.8 0.8 0.4 0.8 0.5 0.5 0.9 0.8 0.2 0.7 0.7 1.0 1.0 0.5 0.9 0.8 1.0 All patients
MALARIA-WHITE
s
Quinine
10.3 11.8 9.3 11.0 10.7 9.8 15.0 11.6 17.4 10.8 6.9 18.6 15.0 16.7 7.7 6.9 9.2 9.8 were unconscious
Plasma Concentration during Seven Days’ Treatment.
0.08 0.05 0.09 0.04 0.06 0.05 0.03 0.08 0.05 0.02 0.10 0.04 0.06 0.06 0.07 0.13 0.07 0.10 at the time of lumbar
puncture.
chemical derangement. The mean blood urea nitrogen level was 62.2 mg/dl, mean serum creatinine level was 5.6 mg/dl, mean total bilirubin level was 16.1 mg/dl, and mean serum aspartate aminoiransferase level was 128 Reitman-Frankel units/dl. The cause of death was multifactorial; in two patients, pulmonary edema developed, in association with metabolic acidosis and acute renal failure in both; in one patient, septicemia and pneumonia developed; in the other, the immediate cause of death was uncertain. The time from the beginning of intravenous quinine to death, the plasma quinine level immediately before death, and the pharmacokinetic parameters in these patients are given in Table VI. The elimination half-times were significantly shorter than those of the remaining patients (p = 0.013). One patient had previously received oral quinine, and quinine was detectable in the baseline blood sample. Total clearance and Vd values in the remaining three patients were not significantly different from those in the surviving patients. The rise in plasma quinine concentration and delayed fall in parasite count in Patient 3 are shown in Figure 1. TABLE VI
566
October
Quinine Pharmacokinetics: Fatal Cases
Palienl Number
Elimination t/2 (hours)
1 2 3 4
16.1 12.5 7.5 13.7
1962
Plasma concentrations reached a peak on the third day of treatment (60 f 24.6 hours, mean f 1 S.D.) and then declined to plateau at plasma levels approximately 65 percent of the peak plasma level, although there was a wide variation. The decline in plasma levels coincided with clinical improvement and also the change from intravenous to oral medication. The bioavailability of quinine in these tablets (quinine sulphate) was not investigated. Fatal Case: Previous Treatment. Seven of the patients with cerebral malaria and two with uncomplicated malaria had quinine in the baseline blood sample. In a further five patients sulphonamide was detected in the urine by the lignin test. Two patients had a positive result on Wilson Edeson urine testing, but their baseline blood contained no quinine. They had presumably previously received chloroquine. Dosage Accuracy. The concentration of quinine in the intravenous preparation was assayed and found to be within 3 percent of the quoted value in five random samples. Dispensing accuracy using a 1 ml syringe was f2 percent. Toxicity. There was no evidence of cardiovascular toxicity in either cerebral or uncomplicated malaria. There was no relationship between mean arterial pressure and plasma quinine concentration. Changes in pulse or blood pressure could not be related to the administration of quinine or to peak plasma concentrations. No dysrythmias were observed. Prolongation of the rate-corrected electrocardiographic Q-T interval (Q-T/d/R - R) by more than 5 percent of the baseline value was observed in 65 percent of the patients with cerebral malaria. The mean maximal lengthening f 1 S.D. of the Q-Tc interval of all the patients was 10.3 f 7.7 percent. The longest recorded Cl-Tc interval was 0.537 seconds in a patient with a plasma concentration of 9.3 mg/liter. Overall, there was no correlation between the change in plasma concentration and the prolongation of the Q-Tc interval. The most consistent electrocardiographic abnormality in all patients was a change in T-wave morphology, which usually became flatter. Slight QRS widening (not exceeding 0.03 seconds) was seen in 11 patients with cerebral malaria, but no other changes were noted.
The American
(IitZkg)
Total clearance (mllminlkg)
1.06 1.15 1.41
0.76 1.06 2.16
Journal
of Medicine
Plasma Quinine Level before Death (mglliter)
Time from Start of Quinine Therapy lo Death (hours)
9.1 11.2 16.0 13.1
22 24 76 44
Volume 73
QUININE IN FALCIPARUM
MALARIA--WHITE
1:
ET AL
IO4 PARASITE COUNT/,,
PLASMA OUININE mg/l
103 1 O2
IO
Figure 1. Response of parasitemia (number of parasitized red cells/p I) to rise in plasma quinine concentration in a fatal case of cerebral malaria (Patient 3).
24
There was no evidence of permanent neurotoxicity; daily examination of the eye revealed no evidence of retinal vasoconstriction and no abnormalities in the light reflex. In patients with cerebral malaria, no II or VIII nerve toxicity was clinically detectable when they regained consciousness, and all were completely normal on discharge from hospital. In one patient, a six year old boy, neuromuscular weakness developed on the third day following admission. This was partially reversed by edrophonium and neostigmine. The electrocardiographic results were normal, and the plasma quinine concentration was 10.2 mg/liter. This may have been quinine-induced myasthenia. Vomiting was not associated with high cerebrospinal fluid concentrations of quinine nor temporally with the highest plasma quinine concentrations. The majority of the patients who vomited did so within the first 24 hours of admission. Eight of the eleven conscious patients had tinnitus when blood levels exceeded 5 mg/liter, two patients complained of nausea and vomited within six hours of admission when their temperature exceeded 39OC. Six of the eleven patients experienced improvement in’ headache. Tinnitus was usually mild or absent on recovery from cerebral malaria, and these patients usually had no complaints other than slight anorexia. Thrombocytopenia is common in cerebral malaria; 25 percent of the patients with cerebral malaria had platelet counts of less than 6O,OOO/~l with a nadir on the third or fourth day following admission. This was attributed to the disease rather than quinine. COMMENTS
Cerebral malaria is a severe manifestation of falciparum malaria, with a treated mortality of up to 50 percent
HOURS
48
72
[6,7]. Parenteral antimalarials are the only treatment of proved value, and therefore it is essential that chemotherapy is optimized through a better understanding of the pharmacokinetics and therapeutic modes of action of the currently available drugs. Recovery of the patient can come only through death of the parasites. There is no convincing evidence that rapid killing of parasites is harmful, so drug treatment should be designed to exceed the parasites’ minimal inhibitory concentration as soon as possible and thereafter maintain concentrations within the therapeutic ratio for the remainder of the treatment course. These same principles are fundamental to the treatment of bacterial infections. In a recent study of 100 cases of cerebral malaria, it was found that patients with fatal disease began receiving quinine significantly later in the course of their illness than survivors [ 71. Quinine clearance was reduced in acute falciparum malaria, and was significantly lower in cerebral malaria than in uncomplicated cases. Thus the reduction in quinine clearance was related to the severity of the disease, and plasma levels of quinine were therefore higher in cerebral malaria. There was a consistent contraction in Vd and variable prolongation of the elimination half-time. The contraction in Vd may be related to several factors such as dehydration, obstruction of the capillary bed by parasitized red cells, and alterations in tissue binding. None of the patients in this study was clinically hypovolemic, but minor degrees of dehydration cannot be excluded. Three patients with cerebral malaria who were readmitted and found to have urinary tract infections also had low quinine clearances and Vd values. Perhaps fever was a common factor contributing to these changes in both bac-
October 1982
The American Journal of Medlclne
Volume 73
569
QUININE IN FALCIPARUM MALARIA-WHITE
ET AL.
terial and malarial infections, although it cannot explain the differences between patients with cerebral and uncomplicated malaria, as temperatures were not significantly different between the two groups. Etiocholanolone fever has also been shown to increase plasma concentrations of quinine [8]. The relative importance of fever and these other factors cannot be assessed from available evidence. Because of the potential danger of interrupted treatment in the acute phase of this severe infection, it was not possible to delay the second dose of quinine in order to collect more blood samples and to obtain a more satisfactory estimate of elimination half-time. However, when the decline in plasma concentration was followed for 30 hours at the end of the treatment course, and in convalescent patients, it always continued monoexponentially. The absence of a measurable distribution phase, and the relatively long elimination half-times make it unlikely that there are large errors in the calculation of Vd values [9]. This suggests that a one-compartment model is adequate to describe the elimination kinetics following a four-hour intravenous infusion of quinine, and that calculation of the elimination half-time from plasma concentrations over a six-hour period gives a reasonable estimate of the true value. The major route of quinine elimination is via liver metabolism [lo]. Renal clearance of unchanged quinine accounted for approximately 20 percent of the total clearance in both uncomplicated and cerebral malaria. Renal and total clearances were reduced in parallel. The urinary quinine clearance was not correlated with rate of urine flow; however, in cerebral malaria and after recovery from uncomplicated malaria, it was significantly correlated with creatinine clearance. At low creatinine clearances, the ratio of urinary quinine clearance to creatinine clearance increased-in some cases this ratio exceeded unity. This suggests that quinine is filtered and secreted by the normal kidney, with secretion becoming relatively more important at low glomerular filtration rates. It is widely recommended [ 1 l-131 that the dose of quinine should be reduced if there is renal impairment. There was, however, no difference in total clearance or Vd values between patients with cerebral malaria who had increased serum creatinine levels (>3 mg/ 100 ml) and the remainder of this group; indeed, nonrenal clearance was greater in the former group, although this was not significant at the 5 percent level (p = 0.051). Quinine dosage in renal failure requires further study. Liver dysfunction in malaria has been widely reported, but is difficult to measure. None of the conventional indexes of liver function correlated with total clearance or prolongation of the quinine elimination half-time.
570
October 1982
The American Journal of Medicine
Trenholme et al. [8] have measured the elimination half-time of quinine at the end of a treatment course for mild artificially induced malaria. They compared the benzene extraction method, which measures quinine alone, with a metaphosphoric acid method, which measures both quinine and its metabolites. They concluded that hepatic metabolism of quinine was reduced in malaria while renal excretion was not. Hepatic blood flow is reduced in experimental malaria [ 141, and there is sometimes histopathologic evidence of liver damage [ 151. Reduction in quinine clearance is probably related to impaired liver function, but this cannot be predicted from the results of biochemical investigations. Quinine plasma concentrations are high in the first two to four days of treatment [ 161, and then fall as the clinical condition improves. When quinine was given every eight hours, plasma concentrations rose to reach a peak 60 f 25 hours (mean f 1 S.D.) after treatment was begun. True steady-state values were never attained in the early phase of treatment because clearance and Vd values were constantly changing; however, at peak plasma concentrations, drug administered must equal drug eliminated, albeit for a short time. In cerebral malaria, estimated Vd values at this stage were 33 percent larger than those measured at the beginning of treatment. Thereafter, levels declined to plateau around 65 percent of the peak value. The fall in plasma quinine levels in the second half of the treatment course may result not only from an increase in total clearance and expansion of the Vd, but also the substitution of oral for intravenous quinine, with probable reduction in bioavailability for the same dose. The development of quinine resistance in Thailand has entailed an increase in the length of treatment and total dosage required to cure P. falciparum malaria [ 171. The in vitro minimal inhibitory concentrations for some strains have now reached 10 mg/liter [ 18,191. This was previously considered a toxic plasma concentration, although in cerebral malaria, by the third day of treatment, plasma concentrations invariably exceed this value. Death from cerebral malaria usually occurs within the first 48 hours [7]; thus, although high plasma concentrations of quinine occur, they are reached at a time when the patient is usually improving. Conversely, inadequate plasma concentrations may be obtained on the critical first day of treatment. The slow and delayed fall in parasite count in the longest surviving patient who died, associated with rapid clearance and consequent slow rise in plasma quinine concentrations, clearly illustrates this point (Figure 1). If the dose of quinine were reduced in cerebral malaria as has been advocated [ 1,2], therapeutic plasma concentrations would take even longer to be achieved. Despite plasma levels of quinine exceeding 15
Volume 73
QUININE IN FALCIPARUM
mg/liter, and in some cases 20 mg/liter, there was no evidence of cardiotoxicity. Quinine-induced electrocardiographic changes were small. It is interesting that the “quinidine effect” of marked electrocardiographic D-T prolongation is much less pronounced with quinine the diastereomer, despite comparatively high plasma concentrations, Apart from one patient who may have had a transient quinine-related myasthenic syndrome, there were no neurologic sequelae. In particular, there was no deafness or blindness. Daily ophthalmoscopy failed to demonstrate the constriction of the retinal vessels seen in quinine poisoning. Clearly, unconscious patients are unable to complain of “cinchonism,” but this symptom complex was usually mild or absent when they recovered consciousness, and was not a problem in the patients with uncomplicated malaria or convalescent patients. Vomiting was relatively frequent during the acute illness, particularly in patients with high fever. This was not related to the cerebrospinal fluid or plasma quinine concentration. Upper gastrointestinal side effects may be related to the introduction of quinine and consequent rise in plasma concentration. There have been several studies relating plasma levels to side effects. Powell and McNamara [20] found that minor side effects were common with quinine treatment, occurring most frequently when plasma concentrations exceeded 10 mg/liter. Plasma concentrations exceeding this value were also associated with minor toxicity by Taggart et al. [ 2 11. In the past, serious cardiovascular side effects were associated with the practice of giving intravenous injections of quinine over a short time [22]. This is still advocated in some textbooks of tropical medicine [23-251, and the side effects are probably due to
MALARIA-WHITE
ET AL.
transient high concentrations of unbound drug early in the distribution phase causing cardiovascular toxicity. Slow infusion over two or four hours appears to prevent these adverse effects. Serious neurologic side effects have been reported occasionally [ 31. Some of the ones reported (e.g., focal hemisphere deficits) are extremely unusual in quinine self-poisoning and may well have resulted from malaria rather than drug toxicity. Recommendations on quinine dosage that have been limited by observations of minor toxicity [2,8,20] are not appropriate in severely ill patients who require urgent treatment. Furthermore, although the major alterations in quinine pharmacokinetics in severely ill patients result in high plasma levels, these do not appear to produce serious toxicity and may be life-saving. This is particularly important in Thailand, where the minimal inhibitory concentrations of quinine for P. falciparum continue to rise. In cerebral malaria, distribution of quinine throughout the capillary bed, particularly in the brain, may be inadequate because of obstruction by parasitized red cells. It is in these regions that parasiticidal concentrations are most needed. There is a definite risk of inadequate treatment in cerebral malaria if the initial doses of quinine are reduced. ACKNOWLEDGMENT We are grateful to the Director of Pra Pokklao Hospital, Dr. Chaisit Dharakul and his staff, and Dr. Damrong Bhanthumkosol and his laboratory staff, and especially the nursing staff of the Intensive Care Unit. We thank the staff and nurses of the Bangkok Hospita! for Tropical Diseases, Mrs. Patchari Prakongpan, and Miss Kamolrat Lavansiri for their help throughout this study.
REFERENCES 1. 2. 3.
a. 9.
Hall AP: The treatment of malaria. Br Med J 1976; 1: 32332a. Hall AP: The treatment of severe falciparum malaria. Trans R Sot Trop Med Hyg 1977; 71: 367-379. Hall AP, Hanchalay S, Doberstyn EB, Bumnetphund S: Quinine dosage and serum levels in falciparum malaria. Annu Rep SEATO Med Res Labs 1975: 241-250. Cramer G, lsaksson B: Quantitative determination of quinidine in plasma. Stand J Clin Lab Invest 1963; 15: 553-556. Goldstein A, Aronow L, Kalman SM: Principles of drug action, 2nd ed. New York: John Wiley, 1974; 314. Daroff RB, Deller JJ, Kastl AJ, Blocker WW: Cerebral malaria. JAMA 1967; 202: 679-662. Warrell DA, Looareesuwan S, Warrell MJ. et al: Dexamethasone proves deleterious in cerebral malaria: double blind trial in 100 patients. N Engl J Med 1962; 306: 313-319. Trenholme GM, Williams RL, Rieckmann KH, Frischer H, Carson PE: Quinine disposition during malaria and during induced fever. Clin Pharmacol Ther 1976; 19: 459-467. Riegelman S, Loo J, Rowland M: Concept of a volume of distribution and possible errors in evaluation of this parameter.
October
10.
il.
12.
13.
14.
15.
1962
J Pharm Sci 1966: 57: 126-133. Brodie BB, Baer JE, Craig LC: Metabolic products of the cinchona alkaloids in human urine. J Biol Chem 1951; 166: 567-581. Miller LH: Malaria. In: Hoeprich PD, ed. Infectious diseases. 2nd ed. Hagerstown, Maryland: Harper & Row, 1977; 1086. Donadio JV, Whelton A, Kazyak L: Quinine therapy and peritoneal dialysis in acute renal failure complicating malarial haemoglobinuria. Lancet 1966; I: 375-379. Canfield Cl, Miller LH, Bastelloni PJ, Eichler P, Barry KG: Acute renal failure in P&wnodium fdcipatum malaria. Arch Intern Med 1966; 122: 199-203. Skirrow MR, Chongsuphajaisiddhi T, Maegraith BG: The circulation in malaria. II. Portal angiography in monkeys (Macaca mulaffa) with P. know/&and in shock following manipulation of the gut. Ann Trop Med Parasitol 1964; 56: 502-510. Aikawa M. Susuki M, Gutierrez Y: Pathology of malaria. In: Kreier JP. ad. Malaria, vol 2. New York: Academic Press, 1980; 47-102.
The American Journal of Medicine
Volunw 73
571
QUININE
16.
17.
18.
19. 20.
572
IN FALCIPARUM
MALARIA-WHITE
ET AL
Brooks MH, Malloy JP, Bartelloni PJ, Sheehy TW, Barry KG: Quinine, pyrimethanine, and sulphorthodimethoxine: clinical response, plasma levels, and urinary excretion during the initial attack of naturally acquired falciparum malaria. Clin Pharmacol Ther 1969; 10: 85-91. Bunnag D, Harinasuta T, Suntharasamai P, et al.: In vivo response of Plasmodium falciparum to different regimes of antimalarial drugs. Mahidol University Annual Research Abstracts, 1980; 362. Chongsuphajaisiddhi T, Subcharoen A, Attanath P: In viva and in vitro sensitivity of falciparum malaria to quinine in Thai children. Ann Trop Paediatr 1981; 1: 21-26. Thaithong S: Personal communication. Powell RD. McNamara JV: Quinine side effects and plasma
October 1992
The American Journal of Medicine
21.
22.
23. 24. 25.
Volume 73
levels. Proc Helminth Sot Wash 1972; 39: 331-338. Taggart JV, Earle DP, Berliner RW, et al.: Studies on the chemotherapy of the human malarias. Ill. The physiological disposition and antimalarial activity of the cinchona alkaloids. J Clin Invest 1948; 27: 80-86. Strahan JH: Quinine by continuous intravenous drip in the treatment of acute falciparum malaria. Trans R Sot Trop Med Hyg 1948; 41: 669-676. Wilcocks C, Manson-Bahr PEC: Manson’s tropical diseases, 17th ed. London: Bailliere Tindall, 1972; 69. Adams ARD, Maegraith BG: Clinical tropical diseases, 6th ed. Oxford: Blackwell Scientific, 1980; 266. Woodruff AW: Medicine in the tropics. Edinburgh: Churchill Livingstone, 1974; 82.