- Email: [email protected]
Endotoxaemia in complicated falciparum malaria
T~~sacmms
OF THE ROYAL Socr~rv
Endotoxaemia
OF TROPICAL MEDIUNE
in complicated
AND
HYGIENE(1988)82, 513-514
falciparum
513
malaria
and Myo-Thwe3 Aung-Kyaw-Zaw’ Khin-Maung-U2 ‘Defence Set&es General Hospital, Mingaladon, Rangoon, Burma; %Xnical Research Division, Department of Medical Research, No. 5, Zafar Shah Road, Dagon P.O., Rangoon, Burma; 3Medicine Department, Institute of Medicine (Z), North Okkalapa General Hospital, Rango?, Burma
Abstract
In a prospective hospital-based study, endotoxin was detected bv amoebocvte limulus lvsate test in the blood of 18 of-20 patients with com$cated Plasmadium falcibarum (16 with cerebral malaria. 2 with black&ate; fever,‘ one with acute malarial ‘hepatitis and one with hepatorenal failure) and in all 5 patients with uncomplicated malaria tested, but in none of 5 healthy volunteers. There were 4 deaths among the 18 patients with complicated malaria and endotoxaemia. No correlation between endotoxaemia and presenceof complications, clinical severity, or degree of parasitaemia was found. A concomitant bacterial infection could account for endotoxaemia in 11 of the 16 patients with cerebral malaria and endotoxaemia; in the other 5 patients with cerebral malaria, 4 with other complications, and 5 with uncomplicated malaria, endotoxin was detected in the blood without any evidence of bacterial infection. Introduction
Many of the clinical, pathological and pathophysiolonical features of comulicated Plasmodium falcipar& malaria (such as cerebral malaria) resemble those of endotoxaemia. Raised endotoxin levels in plasma have long been known in animal models (CLARK, 1978), tid, quite recently, in human malaria natients (TUBBS, 1980). It has been pronosed that bathologi&l changes & fulminant n&r& infection are mediated through the action of raised endotoxin (CLARK, 1981). In this study, the role of endotoxaemia in relation to complicated falciparum malaria, its clinical severity and outcome, is evaluated. Patients
and Methods
20 adult patients with complicated falciparum malaria (16 with cerebral malaria. 2 with blackwater fever, oie with acute malarial hepatitis and one with hepaiorenal failure) and 5 patients with uncomplicated falciparum malaria admitted to Defense Services General Hospital and North Okkalapa General Ho&al. and 5 healthv adults were studied. P&e& were defined as having cerebral malaria when they showed altered consciousness (Glasgow coma scale 3-15) (TEASDALE & JENNETT,1974)fits or meningeal irritation or psychiatric manifestations (such as acute excitement, frank psychosis), but no other pathology (by clinical assessmentand relevant investigations including lumbar puncture), and asexual forms of P. falciparum in blood films. 2 patients with blackwater fever had clinical features of acute malaria and evidence of intravascular Correspondence and requests for reprints to: Dr KhinMaung-U .
haemolysis (free plasma haemoglobin and free urine haemoglobin), but no asexual forms of P. falciparum in their peripheral blood fihn (only 50% of patients with blackwater fever have positive blood smears: BRUCE-CHWATT, 19581.
The patient -with acute malarial hepatitis had asexual forms of P. falcibarum in his wrinheral blood, clinical jaundice (&rum bilirub& o;er 30 pmol/litre), raised serum alanine aminotransferase (over 30 i.u./litre),and an enlarged tender liver. The patient with hepatorenal failure had all the features of the previous patient and signs of hepatic failure and anuria. Drug-induced hepatitis was excluded by history taking, hepatitis-B by absenceof hepatitis B surface antigen in blood, and leptospirosis by examination of acute phase blood and urine under dark ground illumination and negative serological tests for leptospiral antibody in convalescent serum. 5 control patients with uncomplicated malaria had asexual forms of P. falciparum in their blood, clinical features of acute malaria and no evidence of a concomitant bacterial infection. 5 healthy volunteers were included as controls; they subjected voluntarily to collection of blood samples for endotoxin assay. Blood samples were collected under aseptic conditions in sterile pyrogen-free disposable syringes, transferred to sterile pyrogen-free heparinized vials, transported in ice to the Clinical Research Division laboratory within 40 minutes, and stored at -20°C after oromnt separation of plasma. Samples were assayed foi end&toxin by t&e limulus tioebocyte lysate gelation test described by LEVIN et al. (1970), using chloroform extraction of the plasma inhibitors. The limulus lvsate (Etoxate) was obtained from Sigma. Shigella j&x& endotoh (2 pg/ml) was used as standard, and the limulus amoebocyte lysate was sensitive to detect endotoxin in plasma to levels between O-25and 1.0 mg/ml. All batches of glassware, syringes, pipettes, heptiin, chloroform and pyrogenfree water were also checked for endotoxin and found to be negative. Results
Endotoxin was detected in 18 (90%) of 20 patients with complicated falciparum malaria and all 5 patients with uncomplicated malaria, but in none of the healthy controls. There was no correlation between the degree of endotoxaemia and presence or absence of complications in malaria, clinical severity or degree of parasitaemia. There were 4 deaths (22.2%) among 18 patients with complicated malaria and endotoxaemia; neither of 2 patients with complicated malaria without endotoxaemia, and none of 5 patients with
514 uncomplicated malaria (and detectable endotoxin in blood) died. A concomitant bacterial infection (pneumonia, a positive urine culture, pressure sore) that could account for the presence of endotoxin was present in 11 of 16 patients with cerebral malaria. There were 5 patients with cerebral malaria in whom no clinical or laboratory evidence of a concomitant bacterial infection was present. Among these 5 patients, endotoxaemia was present in 4 (of whom one died). Discussion In a study of 281 patients suspected of Gramnegative sepsis (LEVIN et al., 1972), endotoxin was detected by the limulus lysate test m 14%, positive blood cultures in 12%, and both endotoxaemia and bacteraemiaoccurred in 7%. Bacteraemiaand endotoxaemia are transient events that do not necessarily occur simultaneously. Gram-negative organisms in tissue sites may releaseendotoxin which, rather than bacteraemia, could produce symptoms and signs of seusis (LEVIN et al., 1972). -TUBI~ (1980) showed that, among 24 mouse pairs with different eradesof P. bewhei infection tested for endotoxin u&g the limulus &ate test, 21 pairs gave positive results. TUBBS(1980) also demonstrated that endotoxin was present in the blood of all of 16 African children with acute falciparum malaria, of whom 5 had cerebral malaria. In his study there was no attempt to correlate endotoxin levels and the degreeof parasitaemia. In some of our patients with complicated malaria, especially cerebral malaria, a concomitant bacterial infection (such ashypostatic pneumonia, urinary tract infection from indwelling catheters, or pressure sores) could have been the source of endotoxaemia (LEVIN et al., 1972). In our 5 patients with complicated malaria, in whom there was no clinical or laboratory evidence of a coexisting bacterial infection, it is possible that suppressed cellular immunity resulting from acute malarial infection (BRASSEUR et al., 1983; GILBREATH et al., 1983) or altered local flow in the gut could have allowed endotoxin produced by the gut flora to reach the blood stream (CLARK. 1982). The same exnlanation was probably true . also for the presence of endotoxin in blood in 5 patients with uncomplicated malaria. The possibility of the malarial parasite itself
producing endotoxin or endotoxin-like material is still hypothetical. Acknowledgements We thank U Shwe Ni
and Ma Pyone Aye of the
Departmentof Medical Researchfor their technicalassistancewith endotoxinassays,and the medical,laboratoryand nursingstaffof DefenseServicesGeneralHospital,Mingaladon, for their co-operation. References
Brasseur,P., Agraphart., M., Ballet, J. J., Druilhe, P., Warrell, M. J. & Tharavanij, S. (1983). Impaired cell-mediatedimmunitv in Plasmodium falci~anm in-
fected uatients with hiah uarasitaemia. and cerebral malaria.~ Clinical I mmuna&yaand hmunopatholo~, 27, 38-50. BruceChwatt, L. J. (1958). Parasite density index in malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 52, 389. Clark, I. A. (1978). Does endotoxin cause both the disease and parasite death in acute malaria and babesiosis? Lancet, ii, 75. Clark, I. A. (1981). Thromboxane may be important in the organ damageand hypotension of malaria. Medical Hypothesis, 5, 625-631. Clark, 1. A. (1982). Correlation between susceptibility to malaria and Babesia parasites and to endotoxicity. Transactiom of the Royal Socie& of Tropical Medicine and Hygiene, 76, 4-7. Gilbreath, M. J., Pavanand, K., McDermott, R. P., Phisphumvthi, P., Permpanixh, B. & Wimonwatrowateet (1983). Deficient spontaneous cell-mediated cytotoxicity and lectin-induced cellular cytotoxicity by peripheral blood mononuclear cells from Thai adults naturally infected with malaria. Journal of Clinical Microbiology, ii, 296-304. Levin, J., Tomasulo, P. A. & Oser, R. S. (1970). Detection of endotoxin in human blood and demonstration of an inhibitor.~muxul of Laboratory and Clinical Medicine, 75, 903-911. Levin, J., Poore, T. E., Young, N. S., Margolis, S., Zauber, N. P., Townes, A. S. & Bell, W. R. (1972). Gram-negative sepsis: detection of endotoxaemia with the limulus test. Annak of Internal Medicine, 76, l-7. Teasdale? G. & Jennett, B. (1974). Assessment of coma and impaired consciousness: a uractical scale, Lancet,_ ii,. 81184. Tubbs, H. (1980). Endotoxin in human and murine malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 74, 121-3. Received 10 February 1987; accepted for publication 19 July 1987 (delayed in proof)
OF THE ROYAL Socr~rv
Endotoxaemia
OF TROPICAL MEDIUNE
in complicated
AND
HYGIENE(1988)82, 513-514
falciparum
513
malaria
and Myo-Thwe3 Aung-Kyaw-Zaw’ Khin-Maung-U2 ‘Defence Set&es General Hospital, Mingaladon, Rangoon, Burma; %Xnical Research Division, Department of Medical Research, No. 5, Zafar Shah Road, Dagon P.O., Rangoon, Burma; 3Medicine Department, Institute of Medicine (Z), North Okkalapa General Hospital, Rango?, Burma
Abstract
In a prospective hospital-based study, endotoxin was detected bv amoebocvte limulus lvsate test in the blood of 18 of-20 patients with com$cated Plasmadium falcibarum (16 with cerebral malaria. 2 with black&ate; fever,‘ one with acute malarial ‘hepatitis and one with hepatorenal failure) and in all 5 patients with uncomplicated malaria tested, but in none of 5 healthy volunteers. There were 4 deaths among the 18 patients with complicated malaria and endotoxaemia. No correlation between endotoxaemia and presenceof complications, clinical severity, or degree of parasitaemia was found. A concomitant bacterial infection could account for endotoxaemia in 11 of the 16 patients with cerebral malaria and endotoxaemia; in the other 5 patients with cerebral malaria, 4 with other complications, and 5 with uncomplicated malaria, endotoxin was detected in the blood without any evidence of bacterial infection. Introduction
Many of the clinical, pathological and pathophysiolonical features of comulicated Plasmodium falcipar& malaria (such as cerebral malaria) resemble those of endotoxaemia. Raised endotoxin levels in plasma have long been known in animal models (CLARK, 1978), tid, quite recently, in human malaria natients (TUBBS, 1980). It has been pronosed that bathologi&l changes & fulminant n&r& infection are mediated through the action of raised endotoxin (CLARK, 1981). In this study, the role of endotoxaemia in relation to complicated falciparum malaria, its clinical severity and outcome, is evaluated. Patients
and Methods
20 adult patients with complicated falciparum malaria (16 with cerebral malaria. 2 with blackwater fever, oie with acute malarial hepatitis and one with hepaiorenal failure) and 5 patients with uncomplicated falciparum malaria admitted to Defense Services General Hospital and North Okkalapa General Ho&al. and 5 healthv adults were studied. P&e& were defined as having cerebral malaria when they showed altered consciousness (Glasgow coma scale 3-15) (TEASDALE & JENNETT,1974)fits or meningeal irritation or psychiatric manifestations (such as acute excitement, frank psychosis), but no other pathology (by clinical assessmentand relevant investigations including lumbar puncture), and asexual forms of P. falciparum in blood films. 2 patients with blackwater fever had clinical features of acute malaria and evidence of intravascular Correspondence and requests for reprints to: Dr KhinMaung-U .
haemolysis (free plasma haemoglobin and free urine haemoglobin), but no asexual forms of P. falciparum in their peripheral blood fihn (only 50% of patients with blackwater fever have positive blood smears: BRUCE-CHWATT, 19581.
The patient -with acute malarial hepatitis had asexual forms of P. falcibarum in his wrinheral blood, clinical jaundice (&rum bilirub& o;er 30 pmol/litre), raised serum alanine aminotransferase (over 30 i.u./litre),and an enlarged tender liver. The patient with hepatorenal failure had all the features of the previous patient and signs of hepatic failure and anuria. Drug-induced hepatitis was excluded by history taking, hepatitis-B by absenceof hepatitis B surface antigen in blood, and leptospirosis by examination of acute phase blood and urine under dark ground illumination and negative serological tests for leptospiral antibody in convalescent serum. 5 control patients with uncomplicated malaria had asexual forms of P. falciparum in their blood, clinical features of acute malaria and no evidence of a concomitant bacterial infection. 5 healthy volunteers were included as controls; they subjected voluntarily to collection of blood samples for endotoxin assay. Blood samples were collected under aseptic conditions in sterile pyrogen-free disposable syringes, transferred to sterile pyrogen-free heparinized vials, transported in ice to the Clinical Research Division laboratory within 40 minutes, and stored at -20°C after oromnt separation of plasma. Samples were assayed foi end&toxin by t&e limulus tioebocyte lysate gelation test described by LEVIN et al. (1970), using chloroform extraction of the plasma inhibitors. The limulus lvsate (Etoxate) was obtained from Sigma. Shigella j&x& endotoh (2 pg/ml) was used as standard, and the limulus amoebocyte lysate was sensitive to detect endotoxin in plasma to levels between O-25and 1.0 mg/ml. All batches of glassware, syringes, pipettes, heptiin, chloroform and pyrogenfree water were also checked for endotoxin and found to be negative. Results
Endotoxin was detected in 18 (90%) of 20 patients with complicated falciparum malaria and all 5 patients with uncomplicated malaria, but in none of the healthy controls. There was no correlation between the degree of endotoxaemia and presence or absence of complications in malaria, clinical severity or degree of parasitaemia. There were 4 deaths (22.2%) among 18 patients with complicated malaria and endotoxaemia; neither of 2 patients with complicated malaria without endotoxaemia, and none of 5 patients with
514 uncomplicated malaria (and detectable endotoxin in blood) died. A concomitant bacterial infection (pneumonia, a positive urine culture, pressure sore) that could account for the presence of endotoxin was present in 11 of 16 patients with cerebral malaria. There were 5 patients with cerebral malaria in whom no clinical or laboratory evidence of a concomitant bacterial infection was present. Among these 5 patients, endotoxaemia was present in 4 (of whom one died). Discussion In a study of 281 patients suspected of Gramnegative sepsis (LEVIN et al., 1972), endotoxin was detected by the limulus lysate test m 14%, positive blood cultures in 12%, and both endotoxaemia and bacteraemiaoccurred in 7%. Bacteraemiaand endotoxaemia are transient events that do not necessarily occur simultaneously. Gram-negative organisms in tissue sites may releaseendotoxin which, rather than bacteraemia, could produce symptoms and signs of seusis (LEVIN et al., 1972). -TUBI~ (1980) showed that, among 24 mouse pairs with different eradesof P. bewhei infection tested for endotoxin u&g the limulus &ate test, 21 pairs gave positive results. TUBBS(1980) also demonstrated that endotoxin was present in the blood of all of 16 African children with acute falciparum malaria, of whom 5 had cerebral malaria. In his study there was no attempt to correlate endotoxin levels and the degreeof parasitaemia. In some of our patients with complicated malaria, especially cerebral malaria, a concomitant bacterial infection (such ashypostatic pneumonia, urinary tract infection from indwelling catheters, or pressure sores) could have been the source of endotoxaemia (LEVIN et al., 1972). In our 5 patients with complicated malaria, in whom there was no clinical or laboratory evidence of a coexisting bacterial infection, it is possible that suppressed cellular immunity resulting from acute malarial infection (BRASSEUR et al., 1983; GILBREATH et al., 1983) or altered local flow in the gut could have allowed endotoxin produced by the gut flora to reach the blood stream (CLARK. 1982). The same exnlanation was probably true . also for the presence of endotoxin in blood in 5 patients with uncomplicated malaria. The possibility of the malarial parasite itself
producing endotoxin or endotoxin-like material is still hypothetical. Acknowledgements We thank U Shwe Ni
and Ma Pyone Aye of the
Departmentof Medical Researchfor their technicalassistancewith endotoxinassays,and the medical,laboratoryand nursingstaffof DefenseServicesGeneralHospital,Mingaladon, for their co-operation. References
Brasseur,P., Agraphart., M., Ballet, J. J., Druilhe, P., Warrell, M. J. & Tharavanij, S. (1983). Impaired cell-mediatedimmunitv in Plasmodium falci~anm in-
fected uatients with hiah uarasitaemia. and cerebral malaria.~ Clinical I mmuna&yaand hmunopatholo~, 27, 38-50. BruceChwatt, L. J. (1958). Parasite density index in malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 52, 389. Clark, I. A. (1978). Does endotoxin cause both the disease and parasite death in acute malaria and babesiosis? Lancet, ii, 75. Clark, I. A. (1981). Thromboxane may be important in the organ damageand hypotension of malaria. Medical Hypothesis, 5, 625-631. Clark, 1. A. (1982). Correlation between susceptibility to malaria and Babesia parasites and to endotoxicity. Transactiom of the Royal Socie& of Tropical Medicine and Hygiene, 76, 4-7. Gilbreath, M. J., Pavanand, K., McDermott, R. P., Phisphumvthi, P., Permpanixh, B. & Wimonwatrowateet (1983). Deficient spontaneous cell-mediated cytotoxicity and lectin-induced cellular cytotoxicity by peripheral blood mononuclear cells from Thai adults naturally infected with malaria. Journal of Clinical Microbiology, ii, 296-304. Levin, J., Tomasulo, P. A. & Oser, R. S. (1970). Detection of endotoxin in human blood and demonstration of an inhibitor.~muxul of Laboratory and Clinical Medicine, 75, 903-911. Levin, J., Poore, T. E., Young, N. S., Margolis, S., Zauber, N. P., Townes, A. S. & Bell, W. R. (1972). Gram-negative sepsis: detection of endotoxaemia with the limulus test. Annak of Internal Medicine, 76, l-7. Teasdale? G. & Jennett, B. (1974). Assessment of coma and impaired consciousness: a uractical scale, Lancet,_ ii,. 81184. Tubbs, H. (1980). Endotoxin in human and murine malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 74, 121-3. Received 10 February 1987; accepted for publication 19 July 1987 (delayed in proof)