Acute Lung Injury Complicating Imported Plasmodium falciparum Malaria

Acute Lung Injury Complicating Imported Plasmodium falciparum Malaria* Bertrand Gachot, MD, PhD; Michel Wolff, MD; Gisele Nissack, MD; Benoit Veber, MD; and Franyois Vachon, MD Study objective: To characterize adult patients with acute lung injury complicating severe imported Plasmodium falciparum malaria. Design and setting: Retrospective study of patients with severe P falciparum malaria admitted to the medical ICU of a university hospital infectious diseases department. Patients: Forty adults with complicated malaria, with (group 1, 12 patients) or without (group 2, 28 patients) acute lung injury. Results: Patients with acute lung injury had a higher simplified acute physiology score on admission (24.2± 3.2 vs 13.7 ±0. 7 in group 2, p<0.0001) and a longer time interval to adequate antimalarial therapy (8.8±2.5 vs 4.9±0.6 days in group 2, p=0.046). Of the nine group 1 patients given mechanical ventilation, eight had a Pa02fFio2:s200 mm Hg. Two patients with moderate hypoxemia received oxygen through a nasal tube and one received continuous positive airway pressure via a face mask. Acute renal failure, unrousable coma, metabolic acidosis, and shock were significantly more common among group 1 patients. The number of

pulmonary edema and ARDS are severe, often fatal manifestations of Plasrrwdium falciparum malaria in adults. 1·2 Although this syndrome is among the major criteria for severe and complicated malaria, 1 it corresponds to a range of parenchymal lung damage and clearly needs more accurate definition. 3 Moreover, many complications of severe malaria may contribute to the genesis of hypoxemia, including sequestration of parasitized RBCs in the pulmonary vasculature, aspiration pneumonia, fluid overload, and concomitant Gram-negative bacteremia.l·2 The aim of this study was to characterize adult patients with P falciparum malaria who met well-defined criteria for acute lung injury. Factors that may contribute to acute lung injury were also sought.

*From the Clinique de Reanimation des Maladies Infectieuses, Groupe Hospitafier Bichat-Claude Bernard, Paris, France. Manuscript received August 4, 1994; revision accepted April 21, 1995. R~rint requests: Dr.; Cacho!, Clinique de Reanimation des Maladtes Infectteuses, Hopttal Btchat-C[aude Bernard, 46 rue HenriHuchard, 75018 Pans, France

746

complications of malaria was significantly higher in patients with acute lung injury (4.7±0.5 vs 1.6±0.1 in group 2, p<0.0001). Five patients, including four with acute lung injury, had evidence of bacterial infection (pneumonia or primary bacteremia) at ICU admission. Four patients with acute lung injury died (33%) vs one patient without acute lung injury (4%, p=0.022). Conclusions: Acute lung injury is more likely to occur in patients with extremely severe, multisystemic P falciparum malaria. In patients with acute lung injury and septic shock, bacterial coinfection should be suspected and treated empirically since it contributes substantially to early mortality. (CHEST 1995; 108:746-49)

PEEP=positive end-expiratory pressure

Key words: adult respiratory distress syndrome; bacteremia; malaria; Plasmodium falciparum; pneumonia; pulmonary edema; septic shock

METHODS

The records of all patients with severe imported P falciparum malaria admitted to the medical ICU of the Bichat-Claude Bernard Hospital from January 1988 to May 1993 were reviewed retrospectively. Severe and complicated malaria was defined according to 1990 World Health Organization criteria1 In brief, in the presence of asexual forms of Pfalciparum in the blood, presence of one or more of the following features defined severe malaria: (1) cerebral malaria \vith unrousable coma, ie, a Glasgow coma score :58;4 (2) severe anemia (hemoglobin <5 g/dL); (3) oliguric (urine output <400 mU24 h) renal failure (serum creatinine >265 pmoi!L); (4) hypoglycemia (blood glucose <2.2 mmoi!L); (5) acidemia (arterial pH <7.25) or acidosis (plasma bicarbonate concentration <15 mmoi!L); (6) circulatory collapse or shock with systolic blood pressure <70 mm Hg, cold clammy skin, and persistent hypotension (systolic blood pressure <90 mm Hg) despite adequate volume repletion; (7) spontaneous bleeding and/or evidence of disseminated intravascular coagulation with a prothrombin time <70%, a partial thromboplastin time 2::1.2 times the upper limit of normal, and a fibrinogen level <2 giL. The presence of jaundice ( detected clinically or serum bilirubin concentration >50 pmoi!L) was recorded but was not considered a major criterion.l Acute lung injury was defined according to new international criteria as the acute onset of bilateral pulmonary infiltrates with a PaOz/Fio2 :5300 mm Hg regardless of positive end-expiratory Clinical Investigations

Table !-Characteristi cs of Patients

Patients n Age, y Male sex(%) Nonimmune travelers {%) Simplified acute physiology score Length of ICU stay, d Time to treatment,* d

RESULTS

Group 1 Acute Lung Injury

Group 2 No Acute Lung Injury

p Value

12 39.2::'::3.7 6 (50) 8 (67) 24.2::'::3.2

28 38.3::'::0.2 17 (61 ) 18 {64) 13.7::'::0.7

0.842 0.530 0.885 <0.0001

12.3::'::0.3 8.8::'::2.5

10.6::':: 1.0 4.9::'::0.6

0.459 0.046

*Indicates time from the onset of fever to initiation of adequate antimalarial therapy.

pressure (PEEP) and a pulmonary artery occlusion pressure :siS mm Hg or no clinical evidence of left atrial hypertension; patients with acute lung injury and a PaOz/Flo2 :S200 mm Hg were classified as having ARDSa A lung injury score was calculated on the basis of the extent of roentgenographic densities, severity of hypoxemia, and PEEP value.5•6 The severity of the clinical condition on admission to the ICU was assessed based on the simplified acute physiology score, which does not take into account the degree of hypoxemia 7 Other information abstracted from the medical records included demographic data, complete hematologic counts, initial parasitemia, and evidence of bacterial infection on admission. Pneumonia was defined as the presence of a purulent tracheal aspirate with a new persistent infiltrate on the chest radiograph and growth in culture of at least l
AntirMlarial Treatment Antimalarial treatment included intravenous administration of quinine formiate, 10 mglkg every 8 h for 7days. A loading dose of 20 mglkg was given to patients who had received neither quinine nor mefloquine before admission. In patients with renal failure, the maintenance dose was routinely reduced by one third after the first 2 days. Serum quinine levels were monitored on a daily basis and maintained between lO and 12 mg!L.

Statistical Analysis Quantitative data were expressed as means±SEMs and compared using Student's unpaired t test. Percentages were compared using either the 2 test with Yates' correction or Fisher's exact test. Values of p<0.05 were conside red significant.

Of the 40 patients with severe imported P falciparum malaria who were admitted to the ICU during the study period, 12 (30%) had evidence of acute lung injury. Thirty-five patients were white and five were black Africans. The infection was acquired in sub-Saharan Africa (38 patients) or South America (2 patients). General characteristics of patients with (group 1) or without (group 2) acute lung injury are presented in Table 1. Two-thirds of the patients were nonimmune travelers. Patients with acute lung injury were in more critical condition on admission, as reflected by a significantly higher simplified acute physiology score. Acute lung injury appeared to be associated with a longer time interval to initiation of adequate antimalarial therapy. Table 2 lists laboratory test results on admission in our patients. Moderate anemia, marked thrombocytopenia, hyperparasitemia, and high serum bilirubin levels were found on admission in both groups. Among these parameters, only serum bilirubin level showed a significant difference between the two groups, with higher levels in group 1 patients. Acute lung injury developed within 24 h of initiation of intravenous quinine therapy in seven patients and 2 to 4 days after initiation of specific therapy in five patients. At diagnosis of acute lung injury, five patients were receiving mechanical ventilation because of coma. In four patients, including three with impaired consciousness, acute lung injury was one of the reasons that intubation and mechanical ventilation were used. Two patients had moderate hypoxemia and were administered oxygen through a nasal tube. One patient was treated with continuous positive airway pressure through a face mask. The lowest Pa0z/F'Io2 ratio in the eight mechanically ventilated patients was 101±16 mm Hg. According to the new definition, 3 eight patients had ARDS. Seven had a lung injury score above 2.5. The other complications of malaria in both groups are shown in Table 3. Patients with acute lung injury Table 3-0ther Complications of P falciparum Malaria

x

Table 2-Laboratory Test Data on Admission

Patients

Group 1 Acute Lung Injury

Group 2 No Acute Lung Injury

p Value

n Hemoglobin, gldL Platelets, X 1dl/mm 3 Parasitemia, % Serum bilirubin, pmol!L

12 9.8::'::0.6 37.1 ::':: 7.5 18.3::'::3.8 255::'::64

28 10.8::'::0.6 38.0::'::4.2 15.7::'::3.1 129::'::20

0.346 0.910 0.645 0.018

Patients

Group 1 Acute Lung Injury

Group 2 No Acute Lung Injury

p Value

n Acute renal failure (% ) Cerebral malaria (%) DIC* (%) Metabolic acidosis (%) Shock{%) Hypoglycemia (%) Anemia (%) No. of complications

12 10 (83) 8 (67) 7 (58) 7 (58) 8 (67) 2 (17) 1 (8) 4.7::'::0.5

28 12 (43) 7 25) ( 7 25) ( 4 14) ( 2 (7) 2 (7) 2 (7) 1.6::'::0.1

0.018 0.012 0.043 0.010 <0.0001 0.380 0.900 <0.0001

*DIC=disseminated intravascular coagulation . CHEST / 108 / 3 / SEPTEMBER, 1995

747

were more likely to have acute renal failure, unrousable coma, metabolic acidosis, or shock. The mean number of complications was markedly higher in the acute lung injury group. Although disseminated intravascular coagulation was fairly common, no patients had clinically significant bleeding. Six patients who subsequently developed acute lung injury required volume expansion because of hypotension; among them, four received blood products. All patients with acute lung injury and clinical shock required vasoactive drugs, including dopamine 2::10 pglkglmin (five patients), epinephrine (five patients), and/or norepinephrine (one patient) . Five of these patients underwent pulmonary artery catheterization, which yielded the following initial values: mean arterial pressure 58±5 mm Hg, mean pulmonary artery pressure 21 ±2 mm Hg, mean pulmonary artery occlusion pressure 11±2 mm Hg, cardiac index 6.5±0.8 Umin!m 2 , systemic vascular resistance index 601±100 dynes·s·cm - 5/m 2, and pulmonary vascular resistance index 137±77 dynes·s·cm- 5/m 2 . Five patients, including four with acute lung injury (of whom three had ARDS), had evidence of a concomitant bacterial infection at admission to the ICU. Four of these five patients, all of whom were in group 1, also had septic shock. The bacterial infections included two early-onset pneumonias with positive blood cultures (Streptococcus pneurrwniae and Acinetobacter baum.annii) and three primary bacteremias (Escherichia coli, Salmonella enteritidis, and Enterococcus faecal is). Except in the patient with A baum.annii pneumonia and bacteremia, appropriate antibiotic therapy was given. Four patients with acute lung injury died (33%), whereas there was only one death in group 2 (4%, p=0.022). Three patients with ARDS, septic shock, and concomitant bacterial infection died within 24 h after admission. In the remaining two patients (one in each group), brain death was the primary cause of the fatal outcome. DISCUSSION

Pulmonary edema and ARDS are considered common complications in adults with severe P falciparum malaria. 1 Yet, most studies focusing on pulmonary involvement were case reports or included small numbers of patients, and none emRhasized the importance of bacterial superinfection. 9- We reported detailed data on 12 patients with severe P falciparum malaria and well-defined acute lung injury, and compared these patients with patients with severe malaria without lung injury who were admitted to our ICU during the same period. We found a high incidence of lung involvement. Patients with acute lung injury had more severe disease, were more likely to develop septic shock, and had a high incidence of early pneumonia or 748

primary bacteremia. In a large study of cerebral malaria, 10% of the patients (6 to 70 years of age) had frank pulmonary edema.12 In contrast, none of 131 children with cerebral malaria had respiratory involvement. 13 In our study, the incidence oflung injury was particularly high, since almost one third of the patients developed this complication and 20% overall had ARDS. This high incidence may be partly as~.:ribable to older age (our population was composed only of adults) . Selection bias is another possible factor: because our ICU is specialized in infectious diseases, it receives patients referred by other medical ICUs because of a failure to respond to appropriate therapy. The occurrence of acute lung injury in patients with P falciparum malaria appears to characterize a subgroup of patients whose illness is particularly severe, as assessed by the simplified acute physiology score. Mortality in this sub~roup was higher, in keeping with previous reports.l·9,J· Other complications of malaria, such as coma, shock, and renal failure , were more frequent in patients with acute lung injury, and the mean number of malaria complications in this subgroup was nearly five, vs 1.6 in patients without lung injury. Although some complications are obviously interrelated (ie, metabolic acidosis may be associated with shock and/or renal failure) , our data show that acute lung injury in P falciparum malaria is most often prut of the multiple organ dysfunction syndrome that is known to complicate the course of sepsis irrespective of the causative agent. 14 Parasitemia did not differ between the two groups, suggesting that the occurrence of multiple organ dysfunction in our patients was not related to a greater parasite burden. However, appropriate treatment seems to have been initiated later in group 1 patients. The importance ofprompt diagnosis and treatment in severe P falciparum malaria has already been emphasized, particularly in nonimmune subjects. 1·10 Bacterial infection is known to complicate the course of patients with severe malaria,1 but in most studies, the distinction between early and nosocomial infections was not made. Early infections include aspiration pneumonia and Gram-negative bacteremia probably due to impaired splanchnic perfusion. 15 The related production of endotoxin and/or cytokines may be responsible for amplified cytoadherence of parasitized red blood cells, contributing to organ dysfunction.16 In the control groups of the two steroid studies, 10 to 36% of ~atients developed either pneumonia or bacteremia.12· 7 However, the bacteriologic data were not reported in detail and the diagnostic criteria for pneumonia were not provided. In our study, 12% of patients had evidence of septicemia and/or well-documented pneumonia on admission to the ICU, and bacterial infection clearly contributed to death in three Clinical Investigations

patients. Three patients had early Gram-negative or Enterococcus bacteremia without any detectable focus. The case of early-onset A baumannii sepsis was unusual, although a similar case has been described with Pseudonwnas aeruginosa. 18 The authors of the latter case report suggested that the involvement of neutrophils in the phagocytosis of P falciparum may have abolished phagocytic defenses against P aeruginosa, resulting in bacteremia. In P falciparum malaria, shock so-called ( "algid malaria") is often associated with bactetial coinfection.1·11 In our study, six of ten patients with shock had negative blood cultures and no patent focus of bacterial infection. The hemodynamic profile in five patients was that of hyperdynamic septic shock. Although endotoxemia without bacteremia cannot be excluded, 19 these data suggest that Pfalciparum malaria itself may be responsible for septic shock. However, despite very high plasma tumor necrosis factor concentrations in severe malaria, shock remains fairly uncommon, particularly in children.20 Patients with severe P falciparum malaria and shock should therefore be treated empirically with broad-spectrum antibiotics, until the results of cultures are available. In conclusion, acute lung injury is most likely to occur in patients with extremely severe, multisystemic P falciparum malaria, and appears to contribute to mortality in adults. The most severe forms of lung injury, ie, ARDS, are often associated with septic shock. In this situation, pneumonia and/or bacteremia should be suspected and empirically treated since they may contribute to a fatal outcome. ACKNOWLEDGMENT: The authors thank Didier Dreyfuss, MD, PhD, for his stimulating criticism. REFERENCES

1 World Health Organization, Division of Control of Tropical Diseases. Severe and complicated malaria. 2nd ed. Trans H. Soc Trop Med Hyg 1990; 84(suppl 2): 1-65 2 White NJ, Ho M. The pathophysiology of malaria. Adv Parasitol 1992; 31:83-173 3 Bemard GH, Artigas A, Brigham KL, eta!. Heport of the American-European consensus conference on ARDS: definitions, mechanisms, relevant outcomes and clinical bial coordination.

Intensive Care Med 1994; 20:225-32 4 Leaver HJ, De Baetselier H ,Bagshawe A, e al. t Cerebral malaria: what is unrousable coma? Lancet 1990: 335:44-56 5 Murray JF, Matthay MA, Luce JM, e tal. An expanded d efinition of the adult respiratory distress syndrome. Am Hev Hespir Dis 1988; 138:720-23 6 Erratum. Am Hev Hespir Dis 1989; 139:1065 7 Le Gall JH, Loirat P, Alperovitch A, et al. A simplified acute physiology score for ICU patients. Ctit Care Med 1984; 12: 975-77 8 Pham LH, Brun-Buisson C, Legrand P, et al. Diagnosis of nosocomial pneumonia in mechanically ventilated patients: comparison of a plugged t elescoping catheter \vith the protected specimen brush. Am Rev Hespir Dis 1991; 143:1055-61 9 Brooks MH, Kiel FW, Sheehy TW, et al. Acute pulmonary edema in falciparum malaria: a clinicopathological correlation. N Eng! J Med 1968; 279:732-37 10 Feldman HM , Singer C. Noncardiogenic pulmonary edema and pulmonary fibrosis in falciparum malaria. H.ev Infect Dis 1987; 9:134-39 11 Charoenpan P, Indraprasit S, Kiatboonsri S, et al. Pulmonary edema in severe falciparum malaria: hemodynamic study and clinicophysiologic correlation . Chest 1990; 97:1190-97 12 Warrell DA, Looareesuwan S, Warrell MJ, e tal. Dexamethasone proves deleterious in cerebral malaria: a double-blind trial in 100 comatose patients. N Eng! J Med 1982; 306:313-19 13 Molyneux ME, Taylor TE, Wirima JJ, et al. Clinical features and prognostic indicators in paediatric cerebral malaria: a study of 131 comatose Malawian children. Q J M ed 1989; 71:441-59 14 Bone RC, Balk HA, Cerra FB, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative theraptL·, Ill sepsis. Chest 1992; 101:1644-55 15 Molyneux ME, Looareesuwan S, Menzies IS, et al. Reduced hepatic blood flow and intestinal malabsorption in severe falciparum malaria. Am J Trop Med Hyg 1989; 40:470-76 16 Hommel M. Amplification ofcytoadherence in cerebral malaria: towards a more rational explanation of disease pathophysiology. Ann Trop Med Parasitol 1993; 87:627-35 17 Hoffman SL, Hustama D, Punjabi NH, et al. High-dose dexamethasone in quinine-treated patients with cerebral malaria: a double-blind, placebo-controlled trial . J Infect Dis 1988; 158: 325-31 18 Kharazmi A, Hoiby N, Theander TG. Pseudonwnas aerugirwsa septicaemia in a patient with severe Plasnwdium falciparum. Trans H Soc Trop Med Hyg 1987; 81:49-50 19 Usawattanakul W, Tharavanij S, Warrell DA, e t al. Factors contributing to the development of cerebral malaria: II. Endotoxin. Clin Exp Immunol 1985; 61:562-68 20 Grau GE, Taylor TE, Molyneux ME, et a!. Tumor necrosis factor and disease severity in children with falciparum malaria. N Eng! J Med 1989; 320:1586-91

CHEST I 108 1 3 I SEPTEMBER, 1995

749