Background
Severe malaria, caused by
Plasmodium falciparum, is associated with high mortality despite appropriate management at intensive care facilities. Few therapeutic interventions have been shown effective to improve survival, including parenteral artesunate, mechanical ventilation and renal replacement therapy [
1‐
6]. Other adjunctive treatments (e.g. application of corticosteroids to reduce cerebral oedema, anticonvulsive drugs in cerebral malaria or antipyretics.) did not improve patient’s outcome, some of which were even associated with increased rates of sequelae and death [
7‐
9].
Whole blood exchange transfusion has been proposed as a method to rapidly reduce parasite biomass [
10]. A meta-analysis of eight case–control studies including 279 patients demonstrated that whole blood exchange transfusion was not superior to parenteral anti-malarial chemotherapy alone, discouraging the use of this adjunctive treatment [
11].
Automated red blood cell (RBC) exchange is a technique that may potentially overcome problems associated with whole blood exchange [
12‐
16]. With this procedure, RBCs are separated automatically from the patient’s whole blood, while plasma, platelets and white blood cells are retained. This procedure is known to be better tolerated in various indications compared to whole blood exchange transfusion with regards to volume alterations, coagulation and electrolyte disturbances, and susceptibility to infection [
17‐
19]. RBC exchange has been successfully used in patients with sickle cell crises [
20] and has been advocated as adjunctive treatment option for severe malaria. First case reports and case series have supported the use of automated RBC exchange and several national guidelines acknowledge the therapeutic potential for selected patients [
21‐
25]. However, firm evidence for RBC exchange derived from randomized controlled clinical trials is still lacking. In the absence of such information, more data from case series and patient cohorts at different institutions are needed to increase the level of evidence. This study reports the use of RBC exchange as an adjunctive treatment for severe falciparum malaria in a cohort of patients treated at the Medical University of Vienna in Austria between 2000 and 2010.
Discussion
This study describes a cohort of patients undergoing RBC exchange in a tertiary health care institution in Austria. Out of 146 patients treated for malaria at the Medical University of Vienna, 13 were treated at an ICU department. Eleven of them fulfilled institutional criteria to consider RBC exchange, but only five patients were actually treated with automated RBC exchange in a ten-year period. This constitutes less than half of the patients (45 %) who were potentially eligible according to institution guidelines [
24]. This discrepancy is most likely explained by logistical constraints of automated RBC exchange rather than by medical contraindications. Challenges include rapid transfer of patients to specialized medical centres, quick and reliable blood smear assessment, diagnosis by experienced physicians and established collaborations between infectious disease departments, departments for Transfusion Medicine, and the ICU. Furthermore, due to the lack of high quality evidence for automated RBC exchange and the questionable benefit of whole blood exchange [
11], some physicians may opt against the use of this adjunctive treatment in severe malaria or may consider it too late for any potential benefit. A well-organized interdisciplinary approach involving staff of the departments of infectious diseases, ICU, and Transfusion Medicine is needed to address these logistical difficulties in a timely manner and encourage physicians to consider automated RBC exchange as an adjunctive treatment option.
Automated RBC exchange was commenced within an acceptable time period and the procedure was well tolerated without any signs and symptoms of clinical deterioration, electrolyte disturbances or bleeding complications. The efficacy of physical removal of parasites was shown by rapid clearance of peripheral parasitaemia. Despite the exchange volume of approximately 1–1.5 times of patients total RBC pool, apheresis did not lead to problems in fluid overload and haemodynamic distress, a phenomenon which was more frequently observed in patients undergoing whole blood exchange transfusion [
17‐
19]. Moreover, during or shortly after automated RBC exchange no specific medical intervention became necessary to maintain haemodynamic and respiratory stability. The occurrence of pneumonia and subsequent lung failure in the patient with prolonged recovery was related to mechanical ventilation. Based on the clinical characteristics and the time course a causal association of this complication with RBC exchange was deemed unlikely. Several weeks after RBC exchange and completion of anti-malarial treatment, this patient experienced fungal sepsis, which was again a complication of prolonged hospitalization at an ICU. One patient in the adjunctive treatment group presented with advanced cerebral malaria and died at the ICU.
In summary, one out of five patients treated with automated RBC exchange died whereas all eight patients in the control group survived in this retrospective cohort study (Table
1). However, no direct comparison of survival rates between groups is justified since the absence of randomization led to preferential inclusion of patients with more advanced diseases in the RBC exchange group than in the conventional treatment arm as evidenced by the APACHE II score and lactate levels (Table
1).
Automated RBC exchange has been shown to be a safe method to rapidly reduce parasitaemia in critical ill patients. Since high-quality evidence from randomized controlled trials is missing, a final judgement on the risk benefit ratio of this adjunctive treatment for severe malaria is not possible to date. Although total biomass of parasites has been shown to correlate with severity of falciparum malaria [
28,
29], it remains unclear whether physical removal of parasites can improve patient’s outcome. However, it may be argued that rapid reduction of parasitaemia may benefit most patient populations at highest risk for adverse outcome including non-immune travellers and people living in regions of unstable malaria transmission [
11,
30]. Importantly, this risk benefit balance may depend on the employed anti-malarial drug and its speed of action. Whereas artemisinins show rapid parasite reduction in most endemic areas, quinine and clindamycin have a significantly slower onset of activity [
2,
6,
26,
31,
32]. However, parasite clearance times of artemisinin derivatives are dramatically prolonged in regions of South East Asia harbouring artemisinin resistant parasite strains, indicating again a potential shift in the risk benefit ratio of this potential adjunctive treatment in different regions of the world [
33,
34].
Limitations of this report include most importantly the retrospective study design since this does not guarantee comparability of patient groups in the absence of randomization and unified consistent definitions in the malaria patients’ general care (e.g indications for respiratory support, ventilation or renal replacement therapy may have slightly varied within ten years and between different intensive care facilities). Improving the level of evidence is therefore important; however the conduction of prospective studies evaluating this intervention is inherently difficult since high transmission regions often lack resources for purchasing and maintenance of equipment and the availability of safe blood products. Resource rich regions on the other hand generally do not have adequate numbers of patients with this condition. It is, therefore, likely that cohort studies will remain the main source of evidence in the near future.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
LAH, NW, SW and MR have designed the study, were responsible for data acquisition, performed data analysis and drafted the manuscript. All authors have reviewed and approved the final version of the manuscript.