Association between oral antimalarial medication administration and mortality among patients with Ebola virus disease: a multisite cohort study

The 2014–16 Ebola virus disease (EVD) epidemic in West Africa was the largest since the discovery of the virus in 1976 [1]. The mortality of EVD is high, averaging around 50%, although varies greatly, ranging from 25 to 90% across outbreak and treatment settings [1]. EVD is found in areas endemic to other infectious diseases and it can be difficult to differentiate EVD from other infectious diseases during initial evaluations [1]. Malaria, which is highly endemic in West Africa, can present with similar symptoms to EVD and was common among patients presenting to Ebola Treatment Units (ETUs), including among EVD-positive patients during the 2014–16 epidemic [1‐7]. The World Health Organization (WHO) guidelines recommend empiric antimalarial treatment for patients with suspected EVD [8], and as such antimalarial medications are commonly given in EVD treatment protocols [2‐4, 9‐23].

Multiple antimalarial medications have been shown to have activity against the Ebola Virus (EBOV) in laboratory studies. Amodiaquine, mefloquine, and chloroquine inhibit cellular entry of EBOV in vitro [24‐30]. However, conflicting results on the survival impacts with chloroquine treatment in EVD animal models have been reported [24‐26, 31‐34]. Furthermore, there exists limited clinical evidence for improvements in patient-centered outcomes with the use of empiric antimalarials in EVD care from outbreak settings [13, 23, 35].

During the course of the West Africa EVD outbreak, International Medical Corps (IMC) operated five ETUs, three in Sierra Leone and two in Liberia, with local partners, providing care to over 2,500 patients, including 478 with confirmed EVD. The three ETUs in Sierra Leone received patients who were triaged by a government run District Ebola Response Center, and the two ETUs in Liberia received all patients in their respective geographic areas. All of the ETUs followed IMC guidelines that were developed based on guidance from the WHO and Médecins Sans Frontières collected in previous outbreak settings [36]. Given the frequent use of antimalarial medications in EVD treatment and the limited data on clinical impact, this study evaluated the association between early treatment with the oral antimalarial agent combination artemether-lumefantrine on mortality in patients with EVD using the multinational IMC database.

From patients treated at the five ETUs, 478 had EVD, and of these 424 were eligible for inclusion (Fig. 1). The average age was 30.5 years (SD ± 18.7 years), and the population was 59.7% female. Overall, the most commonly observed clinical signs and symptoms were diarrhea 298 (70.3%), dysphagia 171 (40.3%), and dyspnea 135 (31.8%). High CT value (low viral loads) were found in 122 (43.4%) of the 281 patients with CT value results, while 159 (56.6%) had low CT values (high viral load). Malaria RDT assessments were carried out on 243 (57.3%) patients, with 48 (19.8%) having a positive result (Table 1).

Although there now exist effective, targeted therapeutics for use in EVD [41], supportive care and utilization of non-EVD specific antimicrobials is recommended in clinical management. However, many of these recommendations, including empirical treatment for malaria infection in EVD outbreaks, are based on limited patient-centered evidence. The data from the current multisite analysis demonstrates that when controlling for confounding factors, there is evidence that treatment with early oral antimalarials was associated with a decrease in mortality. These findings support current treatment guidelines which call for empiric antimalarial medications provided to patients being cared for in EVD outbreak settings.

There is only minimal and conflicting existing data on the impact of antimalarial agents on patient-centered outcomes in EVD [13]. A serologic analysis from a single ETU in Liberia found that plasmodium species parasitemia was associated with increased survival in EVD cases independent of treatment with antimalarial exposure [46]. Gignoux et al. found that among EVD-positive patients, the adjusted risk of death was reduced in patients treated with artesunate-amodiaquine versus artemether-lumefantrine [13]. This effect was strongest in EVD-positive patients without malaria co-infection [13]. The authors suggest that this effect may be explained either by activity of amodiaquine against the EBOV, as has been shown in vitro, or another confounding effect [13, 27, 29]. Garbern et al. found that, in analyzing data from a population that received community-level mass drug administrations of antimalarial treatments in West Africa, there was a no evidence of change in mortality among patients treated with artesunate-amodiaquine [35]. However, the findings from the current data support the hypothesis that antimalarials confer a beneficial effect in patients with EVD. Furthermore, in the sensitivity analysis, evidence of a protective association was found among patients lacking RDT results and a lack of protective association was noted using the same predictive model with paracetamol, a medication which had statistically significant association with oral antimalarial administration. Insignificant protective effects were noted across sensitivity analyses stratified by all CT value results as well as early in those receiving oral antimalarial treatment with artemether-lumefantrine in the subgroup of patients with a negative malaria RDT result in which the point estimates were in favor of mortality benefit with early oral antimalarial treatment. The lack of evidence in the sensitivity analysis may be due to a type II error derived from the smaller sample sizes in the stratified analyses or could be due to statistical chance. Although there was evidence of an association between a reduction in mortality and early oral anti-malarial treatment, the sensitivity analysis did not yield evidence of this association in RDT negative patients and the literature base is conflicting. The current results should be reinforced by additional prospective studies, drawn from EBV outbreak settings applying standardized international guidelines, for external validation evaluating treatment with artemether-lumefantrine among cases with and without malarial disease in EVD outbreak response care.

A study using computer modeling by Ekins et al. found that the antimalarials chloroquine and amodiaquine had favorable antimalarial binding to a protein site likely in EBOV and other filoviruses [47]. Similarly, Goyal et al. demonstrated in vitro inhibition of EBOV by chloroquine and amodiaquine [48]. Although research on artemether-lumefantrine does not specifically exist, the laboratory data indicate that there is a potential biologic activity for EBOV treatment with antimalarial drugs. Given the limited data on the patient-centered impacts of other antimalarial medications, further studies of not only artemether-lumefantrine but other commonly available malaria treatment regimens would be beneficial to inform this area of research.

IV administration of antimalarial medications are the preferred route when patients are experiencing severe illness due to malaria [49]. In severe malaria, gastrointestinal intolerance and erratic intestinal absorption can occur, which may make the use of oral medications unreliable or ineffective [50]. This can be further compounded by coinfection with EVD, where large-volume gastrointestinal fluid loss can occur [49]. Although the regression analysis evaluation of the addition of IV antimalarial treatment found no evidence of impact on mortality, the number of subjects receiving IV treatment were small and those receiving IV medications were likely to have more severe illness, which may not have been adequately controlled for in the modeling approach. As such, definitive understanding of the role of IV antimalarial treatments cannot be concluded from the available data and this treatment approach warrants further study.

This current study found that artemether-lumefantrine administered early in ETU care in an outbreak setting was associated with decreased mortality in patients with EVD. These data support the empiric provision of antimalarial medications, as currently recommended by international guidelines, however additional prospective studies, in epidemic settings following international guidelines, would be beneficial to validate these findings and enhance our understanding of the impacts of other routes of administration and alternative antimalarial treatment regimes.