sign in sign up
<<
Home , Kivu Ebola epidemic, Marburgvirus, Marburg virus disease, Reston virus

The new england journal of

Review Article

Edward W. Campion, M.D., Editor

Heinz Feldmann, M.D., Armand Sprecher, M.D., and Thomas W. Geisbert, Ph.D.​​

From the of Virology, Division bola (EBOV) was the best-known and most extensively stud- of Intramural Research, National Insti­ ied member of the family ( order) long before tute of and Infectious Diseases, Rocky Mountain , Hamilton, the shattering 2013–2016 West African . The virologic taxon Filo- MT (H.F.); Médecins sans Frontières, E viridae was defined in 1982 and subsequently amended regularly to accommodate Brussels (A.S.); and the Department of changes.1,2 These amendments and the unfortunate renaming of commonly used Microbiology and and Gal­ veston National Laboratory, University terms has made the filovirus confusing. Today, EBOV refers to the spe- of Medical Branch at Galveston, cific member virus of the type species in the ebolavirus Galveston (T.W.G.). Address reprint re­ (Fig. 1). quests to Dr. Feldmann at Rocky Moun­ tain Laboratories, 903 S. 4th St., Hamil­ The history of filoviruses largely involves outbreaks (Fig. 1). ton, MT 59840, or at feldmannh@­ ​­niaid​ virus (MARV) was the first filovirus to be discovered, in 1967.3 EBOV and .­nih​.­gov. virus (SUDV) were codiscovered in 1976 in the Democratic Republic of Congo N Engl J Med 2020;382:1832-42. (DRC) and , respectively.4,5 Subsequently, two additional DOI: 10.1056/NEJMra1901594 were found to be pathogenic in : Taï Forest virus (TAFV) in Côte d’Ivoire Copyright © 2020 Massachusetts Medical Society. in 1994 and Bundibugyo virus (BDBV) in in 2007.6,7 (RESTV), imported into the from the in 1989–1990, has long been the exception, since it appears to infect humans only subclinically.8,9 Unexpect- edly, it emerged in swine in the Philippines, and RESTV sequences were detected in pigs in China, raising fear about food safety.10,11 The zoonotic potential of RESTV remains unclear, and investigation of that potential is urgently needed. More recently, of new filoviruses were detected in and fish species. Lloviu virus (LLOV), genus cuevavirus, was sequenced from (Miniopterus schreibersii) in Spain and Hungary.12 Měnglà virus (MLAV) sequences were found in Chinese rousettus species representing the newly proposed genus, dianlovirus.13 Bombali virus (BOMV) sequences were discovered in bats from Sierra Leone, Guinea, and ; the virus is considered to be a new ebolavirus species.14 Fi- nally, fish-derived filoviruses constitute members of two new genera, striavirus and thamnovirus.2,15 Since no isolates are available, the unknown zoonotic and pathogenic potential of these new filoviruses is a concern.

Virologic Features Filoviruses are enveloped, filamentous particles with a nonsegmented, negative- sense RNA (Fig. 1).2,16-18 The genomic RNA is encapsidated by the nucleo- and, together with polymerase L, polymerase cofactor virion protein (VP) 35, and activator VP30, constitutes the nucleocapsid with replicase and transcriptase function. This structure interacts with the nucleocapsid-associ- ated VP24 and is surrounded by the matrix protein VP40, the driver of particle formation. The viral spike is formed by the sole trimeric transmembrane glycopro- tein and mediates viral entry; it also represents an important target for host im- mune responses.2,16-18 With all ebolaviruses, unlike , RNA editing

1832 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola

A Transmembrane glycoproteins Matrix (VP40 and VP24) Single-stranded, negative-sense RNA (NP encapsidation and VP30)

Polymerase complex Viral (VP35 and L) envelope

Nucleocapsid

B Epidemiology and Taxonomy

GUINEA IVORY SOUTH REPUBLIC OF COAST SUDAN THE CONGO SIERRA UGANDA LEONE LIBERIA

KENYA Genus Species Virus Country Origin (species) Isolate DEMOCRATIC GABON REPUBLIC OF CONGO Marburg , Kenya, Uganda, , Yes (MARV) human Marburgvirus Marburg Ravn virus Egyptian fruit bat, DRC, Kenya, Uganda Yes marburgvirus (RAVV) human ANGOLA

Bundibugyo Bundibugyo virus DRC, Uganda Human Yes ebolavirus (BDBV) ZIMBABWE Sudan Sudan virus DRC, South Sudan Human Yes ebolavirus (SUDV) Ebolavirus Taï Forest Taï Forest virus Ivory Coast Human Yes ebolavirus (TAFV)

Zaire Ebola virus DRC, Gabon, Guinea, Human Yes ebolavirus (EBOV) Liberia, RC, Sierra Leone

Figure 1. Biology, Epidemiology, and Taxonomy of Filoviruses. In Panel A, the electron micrograph (left) includes a computer­annotated viral particle (Ebola virus, Makona strain) showing the typical filamentous shape (blue) and the core structure (yellow). The core structure (right) comprises the genomic RNA encapsidated with the viral (NP) and linked with the viral transcriptase–replicase complex, which consists of virion 30 and 35 (VP30 and VP35) and RNA­dependent RNA polymerase, which is further associated with VP24. The structure is surrounded by a cell­derived membrane associated with VP40 on the inside, with the glycoprotein forming spikes on the outside of the viral envelope. In Panel B, the table shows the taxonomy and some epidemiologic, ecologic, and biologic properties of the members of the Filoviridae family that are pathogenic in humans. The map shows regions in Africa with reported outbreaks. DRC denotes Democratic Republic of Congo, and RC Republic of the Congo. results in glycoprotein expression. The primary macropinocytosis. Subsequently, cysteine pro- product of the glycoprotein is a secreted, teases in the endosome cleave the glycoprotein, nonstructural, soluble glycoprotein that has been allowing it to bind to the receptor Niemann– implicated in antigenic subversion.16 Pick C1 and initiating membrane fusion. This Filoviruses replicate in the cytoplasm of their process releases the genome into the , target cells.16-18 Viral particles attach to the cell where transcription and replication by the viral surface through the binding of glycoprotein to replicase occur through a positive-sense antige- multiple attachment factors, such as C-type lec- nome intermediate that functions as the tem- tins, and cell uptake occurs largely through plate for progeny negative-sense genomes. Viral

n engl j med 382;19 nejm.org May 7, 2020 1833 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Pre-Outbreak Outbreak Post-Outbreak

Outbreak end surveillance Education Capacity building Surveillance Education

Bat Reservoir host

Index case

Zoonotic Human-to-human Safe transmission transmission Education Vaccination

Intermediate or amplifying hosts (end hosts) Triage Convalescent Isolation Follow-up Chimpanzee prevention and control Counseling Treatment Treatment Counseling Gorilla (persistence)

Duiker

Figure 2. Outbreak Phases. Shown are the key elements of the three phases of an Ebola virus outbreak, including control measures.

positive-sense RNA transcripts are translated by foster therapeutic development. Future efforts host ribosomes, leading to intracellular assem- should focus on studying differences between bly of the nucleocapsid and budding from the these and the more recently discovered plasma membrane.2,16-18 filoviruses with unknown pathogenic potential. Knowledge of filoviruses is largely based on studies of EBOV and MARV, but all filoviruses Epidemiologic and Ecologic are thought to follow similar principles in cell Features biology, with certain distinctions such as altera- tions in genomic structure, interferon antago- Filoviruses are zoonotic maintained nistic property, and RNA editing.2,16-18 Over the in reservoir species, perhaps bats, with occasional years, life-cycle modeling systems have been es- spillover into humans and other mammals, tablished for EBOV and MARV that can be safely which may serve as end, intermediate, or ampli- used in level 1 and 2 laboratories.17 fying hosts (Fig. 2).19 This concept, however, has These systems have been instrumental in the been established only for MARV, with isolation understanding of filovirus replication and will from Rousettus aegyptiacus.20 Multiple bat species

1834 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola have been implicated as harboring ebolaviruses, route.18,25,26 Initially, infection is manifested as a but viral isolation has yet to be successful.21 This nonspecific febrile illness characterized by mal- is rather uncommon and may be explained by aise, fatigue, and . A few days later, low viral loads, low susceptibility of cell lines, or gastrointestinal manifestations develop in many inhibitors in bat tissue. Since bats are probably patients, with anorexia, nausea, vomiting, and hunted for food consumption in African coun- diarrhea (Fig. 3). Fluid losses can be substantial tries,22 the lack of reservoir identification is a — up to 10 liters per day. Other common signs concern with respect to preventive measures. and symptoms are dysphagia, headache, con- Human pathogenic filoviruses appear to be junctival injection, abdominal pain, , epizootic in regions close to the African equator and a maculopapular . abnormali- (Fig. 1). EBOV has caused most of the outbreaks ties occur in less than half of affected patients, in central and western African countries, where- usually manifested as bleeding from the gums, as SUDV, BDBV, and the marburgviruses MARV petechiae, oozing from sites, sub- and Ravn virus (RAVV) are more likely to cause conjunctival hemorrhage, and in vomitus disease in eastern Africa.16 RESTV is known to and stool.18,25-27 circulate in the Philippines and is likely to circu- Filovirus usually begin with the late elsewhere in Asia.9 With ongoing - deposition of viral particles on mucous mem- detection programs, new filoviruses will proba- branes and perhaps skin; occasionally infection bly be discovered. Evaluation of the pathogenic occurs percutaneously.18 After uptake of the viral potential for humans will help to determine the particles by dendritic cells and macrophages, public health threat posed by these filoviruses. filovirus replication potently shuts down early Because tools for modeling and predicting out- innate immune responses by blocking interferon breaks have become more sophisticated, future production and signaling.28,29 Dissemination prob- research should be able to focus on predicting ably occurs through the migration of dendritic the appearance of filoviruses in order to facili- cells to lymphoid tissues and release of virus tate public health preparedness. into the circulation, leading to infection of fixed Epizootic and endemic viruses circulate in macrophages in the liver, spleen, and other tis- animals and humans, respectively, but this has sues. Infection then spreads to adjacent hepato- not been convincingly shown for any filovirus.23 cytes, fibroblasts, and other cells.18,26,28-30 The frequent reemergence of EBOV in the DRC Disease is caused by direct effects of viral and Gabon around 2000 and that of MARV in replication and host responses to infection.18 Uganda a decade later supports the hypothesis Viral replication leads to the formation of intra- that these filoviruses are regionally epizootic.16 cellular inclusion bodies, followed by cell ly- The discovery of EBOV persistence in humans sis.30,31 Islands of appear in the liver, may indicate a potential to circulate temporarily with a commensurate elevation in liver enzyme in persons.24 Currently, however, neither EBOV levels. Myositis causes muscle aches and weak- nor other filoviruses can be considered to be ness, coupled with elevation of creatine kinase endemic anywhere; if they were, continuing hu- and aspartate aminotransferase levels. Renal man-to-human transmission could result, a dis- tubular cells and glomerular epithelium are af- turbing thought.23 fected, contributing to renal dysfunction (Fig. 3). Host responses include the production of pro- Disease and Pathogenesis inflammatory cytokines and chemokines by in- fected dendritic cells, macrophages, and mono- The clinical disease is no longer referred to as cytes.28-30,32 These immune responses cause T-cell Ebola or Marburg hemorrhagic fever but rather activation, which is rendered ineffective in se- as Ebola or (EVD or vere or fatal cases because of T-cell exhaustion MVD), which better reflects the variable symp- and apoptosis, followed by an impaired adaptive toms and downplays bleeding as a clinical hall- immune response. Proinflammatory mediators mark. The is 2 to 21 days cause endothelial-cell dysfunction, followed by (typically, 6 to 10 days) and probably depends on increased vascular permeability and fluid extrav­ the filovirus, as well as the exposure dose and asation.18,26,28,29,32 Infected macrophages produce

n engl j med 382;19 nejm.org May 7, 2020 1835 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Conjunctival injection, Psychological support Depression, anxiety

Antiemetic agents and rehydration Encephalitis Nausea and vomiting

Dysphagia Analgesic agents Headache Arthralgia Myositis Hepatic dysfunction

Pulmonary Rehydration

Renal dysfunction and antimalarial agents Endothelial dysfunction, vascular Bacteremia, parasitemia permeability, Gastrointestinal dysfunction Diarrhea and infection

Figure 3. Clinical Symptoms and Treatment. Shown are key organs affected during Ebola virus infections, as well as treatment options for management of .

tissue factor with fibrin deposition in the spleen, tection, including assays for the detection of vi- lymphoid tissues, glomeruli, and renal proximal ral genome, viral antigen, and host immune re- tubules. Consumption of clotting factors (from sponses, even in field operations.37,38 In the West disseminated microthrombi), endothelial dys- African EBOV epidemic, on-site, high-end sequenc- function, and inhibition of platelet function ing technology was implemented to improve contribute to coagulopathy.18,26,28-30,32 Microvas- outbreak response.39 In addition, simple bedside cular anomalies, , and further fluid tests to detect viral antigen have become avail- losses through vomiting and diarrhea ultimately able.37,38 The most widely used technique to diag- lead to tissue hypoperfusion and multiorgan nose acute infections is a quantitative real-time failure (Fig. 3).18,25,26,29,32 polymerase-chain-reaction assay (qRT-PCR), pref- During the 2013–2016 EBOV epidemic, muscu- erably targeting two distinct genome locations loskeletal pain, headache, encephalitis, and ocu- to minimize false negative results due to evolv- lar problems were noted in survivors and were ing genome mutations. The qRT-PCR assay is referred to collectively as the “post-Ebola syn- expected to be positive in symptomatic patients, drome.”33 Historically, filoviruses have been de- with increasing viremia in fatal cases. Since the tected in multiple body fluids, including breast assay may be negative early in the disease milk and semen, in survivors of infection. The course, however, follow-up testing is warranted persistence in semen, with the potential for in patients with initially negative tests who have sexual transmission more than 500 days after continuing symptoms. In the past, negative re- disease onset, is a serious concern. However, sults on at least two sequential tests have been transmission this long after disease onset is very required for discharge from the treatment cen- rare, with undetermined effects.34-36 ter. Despite improved laboratory diagnosis, indi- vidual EVD or MVD cases may still be difficult Diagnosis to diagnose, since clinical assessment is critical. In ill-prepared primary health care settings, di- Multiple techniques have been established for agnosis is further hampered by lack of technical laboratory diagnostic methods of filovirus de- capabilities. Technology transfer and training

1836 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola are still in their infancy in many African coun- investigational drugs, the monoclonal- tries, but awareness of filoviruses has grown, cocktails ZMapp (Mapp Biopharmaceutical) and and with simple, more reliable technologies, REGN-EB3 (Regeneron Pharmaceuticals), a sin- there is a prospect for improvement. gle (MAb114, Ridgeback Biotherapeutics), and (Gilead Sciences), Patient Care a small-molecule antiviral drug, have been given to hundreds of patients under the Monitored Emer- Care of patients has traditionally had three com- gency Use of Unregistered and Investigational ponents: supportive care to maintain or restore Interventions (MEURI) framework and in a ran- normal physiology, treatment of discomfort or domized, (Table 1).55,56 The interim distress, and presumptive treatment of any con- results of the Pamoja Tulinde Maisha (PALM) current, undiagnosed infections (Fig. 3).40,41 The trial54 suggested significantly improved survival most serious derangement is hypoperfusion for patients receiving MAb114 or REGN-EB3, as stemming from volume deficits due to gastroin- compared with those receiving remdesivir or testinal losses and vascular leakage, as well as ZMapp; the patients receiving ZMapp served as a intravascular . Thus, volume replace- control group on the basis of the results from ment through oral rehydration or intravenous the Partnership for Research on Ebola Virus in crystalloid infusion is the primary intervention. Liberia II (PREVAIL II) trial.46 Surprisingly, ZMapp With the advent of on-site biochemical testing, performed worse in the PALM trial than in the care now involves correcting levels PREVAIL II trial. The reasons remain unclear, and hypoglycemia, as well as meeting nutri- and further analysis may be needed to shed tional needs (Fig. 3). In advanced critical care more light on potential differences among treat- settings, additional support may be used, such ment groups. Interestingly, patients receiving care as parenteral nutrition or renal-replacement ther- and treatment earlier in the course of illness apy. Patients with headache, , and ar- fared better than those who entered Ebola treat- thralgias may be offered analgesic agents. Nau- ment units later, which had not been the case sea and vomiting may be relieved with the use of previously.57 Also surprising was the fact that antiemetic agents. No less important is psycho- extremes of age, which had adversely affected logical support to help patients cope with anxi- outcomes in past outbreaks,58 were not associ- ety, stress, and fear (Fig. 3).40,41 ated with differences in outcome in the PALM , whether a coincident tropical ill- trial.54 Patients in whom EVD developed despite ness or an infection due to EVD or MVD, has previous vaccination for EBOV had much better been documented.42,43 The frequency of coinfec- outcomes than patients who had not been vac- tions and the limited diagnostic options in most cinated.54 An analysis involving the patients who care settings have led to presumptive treatment received the same therapeutics under MEURI with antimalarial agents and broad-spectrum that were provided in the PALM trial showed . Future advances in diagnostics should strikingly similar results, despite the lack of allow for more targeted antimicrobial treatment. randomized treatment assignments and trial procedures. Investigational Treatment Overall, under MEURI and in the PALM trial, antibody specificity, initially higher antibody Efforts to develop specific treatments began doses, and perhaps the more favorable pharma- shortly after the discovery of filoviruses, but cokinetics of human may have con- when the West African Ebola epidemic struck, ferred an advantage. Remdesivir may have a efficacy had been shown only in preclinical delayed onset of action as compared with anti- studies. Clinical trials of promising bodies; however, the drug has broader applica- were carried out during the West African out- bility and, with antibodies, may represent a break, including convalescent plasma or whole synergistic therapeutic approach. The prelimi- blood,44,45 antibodies,46 small interfering ,47 nary results of MEURI and the PALM trial pro- and small-molecule inhibitors (),48 but vide hope for further incremental improvements none showed significant efficacy49,50 (Table 1). in the treatment of EVD with newer and im- In the recent Ebola outbreak in the DRC, four proved investigational therapeutics or different

n engl j med 382;19 nejm.org May 7, 2020 1837 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. The new england journal of medicine 51 44 53 54 54 54 54 47 52 52 48 55 46 45 Study Organization Group Siegel et al. et Siegel Mulangu et al. et Mulangu al. et Dunning Restrepo et al. et Henao - Restrepo al. et Kennedy Health World Kennedy et al. et Kennedy van Griensven et al. et Griensven van al. et Sahr al. et Mulangu PREVAIL II Writing PREVAIL II al. et Mulangu al. et Mulangu al. et Sissoko of 60 survivors to assess viral assess to survivors 60 of semen in shedding survival rate) survival benefit cy, 100% (95% CI, 79.3–100.0) CI, (95% 100% cy, in Ebola of decline of because Liberia) 95.8–98.5) CI, (95% because of decline of Ebola in Ebola of decline of because Liberia) benefit benefit rate) survival survival rate) survival rate) survival rate) survival benefit No. of Patients and Outcome Ongoing, with planned enrollment planned with Ongoing, 155 enrolled, 103 survived (66.5% survived 103 enrolled, 155 survival significant no enrolled; 12 5837 vaccinated; estimated effica ­ estimated vaccinated; 5837 eliminated 3 (phase vaccinated 500 97.5% efficacy, vaccinated; 93,965 500 vaccinated (phase 3 eliminated 3 (phase vaccinated 500 84 enrolled; no significant survival significant no enrolled; 84 survival significant no enrolled; 43 (64.9% survived 113 enrolled, 174 36 enrolled, 28 survived (77.8% survived 28 enrolled, 36 (49.5% survived 160 enrolled, 323 (46.9% survived 82 enrolled, 175 survival significant no enrolled; 126

Regimen (IV) sions of 200–250 ml 200–250 of sions each period a over given ml hr 1–4 of for a total of three doses three of total a for doses three of total a for (IV) (oral) 1–9 days on Once daily for 5 days (IV) days 5 for daily Once One dose (IV) dose One days 7 to up for daily Once Single injection (IM) injection Single (IM) injection Single (IM) injection Single Single injection (IM) injection Single Two consecutive IV transfu ­ IV consecutive Two of 450 transfusion IV One One dose every 3 days (IV) days 3 every dose One (IV) days 3 every dose One (IV) dose One days 9–13 for daily Once mg doses daily doses 1200 - mg Two PFU PFU PFU 7 7 7 particle 11 units Dose 100 mg 100 50 mg/kg 50 mg/kg 50 mg/kg 50 Unknown Unknown thereafter thereafter 0.3 mg/kg 0.3 150 mg/kg 150 2 × 10 2 × 10 2 × 10 dose; 100 mg dose; dose; 2400 mg dose; 2 × 10 200 mg loading mg 200 6000 mg loading mg 6000 Ebola () Ebola (Kivu) Ebola (Kivu) Ebola (Kivu) Ebola (Kivu) Ebola Ebola (Makona) Ebola Ebola (Makona) Ebola (Makona) Ebola (Makona) Ebola (Makona) Ebola (Makona) Ebola (Makona) Ebola (Makona) Ebola Ebola (Makona) Ebola Filovirus Species (Strain)

trial of (Guinea) vaccination ring of trial phase 2–3 trial (Liberia) trial 2–3 phase (DRC) vaccination ring of trial (Liberia) trial 2–3 phase study comparative study comparative States) United Guinea, (Liberia) trial history natural controlled, Leone) (Sierra controls historical historical controls (Guinea) controls historical Treatment and Study Design (Country) randomized cluster, open - label, rVSV - ZEBOV; ZEBOV; randomized, placebo - controlled randomized, rVSV - ZEBOV; randomized cluster, open - label, rVSV - ZEBOV; placebo - controlled randomized, ChAd3 - EBO Z; Antiviral nonrandomized plasma; Convalescent nonrandomized blood; Convalescent Leone, Sierra (Liberia, trial 2–3 phase ZMapp; (DRC) trial PALM ZMapp; (DRC) trial PALM MAb114; (DRC) trial PALM REGN - EB3; placebo - - blind, double (GS - 5734); Remdesivir (DRC) trial PALM (GS - 5734); Remdesivir with trial - group single (T - 705); Favipiravir with trial 2 phase single - group, TKM - 130803; Clinical Trials of and Antiviral Therapies for Ebola Virus Infection in Humans.* in Infection Virus Ebola for Therapies Antiviral and Vaccines of Trials Clinical 1. Table CI denotes confidence interval, IM intramuscular, IV intravenous, PALM Pamoja Tulinde Maisha, and PFU plaque - forming units. * 

1838 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola approaches (e.g., combination therapy). Notably, ample, when the objective is to target contacts of ZMapp, REGN-EB3, and MAb114 provide protec- infected patients, as well as potential future tion against EBOV alone, whereas more recent contacts of current contacts, well ahead of their preclinical success with strategically engineered, exposure.51,53 The prime–boost regimen may pro- next-generation human antibodies (i.e., MBP134, vide a more durable immune response, which FVM04, and CA45) has shown protection against takes longer to develop. Since the MVA boost EBOV, SUDV, and BDBV — a promising ad- contains glycoprotein sequences for multiple vance.59,60 filoviruses and a nucleoprotein sequence for TAFV,64 it may provide cross-protection. In gen- Vaccines eral, a prime–boost approach may be preferable for persons who are at risk for exposure because Vaccine development started in the 1970s with of their occupation, such as health care workers, inactivated viral preparations and was followed but the level of efficacy that can be achieved in the 1980s and 1990s by subunit and DNA vac- with the prime regimen alone is unknown. Nev- cine approaches.61,62 The past two decades have ertheless, the prime–boost regimen was recently seen intensified use of vectored vaccines and added to ring vaccination in the DRC as a second combined approaches. Except for the EBOV DNA approach, in the form of pop-up vaccination and and adenovirus-based vaccines, none of these targeted geographic vaccination to address secu- vaccine candidates had made it past the pre- rity concerns and community tensions (www​.­who​ clinical stage when the West African Ebola epi- .int/­ immunization/​­ policy/​­ position_papers/​­ interim​­ demic hit.61,62 This lack of preparedness for _ebola_recommendations_may_2019.​ pdf),­ in areas EBOV was finally corrected with several ap- where there is no active transmission.65 In the proaches that quickly moved to clinical trials future, issues such as vaccine efficacy, stability, (Table 1). One of the approaches is a single-shot, storage, transport, and administration, as well live-attenuated, vectored vaccine based on a re- as supply adequacy, need to be addressed for combinant vesicular stomatitis virus expressing several of the vaccine products. the glycoprotein (rVSV-ZEBOV-GP The protective immune responses to filovi- [ERVEBO, Merck]), which was successfully test- rus infections in nature are still not defined, ed for efficacy in a randomized trial in Guinea and correlates or even mechanisms of protec- during the West African epidemic.51,52 The vac- tion are unknown.61,62 Furthermore, the protec- cine, which was approved by the European Medi- tive immune response provided by vaccination cines Agency and the U.S. Food and Drug Ad- may well differ among vaccine candidates and ministration, has been widely administered in may also differ from the immune response to the DRC EBOV outbreak, with promising pre- natural infection. The closest correlate today liminary results (97.5% efficacy for vaccinees appears to be the total IgG response to EBOV with an onset of illness more than 10 days after glycoprotein.61,62 vaccination, and 88.1% for all those with EVD, 53,63 regardless of the timing of illness onset). Outbreak Management Advances in the development of other vaccines, such as the chimpanzee adenovirus 3 vaccine A comprehensive response to a filovirus out- (ChAd3-EBO-Z, GlaxoSmithKline)52 and the het- break is a complex undertaking (Fig. 2). The erologous prime–boost regimen containing the principle objectives are identifying and isolating Janssen AdVac for priming, followed by Bavarian suspected cases to prevent transmission and car- Nordic modified vaccinia Ankara (MVA-BN) ing for patients with EVD or MVD in order to technologies for boosting (Johnson & Johnson), save lives. Given the vague early clinical presen- are closing the gap between investigational and tation and its similarities to common tropical clinical use.64,65 illnesses, case identification requires reliable case Since safe and immunogenic vaccine candi- definitions, epidemiologic linkage, and labora- dates are available, the question remains what tory confirmation. strategy to choose for a specific target group. Another important component of the response The single-shot rVSV-ZEBOV appears to be valu- to an outbreak is follow-up of contacts of in- able when rapid immunity is needed — for ex- fected patients (Fig. 2). Contacts, historically

n engl j med 382;19 nejm.org May 7, 2020 1839 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

averaging 10 to 15 per patient, are monitored ment centers where experimental drugs are used, daily for 21 days, the maximum incubation pe- rumors of unsavory experimentation may begin riod, making contact follow-up a resource-inten- to spread. Unless the ground is prepared for sive effort.66 There is a high risk of disease intervention, actions are explained, and ques- transmission during traditional funerals, requir- tions answered, these new developments may be ing burial practices that minimize the risk of regarded as a threat. transmission while respecting cultural values.67 Disinfection of the environment (e.g., at funerals Perspective on the Future and treatment centers), contaminated by infect- ed persons and deceased bodies, is another im- We have come a long way since the epidemic portant disease-control activity.68 that devastated West Africa. We have managed Complicating nearly every aspect of outbreak to translate the fruits of laboratory research management is the crucial need to protect into new diagnostics, therapies, and vaccines. health care workers.69,70 With the small infec- Now we are facing the challenges of producing tious inoculum, few treatments, and severe dis- and implementing these tools and moving ease, a zero-tolerance practice has evolved. The them toward licensure, which has recently been iconic image of health care workers is their achieved for ERVEBO.75,76 To meet these chal- personal protective equipment. This equipment lenges, programs such as the Coalition for Epi- has many inconveniences, but none greater than demic Preparedness Innovations will be help- heat stress, which severely limits the time safely ful. The next great challenge is successfully spent in a care setting under tropical conditions. using these tools to help control outbreaks. The introduction of effective vaccines and Providing resources that are available to those therapeutics has great potential not only to im- most in need requires the trust of the recipient prove outcomes for patients but also to improve population. The current DRC outbreak shows outbreak control. The availability of these agents that trust is not a given and that the value of can provide a strong incentive for patients with these tools is not self-evident. EVD or MVD to rapidly bring themselves to the EBOV is just one of many neglected patho- attention of surveillance systems and for persons gens for which we are ill prepared. Can we gen- at risk for exposure to respond to contact trac- erate the impetus and secure the funding to ing, so long as the benefit of vaccines is clearly tackle other neglected tropical diseases? To do perceived. so would require an environment in which Nothing may be as important as community global strategies addressing infectious disease engagement and public perception.71 Disease and health could be rapidly and effectively im- transmission stops only when the community is plemented. Efforts to achieve this goal start with no longer caring for the sick in unprotected set- surveillance and include rapid communication; tings and burying the dead in an unsafe man- unrestricted information and reagent sharing; ner. Programs are aimed at encouraging the early collaborative engagement of government, population to quickly alert authorities about fe- industry, and academia; rapid and coordinated brile cases or unexplained rather than responses; community education and the foster- provide care at home or engage in unprotected ing of trust; and finally, the establishment and .67,72 Another important component is pub- maintenance of local response capacities. Are we lic education about the disease and control mea- there yet? No, but we are moving slowly in the sures. These messages will be effective only if right direction. the community trusts the messenger. Outbreaks The opinions, conclusions, and recommendations in this may occur in locations where mistrust of the report are those of the authors and do not necessarily repre- national government and outside intervention is sent the official positions of the National Institute of Allergy 73,74 and Infectious Diseases (NIAID) at the National Institutes of very high, as currently seen in the DRC. In Health, the University of Texas Medical Branch, or Médecins environments of mistrust, the introduction of sans Frontières. experimental countermeasures may actually bol- Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. ster further mistrust. When people die in vil- We thank Ryan Kissinger, NIAID Visual and Medical Arts, for lages where vaccines were deployed and in treat- help with an earlier version of the figures.

1840 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola

References 1. Kiley MP, Bowen ET, Eddy GA, et al. 18. Baseler L, Chertow DS, Johnson KM, Jacobs M, Bausch DG, Kaiser L. Ebola vi- Filoviridae: a taxonomic home for Mar- Feldmann H, Morens DM. The pathogen- rus shedding and transmission: review of burg and Ebola viruses? Intervirology esis of Ebola virus disease. Annu Rev current evidence. J Infect Dis 2016;​214:​ 1982;18:​ 24-32.​ Pathol 2017;​12:​387-418. Suppl 3:S177-S184.​ 2. Kuhn JH, Amarasinghe GK, Basler CF, 19. Munster VJ, Bausch DG, de Wit E, et al. 35. Sissoko D, Keïta M, Diallo B, et al. et al. ICTV virus taxonomy profile: Filo- Outbreaks in a rapidly changing Central Ebola virus persistence in breast milk af- viridae. J Gen Virol 2019;​100:​911-2. Africa — lessons from Ebola. N Engl J ter no reported illness: a likely source of 3. Siegert R, Shu HL, Slenczka W, Peters Med 2018;​379:​1198-201. virus transmission from mother to child. D, Müller G. On the etiology of an un- 20. Towner JS, Amman BR, Sealy TK, et al. Clin Infect Dis 2017;​64:​513-6. known human infection originating from Isolation of genetically diverse Marburg 36. Diallo B, Sissoko D, Loman NJ, et al. monkeys. Dtsch Med Wochenschr 1967;​ viruses from Egyptian fruit bats. PLoS Resurgence of Ebola virus in Guinea 92:​2341-3. (In German.) Pathog 2009;​5(7):​e1000536. linked to a survivor with virus persistence 4. Ebola haemorrhagic fever in Zaire, 21. Olival KJ, Hayman DT. Filoviruses in in seminal fluid for more than 500 days. 1976. Bull World Health Organ 1978;​56:​ bats: current knowledge and future direc- Clin Infect Dis 2016;​63:​1353-6. 271-93. tions. Viruses 2014;​6:​1759-88. 37. Coarsey CT, Esiobu N, Narayanan R, 5. Ebola haemorrhagic fever in Sudan, 22. Frick WF, Kingston T, Flanders J. A re- Pavlovic M, Shafiee H, Asghar W. Strate- 1976: report of a WHO/International view of the major threats and challenges to gies in Ebola virus disease (EVD) diagnos- Study Team. Bull World Health Organ global bat conservation. Ann N Y Acad Sci tics at the point of care. Crit Rev Micro- 1978;56:​ 247-70.​ 2019 April 2 (Epub ahead of print). biol 2017;43:​ 779-98.​ 6. Le Guenno B, Formenty P, Wyers M, 23. Sprecher A, Feldmann H, Hensley LE, 38. Shorten RJ, Brown CS, Jacobs M, Rat- Gounon P, Walker F, Boesch C. Isolation et al. Ebola virus is unlikely to become tenbury S, Simpson AJ, Mepham S. Diag- and partial characterisation of a new endemic in West Africa. Nat Microbiol nostics in Ebola virus disease in resource- strain of Ebola virus. Lancet 1995;​345:​ 2016;1:​ 16007.​ rich and resource-limited settings. PLoS 1271-4. 24. Dokubo EK, Wendland A, Mate SE, Negl Trop Dis 2016;​10(10):​e0004948. 7. Towner JS, Sealy TK, Khristova ML, et al. Persistence of Ebola virus after the 39. Quick J, Loman NJ, Duraffour S, et al. et al. Newly discovered Ebola virus asso- end of widespread transmission in Libe- Real-time, portable genome sequencing ciated with hemorrhagic fever outbreak ria: an outbreak report. Lancet Infect Dis for Ebola surveillance. Nature 2016;​530:​ in Uganda. PLoS Pathog 2008;​4(11):​ 2018;​18:​1015-24. 228-32. e1000212. 25. Leligdowicz A, Fischer WA II, Uyeki 40. Uyeki TM, Mehta AK, Davey RT Jr, 8. Jahrling PB, Geisbert TW, Dalgard TM, et al. Ebola virus disease and critical et al. Clinical management of Ebola virus DW, et al. Preliminary report: isolation illness. Crit Care 2016;​20:​217. disease in the United States and Europe. of Ebola virus from monkeys imported to 26. McElroy A. Understanding bleeding N Engl J Med 2016;​374:​636-46. USA. Lancet 1990;​335:​502-5. in Ebola virus disease. Clin Adv Hematol 41. Case definition recommendations for 9. Cantoni D, Hamlet A, Michaelis M, Oncol 2015;13:​ 29-31.​ Ebola or Marburg virus diseases. Geneva:​ Wass MN, Rossman JS. Risks posed by 27. Nicastri E, Kobinger G, Vairo F, et al. World Health Organization, August 9, Reston, the forgotten Ebolavirus. Ebola virus disease: epidemiology, clini- 2014 (https://apps​.who​.int/​iris/​bitstream/​ mSphere 2016;1(6):​ e00322-16.​ cal features, management, and preven- handle/​10665/​146397/​WHO_EVD_CaseDef 10. Barrette RW, Metwally SA, Rowland tion. Infect Dis Clin North Am 2019;​33:​ _14.1_eng​ .pdf?sequence=1).​ JM, et al. Discovery of swine as a host for 953-76. 42. de Wit E, Falzarano D, Onyango C, the Reston ebolavirus. Science 2009;​325:​ 28. Prescott JB, Marzi A, Safronetz D, et al. The merits of diagnostics 204-6. Robertson SJ, Feldmann H, Best SM. Im- during an Ebola virus . 11. Pan Y, Zhang W, Cui L, Hua X, Wang munobiology of Ebola and Lassa virus Emerg Infect Dis 2016;​22:​323-6. M, Zeng Q. Reston virus in domestic pigs infections. Nat Rev Immunol 2017;​17:195-​ 43. Carroll MW, Haldenby S, Rickett NY, in China. Arch Virol 2014;​159:​1129-32. 207. et al. Deep sequencing of RNA from blood 12. Negredo A, Palacios G, Vázquez- 29. Muñoz-Fontela C, McElroy AK. Ebola and oral swab samples reveals the pres- Morón S, et al. Discovery of an ebolavirus- virus disease in humans: pathophysiology ence of nucleic acid from a number of like filovirus in Europe. PLoS Pathog and immunity. Curr Top Microbiol Immu- pathogens in patients with acute Ebola 2011;7(10):​ e1002304.​ nol 2017;​411:​141-69. virus disease and is consistent with bacte- 13. Yang XL, Tan CW, Anderson DE, et al. 30. Martines RB, Ng DL, Greer PW, Rollin rial translocation across the gut. mSphere Characterization of a filovirus (Měnglà PE, Zaki SR. Tissue and cellular tropism, 2017;2(4):​ e00325-17.​ virus) from Rousettus bats in China. Nat and pathogenesis of Ebola and 44. van Griensven J, Edwards T, de Lam- Microbiol 2019;​4:​390-5. Marburg viruses. J Pathol 2015;​235:​153- ballerie X, et al. Evaluation of convales- 14. Goldstein T, Anthony SJ, Gbakima A, 74. cent plasma for Ebola virus disease in et al. The discovery of Bombali virus adds 31. Hoenen T, Shabman RS, Groseth A, Guinea. N Engl J Med 2016;​374:​33-42. further support for bats as hosts of ebola- et al. Inclusion bodies are a site of ebola- 45. Sahr F, Ansumana R, Massaquoi TA, viruses. Nat Microbiol 2018;​3:​1084-9. virus replication. J Virol 2012;​86:​11779- et al. Evaluation of convalescent whole 15. Shi M, Lin XD, Chen X, et al. The evo- 88. blood for treating Ebola virus disease in lutionary history of vertebrate RNA vi- 32. Falasca L, Agrati C, Petrosillo N, et al. Freetown, Sierra Leone. J Infect 2017;​74:​ ruses. Nature 2018;556:​ 197-202.​ Molecular mechanisms of Ebola virus 302-9. 16. Emanuel J, Marzi A, Feldmann H. pathogenesis: focus on cell death. Cell 46. The PREVAIL II Writing Group for the Filoviruses: ecology, molecular biology, Death Differ 2015;​22:​1250-9. Multi-National PREVAIL II Study Team. A and evolution. Adv Virus Res 2018;​100:​ 33. Jagadesh S, Sevalie S, Fatoma R, et al. randomized, controlled trial of ZMapp for 189-221. Disability among Ebola survivors and their Ebola virus infection. N Engl J Med 2016;​ 17. Hoenen T, Groseth A, Feldmann H. close contacts in Sierra Leone: a retrospec- 375:​1448-56. Therapeutic strategies to target the Ebola tive case-controlled cohort study. Clin In- 47. Dunning J, Sahr F, Rojek A, et al. Ex- virus life cycle. Nat Rev Microbiol 2019;​ fect Dis 2018;​66:​131-3. perimental treatment of Ebola virus dis- 17:593-606.​ 34. Vetter P, Fischer WA II, Schibler M, ease with TKM-130803: a single-arm

n engl j med 382;19 nejm.org May 7, 2020 1841 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved. Ebola

phase 2 clinical trial. PLoS Med 2016;​13(4):​ 58. Cherif MS, Koonrungsesomboon N, WHO_EVD_GUIDANCE_Burials_14​.2_eng​ e1001997. Diallo MP, et al. The predictor of mortal- .pdf;jsessionid=EF497CD6DB66A714B02F 48. Sissoko D, Laouenan C, Folkesson E, ity outcome in adult patients with Ebola 87F50B348C43?sequence=1). et al. Experimental treatment with favipi- virus disease during the 2014-2015 out- 68. Poliquin PG, Vogt F, Kasztura M, et al. ravir for Ebola virus disease (the JIKI Tri- break in Guinea. Eur J Clin Microbiol In- Environmental contamination and persis- al): a historically controlled, single-arm fect Dis 2017;​36:​689-95. tence of Ebola virus RNA in an Ebola proof-of-concept trial in Guinea. PLoS 59. Bornholdt ZA, Herbert AS, Mire CE, treatment center. J Infect Dis 2016;​214:​ Med 2016;​13(3):​e1001967. et al. A two-antibody pan-Ebolavirus cock- Suppl 3:S145-S152.​ 49. Cross RW, Mire CE, Feldmann H, tail confers broad therapeutic protection 69. Coca A, DiLeo T, Kim JH, Roberge R, Geisbert TW. Post-exposure treatments in ferrets and nonhuman . Cell Shaffer R. Baseline evaluation with a for Ebola and Marburg virus infections. Host Microbe 2019;​25(1):​49-58.e5. sweating thermal manikin of personal Nat Rev Drug Discov 2018;​17:​413-34. 60. Brannan JM, He S, Howell KA, et al. protective ensembles recommended for 50. Dodd LE, Follmann D, Proschan M, Post-exposure immunotherapy for two use in West Africa. Disaster Med Public et al. A meta-analysis of clinical studies ebolaviruses and Marburg virus in non- Health Prep 2015;9:​ 536-42.​ conducted during the West Africa Ebola human primates. Nat Commun 2019;​10:​ 70. Interim infection prevention and con- virus disease outbreak confirms the need 105. trol guidance for care of patients with for randomized control groups. Sci Transl 61. Feldmann H, Feldmann F, Marzi A. suspected or confirmed filovirus haemor- Med 2019;​11(520):​eaaw1049. Ebola: lessons on vaccine development. rhagic fever in health-care settings, with 51. Henao-Restrepo AM, Camacho A, Annu Rev Microbiol 2018;​72:​423-46. focus on Ebola. Geneva: ​World Health Longini IM, et al. Efficacy and effective- 62. Gross L, Lhomme E, Pasin C, Richert Organization, December 2014 (https:// ness of an rVSV-vectored vaccine in pre- L, Thiebaut R. development: apps​.who​.int/​iris/​bitstream/​handle/​10665/​ venting Ebola virus disease: final results systematic review of pre-clinical and clin- 130596/;jsessionid=3DB5CDD877C5B452​ from the Guinea ring vaccination, open- ical studies, and meta-analysis of deter- E1BBAC11BB462A62?sequence=1). label, cluster-randomised trial (Ebola Ça minants of antibody response variability 71. Wilkinson A, Parker M, Martineau F, Suffit!). Lancet 2017;​389:​505-18. after vaccination. Int J Infect Dis 2018;​74:​ Leach M. Engaging ‘communities’: an- 52. Kennedy SB, Bolay F, Kieh M, et al. 83-96. thropological insights from the West Af- Phase 2 placebo-controlled trial of two 63. Ebola virus disease, Democratic Re- rican Ebola epidemic. Philos Trans R Soc vaccines to prevent Ebola in Liberia. N Engl public of the Congo:​ external situation Lond B Biol Sci 2017;​372:​20160305. J Med 2017;377:​ 1438-47.​ report 74/2019. Geneva:​ World Health Or- 72. Agua-Agum J, Ariyarajah A, Aylward 53. Preliminary results on the efficacy of ganization, January 7, 2020 (https://www​ B, et al. Exposure patterns driving Ebola rVSV-ZEBOV-GP Ebola vaccine using the .who.int/​ publications​ -​ detail/­ ebola​ -​ virus­ ​ transmission in West Africa: a retrospec- ring vaccination strategy in the control of -disease­ -​ democratic­ -​ republic­ -​ of­ -​ congo­ ​ tive observational study. PLoS Med 2016;​ an Ebola outbreak in the Democratic Re- -­external​-­situation​-­report​-­74​-­2019). 13(11):​e1002170. public of the Congo:​ an example of inte- 64. Callendret B, Vellinga J, Wunderlich K, 73. Vinck P, Pham PN, Bindu KK, Bedford gration of research into epidemic response. et al. A prophylactic multivalent vaccine J, Nilles EJ. Institutional trust and misin- Geneva:​ World Health Organization, 2019 against different filovirus species is im- formation in the response to the 2018-19 (https://www​.who​.int/​csr/​resources/​ munogenic and provides protection from Ebola outbreak in , DR Congo: publications/ebola/​ ebola​ -​ ring­ -​ vaccination­ ​ lethal infections with Ebolavirus and a population-based survey. Lancet Infect -­results​-­12​-­april​-­2019​.pdf). Marburgvirus species in non-human pri- Dis.201919:52936.​ 54. Mulangu S, Dodd LE, Davey RT Jr, mates. PLoS One 2018;​13(2):​e0192312. 74. Kudra Maliro A-H, Anna C. Attackers et al. A randomized, controlled trial of 65. WHO adapts Ebola vaccination strat- kill doctor at hospital in Congo’s Ebola Ebola virus disease therapeutics. N Engl J egy in the Democratic Republic of the epicenter. AP News. April 19, 2019 (https:// Med 2019;​381:​2293-303. Congo to account for insecurity and com- apnews​.com/​3b99e6c8b646404288dc4a 55. Siegel D, Hui HC, Doerffler E, et al. munity feedback. Geneva:​ World Health 288b402044). Discovery and synthesis of a phosphorami- Organization, May 7, 2019 (https://www​ 75. Terry M. Merck’s Ervebo is the world’s date prodrug of a pyrrolo[2,1-f][triazin- .who​.int/​news​-­room/​detail/​07​-­05​-­2019​-­who​ first approved Ebola vaccine. BioSpace. 4-amino] adenine C-nucleoside (GS-5734) -adapts­ -​ ebola­ -​ vaccination­ -​ strategy­ -​ in­ -​ the­ ​ November 14, 2019 (https://www​.biospace​ for the treatment of Ebola and emerg- -­democratic​-­republic​-­of​-­the​-­congo​-­to​ .com/article/​ merck​ -​ s­ -​ ebola­ -​ vaccine­ ​ ing viruses. J Med Chem 2017;​60:​1648- -­account​-­for​-­insecurity​-­and​-­community​ -approved­ -​ in­ -​ europe/­ ).​ 61. -­feedback). 76. First FDA-approved vaccine for the 56. Nakkazi E. Randomised controlled 66. Saurabh S, Prateek S. Role of contact prevention of Ebola virus disease, mark- trial begins for Ebola therapeutics. Lancet tracing in containing the 2014 Ebola out- ing a critical milestone in public health 2018;​392:​2338. break: a review. Afr Health Sci 2017;​17:​ preparedness and response. Silver Spring, 57. Fitzpatrick G, Vogt F, Moi Gbabai OB, 225-36. MD:​ Food and Drug Administration, De- et al. The contribution of Ebola viral load 67. How to conduct safe and dignified cember 19, 2009 (https://www​.fda​.gov/​news​ at admission and other patient character- burial of a patient who has died from sus- -­events/​press​-­announcements/​first​-­fda​ istics to mortality in a Médecins Sans pected or confirmed Ebola or Marburg -approved­ -​ vaccine­ -​ prevention­ -​ ebola­ -​ virus­ ​ Frontières Ebola case management cen- virus disease. Geneva:​ World Health Or- -disease­ -​ marking­ -​ critical­ -​ milestone­ ​ tre, Kailahun, Sierra Leone, June-October ganization, October 2017 (https://apps​.who​ -public­ -​ health).­ 2014. J Infect Dis 2015;​212:​1752-8. .int/iris/​ bitstream/​ handle/​ 10665/​ 137379/​ ​ Copyright © 2020 Massachusetts Medical Society.

1842 n engl j med 382;19 nejm.org May 7, 2020 The New England Journal of Medicine Downloaded from nejm.org on October 27, 2020. For personal use only. No other uses without permission. Copyright © 2020 Massachusetts Medical Society. All rights reserved.