Skip Navigation

Evaluation of Immune Globulin and Recombinant Interferon-α2b for Treatment of Experimental Ebola Virus Infections

  1. P. B. Jahrling,
  2. T. W. Geisbert,
  3. J. B. Geisbert,
  4. J. R. Swearengen,
  5. M. Bray,
  6. N. K. Jaax,
  7. J. W. Huggins,
  8. J. W. LeDuc and
  9. C. J. Peters
  1. United States Army Research Institute of Infectious Diseases, Fort Detrick, Frederick, Maryland; Special Pathogens Branch, Centers for Disease Control and Prevention, Atlanta, Georgia
  1. Reprints or correspondence: Dr. Peter B. Jahrling, Scientific Advisor, Headquarters, USAMRIID, Ft. Detrick, Frederick, MD 21702-5011 (PBJ{at}Detrick.Army.Mil).
 
Next Section

Abstract

A passive immunization strategy for treating Ebola virus infections was evaluated using BALB/c mice, strain 13 guinea pigs, and cynomolgus monkeys. Guinea pigs were completely protected by injection of hyperimmune equine IgG when treatment was initiated early but not after viremia had developed. In contrast, mice were incompletely protected even when treatment was initiated on day 0, the day of virus inoculation. In monkeys treated with one dose of IgG on day 0, onset of illness and viremia was delayed, but all treated animals died. A second dose of IgG on day 5 had no additional beneficial effect. Pretreatment of monkeys delayed onset of viremia and delayed death several additional days. Interferon-α2b (2 × 107 IU/kg/day) had a similar effect in monkeys, delaying viremia and death by only several days. Effective treatment of Ebola infections may require a combination of drugs that inhibit viral replication in monocyte/macrophage-like cells while reversing the pathologic effects (e.g., coagulopathy) consequent to this replication.

The reemergence of Ebola (EBO) virus in the Democratic Republic of the Congo in 1995 after a 19-year hiatus brought worldwide attention to the virulence of this viral infection and to the lack of effective antiviral drugs or treatment strategies. The antiviral drug ribavirin has been used in controlled clinical trials to successfully treat viral hemorrhagic fevers (VHFs) caused by Lassa virus [1] and Hantaan virus [2]. Limited trials have shown promise for treatment of Junin virus infection [3, 4], Crimean-Congo hemorrhagic fever [5, 6], and Sabia virus infection [7], and significant activity was seen in primate models of Rift Valley fever [8] and Machupo viruses [9]. However, ribavirin was totally ineffective against filoviruses [10, 11]. Likewise, convalescent immune plasma has been used against other VHF agents with mixed success [12]. For EBO, passive immunization was tried only once to treat a laboratory worker who also received interferon (IFN) [13]; he survived, but the role of immune plasma in his recovery was not clear.

The efficacy of passive immunization for some VHFs is thought to correlate with the concentration of antibodies that neutralize viral infectivity [14, 15]. Neutralizing antibodies (NAs) typically evolve late after the acute infection; in contrast, specific antibodies detected by binding assays (e.g., immunofluorescence or ELISA) evolve early but do not predict NA concentrations.

At the time of the 1995 outbreak, essentially no information regarding passive immunization strategies for EBO virus was available. However, the existence of a potentially beneficial IgG that had been prepared by Russian investigators became known to the World Health Organization (WHO). These investigators had hyperimmunized horses with EBO virus [16] and prepared IgG, which had been used with reported success to treat experimentally infected baboons [17, 18]. Several hundred doses of this product were procured by WHO from the Russian association Epidbiomed; a portion of this material was submitted to United States Army Research Institute of Infectious Diseases (USAMRIID) for evaluation in cell culture neutralization assays and for protective efficacy studies in various animal models (mice, guinea pigs, and cynomolgus monkeys) of EBO infection. The present report expands upon our preliminary publication [19] of the initial testing of this IgG in cell culture and in monkeys treated with a single dose on the day of infection.

In addition to evaluating equine IgG, we also tested recombinant IFN-α2b in EBO-infected monkeys. Previous studies with IFN-α showed only a 10- to 100-fold reduction of virus replication in cell culture. It is not known whether a similar reduction in EBO virus replication in vivo correlates with clinical improvement; however, for respiratory syncytial virus, a 10- to 100-fold inhibition is beneficial [20]. Previous studies in primates did not test commercial IFN preparations at the maximum tolerated dose, which is the dose that produces the maximum antiviral effect. IFN-α2b is tolerated by primates at very high doses, allowing for evaluation at concentrations > 10-fold higher than those previously evaluated. Thus, in late 1996, IFN-2α2b was tested in a small number of cynomolgus monkeys when there was a threat of another EBO virus outbreak because of 2 infected patients in a Johannesburg, South Africa, hospital [21].

Previous SectionNext Section

Materials and Methods

IFN-α2b was obtained as a gift from Schering-Plough (Union, NJ). It was supplied as a dry powder and was reconstituted at 108 IU/mL. Monkeys weighing 2.9–3.1 kg received daily intramuscular (im) 0.6-mL injections containing 6 × 107 IU of IFN-α2b, beginning 18 h after virus inoculation. The IgG was obtained from WHO, which procured the material as previously detailed [19] from Epidbiomed, a Russian consortium. It was received as a frozen liquid in 3-mL glass vials labeled “immunoglobulin against EBO fever, from horse serum, liquid, (basic preparation), lot N34.”

For IgG protective efficacy studies, we used mice, strain 13 guinea pigs, and cynomolgus monkeys. For mouse studies, groups of 20 4-week-old BALB/c mice were inoculated im with 30 LD50 of mouse-adapted EBO virus [22]. They were treated 20–30 min later by subcutaneous (sc) inoculation of equine IgG in doses of 3.0, 0.3, and 0.03 mL/kg and observed along with infected, untreated controls for illness and death.

For guinea pig studies, groups of 5 strain 13 guinea pigs weighing 400–450 g were inoculated sc with 104 pfu of guinea pig—adapted EBO virus [23] and treated with 0.42 mL IgG (1.0 mL/kg) im several minutes or 4 days after virus inoculation. One group received IgG on both days 0 and 3.

For primate studies, cynomolgus macaques weighing 5–6 kg were inoculated im in the leg with 1000 pfu of the Zaire subtype of EBO (EBO-Z) virus, which was isolated from a human patient in 1995, and treated with 6 mL IgG distributed in three sites (both arms plus the leg that was not inoculated with virus). Six monkeys were treated on day 0 (the day of virus inoculation); 3 additional monkeys were treated on days 0 and 5. In a pretreatment study, smaller monkeys (3.0–3.2 kg) had been given the same 6-mL IgG dose to achieve a higher passive antibody titer. Blood was obtained for all monkeys under Telazol anesthesia at 2- or 3-day intervals for determination of infectious viremia, passive and active antibody titers, and standard hematologic and clinical pathology parameters. All terminally ill monkeys were euthanatized and necropsied for pathologic examination. Viral infectivity assays on plasma and tissue homogenates were performed as described by plaquing on Vero cell monolayers [24].

Passive anti-EBO antibody titers were obtained for the equine IgG—treated and for the experimentally infected animals by the indirect fluorescent antibody test (IFAT), using EBO—infected Vero cells [25], and by ELISA, using EBO-Z—infected Vero cell lysate as antigen [26] with appropriate species-specific secondary antibodies. ELISA was used to quantitate total horse IgG in monkey plasma, using a reference standard (Cappel Laboratories, Durham, NC) to generate a standard curve, as described [19]. Neutralization tests were performed by measuring plaque reduction in both “virus dilution:constant serum” and “constant virus:serum dilution” formats as previously described [25].

Tissues for transmission electron microscopy (TEM) from necropsied animals were immersion-fixed in 2% glutaraldehyde in 0.1 M Millonig's phosphate buffer. The TEM samples were post-fixed in 1% osmium tetroxide in 0.1 M Millonig's phosphate buffer, rinsed, stained with 0.5% uranyl acetate in ethanol, dehydrated in ethanol and propylene oxide, and embedded in POLY/BED 812 (Polyscience, Warrington, PA) resin as described [27]. Ultrathin sections were cut, placed on 200-mesh copper electron microscopy grids, stained with uranyl acetate and lead citrate, and examined by use of an electron microscope (100 CX; JEOL Ltd., Peabody, MA) at 80 KV.

Previous SectionNext Section

Results

Serologic testing of convalescent plasma and IgG

Following the original outbreak of EBO virus in 1976, plasma was obtained from convalescent patients. Some of this plasma had been sent to USAMRIID and was available for serologic testing. The titers for these plasmas were positive but relatively low in comparison with experimentally infected animal sera when tested by binding assays (ELISA or fluorescent antibody) (table 1). NA titers, expressed as a log10 NA index (LNI), were also positive but low; all neutralized <1.0 log10 pfu of EBOZ. Subsequent testing of three additional convalescent human plasmas, which were obtained 1 year after the 1995 outbreak in Kikwit, had similar unimpressive titers. In contrast, a hyperimmune guinea pig plasma had substantially higher ELISA and fluorescent antibody titers and an LNI of 3.3. The commercially available equine IgG was even more potent by these criteria and contained remarkably high NAs (LNI = 4.2; table 1).

Evolution of the NA response

Although experimentally infected, untreated cynomolgus monkeys never survive infection with EBO-Z, occasionally they survive infection with the Reston subtype of EBO (EBO-R) and develop serologic responses to the infection (figure 1). Despite an early indirect fluorescent antibody test response (detectable within 7 days following infection), the LNI response in a cynomolgus monkey that was a typical EBO-R survivor was delayed, evolving to detectable titers only after 75–90 days following infection. One year after infection, LNI titers >4.0 against EBO-R were measured, and LNI titers against EBO-Z were ∼ 1.1, which was higher than those observed in human convalescent plasmas. To determine whether this level of immunity corresponded with protection, we challenged the monkey with 1000 pfu EBO-Z virus. The monkey died 6 days after inoculation, with a viremia of 8.2 log10 pfu/mL, which was similar to viremia in control cynomolgus monkeys, as described in a later section.

Figure 1
View larger version:
Figure 1

Sequential titers as determined by indirect fluorescent antibody test (IFAT) and neutralizing antibody techniques (LNI; log10 neutralization index) for 1 monkey that survived experimental infection with EBO-R virus.

Efficacy testing of IgG in adult mice

We reported previously that the commercially available equine IgG was exceptionally pure, monomeric IgG with an LNI of 4.2 against EBOZ [19]. Since this NA titer was far greater than any other material available, it was reasonable to test it for protective efficacy in a variety of animal models. Our initial studies used adult BALB/c mice challenged with 30 LD50 or 1 pfu of mouseadapted EBO virus [22]. Mice were injected sc with equine IgG immediately after an intraperitoneal virus challenge. Figure 2 shows a typical outcome from a series of 4 experiments. The highest dose of IgG (3 mL/kg) lengthened by roughly 1 day the mean time to death of the treated mice that died, relative to that of untreated animals. However, all mice became ill, and only 5 of 20 survived. The morbidity and mortality of groups treated with 0.3 or 0.03 mL/kg did not differ from that of untreated controls.

Figure 2
View larger version:
Figure 2

% of BALB/c mice that survived intraperitoneal inoculation with 30 LD50 Ebola virus (mouse-adapted) followed 20 min later by subcutaneous treatment with various doses of equine anti-Ebola IgG or medium; each group had 20 mice.

Efficacy testing of equine IgG in guinea pigs

Although for humans, the supplier's recommended dose of equine IgG was 6.0 mL (one dose at ∼0.1 mL/kg), we selected a 10-fold higher dose (1.0 mL/kg) for our initial studies in guinea pigs. To maximize the chances for treatment success, we administered one dose to 2 groups of animals on day 0, and 3 days later, we administered another dose to 1 of the groups: Both groups of guinea pigs survived (table 2). In contrast, animals treated initially on day 4, after viremia had developed, failed to survive.

An examination of passive antibody titers and viremias in the IgG recipients revealed some correlates of protection (figure 3). Untreated, control guinea pigs developed viremias exceeding 4 log10 pfu/mL on day 4, and they developed 5 log10 pfu/mL by the day of death (mean, 8.7 days). In contrast, guinea pigs that received IgG on day 0 failed to develop detectable viremias at any time. Two days after receiving IgG, these animals had passive ELISA titers of 80–160 against EBO, and their passive titers of total equine IgG were ∼64,000. In contrast, guinea pigs that received their first IgG dose on day 4 were not protected (figure 4). In these animals, the IgG infusion had the immediate effect of reducing the viremia 10-fold, from the pretreatment titer of 4.2 log10 pfu/mL to 3.2 log10 pfu/mL 1 h later. At this time, passive total equine IgG levels were as high as those observed in the group treated on day 0 (64,000). However, specific anti-EBO ELISA titers were undetectable, suggesting that the specific IgG had combined with infectious virus. Suppression of viremia in this group was transient, however, since viremias increased to untreated control values by day 7, and mean time to death was not significantly delayed in this group. Although the treatment data for day 4 were disappointing, the successful treatment of guinea pigs with IgG on day 0 was sufficiently encouraging to justify extending these studies to primates.

Figure 3
View larger version:
Figure 3

Infectious viremias (log10 pfu/mL) in strain 13 guinea pigs that were inoculated subcutaneously with 10,000 pfu guinea pig-adapted EBO-Z virus and not treated (controls) [23] compared with viremias in inoculated animals that were treated intramuscularly with 1.0 mL/kg equine IgG on day of virus inoculation (n = 5 animals/group). Passively acquired total equine IgG and specific equine anti-Ebola IgG are shown for treated animals. Points are geometric means.

Figure 4
View larger version:
Figure 4

Infectious viremias (log10 pfu/mL) in 5 strain 13 guinea pigs that were inoculated subcutaneously with 10,000 pfu guinea pig-adapted EBO-Z virus [23] and treated intramuscularly with equine IgG (1.0 mL/kg) 4 days after virus inoculation. Passively acquired total equine IgG and specific equine anti-Ebola IgG are shown. Points are geometric means.

Efficacy testing of equine IgG in primates

Cynomolgus monkeys were treated im with 6.0 mL IgG at various times following EBO-Z inoculation. The first treatment group received IgG immediately after virus inoculation on day 0. As with the guinea pigs, these monkeys had good passive antibody titers on day 2, both specific anti-EBO IgG and total equine IgG (figure 5). As reported in our preliminary communication [19], the initial observations for this group were very encouraging. On day 5, the untreated controls were viremic and exhibited objective signs of illness, including fever, anorexia, and pronounced neutrophilia. In contrast, IgG recipients had no demonstrable viremia on day 5 and were clinically normal. However, by day 7, the IgG recipients were viremic and febrile, and their specific anti-EBO antibody titers had disappeared, leaving the impression that virus must have combined with the passively acquired antibody, thus depleting it and permitting viremia to evolve. These animals all died on days 7 and 8. The results suggested that a second infusion of IgG around day 5 might be sufficient to delay viremia onset by several additional days and permit the host protective immune defenses to fully activate. Before testing this hypothesis, however, it was necessary to investigate the pharmacokinetics of the IgG following the first and second infusions in uninfected monkeys.

Figure 5
View larger version:
Figure 5

Infectious viremias (log10 pfu/mL) in cynomolgus monkeys simultaneously inoculated intramuscularly with 1000 pfu EBO-Z virus and equine IgG (6.0 mL/6-kg animal) compared with viremias for infected, untreated controls (n = 6/group). Passively acquired total equine IgG and specific equine anti-Ebola IgG are shown for treated animals. Points are geometric means.

Three uninfected cynomolgus monkeys were inoculated with 6.0 mL IgG, and blood samples were obtained daily for determination of total equine IgG values. Following the first infusion, titers were maintained at ⩾80% of time zero values for ∼8 days (figure 6). After day 8, clearance was more rapid, suggesting immune clearance of IgG. Thus, immune clearance might not be a problem if the second infusion was given at day 5, but later infusions could be more problematic. To test the effect of a later infusion, we administered a second dose 2 months later to the same 3 monkeys. While the initial titers 1 day after infusion were the same, clearance of the second IgG was more rapid, although modest levels (>50%) were maintained for the first 4 days. Most important, however, was the fact that there were no signs of serum sickness following either the first or second dose of equine IgG. Thus, it was reasonable to test whether a second dose of IgG on day 5 following EBO infection might be beneficial.

Figure 6
View larger version:
Figure 6

Clearance of equine IgG from plasma of cynomolgus monkeys that were not infected with Ebola virus. Animals were inoculated intramuscularly with 6.0 mL of IgG on day 0, and blood was obtained daily for 10 days (first injection). Same animals received second injection 45 days after first injection. Values on day 1 were same following first and second injections. Subsequent values are expressed as fraction of day-1 values.

Viremia and passive antibody curves for the group receiving IgG on both days 0 and 5 indicated that viremia was totally suppressed beyond day 7 in these animals (figure 7), and 1 of the 3 treated monkeys did survive. However, the 2 remaining monkeys developed high viremia and died 2 days later. As before, the specific IgG titers fell as viremia increased. The 1 survivor never developed viremia and never seroconverted.

Figure 7
View larger version:
Figure 7

Infectious viremias (log10 pfu/mL) in 3 cynomolgus monkeys inoculated intramuscularly (im) with 1000 pfu EBO-Z virus and treated on infection days 0 and 5 with equine IgG (6.0 mL/6-kg animal, im). Passively acquired total equine IgG and specific equine anti-Ebola IgG are shown. Points are geometric means.

In an attempt to demonstrate more conclusively that IgG could afford more uniform protection, as measured by survival of a group of monkeys, we initiated a third treatment group in which monkeys received one IgG dose 2 days before challenge. Indeed, the pretreated animals were free of viremia past day 7 (figure 8); however, all 3 pretreated animals died on days 10–12 after virus inoculation.

Figure 8
View larger version:
Figure 8

Infectious viremias (log10 pfu/mL) in 3 cynomolgus monkeys pretreated with intramuscular (im) injection of equine IgG (6.0 mL/3-kg animal) 2 days before im inoculation of 1000 pfu EBO-Z virus. Passively acquired total equine IgG and specific equine anti-Ebola IgG are shown. Points are geometric means.

Table 3 compares 3 treatment groups with controls. In the first group, in which treatment was initiated on the day of challenge, all 6 monkeys died, with only a slight delay (compared with controls) in onset of infection and time to death. In the 2 subsequent groups, higher passive antibody titers were obtained because, in an attempt to maximize the chances for treatment success, we used smaller monkeys and the same dose of IgG (6 mL). Yet despite passive antibody titers of 640–1280, 2 of the 3 monkeys receiving IgG on both days 0 and 5 died, as did all 3 monkeys given IgG 2 days before challenge. These pretreated monkeys died later on days 10–12; however, on day 7, they were devoid of infectious viremia.

Figure 9
View larger version:
Figure 9

Electron micrograph of thin section through mesenteric lymph node of Ebola-infected monkey treated with anti-Ebola immunoglobulin on infection days 0 and 5. Ebola virus particle (arrow) is seen among fibrillar components of reticular fiber, while fiber has completely lost its protective sheath. Note Ebola virus inclusion material (*) in cytoplasm of degenerate fibroblast-like cell associated with fiber. Bar = 2.5 µm.

Figure 10
View larger version:
Figure 10

Infectious viremias (log10 pfu/mL) in cynomolgus monkeys inoculated intramuscularly with 1000 pfu EBO-Z virus and left untreated (controls; n = 3) or treated with interferon (IFN)-α2b (20 × 106 IU/kg; IFN recipients; n = 4) daily, beginning 18 h after virus inoculation, until death. Points are geometric means.

Table 1
View larger version:
Table 1

Antibody titers against Ebola virus in human plasmas and reference materials.

Table 2
View larger version:
Table 2

Efficacy of equine IgG as treatment for Ebola virus infection in guinea pigs.

Table 3
View larger version:
Table 3

Efficacy of equine IgG as treatment for Ebola virus infection in cynomolgus monkeys.

The concentrations of infectious EBO virus in the tissues of terminally ill monkeys that received IgG were the same as those in untreated controls (table 4). Likewise, conventional histopathologic and electron microscopic examination of these tissues revealed no differences in lesions between groups. The association of virus with fibrillar components of degenerate follicular reticular cells in lymph nodes of IgG recipients (figure 9) was striking and similar to that in controls.

Table 4
View larger version:
Table 4

Virus concentrations in tissues of monkeys experimentally infected with Ebola virus.

The EBO virus present in tissues of monkeys that received IgG was titrated against equine IgG to determine if it was neutralized as effectively as the inoculum virus. In the virusdilution format, a 1:100 dilution of equine IgG neutralized 4.0–4.2 log10 pfu of 3 virus isolates obtained from the livers of IgG recipients, the same neutralization achieved by inoculum virus. There was further evidence that viruses appearing late in IgG recipients were not neutralization escape mutants: Viremic plasma was obtained from IgG recipients 8–12 days after virus inoculation and used as the challenge virus in a serum dilution neutralization test to determine passive titers in recipients 3 days after administration of IgG. In all 6 animals tested, the same titers were obtained using both the potential “breakthrough” virus and the inoculum.

Finally, a small experiment was done to test the protective efficacy of recombinant IFN-α2b at a very high dose (20 × 106 U/kg), which was administered im every day (beginning 18 h after virus inoculation) to 4 cynomolgus monkeys. In contrast to the 2 controls, which both died on day 6, the 4 IFN recipients died on days 7, 7, 8, and 8, respectively. This slight extension in time to death was accompanied by a delay in the development of viremia in the IFN recipients (figure 10).

Previous SectionNext Section

Discussion

Treatment of EBO virus-inoculated cynomolgus monkeys with high-titered neutralizing IgG prepared from hyperimmunized horse serum delayed the onset of clinical signs and viremia. However, 11 of 12 IgG recipient monkeys eventually died with a disease clinically indistinguishable from EBO in untreated controls. In a preliminary paper [19], we reported that one dose of IgG (1.0 mL/kg on the day of infection) only delayed clinical onset; however, we speculated that pretreatment of monkeys before virus exposure might be beneficial. We further speculated that postexposure treatment efficacy might be augmented by a second IgG infusion ∼5 days after infection. Thus, we were somewhat disappointed to find, in the present study, that the beneficial effects of passive immunization were limited to a delay in onset of clinical signs and a postponement of death in the cynomolgus monkey model.

We tested this equine IgG preparation for protective efficacy in other animal models and obtained mixed results. Using mice, we obtained results similar to those for the monkeys, despite differences in routes of IgG and virus inoculation. Mice injected sc with 3 mL/kg of equine IgG within minutes after a 30 LD50 intraperitoneal virus challenge showed a 1-day delay in the onset of illness and death, and only 25% survived infection. This very large dose of equine IgG is equivalent, on a volume-to-weight basis, to inoculation of >200 mL in a 70-kg human. Smaller doses were totally ineffective in the mouse (0% survived); this argues against the possibility that reduced efficacy of high IgG doses might be related to a prozone phenomenon, as described for other passive immunization models [28].

In contrast, we were successful in demonstrating a beneficial effect of the IgG in the strain 13 guinea pig model. All guinea pigs survived when they were treated with IgG at 1.0 mL/kg im on the day of challenge with 10,000 pfu of guinea pig—adapted EBO-Z [23]. However, 5-fold lower IgG doses (0.2 mL/kg) were ineffective. Likewise, when treatment was delayed until viremia had developed on day 4, IgG was without benefit. Successful treatment of guinea pigs could relate to lower viremias and infectious virus burdens compared with those in mice and monkeys. Differences in host susceptibility might also explain why similar IgG preparations were reported to be effective in the Hamadryl baboon model [17, 18]. Even in baboons treated with higher-titered IgG than we tested, IgG efficacy was marginal; treatment was effective only when initiated within 1 h of a low-challenge inoculum (30 LD50), and IgG volumes <1.0 mL/kg were not tested (0.1 mL/kg is the manufacturer's recommended dose).

The failure of the two-dose IgG regimen (days 0 plus 5) to significantly prolong survival or delay viremia more than 1 additional day relative to the day 0—only regimen suggests that target tissues were already seeded with virus on day 5 and that this virus burden overwhelmed the passively acquired antibody. More surprising was the failure of pretreatment to prevent infection of the primary target tissues. Perhaps virus replicates initially in follicular reticular cells at the inoculation site [28] and through association with collagenous fibers (as illustrated in figure 10), spreads throughout the body without exposure to the bloodborne, passively acquired NAs. Sequential sacrifice studies, using mice and immunohistologic techniques for viral antigen and genome localization, are planned to address the question of viral spread in the presence of NAs. Marburg virus has been hypothesized to escape immunologic surveillance by replication in cells of the monocyte/macrophage series [30].

The results of these passive immunization experiments suggest that the presence of NAs is desirable but neither sufficient nor necessary to clear intracellular virus. It is probable that an effective cell-mediated immune response is necessary to clear intracellular virus from critical viral targets in the infected animal. Mice successfully protected from lethal EBO challenge by multiple vaccine approaches did not develop significant NA responses, even when challenged 45–90 days later with > 10,000 LD50 of virus (data not shown). The same was true of mice successfully treated with the antiviral drug carbocyclic 3-deazaadenosine when the spleen and liver had >107 pfu/g of virus at the onset of therapy. They only developed lower titers (10–20) of NA that did not significantly increase on rechallenge (Huggins JW, unpublished observations).

A role for cell-mediated immune response has been hypothesized to explain the reports concerning the successful treatment of 8 EBO patients with convalescent whole blood in the waning days of the Kikwit outbreak in 1995 [31]. If these transfusions were truly responsible for the recovery of the patients, it is unlikely that NA was the effector. The evolution of NAs is very slow (as modeled in experimentally infected monkeys; figure 1), and robust LNI titers are rarely found even in late convalescent sera (table 1). If cells are the important effectors, the problems of acquiring, processing, storing, and safely utilizing viable cells from recently convalescent EBO patients would remain. It is possible that more effective antibody therapeutics will be developed in the future; the potential for monoclonal antibodies developed from convalescent human bone marrow cells is especially promising, and such materials will be tested when they become available.

Far preferable to the use of an empirically defined and poorly standardized biologic potion would be an effective antiviral drug or a recombinant protein with defined immunologic activity. For these reasons, recombinant IFN-α2b was tested. In Vero cell culture, IFN-α2b inhibited EBO virus replication > 100-fold at 200 IU/mL. For treatment of monkeys, a high dose (20 × 106 IU/kg/day) was selected, based on the high tolerance for IFN-α2b in macaques [32]. Despite this high IFN-α2b dose, EBO replication, as reflected by viremia, was only slightly delayed, and IFN recipients died only 1–2 days later than controls (figure 10). Far more promising are some antiviral drugs, especially those of the S-adenosylhomocysteine hydrolase inhibitor class [33]. Independently, IgG and IFN had beneficial effects in delaying the onset of viremia and extending the time to death. However, the progression of disease was only delayed, not eliminated.

Successful elimination of EBO virus following infection may require combination therapy. For Lassa virus, experimentally infected monkeys were cured by a combination of immune plasma plus ribavirin (initiated late in the disease course); neither treatment was effective alone [34]. By analogy, for EBO infections, IgG might work synergistically with an antiviral drug. Likewise, effective treatment might require both an inhibitor of viral replication and an inhibitor of the process of disseminated intravascular coagulation (DIC) that characterizes terminally ill patients. The association of virus with collagen (figure 9) and evidence of fibrin deposition in EBO virus—infected cynomolgus monkeys [35, 36] are compatible with DIC, although no formal study has been done of DIC and coagulopathy in this animal model or in human patients.

We are presently conducting a therapeutic efficacy evaluation of recombinant tissue factor antagonists both alone and in combination with IgG and candidate antiviral drugs. While effective treatment strategies may emerge through this empirical approach, systematic design will require a detailed understanding of the pathogenesis of filovirus disease coupled with the pharmacokinetics of candidate therapeutic formulations in realistic animal models. Antibody-based therapeutics are still viable candidates, but it is likely that mixtures of human monoclonal antibodies will have higher specific activities and more favorable pharmacokinetic properties than the currently available equine IgG. An investment in such human monoclonals is warranted for use either alone or in combination with another antiviral or palliative drug.

Previous Section
 

References

  1. 1.↵
    1. McCormick JB,
    2. King IJ,
    3. Webb PA,
    4. et al
    . Lassa fever. Effective therapy with ribavirin. N Engl J Med 1986;314:20-6.
    MedlineWeb of Science
  2. 2.↵
    1. Huggins JW,
    2. Hsiang CM,
    3. Cosgriff TM,
    4. et al
    . Prospective, double-blind concurrent, placebo-controlled clinical trial of intravenous ribavirin therapy of hemorrhagic fever with renal syndrome. J Infect Dis 1991;164:1119-27.
    Abstract/FREE Full Text
  3. 3.↵
    1. McKee KT Jr.,
    2. Huggins JW,
    3. Trahan CJ,
    4. Mahlandt BG
    . Ribavirin prophylaxis and therapy for experimental Argentine hemorrhagic fever. Anti-microb Agents Chemother 1988;32:1304-9.
    Abstract/FREE Full Text
  4. 4.↵
    1. Enria DA,
    2. Briggiler AM,
    3. Levis S,
    4. Vallejos E,
    5. Maiztegui JI,
    6. Canonico PG
    . Tolerance and antiviral effect of ribavirin in patients with Argentine hemorrhagic fever. Antiviral Res 1988;32:353-9.
  5. 5.↵
    1. van de Wal BW,
    2. Joubert JR,
    3. van Eeden PJ,
    4. King JB
    . A nosocomial outbreak of Crimean-Congo hemorrhagic fever at Tygerberg Hospital. Part IV. Preventive and prophylactic measures. S Afr Med J 1985;68:729-32.
    MedlineWeb of Science
  6. 6.↵
    1. Fisher-Hoch SP,
    2. Khan JA,
    3. Rehman S,
    4. Mirza S,
    5. Khurshid M,
    6. McCormick JB
    . Crimean Congo-haemorrhagic fever treated with oral ribavirin. Lancet 1995;346:472-5.
    CrossRefMedlineWeb of Science
  7. 7.↵
    1. Barry M,
    2. Russi M,
    3. Armstrong L,
    4. et al
    . Brief report: treatment of a laboratory-acquired Sabia virus infection. N Engl J Med 1995;333:294-6.
    CrossRefMedlineWeb of Science
  8. 8.↵
    1. Peters CJ,
    2. Reynolds JA,
    3. Slone TW,
    4. Jones DE,
    5. Stephen EL
    . Prophylaxis of Rift Valley fever with antiviral drugs, immune serum, an interferon inducer, and a macrophage activator. Antiviral Res 1986;6:285-97.
    CrossRefMedlineWeb of Science
  9. 9.↵
    1. Stephen EL,
    2. Jones DE,
    3. Peters CJ,
    4. Eddy GA,
    5. Loizeaux PS,
    6. Jahrling PB,
    7. Smith RA,
    8. Kirkpatrick W
    . Ribavirin: a broad spectrum antiviral agent. New York: Academic Press; 1980. Ribavirin treatment of toga-, arena-, and bunyavirus infections in subhuman primates and other animal species; p. 169-83.
  10. 10.↵
    1. Huggins JW
    . Prospects for treatment of viral hemorrhagic fevers with ribavirin, a broad-spectrum antiviral drug. Rev Infect Dis 1989: suppl 11:S750-61.
  11. 11.↵
    1. Ignat'ev GM,
    2. Strel'tsova MA,
    3. Agafonov AP,
    4. Kashentseva EA,
    5. Prozorov-skii NS
    . Experimental study of possible treatment of Marburg hemorrhagic fever with desferal ribavirin, and homologous interferon. Vopr Virusol 1996;41:206-9.
    MedlineWeb of Science
  12. 12.↵
    1. Monath TP
    . Ribavirin interferon, and antibody approaches to prophylaxis and therapy for viral hemorrhagic fevers. Curr Opin Infect Dis 1990;3:824-33.
    CrossRefWeb of Science
  13. 13.↵
    1. Bowen ETW,
    2. Lloyd G,
    3. Platt G,
    4. Mcardell LB,
    5. Webb PA,
    6. Simpson DIH,
    7. Pattyn SR
    . Ebola virus hemorrhagic fever. Amsterdam: Elsevier/North-Holland Biomedical Press; 1978. Virological studies on a case of Ebola virus infection in man and monkeys; p. 95-100.
  14. 14.↵
    1. Jahrling PB,
    2. Peters CJ
    . Passive antibody therapy of Lassa fever in cynomolgus monkeys: importance of neutralizing antibody and virus strain. Infect Immun 1984;44:528-33.
    Abstract/FREE Full Text
  15. 15.↵
    1. Maiztegui JI,
    2. Fernandez NJ,
    3. deDamilano AJ
    . Efficacy of immune plasma in treatment of Argentine hemorrhagic fever and association between treatment and a late neurological syndrome. Lancet 1979;2:1216-7.
    CrossRefMedlineWeb of Science
  16. 16.↵
    1. Krasnyanskii VP,
    2. Mikhailov VV,
    3. Borisevich IV,
    4. Gradoboev VN,
    5. Evseev AA,
    6. Pshenichnov VA
    . Preparation of hyperimmune horse serum to Ebola virus. Vopr Virusol 1994;39:91-2.
    MedlineWeb of Science
  17. 17.↵
    1. Borisevich V,
    2. Mikhailov VV,
    3. Krasnyansky VP,
    4. et al
    . Creation and study of immunoglobulin to Ebola fever. Vopr Virusol 1995;40:270-3.
    MedlineWeb of Science
  18. 18.↵
    1. Mikhailov VV,
    2. Borisevich IV,
    3. Chernikova NK,
    4. Potryvaeva Krasnyanskii VP
    . An evaluation of the possibility of Ebola fever specific prophylaxis in baboons (Papio hamadryas). Vopr Virusol 1994;39:82-4.
    MedlineWeb of Science
  19. 19.↵
    1. Jahrling PB,
    2. Geisbert J,
    3. Swearengen JR,
    4. et al
    . Passive immunization of Ebola virus—infected cynomolgus monkeys with immunoglobulin from hyperimmune horses. Arch Virol 1996;11 (suppl):135-40.
  20. 20.↵
    1. Englund JA,
    2. Piedra PA,
    3. Ahn YM,
    4. Gilbert BE,
    5. Hiatt P
    . High-dose, short-duration ribavirin aerosol therapy compared with standard ribavirin therapy in children with suspected respiratory syncytial virus infection. J Pediatr 1994;125:635-41.
    CrossRefMedlineWeb of Science
  21. 21.↵
    WHO. Ebola haemorrhagic fever. Wkly Epidemiol Rec 1996;48:366.
  22. 22.↵
    1. Bray M,
    2. Davis K,
    3. Geisbert T,
    4. Schmaljohn C,
    5. Huggins J
    . A mouse model for evaluation of prophylaxis and therapy of Ebola hemor-rhagic fever. J Infect Dis 1998;178:651-61. 1999; 179 (suppl 1):S248–58.
    Abstract/FREE Full Text
  23. 23.↵
    1. Connolly BM,
    2. Steele KE,
    3. Davis KJ,
    4. et al
    . Pathogenesis of experimental Ebola virus infection in guinea pigs. J Infect Dis 1999;179 suppl 1:S203-17.
    Abstract/FREE Full Text
  24. 24.↵
    1. Jahrling PB,
    2. Geisbert TW,
    3. Dalgard DW,
    4. et al
    . Preliminary report: isolation of Ebola virus from monkeys imported to the USA. Lancet 1990;335:502-5.
    CrossRefMedlineWeb of Science
  25. 25.↵
    1. Jahrling PB,
    2. Murray PR,
    3. Baron EJ,
    4. Pfaller MA,
    5. Tenover FC,
    6. Yolken RH
    . Manual of clinical microbiology. 6th ed. Washington, DC: ASM Press; 1995. Filoviruses and arenaviruses; p. 1068-81.
  26. 26.↵
    1. Ksiazek TG
    . Laboratory diagnosis of filovirus infections in nonhuman primates. Lab Animal 1991;20:34-46.
  27. 27.↵
    1. Geisbert TW,
    2. Jahrling PB,
    3. Hanes MA,
    4. Zack PM
    . Association of Ebola-related Reston virus particles and antigen with tissue lesions of monkeys imported to the United States. J Comp Pathol 1992;106:137-52.
    CrossRefMedlineWeb of Science
  28. 28.↵
    1. Casadevall A,
    2. Scharff MD
    . Serum therapy revisited: animal models of infection and development of passive antibody therapy. Antimicrob Agents Chemother 1994;38:1695-702.
    FREE Full Text
  29. 29.
    1. Davis KJ,
    2. Anderson AO,
    3. Geisbert TW,
    4. et al
    . Pathology of experimental Ebola virus infection in African green monkeys: involvement of fibro-blastic reticular cells. Arch Pathol Lab Med 1997;121:805-19.
    MedlineWeb of Science
  30. 30.↵
    1. Feldmann H,
    2. Bugany H,
    3. Mahner F,
    4. Klenk H,
    5. Drenckhahn D,
    6. Schnittler H
    . Filovirus-induced endothelial leakage triggered by infected monocytes/macrophages. J Virol 1996:2208-14.
  31. 31.↵
    1. Mupapa K,
    2. Massamba M,
    3. Kibadi K,
    4. et al
    . Treatment of Ebola hemorrhagic fever with blood transfusions from convalescent patients. J Infect Dis 1999;179 suppl 1:S18-23.
    Abstract/FREE Full Text
  32. 32.↵
    1. Loginova SI,
    2. Bogatikov GV,
    3. Prikhod'ko AV,
    4. Polozov AI
    . Study of the effectiveness of recombinant α2-interferon and its inducers in primates with arthopoxvirus infection. Vopr Virusol 1997;42:186-99.
    MedlineWeb of Science
  33. 33.↵
    1. Huggins J,
    2. Zhang ZX,
    3. Bray M
    . Antiviral drug therapy of filovirus infections: S-adenosylhomocysteine hydrolase inhibitors inhibit Ebola virus in vitro and in a lethal mouse model. J Infect Dis 1999;179 suppl 1:S240-7.
    Abstract/FREE Full Text
  34. 34.↵
    1. Jahrling PB,
    2. Peter CJ,
    3. Stephen EL
    . Enhanced treatment of Lassa fever by immune plasma combined with ribavirin in cynomolgus monkeys. J Infect Dis 1984;149:420-7.
    Abstract/FREE Full Text
  35. 35.↵
    1. Jaax NK,
    2. Davis KJ,
    3. Geisbert TW,
    4. et al
    . Lethal experimental infection of rhesus monkeys with Ebola-Zaire (Mayinga) virus by the oral and conjunctival route of exposure. Arch Pathol Lab Med 1996;120:140-55.
    MedlineWeb of Science
  36. 36.↵
    1. Jahrling PB,
    2. Geisbert TW,
    3. Jaax NK,
    4. Hanes MA,
    5. Ksiazek TG,
    6. Peter CJ
    . Experimental infection of cynomolgus macaques with Ebola-Reston filoviruses from the 1989–1990 US epizootic. Arch Virol 1996;11(suppl):115-43.
| Table of Contents

This Article

  1. J Infect Dis. (1999) 179 (Supplement 1): S224-S234. doi: 10.1086/514310
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most