VIROLOGICAL DIAGNOSIS OF EBOLA VIRUS INFECTION
Universitaire Instelling Antwerpen & Instituut voor Tropische Geneeskunde,
Laboratory of Bacteriology & Virology, Nationalestraat 155, 2000 Antwerpen, Belgium.
It is impossible to consider the virological diagnosis of Ebola virus infection loose from the diagnosis of haemorrhagic fevers in general. The clinical picture of the disease indeed is too nonspecific to allow any hypothesis as to which virus may be responsible for any given case. At present, only the geographic origin of a specimen may give some indication as to the identity of the viruses involved (table 1). However, data of this nature may change with time and each case may represent the first occurrence of a known virus in a geographic area where it had not been encountered or even represent an entirely new virus as was the case with Ebola virus. It is for this reason that, when we received the first sample of the Zaire Haemorrhagic Fever epidemic of 1976, we decided to inoculate it on 3 different substrates in order to cover the maximum number of possibilities (table 2).
VIRUS ETIOLOGY OF HAEMORRHAGIC FEVER RELATED TO GEOGRAPHIC ORIGIN
a: Although numerous strains of Congo virus have been isolated in Africa, mostly from arthropods, the virus has never been identified as a cause of H.F. in that Continent.
FIVE SUBSTRATES FOR VIRUS ISOLATION FROM CASES OF HEMORRHAGIC FEVER
Newborn mice (I.C.)
Weanling mice (I.C. + I.P.)
TECHNIQUES USED IN THREE LABORATORIES WHO ISOLATED AND RECOGNIZED THE EBOLA VIRUS FROM THE 1976 EPIDEMIC
Signal for positivity
Time in days
Newborn mice IC
Weanling mice IC+IP
Newborn mice IC+IP
E.M. liver (U.S.)
IF serol. ident.
CPE cytopathic effect.
N.P. not performed.
f.t. floating technique.
U.S. ultra thin sections.
Where the weanling mice and newborn mice died respectively on the 5th and 7th days it became apparent that the virus involved was most probably not Lassa virus since the latter as a rule is not pathogenic for newborn mice.
Since material was forwarded to the laboratories of Porton Down and CDC Atlanta, it is worthwile to compare the techniques used in the 3 laboratories as available from published evidence 1,2,3 (table 3). All three laboratories characterized the Marburg like agent in their Vero cell cultures by electron microscopy, either by the simple and rapid floating technique or in ultra-thin sections of the infected cells.
Furthermore three laboratories also recognized the virus in ultrathin sections of liver tissue when this became available.
Finally the laboratory in Atlanta was able to perform the serological characterization of the virus within 24 hours of its isolation by the application of indirect immunofluorescence technique2.
From these results a number of important conclusions can be made for future rapid virus diagnosis of hemorrhagic fevers (table 4).
1. Continue to inoculate s.m. from cases of H.F.
2. Use low passage level of Vero cells.
3. Cell culture medium may be important.
4. Do not wait for CPE to do E.M. in emergency : daily E.M.
5. Consider E.M. of the patients serum.
6. Guinea pigs for highly contaminated specimens.
7. I.F. for rapid serological identification.
8. I.F. for virus detection.
9. Autopsy material also in glutaraldehyde.
10.Research optimum diagnostic strategy to allow rational prevention and protection.
1. We think that it is still necessary to inoculate newborn mice with diagnostic material from cases of H.F. Since no clues can be obtained concerning the nature of the virus responsible for a new case or a new epidemic of H.F. Indeed, many arboviruses especially flaviviruses do not multiply in Vero cells and even if they do they not necessarily produce CPE.
2-3. Not all passage levels of Vero cells are equally susceptible to some viruses while the culture medium may also be important . The Vero cells used in the Atlanta laboratory were of a lower passage number than those in Antwerp, and after inoculation with Ebola virus they show a cytopathic effect much sooner. It is thus advisable that Vero cells with a lower passage number be used. In our laboratory a complete CPE was observed in Vero cells maintained in a succinate/succinic acid buffered serumless medium.
4. Electron microscopy has become an indispensable tool for rapid virus diagnosis certainly in this field. It is not necessary to wait for the appearance of a CPE, to examine the supernatant in the E.M. for the presence of virus. In an emergency it can be considered to perform E.M. daily on a drop of the supernatant.
5. Direct E.M. examination of the patients serum should be done. As far as I know, this has not yet been tried and might offer the most rapid diagnosis possible.
6. The guinea pig is necessary only when highly contaminated specimens are submitted. This procedure is analogous to what is done to isolate other fastidious organisms from highly contaminated material as is done for leptospira (through inoculation into hamsters), borreliae (in mice) Mycobacterium tuberculosis and Veterans disease bacteria (in guinea pigs).
7. If E.M. is becoming at long last the standard procedure to recognize morphologically a virus in the same way the Gram stain does in diagnostic bacteriology, the indirect immunofluorescence in virology on its turn is becoming the equivalent of the slide agglutination test in bacteriology, as was beautifully illustrated by Johnson and coworkers when they found that Ebola was serologically different from Marburg(2). This implies that laboratories specializing in the diagnosis of H.F. should have at their disposal the necessary antisera to identify the possible agents.
8. I.F. could as well be applied to look for antigens in inoculated tissue culture and perhaps even in human specimens.
9. The experience with Ebola virus has clearly shown that autopsy material also may reveal rapidly the virus group to which the responsible virus belongs. Therefore it will be necessary to provide those who collect specimens with a fixative suitable for E.M.
10. Finally it will be necessary to set up the optimum strategies to cover all possible H.F. diagnoses. For this purpose additional research into some of the above mentioned aspects will be necessary.
I would like to emphasize that all this is only useful, if for the different H.F. their mode of transmission is known and the mode of excretion of the virus, so that a rapid diagnosis allows rational preventive and protective measures to be taken. But these matters are intimately intermingled : more information on transmission and contamination will only become available if rapid and relatively simple diagnostic procedures are available.
1. Bowen, E.T.W., Platt, G.S., Lloyd, G., Baskerville, A., Harris, W.J., Vella, E.E. (1977) Viral Hemorrhagic Fever in Southern Sudan and Northern Zaire. Preliminary studies on the aetiologic agent. Lancet 1, 571-573.
2. Johnson, K.M., Webb, P.A., Larige, V.E., Murphy, F.A. (1977) Isolation and partial characterization of a new virus causing acute hemorrhagic fever in Zaire. Lancet, 1, 569-571.
3. Pattyn, S.R., Jacob, W., Van der Groen, G., Piot, P., Courteille, G. (1977) Isolation of Marburg-like virus from a case of hemorrhagic fever in Zaire.Lancet, 1, 573-574.