Previous PageIndexNext Page


Bernhard Nocht Institute for Naval and Tropical Diseases, Clinical Department and Departments of Pathology and Virology, Bernhard Nochtstrasse 74, 2000 Hamburg 4, Germany.

In 1967 cases of a hitherto unknown hemorrhagic fever occurred in Marburg. Pathology of Marburg disease could be investigated well enough to become informed about specific organ damage and subsequent clinical syndrome. In Sudan an outbreak of a hemorrhagic fever similar to Marburg virus disease occurred. In contrast to 1967, local circumstances and logistics in the Sudan did not allow elaborate routine nor scientific tests. Thus, only limited material was to be investigated leaving many questions unanswered.


Postmortem biopsies were performed in two cases only to very limited extent. Biopsies of liver, heart, lung, spleen, kidney, and brain preserved in glutaraldehyde were processed by routine methods for histology and stained by Hematoxylin-Eosin, Prussian Blue, Ladewig, and Sudan-III-stain. Electron microscopic investigations were performed on EPON embedded material under standard conditions.

Peripheral blood smears of 11 patients were available at different stages of disease. Bone marrow aspirates of two patients could be evaluated. Blood smears and bone marrow aspirates were stained by panoptic staining May-Grunwald-Giemsa (Pappenheim).

Histology :

Liver: There was moderate hyperemia and marked edema in the center of the lobules with atrophy and dissociation of the liver cell cords in this area. In the periphery of the lobules the liver cells were loaded with fat droplets, as shown by Sudan III staining. Within the terminal plate - rarely in other areas of the lobule - individual or small groups of liver cells had undergone eosinophilic degeneration or necrosis. The portal tracts were enlarged and rather intensely infiltrated with lymphoid cells, histiocytes and - less intensely - plasma cells and eosinophils, focally interspersed with basophilic medullary junction.

Spleen: There was marked hyperemia and cellular depletion of the red pulp, and marked atrophy of the lymphoid follicles.

Myocardium (left ventricle) : There was a general, rather proteinaceous edema of locally varying degree of the interstitial connective tissue with focal and rather inconspicuous accumulations of inflammatory cells, similar to those found in the interstitial connective tissue of the liver and kidney.

Lung: The alveoli were focally dys- or atelectatic. The alveolar walls were generally and moderately thickened, due to an increase in cellularity and the deposit of fibrinoid material within and adjacent to the surface of the alveolar wall.

Brain: In the very small fragment of topographically not definable brain tissue no evidently pathological alterations could be recognized.

Studies of heart, lung, spleen, and kidney by electron microscopy did not show any virus particle or inclusion bodies.

A comparison of our histopathological observations with those previously made in a case of Ebola virus infection by Prof. Gigase at the "Institut de Médecine Tropicale 'Prince Leopold' at Antwerp , suggests almost identity as far as the parenchymal lesions are concerned. Furthermore, the liver lesions in both cases correspond well with the liver lesions which have been described in Marburg virus infections. The same can be stated with regard to the kidney, spleen, and heart lesions. The histopathology of the liver differs quite clearly from the histopathology observed in other virus diseases, e.g. infectious hepatitis and yellow fever. However, a differentiation from infections caused by Arena viruses (Lassa fever, Bolivian hemorrhagic fever, Argentinian hemorrhagic fever) appears to be difficult - if not impossible - by means of light microscopy only.


Slides of peripheral blood were available, taken at random at 20 different days of 11 patients. Considerable changes of the peripheral blood cells were seen. In some cases there was slight anisocytosis and occasionally small percentage of schistocytes. Evidently leukopenia existed at the beginning of the disease with increasing cell counts later predominantly of granulocytes. The most prominent finding was a shift to the left in the granulocytes, and up to 33% pseudo-Pelger forms. Very large cells with dark blue cytoplasm were identified as activated lymphocytes or lymphoblasts - also named "virocytes". Platelets were markedly decreased in some cases. The almost characteristic finding of pseudo-Pelger and so-called "virocytes'', as well as an increase of granules (probably remnants of nuclear decay), mainly in areas bordering the terminal plate of the lobule. The Kupffer cells were enlarged and they contained small granules of dark brown or black, iron-negative pigment. A few small granuloma enclosing fragments of Schistosoma eggs were also seen.

Electron microscopy: Hepatocytes showed enlarged mitochondria without cristae containing coarse granules or being empty. Many empty vacuoles indicating fat droplets in cytoplasm were seen. The spaces of Dissé were considerably enlarged. Frequently the plasmatic membranes of hepatocytes were not recognizable. Microvilli were absent. Inside the cytoplasm of altered hepatocytes meander like inclusions of 1 microm as well as single filamentous particles were found. However, inclusion bodies of nucleocapsids with regular formation were not present. The extracellular space contained virus particles and nucleocapsids not to be distinguished from Marburg virus particles.

Fig. 1. Ebola (Maridi) virus in human liver, 60.000 x

Kidney: The glomerula were inconspicuous. The epithelial cells of the tubules, particularly of the proximal portion of the nephron, exhibited varying degrees of granular, hydropic, and fatty degeneration and - focally necrosis and desquamation. The Bowman's space at the glomeruli and the lumen of the tubules were irregularly filled with amorphous proteinaceous precipitate. -ocal cellular infiltrations - analogous to those found in the portal tracts 'if the liver - were seen around blood vessels, particularly at the corticolymphocytes had its peak between the 6th and 10th day, though these features could be observed in smears from 3rd to 24th day of illness.

A 28 years old patient, who died at day 8 of his illness, showed at the 3rd day leukopenia, thrombocytopenia, 8% pseudo-Pelger, and 12' atypical lymphocytes. One day later leukopenia and thrombocytopenia were found again. There was a marked shift to the left of the granulocytes, and also a finding of 30' pseudo-Pelger forms.

The bone-marrow aspirate showed normal to increased cellularity. Red blood cell precursors were seen with little morphological changes only, such as atypical mitoses. Megakaryocytes were not at all decreased in number, and did not exhibit significant alterations morphologically. There was an increase in monocytes and plasma cells, as well as in eosinophiles to be expected in inhabitants of an area with high risk of parasitic diseases. However, the granulocyte precursors showed major alterations as described in detail. There were blocked mitoses, an 1 . increased number of necrotising cells, and a marked vacuolisation of granulocytic precursors, predominantly myelocytes and promyelocytes.

Some of the granulocyte precursors contained fragmented or very bizarre nuclei or twin nuclei.

Fig. 2. Bone marrow: twin nucleus and vacuolization of metamyelocyte.

Further morphologic peculiarities were seen in storage cells having phagozised decay and greenish inclusions, which could not be identified.

By morphological means only it is impossible to draw definite conclusions concerning the pathogenesis of the hematological changes. However, the marked alterations of the granulocyte precursors may suggest direct effect of virus or virus particle. The normal, at least not decreased megakaryocyte count and the peripheral thrombocytopenia may be correlated with peripheral sequestration of platelets, but is not sufficient to state this mechanism of thrombocytopenia. All described hematological alterations are very similar, if not identical, to those observed in cases with Marburg virus disease.


The description of pathologic anatomy of the Marburg virus disease shows many similarities with the results above(1). In addition, clinical features, morphology of virus, and hematological investigations parallel each other in both diseases, Marburg as well as Ebola (Maridi) virus disease(2). Thus, it may be permissible to use the information learned by the Marburg virus disease for understanding pathogenesis of clinical symptoms that occurred in Ebola (Maridi) virus infection. Unfortunately, mortality in Ebola (Maridi) infection was considerably higher than in Marburg virus disease. This could mean that, despite the similarity of morphological features and some similarities in the clinical course, Ebola (Maridi) virus infection may effect different target organs, as f.i. myocardium, than in Marburg disease. Therefore, more informations on pathology are needed urgently in order to improve possible clinical treatment for better survival.


Histo-pathological studies of biopsy material from two cases in Sudan by light and electron microscopy showed lesions in liver, heart, lung, spleen, and-kidney, most similar to the observations in Marburg virus disease. Furth more, peripheral blood smears and bone-marrow aspirates showed morphological alterations and quantitative changes of peripheral blood cells, that can be used to assist in the diagnosis of suspected hemorrhagic fever. Unfortunately they are inconclusive to explain the pathogenesis of some clinical symptoms, specifically the cause of the bleeding tendency.

1. Gedigk, P., Bechtelsheimer, H., Korb, G. (1971) Pathologic anatomy of the Marburg virus disease, editor G.A. Martini, R. Siegert, Springer, Berlin-Heidelberg-New York, 50-53.
2. Martini, G.A., and Siegert, R. (1971) Marburg Virus Disease, Springer, Berlin-Heidelberg-New York.

Previous PageTop Of PageNext Page