Hepatic ultrasound in a population with high incidence of invasive amoebiasis: evidence for subclinical, self-limited amoebic liver abscesses



About 10% of successfully treated amoebic liver abscesses (ALA) do not completely resolve and can be detected by ultrasound as typical residual liver lesions. The frequency of these residues should be an indicator for the prevalence of ALA in a given population, and may help to solve the question whether non-clinical, self-healing ALAs occur. We have performed hepatic ultrasound in 1036 adult individuals living in a high-risk area for ALA in Central Vietnam and identified typical ALA residual lesions in about 1.2% of the subjects. As expected, these lesions were associated with positive amoeba serology and were found in 11.9% of individuals with a previous ALA history. However, more than 50% of the residues were identified in individuals who had never developed symptoms suspected for ALA and who never received any ALA specific treatment, suggesting that subclinical, self-limited hepatic amoeba abscesses truly exist.

Amoebic liver abscess (ALA) is the most common extraintestinal manifestation of invasive amoebiasis. In contrast to intestinal amoebic disease, which has a broad spectrum of clinical presentations, ranging from asymptomatic gut ulcers and mild diarrhoea to haemorrhagic and fulminant colitis, the clinical course of ALA appears to be more unique. More than 90% of patients experience weight loss, fever and abdominal pain, and most abscesses (>80%) are located in the right liver lobe with strong preference for segments 6–8 (Maltz & Knauer 1991; Pham Van et al. 1996). There is a strong bias towards gender and age as more than 80% of ALA patients are adult males; the peak incidence occurs around the age of 40 years (Pham Van et al. 1996; Blessmann et al. 2002). ALA can be efficiently treated with metronidazole and in the vast majority abscesses completely resolve within a few months to a normal parenchymal pattern as revealed by hepatic ultrasound. However, in about 10% of successfully treated patients, residual lesions remain with the sonographic finding of round, hypoechoic areas with a hyperechoic wall (Ralls et al. 1983; Sheen et al. 1989; Sharma et al. 1995).

Most of the present knowledge about ALA has been accumulated based on the analysis of patients with respective clinical presentations. But these patients might represent only the extreme end of amoeba-induced liver involvement. Whether for example subclinical ALA exists, and whether ALA can heal spontaneously, as was suggested by Rogers as early as 1902 before the advent of specific therapy (Rogers 1902), has not been determined.

Using a portable ultrasonograph (Shimadzu SDU – 350A equipped with a convex 3.75 MHz/57R transducer), we examined the liver of 1036 adult individuals living in the commune Phu Cat, an area of high incidence of ALA within the City of Hué, Central Vietnam (Pham Van et al. 1996; Blessmann et al. 2002). The total yearly incidence of ALA in this area is about 0.14%, but exceedingly higher in adult males. As much as 8% of all adult males in the studied population have been treated for ALA within the last 10 years. We examined 450 males and 586 females from randomly selected households. Forty-two of the subjects (41 males, one female) had a history of ALA whereas the remaining 994 had never developed symptoms suspected for ALA and never received ALA-specific treatment.

Ultrasound examinations revealed various liver abnormalities within the studied population of which haemangiomas, congenital cysts and small focal parenchyma calcifications were most common. The frequency of these alterations was evenly distributed between males and females and there was no association of any of these alterations with known ALA history (Table 1). We also detected 12 subjects who had liver lesions consistent with those reported as ALA residues (Sheen et al. 1989). Sonographically, these lesions were round with a hyperechoic wall and a hypoechoic or mixed hyper- and hypoechoic interior, some with wall calcifications. The diameter ranged from 1 to 4 cm and all were located in the right liver lobe (segments 6–8). They were exclusively found in males between 31 and 58 years (mean 43 years), and were significantly more prevalent in subjects with a history of ALA (P < 0.001) (Table 1). As far as information was available, the number and location of these lesions in subjects with an ALA history matched those previously seen by physicians during the acute phase of the disease. In addition, seven of 12 individuals with suspected ALA residues had serum antibodies against Entamoeba histolytica antigen (58%) compared with only 20% in the studied population. Interestingly, seven cases of suspected ALA residual lesions were identified in subjects with no previous ALA history. Their sonographic appearances were virtually indistinguishable from those found in individuals with a known ALA history (Figure 1). All subjects were free of clinical symptoms and negative for C-reactive protein, α-fetoprotein or echinococcus serology, which excludes other possible causes that might have been responsible for the sonographic findings. Regular follow-up examinations over a period of 15 months indicated no changes in the sonographic appearance, which is consistent with inactive residual liver lesions.

Table 1.  Frequency of sonographically detected liver lesions in 1036 adult residents of Phu Cat, Hué City with regard to gender or previous history of ALA
Type of lesionFemale
(n = 586) (%)
(n = 450) (%)
History of
ALA (n = 42) (%)
No history of
ALA (n = 994) (%)
  • *

    No significant difference between males and females or between history of ALA and no history of ALA.

  • † 

    Significant difference between males and females or between history of ALA and no history of ALA; for both cases P < 0.001. Differences were tested by means of Fischer's exact test.

Haemangioma7 (1.2)5 (1.1)*012 (1.2)*
Congenital cyst4 (0.7)2 (0.4)*06 (0.6)*
Focal calcification23 (3.9)18 (4.0)*2 (4.8)39 (3.9)*
ALA suspected residue012 (2.7)5 (11.9)7 (0.7)†
Figure 1.

Sonographic images of the liver with suspected ALA residues from individuals with or without ALA history. Shown are three representative cases of the two groups. Panel 1 (1a–c), individuals with previous ALA history; panel 2 (2a–c), individuals never suspected or treated for ALA. Note: sonographic appearances of lesions between the two groups are virtually indistinguishable. The examples comprise lesions with a mixed hyper- and hypoechoic interior with (1a, 2a) or with little (1b, 2b) wall calcification or lesions with a hypoechoic interior with no wall calcification (1c, 2c).

To our knowledge, this study represents the first population-based survey on sonographic liver findings in residents living in a high-risk area for amoebiasis. The results confirm previous observations of typical residues in about 10% of successfully treated ALAs. In addition, the same type of lesions was found in a number of individuals with no previous ALA history. These lesions were most likely ALA residues as they were identified in individuals of highest risk for ALA (males around 40 years), located at the preferred sites for ALA within the liver (right lobe, segments 6–8), associated with positive amoeba serology (58%vs. 20%) and in addition, other possible causes for their sonographic appearance were unlikely. Thus it seems reasonable to assume the existence of asymptomatic or at least subclinical ALAs, which are self-limited as they obviously healed or at least became inactive without specific treatment. Although spontaneous healing of ALAs has been suggested previously (Rogers 1902), there has been no documented case reported so far. Presently there is no information about the frequency by which this kind of abscesses may occur. However, our results indicate that post-ALA residues should be considered in the differential diagnosis of space-occupying liver lesions, especially in males from amoebiasis endemic areas even without ALA history.


The study was approved by the Scientific Review Board of the Medical College, University of Hué, Vietnam. Assistance by Dr Ton Nu, Dr Tran Dinh, Dr Viet Quynh, Mrs Buss and Mrs Ha is gratefully acknowledged. The study was supported by the Volkswagen Foundation.

Dr Jörg Blessmann, Department of Molecular Parasitology, Bernhard Nocht Institute for Tropical Medicine, Bernhard Nocht Str. 74, 20359 Hamburg, Germany. E-mail: blessmann@bni.uni-hamburg.de
Prof. Egbert Tannich, Department of Molecular Parasitology, Bernhard Nocht Institute for Tropical Medicine, Bernhard Nocht Str. 74, 20359 Hamburg, Germany. Tel.: +49 40 42 818 477; Fax: +49 40 42 818 512; E-mail: tannich@bni.uni-hamburg.de (corresponding author).
Dr An Le Van, University of Hué, 3 Le Loi st, Hué City, Vietnam.