False-Negative Serologies in Amebic Liver Abscess: Report of Two Cases


Corresponding Author: Marie-Pierre Otto, MD, Department of Biology, HIA Desgenettes, 108 Boulevard Pinel, 69003 Lyon, France. E-mail: mp0207@hotmail.fr


Amebiasis, the parasitic disease caused by Entamoeba histolytica, may result in extra-intestinal diseases among which liver abscess is the most common manifestation. We report two cases of amebic liver abscess illustrating the inequal sensitivity of serologic tests detecting anti-amebic antibodies.

Entamoeba histolytica is the protozoan responsible for amebiasis in humans, causing colitis and dysentery or amebic liver abscess (ALA), which represents the most frequent extra-intestinal amebiasis manifestation. It must be distinguished from Entamoeba dispar, more frequently found in stools, which is not pathogenic and does not induce serum anti-amebic antibodies. Entamoeba histolytica could be responsible for 40,000–100,000 deaths yearly in countries with poor sanitary conditions where seroprevalence can reach 50%. In developed countries, groups at high risk for amebiasis are travelers, recent immigrants from endemic areas, men who have sex with men, institutionalized patients, and those in contact with amebiasis patients.[1, 2] For adequate management, it is essential to rapidly diagnose ALA and to distinguish it from other causes of liver damage, particularly pyogenic liver abscesses (PLA). The clinical[3] and imaging data have poor sensitivity and specificity. Therefore, the confirmation of ALA is based on laboratory diagnostic methods: serological tests are the most helpful especially in an emergency context, thanks to rapid and specific E histolytica antibody tests.[1, 4]

Case Reports

Case Report 1

A 27-year-old French male had returned 6 months earlier from a 6-month journey through Nepal and had spent 6 months in Senegal 2 years previously. He was complaining of night and day sweats and lower-thoracic pain for the previous 7 days.

His physical examination only revealed a body temperature of 37.5°C. Laboratory studies of blood showed elevated white blood cell (WBC) count, 35,000/μL (85% neutrophils), an inflammatory syndrome, and alkaline phosphatase level at 1.5 times the normal value. Blood culture remained sterile. An abdominal computerized tomography (CT) scan revealed a single hypodense lesion in the right lobe of the liver (diameter 9.2 cm) consistent with a hepatic abscess. An amebic etiology was suspected, but latex agglutination test (LAT) (Bichro-Latex Amibe, Fumouze, Levallois-Perret, France) on serum was negative on day 1 (threshold at 1 : 5).

The patient was given a first standard course of empiric intravenous antibiotherapy against pyogenic organisms and ameba: co-amoxiclav (3 g/day) and metronidazole (1.5 g/day). Because of risk of spontaneous rupture, drainage of the liver abscess was performed as an emergency (Figure 1). Microscopic examination of the chocolate brown aspiration fluid revealed neither cysts and trophozoites of Entamoeba sp. nor bacteria after Gram coloration.

Figure 1.

Amebic liver abscess during drainage in sagittal section (computerized tomography scan).

Quantitative indirect hemagglutination assay test (IHAT) (Amibiase HAI, Fumouze) and immunofluorescence assay test (IFAT) (Amoeba-Spot IF, bioMérieux) for the detection of antibodies to E histolytica were both positive: IHAT 1 : 640 (threshold at 1 : 320) and IFAT 1 : 640 (threshold at 1 : 160). The negative result with LAT was confirmed by a new analysis done with a new lot of the same kit and a prozone phenomenon was excluded. Serology was controlled on day 6. The results of serological tests on day 6 compared with day 1 in the same run were respectively 0 (day 1) and 1 : 20 (day 6) for LAT, 1 : 640 and >1 : 2560 for IHAT, and 1 : 320 and 1 : 640 for IFAT.

The result of real-time polymerase chain reaction (PCR) to detect E histolytica DNA directly in pus was positive.

Co-amoxiclav was stopped, metronidazole was maintained for 10 days and tiliquinol was added for 10 days. The patient left the hospital on day 7.

Case Report 2

Three weeks after his arrival in Tchad, a 45-year-old French male suffered from a sudden pain in the right hypochondrium, hyperthermia (40°C), and cholestatic jaundice. Abdominal ultrasound revealed a liver abscess compressing bile ducts. Empiric parenteral antibiotherapy was started (day 1): cefotaxim (3 g/day), gentamicin (200 mg/day), and metronidazole (1.5 g/day). On day 10, the patient was repatriated back to France. Laboratory studies of blood showed WBC count of 16,500/μL (71% neutrophils), inflammatory syndrome, and elevated gamma glutamyltransferase. Blood cultures remained negative. Abdominal CT scan revealed a focal bilobate lesion measuring 6 cm.

Serologic tests for detection of anti-amebic antibodies with LAT, IHAT, and electrosyneresis (ES) were negative. IFAT was found slightly positive: 1 : 80 (threshold at 1 : 80) in a laboratory and 1 : 300 (threshold at 1 : 150) in a reference center.

Because of undetermined etiology, drainage of the liver abscess was required. Microscopic examination of the hematic aspiration fluid remained negative whereas histological examination of the liver fragment taken during the aspiration revealed amebae in the abscess capsule leading to the diagnosis of ALA.

Clinical and biologic outcomes were good after 20 days of treatment with metronidazole.

The serology was followed up for 4 months: LAT, IHA, and ES remained negative, IFAT result was twice the threshold on day 7, 14, and 37 and negative on the fourth month.


The two cases reported here emphasize the difficulties in diagnosing the etiology of a liver abscess. Therefore, at the presentation of a patient with liver abcess, both PLA and ALA should be considered according to the patient's clinical parameters. Furthermore, when no liver abscess puncture is performed the treatment must also cover anaerobes, and therefore metronidazole was used which is efficient against E histolytica.

In industrialized nations where amebiasis is not endemic, serologic tests are essential for the diagnosis of ALA.

Current methods include IFAT, IHAT, enzyme-linked immunosorbent assay (ELISA), counterimmunoelectrophoresis (CIE), ES, and LAT. If IFAT, IHAT, ELISA, CIE, and ES are time-consuming methods requiring trained personnel and specialized equipment, LAT is a bedside test, easy to perform and gives rapid results (5 minutes). Therefore, LAT is often used in first-line in an emergency context. IFAT uses whole antigen whereas other tests use soluble antigens. Most of the tests are marketed, others are home made. Usually, the antigens come from cultivated axenic strains whereas recombinant proteins are exceptionally used.

Altogether, for the diagnosis of liver abscesses, amebic serology is considered as highly sensitive (>94%) and highly specific (>95%).[1] In the literature, sensitivity, specificity, and positive and negative predictive values of serologic tests used to diagnose amebiasis are similar whatever the method used. However, the following study limitations must be emphasized: retrospective studies on serum bank, lack of gold standard, and high pre-test probability. Thus, results of positive and negative predictive values are not very reliable data.

Limits of amebic serology in ALA diagnosis exist. False positives decrease specificity and positive predictive value. It has been frequently observed that current serologic tests measure long-persisting antibodies in amebiasis.[5] Thus, in areas where amebiasis is endemic, serologic testing cannot distinguish past from current infections and confirm an extra-intestinal amebiasis except if positivity breakpoint was higher. However, real false positives in industrialized countries are rare. Among 1000 amebic serologies, Laverdant and colleagues reported only two cases of false positives concerning patients with hepatocarcinoma.[6] Thus, in industrialized countries, amebic serology must be performed only on patients with a hepatic abscess who have stayed in an endemic area. False negatives decrease sensitivity and negative predictive value. They can be due to patients' immune response, the type of serologic test, or the pathogen strain. Sensitivity seems to be less important with serums obtained during acute illness (70%–80%) than those obtained during convalescence (>90%).[4] Indeed, a false-negative result can be obtained with a serologic test done within the first 7 to 10 days of the infection, but when repeated later, the test usually becomes positive: most of the time, seroconversion occurs before the 15th day. Furthermore, discrepant results can be seen between different assays done on the same sample. It has been described between LAT (negative) and IHAT (positive) in several publications (1/15 for Cummins[7] and 4/42 for Kraoul[8]), although these two assays use the same antigen. A similar result has been obtained between LAT and EIA (2/27 for van Doorn[9]). In this last case, initially negative samples in LAT became positive 3 to 6 days later. Furthermore, E histolytica wild-type strains compared to strains developed in cultures used to make serologic tests could present differences in their antigenic profile.

Tanyuksel and colleagues pointed out that the lack of an accurately defined “gold standard” has impeded an objective assessment of the sensitivity of the antibody detection tests currently in use.[10] The accuracy of serology may be overestimated and the use of PCR methods tends to confirm this hypothesis.[11] Furthermore, no studies have been found concerning evaluation of amebic serology performance compared to a “gold standard” with likelihood-ratio test that limit the impact of prevalence.


These observations lead to the conclusion that, in a highly evocative context with negative blood cultures, ALA should be considered despite a first negative amebic serology.

Several propositions can be made to confirm the hypothesis of ALA. First, two or more screening serologic tests must be made. Second, if the initial serologic tests are negative, it is necessary to repeat the assay 7 to 10 days later. Third, the direct detection methods based on PCR gene amplification techniques realized in the abscess liver pus (when the aspiration is required) and also in blood, saliva, and urine samples appear to be very helpful and should be more systematically performed.[12]

Declaration of Interests

The authors state that they have no conflicts of interest to declare.