Background

Amebic pericarditis was first reported by Rouis in 1860 during an autopsy of a patient with an hepatic amebiasis [1]. The three stages of amebic pericarditis are (1) “pre-suppurative pericarditis,” an initial sympathetic or reactive effusion, (2) “suppurative pericarditis,” a rupture of a liver lesion into the pericardial cavity with the characteristic anchovy paste appearance, often resulting in cardiac tamponade, and (3) “constrictive pericarditis,” a late complication of amebic pericarditis which develops over weeks to months and requires pericardiectomy [2, 3]. The pre-suppurative form should be considered a potential precursor of the suppurative form. Amebic pericardial effusion usually results from the extension of a left lobe liver lesion through the diaphragm into the pericardium, leading to a hepatopericardial fistula [4, 5]. Hematogenous dissemination can also occur, although less frequently.

Entamoeba histolytica (E. histolytica) is rarely identified as a cause of pericarditis. We report a case of amebic pericarditis in which the cause of the cardiac tamponade was not identified initially, and the diagnosis was made retrospectively when the patient deteriorated 1 month later, by performing PCR on a stored sample of pericardial fluid obtained during the first admission.

Case presentation

A 71-year-old man without a history of travel to amebiasis endemic areas presented to our emergency department with a 2-week history of left pleuritic chest pain and a non-productive cough. The patient had a recent past medical history of essential thrombocytopenia and myelodysplastic syndrome/myeloproliferative neoplasm was treated with hydroxyurea (500 mg) followed by anagrelide hydrochloride hydrate (3 mg). Chest radiograph revealed cardiomegaly, pulmonary congestion, and left pleural effusion. The transthoracic echocardiogram (TTE) showed normal left ventricular wall motion with a small amount of pericardial effusion near the left ventricular posterior wall. The patient was admitted to the cardiology department on suspicion of anagrelide-induced acute heart failure; thus, treatment with anagrelide was stopped. The hospital course was favorable until hospital day 9 when the patient developed chest pain and dyspnea. The patient was in severe distress and had a blood pressure of 80/66 mmHg, heart rate of 104 beats per minute, respiratory rate of 24 breaths per minute, and an oxygen saturation of 77% on room air. Electrocardiogram showed ST-segment elevation in V2–6 and PR-segment depression in V3–6. TTE and computed tomography (CT) without contrast showed the increased pericardial fluid effusion. The patient was diagnosed with pericardial tamponade and required an urgent pericardiocentesis. The pericardial fluid was blood-colored with cell count of 4250 cells/µL, predominantly neutrophils (88%) with lymphocytes (6.5%), and eosinophils (0.5%). Many clusters of dead or degenerated cells were observed in cytology. We did not identify amebic trophozoites. The bacterial and mycobacterial cultures of the pericardial fluid were negative, as were serum IgM antibodies for viruses including Echovirus, Coxsackievirus, and Parvovirus B19. Rheumatoid and antinuclear antibodies were also negative. Although the cause could not be identified, chest pain and cough gradually improved with ibuprofen and colchicine. The patient was discharged on hospital day 13 after the pericardial effusion had resolved. However, 1 month after discharge and 5 days before the second admission, the patient presented with acute onset of fever, right upper abdominal pain, nausea, and loss of appetite. TTE revealed small amount of pericardial effusion, and contrast-enhanced CT scan of the abdomen revealed multiple liver lesions.

We empirically started with intravenous (IV) ceftriaxone (2 g every 24 h) and metronidazole (1 g every 8 h). Percutaneous drainage of the liver lesion was performed on hospital day 2. The liver lesion was neutrophil-dominant, with erythrocytes, lymphocytes, and histiocytes visible. Many clusters of dead or degenerated cells were observed on cytology as well as in the pericardial fluid. After drainage, the patient’s signs and symptoms gradually improved. Notably, on hospital day 12, serum enzyme immunoassay (EIA) was positive for E. histolytica IgM and IgG antibodies. Moreover, we performed conventional qualitative polymerase chain reaction (PCR) testing of DNA samples extracted from tissue samples of the liver lesion, and pericardial and pleural effusion fluid collected and stored during the first admission, using a dysentery amoeba-specific primer that partially amplified the 18S ribosomal DNA region. The PCR test result was positive. We diagnosed an amebic liver lesion associated with amebic pericarditis. We reported the positive PCR result to the pathologists and asked them to check again, but unfortunately, they could not confirm free living ameba. Of note, the patient initially denied any recent sexual history, but when we interviewed him again after he was diagnosed with amebiasis, he admitted that he had had sexual intercourse with several commercial sex workers 4 months previously. He denied any sexual contact with men, but he stated he had oral-anal sex. The dose of metronidazole was increased (to 750 mg IV every 8 h) and continued for 14 days, followed by oral paromomycin (500 mg every 8 h) for 10 days. The patient was clinically stable and was discharged on day 18 of hospitalization. Follow-up TTE and CT abdomen revealed no evidence of recurrence at the 1-year follow-up.

Systematic review

A literature search of the PubMed database (up to November 2019, updated in April 2021) was conducted using the keywords (((“Amoeba“[Mesh] OR amoeba[TIAB] OR ameba[TIAB] OR amebic[TIAB]) OR (“Amebiasis“[Mesh] OR amebias*[TIAB] OR amoebias*[TIAB])) AND (“Pericarditis“[Mesh] OR pericarditis[TIAB])) OR ((“Heart Diseases/parasitology“[Mesh]) AND ((“Amoeba“[Mesh] OR amoeba[TIAB] OR ameba[TIAB] OR amebic[TIAB]) OR (“Amebiasis“[Mesh] OR amebias*[TIAB] OR amoebias*[TIAB]))). The Embase database was also searched using the keywords (((‘amoeba’/exp OR amoeba) OR ameba OR amebic OR ‘amebiasis’ OR amebias* OR amoebias*) AND ‘pericarditis’) NOT (((‘amoeba’/exp OR amoeba) OR ameba OR amebic OR ‘amebiasis’ OR amebias* OR amoebias*) AND ‘pericarditis’ AND ([medline]/lim OR [pubmed-not-medline]/lim)) (Fig. 1). The database including titles, abstracts, and languages was firstly created by FK using the abovementioned formula. For the literature review, the database records were split in half (A and B). Database A was independently reviewed by KI and RH and database B was independently reviewed by TS and TF. Full-text text was used for eligibility assessment, and detailed information on each case was extracted. Finally, each database was checked by KK and TM to create the combined list of cases. Case reports that were not regarding amoebic pericarditis, and reviews and clinical studies that did not present cases of amoebic pericarditis were excluded. Additionally, reports for which abstracts or full texts were in languages other than English were excluded.

Fig. 1
figure 1

 A flow chart of the systematic review data collected for this study

Full size image

We found 53 papers describing 121 cases of pericarditis caused by E. histolytica [1, 3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55]. The detailed information on the clinical characteristics of these 122 cases, which includes our case, are shown in Table 1. Of those cases, 90.4% were male, the median age was 46 years, and only six cases, including this case, were reported after 2000. Moreover, 45.1% (n = 55) of the cases were reported from Asia, followed by Africa (n = 34, 27.9%), Central and South America (n = 11, 9.0%), North America (n = 6, 4.9%), and Europe (n = 3, 2.5%). Liver lesions were noted in 75% (n = 60) of cases, mostly in the left lobe (n = 32, 53.3%), followed by location data not available (n = 20, 33.3%), the right lobe (n = 7, 11.7%), and both lobes (n = 1, 1.7%). Six cases (2.4%) did not have a liver lesion. The median duration of antimicrobial therapy was 14 days, and 31.1% (n = 38) were fatal.

Table 1 Characteristics of patients with amebic pericarditis
Full size table

Discussion and conclusions

To the best of our knowledge, this is the first systematic review of amebic pericarditis as a rare complication of amebiasis. A strength of this review is that, in addition to providing patient demographics for 122 cases over the past 80 years, it also focuses on the use of PCR, which has recently been shown to be useful for diagnosis for amebiasis. The learning point from this case is that had the diagnosis had been made at the time of the first cardiac tamponade, the second hospitalization might have been avoided. The diagnosis was missed initially and confirmed retrospectively when the patient deteriorated 1 month later, by performing PCR on a stored sample of pericardial fluid obtained during the first admission. It is worth noting that although we did not suspect amebiasis because the patient initially denied a history of sexual intercourse and did not have a history of travel to an endemic area. We learned that if the cause of pericarditis is unclear, a detailed interview should be conducted and amebiasis should be suspected.

Most cases of amebic pericarditis were reported from developing countries, especially the tropics and subtropics in Asia and Africa, where there is inadequate hygiene and access to sanitation [58]. This trend is similar to that of all amebic dysentery and amebic liver lesion [58] The previous study in Egypt showed that approximately 40% of patients with acute diarrhea had amebic colitis [59]. The reason why the number of case reports peaked in the 1970s and has been gradually decreasing subsequently, may be because treatment has been standardized and is generally successful, so clinicians are less likely to report cases. In developing countries where amebiasis are endemic, it is still possible that this has not yet been reported due to underdevelopment of surveillance and diagnostic techniques [58]. Even not in endemic areas, as our case, sexual intercourse could be the risk factors of amebiasis.

Regarding the location of the liver lesion, in our literature review, we found that 32 cases (53.3%) had left lobe liver lesion, seven (11.7%) with only right lobe liver lesion, and six case (2.4%) without a liver lesion was observed. Although it has been known that amebic pericarditis is often associated with left-sided liver lesions, this result suggested that amebic pericarditis is not necessarily only secondary to the left lobe of the liver. The absence of an initial liver lesion in our case made the diagnosis of amebic pericarditis difficult, but this case taught us that PCR testing for amebiasis should be performed promptly when investigating cases of pericarditis of unknown origin. The fact that the pericarditis preceded the liver lesion in our case, and that the pericardial fluid was bloody, not like anchovy paste, suggests that the mechanism of spread to the pericardium was by hematogenous dissemination. Anchovy paste-like liquid is seen in cases of liver lesion, and consists of ameba bound to necrotic liver cells. In fact, considering that other extraintestinal amebiasis, such as central nervous system lesions and pulmonary amebiasis without liver lesions are assumed to be disseminated hematogenously [58, 60], trophozoites could have disseminated into the pericardial sac via the bloodstream.

In terms of diagnosis, anchovy paste-like fluid by pericardiocentesis was reported to be the main characteristic and was found in about half of the cases evaluated in this review. Since the 1970 s, there has been an increase in the number of cases in which a positive antibody test was combined with pericardiocentesis [3, 18]. Until then, most cases had been clinically diagnosed only by the characteristic pericardial fluid findings of “anchovy paste” and response to treatment. Considering that trophozoites could be seen microscopically in only two cases (7.1%), the absence of trophozoites could not rule out amebic pericarditis. Notably, the most recent reports revealed the usefulness of PCR detection of E. histolytica in pericardial fluid [49, 51]. Since PCR for E. histolytica has been reported to be useful in intestinal and extraintestinal amebiasis [61], clinicians should consider using PCR methods to confirm E. histolytica in patients with pericarditis of unknown origin.

The conventional qualitative PCR method used in this study has been shown to be highly specific and sensitive by testing with DNA from a variety of pathogens, including bacteria and other protozoa [62]. Unfortunately, we did not perform serum quantitative PCR in this case. It would have been informative to compare the results of serum quantitative PCR between first and second admissions. Compared with conventional PCR, real-time PCR has the following advantages: (1) the results are easier to interpret numerically than by visually checking the stained gels as in conventional PCR; (2) the sensitivity is higher; (3) no post-amplification analysis is required, minimizing the risk of contamination of the laboratory; (4) the duplex profile can distinguish between E. histolytica and E. dispar infections [63]. Nevertheless, real-time PCR is more costly than morphological stool examination and antigen-based detection tests. Therefore, real-time PCR is unaffordable in many of the countries in which E. histolytica is endemic. Instead, this technique may be useful in developed countries for diagnosing amebiasis in travelers from endemic areas.

Our literature review showed that, in most cases, pericardiocentesis was performed to investigate the cause of pericarditis. However, as in our case, it is essential to note that cardiac tamponade may occur, requiring urgent puncture or pericardiotomy. In this review, approximately 30% of patients died. In patients with tamponade early surgical intervention is essential, regardless of the cause. Regarding antimicrobial options, metronidazole was most commonly used, followed by emetine. The dosage and optimal duration of metronidazole varied widely. Still, as recommended for other amebic infections [58], most of the patients received 500–750 mg three times a day for 10–14 days (median = 14 days).

An important limitation of this study is that many of the cases were old and could not be directly compared due to differences in diagnostic tools. Due to the unavailability of accurate diagnostic tests, there are likely to have been many additional cases that were not diagnosed or reported. Furthermore, it should be noted as a limitation that manuscripts published in languages other than English are excluded in this review, as shown in Fig. 1. There may be differences in the characteristics of amoebiasis in countries where English is not the native language. Moreover, the recurrence rate or long-term prognosis could not be determined because of a lack of information. Further cases should be accumulated, and research on the appropriate diagnosis and management of amebic pericarditis is warranted.

In summary, although amebic pericarditis is rare, it should be promptly diagnosed and treated because of its high mortality rate. In addition to serum antibodies, PCR obtained by pericardiocentesis should be considered for patients with pericarditis of unknown origin and high pretest probability. It is also important for clinicians to be aware that pericarditis can be associated not only with the typical left lobe, but also with the right lobe, in the absence of liver lesion, or with delayed appearance, as in our case. This study highlights amebic pericarditis and contributes to the knowledge of its diagnosis.