Hepatic parasitic abscess resistant to traditional anthelmintics: difficulty of diagnosis and treatment of fascioliasis in the USA

Abstract

We report a case of human fascioliasis in the USA that encountered many diagnostic uncertainties. Numerous tests available for detection of fascioliasis were utilised but the diagnosis remained elusive. Confounders included three negative stool ova and parasite examinations, positive hepatitis A virus IgM antibody, cross-reactive false-positive Echinococcus IgG antibody, absence of characteristic image findings and unrevealing liver biopsy. Praziquantel was started as empiric treatment for helminth infections, but was ineffective. Due to the rarity of the disease in the USA, serologic testing and triclabendazole were only available from the Centers for Disease Control and Prevention, which led to a delay in diagnosis and treatment.

Background

Fascioliasis is a parasitic infection primarily of the hepatobiliary system caused by one of two liver flukes, Fasciola hepatica or Fasciola gigantica. This zoonotic disease is usually acquired through ingestion of raw aquatic vegetation contaminated with immature parasite larvae. A review of literature showed numerous tests are available for detection of fascioliasis: stool ova and parasite examination, serologic detection of antibodies against Fasciola antigen, Fasciola antigen detection assay, CT, ultrasound (US), MRI, microscopic examination of duodenal or biliary aspirates, and liver biopsy.^1–3 Despite the availability of many diagnostic tests, there is still considerable difficulty in diagnosing and treating fascioliasis in the USA.

Case presentation

We present a case report of a 29-year-old woman of Vietnamese descent with no significant medical history who was diagnosed with fascioliasis. She had travelled to Vietnam and Cambodia for approximately 1 month between December and January. During her travel, patient had consumed watercress as part of salads, but denied eating raw meats and seafood. She presented to the emergency department 5–6 months later with complaints of fever, chills, fatigue, bloody stool, loss of appetite, weight loss of 3–4 pounds, and right-sided abdominal pain. She denied nausea, vomiting, diarrhoea, rash, pruritus, headaches, chest pain, and joint pain.

Investigations

Initial vitals included temperature 37.2°C, heart rate 83 beats per minute, respiratory rate 18 breaths per minute, blood pressure 111/72 mmHg, and oxygen saturation 100% on room air. Physical examination demonstrated tenderness to palpation in the right upper quadrant of abdomen. Laboratory results showed hepatitis A IgM antibody reactive, hepatitis B surface antigen non-reactive, hepatitis B core antibody reactive, hepatitis B virus DNA undetected, hepatitis C antibody non-reactive, HIV antigen and antibody screen negative, Echinococcus IgG antibody positive, haemoglobin 9.6, platelet 244, white blood cell count (WBC) 12.64 with differential (neutrophil 42.2%, lymphocyte 7.3%, monocyte 3.7%, basophil 0%, eosinophil 46.8%), alanine aminotransferase 42, aspartate transaminase 27, total bilirubin 0.4, alkaline phosphatase 109, partial thromboplastin time 32, prothrombin time 14.4, international normalised ratio 1.1, erythrocyte sedimentation rate 65, ferritin 85, peripheral smear with few schistocytes and few ovalocytes, blood cultures were negative, and three separate stool specimens collected over a 10-day period for ova and parasite examination were negative. Unenhanced CT imaging exhibited an ill-defined collection in the right liver lobe. CT-guided needle biopsy of the right liver lobe with aspiration of lesion was negative for acid-fast bacilli, anaerobe and fungi. Pathology of liver biopsy revealed necroinflammatory infiltrate associated with granulation-type tissue and inflammatory infiltrate rich in eosinophils. MRI of liver uncovered a 4.2×7.2 cm ill-defined area in the right hepatic lobe with T1 hyperintense haemorrhagic fluid likely from haemobilia that contained numerous confluent cysts, which was concerning for microabscesses (figure 1A–C).

Figure 1.

MRI (A–C) with contrast demonstrates an ill-defined 7 cm right hepatic lobe area containing haemorrhagic debris and fluid (white arrows). This collection is comprised of numerous small confluent cystic structures concerning for microabscesses (circle).

Differential diagnosis

Infectious disease team deduced that the Echinococcus IgG antibody was likely cross-reactive false positive because the liver lesion on imaging did not have the characteristic appearance of Echinococcus. A previous study demonstrated that false-positive detection of antibodies to Echinococcus can occur in the setting of previous Echinococcus infection, other helminth infections, cancer, tuberculosis and liver cirrhosis.⁴

Treatment

Praziquantel is no longer recommended for treatment of human fascioliasis. Our infectious disease team had recommended it as empiric treatment for helminth infections because triclabendazole, the drug of choice for treatment of fascioliasis, could only be acquired through the Centers for Disease Control and Prevention (CDC) and therefore not readily available. The eosinophilia persisted despite treatment with praziquantel. The CDC was then contacted for serologic testing, which confirmed antibodies against Fasciola antigen. After the diagnosis of fascioliasis was established, the CDC provided two doses of triclabendazole, which led to rapid resolution of eosinophilia.

Outcome and follow-up

After receiving two doses of triclabendazole, patient was followed up clinically and radiographically at 1, 3, 6 and 12 months. At 1 month, the hepatic lobe collection measured 8.5×6.1×6.6 cm (figure 2A, B) and the patient reported no fever, chills, night sweat, fatigue and loss of appetite. However, she felt occasional right upper quadrant abdominal pain that resolved with tramadol. Her WBC was 6.48 with absolute eosinophil 0.71. At 3 months, the hepatic lobe collection measured 6.8×5.5×5.7 cm and the patient continued to experience occasional mild abdominal pain in the right upper quadrant once every 1–2 weeks with each episode lasting 1–2 hours. Overall, the right upper quadrant lessened in intensity and the patient did not require any pain medications. Her WBC was 4.77 with absolute eosinophil 0.19. At 6 months, the hepatic lobe collection measured 7.1×5.1×7.5 cm but was less visible and therefore more difficult to measure on imaging. The patient reported occasional right upper quadrant abdominal discomfort instead of pain. The infectious disease specialist noted that although the lesion was persistent and still large, the fact that it is less distinct and harder to measure is a sign of it resolving. Her WBC was 5.74 with absolute eosinophil 0.17. At 12 months, the liver was normal in size, shape and echotexture. In the right hepatic dome in region of prior abnormality, there was no discrete residual with only minimal heterogeneity (figure 2C).

Figure 2.

Ultrasound imaging (A, B) after 1 month of treatment demonstrates persistent ill-defined hyperechoic mass with confluent cystic spaces. No blood flow within the cystic spaces on colour Doppler imaging. Follow-up ultrasound after treatment demonstrates involution of the right hepatic lobe irregular collection (C).

Discussion

Parasitic infection should be on the differential diagnoses for any patient presenting with the combination of right upper quadrant abdominal pain, transaminitis and peripheral eosinophilia. A history of travel to endemic areas of Europe, Asia or the Pacific raises even more suspicion.⁵ Our case report illustrates the difficulty of diagnosis and treatment of fascioliasis in the USA despite having heightened suspicion for parasitic infection. Many tests available for detection of fascioliasis were utilised but the diagnosis remained elusive. Confounders included three negative stool O&P examinations, positive hepatitis A virus IgM antibody, cross-reactive false-positive Echinococcus IgG antibody, absence of characteristic image findings and negative liver biopsy.

The clinical course of fascioliasis entails two phases: a hepatic parenchymal phase followed by a hepatic ductal phase.² In the parenchymal phase, common image findings consist of multiple small nodular and branching linear lesions in the subcapsular areas of the liver parenchyma,⁶ which are tortuous tracks left behind by the migrating parasites.⁷ In the ductal phase phase, CT image demonstrates dilated biliary ducts, and MRI shows irregular heterogeneity as a result of capsular and subcapsular fibrotic scars.² A unique imaging feature in our case is the haemobilia seen on the MRI. The presence of a subcapsular haematoma of the liver is suggestive of invasive hepatic fascioliasis.² However, these characteristic imaging findings of parenchymal and ductal phases were only appreciated in retrospect in our case.

The absence of eggs in the stool samples may be attributed due to several factors including incapability of Fasciola to produce eggs, lack of adaptation of the fluke to the host, encapsulation of eggs in the liver granuloma, low egg releasing because of low infection burden, intermittent egg production of the parasite and cessation of egg shedding in the advanced chronic phase of the disease.⁸ Therefore, multiple negative stool samples do not necessarily exclude the diagnosis. Additionally, histopathology of liver biopsy specimen did not reveal any microorganisms, eggs or granulomas with eosinophilic walls and multiple cystic cavitating structures that would be suggestive of Fasciola.^{7 9}

A case of domestically acquired fascioliasis in California further highlights the difficulty of diagnosis and treatment of fascioliasis in the USA.¹⁰ In the referenced case, the patient had eaten local watercress in rural Northern California and then developed abdominal pain 2 months later. Similar to our case, their investigation revealed marked eosinophilia, an ill-defined area of heterogeneous attenuation and enhancement of the liver on CT imaging, and three stools that were negative for ova and parasites. The patient was treated with a 14-day course of nitazoxanide, which was considered but not given in our case. A repeat CT of the liver was unchanged a week later after initiation of nitazoxanide. Serial Fasciola serologies at a commercial laboratory by ELISA to crude F. hepatica extract were all negative. Finally, F. hepatica serology was sent to the CDC’s reference laboratory for Fasciola testing, which confirmed the diagnosis.

In summary, triclabendazole is the only treatment recommended by the World Health Organization for Fasciola species. Triclabendazole and serologic testing are not readily available in the USA since they must be obtained from the CDC. In both our case and the above-referenced case of domestically acquired fascioliasis, there was high suspicion of parasitic disease and fascioliasis was considered a possibility. However, initial treatments and testing were ordered based on the immediate availability of the treatments and testing, which led to a delay in diagnosis and treatment of fascioliasis.

Learning points.

• Triclabendazole is the only treatment recommended by the World Health Organization for fascioliasis. But this medication is not approved for use in the USA.

• Due to the rarity of fascioliasis in the USA, serologic testing and triclabendazole are only available from the CDC, which can lead to delay in diagnosis and treatment.

• The absence of eggs in the stool samples may be due to attributed to several factors including incapability of Fasciola to produce eggs, lack of adaptation of the fluke to the host, encapsulation of eggs in the liver granuloma, low egg releasing because of low infection burden, intermittent egg production of the parasite and cessation of egg shedding in the advanced chronic phase of the disease.

Ethics statements

Patient consent for publication

Obtained.