Use of a rapid faeces multiplex PCR assay for diagnosis of amoebic liver abscess.

Sir,

Entamoeba histolytica is an intestinal, protozoan parasite endemic to non-industrialised parts of Latin America, Africa and the Indian subcontinent. It causes an estimated 40,000–100,000 deaths each year, primarily from amoebic colitis and amoebic liver abscess (ALA). Although morphologically indistinguishable, strains of E. histolytica causing ALA often differ genetically from those causing colitis. , Colonic carriage can also be asymptomatic, as is always the case for other apathogenic species such as E. dispar and E. moshkovskii. Therefore, clinical history of colitis or detection of Entamoeba in stool do not provide reliable clues for the diagnosis of ALA.

To date, microscopic examination of abscess fluid and serological testing have been the mainstays of diagnosis of ALA. Both techniques have important deficiencies. Examination of liver abscess pus only gives a microscopic diagnosis in ≤20% of cases as amoebae are typically found in the periphery of abscesses rather than in the aspirated pus more centrally. Serology may be falsely negative early in the course of ALA, has comparatively poor specificity in areas of high seroprevalence (due to persistence of antibodies following recovery from invasive amoebiasis) and, because of centralised testing in Australia, often yields an unacceptably delayed result. ,

Targeted polymerase chain reaction (PCR)-based assays on abscess pus have been found to be both sensitive and specific for the diagnosis of E. histolytica-associated ALA but are only available in research settings and are relatively laborious. , Antigen detection methods have been used both in stool and liver abscess pus but are not commercially available and are significantly less sensitive and specific than the aforementioned molecular assays. Therefore, there remains an important need for new, rapid and accurate microbiological methods for diagnosing ALA.

In March 2020, a male from India, aged in his 30s, presented to Royal Darwin Hospital with fever, upper abdominal pain and mild diarrhoea. Examination revealed tachycardia, a temperature of 38.1°C and right upper quadrant tenderness without guarding. Initial blood testing showed an elevated C-reactive protein (408.2 mg/L), a raised neutrophil count (19.7×109/L) and a raised alkaline phosphatase (317 U/L). A computed tomography (CT) scan of the abdomen showed two hypodense hepatic lesions in segments 7 and 8, respectively, the largest of which measured 100×58×99 mm, with appearances suggestive of liver abscesses (Fig. 1 ). Urine, faeces and blood culture were negative for bacterial and fungal growth and no ova, cysts or parasites were seen in two stool specimens. Serum was sent to an offsite reference laboratory for E. histolytica serology but results were not available during this patient's admission. A total of approximately 200 mL of non-odorous ‘anchovy sauce’ pus was aspirated from both liver lesions on day one of admission (Fig. 1). This contained amorphous material and many neutrophils, but no bacterial, fungal or parasitic organisms were seen on microscopy and there was no growth after appropriate incubation. Given the strong suspicion of ALA and the expected delay in the results of serology, we performed a Biofire FilmArray Gastrointestinal multiplex PCR panel (BioMerieux, France), designed for detection of a range of different stool pathogens, including E. histolytica, on 200 μL of the liver abscess pus. This gave a positive result for E. histolytica within one hour. Treatment was rationalised to high dose metronidazole (and subsequent paromomycin), as per Australian Therapeutic Guidelines, with a rapid clinical response. The E. histolytica indirect haemagglutination titre was subsequently reported at ≥1:2560, 20 days after initial diagnosis. Prolonged turnaround times for send-away tests from our hospital in the tropical north of Australia are not unusual but, in this case, air courier service disruption during the COVID-19 outbreak probably exacerbated the time delay.

Fig. 1

(A) Post-contrast computed tomography (CT) scan of the upper abdomen showing two large hepatic abscesses in segments 7 and 8, respectively. (B) Macroscopic appearance of the aspirated pus.

In our patient, analysis of liver abscess pus using a commercially available, multiplex, nested PCR assay designed for stool, provided a rapid result and allowed immediate rationalisation of therapy targeting a single pathogen. To our knowledge, this is the third published use of a stool multiplex PCR assay and the second of the Biofire assay specifically for diagnosis of ALA using abscess pus. Weitzel et al. used the Rida Gene Stool Panel (R-Biopharm, Germany) and the Biofire panel on liver abscess pus for rapid diagnosis of ALA in a 34-year-old man from Chile with a 50 mm left lobe liver lesion. Both assays confirmed the presence of E. histolytica and subsequent serology was consistent with this diagnosis. Bernet Sánchez et al. used the Allplex Gastrointestinal Panel 4 Assay (Seegene, South Korea) in the same manner to diagnose a 55-year-old immunosuppressed female with multiple large ALAs, again with consistent serological results. In both of these cases, the PCR panels were performed prior to initiation of metronidazole, which has been shown previously to improve detection rates of targeted PCR panels on liver abscess pus.

There are several commercially available multiplex PCR panels containing primers for E. histolytica and their use for analysis of liver abscess pus has many theoretical advantages over conventional methods for the diagnosis of ALA: simplicity, widespread laboratory availability, rapid turnaround time and the ability to concomitantly exclude several other potential microbiological causes of liver abscess. One would assume that the sensitivity and specificity of nested PCR assays, such as the Biofire panel, would be much greater than standard microscopy and serology. Nevertheless, their use on liver abscess pus remains unvalidated and their false positivity and negativity rates are unknown. We plan to continue using and prospectively validating the Biofire Gastrointestinal multiplex panel on liver abscess pus in patients with suspected amoebic liver abscess and encourage others to consider using one of the commercially available multiplex assays to do the same.

Conflicts of interest and sources of funding

The manufacturer of the Biofire FilmArray Gastrointestinal multiplex panel had no involvement in this study. The authors state that there are no conflicts of interest to disclose.