Percutaneous catheter drainage in drug-resistant amoebic liver abscess


correspondence R. M. Hanna, P. O. Box 25415, Safat 13115, Kuwait


Summary This communication records our experience with the percutaneous catheter drainage (PCD) of 22 amoebic liver abscesses in 19 patients who had failed to respond to amoebicidal therapy. In one patient with a left lobe abscess, imminent rupture was an additional indication for drainage. PCD combined with amoebicidal therapy not only expedited recovery, but was curative in all 19 patients. There were no complications. We conclude that PCD is a most useful adjunct to drug therapy and recommend its routine use in the management of drug-resistant amoebic liver abscesses.


Liver abscess is the most common extraintestinal complication of amoebiasis and occurs in approximately 3% to 9% of patients with amoebic infection ( Debakey & Ochsner 1959; Landay et al. 1980 ; Peters et al. 1981 ). Due to the effectiveness of amoebicidal therapy in the management of these abscesses, aspiration or catheter drainage is rarely necessary ( Ralls 1998). However, resistance to drug therapy occurs in some patients ( Singh & Kashyap 1989), who may consequently need alternative or additional methods of treatment. Although needle aspiration has been used as an adjunct, percutaneous catheter drainage (PCD) has been found to be the most effective form of drainage in such circumstances ( Merten & Kirks 1984; Singh & Kashyap 1989; Moazam & Nasir 1998; Rajak et al. 1998 ). We have retrospectively reviewed our cases of amoebic liver abscess who underwent PCD due to lack of response to drug therapy and present our results here.

Patients and methods

We reviewed the cases of 19 patients, 16 males and three females, with a diagnosis of amoebic liver abscess made on the basis of the clinical history and physical findings, and confirmed by serological, ultrasonographic and/or computed tomographic examinations. Their ages ranged from 9 to 56 years (mean 36 years). In all patients, pain, fever and abdominal tenderness had persisted after two weeks of treatment with oral metronidazole (Rhone Poulenc Rorer, Doma, France) 800 mg three times daily in adults and 400 mg three times daily in the single child in the series. Drainage was carried out under ultrasonographic (US) guidance with 8–12 French catheters introduced into the abscess cavity by Seldinger technique and under basal sedation (Sosegun 60 mg; Pentazocine hydrochloride 30 mg/ml, Sanofi Winthrop, Raspail, France) and local anaesthesia (xylocaine 2%–20 mg/ml, Austra, Sodertalje, Sweden). The patients with two abscesses had them drained by two separate catheters. Aspirates were sent to the laboratory for gram stain testing, culture and sensitivity and for ova and parasites. The catheter was left in situ until there was no more fluid draining from it. This period varied from 24 h to 48 h in all patients. No amoebicidal drug was injected into the abscess cavities. Ultrasonography (US) was done just before removal of the catheter to confirm that the abscess cavity was empty. Patients were discharged 3 to 5 days later. The volume of fluid drained was measured in each case. During hospitalization, all patients underwent US daily. Amoebicidal therapy was continued for 7–22 days.


There were a total of 22 abscesses. Fifteen patients had a single abscess in the right lobe of the liver ( Figure 1), three patients a single abscess in the left lobe ( Figure 2a) and two patients one abscess in each lobe. The size of the abscesses varied from 5 to 16 cm in diameter (mean 8.5 cm). The volume of fluid drained ranged from 55 ml to 290 ml. Twelve patients experienced no pain after the procedure, five complained of mild pain and two of moderately severe pain. There were no serious complications. All abscesses were successfully drained. Clinical improvement in terms of pain reduction and abatement of fever was observed in all patients 24–48 h after drainage. On follow-up after discharge, all abscesses had completely resolved after 2–4 weeks. The procedure not only expedited recovery, but was curative in all 19 patients.

Figure 1.

Contrast enhanced axial CT of the liver showing a large abscess in the right lobe. The patient was a 39-year-old male who presented with right hypochondrial pain and fever following a bout of diarrhoea.

Figure 2.

(a) Contrast enhanced CT of the liver. 54 year-old male with upper abdominal pain and fever. A multiloculated abscess with internal septations is visible in the left lobe.

Figure 2(b) Contrast enhanced CT of the same patient after PCD: The abscess has resolved completely.


Infection with the protozoan Entamoeba histolytica occurs worldwide. 10% of the world's population is said to be infected with this organism ( Merten & Kirks 1984). Liver abscess is the most common extraintestinal manifestation of amoebiasis and occurs in 3% to 9% of patients with amoebic infection ( Debakey & Ochsner 1959; Landay et al. 1980 ; Peters et al. 1981 ).

Diagnosis of amoebic liver abscess is usually established on the basis of the clinical, epidemiological, aetiological and imaging findings ( Ralls 1998). Serological tests, however, have false negative rates varying from 2% to 18% depending on the type of test ( Van Sonnenberg et al. 1985 ). Negative serology does not therefore conclusively exclude the diagnosis. Both CT scan and sonography are sensitive and accurate in diagnosing amoebic liver abscess ( Ralls 1998). The sonographic and CT features have been extensively described ( Ralls et al. 1982 ; 1987a; Barnes et al. 1987 ; Radin et al. 1988 ; Ralls 1998). But imaging is not specific and may be confused with that of pyogenic abscesses or other cystic lesions including tumours ( Ralls et al. 1987a ). Based on clinical features and imaging findings, it may be difficult to distinguish an amoebic from a pyogenic abscess. Abdominal pain and diarrhoea are more common in patients with amoebic abscess. Although it is well known that patients with amoebic abscess are more often acutely ill and toxic than patients with pyogenic abscess ( Ralls 1998), most of our patients (13/19) were neither acutely ill nor toxic. Amoebic liver abscess usually presents as primary condition, whilst pyogenic abscesses are generally a complication of pre-existing illness ( Ralls 1998). Patients with amoebic liver abscess are likely to originate from endemic areas (as did all our patients) or are recent travellers to high prevalence areas. Characteristic sonographic features of amoebic abscesses include absence of significant wall echoes, round or oval shape, homogeneous hypoechoic lesion with low-level internal echoes, near or touching the liver capsule and distal acoustic enhancement ( Ralls et al. 1982 ). The most reliable of these features are rounded or oval shape and a hypoechoic appearance with homogeneous low-level echoes ( Figure 3) ( Ralls et al. 1987a ).

Figure 3.

Abdominal ultrasonography of a 41-years-old male with right upper quadrant pain and fever following an attack of diarrhoea. An abscess measuring 79 × 58 mm, with low level internal echoes is shown in the right lobe of the liver.

CT features of amoebic liver abscesses include rounded, well-defined hypodense lesions with an enhancing wall and a peripheral zone of oedema. Internal septations ( Figure 2a) or fluid debris levels may be seen. Because of the nonspecificity of these imaging features, image-guided needle aspiration is occasionally required to establish the diagnosis and exclude pyogenic abscess. Although fluid aspirated from an amoebic abscess is quite variable in appearance ( Ralls 1998), the fluid from all our patients had the classical reddish brown anchovy sauce appearance.

Once the diagnosis of amoebic liver abscess is established, prompt treatment becomes necessary since failure to treat may result in rupture of the abscess, a complication that is associated with high mortality ( Baijal et al. 1995 ). Medical treatment is generally effective in the management. All 96 patients in one series were successfully treated medically ( Ralls et al. 1987b ). Thirteen of them (13.5%) had sonographically guided needle aspiration because of nonspecific imaging features. Only one patient required therapeutic aspiration because response to medical therapy was slow. No patient required therapeutic drainage.

Opinion differs on the role of drainage (by needle aspiration or catheter) in the management of amoebic liver abscess ( Rajak et al. 1998 ). Some believe that drainage is a helpful and necessary adjunct to amoebicidal therapy ( Van Sonnenberg et al. 1985 ), others doubt its usefulness and insist that drainage is very rarely necessary ( Ralls 1998). Resistance to drug therapy is, however, a well-known complication ( Singh & Kashyap 1989), and in such circumstances, most would consider drainage imperative. When drainage is necessary, PCD produces the best results. In a comparative study of 56 patients with drug-resistant abscesses ( Singh & Kashyap 1989), PCD produced the most impressive results in terms of cure compared to repeat trial of amoebicidal therapy, therapeutic needle aspiration and surgical drainage. PCD has also been successful in managing the sequelae of rupture of these abscesses ( Baijal et al. 1995 ) and, in cases of amoebic abscess-biliary communication, although a prolonged period of drainage may be necessary in such instances ( Agarwal et al. 1995 ).

Owing to the lack of adequate response to amoebicidal therapy in our patients, PCD was carried out as an adjunct to drug therapy. There was rapid clinical improvement in all 19 patients, recovery was expedited and all abscesses eventually completely resolved. Although this was not a comparative study and there were no controls, we feel justified, in view of our excellent results, to advocate prompt PCD in all cases of amoebic liver abscess that fail to respond to amoebicidal therapy. Finally, despite the controversies, there are certain other clear-cut situations when drainage of amoebic liver abscesses becomes mandatory. These include bacterial superinfection of the abscess ( Donovan et al. 1991 ) and imminent intrapericardial rupture ( Van Sonnenberg et al. 1985 ; Singh & Kashyap 1989; Ralls 1998) although these are extremely rare events ( Ibarra-Perez 1981).


We thank Mrs. Viyena Rodrigues for secretarial help.