Outcomes of a conservative approach to management in amoebic liver abscess.

CONTEXT: Unfortunately, there is confusion among the medical community regarding the management of amoebic liver abscess (ALA). Therapeutic options range from simple pharmacotherapy to use of interventions like a needle or catheter aspiration under ultrasound guidance to surgical intervention. There is a plethora of thresholds for parameters such as the maximum diameter of the abscess and volume on ultrasound examination suggested by various authors to serve as a criterion to help to decide when to use which modality in these cases. AIMS: To assess the outcome of patients with uncomplicated ALA treated using a conservative approach. Moreover, to identify factors associated with its failure. SETTINGS AND
DESIGN: A prospective, observational study was carried out at a large municipal urban health care center over a period of 3-year (2011-2014) in India.
MATERIALS AND METHODS: Patients with uncomplicated ALA were recruited. All patients were managed with pharmacotherapy initially for a period of 72 h. Response to treatment was assessed by resolution of symptoms within the given time frame. Failure to respond was considered an indication for intervention. Needle aspiration was offered to these patients and response assessed within 72 h. Failure to respond to aspiration was considered an indication for catheter drainage. STATISTICAL ANALYSIS USED: Data recorded were entered in a Microsoft Office Excel Sheet and analyzed using the SPSS version 16.0 (IBM).
RESULTS: Sixty patients with ALA were included in the study over its duration. Forty-nine (81.67%) patients were managed conservatively, while 11 (18.33%) patients needed an intervention for relief. Patients who required intervention had deranged liver function at presentation, a larger abscess diameter (10.09 ± 2.23 vs. 6.33 ± 1.69 cm P < 0.001) and volume (399.73 ± 244.46 vs. 138.34 ± 117.85 ml, P < 0.001) compared to those who did not need it. Patients that required intervention had a longer length of hospital stay (7.1 ± 2.4 vs. 4.8 ± 0.9 days, P < 0.001). On post hoc analysis, a maximum diameter of >7.7 cm was found to be the optimal criterion to predict the need of intervention in cases of ALA.
CONCLUSIONS: A conservative approach is effective in the management of ALA for a majority of patients. Failure of conservative management was predicted by the size of the abscess (maximum diameter >7.7 cm). Even in the cases of failure, a gradual step-up with interventions was found to be safe and effective.

Introduction

Amoebic liver abscess (ALA) should be excluded in all patients presenting with right-sided upper abdominal pain or right lower thoracic pain, with or without fever.[1] If diagnosed early, ALA is readily treatable and its mortality is negligible. When left untreated, it may lead to life-threatening complications such as rupture into the peritoneal, pleural, or pericardial cavity. Drugs from the nitroimidazole group of antibiotics are the agents of choice against amoebae. Apart from pharmacotherapy, another aspect of treatment of ALA is the management of necrotic debris or pus within the abscess cavity. Unfortunately, there is a lack of consensus among the medical community regarding the need of intervention in such cases. This is made evident by the various recommendations seen in current literature. Recommendations have ranged from conservative thresholds such as maximum diameter >5 cm, bilobar involvement, and age >55 years,[2] while some authors have had more liberal thresholds of maximum diameter >10.7 cm and intervention only in the absence of response to drugs.[3] In the past varied volume, thresholds have been used to assess the need of intervention however the thresholds are often arbitrary and may not always have a strong evidence base.[4] In the authors’ view due to the confusion in available literature, the only reliable criterion to decide when to intervene in cases of uncomplicated ALA seems to be a lack of clinical response to pharmacologic agents. Response to treatment has been assessed by symptomatic relief, absence of fever, and right upper quadrant tenderness at the end of the first 72 h of hospital stay.[5] Thereafter, in cases that fail to respond, interventions such as needle aspiration, catheter drainage, or surgical interventions can be employed as required. Therefore, a decision was made to carry out a study to assess the outcomes of a conservative management policy in the cases of ALA.

Materials and Methods

A prospective, observational study was carried out at a large municipal health care center in Mumbai, India, over a period of 3-year from September 2011 to July 2014. Approval from the Institutional Ethics Committee was obtained prior to the commencement of the study. Consecutive adult (>18 years) patients admitted to the general surgical ward with the diagnosis of ALA were screened for eligibility to participate in the study. The diagnosis of ALA was confirmed by a positive enzyme-linked immunosorbent assay (ELISA) and ultrasonographic (USG) evidence of liver abscess. An informed written consent for the study participation was obtained from each patient. Patient selection and management algorithm is shown in Figure 1. All patients in the study were admitted under the care of a single surgical team with a large experience in general and hepatobiliary surgery.

Figure 1

Flowchart showing patient selection and management algorithm

At USG, the abscess was characterized by site (lobe), size (dimensions, recorded as the largest dimension of the cavity), number (one vs. multiple), distance from the liver capsule (measured from the liver capsule to the point of the abscess closest to it), vascular, and biliary structures. In cases with multiple abscesses, the site and size of the larger abscess cavity were considered as the dimension of interest. ALA with subcapsular or free peritoneal/pleural rupture was considered complicated. ELISA was used to detect the antiamoebic IgG antibody in the serum. A value of more than 0.4 optical density units obtained by ELISA was considered positive. Metronidazole therapy was administered as an oral formulation 750 mg 3 times a day to all patients, irrespective of response or need for intervention, for a total of 10 days. Parenteral formulation (500 mg, 3 times a day intravenous) was administered in case of inability to consume the medication orally. Oral tablets were reinstated once tolerated. Therapy was continued under supervision as long as patients remained in hospital after which oral formulation was given to them for home use. Compliance was assessed by the return of empty drug packets at follow-up. Ultrasound-guided needle aspiration or pigtail catheter insertion was considered interventions. Aspiration was done with a long 18 gauge needle till the evacuation of all liquefied contents. A pigtail catheter (12–18 Fr) was used to drain the abscess when deemed necessary. It was left in situ till 24 h output decreased to <30 ml or an ultrasound revealed no residual liquid content within the abscess cavity. A contingency plan for the management of an unforeseen complication (free peritoneal rupture) while patients were enrolled in the study was put in place.

Data recorded were entered in a Microsoft Office Excel Sheet and analyzed using the SPSS version 16.0 (IBM). The data were presented using frequency, percentage, and cross tables. Patients were grouped based on whether they responded to conservative treatment or required intervention. Subsequently, the Fischer's test, Chi-square test for qualitative data, and Student's t-test for quantitative data were used as tests of significance for appropriate comparison and P < 0.05 was taken as significance. Logistic regression was carried out using an independent variable with P < 0.15 to analyze whether they significantly associated with the final outcome. A receiver operating characteristic (ROC) curve was plotted with the parameter of interest.

Results

Over the duration of the study at our center, 60 out of 86 patients of ALA were recruited for the study. The clinical profile of the study population is shown in Table 1. The distribution of patients according to their response to therapy is shown in Figure 2. Post hoc analysis was carried out comparing the group that responded to only conservative management (Group 1) versus the group that needed some form of intervention (Group 2). The comparison of the demographic profiles and laboratory investigations on admission are shown in Table 2. The USG features observed at index examination and the comparison between the groups are shown in Table 3. The coefficient of correlation between the maximum diameter and volume was + 0.9. The results of logistic regression analysis are shown in Table 4. A ROC curve was plotted [Figure 3] using the maximum diameter of the abscess. The criterion associated with the Youden index (a point on the curve where the trade-off between sensitivity and specificity was the least) for this curve was at >7.7 cm (95% confidence interval - >7.5 to >10 cm, sensitivity - 100%, and specificity - 81.63%). The length of hospital stay in Group 1 was 4.8 ± 0.9 days versus 7.1 ± 2.4 days in Group 2, which was significantly different (P < 0.001, by Student's t-test). There were no untoward complications in those who underwent conservative management or intervention. We had a 100% follow-up until 6 weeks after discharge. All patients were well at the time.

Table 1

Clinical profile of study population

Parameter	Result
Age
Mean±SD (years)	38.5±13.5
Sex (n)
Male	58
Female	2
Symptomatology (n)
Pain in abdomen	56
Fever	38
Anorexia	32
Vomiting	10
Diarrhea	11
Jaundice	6
Respiratory complaints	5

SD: Standard deviation

Figure 2

Flowchart showing the distribution of patients as per therapy offered

Table 2

Demographic profile and laboratory parameters in study population

Parameter	Variable	Group 1 (n=49)	Group 2 (n=11)	P
Demographic
Sex*	Male	47	11	0.31
	Female	2	0
Age (in years), mean±SD#		38.9±13.2	42.9±12.7	0.37
Laboratory# (mean±SD)	Hemoglobin	11.74±01.96	12.00±02.63	0.72
	Leukocyte count	16,041.25±6281.18	16,436.36±8090.52	0.86
	Platelet count	254,361.22±127,959.54	214,272.73±149,715.79	0.41
	Albumin	3.12±00.41	2.79±00.47	0.02a
	Globulin	3.44±00.56	3.57±00.69	0.49
	Total bilirubin	0.58±00.55	1.79±01.35	0.04a
	AST	46.86±42.41	118.09±172.89	0.02a
	ALT	42.22±25.86	92.00±131.91	0.02a
	ALP	221.14±103.60	358.55±310.89	0.01a
	Prothrombin time	16.31±04.36	16.27±02.72	0.97

*Analyzed using Fischer's test, aP<0.05, significant difference, #Analyzed using Student's t-test. AST: Aspartate aminotransferase, ALT: Alanine transaminase, ALP: Alkaline phosphatase, SD: Standard deviation

Table 3

Ultrasonographic parameters in study population

Parameter	Variable	Group 1 (n=49)	Group 2 (n=11)	P
Number*	Single	35	11	0.61
	Multiple	14	0
Site*	Right lobe	41	9	0.16
	Left lobe	4	2
	Both	4	0
Maximum diameter# (cm), mean±SD		6.33±1.69	10.09±2.23	<0.001a
Volume (ml)#, mean±SD		138.34±117.85399.73±244.46<0.001a

*Analyzed using Fischer's test, aP<0.05, significant difference, #Analyzed using Student's t-test. SD: Standard deviation

Table 4

Result of logistic regression analyses on parameters associated with outcome

Variable	Coefficient (B)	SE	Wald statistic (B/SE)2	OR	95% CI	P
Albumin	−2.06378	1.3881	2.2105	0.127	0.0084-1.9288	0.1371
Bilirubin	0.33326	0.4744	0.4935	1.3955	0.5507-3.5363	0.4824
AST	−0.018508	0.014714	1.5823	0.9817	0.9538-1.0104	0.2084
ALT	0.053971	0.035723	2.2826	1.0555	0.9841-1.1320	0.1308
Maximum diameter	3.60407	1.62104	4.9431	36.7476	1.5324-881.2223	0.0262a
Volume	−0.026434	0.015155	3.0423	0.9739	0.9454-1.0033	0.0811

aP<0.05, significant association. Overall model fit significance level P<0.001. AST: Aspartate aminotransferase, ALT: Alanine transaminase, SE: Standard error, OR: Odds ratio, CI: Confidence interval

Figure 3

Receiver operating characteristic curve plotted using maximum diameter, table below shows the area under curve, standard error, confidence interval, and P value

Discussion

The study shows that 81.67% of patients presenting with uncomplicated ALA can successfully be treated on pharmacotherapy alone. Only 18.33% of patients that presented with ALA ultimately required intervention. Independent variables that were significantly associated with patients who underwent intervention were deranged liver function tests, higher maximum diameter, and volumes as assessed by ultrasound examination. A possible explanation for this is given by the fact that the group that needed intervention had larger abscesses. There are several hypotheses regarding how large abscesses lead to derangement in liver function, external compression of biliary apparatus, pressure on surrounding parenchyma, hepatic necrosis at the margins of the abscess, and vascular or biliary invasion.[678] Larger abscesses are also known to lead to increased pain and discomfort due to pressure effects and the stretch they produce on the Glisson's capsule. On the contrary, patients that did not need intervention had smaller abscesses (6.33 ± 1.69 cm in diameter and 138.34 ± 117.85 ml by volume). Further analysis revealed that a maximum diameter of >7.7 cm was able to predict the need for intervention accurately (sensitivity - 100% and specificity - 81.63%). Even though these patients ultimately required intervention, it was found safe to wait for 72 h in these patients.

The apparent success of conservative management, we believe, lays in the timely administration of effective pharmacologic agents, namely drugs from the nitroimidazole groups. This may be the basis on which modern treatment algorithms for ALA can be formulated. In them, we can afford to heavily rely on conservative management. The response is noticeable even with a single dose of these drugs, but it is advisable to wait for 72–96 h to judge symptomatic or clinical resolution.[5] The response may be judged by a reduction in pain, resolution of fever, and an increased sense of well-being.[5] In one particular study, the drugs are said to be so effective that there was no significant difference in clinical response in the group treated with metronidazole alone versus the group treated with metronidazole and percutaneous aspiration.[9] An interesting question raised by the above observation is why are drugs alone not effective in the treatment of large abscesses? Since most studies including ours assess abscess size at presentation, it is representative of the condition in a drug-naïve state. Once drugs are given, it leads to eradication of amoebae within liver tissue usually within 72 h of beginning therapy.[5] However, in large abscesses, this does not lead to an immediate reversal of pressure effects which result in pain/discomfort, as well as other symptoms like jaundice.[7] Since the drugs have no action in terms of mechanical clearance of necrotic debris, drainage is the only option that leads to relief. Drainage leads to decreased pressure on surrounding parenchyma and leads to a reduction in pain and discomfort. In smaller abscesses, the sterile pus (after drug therapy) is cleared by the normal immune clearance pathways. This is a slow process evident by the slow resolution of ALA noted on subsequent USG. It usually requires pharmacotherapy for a limited period of time and routine follow-up and need not be monitored by serial USG in the follow-up period.[4]

There are some obvious limitations of the study in question. The sample size is small, and we had a short follow-up period. It includes patients with uncomplicated ALA and excludes patients with subcapsular/free peritoneal rupture. While the diagnosis of ALA in the presence of positive serology and the USG findings in areas of high endemicity is relatively straightforward, a rare problem encountered is when an abscess detected on the USG yields negative results on serology. We believe in such cases that diagnostic aspiration should be carried out. An unfortunate but expected drawback of using a conservative management protocol has been a prolonged length of hospital stay in the group of patients that ultimately require intervention. Several studies have shown a shorter length of hospital stay in early aspiration groups, but results from a systematic review from the Cochrane database suggests that current evidence is not sufficient to support or refute that aspiration hastens recovery in uncomplicated ALA.[10] Therefore, future studies should be well-planned and use liberal thresholds like maximum diameter >7.7 cm to decide when to intervene early. It would be interesting to note if this could reduce the mean LOS in patients that ultimately require intervention.

Conclusion

A conservative approach to the management of uncomplicated ALA is safe and effective. Treatment algorithms may be tailored to give an appropriate trial of conservative management before intervening.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.