Patients with culture negative pyogenic liver abscess have the same outcomes compared to those with Klebsiella pneumoniae pyogenic liver abscess

Vishal G Shelat， Qiao Wang， Clement LK Chia， Zhongkai Wang，Jee Keem Low and Winston WL Woon

Singapore， Singapore

Patients with culture negative pyogenic liver abscess have the same outcomes compared to those with Klebsiella pneumoniae pyogenic liver abscess

Vishal G Shelat， Qiao Wang， Clement LK Chia， Zhongkai Wang，Jee Keem Low and Winston WL Woon

Singapore， Singapore

BACKGROUND： Etiologic organism is not frequently isolated despite multiple blood and fluid cultures during management of pyogenic liver abscess （PLA）. Such culture negative pyogenic liver abscess （CNPLA） is routinely managed by antibiotics targeted to Klebsiella pneumoniae. In this study， we evaluated the outcomes of such clinical practice.

METHODS： All the patients with CNPLA and Klebsiella pneumoniae PLA （KPPLA） admitted from January 2003 to December 2011 were included in the study. A retrospective review of medical records was performed and demographic， clinical and outcome data were collected.

RESULTS： A total of 528 patients were treated as CNPLA or KPPLA over the study period. CNPLA presented more commonly with abdominal pain （P=0.024）. KPPLA was more common in older age （P=0.029） and was associated with thrombocytopenia （P=0.001）， elevated creatinine （P=0.002）， bilirubin（P=0.001）， alanine aminotransferase （P=0.006） and C-reactive protein level （P=0.036）. CNPLA patients tend to have anemia（P=0.015） and smaller abscess （P=0.008）. There was no difference in hospital stay （15.7 vs 16.8 days） or mortality （14.0% vs 11.0%）. No patients required surgical drainage after initiation of medical therapy.

CONCLUSION： Despite demographic and clinical differences between CNPLA and KPPLA， overall outcomes are not different.

（Hepatobiliary Pancreat Dis Int 2016；15：504-511）

culture negative；

Klebsiella pneumoniae；

pyogenic liver abscess

Introduction

Pyogenic liver abscess （PLA） is a potentially lifethreatening infection with mortality rate ranging from 4%-30%.［1-3］The pathogens for this condition are detected by culture of blood and/or pus drained from the abscess. Among all pathogens found in PLA，Klebsiella pneumoniae is the commonest and the features of Klebsiella pneumoniae PLA （KPPLA） have been widely described， particularly from Southeast Asian countries.［4-7］Absence of routine widespread application of advanced microbial detection techniques，［8，9］noncompliance to sepsis guidelines［10］and absence of routine testing for fungal infections result in inability to identify the etiologic microorganism. Such culture negative pyogenic liver abscess （CNPLA） deserves a special recognition as the selection of antibiotic is not standardized. The incidence of CNPLA varies from 15%-80%［11-13］and despite being commonly encountered， it has not received the same attention as KPPLA.［14］It is a common practice to empirically treat CNPLA patients with the same antibiotics which are commonly used for KPPLA patients. It remains unclear if this approach is safe. Furthermore，CNPLA is more common when an attempt to obtain pus culture by aspiration or drainage has not been done. It is evident that outcomes of patients whose PLA is nottreated by drainage are inferior compared to drainage.［15］Also， routinely patients with positive blood cultures are treated with minimum 10-14 days of intravenous antibiotics and it is possible that CNPLA patients are undertreated. We conducted this study to evaluate if CNPLA results in adverse outcomes comparable to KPPLA.

Methods

Patients

A diagnosis of PLA was established based on clinical presentation and imaging. Patients with age ＜18 years，amoebic liver abscess， infected liver cyst， infected hydatid cyst， ruptured liver abscess requiring an urgent surgical intervention and culture positive for any bacteria other than Klebsiella pneumoniae were excluded. Constitutional symptom was defined collectively for vague presenting complains of weakness， fatigue， fever and feeling unwell. All patients with PLA had at least one set of blood culture taken prior to the initiation of empirical broad spectrum intravenous antibiotics. Patients underwent percutaneous drainage （or aspiration） of PLA when any of the following criteria was met： （1） size of PLA ＞4 cm， solitary or dominant； （2） presence of hemodynamic instability or need for inotropic support on admission；（3） gas within the abscess cavity regardless of size and（4） failure of antibiotic therapy for PLA ＜4 cm. Drain fluid was routinely sent for microbiology testing. Drains were flushed with 10 mL saline to prevent blockages and accurate logs were maintained for the quality and quantity of effluents. A contrast dye study via the drain was routinely obtained when the drain output was ＜10 mL/24 h for at least two days. The percutaneous drains were removed upon resolution of sepsis as evidenced by stable vital parameters， total white blood cell count and/ or C-reactive protein （CRP） levels down trending and＜10 mL/24 h drain aspirates for at least two consecutive days.

Groups and treatment

The specimens obtained were processed for Gram stain， bacterial cultures （aerobic and anaerobic） and tests for antibiotic susceptibility ［organism identification was performed using a mix of Microbact? 12A （Oxoid， UK），VITEK 2 system （bioMe'rieux）， and the Bruker Biotyper（version 2.0） matrix-assisted laser desorption/ionization time-of-flight （MALDI-TOF） mass spectrometry system；susceptibility testing was performed using the VITEK 2 system （bioMe'rieux） and Kirby-Bauer method］. Antibiotic therapy was subsequently tailored to culture and sensitivity results. In KPPLA patients， either blood or pus culture was positive for Klebsiella penumoniae. CNPLA was defined when no organism was evident on any blood or pus samples. PLA patients were initially taken empirical broad spectrum intravenous antibiotics， usually a combination of a third generation cephalosporin， such as ceftriaxone 1 g intravenous 8 hourly， along with metronidazole 500 mg intravenous 8 hourly. Total duration of antibiotic therapy varied according to the clinical and radiological response. Patient monitoring and blood investigations were done based on individual surgeons' preference. All the patients underwent a repeat radiologic imaging at 2-3 weeks interval to evaluate radiological response. Antibiotic therapy was switched from intravenous to oral based on clinical response and discontinued when there was a near complete to complete radiological resolution of abscess. Patients with positive blood culture were routinely treated with minimum 10-14 days of intravenous antibiotics before converting to oral. All patients with persistent sepsis were evaluated according to surviving sepsis guidelines with repeat imaging and culture directed escalation of antibiotic treatment. It is a local practice to escalate antibiotic treatment to either piperacillin-tazobactam 4.5 g intravenous 8 hourly or carbapenems in septic CNPLA patients who do not respond to first line treatment. In patients with negative cultures， metronidazole was discontinued at the time of switching intravenous antibiotics to oral amoxicillin clavulanate. The medical records of all patients were studied for demographic profile， co-morbidities， clinical features， laboratory and radiologic investigations， imaging characteristics of abscesses， treatment， hospital stay，and mortality. All the cut-off values of laboratory investigations were based on our hospital laboratory records except for CRP where an arbitrary cut-off value of 100 mg/L was determined. Total duration of antibiotic treatment was determined by adding the intravenous and oral days of all types of antibiotics prescribed. Mortality was defined as death due to any cause during hospitalization or within 30 days of diagnosis of PLA.

Statistical analysis

Data was analyzed using commercial statistical software （SPSS for Windows， version 14.0， Chicago， IL， USA）. All continuous data were expressed as mean±standard deviation （SD） and analyzed by independent samples t-test. All categorical variables were described as percentage and compared by either Chi-square or Fisher's exact test based on the feature of the data. Multivariate analysis for mortality was processed by logistic regression using stepwise forward method， all factors analyzed in the single factor study were included as inputs. A P value of＜0.05 was considered statistically significant.

Results

Baseline characteristics

During January 2003 and December 2011， 635 patients were treated for PLA at Tan Tock Seng Hospital in Singapore. Fig. 1 demonstrates the patient selection. 528 patients were treated for CNPLA （264 patients） or KPPLA （264 patients）. This equal number is purely coincidental. The mean age of patients was 60.7 years （range 15-99） and 322 （61.0%） patients were male. 191 （36.2%）patients had history of hepato-pancreatico-biliary diseases， in which， the majority was gallstones （74.3%）. Table 1 shows the baseline characteristics of patients with CNPLA and KPPLA. KPPLA patients had blood or pus culture done for a median of 4 times （range 2-13） and CNPLA patients had blood or pus culture taken for a median of 4 times （range 2-15）.

Clinical presentation and investigations

Clinical features， biochemical results and imagingfeatures are shown in Table 2. Fever and constitutional symptoms were the most common clinical manifestations in both groups but significantly more in KPPLA patients （85.2% vs 68.6%， P=0.001， and 83.0% vs 68.9%，P=0.001， respectively）. Abdominal pain was more common in CNPLA group （53.0% vs 43.2%， P=0.024）. KPPLA patients showed evidence of organ dysfunction withmore patients having thrombocytopenia （platelet ＜150 × 109/L） （35.6% vs 15.5%， P=0.001）， elevated creatinine（＞130 μmoI/L） （32.2% vs 20.1%， P=0.002）， elevated bilirubin （＞22 μmol/L） （56.1% vs 41.7%， P=0.001）， raised aspartate aminotransferase （AST） （＞34 IU/L） （76.9% vs 67.8%， P=0.025） and raised alanine aminotransferase （ALT） （＞44 IU/L） （65.2% vs 53.0%， P=0.006） level. High CRP level was also more common in KPPLA group（68.9% vs 60.2%， P=0.036）. CNPLA patients tend to have anemia （17.0% vs 9.8%， P=0.015） and smaller abscesses（5.4 vs 6.2 cm， P=0.008）. There was equal distribution of multiple lesions （29.9% vs 30.3%， P=0.992）.

Fig. 1. Flowchart of patient selection.

Table 1. Baseline characteristics of patients with CNPLA and KPPLA

Table 2. Clinical presentation and investigations of CNPLA and KPPLA patients

Table 3. Management and clinical outcomes of CNPLA and KPPLA patients （n， %）

Management and clinical outcomes

KPPLA patients required percutaneous drainag （63.6% vs 34.8%， P=0.001） more frequently. No patients required surgical drainage after initiation of medical therapy. There was no difference in hospital stay （15.7 vs 16.8 days， P=0.397）. Overall 30-day mortality of CNPLA and KPPLA patients（14.0% vs 11.0%， P=0.292） was similar. Mortality in patients with percutaneous drainage was 8.7% and 9.5%（P=0.825） and of those who were treated with antibiotics alone was 16.9% and 13.5% （P=0.474） for CNPLA and KPPLA groups， respectively （Table 3）.

CNPLA patients outcomes

Those patients who had the abscess drained in the CNPLA group， were younger （P=0.05） with larger abscess （7.4 vs 4.2 cm， P＜0.001） and fever （78.3% vs 63.4%，P=0.013） as common presenting complain. CNPLA patients treated with drainage also had deranged （＞12 × 09/L or ＜4×109/L） white blood cells （71.7% vs 57.0%，P=0.018） and elevated gamma-glutamyl transferase（GGT） levels （89.1% vs 76.7%， P=0.014） compared to CNPLA patients treated with antibiotic alone. Mortality was lower in patients treated with drainage but not significant （8.7% vs 16.9%， P=0.069） （Table 4）. The multivariate analysis showed that old age， absence of fever，elevated blood urea and coagulopathy as evidenced by elevated international normalized ratio （INR） predicted mortality in CNPLA patients （Table 5）.

Table 4. Demographic， clinical， radiological profile and outcomes of CNPLA patients treated with and without percutaneous drainage

Discussion

CNPLA is common but poorly studied and under reported. In this study involving 264 patients with CNPLA，we are the first to demonstrate that empirical treatment of CNPLA according to KPPLA is an acceptable clinical practice. This study also confirms that CNPLA patients with percutaneous drainage have better outcomes com-pared to those treated with antibiotics alone. This is the first study to report variables that predict mortality in CNPLA patients by performing a multivariate analysis on a large population.

Table 5. Univariate and multivariate analysis of CNPLA patients for mortality

Antibiotic treatment is a minimum universal requirement of all PLA patients. Ideally antibiotic therapy should be targeted to the culprit microorganism. In PLA patients， blood culture is the prime source for microbial identification. Chemaly et al［16］have found that the sensitivities of the blood cultures for any Gram-positive cocci and Gram-negative bacilli present in PLA were 30% and 39% respectively， which means that PLA pathogens are usually underrepresented in blood cultures. Hence alternative means of microbial identification are important. Percutaneous aspiration or drainage provides pus which serves dual purpose of microbial isolation and therapeutic abscess drainage. Patients who have PLA aspiration or drainage will have pus available for bacteriology testing and hence will have opportunity of microbial isolation and targeted therapy. In our previous experience of managing 741 patients with PLA， only 60.1% （220/366）had positive pus cultures.［17］This is likely because antibiotics are almost always prescribed prior to drainage. Recently， advanced microbial detection techniques like polymerase chain reaction （PCR） have been employed not only to enhance the diagnostic yield but also to facilitate early bacteriological diagnosis.［18］Microorganism detection rate of PCR along with routine culture meth-ods was significantly better compared to each modality alone.［18］Higher incidence of CNPLA in patients treated without percutaneous drainage is confirmed in our study. Also， in clinical practice， more sick patients are managed aggressively with repeated testing for bacteriology. This can enhance the bacterial yield in such patients. This is confirmed in our experience where more sick patients belong to KPPLA group. It is our routine practice to treat CNPLA patients empirically according to the treatment of the commonest pathogen i.e. Klebsiella pneumoniae. It is challenging that some patients may fail to respond to treatment， especially in the absence of bacteriology results. This could increase treatment failures and adverse outcomes in CNPLA patients. It is hence important to evaluate if local practice of empirical treatment of CNPLA patients according to KPPLA algorithm is acceptable. Fig. 2 shows the dilemma associated with CNPLA.

Fig. 2. Dilemma of culture negative pyogenic liver abscess.

CNPLA patients universally receive empirical antibiotics and selectively receive percutaneous aspiration/ drainage. Khan et al［14］have reported that predictive factors for aspiration of liver abscess included age ≥55 years，size of abscess ≥5 cm， involvement of both lobes of the liver， and duration of symptoms ≥7 days. Considerable success rate of needle aspiration of PLA， even in multiple abscesses， was reported in different series.［19］However，up to 40% of PLA patients require repeated aspiration sessions.［20］Percutaneous drainage is a safe intervention which is tolerated well by patients and thick pus can be drained more efficiently.［21，22］We do not perform percutaneous aspiration but offer percutaneous drainage in selected patients based on abscess size， sepsis severity，presence of gas in abscess cavity and failure of antibiotic treatment. A 4-cm size cut-off is widely recommended for acute colonic diverticular abscess and recently a new classification system is proposed based on this size criterion.［23］We have previously published 4 cm cut-off for consideration of percutaneous drainage of PLA.［24］Our results showed that KPPLA patients present with severe sepsis and larger abscess size compared to CNPLA patients. However due to retrospective nature of our study，the cause and effect relationship cannot be established as it is possible that more sick patients underwent aggressive management with repeated tests for bacteriology increasing the microbial isolation. Also patients with larger abscess size underwent more frequent percutaneous drainage providing additional source of microbial isolation and increasing yield. We do not consider large size as an indication of surgery and have reported that percutaneous drainage is safe and sufficient even in ≥10 cm PLA.［25］In our study， almost twice the number of patients received percutaneous drainage in the KPPLA group. Despite the lower proportion of patients undergoing percutaneous drainage in CNPLA group，the overall mortality was similar to KPPLA group. In a subgroup of patients treated with antibiotic alone， there was no difference in mortality among the two groups. Similarly in a subgroup of patients treated with drainage， there was no difference in mortality among the two groups either. However a subgroup analysis of CNPLA patients demonstrated that， in patients treated with antibiotics alone， mortality was higher compared to CNPLA patients treated with drainage. This may suggest a more liberal policy of drainage for therapeutic efficacy rather than simple aspiration for improving the microbial yield. Moreover， these results also suggest that a retroactive strategy of percutaneous drainage in cases of failure of antibiotic treatment could be a missed opportunity to improve outcomes by early drainage. Some reports show that Klebsiella pneumoniae is more likely to appear as single abscess［26］and uni-lobular involvements［27］than non-Klebsiella pneumoniae liver abscess， but this was not confirmed in our study and hence the outcomes are not determined by number and focality of PLA location. Prospective study is required to confirm these findings.

KPPLA is reported to have a mortality rate of 4%-12%［4-7］and there is limited data for CNPLA. Khan et al［14］reported 2.3% mortality rate in a series of 966 patients including 661 （68%） of patients with amoebic etiology. In our series， the overall mortality rate was 11.0% and 14.0% for KPPLA and CNPLA groups， respectively. Khan's study predominantly included young patients （71% were ＜55 years）， only 17.5% incidence of diabetes and none patient with PLA having cholangitis. Advanced age， presence of sepsis， underlying malignancy，poor physiological state and co-morbidities， are predictors of mortality in PLA.［2，11，28-30］In our study， various factors had been analyzed to reveal the risk of mortality for patients with CNPLA. Multivariate analysis revealedold age， fever， coagulopathy and elevated urea predict mortality in CNPLA patients. These results emphasize the fact that elderly patients who are unable to mount a febrile response to sepsis are vulnerable and high index of suspicion with prompt diagnosis and management are recommended. Elevated urea has been shown to be a predictor of poor outcomes including mortality in patients with acute pancreatitis， acute cholecystitis and perforated peptic ulcer.［31-33］Blood urea nitrogen is a surrogate marker of intravascular volume status and provides valuable feedback with regards to fluid resuscitation. Coagulopthy is a marker of organ dysfunction in severe sepsis and has been shown to be associated with mortality in PLA.［34］It is possible that percutaneous drainage is delayed until the coagulopathy is artificially corrected. In a study of 103 elderly patients with acute cholecystitis，we have shown that early percutaneous cholecystostomy improves outcomes.［35］While this large single institution study provides novel information on under reported and poorly studied disease of CNPLA， it is limited by retrospective nature.

In conclusion， KPPLA and CNPLA have several demographic and clinical differences. CNPLA patients are less likely to present with severe sepsis and renal dysfunction. CNPLA patients are also more likely to have a smaller abscess size which is less likely to be drained. There is an increased mortality in CNPLA patients who did not receive percutaneous drainage. The overall outcomes of CNPLA patients are equivalent to KPPLA patients and hence it is justified to treat CNPLA patients with empirical antibiotics targeted to Klebsiella pneumoniae.

Contributors： SVG proposed the study. WQ and CCLK performed the research and wrote the first draft. WQ collected and analyzed the data. All authors contributed to the design and interpretation of the study and to further drafts. SVG and WQ contributed equally to this study. SVG is the guarantor.

Funding： None.

Ethical approval： This study was approved by Institutional Review Board of Tan Tock Seng Hospital.

Competing interest： No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

1 Alvarez Pérez JA， González JJ， Baldonedo RF， Sanz L， Carre?o G， Junco A， et al. Clinical course， treatment， and multivariate analysis of risk factors for pyogenic liver abscess. Am J Surg 2001；181：177-186.

2 Lee KT， Wong SR， Sheen PC. Pyogenic liver abscess： an audit of 10 years' experience and analysis of risk factors. Dig Surg 2001；18：459-466.

3 Ruiz-Hernández JJ， León-Mazorra M， Conde-Martel A， Marchena-Gómez J， Hemmersbach-Miller M， Betancor-León P. Pyogenic liver abscesses： mortality-related factors. Eur J Gastroenterol Hepatol 2007；19：853-858.

4 Cerwenka H. Pyogenic liver abscess： differences in etiology and treatment in Southeast Asia and Central Europe. World J Gastroenterol 2010；16：2458-2462.

5 Pang TC， Fung T， Samra J， Hugh TJ， Smith RC. Pyogenic liver abscess： an audit of 10 years' experience. World J Gastroenterol 2011；17：1622-1630.

6 Chan DS， Archuleta S， Llorin RM， Lye DC， Fisher D. Standardized outpatient management of Klebsiella pneumoniae liver abscesses. Int J Infect Dis 2013；17：e185-188.

7 Lederman ER， Crum NF. Pyogenic liver abscess with a focus on Klebsiella pneumoniae as a primary pathogen： an emerging disease with unique clinical characteristics. Am J Gastroenterol 2005；100：322-331.

8 Liao CH， Huang YT， Chang CY， Hsu HS， Hsueh PR. Capsular serotypes and multilocus sequence types of bacteremic Klebsiella pneumoniae isolates associated with different types of infections. Eur J Clin Microbiol Infect Dis 2014；33：365-369.

9 Orosz E， Perkátai K， Kapusinszky B， Farkas A， Kucsera I. Realtime PCR assay for rapid qualitative and quantitative detection of Entamoeba histolytica. Acta Microbiol Immunol Hung 2012；59：451-460.

10 Berg GM， Vasquez DG， Hale LS， Nyberg SM， Moran DA. Evaluation of process variations in noncompliance in the implementation of evidence-based sepsis care. J Healthc Qual 2013；35：60-69.

11 Meddings L， Myers RP， Hubbard J， Shaheen AA， Laupland KB，Dixon E， et al. A population-based study of pyogenic liver abscesses in the United States： incidence， mortality， and temporal trends. Am J Gastroenterol 2010；105：117-124.

12 Zibari GB， Maguire S， Aultman DF， McMillan RW， McDonald JC. Pyogenic liver abscess. Surg Infect （Larchmt） 2000；1：15-21.

13 Mangukiya DO， Darshan JR， Kanani VK， Gupta ST. A prospective series case study of pyogenic liver abscess： recent trands in etiology and management. Indian J Surg 2012；74：385-390.

14 Khan R， Hamid S， Abid S， Jafri W， Abbas Z， Islam M， et al. Predictive factors for early aspiration in liver abscess. World J Gastroenterol 2008；14：2089-2093.

15 Cheng HP， Siu LK， Chang FY. Extended-spectrum cephalosporin compared to cefazolin for treatment of Klebsiella pneumoniae-caused liver abscess. Antimicrob Agents Chemother 2003；47：2088-2092.

16 Chemaly RF， Hall GS， Keys TF， Procop GW. Microbiology of liver abscesses and the predictive value of abscess gram stain and associated blood cultures. Diagn Microbiol Infect Dis 2003；46：245-248.

17 Lo JZ， Leow JJ， Ng PL， Lee HQ， Mohd Noor NA， Low JK， et al. Predictors of therapy failure in a series of 741 adult pyogenic liver abscesses. J Hepatobiliary Pancreat Sci 2015；22：156-165.

18 Pletz MW， Wellinghausen N， Welte T. Will polymerase chain reaction （PCR）-based diagnostics improve outcome in septic patients？ A clinical view. Intensive Care Med 2011；37：1069-1076.

19 Giorgio A， de Stefano G， Di Sarno A， Liorre G， Ferraioli G. Percutaneous needle aspiration of multiple pyogenic abscesses of the liver： 13-year single-center experience. AJR Am J Roentgenol 2006；187：1585-1590.

20 Zerem E， Hadzic A. Sonographically guided percutaneous catheter drainage versus needle aspiration in the managementof pyogenic liver abscess. AJR Am J Roentgenol 2007；189：138-142.

21 Liao WI， Tsai SH， Yu CY， Huang GS， Lin YY， Hsu CW， et al. Pyogenic liver abscess treated by percutaneous catheter drainage： MDCT measurement for treatment outcome. Eur J Radiol 2012；81：609-615.

22 Yu SC， Ho SS， Lau WY， Yeung DT， Yuen EH， Lee PS， et al. Treatment of pyogenic liver abscess： prospective randomized comparison of catheter drainage and needle aspiration. Hepatology 2004；39：932-938.

23 Sartelli M， Moore FA， Ansaloni L， Di Saverio S， Coccolini F，Griffiths EA， et al. A proposal for a CT driven classification of left colon acute diverticulitis. World J Emerg Surg 2015；10：3.

24 Shelat VG， Chia CL， Yeo CS， Qiao W， Woon W， Junnarkar SP. Pyogenic liver abscess： does Escherichia coli cause more adverse outcomes than Klebsiella pneumoniae？ World J Surg 2015；39：2535-2542.

25 Ahmed S， Chia CL， Junnarkar SP， Woon W， Shelat VG. Percutaneous drainage for giant pyogenic liver abscess--is it safe and sufficient？ Am J Surg 2016；211：95-101.

26 Yang CC， Yen CH， Ho MW， Wang JH. Comparison of pyogenic liver abscess caused by non-Klebsiella pneumoniae and Klebsiella pneumoniae. J Microbiol Immunol Infect 2004；37：176-184.

27 Alsaif HS， Venkatesh SK， Chan DS， Archuleta S. CT appearance of pyogenic liver abscesses caused by Klebsiella pneumoniae. Radiology 2011；260：129-138.

28 Alkofer B， Dufay C， Parienti JJ， Lepennec V， Dargere S， Chiche L. Are pyogenic liver abscesses still a surgical concern？ A Western experience. HPB Surg 2012；2012：316013.

29 Chen SC， Huang CC， Tsai SJ， Yen CH， Lin DB， Wang PH， et al. Severity of disease as main predictor for mortality in patients with pyogenic liver abscess. Am J Surg 2009；198：164-172.

30 Lok KH， Li KF， Li KK， Szeto ML. Pyogenic liver abscess：clinical profile， microbiological characteristics， and management in a Hong Kong hospital. J Microbiol Immunol Infect 2008；41：483-490.

31 Wu BU， Johannes RS， Sun X， Conwell DL， Banks PA. Early changes in blood urea nitrogen predict mortality in acute pancreatitis. Gastroenterology 2009；137：129-135.

32 Ransohoff DF， Miller GL， Forsythe SB， Hermann RE. Outcome of acute cholecystitis in patients with diabetes mellitus. Ann Intern Med 1987；106：829-832.

33 Menekse E， Kocer B， Topcu R， Olmez A， Tez M， Kayaalp C. A practical scoring system to predict mortality in patients with perforated peptic ulcer. World J Emerg Surg 2015；10：7.

34 Chu KM， Fan ST， Lai EC， Lo CM， Wong J. Pyogenic liver abscess. An audit of experience over the past decade. Arch Surg 1996；131：148-152.

35 Yeo CS， Tay VW， Low JK， Woon WW， Punamiya SJ， Shelat VG. Outcomes of percutaneous cholecystostomy and predictors of eventual cholecystectomy. J Hepatobiliary Pancreat Sci 2016；23：65-73.

Accepted after revision May 12， 2016

Strive not to be a success， but rather to be of value.

—Albert Einstein

December 2， 2105

Author Affiliations： Department of General Surgery， Tan Tock Seng Hospital， Singapore 308433， Singapore （Shelat VG， Low JK and Woon WWL）；Ministry of Health Holdings Pte Ltd.， Singapore 099253， Singapore（Wang Q， Chia CLK and Wang Z）

Vishal G Shelat， FRCS， FEBS， Consultant， Department of General Surgery， Tan Tock Seng Hospital， Singapore 308433，Singapore （Tel： +65-635-77807； Fax： +65-635-77809； Email： vgshelat@ gmail.com）

? 2016， Hepatobiliary Pancreat Dis Int. All rights reserved.

10.1016/S1499-3872（16）60127-3

Published online August 25， 2016.