Microbiological profile of pathogens in spontaneous bacterial peritonitis secondary to liver cirrhosis: a retrospective study

Background

Antibiotic prophylaxis is advised for patients with chronic liver disease, who have had a previous spontaneous bacterial peritonitis (SBP) and for those with low albumin levels in their ascitic fluid. More Gram-positive and multidrug-resistant bacteria are isolated in cases of SBP, possibly related to the use of prophylactic antibiotics and the rising incidence of nosocomial infections in hospitalised cases secondary to an increase in invasive procedures. As SBP worsens the prognosis in cirrhosis, it is necessary to treat it. However, the poor yield of ascitic fluid culture leads to empiric treatment, usually with third-generation cephalosporins, which is often inadequate.

Methods

Ours was a retrospective case record study over a three-year period from July 2014 to July 2017. After obtaining ethics approval from our institute’s ethical committee at a tertiary care centre in South India, we collected 721 consecutive cases recorded with a clinical diagnosis of spontaneous bacterial peritonitis. The diagnosis of SBP in a known cirrhotic patient was based on the presence of fever, abdominal pain and free fluid in the abdomen and an ascitic fluid polymorphonuclear cell count >250 cells/mm³. Data were collected from the microbiology reports of the ascitic fluid culture (whether negative or positive, and if so the organism isolated) and the bacterial susceptibility to chosen antibiotics. Isolates exhibiting resistance to at least one antibiotic in three or more antibiotic classes were labelled as multidrug-resistant (MDR). Ascitic fluid was cultured by inoculating at least 10 mL in a blood culture bottle. Percentages of Gram-positive and negative bacteria in ascitic fluid and their resistance patterns to seven classes of antibiotics—namely Cephalosporins, Quinolones, Penicillins, Carbapenems, Aminoglycosides, Tetracyclines and Macrolides—were documented. Descriptive statistics were calculated using the SPSS software version 20.0.

Results

Of the 721cases of SBP, 276 (38.2%) had positive ascitic fluid cultures. Gram-negative bacteria (GNB) were isolated from 205 (74.2%); 58 (21%) grew Gram-positive bacteria and 13 (4.7%) had fungal isolates. Common organisms found were Escherichia coli, Acinetobacter baumannii, Enterococcus spp. (including E. faecalis and E. faecium), Klebsiella pneumoniae, Pseudomonas, Enterobacter spp., Staphylococcus aureus, Streptococcus spp. and other GNB. Enterococcus spp. was the most common pathogen (11.9%) among Gram-positive bacteria, followed by Staphylococcus aureus (5.3%) and Streptococcus (3.6%). Candida spp. contributed to most of fungal isolates. Of all the antibiotics tested, the best susceptibility was seen towards Amikacin. Ceftriaxone and Ceftazidime had high resistance rates (Table 1). Of all Acinetobacter baumannii isolates, 71.4% were MDR, as were 30%–60% of K. pneumoniae, Enterobacter spp. and E. coli isolates. Enterococcus faecalis showed >50% sensitivity to Ampicillin, and high levels of Gentamicin and Tetracycline. On the other hand, E. faecium showed a resistance of >50% to these drugs. The antimicrobial activity of Penicillin and Ciprofloxacin against S. aureus was poor, with a resistance of >90%, but erythromycin showed lower rates of resistance. Of the 15 S. aureus isolates, 7 (46.6%) were methicillin-resistant (MRSA). Streptococcus spp. showed 20%–40% resistance to ampicillin. Vancomycin resistance was observed in only a few E. faecium isolates (Table 2). Of Candida spp., 16.6% were resistant to fluconazole, while all were sensitive to amphotericin-B.

OPEN IN VIEWER

Table 1.

Percentage susceptibility of Gram-negative bacteria and proportion of MDR (percentages) in SBP.

Study no.	Antibiotic organism	CTX	CIPRO	CTZ	CFS	MERO	AMI	GENTA	MDR
1	Escherichia coli (n = 78, 28.2%)	16.7	25	22.4	85.8	85.8	93.6	62.2	38.4
2	Acinetobacter baumannii (n = 42, 15.2%)	NT	38.1	23.8	85.8	27.3	47.6	33.3	71.4
3	Klebsiella pneumoniae (n = 31, 11.2%)	29	41.9	29	45	46.2	54.9	45.2	51.6
4	Pseudomonas spp. (n = 28, 10.1%)	NT	68.9	47.2	67.1	54.1	79.4	68.3	32.1
5	Enterobacter species (n = 20, 7.2%)	25	41.9	20	53.4	58.4	75	55	45
6	Citrobacter koseri (n = 4, 1.4%)	25	50	50	0	66.7	75	25	50
7	Flavobacterium species (n = 1, 0.3%)	NT	0	0	NT	0	100	100
8	Non-fermenting Gram-negative bacilli (NFGNB) (n = 1, 0.3%)	NT	0	0	NT	0	0	0

OPEN IN VIEWER

Table 2.

Percentage susceptibility(s) of Gram-positive bacteria and proportion of MDR (percentages) in SBP.

Study no.	Antibiotic organism	AMPI	HLG	TETRA	VANCO	OXA	PEN	CIPRO	ERYTHRO	MDR
1	E. faecalis (n = 19, 6.7%)	68.5	68.8	66.7	100	NT	NT	NT	NT	10.5
2	S. aureus (n = 15, 5.3%)	NT	NT	NT	100	60	6.7	7.7	40	46.6
3	E. faecium (n = 14, 5%)	7.2	21.5	28.6	91	NT	NT	NT	NT	64.2
4	Strep. spp. (n = 9, 3.2%)	80	75	55.6	100	NT	NT	NT	NT	0
5	Beta hemolytic Streptococci (n = 1, 0.3%)	NT	NT	0	NT	NT	100	NT	100	0

AMPI, Ampicillin; Beta-haemolytic, beta-haemolytic streptococcus species; CIPRO, Ciprofloxacin; E. faecalis, Enterococcus faecalis; E. faecium, Enterococcus faecium; ERYTHRO, Erythromycin; HLG, high-level gentamicin; MDR, multidrug resistance; NT, not tested; OXA, Oxacillin; PEN, Penicillin; S. aureus, Staphylococcus aureus; SBP, spontaneous bacterial peritonitis; Strep. spp., Streptococcus species; TETRA, Tetracycline; VANCO, Vancomycin.

Discussion

Spontaneous bacterial peritonitis is treated with antibiotics, once the ascitic fluid white cell count exceeds 250 cells/mm³. ¹ The outcome depends on the causative organism and its antibiotic susceptibility. Our frequency of Gram-negative organisms was similar to other studies done worldwide.^2,3 Recent reports from Asia and European nations also suggest that an increasing proportion of SBP is caused by gram-positive pathogens.^4–7

Poor antimicrobial coverage with cephalosporins has also been reported earlier, but in the range of 10%–45%.^8–11 As ours is a tertiary care referral centre, it is possible that patients may have been partially treated elsewhere previously. Nonetheless, vancomycin resistant Enterococci were fewer than in a German study, ¹² and we did not see Vancomycin resistance in Staphylococci as in a Chinese study. ¹³

The distribution of multidrug resistance differs in various countries and healthcare facilities, and depends on the antibiotics used and epidemiology of the causative pathogens in specific areas.

Thus, empirical treatment strategies need regular local re-evaluation.^14,15

One of the major limitations of this study was the lack of data on the proportion of patients who were on quinolone prophylaxis for SBP. Management strategies and outcome of such should be the focus of future research.