Clinical characteristics of bacteraemia caused by Lactobacillus spp. and antimicrobial susceptibilities of the isolates at a medical centre in Taiwan, 2000–2014

International Journal of Antimicrobial Agents 46 (2015) 439–445

Contents lists available at ScienceDirect

International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag

Clinical characteristics of bacteraemia caused by Lactobacillus spp. and antimicrobial susceptibilities of the isolates at a medical centre in Taiwan, 2000–2014 Meng-Rui Lee a,b,c , Chia-Jung Tsai d , Sheng-Kai Liang a,b , Ching-Kai Lin a,b , Yu-Tsung Huang e , Po-Ren Hsueh b,f,∗ a

Department of Internal Medicine, National Taiwan University Hospital Hsin-Chu Branch, Hsin-Chu, Taiwan Department of Internal Medicine, National Taiwan University Hospital, National Taiwan University College of Medicine, Taipei, Taiwan c Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan d Department of Internal Medicine, Taoyuan General Hospital, Ministry of Health and Welfare, Taoyuan City, Taiwan e Department of Internal Medicine, Far Eastern Memorial Hospital, New Taipei City, Taiwan f Department of Laboratory Medicine, National Taiwan University College of Medicine, National Taiwan University Hospital, Taipei, Taiwan b

a r t i c l e

i n f o

Article history: Received 20 May 2015 Accepted 25 June 2015 Keywords: Bacteraemia Lactobacillus Lactobacillus salivarius Matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry Mortality Probiotics

a b s t r a c t The clinical characteristics of 89 patients with Lactobacillus bacteraemia treated at a university-affiliated hospital in northern Taiwan during 2000–2014 were retrospectively evaluated. Lactobacillus spp. were identified by 16S rRNA sequencing analysis and matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF/MS). Antimicrobial susceptibilities of the isolates were determined by broth microdilution. The most commonly isolated species was Lactobacillus salivarius (n = 21), followed by Lactobacillus paracasei (n = 16) and Lactobacillus fermentum (n = 13). Excluding three isolates with lower 16S rRNA sequence similarity, MALDI-TOF/MS provided correct identification for 84.9% (73/86) of Lactobacillus isolates. Concordant identification was lowest for Lactobacillus casei (11%). The main infection foci were intra-abdominal infection (49%) and catheter-related bloodstream infection (17%). Only onehalf of the patients received adequate antibiotic treatment during the bacteraemic episode. The majority of patients with Lactobacillus bacteraemia were immunocompromised. The 7-day and in-hospital mortality rates were 21% and 62%, respectively, and underlying malignancy was associated with a higher in-hospital mortality rate (odds ratio = 2.666). There were no significant differences in mortality (7-day, 14-day, 30-day and in-hospital) among patients with bacteraemia due to different Lactobacillus spp. Minimum inhibitory concentrations were highest for glycopeptides, cephalosporins and fluoroquinolones and were lowest for carbapenems and aminopenicillins. Lactobacillus bacteraemia was associated with a high mortality rate, and patient outcome was associated with underlying malignancy. MALDI-TOF/MS was able to accurately identify 84.9% of the Lactobacillus isolates, and L. salivarius was the predominant pathogen. The accuracy rate for identification of Lactobacillus spp. by MALDI-TOF/MS was lowest for L. casei. © 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

1. Introduction The Lactobacillus genus is a group of Gram-positive, catalasenegative rods or coccobacilli that produce lactic acid as the end product of fermentation. The genus comprises more than 100

∗ Corresponding author at: Departments of Laboratory Medicine and Internal Medicine, National Taiwan University College of Medicine, National Taiwan University Hospital, Taipei, Taiwan. Tel.: +886 2 2312 3456x65355; fax: +886 2 2322 4263. E-mail address: [email protected] (P.-R. Hsueh).

species and all are widely distributed in the environment, animals and humans [1]. In humans, Lactobacillus spp. are common inhabitants of the oral cavity, the gastrointestinal tract and the female genital tract [2]. These organisms are also present in various food products, such as dairy products, fermented meat, vegetables, fruits and beverages, and in the food industry some Lactobacillus spp. are used for probiotic applications [1,3]. Lactobacillus spp. are traditionally considered to be of low virulence, however they can cause infections ranging from localised infection to bacteraemia. Lactobacillus bacteraemia usually occurs in immunocompromised hosts with significant co-morbidities

http://dx.doi.org/10.1016/j.ijantimicag.2015.06.017 0924-8579/© 2015 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.

440

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

and has been linked with endocarditis in previous reports [4,5]. Although the clinical relevance of Lactobacillus spp. has been questioned, studies have shown that inadequate antibiotic therapy was associated with adverse outcomes [6]. In Finland and France, Lactobacillus rhamnosus was the most common species involved in Lactobacillus bacteraemia [7]. The exact distribution of Lactobacillus spp. among bacteraemia isolates remains to be investigated and the results could have a clinical impact on the utilisation of Lactobacillus in the food industry. Members of the Lactobacillus genus are also considered to be underestimated pathogens since conventional methods have only been able to identify a limited number of Lactobacillus spp. [8]. Correct identification of Lactobacillus spp. relies on 16S rRNA sequencing methods. Although matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF/MS) is a routine laboratory method for identifying bacterial species from a number of genera, its utilisation for identifying Lactobacillus spp. remains largely limited to the field of food microbiology [9]. Very few studies have been published on the use of MALDI-TOF/MS analysis for the identification of pathogenic lactobacilli in the clinical laboratory [10]. In this study, the incidence of Lactobacillus bacteraemia was investigated at a university-affiliated hospital in Taiwan during the period 2000–2014. In addition, the accuracy of MALDI-TOF/MS was compared with that of 16S rRNA gene sequencing in identifying isolates of Lactobacillus to species level as well as the susceptibility of the isolates to various antimicrobial agents was analysed. 2. Patients and methods 2.1. Bacterial isolates, patients and setting This study was conducted at National Taiwan University Hospital, a 2900-bed tertiary care centre in northern Taiwan. Information on patients with bacteraemia caused by Lactobacillus spp. was obtained from the Clinical Microbiology Laboratory of the hospital. The incidence of bacteraemia caused by Lactobacillus spp. per year was also analysed. Conventional biochemical methods were used for identification of Lactobacillus spp. at the Clinical Microbiology Laboratory of the hospital. Of the patients with Lactobacillus bacteraemia in this study, five have been reported previously [11]. One isolate initially identified as Leuconostoc spp. by conventional methods was subsequently confirmed to be Lactobacillus spp. by 16S rRNA sequencing and was added into this study [12]. 2.2. Clinical characteristics of the patients A standardised case record form was used to collect demographic and clinical data, including patient age, sex, underlying status, the presence or absence of central vascular access, surgical procedure received and empirical antibiotic usage. The primary outcome measure was in-hospital mortality. Crude mortality rates at Days 7, 14 and 30 were also recorded. 2.3. 16S rRNA gene sequencing analysis Partial sequencing of the 16S rRNA gene was performed on all blood isolates obtained during the study period. A 1475-bp fragment was amplified and was then sequenced using the following primers: 8FLP, 5 -AGAGTTTGATCCTGGCTCAG-3 ; and 1492RPL, 5 -GGTTACCTTGTTACGACTT-3 . The results were compared with published sequences in the GenBank database using the BLASTN algorithm (http://www.ncbi.nlm.nih.gov/blast) [13]. The closest matches and GenBank accession no. were obtained. Lactobacillus lactis ATCC 19256 was used as the control strain in each test.

2.4. Identification of Lactobacillus spp. isolates by MALDI-TOF/MS Samples of isolates to be analysed by the Bruker MALDI-TOF Biotyper system were prepared as previously described [14]. Briefly, all isolates were incubated on trypticase soy agar with 5% sheep blood (Becton Dickinson Microbiology Systems, Sparks, MD) and were incubated for 48 h at 37 ◦ C. Two to three colonies were transferred to a 1.5-mL screw-cap Eppendorf tube containing 300 ␮L of distilled water and were then mixed with 900 ␮L of ethanol by pipetting. The suspension was pelleted by centrifugation at 13 000 rpm for 2 min, was evaporated to dryness and then reconstituted in 50 ␮L of 70% formic acid. Following incubation for 30 s, 50 ␮L of acetonitrile (Sigma–Aldrich) was added. The suspension was then centrifuged at 13 000 rpm for 2 min. Next, 1.0 ␮L of the supernatant was applied to a 96-spot polished steel target plate (Bruker Daltonik GmbH, Bremen, Germany) and was dried. A saturated solution of 1.0 ␮L of MALDI matrix (␣-cyano-4-hydroxycinnamic acid; Bruker Daltonik GmbH) was applied to each sample and dried. Measurements were performed with a Bruker microflexTM LT MALDI-TOF mass spectrometer (Bruker Daltonik GmbH) using flexControl software with Compass Flex Series v.1.3 software and a 60-Hz nitrogen laser (337 nm wavelength). Identification scores of ≥2.000 indicated species-level identification, scores ranging from 1.700 to 1.999 indicated genus-level identification, and scores of <1.700 indicated no reliable identification. The spectra covering the massto-charge ratio (m/z) ranging from 1960 to 20 132 were collected in the linear positive mode and were analysed using Bruker Biotyper automation control and the Bruker Biotyper 3.1 software and library (database 5627 with 5627 entries). All isolates with discrepant identification results between the molecular and Bruker Biotyper methods were re-tested twice as described previously [15,16]. 2.5. Antimicrobial susceptibility testing Minimum inhibitory concentrations (MICs) of nine antimicrobial agents against 86 preserved isolates were determined using the broth microdilution method according to Clinical and Laboratory Standards Institute (CLSI) guidelines [17]. Cation-adjusted Mueller–Hinton broth supplemented with 5% lysed horse blood (Becton Dickinson Microbiology Systems) was used [17]. The antimicrobial agents included amoxicillin/clavulanic acid (AMC), cefuroxime, ceftazidime, cefepime, piperacillin/tazobactam (TZP), imipenem, levofloxacin, vancomycin and teicoplanin. Standard powders of antimicrobial agents were obtained from various manufacturers. The MICs were read following incubation at 35 ◦ C in a 5% CO2 atmosphere for 24 h and were defined as the lowest concentration that inhibited visible growth of the organism. Streptococcus pneumoniae ATCC 49619 and Escherichia coli ATCC 25922 were used as control strains [17]. MIC interpretive breakpoints for imipenem (susceptible, ≤0.5 mg/L; intermediate, 1 mg/L; resistant, ≥2 mg/L) and vancomycin (susceptible ≤2 mg/L; intermediate, 4–8 mg/L; resistant, ≥16 mg/L) against Lactobacillus spp. provided by the CLSI were used. There are no MIC interpretive breakpoints for the remaining agents [17]. 2.6. Definition Bacteraemia was defined as the isolation of an organism in one or more separately obtained blood cultures with compatible clinical features. Lactobacillus bacteraemia was classified as healthcareassociated infection if patients acquired the infection during the course of receiving treatment for other conditions within a healthcare setting. It was otherwise classified as community-associated [12]. Antibiotic treatment was considered effective if the patients

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

441

Table 1 Identification of 86 Lactobacillus isolates to species level by 16S rRNA sequencing analysis and MALDI-TOF/MS.a Species identification by 16S rRNA sequencing analysis

Identification results by MALDI-TOF/MS compared with those by 16S rRNA sequencing analysis

Lactobacillus spp.

No. of isolates

Accession no. (no. of isolates)

No. (%) of isolates with concordant identification

Score values (range) (no. of isolates)

Lactobacillus spp. with discordant identification (no. of isolates)

Score values (range) (no. of isolates)

L. salivarius

19

17 (89.5)

2.017–2.184 (9) 1.77–1.989 (8)

No reliable identification (2)

–

L. paracasei

16

16 (100)

2.266–2.416 (16)

No

–

L. fermentum

13

12 (92.3)

2.006–2.16 (5) 1.821–1.992 (7)

No reliable identification (1)

–

L. rhamnosus

10

AB932531.1 (n = 10) AB932533.1 (n = 5) AB932525.1 (n = 2) AB911537.1 (n = 1) KC817298.1 (n = 1) KJ764645.1 (n = 7) KJ764646.1 (n = 6) KJ702578.1 (n = 2) HM462419.1 (n = 1) KJ690756.1 (n = 8) AJ575812.1 (n = 1) JF414736.1 (n = 1) KF245553.1 (n = 1) KF740705.1 (n = 1) KJ026616.1 (n = 1) KF971890.1 (n = 4) KJ939337.1 (n = 4) AB932536.1 (n = 1) HM462427.1 (n = 1) KJ806307.1 (n = 4) JQ580981.1 (n = 2) KJ806302.1 (n = 2) KF971891.1 (n = 1) AB932530.1 (6) KJ725212.1 (n = 2) KJ918744.1 (n = 1) KJ958428.1 (n = 1) JQ073738.1 (n = 1) JF745115.1 (n = 1) HM596285.1 NR 117574.1 NR 112754.1 AB911532.1 AB911500.1 AB186315.1 AB889731.1 –

10 (100)

2.014–2.171 (5) 1.926–1.991 (5)

No

–

1 (11.1)

1.839 (1)

L. paracasei (8)

2.120–2.381 (8)

5 (83.3) 4 (100)

2.015–2.37 (5) 2.311–2.399 (4)

L. ingluviei (1) No

1.802 (1) –

2 (100)

1.812–2.313 (2)

No

–

1.962 (1) 2.211 (1) 2.105 (1) – 1.873 (1) 2.088 (1) 2.023 (1) –

No No No L. johnsonii (1) No No No 13 (15.1)

– –

L. casei

9

L. gasseri L. plantarum

6 4

L. reuteri

2

L. oris L. johnsonii L. nagelli L. taiwanensis L. ingluviei L. mucosae L. farciminis Total

1 1 1 1 1 1 1 86

1 (100) 1 (100) 1 (100) 0 1 (100) 1 (100) 1 (100) 73 (84.9)

2.348 (1) – – – –

MALDI-TOF/MS, matrix-associated laser desorption/ionisation time-of-flight mass spectrometry. a Three isolates with lower 16S rRNA sequence similarity were excluded from MALDI-TOF/MS analysis. These three isolates were L. salivarius (accession no. AB932533.1; similarity 96.0%), L. salivarius (accession no. AB932533.1; similarity 97.0%) and L. reuteri (accession no. JQ073738.1; similarity 94.8%) by 16S rRNA sequencing analysis and were identified by MALDI-TOF/MS as L. salivarius (score value of 2.037), L. salivarius (1.825) and L. reuteri (2.163), respectively.

received antibiotics with in vitro activity (susceptible) based on the results of the disk diffusion method or the results of MIC testing. Lactobacillus bacteraemia was considered to be primary bacteraemia if the bacteraemia was not secondary to an infection at another body site. (http://www.cdc.gov/nhsn/PDFs/pscManual/ 4PSC CLABScurrent.pdf). The focus of Lactobacillus bacteraemia was considered to be intra-abdominal if there was existing unresolved intra-abdominal infection at the time of Lactobacillus bacteraemia. Lactobacillus bacteraemia was defined as catheter-related if Lactobacillus bacteraemia developed in a patient who had a central line within the 48-h period before the development of bacteraemia and the bacteraemia was not related to an infection at another site [18]. 2.7. Statistical analysis The statistical package SAS for Windows v.9.3 (SAS Institute Inc., Cary, NC) was used for data retrieval and data analysis. The 2 test was applied for categorical data comparisons, and oneway analysis of variance (ANOVA) was applied for continuous variable comparisons. Backward step-wise and manual multivariate logistic regression modelling was performed to determine the

independent predictors of in-hospital mortality. All P-values reported are two-tailed, and the significance level was set at <0.05. 3. Results 3.1. Species identification by 16S rRNA sequencing analysis with regard to the year of isolation During the 15-year period, a total of 89 bacteraemic isolates were identified as Lactobacillus spp. using 16S rRNA sequencing analysis (Table 1). Of the 89 isolates, 5 were retrieved from among the preserved isolates used in our previous study [11]. One bacteraemic isolate was originally identified as Leuconostoc spp. by the conventional method [12]. The annual incidence (per 100 000 admissions) of Lactobacillus bacteraemia was ca. 8.2 during the years 2000–2014 (range, 2.5 in 2011 to 15.9 in 2005) (Fig. 1A). All isolates of Lactobacillus spp. exhibited 99.8–100% sequence similarity by 16S rRNA sequencing analysis, with the exception of two isolates of Lactobacillus salivarius (96.0% and 97.0% similarity, respectively) and one isolate of Lactobacillus reuteri (94.8% similarity). These three isolates were excluded from comparison with identification results by the MALDI-TOF/MS owing to low sequence

442

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

Fig. 1. (A) Annual incidence (per 100 000 admissions) of Lactobacillus bacteraemia from 2000 to 2014. (B) Number of patients with bacteraemia caused by four Lactobacillus spp. in three 5-year time periods from 2000 to 2014. The numbers in parentheses denote the number of patients with bacteraemia caused by all Lactobacillus spp. in each indicated time period.

similarity (<98%) by 16S rRNA sequencing analysis. Lactobacillus salivarius accounted for almost one-quarter of all bacteraemic isolates (21; 23.6%). Other commonly isolated species included Lactobacillus paracasei (16; 18.0%), L. fermentum (13; 14.6%), L. rhamnosus (10; 11.2%) and Lactobacillus casei (9; 10.1%). Lactobacillus gasseri (6; 6.7%), L. plantarum (4; 4.5%) and L. reuteri (3; 3.4%) were less commonly isolated. The distribution of major Lactobacillus spp. causing bacteraemia with regard to the three 5-year time periods in which the isolates were recovered is shown in Fig. 1B. The trend in incidence of major bacteraemic Lactobacillus spp. showed no obvious fluctuations.

3.2. Species identification by MALDI-TOF/MS MALDI-TOF/MS accurately identified 84.9% (73/86) of the Lactobacillus bacteraemia isolates to species level. MALDI-TOF/MS correctly identified all isolates of L. paracasei and L. rhamnosus and correctly identified most isolates of L. salivarius [17/19 (89.5%), with 2 unidentified species] and L. fermentum [12/13 (92.3%), with 1 unidentified species]. MALDI-TOF/MS misidentified most of the L. casei isolates as L. paracasei. Detailed identification results obtained by 16S rRNA and MALDI-TOF/MS are summarised in Table 1.

3.3. Characteristics of patients Of the 89 patients with bacteraemia caused by Lactobacillus spp., most were of advanced age (mean age, 64 years) and had underlying malignancy (n = 57; 64%) (Table 2). Approximately 40% of patients had polymicrobial bacteraemia (n = 35), with Candida spp. (n = 16) and Bacteroides spp. (n = 7) being the most common concomitant pathogens. Most bacteraemia episodes were healthcare-associated (n = 76; 85%). Intra-abdominal infection, primary bacteraemia and catheter-related infections were the most common infectious foci. TZP (n = 23; 26%) and carbapenems (n = 19; 21%) were the two most commonly administered antimicrobial agents. The in-hospital mortality rate was 62% (n = 55). Results of the multivariate logistic regression analysis revealed that malignancy was independently associated with in-hospital mortality (odds ratio = 2.666, 95% confidence interval 1.088–6.537; P = 0.0321).

3.4. Antimicrobial susceptibilities Table 3 shows the susceptibility results of the 86 preserved isolates to nine antimicrobial agents. The majority of the

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

443

Table 2 Clinical characteristics of 89 patients with bacteraemia caused by different Lactobacillus species. Characteristic

All (n = 89)

L. salivarius (n = 21)

L. paracasei (n = 16)

L. fermentum (n = 13)

L. rhamnosus (n = 10)

L. casei (n = 9)

Others (n = 20)

P-value

Male/female (n) Age (years) (mean ± S.D.) Type of bacteraemia [n (%)] Polymicrobial Healthcare-associated Community-acquired IAI-related Catheter-related Primary Underling conditions [n (%)] Malignancy ESRD Diabetes mellitus Liver cirrhosis Recent chemotherapy Abdominal surgery Laboratory findings (mean ± S.D.) WBC count (cells/mm3 ) Serum albumin (mg/L) ALT (IU/L) Serum creatinine (mg/L) Antibiotics used [n (%)] Penicillin AMC TZP Ceftazidime Cefepime Quinolones Carbapenems Vancomycin Teicoplanin Metronidazole Clindamycin Effective antibiotic use [n (%)] Mortality [n (%)] 7-day 14-day 30-day In-hospital

46/43 63.89 ± 15.47

11/10 64.67 ± 14.45

9/7 64.88 ± 15.93

6/7 59 ± 22.62

4/6 62.2 ± 15.25

6/3 63.78 ± 10.20

10/10 66.35 ± 13.60

0.8920 0.8513

35 (39) 76 (85) 13 (15) 44 (49) 15 (17) 30 (34)

11 (52) 15 (71) 6 (29) 11 (52) 1 (5) 9 (43)

10 (63) 14 (88) 2 (13) 7 (44) 4 (25) 5 (31)

6 (46) 9 (69) 4 (31) 6 (46) 2 (15) 5 (38)

2 (20) 10 (100) 0 3 (30) 4 (40) 3 (30)

3 (33) 9 (100) 0 5 (56) 1 (11) 3 (33)

3 (15) 19 (95) 1 (5) 12 (60) 3 (15) 5 (25)

0.0346 0.0557 0.0557 0.7193 0.2167 0.8886

57 (64) 6 (7) 15 (17) 17 (19) 30 (34) 23 (26)

13 (62) 2 (10) 2 (10) 3 (14) 7 (33) 4 (19)

9 (56) 0 1 (6) 3 (19) 5 (31) 5 (31)

8 (62) 1 (8) 1 (8) 5 (38) 2 (15) 1 (8)

7 (70) 1 (10) 1 (10) 2 (20) 5 (50) 2 (20)

6 (67) 1 (11) 4 (44) 1 (11) 3 (33) 7 (78)

14 (70) 1 (5) 6 (30) 3 (15) 8 (40) 4 (20)

0.9635 0.8529 0.0609 0.5402 0.6173 0.0063

12.34 ± 8.1 2.85 ± 0.68 75.36 ± 128.47 1.52 ± 1.51

12.4 ± 7.65 2.76 ± 0.58 49 ± 28.39 1.79 ± 2.11

14.70 ± 11.10 3.19 ± 0.76 120.93 ± 216.59 1.35 ± 0.91

9.71 ± 3.86 2.53 ± 0.59 63.07 ± 41.14 1.59 ± 1.15

14.41 ± 10.2 3.01 ± 0.58 153.2 ± 224.78 1.49 ± 1.34

15.00 ± 9.98 2.98 ± 0.76 33 ± 22.09 1.24 ± 1.38

9.73 ± 4.46 2.66 ± 0.66 49.47 ± 54.21 1.45 ± 1.65

0.2793 0.1479 0.1363 0.9483

5 (6) 6 (7) 23 (26) 4 (4) 11 (12) 2 (2) 19 (21) 11 (12) 2 (2) 7 (8) 1 (1) 48 (54)

2 (10) 0 3 (14) 2 (10) 4 (19) 0 3 (14) 4 (19) 1 (5) 3 (14) 1 (5) 13 (62)

0 0 5 (31) 0 3 (19) 1 (6) 5 (31) 2 (13) 1 (6) 1 (6) 0 7 (44)

1 (8) 2 (15) 3 (23) 0 1 (8) 0 1 (8) 0 0 1 (8) 0 7 (54)

2 (20) 1 (10) 0 1 (10) 1 (10) 0 2 (20) 2 (20) 0 1 (10) 0 3 (30)

0 1 (11) 2 (22) 0 0 0 5 (56) 1 (11) 0 0 0 5 (56)

0 2 (10) 10 (50) 1 (5) 2 (10) 1 (5) 3 (15) 2 (10) 0 1 (5) 0 13 (65)

0.2000 0.4160 0.0306 0.5869 0.6670 0.6884 0.0822 0.0822 0.7136 0.8079 0.6650 0.4791

19 (21) 26 (29) 37 (42) 55 (62)

5 (24) 6 (29) 8 (38) 12 (57)

4 (25) 6 (38) 7 (44) 11 (69)

2 (15) 3 (23) 5 (38) 9 (69)

2 (20) 3 (30) 5 (50) 5 (50)

2 (22) 4 (44) 5 (56) 8 (89)

4 (20) 4 (20) 7 (35) 10 (50)

0.9916 0.7605 0.9060 0.3700

S.D., standard deviation; IAI, intra-abdominal infection; ESRD, end-stage renal disease; WBC, white blood cell; ALT, alanine aminotransferase; AMC, amoxicillin/clavulanic acid; TZP, piperacillin/tazobactam.

commonly isolated Lactobacillus spp. were resistant to vancomycin (89.5%). Lactobacillus gasseri (n = 6), Lactobacillus johnsonii (n = 1) and Lactobacillus taiwanensis (n = 1) had low vancomycin MICs. The overall resistance rate was moderate for imipenem (26.7%). For all isolates, the MIC90 (MIC that inhibits 90% of the isolates) was lower for the aminopenicillins (AMC, MIC90 = 4 mg/L; and TZP, MIC90 = 16 mg/L) and was higher for the cephalosporins (cefuroxime, MIC90 > 64 mg/L; ceftazidime MIC90 > 64 mg/L; and cefepime MIC90 > 64 mg/L). Levofloxacin exhibited an MIC50 (MIC that inhibits 50% of the isolates) of 4 mg/L and a MIC90 of 32 mg/L. 4. Discussion In this study, L. salivarius was the most common species causing Lactobacillus bacteraemia. Lactobacillus bacteraemia tended to occur in patients with immunocompromised status. Mortality was high and was related to underlying co-morbidities. The most common associated infections (portals of entry) were intra-abdominal and catheter-related infections. MALDI-TOF/MS correctly identified the majority of Lactobacillus isolates to species level but misidentified most of the L. casei isolates. MICs were highest for glycopeptides, cephalosporins and fluoroquinolones and were lowest for carbapenems and aminopenicillins. Lactobacillus salivarius is a rare cause of bacteraemia. In one case report, L. salivarius was determined to be the cause of acute bacteraemic cholecystitis [19]. In a study involving 85 blood isolates

with detailed species identification, approximately one-half were L. rhamnosus (22 L. rhamnosus GG type and 24 L. rhamnosus) and only 3 were L. salivarius [8]. In another study, 16 of 28 cases of bacteraemia were found to be secondary to L. rhamnosus and none were due to L. salivarius [7]. None the less, L. salivarius is widely used as a probiotic because of its ability to inhibit other pathogens and because of its immunomodulatory properties [3,20–23]. Lactobacillus salivarius has also been isolated from swine colon and infant faeces in Taiwan, indicating that it is a commensal organism in human and animal populations in that region [24,25]. The current study differed from other reports conducted in Europe that reported L. rhamnosus to be the most commonly isolated species for Lactobacillus bacteraemia [4,7,8]. Given the fact that the majority of L. salivarius bacteraemia was considered to be related to intraabdominal infections or primary bacteraemia, the predominance of L. salivarius in this study might reflect the geographic difference in gastrointestinal commensals. Further studies are warranted to find more evidence to support this hypothesis. Although rare, the high mortality rate associated with L. salivarius bacteraemia is reason enough to ensure that Lactobacillus isolates are accurately identified to species level so that appropriate antibiotics can be administered early. MALDI-TOF/MS has been shown to be a very accurate tool for differentiating and characterising Lactobacillus spp. in the human oral cavity, digestive tract and genital tract [10,26]; however, few studies have investigated the utilisation of MALDI-TOF/MS for the

444

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

Table 3 Antimicrobial susceptibilities of 86 preserved blood isolates of Lactobacillus.a Agent and MICs

MICs (mg/L) [and susceptibility results for indicated Lactobacillus spp.] L. salivarius (n = 20)

Amoxicillin/clavulanic acid 0.25 to >32 MIC range 0.5 MIC50 1 MIC90 Cefuroxime 0.5 to >64 MIC range 1 MIC50 2 MIC90 Piperacillin/tazobactam 0.5 to >64 MIC range MIC50 2 MIC90 2 Ceftazidime 8 to >64 MIC range 16 MIC50 16 MIC90 Cefepime 0.5 to >64 MIC range 1 MIC50 MIC90 4 Imipenem 0.03–1 MIC range 0.03 MIC50 0.25 MIC90 19 (95.0) Susceptible [n (%)] 0 Resistant [n (%)] Levofloxacin 0.25 to >32 MIC range MIC50 2 4 MIC90 Vancomycin 4 to >128 MIC range >128 MIC50 MIC90 >128 0 Susceptible [n (%)] 19 (95.0) Resistant [n (%)] Teicoplanin 32 to >128 MIC range MIC50 >128 >128 MIC90

L. paracasei (n = 16)

L. fermentum (n = 13)

L. rhamnosus (n = 10)

L. casei (n = 8)

L. gasseri (n = 6)

Total (n = 86)

1–8 2 4

0.12–1 0.25 0.5

2–8 2 4

2–4 2 4

0.25–0.5 0.5 0.5

0.12 to >32 1 4

8 to >64 16 64

8 to >64 32 64

16 to >64 32 >64

8 to >64 16 32

1–4 1 2

0.5 to >64 16 >64

2–32 4 16

0.12–2 1 2

2 to >64 4 32

2–16 4 4

0.5–1 0.5 1

0.12 to >64 2 16

32 to >64 >64 >64

32 to >64 >64 >64

64 to >64 >64 >64

64 to >64 >64 >64

8–16 8 16

>64 >64 >64

64 to >64 >64 >64

8 to >64 >64 >64

64 to >64 >64 >64

64 to >64 >64 >64

1–4 1 2

0.12 to >64 64 >64

1–4 2 2 0 10 (62.5)

0.03 0.03 0.03 13 (100) 0

1–8 2 2 0 8 (80)

1–2 2 2 0 5 (62.5)

0.03–0.12 0.06 0.06 6 (100) 0

0.03–8 0.06 2 51 (59.3) 23 (26.7)

2–8 4 8

8 to >32 32 >32

2–16 2 4

1–32 4 8

8 to >32 16 >32

0.25 to >32 4 32

>128 >128 >128 0 16 (100)

>128 >128 >128 0 13 (100)

>128 >128 >128 0 10 (100)

>128 >128 >128 0 8 (100)

0.5–1 1 1 6 (100) 0

0.5 to >128 >128 >128 8 (9.3) 77 (89.5)

>128 >128 >128

8 to >128 >128 >128

>128 >128 >128

>128 >128 >128

0.25–0.5 0.25 0.5

0.12 to >128 >128 >128

MIC, minimum inhibitory concentration; MIC50/90 , MIC that inhibits 50% and 90% of the isolates, respectively. a Lactobacillus spp. with at least six isolates were calculated.

identification and differentiation of clinically relevant lactobacilli [7]. In the current study, MALDI-TOF/MS performed well for the identification of Lactobacillus to species level, with an overall accuracy rate of 84.9%. However, it failed to correctly identify most L. casei species. A previous study found that the results of 16S rRNA sequencing were discordant with those of MALDI-TOF/MS [7]. In another study, MALDI-TOF/MS correctly identified three of four L. casei isolates [26]. More data are needed to further characterise the performance of MALDI-TOF/MS for the identification of Lactobacillus spp. The pathogenic role of Lactobacillus spp. is controversial [27]. Lactobacillus bacteraemia has a grave outcome and tends to occur in severely immunocompromised hosts, making the causal inference between grave prognosis and Lactobacillus bacteraemia questionable. In a previous study, it was found that treatment with antimicrobials that are effective in vitro was associated with lower mortality [27]. In the current study, however, we found that underlying malignancy was associated with higher in-hospital mortality. Salminen et al. also found that severe underlying diseases were a significant predictor of mortality [27]. The portal of entry of Lactobacillus is another issue of interest. Lactobacillus has been reported to be associated with endocarditis [8,27], and L. rhamnosus bacteraemia has been reported to be associated with catheter usage [7]. In the current study, we found that intraabdominal infection was the most common infection foci. This

could be explained by the fact that Lactobacillus spp. are commensals in the human gastrointestinal tract and in the case of immunocompromised status and mucosal breakdown secondary to chemotherapy, Lactobacillus spp. can enter the bloodstream. In a recent report, a patient with ulcerative colitis developed L. rhamnosus GG bacteraemia while taking probiotics containing the same strain at the same time [28]. It should therefore be highlighted that Lactobacillus bacteraemia can be associated with probiotic use in immunocompromised hosts with underlying gastrointestinal defects [28]. The recommended treatment regimen for Lactobacillus infections includes high-dose penicillin and an aminoglycoside [8,27]. However, compliance with recommended therapy is low. This may be because broad-spectrum antibiotics are often needed for patients with prolonged hospital stay, immunocompromised status and polymicrobial bacteraemia. Another possible explanation is that physicians are generally unaware of the pathogenicity of Lactobacillus spp. and tend to consider them mere contaminants [6]. Currently, however, there is a lack of recommended interpretive breakpoints of susceptibility for commonly used broad-spectrum antibiotics for Lactobacillus spp. [17]. Antibiotics with interpretive breakpoints for Lactobacillus spp. include penicillin, ampicillin, imipenem, gentamicin, vancomycin, daptomycin, erythromycin, clindamycin and linezolid [17]. If a broader-spectrum antibiotic is needed, then an antipseudomonal aminopenicillin or carbapenems

M.-R. Lee et al. / International Journal of Antimicrobial Agents 46 (2015) 439–445

may be preferred because of their lower MICs. Glycopeptides would not be advisable because these agents are not effective against the majority of Lactobacillus spp. This study has a few limitations. First, we were unable to obtain a history of probiotic usage in our patients because of the retrospective nature of the study. Second, we were unable to examine the incidence of infective endocarditis due to Lactobacillus spp. because echocardiographic studies at the time of bacteraemia were not performed in the majority of patients. However, there was no clinical evidence of infective endocarditis in any of the patients, making the pre-test probability relatively low. To the best of our knowledge, this is the largest study to provide detailed species identification and antimicrobial susceptibilities of Lactobacillus spp. causing bacteraemia. Further studies are needed to monitor the trend in incidence of infections due to Lactobacillus spp. and the distribution of clinically relevant Lactobacillus spp. causing bacteraemia. Funding: None. Competing interests: None declared. Ethical approval: Not required. References [1] Felis GE, Dellaglio F. Taxonomy of lactobacilli and bifidobacteria. Curr Issues Intest Microbiol 2007;8:44–61. [2] Pavlova SI, Kilic AO, Kilic SS, So JS, Nader-Macias ME, Simoes JA, et al. Genetic diversity of vaginal lactobacilli from women in different countries based on 16S rRNA gene sequences. J Appl Microbiol 2002;92:451–9. [3] Neville BA, O’Toole PW. Probiotic properties of Lactobacillus salivarius and closely related Lactobacillus species. Future Microbiol 2010;5:759–74. [4] Salminen MK, Tynkkynen S, Rautelin H, Saxelin M, Vaara M, Ruutu P, et al. Lactobacillus bacteremia during a rapid increase in probiotic use of Lactobacillus rhamnosus GG in Finland. Clin Infect Dis 2002;35:1155–60. [5] Husni RN, Gordon SM, Washington JA, Longworth DL. Lactobacillus bacteremia and endocarditis: review of 45 cases. Clin Infect Dis 1997;25:1048–55. [6] Cannon JP, Lee TA, Bolanos JT, Danziger LH. Pathogenic relevance of Lactobacillus: a retrospective review of over 200 cases. Eur J Clin Microbiol Infect Dis 2005;24:31–40. [7] Gouriet F, Million M, Henri M, Fournier PE, Raoult D. Lactobacillus rhamnosus bacteremia: an emerging clinical entity. Eur J Clin Microbiol Infect Dis 2012;31:2469–80. [8] Salminen MK, Rautelin H, Tynkkynen S, Poussa T, Saxelin M, Valtonen V, et al. Lactobacillus bacteremia, species identification, and antimicrobial susceptibility of 85 blood isolates. Clin Infect Dis 2006;42:e35–44. [9] Duskova M, Sedo O, Ksicova K, Zdrahal Z, Karpiskova R. Identification of lactobacilli isolated from food by genotypic methods and MALDI-TOF MS. Int J Food Microbiol 2012;159:107–14. [10] Anderson AC, Sanunu M, Schneider C, Clad A, Karygianni L, Hellwig E, et al. Rapid species-level identification of vaginal and oral lactobacilli using MALDI-TOF MS analysis and 16S rDNA sequencing. BMC Microbiol 2014;14:312. [11] Lee MR, Huang YT, Liao CH, Lai CC, Lee PI, Hsueh PR. Bacteraemia caused by Weissella confusa at a university hospital in Taiwan, 1997–2007. Clin Microbiol Infect 2011;17:1226–31.

445

[12] Lee MR, Huang YT, Lee PI, Liao CH, Lai CC, Lee LN, et al. Healthcare-associated bacteraemia caused by Leuconostoc species at a university hospital in Taiwan between 1995 and 2008. J Hosp Infect 2011;78:45–9. [13] Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25:3389–402. [14] Verroken A, Janssens M, Berhin C, Bogaerts P, Huang TD, Wauters G, et al. Evaluation of matrix-assisted laser desorption ionization-time of flight mass spectrometry for identification of Nocardia species. J Clin Microbiol 2010;48:4015–21. [15] Hsueh PR, Lee TF, Du SH, Teng SH, Liao CH, Sheng WH, et al. Bruker biotyper matrix-assisted laser desorption ionization-time of flight mass spectrometry system for identification of Nocardia, Rhodococcus, Kocuria, Gordonia, Tsukamurella, and Listeria species. J Clin Microbiol 2014;52:2371–9. [16] Hsueh PR, Kuo LC, Chang TC, Lee TF, Teng SH, Chuang YC, et al. Evaluation of the Bruker Biotyper matrix-assisted laser desorption ionization-time of flight mass spectrometry system for identification of blood isolates of Acinetobacter species. J Clin Microbiol 2014;52:3095–100. [17] Clinical and Laboratory Standards Institute. Methods for antimicrobial dilution and disk susceptibility testing of infrequently isolated or fastidious bacteria; approved guideline. Document M45-A2. 2nd ed. Wayne, PA: CLSI; 2010. [18] O’Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control 2011;39(Suppl. 1):S1–34. [19] Woo PC, Fung AM, Lau SK, Yuen KY. Identification by 16S rRNA gene sequencing of Lactobacillus salivarius bacteremic cholecystitis. J Clin Microbiol 2002;40:265–7. [20] Moles L, Escribano E, de Andrés J, Montes MT, Rodríguez JM, Jiménez E, et al. Administration of Bifidobacterium breve PS12929 and Lactobacillus salivarius PS12934, two strains isolated from human milk, to very low and extremely low birth weight preterm infants: a pilot study. J Immunol Res 2015;2015: 538171. [21] Drago L, De Vecchi E, Gabrieli A, De Grandi R, Toscano M. Immunomodulatory effects of Lactobacillus salivarius LS01 and Bifidobacterium breve BR03, alone and in combination, on peripheral blood mononuclear cells of allergic asthmatics. Allergy Asthma Immunol Res 2015;7:409–13. [22] Ryan KA, Daly P, Li Y, Hooton C, O’Toole PW. Strain-specific inhibition of Helicobacter pylori by Lactobacillus salivarius and other lactobacilli. J Antimicrob Chemother 2008;61:831–4. [23] Wu KG, Li TH, Peng HJ. Lactobacillus salivarius plus fructo-oligosaccharide is superior to fructo-oligosaccharide alone for treating children with moderate to severe atopic dermatitis: a double-blind, randomized, clinical trial of efficacy and safety. Br J Dermatol 2012;166:129–36. [24] Tsai CC, Lin PP, Hsieh YM. Three Lactobacillus strains from healthy infant stool inhibit enterotoxigenic Escherichia coli grown in vitro. Anaerobe 2008;14:61–7. [25] Chang YC, Tsai YC, Lin CF, Wang YC, Wang IK, Chung TC. Characterization of tetracycline resistance lactobacilli isolated from swine intestines at western area of Taiwan. Anaerobe 2011;17:239–45. [26] Zhang Y, Liu Y, Ma Q, Song Y, Zhang Q, Wang X, et al. Identification of Lactobacillus from the saliva of adult patients with caries using matrixassisted laser desorption/ionization time-of-flight mass spectrometry. PLOS ONE 2014;9:e106185. [27] Salminen MK, Rautelin H, Tynkkynen S, Poussa T, Saxelin M, Valtonen V, et al. Lactobacillus bacteremia, clinical significance, and patient outcome, with special focus on probiotic Lrhamnosus GG. Clin Infect Dis 2004;38:62–9. [28] Meini S, Laureano R, Fani L, Tascini C, Galano A, Antonelli A, et al. Breakthrough Lactobacillus rhamnosus GG bacteremia associated with probiotic use in an adult patient with severe active ulcerative colitis: case report and review of the literature. Infection 2015 [Epub ahead of print].