Is echocardiography mandatory for all Steptococcus gallolyticus bacteremia?

Background: Streptococcus gallolyticus, formerly known as one of the Streptococcus bovis group, is frequently associated with endocarditis. Current guidelines recommended diagnostic work-up for endocarditis among patients with S. gallolyticus bacteremia. However, S. gallolyticus subsp. pasteurianus, was found to be associated with neonatal sepsis and liver diseases and is less commonly associated with endocarditis compared with S. gallolyticus subsp. gallolyticus. Our study aimed to identify the risk factors for S. gallolyticus subsp. pasteurianus endocarditis to help select the patients for echocardiography . Methods: In this retrospective cohort study, medical records from all adult patients with S. gallolyticus subsp. pasteurianus isolated from blood cultures at Phramongkutklao Hospital from 2009 to 2015 were reviewed. Patients who had mixed bacteremia or missing records were excluded from the study. Result: During the study period, S. gallolyticus subsp. pasteurianus were isolated among 106 individuals. Mean age was 66.9±15.6 years. Most patients (61.3%) were male, with cirrhosis as the most common underlying diseases (46.2%), followed by malignancy and chronic kidney disease. Most common manifestations included primary bacteremia (44.3%), followed by spontaneous bacterial peritonitis (23.6%). Infective endocarditis was found among 9 patients. No patients with cirrhosis or single blood specimen of bacteremia had endocarditis (RR 0; p-value 0.003, and RR 1.35; p-value 0.079). The common complications associated with endocarditis were acute respiratory failure (RR 4.32; p-value 0.05), whereas acute kidney injury was a protective factor (RR 0; p-value 0.01). Among 76 patients who had records of 2-year follow-up, no new diagnosis of endocarditis or malignancy was observed. Conclusion: Among patients with S. gallolyticus subsp. pasteurianus bacteremia, echocardiography might not be needed among patients with cirrhosis or transient bacteremia. Key word: Streptococcus

different DNA groups. Common human pathogenic subspecies include S. gallolyticus subsp.
Nevertheless, S. gallolyticus subsp. pasteurianus was found to be associated with neonatal sepsis, liver diseases and malignancies of the digestive tract, i.e., gastric, pancreatic, hepatobiliary and colorectal cancer and is less commonly associated with endocarditis (14-23%) compared with S. gallolyticus subsp. gallolyticus (50-53%). (9-15) Nevertheless, many guidelines still recommend diagnostic work-up for endocarditis in all S. gallolyticus bacteremia cases, which may result in unnecessary procedures and higher costs of hospitalization. (16,17) Because of the lack of large-scale studies investigating the association between S. gallolyticus subsp.
pasteurianus and endocarditis, we aimed to find the risk factors for S. gallolyticus subsp. pasteurianus endocarditis to help select appropriate patients for echocardiography, expected to minimize the unnecessary investigation.

Objectives
Our primary objective was to determine the risk factors for developing endocarditis among patients with S. gallolyticus subsp. pasteurianus bacteremia and secondary objective was to determine the 30day mortality rate of S. gallolyticus subsp. pasteurianus bacteremia

Study setting and design
A retrospective cohort study was conducted. We collected all blood isolates of S. gallolyticus subsp. Thailand. The inclusion criteria were participants aged over 18 years with monomicrobial S. gallolyticus subsp. pasteurianus bacteremia. Participants referred to other hospitals within the first seven days were excluded. Demographic data, patient characteristics, comorbidities, immunosuppressive status, microbiological data and patient outcomes were recorded. The site of infection was defined according to definitions of the Centers for Disease Control and Prevention. (18) Patients whose foci of infection were unidentified were classified as primary bacteremia. Results from echocardiography, ultrasonography, computed tomography and endoscopic examinations searching for endocarditis, hepatobiliary pathology and colonic lesions were collected. All-cause-30-day crude mortality after the onset of bacteremia was recorded. Participants' medical records were followed up to two years focusing on the finding of new malignancy or endocarditis. The study protocol followed the guidelines of the Declaration of Helsinki and ethics approval was obtained from the Institutional Review Board, Royal Thai army Department.

Microbiological analysis
Bacterial isolates were collected by the clinical microbiological laboratory at the study hospital.
Species were identified using the BACTEC system (Becton Dickson, Sparks, MD). The VITEK 2 automated system (bio-Merieux, Hazelwood, MO) was used to identify bacterial and subspecies.
Antimicrobial susceptibility was determined using the disk diffusion test of various antimicrobial agents including penicillin, gentamicin, clindamycin, ceftriaxone, erythromycin, azithromycin, moxifloxacin, levofloxacin, tigecycline, linezolid, daptomycin and vancomycin. Minimum inhibitory concentrations (MICs) of the isolates to penicillin and ceftriaxone were performed using the E-test.
Disk diffusion and MICs were interpreted using the Clinical and Laboratory Standards Institute (CLSI)

Statistical analysis
For categorical variables, Fisher's exact test and the Chi-square test were used. The Mann-Whitney test was used to compare continuous variables. For all analyses, a two-sided p-value of 0.05 was considered significant. Potentially significant predictors in the univariate analyses (p-value <0.10) were included in a forward, stepwise multiple logistic regression model to identify the most important factors related to developing endocarditis. All statistical analyses were performed using Stata 12.0 Software (StataCorp, USA).

Results
During the study period, S. gallolyticus subsp. pasteurianus was isolated in 106 individuals, in whom 191 specimens (88%) blood culture specimens were positive among the overall of 217 specimens taken. The mean age was 66.9±15.6 years. Most patients were male (61.3%), with cirrhosis as the most common underlying disease, followed by diabetes mellitus, malignancy, and chronic kidney disease. Nearly 20% of participants had experienced a malignancy before bacteremic episodes. The most common malignancies included hepatocellular carcinoma, colorectal cancer and hematologic malignancy. The most common clinical manifestations comprised primary bacteremia, followed by spontaneous bacterial peritonitis and endocarditis. Acute kidney injury, septic shock, and respiratory failure were the three most common complications. Almost all participants had bacteria isolated in two blood culture specimens. The 30-day crude mortality rate was 21.7%, and baseline characteristics of all participants are shown in the Table 1.
Thirty-five patients underwent echocardiography based on clinical presentation and physician decision as Figure 1. Nine patients were diagnosed with infective endocarditis as described in Table 2.
All of whom had positive blood culture in more than two specimens (RR 1.35; p-value 0.079). None of the cirrhosis patients had endocarditis (RR 0; p-value 0.003). The common complication associated with endocarditis was acute respiratory failure (RR 4.32; p-value 0.05), whereas acute kidney injury served as a protective factor (RR 0; p-value 0.01) as shown in Table 3. Because of the small number of patients with endocarditis, multivariate analysis was not performed.
Among 76 patients with records of two-year follow-up, no new diagnosis of endocarditis or malignancy was observed.
The majority of S. gallolyticus subsp. pasteurianus was susceptible to antimicrobial agents, with penicillin susceptibility at 97%. However, a high rate of erythromycin resistance (24.5%) was observed as shown in the Table 4.

Discussion
In this cohort study, nearly one half of the patients with S. gallolyticus subsp. pasteurianus bacteremia presented cirrhosis. This finding was consistent with those of related studies reporting that 15 to 39% of patients had liver diseases or cirrhosis. (8,14,15) Regarding the clinical presentations, this study was consistent with others reporting primary bacteremia as the most common clinical presentation, followed by spontaneous bacterial peritonitis, and infective endocarditis. Although this is the first study to demonstrate a correlation between S. gallolyticus subsp.
pasteurianus and infective endocarditis among noncirrhotic patients, it had certain limitations. First, although a large number of participants were enrolled in this study, only 8.5% presented endocarditis, which might have overestimated the true association. This small proportion helps confirm that, unlike S. gallolyticus subsp. gallolyticus, S. gallolyticus subsp. pasteurianus uncommonly causes endocarditis. Second, the ability of species identification using VITEK 2 system to discriminate between the two subspecies of S. gallolyticus is not apparent. Nevertheless, studies comparing the Vitek 2 and sodA sequencing method showed generally acceptable agreement. (26) Finally, due to the retrospective design, factors influencing the physician's reason for the decision to perform colonoscopy and echocardiography was unidentified, which might have underestimated the prevalence of endocarditis and cancer. However, this bias was minimized by the long-term follow-up.

Ethics approval and consent to participate
The study protocol followed the guidelines of the Declaration of Helsinki and ethics approval was obtained from the Institutional Review Board, Royal Thai army Department The Ethics Committee of the Institutional Review Board, Royal Thai army Department waived the need to obtain consent for the collection, analysis and publication of the retrospectively obtained and anonymized data for this non-interventional study.

Availability of data and materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare no potential conflict of interest relevant to this article existed.

Funding
This study was supported by the Division of Infectious Diseases, Department of Internal Medicine, Phramongkutklao Hospital.
The funder supported budget in patient data collection process and all research equipment expenses e.g. office supplies.

Authors' contributions
WN prepared proposal, collected the patient data and was a major contributor in writing the manuscript.
VV analyzed and interpreted the patient data regarding to factors related to developing endocarditis in S. gallolyticus subsp. pasteurianus bacteremia. J T assisted in patient data collection.
DC supervised the manuscript writing and managed the funding.
All authors read and approved the final manuscript.