Epidemiology, clinical characteristics, and outcome of infective endocarditis due to Abiotrophia and Granulicatella in a Tertiary Hospital in China, 2015–2023: a retrospective study

Abstract

Background

Abiotrophia (ABI) and Granulicatella (GRA) are rare causative pathogens in infective endocarditis (IE). This study aims to describe the epidemiology, clinical characteristics, and outcome of ABI/GRA-IE. The main features of ABI/GRA-IE were compared with Viridans group streptococci (VGS) IE.

Methods

From January 2015 to December 2023, a total of 1531 definite IE in Zhongshan Hospital, Fudan University, Shanghai, China were retrospectively enrolled in this study. Clinical and laboratory data were collected.

Results

Forty-five ABI/GRA-IE cases were identified, representing 2.9% of all IE cases in Zhongshan Hospital between 2015 and 2023, compared to 20.1% of VGS-IE. ABI and GRA IE shared similar clinical characteristics. Congenital valvulopathy was reported in 21 (46.7%) ABI/GRA-IE and 85 (28.8%) VGS-IE (P = 0.025). Pulmonary valve was more frequently affected in ABI/GRA-IE (6 [13.3%]) than VGS-IE (7 [2.4%]) (P = 0.002). Congestive heart failure was observed in 30 (66.7%) ABI/GRA-IE and 103 (34.9%) VGS-IE (P < 0.001). Systemic embolization excluding central nervous system (CNS) occurred in 13 (28.9%) ABI/GRA-IE and 39 (13.2%) VGS-IE (P = 0.012). In-hospital mortality was reported as 4.4% in ABI/GRA-IE and 3.7% in VGS-IE (P = 0.854).

Conclusion

GRA/ABI-IE was approximately one-seventh as prevalent as VGS-IE. Congestive heart failure and systemic embolization (excluding CNS) were more frequent in GRA/ABI-IE compared to VGS-IE. Mortality of ABI/GRA-IE in this study was comparable to that of VGS-IE and lower than previously reported results.

Background

Abiotrophia (ABI) and Granulicatella (GRA) are fastidious gram-positive cocci previously described as nutritionally variant streptococci (NVS) because they grow as satellite colonies around other microorganisms or in complex media containing sulfhydryl compounds, such as cysteine or pyridoxal hydrochloride. NVS include one Abiotrophia species (Abiotrophia defective) and three Granulicatella species (Granulicatella adiacens, Granulicatella elegans and Granulicatella balaenopterae) [1]. They are part of the normal oral cavity, urogenital, and intestinal flora, and an important cause of bacteremia and infective endocarditis (IE) [2,3,4,5].

IE is a life-threatening disease with high mortality [6,7,8]. IE due to ABI and GRA represents around 1–3% of all IE, typically presenting with a subacute course [2, 9]. However, epidemiology, clinical characteristics, and outcome of ABI/GRA related IE remain poorly studied.

The largest case series of IE due to ABI or GRA for a single institution was reported [2]. In this single center study, periannular complication were more common in ABI/GRA-IE and overall mortality was low. The largest multicenter prospective cohort study of ABI/GRA-IE was recently published and clinical features of ABI/GRA-IE were compared with Viridans group streptococci (VGS) IE [10]. In this study, patients with ABI/GRA-IE were younger, had similar clinical features and rates of surgery and better prognosis than VGS-IE. However, most of the reported cases of ABI/GRA-IE were from Europe, America, or Australia. Data from Asian patients were lacking.

Here, we retrospectively reviewed all the IE cases admitted to Zhongshan Hospital, Fudan University, Shanghai, China from 2015 to 2023 and identified 45 ABI/GRA-IE. Epidemiology, clinical characteristics, and outcome of ABI/GRA-IE were described and compared with those of VGS-IE.

Methods

Patients

We retrospectively searched the electronic medical history retrieval system in Zhongshan Hospital, Fudan University, Shanghai, from January 2015 to December 2023, using the term “infective endocarditis”. Zhongshan Hospital is a 3000-bed, tertiary-level university hospital and a cardiac surgery center in China.

The diagnosis of IE was made according to the modified Duke criteria [11]. Only definite IE cases were included in this study. All the patients admitted to Zhongshan Hospital were asked on admission whether they agreed to sign the informed consent of biological sample donation and agreed that the donated samples and related information could be used for all medical research. Only patients who had signed the informed consent of biological sample donation were included in this study. The exclusion criteria in this study were patients whose demographic information, clinical presentations, etiological tests, laboratory and echocardiography examinations, complications, treatments, and outcomes were incompletely recorded, not including etiologically unknown IE. As for the etiologically unknown IEs in this study, etiological tests such as blood culture or valve culture were conducted and turned negative, and culture results were precisely recorded in the medical history. A total of 1591 definite IE patients were screened, including 60 patients with incomplete medical history records and 1531 enrolled in this study. Demographic information, medical history, clinical presentations, laboratory and echocardiography examinations, complications, treatments, and outcomes were documented.

Definition

Patients were identified with congestive heart failure if at least one of the following conditions were met: (1) N-terminal pro-B type natriuretic peptide (NT-proBNP) on admission higher than 1500 pg/mL; (2) cardiac function of III or IV class evaluated with New York Heart Association (NYHA) classification; (3) echocardiogram on admission identifying reduced cardiac activity. Duration of symptoms referred to the period from onset until admission. Follow-up time referred to the period from admission until the last follow-up. Cumulative mortality included the period from admission until the last follow-up.

Culture

Blood samples (8–10 mL) of IE patients were injected into aerobic, anaerobic, and fungal blood culture bottles [BD BACTECTM, Becton, Dickinson, and Co. (BD), Franklin Lakes, NJ, USA] and then loaded into an automated continuous monitoring system (BD BACTECTM, BD) for 7 days. For the patients who received cardiac surgery, valve homogenates after surgery were cultured onto blood agar plates, chocolate agar plates, and fungal chromogenic plates for 14 days. If the culture showed growth of microbes, strain identification was conducted by VITEK MALDI-TOF mass spectrometry (bioMérieux, Craponne, France).

In vitro susceptibility assays

The minimal inhibitory concentrations (MICs) of penicillin were determined by the epsilometer test (E-test). The Kirby-Bauer disc diffusion method was used to determine in vitro susceptibility of clindamycin, linezolid, cefepime, ceftriaxone, levofloxacin, vancomycin, and erythromycin.

Statistical analysis

For continuous variables, data were expressed as median (interquartile range, IQR) if they followed a non-normal distribution and comparative analysis between the two groups was conducted by the Mann-Whitney U test. For discrete variables, comparative analysis between the two groups was conducted by the Chi-square test. Data analysis was performed with the statistical software SPSS version 21.0 (IBM Corp., Armonk, NY, USA). All tests were two-tailed, and statistical significance was considered at P < 0.05.

Results

Cases identified

From January 2015 to December 2023, a total of 1591 definite IE patients were admitted to Zhongshan Hospital, Fudan University, Shanghai, China, including 60 patients with incomplete medical history records. Among the remaining 1531 IE cases, 307 (20.1%) were caused by VGS, 45 (2.9%) by ABI/GRA, 534 (34.9%) by other pathogens, and 645 (42.1%) by unknown pathogens, as shown in Fig. 1. Among the 307 VGS-IE cases, 12 patients lost follow-up. Clinical and laboratory data of the remaining 295 VGS-IE cases were used in the final analysis. Among the 45 ABI/GRA-IE cases, no patient lost follow-up. Nineteen (1.2%) cases were due to ABI and 26 (1.7%) cases were due to GRA.

Fig. 1

Flowchart of patient inclusion. A total of 1591 definite IE patients were screened from January 2015 to December 2023 in Zhongshan Hospital, Fudan University, Shanghai, China. Sixty patients with incomplete medical history records were excluded. Among the remaining 1531 IE cases, 307 (20.1%) were caused by VGS, 45 (2.9%) by ABI/GRA, 534 (34.9%) by other pathogens, and 645 (42.1%) by unknown pathogens. Among the 45 ABI/GRA-IE cases, 19 (1.2%) were due to ABI and 26 (1.7%) were due to GRA Abbreviations ABI, Abiotrophia; GRA, Granulicatella; IE, infective endocarditis; VGS, Viridans group streptococci

Abiotrophia and Granulicatella infective endocarditis

The demographic characteristics, baseline comorbidities, underlying cardiopathy, clinical and echocardiographic findings, complications, and outcomes of ABI and GRA IE were summarized in Table 1. As the characteristics mentioned above and outcomes were similar between ABI and GRA IE, they were analyzed together and compared with VGS-IE. The detailed clinical and microbiological characteristics of ABI/GRA-IE patients are provided in Additional file 1. Detailed complications, treatment, and outcome of ABI/GRA-IE cases are provided in Additional file 2.

Table 1 Comparative characteristics of Abiotrophia and Granulicatella Infective Endocarditis cases

Demographic features, type of infective endocarditis, comorbidities and underlying cardiopathy

Table 2 summarized the demographic characteristics, baseline comorbidities, underlying cardiopathy, clinical and echocardiographic findings, complications, and outcomes of ABI/GRA-IE and VGS-IE. The median age in the ABI/GRA-IE was 48 years (IQR, 40, 59) and 51 (IQR, 37.5, 60) in the VGS-IE group (P = 0.832). Most of the cases were male in both groups, 34 (75.6%) in ABI/GRA-IE and 220 (74.6%) in VGS-IE (P = 0.965). All the cases were community acquired in both groups. Most of the cases were native valve endocarditis (NVE) in both groups, 42 (93.3%) in ABI/GRA-IE and 274 (92.9%) in VGS-IE (P = 0.839). Baseline comorbidities including diabetes mellitus, hypertension, malignant tumor, hepatitis B virus infection, renal insufficiency, and immunosuppression were similar in both groups. Congenital valvulopathy was more frequent among ABI/GRA-IE (21 [46.7%]) than VGS-IE (85 [28.8%]) (P = 0.025).

Table 2 Comparative characteristics of Abiotrophia and Granulicatella, and Viridans Group Streptococci Infective Endocarditis Cases

Clinical presentation and echocardiographic findings

Fever was less frequent among ABI/GRA-IE (29 [64.4%]) than VGS-IE (233 [79.0%]) (P = 0.048). Splenomegaly was more frequent among ABI/GRA-IE (14 [31.1%]) than VGS-IE (37 [12.5%]) (P = 0.002). More chest tightness was presented in ABI/GRA-IE (32 [71.1%]) than VGS-IE (142 [48.1%]) (P = 0.007). Pulmonary valve was more frequently affected in ABI/GRA-IE (6 [13.3%]) than VGS-IE (7 [2.4%]) (P = 0.002).

In vitro susceptibility and antimicrobial therapy

In vitro susceptibility results were available in 12 ABI-IE and 12 GRA-IE (9 G. adiacens and 3 G. elegans) cases (1 strain in each case), as summarized in Tables 3 and Fig. 2. Eight (66.7%) ABI strains and 9 (75.0%) GRA strains were sensitive to penicillin. All the 24 ABI/GRA strains were susceptible to vancomycin and linezolid. The antimicrobial susceptibility information of the other 21 ABI/GRA-IE cases was lost during data migration, as the electronic medical history retrieval system in Zhongshan Hospital had several updates and even replacements from 2015 to 2023.

Table 3 In vitro susceptibility results in Abiotrophia and Granulicatella Infective Endocarditis cases

Fig. 2

In vitro susceptibility of 12 ABI-IE and 12 GRA-IE cases. For the 12 ABI-IE cases, eight (66.7%), 7 (58.3%), 12 (100.0%), 9 (75.0%), 9 (75.0%), 11 (91.7%), 12 (100.0%), and 5 (41.7%) were susceptible to penicillin, clindamycin, linezolid, cefepime, ceftriaxone, levofloxacin, vancomycin, and erythromycin, respectively. For the 12 GRA-IE cases, nine (75.0%), 9 (75%), 12 (100.0%), 10 (83.3%), 9 (75.0%), 11 (91.7%), 12 (100.0%), and 6 (50.0%) were susceptible to penicillin, clindamycin, linezolid, cefepime, ceftriaxone, levofloxacin, vancomycin, and erythromycin, respectively. Abbreviations ABI, Abiotrophia; Ctx, ceftriaxone; Da, clindamycin; E, erythromycin; Fep, cefepime; GRA, Granulicatella; IE, infective endocarditis; Lev, levofloxacin; Lzd, linezolid; Peni, penicillin; Van, vancomycin

Vancomycin as monotherapy were used in 21 ABI/GRA-IE cases. Twenty ABI/GRA-IE patients were treated with combination antimicrobial therapy (β-lactam plus quinolone, β-lactam plus aminoglycoside, vancomycin plus β-lactam, or other antibiotics).

Complications

Paravalvular complications including perforation, abscess and prosthetic paravalvular dehiscence were similar between ABI/GRA-IE and VGS-IE, as shown in Table 2. More congestive heart failure was observed in patients with ABI/GRA-IE (30 [66.7%]) than VGS-IE (103 [34.9%]) (P < 0.001). Central nervous system (CNS) involvement was comparable in both groups while systemic embolization (excluding CNS) was more frequent among ABI/GRA-IE (13 [28.9%]) than VGS-IE (39 [13.2%]) (P = 0.012).

Outcome

Follow-up time was longer in the ABI/GRA-IE group. The median follow-up time in ABI/GRA-IE was 360 days (IQR, 157.5, 1297.5) and 270 days (IQR, 90, 670) in VGS-IE (P = 0.020). Forty-two (93.3%) patients with ABI/GRA-IE and 270 (91.5%) patients with VGS-IE received cardiac surgery during hospitalization (P = 0.905). In-hospital death was reported as 4.4% in ABI/GRA-IE and 3.7% in VGS-IE (P = 0.854). Cumulative mortality was 4.4% in ABI/GRA-IE and 4.1% in VGS-IE (P = 0.776).

Discussion

ABI and GRA are relatively rare causative pathogens in IE [12,13,14]. However, there might be underestimation due to the special growth requirements of ABI/GRA. Prolonged incubation might allow for identification of these fastidious microorganisms and improve the positive rate of blood culture [15]. To our knowledge, this is the largest cohort study of ABI/GRA-IE for a single center. In our hospital, ABI and GRA related IE presented similar clinical characteristics and outcomes. ABI/GRA-IE was approximately 7 times less frequent than VGS-IE. Compared to VGS-IE, splenomegaly and chest tightness were more frequently found in the ABI/GRA-IE group while fever was less frequent. Pulmonary valve was more frequently affected in ABI/GRA-IE. Paravalvular complications were quite comparable in both groups while congestive heart failure and systemic emboli (excluding CNS) were more frequent in ABI/GRA-IE.

The aortic and mitral valves were reported to be the most commonly affected [2, 10, 14]. In our study, mitral valves were the most commonly affected (66.7%), followed by aortic valves (55.6%), pulmonary valves (13.3%), and tricuspid valves (8.9%), similar to previous studies.

In a previous study, the penicillin-non-susceptible rate of ABI and GRA was relatively high: 66.7% and 53.7% respectively [16]. However, in another study conducted by Téllez et al.., 84.6% of ABI and 90.9% of GRA strains were penicillin sensitive [2]. María A. Cañas et al.. observed reduced susceptibility to penicillin in both ABI and GRA, with zero of six (0%) and two of nine (22%) strains being completely susceptible, respectively [17]. Antimicrobial susceptibilities of ABI/GRA varied in different studies, partly due to limitations of sample size and regional diversities [1, 18, 19]. In our study, 66.7% of ABI strains and 75.0% of GRA strains were sensitive to penicillin.

In previous studies, in-hospital surgery was performed in 27-70% of ABI/GRA-IE [2, 10, 20, 21]. In our study, surgery was performed in 84.2% in ABI-IE and 100.0% in GRA-IE. The surgery rate in our study is quite high, which could partly be explained by the specific characteristics of our hospital. Zhongshan Hospital is a regional cardiac surgery center, and many patients have already received antibiotic treatment and evaluation before being admitted with indications for surgery. Etiological diagnosis of IE is interfered by previous antibiotic usage, which could partly explain the high portion of etiologically unknown IE (645 of 1531, 42.1%) in this study.

Mortality of ABI/GRA-IE varied from 2.1 to 25% in different previous studies [2, 10, 20]. Bouvet A. et al. reported the mortality as high as 17-20%. In our study, mortality of ABI/GRA-IE was 4.4%, comparable to that of VGS-IE (4.1%) and lower than previously reported results. Precise etiological diagnosis, targeted antibiotic treatment, and high-quality surgery might explain the improved outcomes. On the other hand, as we mentioned above, this study was conducted in a regional cardiac surgery center, and many patients had already received antibiotic treatment and evaluation before being admitted with indications for surgery. These patients were relatively mild and might have a better prognosis. This factor might also contribute to the low mortality of ABI/GRA-IE in this study.

There were some limitations in this study. This was retrospective study in a single center. Patient selection bias may have existed, and some clinical and laboratory data were not comprehensive. Antimicrobial susceptibilities were only available in 24 ABI/GRA-IE cases. In addition, due to the limitations of sample size, we did not construct a predictive model for the prognosis of ABI/GRA-IE.

Conclusions

ABI-IE and GRA-IE seem to have similar clinical characteristics. Patients with ABI/GRA-IE have comparable surgical rate and prognosis with VGS-IE patients, although congestive heart failure and systemic embolization (excluding CNS) were more frequently detected. Mortality of ABI/GRA-IE in this study was comparable to that of VGS-IE and lower than previously reported results. These findings differ from previous reports and provide more understanding in ABI/GRA-IE.