Retrospective Study of Bacteremia Caused by Alistipes Species in Japan, 2016–2023

Study Design and Data Collection

This was a retrospective cohort study of Alistipes bacteremia at Kameda Medical Center in Japan conducted between April 2016 and December 2023. We reviewed 31,875 patients who underwent blood cultures and identified Alistipes cases. For patients with multiple episodes, only the first episode was included in this study. Patient data and microbiological testing results were collected from medical records and a laboratory system. Patient data included age, sex (male or female), infection type (community-acquired, healthcare-associated), comorbidities, symptoms, infection sites, procedures, antimicrobial therapy, hospitalization days from blood collection, and outcomes. Healthcare-associated infections were defined as cases with one of the following conditions: onset after 48 h of hospitalization or patients receiving medical care, such as a nursing home, dialysis, or outpatient chemotherapy. The Ethics Committee of Kameda Medical Center approved this study (agreement no. 23-121). Written informed consent was not required because the data used were anonymized.

Blood Culture and Identification Testing

Blood cultures were performed using BD Bactec™ FX Blood Culture System (Becton, Dickinson and Company, https://www.bd.com). Blood samples were collected using Plus Aerobic and Lytic Anaerobic Media for adult patients and Peds Plus Medium for pediatric patients (Becton, Dickinson and Company). The blood cultures were incubated at 35 °C for 7 days. The number of positive blood culture sets, time to positivity, and detected Alistipes were recorded. Alistipes strains were stored at –80 °C for identification, whole genome sequencing, and antimicrobial susceptibility testing. The preserved strains were cultured anaerobically using Brucella HK agar medium (Kyokuto Pharmaceutical Industrial, https://www.kyokutoseiyaku.co.jp) at 35 °C for 2 days and were used for each test.

The strains were identified using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). MALDI-TOF MS was performed using the MALDI Biotyper system, which included the microflex LT/SH and flexControl version 3.4 (Bruker Daltonics GmbH, https://www.bruker.com). Alistipes colonies that developed on the Brucella HK agar medium were selected and applied to the test plate. Mass spectra were analyzed using the MALDI Biotyper Library, which recorded the highest score for the predicted species.

Identification results were accepted at the species level for scores of ≥ 2.0. In cases where an identification result could not be obtained using the standard cell smear method, the strains were retested using the formic acid on-plate or formic acid extraction method.

Polymerase Chain Reaction (PCR) and Whole Genome Analysis

The identification results were validated by 16S rRNA gene and whole genome analyses of the preserved strains. Non-preserved strains were excluded from this analysis. Genomic DNA was extracted from the strains using magLEAD and Consumable Kits (Precision System Science; https://www.pss.co.jp). GeneAtlas Type G (Astec, https://www.astec-corp.co.jp) was used as the thermal cycler and Premix Taq Hot Start Version (Takara Bio, https://www.takara-bio.com) for the PCR reagent. Using previously reported primers for the 16S rRNA gene (15, 16), PCR was performed according to the literature (17). PCR conditions and primers used are listed in Table S1. Sequencing was performed using FASMAC (https://fasmac.co.jp), and the resulting sequences were analyzed using EzBioCloud (18) for species identification. Additionally, 16S rRNA gene reanalysis was performed on three A. finegoldii strains (3302398, 4401054, and Granada) from previous bacteremia cases (12, 13). The sequence data for 3302398 (accession no. AY547271) , 4401054 (AY643082), and Granada (OM900033) were obtained from GenBank and analyzed using the same methods as the strains collected in this study.

Fragment libraries for the whole genome analysis were prepared using Illumina DNA Prep Tagmentation (M) Beads (Illumina, https://www.illumina.com). Sequencing was performed using the MiSeq Reagent Micro Kit (Illumina) to generate paired-end reads. De novo assembly was conducted using the CLC Genomics Workbench version 22.0.2 (Qiagen, https://www.qiagen.com). CheckM (19) was used as a quality check for the genome, and samples with a contamination rate of ≥ 5% were excluded from subsequent analysis. The average nucleotide identity (ANI) was performed using DFAST (20), identifying bacterial species with a threshold of ≥ 95%. For subspecies analysis, DNA-DNA hybridization (DDH) analysis was performed using the Genome–Genome Distance Calculator 3.0 (https://ggdc.dsmz.de) (21). The threshold for defining subspecies was set at 79% DDH of the previously reported criteria for subspecies classification (22). The presence of antimicrobial resistance genes was analyzed using the Comprehensive Antibiotic Resistance Database (23).

Antimicrobial Susceptibility Testing

The antimicrobial susceptibility of the preserved strains was evaluated using the broth microdilution method with Brucella Broth and Dry Plates (Eiken Chemical, https://www.eiken.co.jp). The inoculum was prepared according to the manufacturer’s instructions and transferred to a microplate at a final volume of approximately 1 × 10⁵ colony-forming unit per well. After inoculation, the microplate was incubated anaerobically at 35–37 °C for 46–48 h. Minimum inhibitory concentrations (MICs) were determined after confirming bacterial growth in the control wells. The Dry Plate included the following antibiotics: PEN, ampicillin-sulbactam (SAM), amoxicillin-clavulanic acid (AMC), piperacillin-tazobactam (TZP), ceftriaxone (CRO), cefoxitin (FOX), imipenem (IPM), clindamycin (CLI), minocycline (MINO), moxifloxacin (MXF), and metronidazole (MTZ). After determining the MICs, the concentrations required to inhibit 50% and 90% of the strains (MIC₅₀ and MIC₉₀) were calculated.

Statistical Analyses

The characteristics of Alistipes cases were summarized using EZR version 1.54 (24). Continuous variables are reported as medians and interquartile ranges (IQRs), and categorical variables are reported as actual numbers. The proportion of patients with Alistipes bacteremia among all patients who underwent blood culture was calculated using 95% confidence intervals. Time to positivity was reported as the median and IQR, excluding the results from blood culture sets with multiple pathogens.

Patient Characteristics and Clinical Courses

Thirteen Alistipes bacteremia were identified over a 7-year period. This number represented 0.04% of the patients with blood cultures, and all cases were single episodes. The characteristics of the patients with Alistipes bacteremia are shown in Table 1. The median patient age was 72 (IQR, 65–85) years, and 10 patients were aged > 65 years. Seven patients were female, and nine patients had comorbid conditions, with a Charlson score ≥ 1. Six patients had solid tumors, and one had a hematological malignancy. The most common clinical symptom was abdominal pain (nine patients), followed by fever (six patients). Of the 13 patients, 11 presented with gastrointestinal symptoms, including abdominal pain, vomiting, diarrhea, and/or abdominal distention. The median time to positivity was 81 (IQR, 71–106) h, and Alistipes strains from three cases were detected in blood cultures that tested positive ≥ 120 h after the start of incubation. Four patients had two sets of positive blood cultures.

The clinical presentations, treatments, and outcomes of Alistipes bacteremia are summarized in Table 2. Eleven of thirteen patients were diagnosed with gastrointestinal tract disease, with infection sites in these areas. The remaining two patients had Alistipes bacteremia with unknown infection sites (cases 7 and 12). The patient in case 7 developed a fever during treatment for a catheter-related bloodstream infection caused by Staphylococcus aureus, and Alistipes indistinctus was detected in a blood culture collected at the time of fever. The patient in case 12 had a history of a cerebral infarction and cholecystectomy. The patient presented with chills, fever, and vomiting, and blood cultures were positive for A. putredinis. All patients were clinically diagnosed with true bacteremia and received antimicrobial therapy. The duration of the antimicrobial therapy ranged from 13 to 36 days. The most commonly prescribed antimicrobials were β-lactam/β-lactamase inhibitors (BLBLIs), with nine patients receiving SAM and six patients receiving TZP. Six patients were positive for multiple pathogens, including Escherichia coli, Klebsiella pneumoniae, Bacteroides species, and others.

Clinical Outcomes

Of the 13 patients, 11 achieved clinical resolution following antimicrobial therapy, whereas the remaining 2 (cases 1 and 8) had a fatal outcome. The patient in fatal case 1 was a woman in her 80s with rectal cancer and liver metastases, who was admitted to the hospital for abdominal pain. Upon arrival at the hospital, the patient experienced septic shock. The patient received intravenous fluids, blood cultures were obtained, and TAZ and VCM were administered. Her condition briefly stabilized following the infusion, but subsequently deteriorated with the development of acidosis, hyperkalemia, and elevated lactic acid levels. The patient was diagnosed with colonic perforation and obstruction. The decision was made to transfer the patient to a hospital near her home for palliative care and to discontinue aggressive treatment. However, her condition deteriorated, and she died within the day of the transfer. After the patient died, A. finegoldii was detected in the blood culture obtained during an episode of septic shock.

The patient in fatal case 8 was a man in his 70s who was hospitalized for a comprehensive evaluation of adenocarcinoma of unknown primary origin. The patient was administered fluconazole for oral candidiasis. During hospitalization, the patient developed septic shock and was administered intravenous fluids. Blood cultures were obtained, and TAZ was administered. Although the patient initially stabilized, he later developed fecal emesis, likely due to intestinal obstruction. One day after the onset, the patient eventually died of septic shock. Blood cultures obtained during a septic shock episode revealed the presence of A. onderdonkii.

Microbiological Characteristics

The microbiological characteristics of the Alistipes cases and strains are presented in Table 3. Alistipes strains formed small colonies on the Brucella HK agar medium (Fig. 1). MALDI-TOF MS identified 13 strains with species-level accuracy, including A. finegoldii (n = 4), A. onderdonkii (n = 4), A. putredinis (n = 3), and A. indistinctus (n = 2). Of the 13 strains, 16S rRNA and whole genome analyses were performed on 10 preserved strains (KML 24001–24010). 16S rRNA gene analysis showed that 10 of the preserved strains were 99.8–100% similar to the species identified by MALDI-TOF MS (Table S2).

• Download figure
• Open in new tab

Fig 1. Small colonies of Alistipes finegoldii grown on agar medium in anaerobic culture for 48 h^*

^*Figure 1 shows A. finegoldii (arrow mark) and Escherichia coli ATCC 25922 colonies grown on Brucella HK agar medium after 2 days of anaerobic culture.

The draft genomes of nine strains (KML 24001-24009) had contamination rates of < 1% and were used for further analysis (Table S3). One remaining strain (KML 24010) had a contamination rate of > 5% and was excluded from further analysis. The genome sizes of A. finegoldii, A. onderdonkii, A. putredinis, and A. indistinctus were 3.6 Mb, 3.3–3.6 Mb, 2.5–2.8 Mb, and 2.8–2.9 Mb, respectively (Table S4). The ANI results of KML 24001-24009 were in agreement with the results of MALDI-TOF MS, with an ANI of 99.8–100% (Table S5). DDH analysis for A. onderdonkii revealed that KML 24001 and KML 24005 showed DDH > 80% only for A. onderdonkii subsp. vulgaris, and KML 24008 showed DDH > 80% only for A. onderdonkii subsp. onderdonkii. The subspecies of each strain were identified (Table S6).

Strains with > 5% contamination were pure-cultured, and the two types of colonies grown were re-identified by MALDI-TOF MS. One strain (KML 24010) was identified as A. finegoldii by MALDI-TOF MS, and this strain showed 99.8% similarity to A. finegoldii by 16S rRNA gene analysis. Another strain (KML 24011) was not identified by MALDI-TOF MS and showed 99.7% similarity to A. ihumii AP11 (JX101692). The 16S rRNA gene sequence and draft genome data were deposited in GenBank via the DNA DataBank of Japan, and the accession numbers are given in Tables S2 and S4.

Reanalysis of the three A. finegoldii strains (3302398, 4401054, and Granada) isolated in previous reports yielded the following results: strain 3302398 showed 100% similarity to A. dispar 5CPEGH6 (accession no. AP019736), and strain 4401054 showed 99.0–100% similarity to A. onderdonkii subsp. onderdonkii DSM 19147 (ARFY01000017). Strain Granada showed 97.6% similarity with A. finegoldii DSM 17242 (CP003274); however, the species could not be confirmed.

Antimicrobial Susceptibility and Resistance Genes

Antimicrobial susceptibility data were available for Alistipes 10 strains (Table 4). Low MIC₉₀ values were observed for SAM (MIC₉₀, 2 μg/mL), AMC (2 μg/mL), TZP (≤ 4 μg/mL), IPM (1 μg/mL), CLI (1 μg/mL), and MTZ (0.5 μg/mL). For CLI, only one A. putredinis strain showed an MIC of 2 μg/mL, and this strain possessed the erm(G) gene. The antibiotics with high MIC₉₀ values were PEN (MIC₉₀, > 1 μg/mL), CRO (> 32 μg/mL), FOX (32 μg/mL), MINO (8 μg/mL), and MXF (> 4 μg/mL). MICs were particularly high for PEN and MINO, with MIC₅₀ values of > 1 μg/mL and 4 μg/mL, respectively. Seven strains with MICs ≥ 4 μg/mL for MINO harbored the adeF and tet(Q) genes. Two strains exhibited MICs of 32 μg/mL for FOX, and these strains were all A. indistinctus. Three strains showed MICs ≥ 4 μg/mL for MXF, all of which were A. onderdonkii.