Clinical characteristics of bloodstream infection by Parvimonas micra: retrospective case series and literature review

Anaerobic bacteria are major components of the microflora on the mucosal membranes in humans, and several hundred distinct species of anaerobic microorganisms have been identified by classical and molecular methods. Gram-positive anaerobes (GPAs) are found in the commensal microbiota of the digestive tract, with some species colonizing the urogenital tract and skin. GPAs include Clostridium spp., Gram-positive anaerobic cocci (GPAC), and Gram-positive non-spore-forming anaerobes (GPNAs) [1]. As reported, Clostridium spp., GPAC, and GPNAs account for 20.7, 14.7, and 6% of all anaerobic hematic isolates, respectively [1].

Parvimonas micra (P. micra) is a member of GPAC normally found in the oral cavity, respiratory system, gastrointestinal tract (GIT), and the female genitourinary tract. Originally known as Peptostreptococcus micros, the organism was reclassified as P. micra in 2006 [2]. P. micra is one of the bacterial species most frequently isolated from infected root canals of teeth with chronic apical periodontitis [3]. P. micra has also been implicated in meningitis [4], cervical and brain abscess [5, 6], infective endocarditis (IE) [7], and spondylodiscitis [8]. Recently, Badri et al. demonstrated that P. micra was one of the most commonly isolated GPAC species and characteristics of Parvimonas spp. bacteremia (mostly P. micra) were discussed [9]. However, due to historical difficulties with laboratory identification, there is a paucity of published data about the clinical characteristics or mortality due to bloodstream infection by P. micra.

A new technique, matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), was reported to correctly identify 91.2% of anaerobic isolates at the species level compared to 16S rRNA sequencing as the gold standard of diagnosis [10]. MALDI-TOF MS enables accurate and timely species classification of anaerobic organisms for appropriate treatment.

In this study, we retrospectively analyzed the causative pathogens of GPA bacteremia before (2013–2015) and after (2016–2018) the introduction of MALDI-TOF MS in a tertiary medical center. P. micra was found to be the second most commonly identified causative organism of GPA bacteremia using MALDI-TOF MS, therefore, we described the clinical features and laboratory diagnosis of P. micra bacteremia. Finally, we conducted a review of literatures on previously reported P. micra bacteremia cases.

In our analysis, 70 clinically relevant cases of GPA-associated bacteremia were detected in 2013–2015, whereas 126 were detected in 2016–2018 after introducing MALDI-TOF MS (sFigure 1). In 2016–2018, Clostridium perfringens was the most frequently identified causative organism (23/126 cases, 18.3%) of GPA bacteremia, followed by P. micra (22/126 cases, 17.5%), and Eggerthella lenta (12/126 cases, 9.5%) (Table 1). In 2016–2018, more cases of GPA bacteremia were identified at the specie level (83.3%, 105/126 cases) compared to those identified in 2013–2015 (62.9%, 44/70 cases). P. micra bacteremia was observed in 23 cases using either MALDI-TOF MS or conventional phenotypic method in our institute.

In addition to these 23 cases, we identified 2 cases of P. micra bacteremia in 2011 and 2019, where medical records data were available and sufficient to analyze the clinical characteristics. Details of a total of 25 cases with bacteremia caused by P. micra that we identified are shown in Table 2. The median age of the 25 patients with P. micra bacteremia was 83.0 years (range, 58–92 years), and 14/25 (56%) patients were male. Most patients (23/25, 92%) had community-onset disease. Nearly all patients (23/25, 92%) presented with fever; however, focal symptoms such as sore throat (1/25, 4%) and abdominal pain (1/25, 4%) were less frequently observed. The most common comorbidities were cancer (9/25, 36%) and diabetes (6/25, 24%). The laryngeal pharynx (7/25, 28%) and GIT (6/25, 24%) were identified as the most likely sources of bacteremia. The infection source was not identified in 9 cases (9/25, 36%). Twelve patients (48%) had polymicrobial bacteremia. The various pathogens associated with concomitant polymicrobial infection included Solobacterium moorei, Staphylococcus capitis, Fusobacterium nucleatum, Pseudomonas citronellolis, Bacillus subtilis, Streptococcus oralis, Streptococcus constellatus, Staphylococcus lugdunensis, Veillonella parvula, Actinomyces meyeri, Prevotella intermedia, Atopobium parvulum, Streptococcus anginosus, and Prevotella oris, Corynebacterium spp. Most P. micra isolates were susceptible to penicillin, ampicillin-sulbactam, piperacillin-tazobactam, and meropenem; however, clindamycin showed slightly elevated MIC (susceptibility rate, 86.7%) (Table 2). Most cases (24/25, 96%) were treated successfully, and surgery was performed in one patient with gastroesophageal junction carcinoma who had infection due to postoperative anastomotic leakage.

Literature review identified 279 articles, and these were assessed for inclusion. Overall, 251 articles were excluded based on the title and/or abstract and/or text or due to duplication. One article was excluded due to duplication with our case series. There were 27 reported cases of P. micra bacteremia (sTable 1, 26 cases in English and 1 case in Japanese). We also found that Badri et al. reported 96 cases of P. micra bacteremia. However, since combined data of clinical characteristics about P. micra and other Parvimonas spp. were demonstrated in that study, we described the results of the literature by Badri et al. in the section of discussion [9].

In 27 literature cases, P. micra bacteremia was frequently associated with spondylodiscitis (8/27, 29.6%), oropharyngeal infection (7/27, 25.9%) intraabdominal abscess (4/27, 14.8%), GIT infection (3/27, 11.1%), IE (3/27, 11.1%), and septic pulmonary emboli (3/27, 11.1%). Five patients (18.5%) had polymicrobial bacteremia, and pathogens associated with concomitant polymicrobial infection included Gemella morbillorum, Atopobium rimae, Fusobacterium nucleatum, Prevotella bivia, and Bacteroides fragilis. In the literature cases, P. micra bacteremia also had a favorable outcome (mortality, 1/27, 3.7%).

In our cases, using MALDI-TOF MS, P. micra was a frequently identified causative organism of GPA bacteremia. However, P. micra have rarely been identified as causative pathogens of GPA bacteremia using conventional methods due to difficulties with the laboratory identification. In both ours and the literature cases, P. micra bacteremia cases were increasingly diagnosed after 2015 (Fig. 1) and were mostly detected using MALDI-TOF MS or 16S rRNA gene sequencing (Table 2 and sTable 1). In a recent population-based study, GPAC bacteremia, identified using MALDI-TOF MS or 16S rRNA gene sequencing, was reported to be much more common than previously reported with an annual incidence of 3.4 cases per 100,000 persons per year [9], while P. micra was the most commonly isolated in the study. Thus, these new techniques might enable accurate diagnosis of P. micra bacteremia for appropriate treatment.

We reviewed 25 patients with P. micra bacteremia using MALDI-TOF MS or conventional phenotypic method from 2010 to 2019. We found that oropharyngeal infections (7/25, 28.0%) and GIT (6/25, 24.0%) were the primary sources of P. micra bacteremia; however, the sources of bacteremia in the remaining 9 patients (36.0%) were not recorded. As the presenting symptoms in these patients were non-specific (e.g., fever, shake, and chills), thorough searches for the primary source of bacteremia may not have been carefully conducted in these 9 cases. In 27 literature cases, P. micra bacteremia were also frequently associated with oropharyngeal infection (7/27, 25.9%) and GIT infection (3/27, 11.1%) [14‐31]. Other common infectious diseases of P. micra in the literature cases were spondylodiscitis (8/27, 29.6%), intra-abdominal abscess (4/27, 14.8%), IE (3.27, 11.1%), and septic pulmonary emboli (3/27, 11.1%) which is inconsistent with our results. Spondylodiscitis caused by P. micra is rare but an important infectious disease, which is sometimes diagnosed from samples of the bone or vertebral disk without positive blood culture [23]. IE due to anaerobic bacteremia account for 2–16% of all cases of IE over the past few decades [32], and P. micra might be one of the important anaerobic pathogens causing IE. Recently, Badri and coworkers reported clinical features about 100 cases of Parvimonas spp. (96 P. micra and 4 Parvimonas spp.), in which common focal infections were in the abdominal tract, bones and joints, the respiratory tract, and in skin/soft tissue, however, oropharyngeal infection were rarely identified [9]. Our case series and literature review indicated that it is necessary to note the GIT, oropharyngeal tract, pulmonary, vertebral spine, intra-abdominal region, and heart valves as infectious sites of P. micra bacteremia.

In our cases, 48% (12/25) of P. micra bacteremia had polymicrobial bacteremia. In the literature cases, 5 patients (18.5%) had polymicrobial bacteremia. Most concomitant pathogens were aerobic and anaerobic members of the oral and gastrointestinal microbiota. Badri et al. showed that 50% of Parvimonas spp. associated bacteremic infections were polymicrobial [9], which was equivalent to our cases. Although there was no difference in the mortality rate between patients with monomicrobial and polymicrobial anaerobic bacteremia [33], it was reported that inappropriate antibiotic selection without attention to the results of anaerobic cultures have serious consequences for patients [34]. Therefore, we should choose the antibiotics targeting identified anaerobic organisms including P. micra.

Most cases of P. micra bacteremia in our case series were successfully treated with antibiotics (mortality, 1/25, 4.0%) (Table 2). Median of Pitt bacteremia score was 1, and two of 25 cases were treated in the intensive care unit. In the literature cases, P. micra bacteremia also had a favorable outcome (mortality, 1/27, 3.7%). In one of the previous cases, a patient died as a consequence of the primary tumor after the completion of antibiotic therapy [21]. In a previous study, in-hospital mortality of GPAC bacteremia including Parvimonas spp. was 10% [9]. In contrast, the overall mortality was as high as 25% in all patients with anaerobic bacteremia [34]. Although P. micra can cause complicated infectious diseases including IE and spondylodiscitis, these results suggest that patients with P. micra bacteremia might have good prognosis following appropriate treatment.

There were some limitations to this study. First, we may have missed some cases at our institute as well as previously reported cases due to the retrospective design. Second, the cases in our case series were diagnosed in a single tertiary medical center in Japan; therefore, regional and institutional differences were not considered. Third, different blood culture bottles were used in the two study periods. Although a prior study found no difference in the detection rate of anaerobic bacteremia using Bact/Alert FN bottles or Bact/Alert FN plus bottles [35], we cannot rule out the potential impact of our transition from Bact/Alert FN to Bact/Alert FA plus bottles on identification. Lastly, the literature review on P. micra bacteremia was dependent on publication biases, making it impossible to address the real prevalence and incidence of P. micra bacteremia in clinical setting.

We reported the causative pathogens of GPAs bacteremia identified by MALDI-TOF MS, and findings showed P. micra as the second most frequently identified GPA. Our case series and literature review showed that P. micra have been mainly identified in blood culture using MALDI-TOF MS and 16 s rRNA sequencing. Infection sites of P. micra were predominantly associated with GIT, oropharyngeal, vertebral spine, intra-abdominal region, pulmonary, and heart valves. Patients with P. micra bacteremia could have good prognosis following appropriate treatment. This study helps improve our understanding of the clinical characteristics of bloodstream infections by P. micra.