Introduction

Pantoea spp. are Gram-negative bacillus that are non-encapsulated and non-spore-forming. Pantoea became its own species in 1989, after being separated from Enterobacteriaceae [1], and is now placed within the Erwiniaceae family. Currently, at least 20 different subspecies have been identified. Pantoea spp. can be found as a colonizer in humans, plants, and the environment, such as water and soil. Its adaptability has led to promising applications in commercial and agricultural sectors, including biotechnology, environmental protection, soil bioremediation, and plant growth stimulation [2]. However, while infrequent, Pantoea species can be pathogenic to humans.

Previous case reports involving Pantoea spp. have included both immunocompetent and immunocompromised patients. Though described as a species with low virulence, fatal incidents have been seen in infants, postoperative patients, or those who are immunocompromised [3, 4]. Of the cases being reported, infections typically occur via traumatic, sinopulmonary, hepatobiliary, or iatrogenic routes and can lead to bacteremia. In this report, we present two cases of Pantoea peritonitis, recurrent peritonitis due to P. dispersa, and refractory peritonitis due to P. agglomerans, each of which developed in a different patient with end-stage kidney disease (ESKD) receiving peritoneal dialysis (PD).

Case presentation

Case 1

A 66-year-old male patient with a past medical history of ESKD, hypertension, type 2 diabetes, and gout was admitted with a 2-day history of worsening diffuse abdominal pain and cloudy PD effluent. Additional symptoms included chills, nausea, emesis, diarrhea, and worsening pedal edema. He had been treated with PD for 4 years and was on continuous cycling PD with two daytime exchanges at the time of presentation. He was a low-average transporter. Vital signs were within normal limits. Physical exam was notable for a tender abdomen to all quadrants. There was no rebound or guarding. His PD catheter exit site did not have erythema, warmth, drainage, or tenderness. Laboratory tests showed mild peripheral leukocytosis (11.9 × 103/mm3: reference, 4.0–10.0 × 103/mm3), but otherwise compatible with the history of ESKD. Analysis of PD effluent revealed an elevated white blood cell (WBC) count (9749 cells/mm3) with neutrophil predominance (98%). Vancomycin and cefepime were empirically initiated due to the high local prevalence of methicillin-resistant Staphylococcus aureus and the susceptibility patterns of Gram-negative bacteria in the region. The loading doses of vancomycin 15 mg/kg and cefepime 1 g were given intravenously, followed by continuous intraperitoneal (IP) doses (i.e., 25 mg/L and 125 mg/L, respectively). The patient reported significant improvement in abdominal pain by day 3 when PD fluid culture grew P. dispersa, identified by Vitek® 2 MS (bioMérieux Co., Nürtingen, Germany), an automated mass spectrometry microbial identification system that uses matrix assisted laser desorption ionization time-of-flight (MALDI-TOF). Susceptibility testing showed low minimum inhibitory concentrations (MIC) to all tested antibiotics including aminoglycosides, trimethoprim/sulfamethoxazole, fluoroquinolones, and cephalosporins except for cefoxitin, which was resistant (Table 1). The antibiotic regimen was changed to continuous IP ceftazidime monotherapy using a concentration of 125 mg/L. A repeat PD fluid analysis on day 4 confirmed a good response to antibiotic therapy (WBC 113 cells/mm3). The patient was discharged and completed a 3-week outpatient antibiotic course of intermittent IP ceftriaxone 1 g/day as outpatient. Another PD fluid analysis was done on the last day of the 3-week antibiotic course and showed a WBC count of 38 cells/mm3. Blood culture remained negative throughout the course (Fig. 1).

Table 1 Susceptibilities of isolated Pantoea species in the two cases
Full size table
Fig. 1
figure 1

Timeline of clinical events from the initial admission for Pantoea dispersa PD-associated peritonitis. PD, peritoneal dialysis; temp, body temperature; IP, intraperitoneal; IV, intravenous

Full size image

However, 12 days after he completed the antibiotics, he required re-hospitalization due to severe, acute-onset abdominal pain associated with nausea and vomiting. He was febrile to 38.0 ℃ and had diffuse, moderate abdominal tenderness. He did not have leukocytosis (WBC 7.8 × 103 cells/mm3). PD effluent showed an elevated WBC count (15,967 cells/mm3) with neutrophil predominance (97%), and the patient was started on the same empirical IP antibiotic regimen with the initial episode. PD fluid culture grew P. dispersa again, confirming recurrent peritonitis. Susceptibility testing showed the same levels of MIC to all the tested antibiotics. The patient again showed good clinical improvement, but the decision was made to remove the patient’s PD catheter given the recurrent nature of peritonitis. The patient underwent PD catheter removal and started intermittent hemodialysis via a tunneled dialysis catheter. He was discharged home and completed a 2-week course of IV ceftazidime 2 g three times a week at his dialysis clinic. He remained stable for 2 weeks after the completion and was cleared for PD catheter placement. However, he made the decision to continue receiving in-center hemodialysis and declined the offer to resume PD.

Case 2

A 42-year-old male patient with a past medical history of ESKD and hypertension was admitted after 6 days of bloody PD fluids without any recent trauma or systemic illness. He had been treated with PD for 2.5 years and was on continuous cycling PD without a last fill at the time of presentation. He was a high-average transporter. He reported no symptoms of abdominal pain, fever, chills, nausea, vomiting, or diarrhea and confirmed he was not on any anticoagulants. At admission, his vitals were largely normal except for a mildly elevated blood pressure at 174/88 mmHg. Computed tomography (CT) imaging showed no significant abnormalities. There was no peripheral leukocytosis (WBC 9.0 × 103 cells/mm; 67% neutrophils), and the whole blood RBC count was 3.22 × 106 cells/mm3. However, a repeat PD fluid analysis showed worsening cell counts with a WBC count of 2235 cells/mm3 (43% neutrophils) and an RBC count of 473,922 cells/mm3, compared with the results from 2 days prior, which showed a WBC count of 11 cells/mm3 (74% neutrophils) and an RBC count of 2556 cells/mm3. Although small hemoperitoneum might have influenced the PD effluent WBC count to some degree, the patient was started on the empirical vancomycin and cefepime for presumed PD-associated peritonitis using the same regimen with case 1. On day 2, PD fluid cultures identified Streptococcus mitis and Pantoea agglomerans, which showed good susceptibility to all tested antibiotics except for penicillin G (intermediate) and cefoxitin (intermediate), respectively, and intraperitoneal cefepime was continued. However, repeat peritoneal dialysis fluid analysis on day 5 indicated a suboptimal response to the antibiotic regimen, showing a WBC count of 473 cells/mm3 with 17% neutrophils. Although a longer observation for antibiotic effect was an option, patient decided to remove the PD catheter and switch his dialysis modality to hemodialysis after discussing potential advantages and disadvantages of each approach. Post-catheter removal, repeat PD fluid culture obtained on day 5 turned out to be negative, and the patient was treated with intravenous ceftriaxone 2 g three times a week at his dialysis clinic for a total of 3 weeks.

Literature research

The literature search was conducted using PubMed, covering the period from January 1989 to January 2024. We used a combination of keywords “Pantoea” and “peritonitis” to identify relevant articles. Additionally, another combination of “Pantoea dispersa” and “infection” was used in a separate search to ensure comprehensive coverage of the topic.

Discussion

We described two unique cases of Pantoea peritonitis in PD, recurrent P. dispersa peritonitis and refractory P. agglomerans peritonitis, both of which resulted in PD catheter removal despite the standard duration of susceptible antibiotic treatment. The route of Pantoea spp. peritonitis remained unclear in both cases despite our best efforts. Previous reports of P. agglomerans soft tissue and blood stream infections suggested ingestion of infected plants or thorn pricks as potential causes. However, both patients denied such events, engaging in gardening activities, or any touch contamination, but reported strict adherence to hygiene techniques. These cases underscore the importance of vigilant monitoring and the potential need for PD catheter removal in Pantoea peritonitis.

The most common isolated Pantoea species in the healthcare setting is P. agglomerans [4], but there is an increasing number of reports on infections attributable to P. dispersa, especially among infants, elderly, and immunocompromised patients [4,5,6,7,8,9]. Interestingly, P. dispersa infection may be more common in certain geographic areas or in rhinosinusitis. A multicenter study in Ethiopia evaluated 426 positive blood culture results confirmed by MALDI-TOF between 2019 and 2020 and reported that 21 samples (5%) were positive for P. dispersa [10]. Of those, 20 (95%, 18 from the neonatal ICU and 2 from the adult emergency department) were isolated at one specific hospital in the northern part of the country where P. dispersa was the second most frequent pathogenic bacteria (15%). Additionally, P. dispersa showed an overall multidrug resistance frequency of 81%. In Italy, however, the prevalence of Pantoea species accounted for only 19 of 4996 G-negative isolates from blood culture, of which only one isolate was identified as ampicillin-resistant P. dispersa [11]. Another case series report from a hospital in Taiwan evaluated 274 rhinosinusitis patients with positive sinus culture and found that P. dispersa was identified in a total of 36 patients (13%) [12]. Patients with P. dispersa, compared with those with other pathogens, received a significantly shorter duration of antibiotic treatment and had lower surgery rate without significant difference in clinical outcomes. No data were available regarding the prevalence of P. dispersa infection among other infections at their institution or in other geographic areas of Taiwan.

Otherwise, there have been only ten reported cases of P. dispersa infection, as summarized in Table 2. Of those, two cases were neonates and two adults were immunocompromised due to acute myeloid leukemia and cirrhosis, respectively, while the remaining were immunocompetent adults. Of eight cases reporting antibiotic sensitivities, five noted single or multidrug resistance. Commonly used antibiotic classes included beta-lactams, carbapenems, macrolides, aminoglycosides, and glycopeptides. Patients with P. dispersa infection were reported to have favorable outcomes [6], but three of the ten patients died [3,4,5].

Table 2 Case summaries of Pantoea dispersa infection
Full size table

The management strategy for Pantoea spp. peritonitis is yet to be established. Literature review identified a total of 17 cases of PD-associated peritonitis caused by Pantoea spp. (Table 3) [13,14,15,16,17,18,19,20,21,22,23,24,25,26], with only one case being P. dispersa PD-associated peritonitis published as a conference abstract [26]. Of those, three had unfavorable outcomes despite the use of antibiotics on the basis of susceptibility; two cases required PD catheter removal because of the lack of clinical improvement [13, 24], and the other case with P. agglomerans died due to septic shock after refusing PD catheter removal [20]. The previous case of P. dispersa PD-associated peritonitis showed multidrug resistance but was successfully treated with IP tobramycin without relapse or complication [26]. Our case responded well to the targeted antibiotic therapy, including ceftazidime and ceftriaxone, and susceptibility testing revealed fairly susceptible organism. Therefore, resistance to antibiotics is unlikely the underlying cause of peritonitis recurrence. Instead, the formation of bacterial biofilm, as seen in the case of P. agglomerans [27, 28], could be a potential reason for the recurrence of PD-associated peritonitis by P. dispersa in our case. Microbiological evaluation of a removed PD catheter should be considered in future cases to examine this hypothesis.

Table 3 Case summaries of peritoneal dialysis-associated peritonitis with Pantoea species infection
Full size table

One of the pitfalls in identifying Pantoea species is the multiple types of analysis that this bacterium needs to undergo to be correctly identified. As done in our case, MALDI-TOF-based identification methods are commonly used in clinical laboratories as a rapid means of identification, but a previous study using cpn60-based molecular typing demonstrated that 24% of such isolates were misidentified and were actually strains of Citrobacter, Enterobacter, Kosakonia, Klebsiella, or Pseudocitrobacter, members of the newly described Erwinia gerundensis, and even several unclassified members of Enterobacteriaceae [29]. However, none of the centers at which 24% isolates were misidentified used the same MALDI-TOF system as our laboratory.

In conclusion, we presented two rare cases of PD-associated peritonitis caused by Pantoea spp. along with literature review. It is imperative to correctly identify this species along its susceptibility, as there have been documented cases of this organism being multidrug resistant. Additionally, further investigation is necessary to evaluate the risk of recurrence and to develop the best treatment strategy including the duration of antibiotic treatment and PD catheter salvage versus removal in the setting of Pantoea peritonitis.