Central Line-Associated Bloodstream Infection Due to Elizabethkingia anophelis: Case Report and Literature Review on Pediatric Infections

1. Introduction

Elizabethkingia species are aerobic, glucose-nonfermenting, catalase-positive, oxidase-positive, and indole-positive Gram-negative bacilli widely distributed in natural environments such as soil, water, and plants, as well as in healthcare settings [1]. The genus Elizabethkingia comprises 6 species, namely, E. meningoseptica, E. anophelis, E. miricola, E. bruuniana, E. ursingii, and E. occulta [1]. E. anophelis is an opportunistic pathogen most commonly affecting infants or critically ill adults with underlying comorbidities [1,2]. It is particularly known to cause neonatal sepsis and meningitis especially in premature newborns and sometimes is involved in outbreaks of life threatening infections, with mortality rates ranging from 24% to 60% [1-5].

Herein, we describe a case of central line-associated bloodstream infection (CLABSI) due to E. anophelis in a 2.5-year-old girl with acute lymphoblastic leukemia and review the literature on pediatric cases caused by E. anophelis.

2. Case Description

A 2.5-year-old girl was diagnosed with acute lymphoblastic leukemia of B lineage (B-ALL). The full blood count (FBC) at diagnosis was WBC: 2,300/mm³, Hb: 6.9g/dL and PLT: 16,000/ mm³. The myelogram showed full infiltration by lymphoblasts and the immonophenotyping revealed common pre-B ALL (EGIL classification). A central venous catheter (CVC) Hickman type was inserted and then commenced on intensive chemotherapy according to ALL IC-BFM 2009 Protocol. Due to prognostic factors and treatment response she was classified to receive treatment of intermediate risk group.

In a febrile neutropenia episode three months post starting intensive chemotherapy, Streptococcus mitis was isolated from blood cultures taken from CVC. Based on the results of the susceptibility testing the patient was given teicoplanin as a loading dose at 10mg/kg every12 hours intravenously for three doses, followed by the maintainance dose of 10mg/kg once daily, along with teicoplanin lock therapy. The CVC was kept in place. She was started again on chemotherapy according to her protocol.

Seven months after diagnosis and a month before ending intensive protocol when she was receiving cytarabine 70mg/m²/d and thioguanine 60mg/kg/m² she became febrile. Blood, urine, stool, and pharyngeal cultures were taken and she was started on empirical treatment with intravenous piperacillin/tazobactam at a dosage of 300mg/kg every 6 hours. Full blood count was WBC: 200/ mm³ with absolute neutrophil count (ANC): 0/μl, Hb: 8,6g/dl, PLT: 38000/μl, and CRP: 4.5mg/dl (normal value <0.5mg/dl). Chemotherapy was stopped during the episode. Blood specimens taken from the CVC the first day of the febrile episode were inoculated into BacT/Alert PF bottles and incubated in a BacT/Alert 3D blood culture system (BioMérieux, Marcy L’ Etoile, France). After 17.2 hours the blood cultures turned positive for a Gram-negative microorganism, Elizabethkingia anophelis as identified by Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (VITEK MS system, BioMérieux; version 3.2). The blood cultures continued to be positive for E. anophelis, and the patient remained febrile for 3days.

The in vitro susceptibility testing performed by E-test revealed that E. anophelis was susceptible to piperacillin/tazobctam, amikacin, minocycline, doxycycline, ciprofloxacin, levofloxacin, trimethoprim-sulfamethoxazole and tigecycline, intermediate to vancomycin, and resistant to piperacillin, ceftriaxone, ceftazidime, imipenem, meropenem, ceftazidime/avibactam, meropenem/ vaborbactam, imipenem/relebactam, eravacycline, plazomicin and tetracycline (Table 1).

Based on the profile of the antibiogram, amikacin 20mg/kg every 24 hours was added. Four days after the start of the episode the patient became afebrile and blood cultures became negative. She continued on antibiotics for a total of 10 days. The CVC was not removed. She is now on maintenance treatment with chemotherapeutical agents per os.

The source of the infection remained undetermined because the microorganism was not isolated from any of the environmental samples (water supplies, surfaces and medical equipment).

3. Discussion

E. anophelis was initially isolated from the midgut of the Anopheles gambiae mosquito in 2011 [6]. The first reported clinical case of E. anophelis infection was meningitis in an 8-day-old girl in the Central African Republic. E. anophelis was identified by 16S-rRNA sequencing [7]. Since this initial report sporadic cases of serious systemic infections in infants and adults and several outbreaks of E. anophelis have been reported in Asia, and the USA. The largest outbreak was registered in the Midwestern United States, resulting in 20 deaths among 65 infected patients [3]. To date, in Europe only 2 adult cases and one outbreak of E. anophelis have been described [4,8,9]. In many previous studies it has been revealed that the incidence of E. anophelis infections was highly underestimated due to misidentification as E. meningoseptica based on phenotypes and prior MALDI-TOF systems not including E. anophelis in their diagnostic databases [1,5]. MALDI-TOF with updated databases and molecular methods such as 16S rRNA sequencing and whole-genome sequencing (WGS) are reliable and accurate in species identification. Our isolate was identified by MALDI-TOF MS (v. 3.2) containing in its database three species of the genus Elizabethkingia, namely E. meningoseptica, E. anophelis and E. miricola.

In a Medline/Pubmed searching the keywords ‘‘Elizabethkingia anophelis pediatric infections’’we found only 21 previously reported cases [5,7,10-19]. Our case is the first pediatric E. anophelis infection described in Europe. Table 2 summarizes the patients’ characteristics (gender, age, country of origin), the clinical manifestation, underlying medical conditions, the type of specimen cultured, antibiotic treatment and outcome.

The majority of cases (59.1%) involved newborns mostly premature. A slight female predominance was observed (1.2:1). Although E. anophelis is ubiquitous in nature with global distribution, most cases (81.8%) have been reported in Asian countries. Meningitis was the most common presentation in newborns. Other clinical manifestations included bloodstream and respiratory infections. The present case was a CLABSI. It has been demonstrated that E. anophelis has the ability to form biofilm that facilitate their establishment in CVCs, complicating treatment [20]. The source of the infection and the route of transmission remain unclear for all cases, except for one of vertical transmission from the mother who had chorioamnionitis to her neonate [11]. The majority of children had their immune system weakened by prematurity, by intensive medical interventions, or by other comorbities. The case fatality rate of the infected children was 33.3%, with deaths being most common among infected neonates. Five children among survivors of E. anophelis meningitis developed neurologic sequelae such as hyrocephalus and hearing loss [10,12,15,19].

E. anophelis has been known to be resistant to multiple antimicrobial agents, including most β-lactams, β-lactam/β-lactamase inhibitors, carbapenems, and polymyxins [1]. The majority of the reported cases were treated with vancomycin combined with other antibiotics, such as rifampicin, ciprofloxacin, trimethoprim-sulfamethoxazole or piperacillin/tazobactam. Our isolate was intermediate to vancomycin, and resistant to β-lactams, carbapenems, and the novel β-lactam/β-lactamase inhibitors such as ceftazidime/avibactam, imipenem/relebactam, and meropenem/vaborbactam because inhibitors have low activity against the metallo-β-lactamases produced by E. anophelis. In our case piperacillin/tazobactam was initially given as empiric therapy and was continued with the addition of amikacin after susceptibility data became available. Notably, four of the nine reported cases of E. anophelis meningitis with favorable outcome were treated with combinations including piperacillin/tazobactam [10,11,14]. Comparable results suggestive of susceptibility to piperacillin/tazobactam have been reported in studies of Han et al., Jian et al. and Perrin et al. [3,21,22]. However, further evaluation of in vivo data and continuous surveillance of antimicrobial resistance are required to make optimal therapeutic decisions. It has been shown that inappropriate empirical antimicrobial therapy is an independent risk factor for increased mortality in patients infected with E. anophelis.¹ Accurate identification is essential for selecting the appropriate antimicrobial therapy because of the varying susceptibility profiles among species.

4. Conclusion

The increasing number of cases of E. anophelis infections which is a result of availability of new, accurate identification methods, highlight the clinical significance of this opportunistic pathogen in pathogenesis of human infections. Early diagnosis, timely and accurate species identification and prompt effective treatment optimize outcomes.