Elizabethkingia anophelis Infection in Infants, Cambodia, 2012–2018

We describe 6 clinical isolates of Elizabethkingia anophelis from a pediatric referral hospital in Cambodia, along with 1 isolate reported from Thailand. Improving diagnostic microbiological methods in resource-limited settings will increase the frequency of reporting for this pathogen. Consensus on therapeutic options is needed, especially for resource-limited settings.

45 × 10 9 /L; C-reactive protein 195 mg/L; and total bilirubin 252 μmol/L. Lumbar puncture was omitted due to thrombocytopenia. A blood culture was transferred with her from the local hospital.
The day after transfer, she experienced symptoms of meningitis, including fever and seizures. We initiated anticonvulsant therapy and changed her antimicrobial therapy to intravenous meropenem (40 mg/kg 3×/d). Blood culture microscopy subsequently showed gram-negative bacilli, identified as E. anophelis on hospitalization day 3 by matrixassisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry using bioMérieux VITEK MS in an in vitro diagnostic mode using the spectrum knowledge base version 3.2.0 (bioMérieux, https://www.biomerieux.com). At this stage antimicrobial drugs were changed to intravenous ciprofloxacin (10 mg/kg, 2×/d) and vancomycin (15 mg/kg, 1×/d); a blood culture collected before the change confirmed bacteremia caused by E. anophelis.
The patient was extubated on day 6 and underwent lumbar puncture because her platelet count had improved. Cerebrospinal fluid was cloudy, with a leukocyte count of 265 cells/µL (75% polymorphs), glucose of 1 mmol/L, and protein of 13 g/L. Gram stain microscopy revealed no organisms, and culture was negative. After 28 days of ciprofloxacin/vancomycin, she was clinically well and discharged home.
At her 1-month follow-up appointment, she displayed clinical features of raised intracranial pressure, including neurologic deficits. Cranial ultrasound showed hydrocephalus, a suspected sequela of meningitis, and she was referred for neurosurgical opinion.
After the case described was identified, we retrieved all isolates in −80°C storage that had been identified since January 2012 as Chryseobacterium meningosepticum, C. miricola, or Elizabethkingia spp. We included in our study the first isolates from a given clinical episode: 4 identified as C. meningosepticum, 3 as E. meningoseptica, and the isolate already identified as E. anophelis. From subculture, we analyzed these using VITEK MS MALDI-TOF mass spectrometry. We identified  We describe 6 clinical isolates of Elizabethkingia anophelis from a pediatric referral hospital in Cambodia, along with 1 isolate reported from Thailand. Improving diagnostic microbiological methods in resource-limited settings will increase the frequency of reporting for this pathogen. Consensus on therapeutic options is needed, especially for resource-limited settings.
To provide regional context for these results, 2 microbiology laboratories in Mae Sot, Thailand, and Vientiane, Laos, also reanalyzed stored clinical isolates as we described. In Mae Sot, a single isolate of E. meningoseptica from a neonatal blood culture was reidentified as E. anophelis. In Vientiane, 9 isolates of C. meningosepticum were reidentified as E. meningoseptica, and the identity of 1 E. meningoseptica isolate remained the same.

Conclusions
Although reports of E. anophelis are rare, cases are reported from countries in southern Asia, including Singapore (3), Taiwan (5), and Hong Kong (6). Our findings are consistent with reports of E. anophelis infection from other countries demonstrating it to be an opportunistic organism affecting more vulnerable patient groups (6). The mortality rate associated with E. anophelis is high (50%), and isolation of E. anophelis from blood in two thirds of the children in this study demonstrates its importance as a human pathogen.
Previous reports of community-and hospital-acquired E. anophelis infection among infants have proposed a range of transmission routes, including vectorborne (Anopheles mosquitoes) (1,2,7), waterborne (8), and vertical transmission (9). With no temporal clustering, and with most cases occurring among older infants, we suspect that unidentified environmental reservoirs are possible sources of these cases.
Previously, studies relied on 16S rRNA testing to identify E. anophelis, with biochemical phenotypic methods unable to distinguish between Elizabethkingia spp. (10). Although this method provides high discriminatory power, its use in diagnostic microbiology is limited to established laboratory settings. It also requires highly trained staff to interpret results, which are rarely available within a clinically useful timeframe. Until late 2017, oxidase-positive gram-negative isolates were identified at the microbiology laboratory at Angkor Hospital for Children by biochemical phenotypic methods (API 20NE, bioMérieux); identification is now done by MAL-DI-TOF mass spectrometry. Misidentification of Elizabethkingia spp. using biochemical methods has been reported (2,6); however, updated MALDI-TOF databases provide reliable differentiation (10). As the resolution that MALDI-TOF mass spectrometry provides in pathogen identification expands, and its use becomes available in low-and middle-income countries, we expect to see higher reported incidence of E. anophelis infection. Conversely, it may become apparent that the burden of E. meningoseptica is not as high as previously thought, with retrospective studies already showing E. anophelis as the predominant species of its genus (6,10,11). In our study, this possibility was not found to be the case in Laos, suggesting possible regional variation. E. anophelis demonstrates phenotypic and genotypic resistance to multiple antimicrobial drugs, and, without epidemiologically based interpretive cutoffs, selection of therapeutic options is challenging (4,5,10,12). High MICs to ceftriaxone are consistent with β-lactam resistance reported elsewhere, and carbapenem resistance should also be expected (4,5,10,12). Following Clinical and Laboratory Standards Institute guidelines (M100-29; 2019) (13) for "other non-Enterobacteriaceae," these isolates were susceptible to ciprofloxacin and sulfamethoxazole/trimethoprim. This finding is not consistent with other regional data that show greater rates of resistance to these drugs (5,10). E. anophelis has been shown to be susceptible to piperacillin/tazobactam and to rifampin (4,10), which were not tested against in this study and are not currently available as therapeutic options in the study setting. It is unusual for gram-negative organisms to exhibit susceptibility to vancomycin, and interpretation of MICs to this drug should be approached with caution. Use of Etest in this study was a methodological limitation; the preferred method of broth microdilution was not available.
In summary, updates of mass spectrometry platforms have enabled identification of clinical E. anophelis isolates in Cambodia and Thailand. As diagnostic microbiology capacity expands in low-and middleincome countries, further reports of this organism are expected. Because of the associated high mortality rates for this pathogen, consensus on therapeutic options for infection caused by E. anophelis is needed, especially in resource-limited settings with restricted choices for antimicrobial drugs.