A 76-year-old Danish male presented to the emergency department at Aalborg University Hospital (Aalborg, Denmark) in May 2017 with right-side otalgia throughout the previous week, and onset of fever and confusion within the previous 24 h. Upon admission, the patient was otherwise healthy, and he used no daily medication. Fifty-three days prior to admission, the patient returned from a 16 day holiday on the east coast of Peninsular Malaysia. Before the journey, he was re-vaccinated against diphtheria, tetanus and hepatitis A. He had not used malaria prophylaxis during his vacation. On examination, the patient had altered mental status with a Glasgow Coma Score of six, neck stiffness and fever (40.0 °C). In accord with national guidelines on the management of suspected bacterial meningitis, the patient had blood cultures performed and was started on high-dose intravenous (iv) benzylpenicillin (1.8 g every 4 h), cefotaxime (3 g every 6 h) and dexamethasone (10 mg every 6 h). A lumbar puncture of the patient was performed, following a computed tomographic scan of the cerebrum, which had proved normal. Laboratory tests showed a C-reactive-protein level of 273 mg l−1, procalcitonin of 10.8 µg l−1 and white blood cells of 19.9×109 l−1. Cerebrospinal fluid (CSF) analysis revealed pleocytosis with white blood cells of 741×106 l−1 (636 polynuclear and 105 mononuclear), a slightly decreased glucose ratio (CSF: serum) of 0.38, an elevated protein level of 1.34 g l−1 and lactate of 7.3 mmol l−1, and the patient was transferred to the Intensive Care Unit (ICU). Overnight culture of the patient’s CSF yielded Gram-negative, non-motile, oxidase- and catalase-positive rods. Moreover, a simultaneous positive blood culture (BD BACTEC; Becton Dickinson) had similar findings. MALDI-TOF MS (matrix-associated laser desorption ionization-time of flight MS) (MALDI Biotyper 3.1, Bruker Daltonics Microflex LT, MBT 6903 MSP Library) could not distinguish the colonies between either E. meningoseptica (score 2.215) or E. miricola (score 2.101), whereas E. anophelis was not present in the MALDI library. API 20 E v5.0 (bioMérieux) gave an identification of E. meningoseptica with a score of 71.6 % identity (numerical profile: 0042004). The isolate was multidrug resistant and positive for metallo-β-lactamase by using the MBL Confirm kit (Rosco Diagnostica). Antimicrobial-susceptibility testing (AST) was performed by using Etests (bioMérieux) according to the European Committee on Antimicrobial Susceptibility Testing guidelines (). We used the pharmacokinetics/pharmacodynamics (non-species related) breakpoints except for trimethoprim/sulfamethoxazole, in which Stenotrophomonas maltophilia breakpoints were used, and for gentamicin and colistin we used Pseudomonas species breakpoints, as performed by Eriksen et al. []. The isolate was susceptible to moxifloxacin (MIC 0.125 mg l−1) and trimethoprim/sulfamethoxazole (MIC 0.25 mg l−1); intermediately susceptible to amoxicillin/clavulanic acid (MIC 6 mg l−1); and resistant to ciprofloxacin (MIC 0.75 mg l−1) and all other drugs tested including: ampicillin, cefuroxime, ceftazidime, meropenem, gentamicin, colistin and tigecycline. An Etest for vancomycin showed a MIC of 12 mg l−1, but we did not make an interpretation of susceptible or resistant. Etests were not available for piperacillin/tazobactam and rifampicin, but the disc diffusion zone (Neo-Sensitabs; Rosco Diagnostica) for piperacillin/tazobactam was 19 mm and for rifampicin was 24 mm, but as for vancomycin we did not make an interpretation of susceptible or resistant. To determine the identity of the isolate to the species level, we performed sequencing with the Illumina MiSeq instrument producing 2×300 bp paired-end reads by using a Nextera XT library preparation kit (Illumina). Reads were assembled using CLC Genomics Workbench (version 11) (QIAGEN Bioinformatics) into 105 contigs, N50 (497, 160), total sequence length 4 047 726 bp, with a G+C content of 35.6mol %. Analysis of the 16S rRNA gene, as well as a k-mer based distance measure, compared to the publicly available strains of E. meningoseptica and E. miricola showed a clear identification of the isolate as E. anophelis (data not shown). Breurec et al. [] and Perrin et al. [] reported a clear division of E. anophelis into 15 sublineages, including 1 associated with the large outbreak of E. anophelis infections that occurred in Wisconsin (USA) in 2015–2016. To subtype our strain and define its sublineage, we used the core-genome multilocus sequence typing (cgMLST) strategy, using the subset of 1546 genes families that are highly conserved within E. anopheles []. The cgMLST profile of our isolate was compared to those publicly available in the Elizabethkingia cgMLST database on the Institut Pasteur server (). Cluster analysis based on cgMLST profiles, see, showed that the isolate belonged to sublineage 11, which was defined with strain CIP 60–59 (CDC 3375; ATCC 13255; NCTC 10586; CCUG 4321; LMG 12873) as a reference. Strain CIP 60–59 was isolated from the CSF of a premature infant who died []; however, the two strains of sublineage 11 have different alleles at 299 loci out of 1513 loci called in both strains. This result clearly shows that AAUH 98722 (our isolate) and CIP 60–59 are genetically distinct. The assemblies were submitted to ResFinder 3.0 () to analyse them for the presence of antimicrobial-resistance genes []. The isolate proved positive for two metallo-β-lactamase genes, namely blaGOB-3 (located on contig 69; hit length 756; 100 % ID; accession no. AF189291), and blaB-3 (located on contig 21; hit length 750; 100 % ID; accession no. AF189299), as well as an extended-spectrum β-lactamase-encoding gene, blaCME-1, (located on contig 41; hit length 784; 100 % ID; accession no. AJ006275). This is consistent with the known conservation of carbapenemase and β-lactamase-encoding genes within E. anophelis []. After the preliminary identification of Elizabethkingia species, the antimicrobial therapy was changed to iv vancomycin combined with iv rifampicin, 600 mg two times a day. However, as MICs values were available the day after, the definite treatment was changed to iv moxifloxacin, 400 mg once per day, combined with iv rifampicin 600 mg two times a day for a total duration of 14 days. The patient improved and after 10 days at the ICU he was transferred to the infectious disease ward for further treatment and rehabilitation, and he was finally discharged after 3 weeks of hospitalization. The only sequela was a partial hearing loss. Due to the severity of infection, the patient was examined for immunodeficiency in an outpatient setting and was found to have a sustained and elevated level of IgM of approximately 14 g l−1 (normal range 0.39–2.1 g l−1) during the following months. Seven months after the meningitis episode, his bone marrow was further investigated, and he was finally diagnosed with lymphoplasmacytic lymphoma (Waldenström macroglobulinaemia).