Emerging Problems in Infectious Diseases Microbiological characteristics of hypermucoviscous Klebsiella pneumoniae isolates from different body fluids

Introduction: Reports of hypermucoviscous Klebsiella pneumoniae (hvKP) isolated from fluids other than blood or abscess are rare. The aim of the study was to compare clinical and microbiological characteristics of hvKP found in blood or abscess fluid with those isolated from other loci. Methodology: A total of 24 non-repetitive hvKP isolates were collected from January 2013 to June 2014 from patients with hvKP infections. There were 15 in Group 1 (fluid other than blood or abscess) and 9 in Group 2 (blood or abscess fluid). Medical records of all patients were reviewed. Capsular polysaccharide (CPS) typing, virulence factor determination, and multilocus sequence typing (MLST) of hvKP isolates were performed. Results: Seventeen sequence types (STs) and 6 capsular serotypes were identified. Type K2 was most commonly identified in Group 1 and type K2 in Group 2. Deletion of pLVPK-derived loci were found in K2 and non-K1/K2 hvKP strains. Two virulent genes, fimH and ycfM, were identified more frequently in Group 2 than in Group 1. There was no difference in the frequency of other virulent genes or serotypes in the two groups. Two imipenem resistant hvKP isolates (cr-hvKP) were found in non-blood or abscess samples. Conclusions: hvKP isolated from different body fluids had similar clinical and microbiological characteristics. cr-hvKP identified in non-blood or abscess samples should raise our attention to the challenging situation and management of hvKP infection.

The hypermucoviscous (hv) phenotype is a distinguishing factor of hvKP strains. hvKP is considered to be more virulent than hv-negative strains and this phenotype is commonly found in K. pneumoniae strains which cause CA-PLA and BSI [8,12,13]. More importantly, they exhibit a higher tendency to metastatic spread. These two traits contribute to the high mortality rate associated with hvKP infection.
hvKP is more resistant to phagocytosis by polymorphonucleated neutrophils, less sensitive to killing by serum complement, and more virulent in animal studies [14][15][16][17][18]. Previous reports have focused on hvKP infections isolated from abscess and blood fluids. We isolated hvKP from urine and tracheal secretions, evaluated their clinical and microbiological characteristics, and compared them to those of isolates obtained from blood and abscess fluid.

Bacterial isolates
All isolates were obtained from patients seen at Peking University Third Hospital (PUTH), a universityaffiliated medical center with a 1,498-bed capacity and 79,000 hospital admissions per year. A total of 533 non-repetitive K. pneumoniae clinical isolates were identified by VITEK GN card (bioMe'rieux, Marcy l'Etoile, France) from January 2013 to June 2014. All samples were stored at −80°C prior to genetic and virulence testing. An hv phenotype test was performed on all isolates to distinguish hvKP from classic K. pneumoniae (cKP) [12]. The presence of a viscous string greater than 5 mm in length with a bacteriology inoculation loop on bacterial cultures grown overnight on 5% sheep blood agar plate at 37°C was regarded as diagnostic of hvKP.
The first hvKP isolate from each patient was used for further investigation. hvKP isolated from asymptomatic patients was classified as colonization and was not included in this study. Infection was considered to be nosocomial if it was diagnosed ≥48 h after admission and the patient had no evidence of clinical infection at the time of admission. Symptomatic infections diagnosed within 48h of admission were considered to be community acquired [19,20]. hvKP strains not isolated from blood or abscess fluid were Group 1. hvKP strains isolated from blood or abscess fluid were Group 2.
The BLAST program at http://www.ncbi.nlm.nih.gov was used for final serotype identification. MLST was performed according the protocols for K. pneumoniae provided on the MLST website (http://bigsdb.web.pasteur.fr/klebsiella/klebsiella.html) . MLST allelic profiles were characterized using the MLST database. Sequence types (STs) were analyzed using eBURST version 3.0. Clonal complexes (CCs) were defined as groups of two or more isolates sharing at least 6 identical alleles.

Statistical analysis
Data were analyzed using the statistical package SPSS 17.0 for Windows. The Fisher's exact test or Chisquare test were used for categorical variables. All statistical tests were 2 tailed and a p value ≤0.05 was considered statistically significant. Descriptive data were reported as mean ± SD.

Ethical approval
The study was approved by the Institutional Review Board of PUTH.
Men older than 60 years were predominantly affected with community acquired infections. Patients were generally toxic on admission and frequently required invasive monitoring or care. Over 80% of patients recovered and were discharged. The clinical characteristics of each hvKP group were similar ( Table  2).

Discussion
All hvKP isolates obtained from a large urban hospital during two years were evaluated in our laboratory. The infrequency of this infection led to the identification of only 24 hvKP isolates from 24 patients during a 2-year period. hvKP isolates comprised 7.5% of all KP isolates from our hospital, 60% of which were associated with infection. We identified hvKP more frequently in urine and abscess, and less frequently in blood, tracheal secretion, and bile than another similar large hospital in China [26]. Most community-acquired infections were blood infections or a presence of different abscess, similar to previous reports [12,16,27]. Out of all, 40% (6/15) hvKP infections in nonblood/abscess group were defined as hospital-acquired versus 11.1% (1/9) in blood/abscess group. This may suggest an elevated risk for hospitalized patients and the potential dissemination of hvKP strains in health care facilities.
Colonization with hvKP is thought to be a first step in developing infection [28]. Patients with pyogenic liver abscesses have been reported to have frequent intestinal (81.4%) and/or pharyngeal (39.5%) colonization with hvKP [29]. We found hvKP colonization in 16 patients. Colonized fluids included tracheal secretions, urine and superficial secretions. Factors leading from colonization to infection are not well understood [8]. The presence of hvKP in the urine may be a potential marker for bacteremia [18]. Therefore, at the time of hvKP isolation, strict adherence to standard hospital infection control precautions should be reinforced to limit its spread.
The hvKP that we evaluated had similar antibiotic resistance patterns as previous reports. An exception to this finding was the presence of carbapenem resistance identified in hvKP isolated from tracheal secretions and urine of patients in Group 1 [12,26]. These two isolates were characterized as K2/ST25 and K2/ST65 with blaKPC-2, respectively. The study from China revealed 5 K1 cr-hvKP isolates with plasmid-borne blaKPC-2 gene and genetic relation between 3 of them [30]. The development of resistant strains of hvKP demonstrates a need for avoiding unnecessary antibiotic use that stimulates the generation of resistant strains. This problem is emphasized by a cr-hvKP isolate obtained from an elderly man receiving chemotherapy who was previously treated because of another systemic infection before developing a fatal cr-hvKP infection [31]. The bacterial capsule is an important virulence factor and serotypes K1 and K2 have been particularly linked to severe bacteraemia and liver abscess [32,33]. These findings have rarely been reported in fluids other than blood and /abscess. In the present study, no difference was found in the distribution of capsular serotypes of blood/abscess and non-blood/abscess groups.
K. pneumoniae virulence determinant pLVPK is a 219,385-bp plasmid isolated from the invasive K2 strain CG43. pLVPK-derived terW-rmpA-iutA-silS loci are independent pathogenicity factors for abscess formation [25]. pLVPK derivatives could be extrachromosomal and carry the repA gene, or be a chromosome-integrated form [25,34]. Gene deficiency of pLVPK derivatives was found for the first time in the hvKP isolates except K1 type. entB, iroN, ybtS, kfuBC, and iutA are part of iron scavenging systems that contribute to bacterial virulence [18]. Similar expression patterns were found in both groups. Both, fimH, which encodes type 1 fimbriae adhesion and ycfM, which encodes an outer membrane lipoprotein, were more common in group 2 and could also contribute to bacterial virulence.
There were several limitations to this study. The small number of hvKP isolates obtained does not support subgroup analyses. Not all patients were tested for colonization at admission and there was limited testing of colonization sites. The use of 48 hours as a cut-off for diagnosing nosocomial infections was arbitrary and could led to an overestimation of nosocomial infections. However, the frequency of community acquired hvKP infections in our patients was similar to that reported from other centers [26].

Conclusions
This study is unique in comparing the clinical and microbiological characteristics of hvKP isolates obtained from blood and abscess to those obtained from other fluids. Bacteria isolated from all areas had similar clinical characteristics and virulence characteristics. Antibiotic resistance is a developing problem with hvKP.