Академический Документы
Профессиональный Документы
Культура Документы
PII: S2213-7165(17)30189-3
DOI: https://doi.org/10.1016/j.jgar.2017.10.002
Reference: JGAR 512
To appear in:
Please cite this article as: Yanning Ma, Chunmei Bao, Jie Liu, Xiuhong Hao, Jingui Cao,
Liyan Ye, Jiyong Yang, Microbiological Characterization of Klebsiella pneumoniae
isolates causing Bloodstream Infection from Five Tertiary Hospitals in Beijing, China
(2010), https://doi.org/10.1016/j.jgar.2017.10.002
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Microbiological Characterization of Klebsiella pneumoniae isolates causing
Yanning Ma1, Chunmei Bao2, Jie Liu3, Xiuhong Hao4, Jingui Cao5, Liyan Ye1 and Jiyong Yang1*
5. Department of Infection Control, Air Force General Hospital, PLA, Beijing, China 100142
*Corresponding author:
Email: yangjy301@hotmail.com
Mailing address: Microbiology Department, 301 Hospital, 28# Fuxing Road, Beijing 100853, China
Highlights
Diverse sequence types have been identified among BSI-causing K. pneumoniae.
ST11 and ST23 were the predominant clones.
The majority (70.4%, 143/203) of BSIKpn isolates were associated with
non-K1/K2/K5/K20/K54/K57 serotypes.
The rmpA was a predominant factor in determining whether or not a BSIKpn strain
had HV phenotype.
The majority of BSIKpn strains exhibited low resistance to commonly used
antimicrobial agents.
ABSTRACT
Background: Klebsiella pneumoniae exhibits prevalence in China. Little is known about the
(BSIs).
Methods: BSI-causing K. pneumoniae (BSIKpn) isolates were collected from five tertiary-care
hospitals in Beijing, China. Their genetic relatedness was analyzed by pulsed-field gel
electrophoresis (PFGE), antimicrobial susceptibility was determined by agar dilution method, and
sequence type (ST) was evaluated by multilocus sequence typing (MLST). The hypermucoviscosity
(HV) phenotype was identified by a positive string test. Carbapenemase, capsular serotypes and
Results: A total of 219 non-duplicated BSIKpn isolates were collected from Dec. 2013 to Dec. 2014
and categorized into 203 types (strains) which had a unique PFGE pattern. Among 203 BSIKpn
strains, 105 different STs were identified. Overall, ST11 and ST23 were the predominant clones and
detected from 14 strains (6.9%), respectively, followed by ST412 (n = 13), ST37 (n = 9), ST65 (n =
7), ST15 (n = 6), ST86 (n = 6), ST592 (n = 5) and ST29 (n = 4). There were 74 STs containing only
a single strain. Approximately 8.4% (17/203) of BSIKpn strains exhibited resistance to carbapenem
and most of them produced K. pneumoniae carbapenemase (KPC). The majority (70.4%, 143/203) of
BSIKpn isolates were associated with non-K1/K2/K5/K20/K54/K57 serotypes. Only 16.7% (34/203)
of BSIKpn strains had K1/K2 serotypes. A total of 66 (32.5%) of the BSIKpn strains exhibited HV
phenotype. The rmpA was a predominant factor in determining whether or not a BSIKpn strain had
HV phenotype.
Conclusions: The majority of BSIKpn strains exhibited high level of genetic diversity and low
resistance to commonly used antimicrobial agents. The specific capsular serotype and HV phenotype
were not the major features of BSIKpn strains from China.
INTRODUCTION
Bloodstream infection (BSI) is one of the most common infections for hospitalized patients. For
patients in intensive care unit, BSIs are the leading healthcare-associated infections which have been
linked to major morbidity and mortality [1]. With the growing incidence of BSIs caused by various
multi-drug resistant or even pan-drug resistant Enterobacteriaceae isolates, BSIs have been
urinary tract infection, bacteremia, and liver abscess. Among Gram-negative pathogens, K.
pneumoniae is second only to E. coli as the causative agent of various bacteremias [3]. The K.
pneumoniae-causing bacteremia can lead to higher mortality (ranged from 27.4 to 37%), prolonged
recognizable hypermucoviscosity (HV), have been identified 20 years ago, causing great concern
worldwide [3]. Recently, hvKP-causing bacteremias have also been reported [4-7]. The HV
phenotype in K. pneumoniae has been found to be related to the presence of the chromosomally
encoded mucoviscosity-associated gene A (magA) [8]. Further studies have revealed that magA is
responsible for capsular serotype K1 of K. pneumoniae, and it is clarified as the capsular polymerase
magA is not detected in any type of other serotypes[12]. Another HV-associated gene, the plasmid
gene regulator of the mucoid phenotype A (rmpA), has been confirmed as a gene regulating capsular
polysaccharide synthesis that confers a mucoid phenotype of clinical K. pneumoniae isolates [13].
However, BSI-causing K. pneumoniae (BSIKpn) isolates are highly diverse, suggesting that HV is
not a reliable indicator of the presence of virulent clones carrying magA and/or rmpA genes [6, 7].
The emergence and prevalence of carbapenem-resistant K. pneumoniae is another concern for K.
pneumoniae-causing BSI which has an impact on increased infection-related mortality and has
become the most important epidemiologic and therapeutic challenge [2, 14]. Various and
overlapping carbapenemases (including KPC, OXA-48 and NDM) have been involved in the
Although some clones (such ST23) account for the majority of BSIKpn, the BSIKpn isolates
exhibit highly sequence type (ST) diversity [4, 6, 7, 15]. The high level of clonal diversity of BSIKpn
Recently, a retrospective multicenter study was conducted to describe phenotypic and genotypic
characteristics of clinical BSIKpn isolates. In total, 219 BSIKpn isolates were collected from five
tertiary-care hospitals in Beijing, China, and were analyzed for their phenotypic and genotypic
characteristics. It was the first multicenter study about clinical BSIKpn isolates in Beijing.
Gram-negative pathogens was performed in five tertiary-care hospitals located in Beijing, China,
including Chinese PLA General Hospital (hospital A), 302nd Hospital of China (hospital B), PLA
Army General Hospital (hospital C), Navy General Hospital, PLA (hospital D) and Air Force
General hospital, PLA (hospital E). A total of 219 non-repetitive clinical BSIKpn isolates were
collected from Dec. 2013 to Dec. 2014. All clinical isolates were identified by VITEK® MS
(bioMérieux SA, Marcy-l'Etoile, France). Escherichia coli ATCC 25922 was used as the quality
control strain for antimicrobial susceptibility test. The Salmonella ser. Braenderup strain H9812 was
used as a reference standard for pulsed-field gel electrophoresis (PFGE). No ethical approval was
obtained for using the clinical samples, because these samples were collected during routine
bacteriologic analyses in public hospitals, and the data were anonymously analyzed.
Antimicrobial susceptibility test. The minimum inhibitory concentrations (MICs) of commonly
used antimicrobial agents (listed in Table 1) were measured by the agar dilution method. All
susceptibility results were interpreted according to the performance standards of 2016 CLSI [16].
PFGE and MLST analyses. PFGE with XbaI was performed for all clinical BSIKpn isolates as
previously described [17]. The PFGE patterns were analyzed by BIONUMERICS software (Applied
Maths NV, Sint-Martens-Latem, Belgium) using the dice similarity coefficient. Isolates were
considered as the same strain (PFGE type) if they possessed a genetic similarity of ≥95%. MLST was
carried out for all BSIKpn strains according to protocols available on the MLST websites
by eBURST v3.0 software (http://eburst.mlst.net/) to determine the clonal relationship of the isolates.
Determination of the HV phenotype. The HV phenotype was identified by a positive string test,
which was defined as the formation of a viscous string >5 mm in length when a colony grown
overnight on a blood agar plate at 37°C was stretched by a bacteriology inoculating loop [4].
Detection of capsular serotypes and HV-associated genes. Capsular serotypes K1, K2, K5, K20,
K54, K57, which have been described as the capsular serotypes of hvKP strains [10], were
determined by PCR detection of K serotype-specific alleles at wzx and wzy loci as described
previously [18]. magA is the serotype K1 wzy allele in BSIKpn [9-11, 18]. rmpA and rmpA2 genes
blaAIM, blaDIM, blaGIM, blaSIM, blaVIM, blaSPM, blaNDM and blaOX-48, were screened for
RESULTS
Prevalence of BSIKpn. A total of 857 BSI-causing Gram-negative bacilli were collected from
Dec. 2013 to Dec. 2014 from five tertiary-care hospitals located in Beijing, China. E. coli, K.
(81/857), 5.02% (43/857) and 7.93% (68/857) of all the Gram-negative bacilli BSI episodes,
respectively. Among 219 non-repetitive BSIKpn isolates, 69 of them were obtained from hospital A,
while 65, 28, 28 and 29 isolates were collected from hospital B, C, D and E, respectively.
Genetic relatedness of BSIKpn. The 219 non-repetitive BSIKpn isolates were categorized into
203 types (strains) which had a unique PFGE pattern. Twelve types contained two BSIKpn isolates,
two types were composed of three isolates, while other 189 types contained only one isolate. Isolates
with same PFGE type were considered as the same strain. Therefore, a total of 203 BSIKpn strains
MLST. A total of 105 different STs were identified from 203 BSIKpn strains. Figure 1shows that
the currently known 89 STs were clustered into 8 clonal complexes (CC) and 66 singletons. Overall,
ST11 and ST23 were the predominant clones, and they were detected from 14 strains (6.9%),
respectively, followed by ST412 (n = 13), ST37 (n = 9), ST65 (n = 7), ST15 (n = 6), ST86 (n = 6),
ST592 (n = 5) and ST29 (n = 4). Six STs (ST36, ST48, ST147, ST375, ST395 and ST462) contained
three strains, and 16 STs (ST12, ST17, ST25, ST35, ST60, ST101, ST200, ST231, ST290, ST367,
ST420, ST722, ST896, ST927, ST1373 and ST2374) were responsible for only two strains. The rest
of the 74 STs including 16 novel STs which have not been submitted to the MLST databases
antimicrobial susceptibility patterns of the BSIKpn strains. Approximately 8.4% (17/203) of BSIKpn
strains exhibited carbapenem-resistant phenotype. Fourteen of them produced KPC, while three
carbapenem-resistant BSIKpn strains did not produced any commonly known carbapenemase genes
(Table 2).
Capsular serotypes and HV-associated genes. The majority (70.4%, 143/203) of BSIKpn strains
strains had K1/K2 serotypes, followed by K54 (2.96%, 6/203), K57 (2.46%, 5/203), K20 (1.97%,
4/203) and K5 (0.49%, 1/203). About one-third (32.5%, 66/203) of the BSIKpn strains exhibited HV
phenotype. Table 3 lists the prevalence of capsular serotypes and HV-associated genes among
BSIKpn strains with and without HV phenotype. The HV positive BSIKpn isolates exhibited high
level of clonal diversity, and only two, five and eight isolates belonged to ST23, ST65 and ST412,
respectively. The majority (10 of 12) of BSIKpn isolates with K1 serotype belonged ST23, and only
DISCUSSION
from five tertiary-care hospitals in Beijing, China. PFGE results showed that 86.3% (189/203) of
clinical BSIKpn isolates had a unique pattern, indicating widespread diversification of BSIKpn
strains in this area. This high level of genetic diversity of clinical BSIKpn isolates also suggested that
possessed the advantage of transmission in the process of causing BSI. Substantial differences in
clonal structure of BSIKpn strains were also confirmed in this study. The predominant clones (ST11
and ST23) accounted for only 6.9% of all BSIKpn strains, respectively, while a large proportion of
strains with other STs were detected (Figure 1). This high level of clonal diversity has also been
found among BSIKpn strains from other areas [4, 7, 15], suggesting the common epidemiological
characteristics of the BSIKpn strains across different region. Therefore, further studies on a larger
scale are necessary to better understand the prevalence and clonal structure of clinical BSIKpn
isolates.
The BSIKpn strains exhibited low resistance to commonly used antimicrobial agents (Table 1).
In this study, the third generation cephalosporin resistance in BSIKpn was close to that of the United
States (about 29%), but it was lower than that of other studies in China (about 34%) [2]. About 8.4%
(17/203) of BSIKpn strains exhibited carbapenem-resistant phenotype, and the majority of them
produced KPC (Table 2). About half (8/17) of them belonged to ST11, which emerged as the major
clone responsible for the prevalence of carbapenemase-producing K. pneumoniae in this area [17, 20].
However, we also detected several non-ST11 BSIKpn clones (Table 2), suggesting that the high level
of clonal diversity of carbapenem-resistant BSIKpn was significantly different from the previous
heterogeneous bacterial pathogen causing various infections, clones of K. pneumoniae are adept at
acquiring additional genetic traits and becoming either hypervirulent or antibiotic resistant. A few
studies have demonstrated that both carbapenem resistance and severity of illnesses are associated
with increased mortality [14, 21], adding urgency to better understand and to prevent the spread of
carbapenem-resistant BSIKpn.
In this study, we detected six serotypes (K1, K2, K5, K20 K54 and K57), which were
considered as highly virulent phenotypes associated with severe infections in humans among the 77
described capsular (K) types of the serotyping scheme [10, 22]. The majority (70.4%) of BSIKpn
strains was associated with non-K1/K2/K5/K20/K54/K57 serotypes, and only 16.7% of BSIKpn
strains had K1/K2 serotypes. Meanwhile, the hypervirulent clones (such as CC23K1 and CC65K2)
were infrequent among BSIKpn strains. Although the K1/K2 capsular serotypes are very common in
some hvKP strains, some studies have revealed that a considerable proportion of the hvKP strains
may have a non-K1/K2 serotype [10, 18, 23, 24], and the majority of non-K1/K2 serotypes are K5,
K20, K54 and K57 [18, 24]. However, it has been confirmed from cases of respiratory infections and
septicemia that pyogenic liver abscess-causing K1 isolates are genetically distinct from other K1
isolates, and the pathogenic potential of liver abscess K1 isolates resides in their genomic
background and surface antigens, but not in capsular serotype [22, 25, 26]. our results also indicated
that serotypes other than K1/K2/K5/K20/K54/K57 also played important roles in the BSI.
invasive syndrome, such as liver abscess, meningitis or endophthalmitis [27-29]. However, only
one-third (32.5%, 66/203) of the BSIKpn strains exhibited HV phenotype in this study, which was
similar to the data from other studies in the same region (31.4-42.4%) [4, 6], but much higher than
the data of other areas (5.4-12.9%) [5, 7]. The prevalence of each serotype exhibited no significant
difference between groups with and without HV phenotype (Table 3). The virulent serotype K1 even
presented lower prevalence among hypermucoviscous strains (1.5% vs. 8%). All these data suggested
that there was no relationship between the capsular type and HV phenotype among clinical BSIKpn
strains. It has been confirmed that an HV phenotype, which has been used as phenotypic
characteristic for identification of hvKP strain, is not enough to recognize those virulent clones [7, 13,
29, 30]. Our results also suggested that factors other than HV phenotype may be required for the
systemic dissemination of K. pneumoniae. Furthermore, the phenotypic string test has poor positive
predictive value and test specificity for hvKP identification [31]. Altogether, it is very unreliable to
determine hvKP strains only by detecting HV phenotype. In other words, the phenotype should not
be used as the unique feature for the identification of hvKP. The prevalence of rmpA gene showed
significant difference between two groups (Table 3), suggesting that the rmpA was a predominant
factor in determining whether or not a BSIKpn strain had an HV phenotype. However, some strains
were positive for rmpA but negative for HV phenotype (Table 3). It has been pointed out that the role
of RmpA/RmpA2 and/or the HV phenotype are difficult to interpret [13, 29, 30]. Further efforts are
required to understand the underlying mechanisms of differential prevalence of BSIKpn and related
determinants.
However, the present study had the following limitations. First, the main drawback of the study
is the paucity of information on the clinical characterization of the patients. Second, other resistant
and virulent determinants of BSIKpn (such as AllS, Kfu, Aerobactin and RepA) did not been
unexplored.
analyzed in this study. The BSIKpn isolates exhibited high level of genetic diversity and low
resistance to commonly used antimicrobial agents. Approximately 8.4% (17/203) of BSIKpn strains
exhibited resistance to carbapenems. ST11 and ST23 were the predominant clones, but only 13.8%.
There were 74 STs containing only a single strain. Only 16.7% of BSIKpn strains had K1/K2
serotypes, while 32.5% of the BSIKpn strains exhibited HV phenotype. Therefore, The specific
capsular serotype and HV phenotype were not the major features of BSIKpn strains from China.
DECLARATIONS
Funding: This study was financially supported by the Special Key Project of Biosafety Technologies
for the National Major Research & Development Program of China (2017YFC1200800).
1 Russotto V, Cortegiani A, Graziano G, et al. Bloodstream infections in intensive care unit patients: Distribution and
antibiotic resistance of bacteria. Infect Drug Resist. 2015; 8: 287-296.
2 Akova M. Epidemiology of antimicrobial resistance in bloodstream infections. Virulence. 2016; 7: 252-266.
3 Paczosa MK, Mecsas J. Klebsiella pneumoniae: Going on the offense with a strong defense. Microbiol Mol Biol Rev.
2016; 80: 629-661.
4 Jung SW, Chae HJ, Park YJ, et al. Microbiological and clinical characteristics of bacteraemia caused by the
hypermucoviscosity phenotype of klebsiella pneumoniae in korea. Epidemiol Infect. 2012: 1-7.
5 Chang L, Bastian I, Warner M. Survey of klebsiella pneumoniae bacteraemia in two south australian hospitals and
detection of hypermucoviscous phenotype and maga/rmpa genotypes in k . Pneumoniae isolates. Infection. 2013; 41:
559-563.
6 Liu YM, Li BB, Zhang YY, et al. Clinical and molecular characteristics of emerging hypervirulent klebsiella pneumoniae
bloodstream infections in mainland china. Antimicrob Agents Chemother. 2014; 58: 5379-5385.
7 Cubero M, Grau I, Tubau F, et al. Hypervirulent klebsiella pneumoniae clones causing bacteraemia in adults in a
teaching hospital in barcelona, spain (2007-2013). Clin Microbiol Infect. 2016; 22: 154-160.
8 Fang CT, Chuang YP, Shun CT, Chang SC, Wang JT. A novel virulence gene in klebsiella pneumoniae strains causing
primary liver abscess and septic metastatic complications. J Exp Med. 2004; 199: 697-705.
9 Yeh KM, Chang FY, Fung CP, Lin JC, Siu LK. Maga is not a specific virulence gene for klebsiella pneumoniae strains
causing liver abscess but is part of the capsular polysaccharide gene cluster of k. Pneumoniae serotype k1. J Med
Microbiol. 2006; 55: 803-804.
10 Shon AS, Bajwa RP, Russo TA. Hypervirulent (hypermucoviscous) klebsiella pneumoniae: A new and dangerous breed.
Virulence. 2013; 4: 107-118.
11 Yeh KM, Lin JC, Yin FY, et al. Revisiting the importance of virulence determinant maga and its surrounding genes in
klebsiella pneumoniae causing pyogenic liver abscesses: Exact role in serotype k1 capsule formation. J Infect Dis.
2010; 201: 1259-1267.
12 Struve C, Bojer M, Nielsen EM, Hansen DS, Krogfelt KA. Investigation of the putative virulence gene maga in a
worldwide collection of 495 klebsiella isolates: Maga is restricted to the gene cluster of klebsiella pneumoniae
capsule serotype k1. J Med Microbiol. 2005; 54: 1111-1113.
13 Nassif X, Fournier JM, Arondel J, Sansonetti PJ. Mucoid phenotype of klebsiella pneumoniae is a plasmid-encoded
virulence factor. Infect Immun. 1989; 57: 546-552.
14 Ben-David D, Kordevani R, Keller N, et al. Outcome of carbapenem resistant klebsiella pneumoniae bloodstream
infections. Clin Microbiol Infect. 2012; 18: 54-60.
15 Gomez-Simmonds A, Greenman M, Sullivan SB, et al. Population structure of klebsiella pneumoniae causing
bloodstream infections at a new york city tertiary care hospital: Diversification of multidrug-resistant isolates. J Clin
Microbiol. 2015; 53: 2060-2067.
16 CLSI. Performance standards for antimicrobial susceptibility testing; twenty-fourth informational supplement.
Clinical and Laboratory Standards Institute, 950 West Valley Road, Suite 2500, Wayne, PA 19087 USA. 2016; 26th ed.
CLSI supplement M100S.
17 Yang J, Ye L, Guo L, et al. A nosocomial outbreak of kpc-2-producing klebsiella pneumoniae in a chinese hospital:
Dissemination of st11 and emergence of st37, st392 and st395. Clin Microbiol Infect. 2013.
18 Fang CT, Lai SY, Yi WC, Hsueh PR, Liu KL, Chang SC. Klebsiella pneumoniae genotype k1: An emerging pathogen that
causes septic ocular or central nervous system complications from pyogenic liver abscess. Clin Infect Dis. 2007; 45:
284-293.
19 Poirel L, Walsh TR, Cuvillier V, Nordmann P. Multiplex pcr for detection of acquired carbapenemase genes. Diagn
Microbiol Infect Dis. 2011; 70: 119-123.
20 Qi Y, Wei Z, Ji S, Du X, Shen P, Yu Y. St11, the dominant clone of kpc-producing klebsiella pneumoniae in china. J
Antimicrob Chemother. 2011; 66: 307-312.
21 Schwaber MJ, Klarfeld-Lidji S, Navon-Venezia S, Schwartz D, Leavitt A, Carmeli Y. Predictors of carbapenem-resistant
klebsiella pneumoniae acquisition among hospitalized adults and effect of acquisition on mortality. Antimicrob
Agents Chemother. 2008; 52: 1028-1033.
22 Brisse S, Fevre C, Passet V, et al. Virulent clones of klebsiella pneumoniae: Identification and evolutionary scenario
based on genomic and phenotypic characterization. PLoS One. 2009; 4: e4982.
23 Luo Y, Wang Y, Ye L, Yang J. Molecular epidemiology and virulence factors of pyogenic liver abscess causing klebsiella
pneumoniae in china. Clin Microbiol Infect. 2014; 20: O818-824.
24 Fung CP, Chang FY, Lee SC, et al. A global emerging disease of klebsiella pneumoniae liver abscess: Is serotype k1 an
important factor for complicated endophthalmitis? Gut. 2002; 50: 420-424.
25 Hsieh PF, Lin TL, Yang FL, et al. Lipopolysaccharide o1 antigen contributes to the virulence in klebsiella pneumoniae
causing pyogenic liver abscess. PLoS One. 2012; 7: e33155.
26 Yeh KM, Chiu SK, Lin CL, et al. Surface antigens contribute differently to the pathophysiological features in serotype
k1 and k2 klebsiella pneumoniae strains isolated from liver abscesses. Gut Pathog. 2016; 8: 4.
27 Lee CH, Liu JW, Su LH, Chien CC, Li CC, Yang KD. Hypermucoviscosity associated with klebsiella pneumoniae-mediated
invasive syndrome: A prospective cross-sectional study in taiwan. Int J Infect Dis. 2010; 14: e688-692.
28 Lee HC, Chuang YC, Yu WL, et al. Clinical implications of hypermucoviscosity phenotype in klebsiella pneumoniae
isolates: Association with invasive syndrome in patients with community-acquired bacteraemia. J Intern Med. 2006;
259: 606-614.
29 Lin YC, Lu MC, Tang HL, et al. Assessment of hypermucoviscosity as a virulence factor for experimental klebsiella
pneumoniae infections: Comparative virulence analysis with hypermucoviscosity-negative strain. BMC Microbiol.
2011; 11: 50.
30 Yu WL, Ko WC, Cheng KC, Lee CC, Lai CC, Chuang YC. Comparison of prevalence of virulence factors for klebsiella
pneumoniae liver abscesses between isolates with capsular k1/k2 and non-k1/k2 serotypes. Diagn Microbiol Infect
Dis. 2008; 62: 1-6.
31 Tan TY, Cheng Y, Ong M, Ng LS. Performance characteristics and clinical predictive value of the string test for
detection of hepato-virulent klebsiella pneumoniae isolated from blood cultures. Diagn Microbiol Infect Dis. 2014; 78:
127-128.
Figure 1. The eBURST diagram of MLST. Lines connect single locus variants. The relative size of
String test
Positive
1 (1.5) 56 (84.8) 12 (18.2) 1 (1.5) 11 (16.7) 0 (0) 2 (3) 2 (3) 3 (4.5) 47 (71.2)
(n=66)
Negative
11 (8) 11 (8.03) 22 (16.1) 11 (8) 10 (7.3) 1 (0.7) 2 (1.5) 4 (2.9) 2 (1.5) 96 (70.1)
(n=137)