Đặc điểm lâm sàng, cận lâm sàng, các yếu tố tiên lượng và một số kiểu gen vi khuẩn ở bệnh nhân nhiễm streptococcus suis điều trị tại bệnh viện bệnh nhiệt đới trung ương (2015 – 2018) ttta

1
A. INTRODUCTION
RATIONALE
Streptococcus suis is a gram positive cocci that could transmit from
animals (mainly pigs) to human via ingestion behaviors of uncooked
foods and non-intact skin exposure during pig butchering and pork meat
processing. It could cause an endemic with a high mortality rate raging
from 12,8% to 27,9%. Patients with S. suis infection after discharge
could have severe complications sush as hearing loss (from 50% to
66,4%), vestibular disorders (22,7%), limb gangrene and amputation.
These complications have a low recovery rate, leading to a great burden
in healthcare system and economic system. In Vietnam, disease due to
S.suis is a topical issue in the field of infectious diseases because of
specific epidemiological factors and severe clinical consequences,
especially in case of inappropriate diagnosis and late treatment. This
disease ranked on the 6th position among 10 infectious diseases with the
highest incidence and mortality rate. Currently, the resistance of S.suis to
almost antibiotics used in farming results in a high risk of antimicrobial
resistance in human. Resistance genes and virulence related genes have
been studied in the world. In Vietnam, major studies are mainly from the
South. Until now, despite a widespread circulation, there has been no
study on comprehensive evaluation of clinical and laboratory
characteristics, prognostic and virulence factors, distribution of these
features by clinical syndromes of S.suis in Vietnam. Therefore, we
conduct this research for three following objectives:
1 Describe clinical and laboratory characteristics, clinical
syndromes in patients infected with Streptococcus suis.
2 Determine prognostic factors of mortality in patients infected with
Streptococcus suis.
3 Evaluate the antibiotic susceptibility and identify some resistance
genes, virulence genes of Streptococcus suis.

URGENCY OF THESIS
Due to a different distribution of clinical syndromes among regions
in Vietnam, there should have been an overall picture of clinical and
laboratory characteristics, mortality prognostic factors in patients with
S.suis infection in order to help physicians avoid a pitfall in practice,
access to proper diagnosis and treatment, reduce the mortality risk. In
addition, the uncontrolled use of antibiotics in farming could lead to an
increase in antibiotic resistance associated with the emergence of high

1


2
virulent strains in pigs, which makes treatment becomes difficulty.
Therefore, it is necessary to understand genotypes, resistance genes and
virulence related genes of S.suis in human.
THESIS STRUCTURE
The length of this thesis was 124 pages, divided into 4 chapters
(Background: 30 pages; Material and Method: 23 pages; Results: 32
pages; Discussion: 34 pages. There were 40 tables; 8 charts; 10 figures;
115 references (both in English and Vietnamese, with 35% documents
from 2014)
B. CONTENT
Chapter 1. BACKGROUND

1

Microbiology features and pathogenesis of Streptococcus suis
In 1950s, Streptococcus suis which caused infection in pig was
initially reported in England and Netherlands. It is facultative gram

positive anaerobes with a spherical shape. It is transmitted via uncooked
ingestion and exposure through wounds in the skin while contact with
infected pigs in butchering and uncooked food processing. S.suis breach
the epithelial barriers, exist in blood then invade multiple organs and
cause inflammatory reactions in these organs. Moreover, S.suis is able to
cross the blood-brain barrier to invade the central nervous system,
leading to meningitis.

2

Epidemiology of Streptococcus suis infection in human
The first case of S.suis infection in human was described in 1968 in
Denmark. Since then, significantly increasing numbers of human cases
have been reported in many countries including Netherlands, Belgium,
Denmark, England, France, Germany, Sweden, Canada. Asia regions
such as Hong Kong, Taiwan, China, Thailand and Vietnam notably have
a great proportion of 90.2%. S.suis infection could cause an endemic,
typically in China where observed two outbreaks in 1999 and 2005 with
huge numbers of infected cases and a high mortality rate. In Vietnam, a
case of S.suis infection was first reported in November 1996 with
meningitis manifestation. Since then, human cases infected by S.suis
have been reported in all three regions in Vietnam. According to General
Department of Preventive Medicine – Ministry of Health, until 2016, the
rate of S.suis infection in human was 0.007 cases per 100,000 persons
and this disease was at the 6 th rank among 10 infectious diseases with the
highest incidence and mortality rate.

3

Clinical syndromes of S. suis infection and prognostic factors

2


3
Streptococcus suis causes a variety of clinical manifestations in
human such as meningitis, sepsis, septic shock, endocarditis…Around
the world, 68% of observed clinical manifestations are meningitis.
Hearing loss is the most common complication in patients with S.suis
meningitis with approximately 50% and a very low recovery rate. Septic
shock usually has a rapid progression with a high mortality rate ranging
from 60% to 80% depending on the research. Factors like coagulation
disorders, multiorgan failure, comorbidities and short incubation period
are prognostic factors in patients with S. Suis infection.

4

Diagnosis and treatment of patients with Streptococcus suis infection
Diagnosis of S.suis infection is relied on clinical manifestations of
meningitis (fever, headache, altered mental status, neck stiffness, nausea,
vomiting…), sepsis (fever, myalgia, hemorrhagic necrotic rash,
tachycardia…) in association with definitive tests for pathogen
identification sush as positive culture or PCR. Specific treatments are
betalactam antibiotics with supportive care like mechanical ventilation,
renal replacement therapy…

5

Antibiotic resistance and virulence factors of S. suis
S. suis is almost resistant to all antibiotics used in the farming like

tetracycline, erythromycin; some strains are resistant to ceftriaxon and
fluoroquinolone. Resistance genes were also identified such as erm(B)
(erythromycin), cat (chloramphenicol), tet(M), tet(O), tet(L)
(tetracyclin). Virulence factors are predominant existence in type 2 S.
suis which transmits directly from animals to human. Many studies in
the world indicated three major virulence factors including: protein
muramidase-released protein (MRP) encoded by mrp gene, suilysin
encoded by sly, extracellular factor (EF) encoded by epf gene.

6

Technique of molecular biology and the applications in S. Suis
research
The discovery of novel gene sequencing technology makes a
significant breakthrough compared to Sanger technology in time, quality
and cost of sequencing process. Whole genome sequencing of S.suis
provides aid for determining serotypes, MLST (Multilocus sequence
typing), resistance genes and virulence related genes.
Chapter 2. MATERIAL AND METHOD
Study population

1

3


4

1
1

2
3
4

2
1
2
3
4

2
3
4

–
•
•

Inclusion criteria: During the study period, we recruited 221 patients
with S. suis infection who met the following criteria:
Age > 16, without gender discrimination
Signs and symptoms of sepsis and/or meningitis
Blood/cerebrospinal fluid (CSF) culture and/or PCR blood/CSF is
positive with S. suis.
Agreement on study participation.
Exclusion criteria
Patients had additional microbiology or molecular results of other
bacteria beside S. suis
Patients had comorbidities including liver and kidney failure before S.
suis infection

Patients had co-infection with HIV, HCV, HBV
Patients refused to participate in the study at any time regardless of
study period.
Study design: Prospective study.
Sample size: Convenience sampling, the study will recruited all patients
who satisfied inclusion criteria meet in 2.1.1, and who were treated at
National hospital of Tropical diseases from May 2015 to May 2018.
Method
Patients with definitive diagnosis of S. suis infection will be
classified into 3 groups: (1) Meningitis, (2) Sepsis with meningitis and
(3) Septic shock. We will not only evaluate and analyse general
characteristics but we also compare features of clinical findings,
laboratory tests and prognostic factors, resistance genes, virulence genes
and serotype.
2.4.1. Study process
a) Objective
1:
Evaluate
general
clinical
characteristics, laboratory results and clinical
syndromes in patients infected with Streptococcus
suis
Clinical syndromes caused by human infection of S. suis: Based on
clinical manifestations and sample culture or PCR results, clinical
syndromes were divided into three following groups:
Meningitis: patients with signs and symptoms of meningitis + CSF
culture and/or PCR is positive with S. suis.
Sepsis with meningitis: patients with clinical manifestations of sepsis
and meningitis had positive result of S. suis in both blood and CSF

4


5
samples.
Septic shock: patients with clinical presentations of septic shock + at
least 1 positive laboratory result of S. suis (culture or PCR in blood
and/or CSF).
⇒ Record clinical syndromes will be statistically represented in appropriate
quantity, percentage and proper tables or charts.
– Clinical characteristics: research was carried out to investigate not only
general clinical characteristics in patient population but also individual
features in each group of clinical syndrome. A comparison of individual
characteristics among 3 clinical syndromes were made based on
exposure history (ingestion, contact..), symptoms (fever, headache,
myalgia, nausea, vomiting, diarrhea...), signs (altered mental status,
neck stiffness, kernig...)
– Laboratory characteristics: we recorded laboratory results in 3 clinical
syndromes and noticed a difference by using mean or median
comparison (chi-squared test and Kruskal-wallis test...). Laboratory tests
included routine tests, tests for pathogen identification and tests for
follow-up treatment.
b) Objective 2: Analyze prognostic factors of mortality in patients
infected with Streptococcus suis
Patients in the study were divided into 02 groups based on
treatment outcome
1 Survival group: patients were cured or made a good recovery and were
transferred to provincial/district hospital in order to satisfy expectations
of patient or family members.

2 Death group: patients were died at the hospital or patients went home for
death because of treatment unresponsiveness or the severity of disease
progression.
We utilized univariate and multivariate regression analysis of
clinical characteristics, laboratory results, exposure history and
comorbidities between two groups to identify prognostic factors of
mortality.
Treatment outcome: Mortality rate, average duration of treatment,
sequelae (hearing loss, paralysis, mental issues, limb amputation due to
gangrene after discharge).
c) Objective 3: Evaluate antibiotic susceptibility patterns and identify
some resistance genes, virulence genes of Streptococcus suis

•

5


6
We identified antibiotic sensitivity of S. suis by susceptibility
testing, MIC calculation with E-test technique (E-test strips
manufactured by Biomerieux-France), which is based on the guidance of
CLSI M100.
Identification of resistance genes, virulence genes, serotypes and
genotypes via MLST were based on analysis of S. suis genome after
performing novel sequencing method by Miseq machine, using primer
from standard kit of NextaraXT library manufactured by Illumia –
American. Results were checked, analysed with experts from Hanoi
Oxford University Clinical Research Unit (OUCRU).
– Evaluate antibiotic sensitivity of S. suis: Determine the sensitivity of

some common antibiotics in clinical practice (ampicillin, penicillin,
ceftriaxon, levofloxacin, linezolid, vancomycin) and in cattle breeding
(erythromycin, clindamycin, tetracycllin). Calculate the percentage of
sensitivity and resistance.
– Determine some resistance genes, plasmids and virulence genes: We
identify resistance genes, virulence genes based on a search of following
genbank:
• For resistance genes: from ARG-ANNOT (Antibiotic Resistance GeneANNOTation), CARD (Comprehensive Antibiotic Resistance Database,
Resfinder);
• For virulence genes: from VFDB (Virulence Factor Database);
– Determine serotype, MLST (Multilocus sequencing type) and
phylogenetic tree: Identification of S. suis serotype was relied on
capsular polysaccharide region where encoded genes in locus cps for the
synthesis of bio-functional proteins. We downloaded database of cps
gene from the system via the following link: https://
github.com/streplab/SsuisSerotyping_pipeline. MLST classification of
isolated S. suis strains were determined owing to 7 basic genes (house
keeping gene). We compared this data with database in the website:
.
⇒ Analyze the distribution of serotypes and genotypes in each clinical
syndrome and general distribution of overall population.
⇒ Compare genome sequences among MSLT strains and with standard
strain.
5 Criteria, technique in our study
– Criteria for diagnosis of sepsis and septic shock: Surviving Sepsis

6


7

Campaign (2012)
– Criteria for clinical diagnosis of meningitis: Based on Principles and
Practice of Infectious Diseases (2013).
– Criteria for alcoholism: Based on AUDIT-C score of WHO (2001) and
developed by Frank (2008). Patients were diagnosed with alcoholism
when they had total score ≥ 4 in men and ≥ 3 in women.
– Technique for Streptococcus suis identification using Realtime PCR:
Use a pair of primers and a probe for cps2 gene to recognize S. suis
o Primers:
• cps2JF (5’- GGTTACTTGCTACTTTTGATGGAAATT-3’)
• cps2JR (5’-CGCACCTCTTTTATCTCTTCCAA-3’)
o Probe: (FAM-TCAAGAATCTGAGCTGCAAAAGTGTCAAATTGATAMRA). Bioresearch (American)
+ Machine: 7500 Fast Real-time PCR (Applied Biosystem – American)
+ A process of real-time PCR for cps2 gene (based on the process
performed at Laboratory Department in National Hospital of Tropical
Diseases and OUCRU in Hanoi
– Culture and identification technique of S. suis: Equipment: Blood
culture system Bactec 9050/9120: Bactec Dickinson, American and
some machines from England, Germany, France; Sheep blood agar,
chocolate agar (Oxoid, England). Bacteria identification was based on
Vitek 2 compact system from Biomeriux, France.
– Susceptibility testing and MIC identification with E-test: Use
equipment and antibiotic strips (E-test) manufactured by Bio merieux,
France. Interpretation and Report based on CLSI M100.
– Novel sequencing technology and data analysis using bioinformatics
sofware: Follow current procedures in National Hospital of Tropical
Diseases and Oxford University Clinical Research Unit.
6 Data collection: Data collection was carried out via designed case report
form (CRF).
7 Data analysis: Software SPSS 16.0, STAT with appropiate statistical

methods were performed for data analysis.
8 Study location: National Hospital of Tropical Diseases.

9

Study duration: 3 years from May 2015 to May 2018.

7


8

10 Medical Ethics: The study was approval by Institutional Review Board
of Hanoi Medical University and National Hospital of Tropical Diseases.
Chapter 3. RESULTS
There were 221 patients who fulfilled the inclusion criteria during
study period. A higher proportion of male patients was observed in the
study (92%). Mean age of patient population was 53.5 with the highest
proportion (69.23%) belonged to age group of 40 – 60. Disease due to
S.suis occurs predominantly from April to October.
1 Clinical and laboratory characteristics of patient population
1 General characteristics
– Pattern of exposure conditions
Table 3.1. Exposure conditions
Frequency
Proportion
History of exposure
(n=221)
%
Contact exposure

59
26.70
• Pig butchering
41
18.55
• Uncooked food processing
11
4.98
• Pig breeding
7
3.17
Ingestion exposure
46
20.81
• Eat “tiet canh”
41
18.55
• Eat other raw foods
5
2.26
Ingestion + Contact
12
5.43
Unknown
104
47.06
There were 117 patients (52.94%) with medical history of
exposure to source of infection. Eating “tiet canh” made the largest
contribution of 23.08%. There were 47.06% patients with unknown
history of exposure.

2 Clinical characteristics
– Clinical syndromes.
Chart 3.1. Clinical syndromes
Meningitis was the most common clinical syndrome, followed
by sepsis with meningitis (35.8%) and septic shock (18.1%).
Table 3.2. General symptoms

8


9
Frequency
Proportion %
(n=221)
Fever
221
100
Headache
197
89.14
Shivering
151
68.33
Nausea
139
62.90
Vomiting
132
59.73
Myalgia

121
55.00
The most common symptom in S.suis patients were fever, headache
and shivering.
Table 3.3. Symptoms by category of clinical syndrome
Meningiti
Meningitis with
Septic
p
s
sepsis
shock
Symptoms
n=102
n=79
n=40
(%)
(%)
(%)
102
79
40
Fever
(100)
(100)
(100)
63
53
35
Shivering

0.01
(61.76)
(67.09)
(87.50)
52
43
26
Myalgia
0.35
(51.49)
(54.43)
(65.00)
96
72
29
Headache
0.001
(94.12)
(91.14)
(72.5)
76
49
14
<0.00
Nausea
1
(74.51)
(62.03)
(35.00)
66

46
20
Vomiting
0.26
(64.71)
(58.23)
(50.00)
9
14
25
<0.00
Diarrhea
1
(8.82)
(17.72)
(62.50)
Specific symptoms in meningitis group were headache, nausea and
vomiting. Patients with septic shock often presented with shivering,
myalgia and diarrhea.
Symptoms

–
9

Signs


10
Table 3.4. Signs
Frequency

Proportio
(n=221)
n%
Kernig (+)
171
77.38
Altered mental status (Glasgow<15)
133
60.18
Refill >2s
78
38.42
Respiratory rate >20 rpm
71
32.13
Hemorrhagic necrotic rash
55
24.89
Kernig and altered mental status were the most common signs in
patients with S. suis infection.
Signs

Table 3.5. Signs by category of clinical syndrome
Meningitis
Signs

Kernig (+)
Altered mental status
(G<15)
Cold extremities

Hemorrhagic necrotic
rash
Refill >2s

n=102
(%)
97
(95.1)
68
(66.67)
1
(0.98)
6
(5.88)
26
(25.49)

Sepsis with
Septic shock
meningitis
n=79
n=40
(%)
(%)
60
14
(75.95)
(35.00)
48
17

(60.76)
(42.05)
5
24
(6.33)
(60.00)
16
33
(20.25)
(82.5)
29
23
(36.71)
(63.89)

p

<0.00
1
<0.00
1
<0.00
1
<0.00
1
0.002

(Chi-squared test)
Major patients with meningitis had Glasgow coma score below
15 and positive kernig sign. In group of septic shock, common signs

were hemorrhagic necrotic rash, refill > 2 seconds and cold extremities.

3
–
10

Laboratory characteristics
Biochemistry results.


11
Table 3.6. Biochemistry features
Total
Meningiti Sepsis with
Septic
p
s
meningitis
shock
(n=221)
Biochemistr
y result
Frequenc Frequency Frequenc Frequency
y (%)
y (%)
(%)
(%)
Creatinin ≥
5
16

37
58
<0.001
120 (µmol/l)
*
(4.90)
(20.25)
(92.5)
(26.24)
Total
21
36
38
95
bilirubin > 17
<0.001
(20.59)
(45.57)
(95.00)
(42.99)
(mmol/l)
52
55
39
146
AST ≥ 40
<0.001
(UI/l)
(50.98)
(69.62)

(97.50)
(66.06)
60
55
37
152
ALT ≥ 37
<0.001
(UI/l)
(58.82)
(69.62)
(92.50)
(68.78)
(Chi-squared test, * Fisher’s exact test)
Patients with septic shock had more significant decrease in liver
and kidney function than that in patients with meningitis.
– Coagulation results
Table 3.7. Coagulation features
Sepsis with
Septic
Meningitis
Total
p
meningitis
shock
Coagulation
tests
Frequency Frequency Frequency Frequency
(%)
(%)

(%)
(%)
30
33
37
100
<0.00
PT< 70 (%)
(29.41)
(41.77)
(92.5)
(45.25)
1
Fibrinogen <
0
7
31
38
<0.00
2 (g/l)
(0)
(8.86)
(77.5)
(17.19)
1
D-Dimer
44
47
37
128

<0.00
>500 (ng/l)
(43.14)
(59.49)
(92.5)
(57.92)
1
(Chi-squared test)
There were a marked decline in prothrombin time, fibrinogen and
noticeable increase in D-dimer, APTTs in the group of septic shock
compared to remaining groups.
4 Treatment outcome
– Mortality rate: The overall mortality rate was 14.03%. The mortality
rate in patients of septic shock was 60%. 35.26% patients were death
within the first 24 hours of admission.
– Sequelae when discharged: Proportion of hearing loss in patient
population was 44.8%.

11


12

2

Prognostic factors of mortality
Table 3.8. Regression analysis of epidemiological factors related to
mortality
p
p

Characteristics OR 95% CI
aOR 95% CI
value
value
Exposure
No ingestion
1
1
3.7
7.5
Ingestion
1.72 8.23 <0.001 3.37 1.52
<0.001
6
1
Alcoholism
No
1
1
2.1
Yes
1 4.63 0.05 1.72 0.77 3.82 0.18
5
Mortality risk in patients with ingestion exposure was 3.37 times
higher than that in patients with other factors.
Table 3.9. Regression analysis of clinical factors related to mortality
Characteristic
p
p
OR 95% CI

aOR 95% CI
s
value
value
Diarrhea
No
1
1
13.9
Yes
6.23 2.78
<0.001
3
1.30 0.45 3.80
0.63
Hemorrhagic necrotic
rash
No
1
1
17.5
44.3
32.9
Yes
6.97
<0.001
9
8
9.59 2.79 9
<0.001

Patients with hemorrhagic necrotic rash had 9.59 times higher in
mortality risk.
Table 3.10. Regression analysis of liver and kidney function related
to mortality
Liver & kidney
function
Ure (mmol/l)
≤7.5
>7.5
Creatinin (μmol/l)
<120

12

OR

Univariate
95 CI

p

aOR

Multivariate
95 CI
p

1
1
21.7

<0.00
0.1
6.36 74.54
1.00
8
1
2
1

8.32

1.00


13
≥120

49.7 14.2 174.2 <0.00 21.7 2.9 162.2 <0.00
8
2
4
1
8
2
3
1

Total Bilirubin (mmol/l)
≤17
1

1
17.1
<0.00
1.0
>17
5.02 58.46
4.30
17.47
3
1
6
AST (UI/l)
<40
1
1
19.1
143.3
0.1
≥40
2.56
0.004 2.41
32.44
4
5
8
ALT (UI/l)
<37
1
1
0.2

≥37
7.90 1.83 34.13 0.01 1.63
12.37
1

0.04

0.51

0.64

Patients with Creatinin ≥120 μmol/l and total bilirubin >17 UI/l had
a higher mortality risk than other patients (OR were 21.78 and 4.3;
respectively, and p<0.05).
Table 3.11. Regression analysis of coagulation factors related to
mortality
Univariate
Multivariate
Coagulation
OR
95% CI
p
aOR 95% CI
p
PT (%)
≥70
1
1
<0.00
<70

15.30 4.49 52.13
4.44 0.70 28.16 0.11
1
Fibrinogen (g/L)
≥2
1
1
<0.00
<2
21.37 8.66 52.72
7.68 2.64 22.32 <0.001
1
D-dimer (ng/l)
≤500
1
1
<0.00
>500
8.40 2.47 28.57
3.12 0.80 12.23 0.10
1
Patients with fibrinogen <2 g/L on admission had 7.68 times higher
risk of mortality than patients with fibrinogen ≥2 g/L (p<0.001)
3 Antibiotic resistance and molecular results.
In our study, 80 isolated samples with full criteria were
performed whole genome sequencing.

13

14

1
–

Antibiotic resistance of S. suis
Susceptibility test

Chart 3.2. Antibiotic susceptibility profile of S. suis
Low susceptibility was observed in erythromycin, clindamycin
and tetracyline whereas susceptibility of tetracyline was 0%.
Susceptibility of remaining antibiotics (peniciline, ampiciline,
ceftriaxone, linezolid, levofloxacin and vancomycin) were 100%.
– Occurrence of resistance genes.

Chart 3.3. Frequency of resistance gene
The most common resistance genes were ant(6) with 78.75%, tet
(B) with 75% and erm(B) with 60%, followed by tet(O) with 28.75% and
aph (3) with 21.25%.

2

Distribution of serotype

–

Distribution of serotype
Table 3.12. Distribution of serotype

14

15
Serotyp
e
2
14
1/2

Meningiti
s
n=37
(%)
35
(94.59)
2
(5.41)
0
(0.00)

Sepsis with
meningitis
n=28
(%)
26
(92.86)
2
(7.14)
0
(0.00)

Septic
shock
n=15
(% )
13
(86.67)
0
(0.00)
2
(13.33)

Total
n=80
(% )
74
(92.50)
4
(5.00)
2
(2.50)

p

0.05

The predominant serotype was 02. The occurrence of serotype 1/2 was
only observed in patients with septic shock (with the proportion of 13.33%).

3

–

Distribution of sequence type (MLST).

Distribution of sequence type
Table 3.13. Distribution of sequence type by group of
clinical syndrome
Sepsis
Meningiti
with
Septic
Total
s
meningiti
shock
ST
p
s
n=37
n=28
n=15
n=80
(%)
(%)
(%)
(%)
65
29
25
11

ST1
(81.25
(78.38)
(89.29)
(73.33)
)
9
2
1
6
ST665
(11.25
(16.22)
(7.14)
(6.67)
)
1
1
0
2
ST105
(2.70)
(3.57)
(0.00)
(2.50)
0
0
1
1
0.08

ST25
(0.00)
(0.00)
(6.67)
(1.25)
1
0
0
1
ST28
(2.70)
(0.00)
(0.00)
(1.25)
1
0
1
2
Undefined
ST
(2.7)
(0.00)
(6.67)
(2.50)

15


16
ST1 were predominant (81.25%). There was no significant

difference in distribution of sequence type among groups of clinical
syndrome. ST25 and ST28 were 2 novel types. There were 02 samples
with undefined sequence type.
4 Occurrence of virulence genes
– Distribution of virulence genes
Table 3.14. Distribution of virulence genes
Meningiti Sepsis with
Septic
Total
s
meningitis
shock
Virulence
p
gene
n=37
n=28
n=15
n=80
(%)
(%)
(%)
(%)
37
28
14
79
mrp
0.11
(100)

(100)
(93.33)
(98.75)
36
28
14
78
epf
0.41
(97.3)
(100)
(93.33)
(97.50)
35
28
14
77
sly
0.42
(94.59)
(100)
(93.33)
(96.25)
32
19
9
60
salK/R
0.08
(86.49)

(67.86)
(60.00)
(75.00)
Major virulence genes in isolated samples were mrp, epf và sly.
There was no significant difference in distribution of virulence genes
among groups of clinical syndrome
Chapter 4. DISCUSSION
4.1. General characteristics of study population
.1.1. Characteristics of age, gender, and monthly distribution
– Age characteristic
Our study on 221 patients showed that the majority of patients was
in their middle age ranging from 40 to 60 years old, constituting 69.23%.
Median age of patients was 53 years old. This finding is consistent with
several previously published studies, as the median age of S. suis
infection was 54 years old in a study in China [4], and 52.9 years old in
another study in Thailand [57]. In Vietnam, a study on 101 patients
conducted by Ho Dang Trung Nghia at Hospital of Tropical Diseases at
Ho Chi Minh City also showed a similar result (a median age of 50 years
old) [21]. Our result confirms again that S. suis infection was common
among middle-aged population (over 40 years old), but rare among
young population (under 30 years old).
– Gender characteristic.
Our study participants were predominantly male (92%). There was
little difference in gender characteristic between our study and previous

16


17
studies. Several meta-analysis indicated that a proportion of male gender

ranged from 72.2% to 80.6% in a variety of previous studies. In
Vietnam, men are the principal component of workforce of pig-related
occupations, such as pig farming and pig butchering, which places them
under a high risk for S. suis infection. Additionally, Vietnamese men also
commonly developed habits of consuming porcine raw and undercooked
products, including “tiet canh”, fermented pork rolls, and undercooked
pork.
– Monthly distribution of S. suis infection
S. suis infection had a seasonal distribution throughout a year,
with higher incidence in months in summer and autumn from April to
October. This is in line with a study by Wertheim which demonstrated
that S. suis infection cases were reported mainly in May and July in
2007. However, further evidence from studies on a larger sample size in
a wider variety of settings is still needed to confirm seasonal distribution
of S. suis infection in Vietnam.
4.1.2. Characteristics of transmission routes and comorbidities
of study population
– Transmission routes.
Transmission routes of S. suis were reported in 52.94% of all
patients. The most common transmission route was exposure through
consumption of porcine food products, mostly consumption of “tiet
canh”, followed by exposure through pig butchering. This is in
agreement with other reports in Vietnam which illustrated that the
majority of S. suis infection was transmitted through eating pig-related
raw food products. A study by Ho Dang Trung Nghia in 2009 indicated
exposure to undercooked porcine food products was reported in 48% of
all cases. A meta-analysis on 21 previous studies showed that
occupational risk factor was reported in 58.6% of all cases in case
reports, and in 25% of all cases in studies on a larger sample size. Our
finding again highlights the significance of preventive prophylaxis of S.

suis infection through consumption of undercooked pig-related food
products to minimize risk of S. suis infection. Strategies to raise public
awareness and to change raw and undercooked pork eating habits should
be intensified, particularly in the context of high incidence of S. suis
infection in Vietnam.
4.2. Clinical and laboratory characteristics and clinical syndromes of
S. suis infection in human.
4.2.1. Clinical syndromes.
Meningitis was the most common clinical syndrome among our
study population, with a prevalence of 46.1%, followed by sepsis with
meningitis (35.7%) and septic shock (18%). The distribution of clinical

17


18
syndromes varies greatly among different countries and studies with
different designs and diagnostic methods. According to
Wangsomboonsiri (2008), meningitis accounted for 52%, followed by
sepsis (27%) and septic shock (12%). In China, septic shock had a
prevalence of 64% in an outbreak in Jiangsu and of 28.9% in another
outbreak in Sichuan. This reflects that pathogenic properties and
distribution of S. suis strains varies among different countries.
4.2.2. Clinical characteristics
a) Symptoms.
All of our patients had fever as the onset symptom, which is
consistent with the outbreak in Sichuan, China. Symptoms on admission
were significantly different between patients with meningitis and those
with septic shock. In patients with only meningitis and patients with sepsis
accompanied by meningitis, the most common symptoms were headache,

nausea and vomiting. Meanwhile, patients with septic shock frequently
presented with shivering, myalgia and diarrhea. Our results of symptoms
on admission in patients with meningitis share significant similarities with
previous studies in Vietnam and other countries. In a study in 2015 by
Anusha van Samkar, headache and nausea/vomiting were presented in
95% and 65% of all patients, respectively. In the 2015 outbreak in Jiangsu,
China, 85.7% of all patients had headache. In Vietnam, a study on 151
patients with meningitis at Hospital of Tropical Diseases at Ho Chi Minh
City by Nguyen Thi Hoang Mai reported that headache and vomiting was
found in 94% and 66.2% of all patients.
b) Signs.
60.18% of our patients had altered mental status with Glasgow
Coma Scale below 15 points. This finding matches with a result of
69.8% patients with altered mental status in a study on 151 patients by
Nguyen Thi Hoang Mai. Other studies in Vietnam and in foreign
countries also reported a similar proportion of meningitis signs: 93% in a
study by Anusha van Samkar, 90.5% in a study by G. Wang and 94% in
the study by Nguyen Thi Hoang Mai.
In our study, typical signs in patients with septic shock were
significantly different from patients with meningitis. The most common
signs were hemorrhagic necrotic rash and cold extremities, with a
respective proportion of 82.5% and 60%. There were few studies on
patients with septic shock by S. suis infection in the literature over the
world, and in Vietnam, no study has been systematically conducted on
this clinical syndrome. Cold extremities and hemorrhagic necrotic rash
were found in 79% and 61% of patients in the Jiangsu outbreak in
China, while a study in Thailand by Fongcom et al demonstrated a
proportion of patients with hemorrhagic necrotic rash of 95%. This is a

18

19
common sign in patients with septic shock by S. suis. Hemorrhagic
necrotic rash in our study was dark purple, macular, confluent rash
which usually begins on outer ear, head, face, neck, then spreads to trunk
and extremities. This could suggest an important clinical evidence to
early diagnose septic shock by S. suis and provide physicians with a
basis for an appropriate management and plan to minimize mortality and
to improve treatment effectiveness in patients with S. suis infection.
4.2.4. Laboratory characteristics
a) Features of serum biochemistry results.
Patients with septic shock had significantly abnormal liver and
kidney function results compared to the remaining groups of patients.
More than 90% of septic shock patients had abnormal liver and kidney
function tests. Our results concur with previous findings in several
studies in Thailand and China. Hongjue Yu et al reported a proportion of
liver and renal failure of 90% and 59% in patients with septic shock,
respectively, which was higher than in patients without septic shock
(68% and 3%). In Thailand, a 2009 study by Fongcon also showed
similar results: patients with septic shock had elevated liver enzymes,
creatinine and total bilirubin level. Severe hepatic and renal injuries in
patients with septic shock correlates with their clinical scenarios. The
majority of our septic shock patients had oliguria or anuria, some of
whom required hemodialysis to replace renal functions.
b) Characteristics of blood coagulation.
Blood coagulation results were also considerably different among
three groups of patients. Patients with septic shock had a significantly
higher proportion of abnormal coagulation tests – namely prothrombine
time, fibrinogen and D-dimer – than remaining groups (meningitis and

sepsis with meningitis). All of our patients with septic shock had
disseminated intravascular coagulation, while this disorder was
presented in only 79.7% of patients in the Sichuan outbreak in China.
This finding suggests that septic shock patients by S. suis infection
develop a severe coagulation disorder, in which widespread
consumption of coagulation factors culminates in hemorrhage. This
disseminated intravascular coagulation, which is manifested by an
elevated D-dimer level, in combination with other factors, brings a
poorer prognosis to patients with septic shock.
4.3. Treatment outcome and mortality prognostic factors
4.3.1. Mortality rate.
Our patients with S. suis infection had an overall mortality rate of
14.03%, 35.26% of whom died within 24 hours after admission. The
mortality rate of patients with septic shock was 60%. Our mortality rate
differs considerably from that of previous studies. In 2018, Vu Thi Lan

19


20
Huong reported a mortality rate of 12.8% in patients with S. suis infection.
The highest mortality rate ever reported was 27.9% in a study by
Fongcom in 2009, in which a mortality rate of patients with septic shock
was 80%. In a 2006 study in China by Hongjie Yu, the mortality rate was
18.1%. This demonstrates a significantly higher mortality rate in our study
compared to previous studies in Vietnam. A plausible explanation for this
phenomenon could be the exclusion of patients with septic shock in those
above-mentioned studies, while 18% of our patients developed septic
shock. Moreover, it should be noted that our study was conducted at
National Hospital of Tropical Diseases, a tertiary referral hospital

specialized in infectious diseases, where most patients are in a severe
condition and beyond the treatment and care capability of provincial and
primary care hospitals. In addition, most patients are referred to national
specialized hospitals in their late stages of disease, which could contribute
to a higher mortality rate in these hospitals.
4.3.2. Complications of patients with S. suis infection.
a) Hearing loss.
Hearing loss prevalence was 44.8% in our study. This finding
correlates with several previous studies in Vietnam and in foreign
countries. Some research reported that hearing loss and vestibular
disorders are two most common complications of S. suis infection in
human. Hearing loss prevalence varies greatly among different
studies, ranging from 6% to 100%, with a mean prevalence of 39%.
In Vietnam, hearing loss had a prevalence of 66.4% and 27.3%,
respectively.
b) Extremity gangrene.
A rare complication in our study that has never been mentioned in
the literature was extremity gangrene by S. suis. In our study, this
complication was found in 6 patients, 4 of whom had a clinical scenario
of septic shock, but only 3 of whom required limb amputation with the
proportion of 1.36%. These three patients, after a period of recovery
from shock or improved clinical scenarios, began to develop cyanosis in
distal extremities, and then turn into dry gangrene in middle and distal
phalanges. In these patients, distal extremities could not be preserved,
and therefore, required amputation.
4.3.3. Prognostic factors of mortality
a) Clinical epidemiological factors
Our study illustrated a higher mortality rate in patients exposed
through food ingestion than in those exposed through farming,
butchering or trading. In a study in Thailand by Wangsomboonsiri,

patients having live pig exposure (through butchering, farming) had a

20


21
higher mortality rate than those with S. suis infection through food
consumption (P = 0.016). This difference could be attributable to the
discrepancies in epidemiological properties and in pathogenic strains
between two studies. Our finding is significant in the context of
Vietnam, where a majority of S. suis infection cases is transmitted
through eating “tiet canh” or other undercooked porcine food products.
Exposure through food ingestion was also identified in over a half of all
cases with documented exposure history in our study.
b) Clinical factors
Among all clinical characteristics, we noticed significant
differences in the presence of diarrhea and hemorrhagic necrotic rash
between survival and death groups. When including these factors into
univariate and multivariate logistic regression analysis, we found an
increase of 9.59 times in mortality risk in patients with hemorrhagic
necrotic rash compared to those without it (p<0.001).
c) Laboratory factors
In our study, patients who died had more severe manifestations of
renal failure than those who survived. When including this factor into
multivariate logistic regression analysis, we noticed a 21.78-time higher
mortality risk in patients with serum creatinine level greater than 120
μmol/L compared to those without renal failure (p<0.01). Severe
coagulation disorder was a recognizable clinical manifestation in nonsurviving patients, with a significantly higher rate of fibrinogen
consumption compared to survivors (P<0.001). This supports previous
findings in 2008 by Wittaya Wangsomboonsiri, which, in multivariate

logistic regression analysis, demonstrated that mortality rate in patients
with increased fibrinogen consumption (fibrinogen level <2g/L) was
7.68 times higher than in those with normal fibrinogen level (fibrinogen
level ≥2 g/L) (P <0.001).
4.4. Antibiotic resistance
All isolated S. suis strains in our study were sensitive to beta-lactam
antibiotics. In addition, all strains were also sensitive to other antibiotics,
including linezolid, levofloxacin và vancomycin, which are typically used
in patients with ventilation associated pneumonia, or other superinfections
probably caused by percutaneous procedures in intensive care unit such as
central venous catheterization and continuous renal replacement therapy.
This implies that broad-spectrum antimicrobial combination therapy in
critical and complicated patients might have no impact on treatment
effectiveness in S. suis infection. Our results correlate well with Ngo Thi
Hoa (2011), who indicated that 100% S. suis strains were sensitive to
penicillin, ceftriaxone and vancomycin. This antibiotic susceptibility was
recently replicated in a 2019 study in Thailand on 27 S. suis strains

21


22
isolated from human, in which all strains were sensitive to beta-lactam
antibiotics, such as ampicillin, penicillin and third-generation
cephalosporins, as well as levofloxacin and vancomycin. This indicates a
similarity in antimicrobial susceptibility of pathogenic S. suis strains in
Vietnam and in other South-east Asian countries, particularly in Thailand.
However, a high proportion of S. suis strains in our study were resistant to
antibiotics that are customarily used in farming, such as erythromycin
(76.3%), clindamycin (73.5%) and tetracycline (97.6%). Our finding was

consistent with Suganya Yongkiettrakul (2019), who demonstrated
proportions of above-mentioned antibiotic-resistant S. suis strains were
70.4% (to erythromycin), 81.5% (to clindamycin) and 100% (to
tetracycline). Another study in 2016 by Athey et al showed 96% and 98%
of S. suis strains were resistant to tetracycline and erythromycin,
respectively.
4.5. Several genotypes of S. suis in our study
4.5.1. Antimicrobial resistance genes of S. suis
In China, Ye et al (2008) found that all S. suis strains isolated from
human were resistant to tetracycline, 97.27% of which had the presence
of tet(O) gene and one of which (0.08%) had the presence of tet(M)
gene. In Italia, Manzin et al (2008) detected the presence of tet(W) gene
in human-derived isolated S. suis strains. In Vietnam, the tet (O), tet (L)
and tet(M) genes were present in 21.6%, 3.3% and 84.3% of S. suis
strains. The erm(B) gene was found in 94.7% of erythromycin-resistant
strains. Our study was in line with previous studies in foreign countries
and in Vietnam, with presence of tet(M), tet (O) and erm(B) in 75%,
28.75% and 60% of isolated strains. The presence of these three genes
associated with resistance to erythromycin and tetracycline in our
patients. A novel finding in our study was that 78.75% of 80 samples
for bacterial isolation carried ant(6) gene – a resistance gene that has
never been mentioned in previous studies in Vietnam. Studies in
China have confirmed a critical role of resistance genes, including
tetO, ermB, ant6 and aphA, in the progress of antimicrobial
multidrug
resistance
to
tetracycline,
erythromycin
and

aminoglycosides.
4.5.2. Serotype of isolated strains
In our study, a majority of isolated strains were serotype 2,
constituting 92.25%. Serotype 14 and serotype 1/2 represented lower
proportions, 5% and 2.5%, and particularly, the latter was only presented
in septic shock patients. These findings are in good agreement with a
study by Nguyen Thi Hoang Mai et al in which serotype 2 was the
principal pathogen of meningitis caused by S. suis in adults in Southern
Vietnam, with 150 out of 151 patients infected with S. suis serotype 2 and

22


23
the only other patient infected with S. suis serotype 24 [8]. The majority of
previous studies in foreign countries and in Vietnam reported that S. suis
infection in human is predominantly caused by serotype 2, along with
other less common serotypes 1, 4, 14, 16.
4.5.3. Sequence type (ST) classification in isolated strains
81.25% of isolated strains in our study were S. suis ST1, and 9
were S. suis ST665, which differs from ST1 only in the allele profile of
gki gene. ST1 were reported to be the primary pathogen in adults in Asia
(Vietnam, Thailand, Hongkong, Japan, China, …), in Europe (along with
ST20 in Netherlands) and in South America (Argentina), whereas in
North America, among three prevailing sequence types of S. suis
serotype 2, ST25 and ST28 were more common than ST1. These two
sequence types were identified in two strains in our study.
4.5.4. Virulence genes
Approximately all isolated strains carried simultaneously all of
three virulence genes, mrp, epf and sly, with respective proportions of

98.75%, 97.5% and 96.25%. 95% and 72.5% of our patients carried a
combination of mrp+/epf+/sly+ and mrp+/epf+/sly+/SalKR+ genes,
which were all four virulence genes investigated in our study. The three
virulence genes, mrp, epf and sly were proven to be significant virulence
markers to distinguish S. suis virulent strains (existing) from less
virulent strains (non-existing) in Europe and Asia. Similar results were
also found in other studies in foreign countries and in Vietnam in which
a majority of isolated serotype 2 strains carried all of three
aforementioned virulence genes.
4.5.5. Identification of two novel strains ST25 and ST28 in Vietnam
It is fascinating that we discovered two strains with ST25 and
ST28, which has never been reported in Vietnam. According to several
recent studies, S. suis infection cases in Europe and Asia differs
markedly from those in North America, where ST25 and ST28 are
predominant. These two sequence types were also found in Thailand and
Japan. Owing to globalization, Vietnam is now globally trading its
domestic agricultural products with other regions, which is a possible
source of exchange of S. suis strains among different corners of the
world. In other words, the two strains ST25 and ST28 found in our study
could have an extrinsic origin. In phylogenetic tree, two strains ST25
and ST28 are distantly related to the other ST strains in our study. This
further consolidates our hypothesis that difference in sequence types
could have a great impact on homology between genome sequences, and
consequently, highlights the significance of identifying sequence types
of S. suis strains in order to assess their genomic homology and

23


24

evolution.
4.5.6. Identification of two strains with undefined sequence types
It is exceptional that we identified two strains with undefined
sequence types in this study. We certified that our sequencing database
were of good quality after a thorough re-examination to ascertain that
those two samples were not mistakenly sequenced to cause
misidentification of sequence types. The genome of those two strains
shared a number of fair similarities to that of the reference strain BM407
and other strains in our study. Those two strains, one isolated from a
patient with septic shock and meningitis and the other isolated from a
patient with meningitis only, all lacked several sequences in regions from
500 to 600 kbps and from 1000 to 1100 kbps compared to the reference
strain BM407. We are planning to contact with the authority of MLST
database to update sequence types of those two strains.
CONCLUSION

–

–

–

–

1. Clinical and laboratory characteristics and clinical syndromes
of S. suis infection in human.
Our study recruited 221 patients with three clinical syndromes of S. suis
infection, namely meningitis, sepsis with meningitis and septic shock.
This disease was predominantly found in men (90%) and in middle-aged
people between 40 and 60 years old (69.23%).

Common clinical symptoms of patients with S. suis infection were fever
(100%), headache (89.14%), shivering (68.33%). Meningitis was
typically presented with headache, nausea and altered mental status.
Hemorrhagic necrotic rash, myalgia and diarrhea were three
characteristic symptoms of septic shock by S. suis infection.
The mortality rate of S. suis infection was high (14.03%). 44.8% of our
patients developed hearing loss as a sequelae, and 1.36% required limb
amputation due to extremity gangrene.
2. Mortality prognostic factors in patients with S. suis infection.
Exposure through food ingestion, alcoholism, diarrhea,
hemorrhagic necrotic rash, hepatic failure, renal failure, and
disseminated intravascular coagulation were all mortality prognostic
factors in patients with S. suis infection.
3. Antibiotic resistance and several genotypes of S. suis.
All isolated S. suis strains were sensitive to penicilin, ampicilin,
ceftriaxone, linezolid, and vancomycin. Three antibiotics with high

24


25

–
–
–
–
–

resistance rate were erythromycin (76.3%), clindamycin (73.5%) and
tetracycline (97.6%).

Antibiotic resistance genes identified in our study were ant(6) (78.75%),
tet(B) (75%), tet(O) (28.75%) and erm(B) (60%).
The most common S. suis serotype was serotype 2 (92.5%), followed by
serotype 14 (5%) and serotype 1/2 (2.5%).
The most common sequence type was ST1 (81.25%), followed by
ST665 (11.25%) and ST105 (2.5%). There were two strains with
undefined sequence types.
Over 90% of isolated strains carried mrp, epf and sly virulence genes.
Two sequence types, namely ST25 and ST28, were identified in Vietnam
for the first time.

25