Pediatr Blood Cancer 2013;60:11541160

High Frequency of Streptococcal Bacteraemia During Childhood AML Therapy

Irrespective of Dose of Cytarabine
Katrine Helle Johannsen, Medical student,1 Mette Mller Handrup, MD, PhD,1 Birgitte Lausen, MD,
Henrik Schrder, MD, Med.Sci,1 and Henrik Hasle, MD, PhD1*
Background. High-dose cytarabine has been associated with a
high frequency of viridans group streptococcal (VGS) bacteraemia.
VGS bacteraemia causes considerable morbidity and mortality. The
Nordic Society of Paediatric Haematology and Oncology
(NOPHO)-AML protocols use higher cumulated doses of cytarabine
and more chemotherapy series (n 6) than other protocols with the
potential increased risk of severe VGS infections. Procedure. Medical records of all Danish children enrolled on the NOPHO-AML2004 protocol between January 2004 and September 2011 (n 45)
were retrospectively reviewed and all febrile episodes were registered. Results. Following 236 courses of chemotherapy, 112 positive blood cultures were registered. VGS were found in 38% of all
positive blood cultures. Infectious episodes with VGS resulted in
more days of fever and intravenous antibiotic therapy compared

Key words:

2
PhD,

with infections caused by other microorganisms. VGS were involved in 45% of all episodes categorized as sepsis (n 51) and
in 67% of life-threatening bloodstream infections (n 9). Penicillin
resistance was found in 40% of the VGS strains. No significant
difference was observed in the proportion of VGS isolated following
courses with low-dose cytarabine versus courses with high-dose
cytarabine. Conclusions. VGS were the most commonly isolated
pathogens causing the most severe infections and the majority of
life-threatening infections. A substantial proportion of the strains
were resistant to penicillin. The high rate of VGS seemed independent of high-dose cytarabine but was more likely caused by the
intensive chemotherapy treatment leading to severe mucositis and
neutropenia. Pediatr Blood Cancer 2013;60:11541160.
2012 Wiley Periodicals, Inc.

acute myeloid leukaemia; bacteraemia; cytarabine; infections; paediatric; viridans group streptococci

INTRODUCTION

METHODS

Intensified induction and consolidation chemotherapy have

improved the overall survival of children with acute myeloid
leukaemia (AML) over the last decades [1]. Increased risk of
infection caused by severe prolonged neutropenia and mucositis
have been the side effects of the intensive chemotherapy. Thus
infections remain an important cause of morbidity and mortality
of children with AML. An infection-related mortality of 5.411%
has been reported [2,3] and second to malignancy infections still
constitute the most common cause of death in children with
AML [4].
Since the 1980s a change in the balance of predominant
pathogens from gram-negative to gram-positive bacteria has
been observed. The increased occurrence of gram-positive bacteria has been explained by the increasingly potent chemotherapeutic regimens, causing more severe oropharyngeal mucositis and
neutropenia, a more resistant gram-positive oral flora caused by
the prophylactic use of sulfamethoxazole/trimethoprim and fluoroquinolones, and the increased use of central venous catheters
[57].
Specifically viridans group streptococci (VGS) have been
reported in paediatric AML [2,3] accounting for up to 25% of
all bacteraemic events in children with AML. VGS bacteraemia
are often more severe than bacteraemia caused by other bacteria
[8,9]. VGS strains are showing increased antibiotic resistance
among children with cancer in particular penicillin resistance is
increasing from 9% of 34 VGS isolates from children with cancer
published in 1991 to 37% penicillin resistance in 38 isolates in a
study published in 2006 [10,11].
High-dose cytarabine has been claimed a risk factor for VGS
bacteraemia [9,1215]. The Nordic Society of Paediatric Haematology and Oncology (NOPHO)-AML-2004 protocol uses higher
cumulated doses of cytarabine than other protocols (49.3 g/m2)
[16]. This study provides an evaluation of infections in children
treated according to the NOPHO-AML-2004 protocol and
includes infectious details of all Danish children with a diagnosis
of AML during an 8-year period.

Patients

2012 Wiley Periodicals, Inc.

DOI 10.1002/pbc.24448
Published online 31 December 2012 in Wiley Online Library
(wileyonlinelibrary.com).

All Danish children 014 years of age diagnosed with AML

between January 1, 2004 and September 1, 2011 were identified
from the NOPHO-AML database. Patients with Down syndrome,
acute promyelocytic leukaemia, isolated granulocytic sarcoma, or
secondary AML were excluded. Patients who died from AMLrelated bleedings within the first week during the first induction
course were also excluded. The patients were treated at one of the
two tertiary referral centres for childhood AML in Denmark: the
paediatric department of haematology and oncology at Aarhus
University Hospital Skejby (Centre A) and at Copenhagen University Hospital Rigshospitalet (Centre B).

Treatment
NOPHO-AML-2004 protocol. All patients were treated
according to the NOPHO-AML-2004 protocol [16]. Treatment
details are summarized in Table I. Patients were treated according
to standard-risk or high-risk. Risk grouping was based on treatment response and cytogenetics. High-risk criteria were bone
marrow (BM) blast cells >15% on Day 15 after first induction
course or not in complete remission after second induction course.
Patients with 11q23 rearrangements other than t(9;11) were
treated as high-risk until June 2009. Patients with FLT3-ITD

Department of Paediatrics, Aarhus University Hospital Skejby, Aarhus, Denmark; 2Department of Paediatrics and Adolescent Medicine,
Rigshospitalet, University of Copenhagen, Copenhagen, Denmark
Grant sponsor: Danish Childhood Cancer Foundation.
Conflict of interest: Nothing to declare.
*Correspondence to: Prof. Henrik Hasle, MD, PhD, Department of
Paediatrics, Aarhus University Hospital Skejby, Brendstrupgaardsvej
100, Aarhus N DK-8200, Denmark. E-mail: hasle@dadlnet.dk
Received 18 September 2012; Accepted 27 November 2012

Streptococcal Bacteraemia During AML Therapy

1155

TABLE I. NOPHO-AML-2004 Treatment Elements

Dose (mg/m2)

Administration

Day

200
12
100
100

24-hour IV
4-hour IV
24 hour IV
Oral every 12 hours

14
2, 4 and 6
14
14

100
12

24-hour IV
30-minute IV

15
13

1,000
10

2-hour IV every 12 hours

30-minute IV

13
35

2,000
100

2-hour IV every 12 hours

60-minute IV

13
25

3,000

2-hour IV every 12 hours

13

2,000
100

2-hour IV every 12 hours

60-minute IV

13
25

Age-adjusted dose 612 mg

Intrathecal injection

On the first day of all courses

Course name and drugs

Induction
AIET
Cytarabine
Idarubicin
Etoposide
6-thioguanin
AM
Cytarabine
Mitoxantrone
Consolidation
HA1M
Cytarabine
Mitoxantrone
HA2E1
Cytarabine
Etoposide
HA3
Cytarabine
HA2E2
Cytarabine
Etoposide
CNS therapy
Methotrexate

IV: intravenous. Patients younger than 1 year of age or weight less than 10 kg received doses calculated according to bodyweight
(1 m2 30 kg).

mutations without simultaneous NPM1 mutations were treated as

high-risk from January 2011. All other patients were treated as
standard-risk patients [16]. Patients in both risk groups received
two induction courses: AIET and AM (Table I). Patients in remission after AIET received AM following haematological recovery.
Patients with blast cells >5% on Day 15 proceeded directly to
AM unless therapy-related complications necessitated a delay.
Patients not in remission after the two induction courses
(n 4) were given fludarabine, cytarabine, and granulocyte colony stimulating factor (FLAG) as a third induction course, whereas those in remission proceeded to consolidation therapy.
Standard-risk patients were given a total of four consolidation
courses (HA1M, HA2E1, HA3 and HA2E2). Each course was
started immediately after clinical and haematological recovery
(absolute neutrophil count (ANC) > 1.0
109/L and platelets
>80
109/L). After the last course of consolidation patients
were eligible for randomisation to receive either two courses of
gemtuzumab ozogamicin (GO) or no further therapy. This randomisation stopped in December 2010 [17]. Patients in the highrisk group proceeded to allogeneic stem cell transplantation
(SCT) as soon as possible after first consolidation. If no donor
was available patients with high-risk disease received the same
consolidation therapy as standard-risk patients.
Supportive care. A central venous catheter (CVC) was
inserted before administration of chemotherapy. The preferred
type of CVC in Centre A was a total implantable subcutaneous
device (Port-A-Cath) with puncture into the jugular vein where
Centre B used a tunnelled catheter with two external lines (Hickman) and puncture into the subclavian vein.
Patients received prophylactic sulfamethoxazole/trimethoprim
against pneumocystis jiroveci pneumonia and prophylactic
Pediatr Blood Cancer DOI 10.1002/pbc

fluconazole against systemic fungal infections until 1 month after

the last course. No prophylactic antibacterial antibiotics or acyclovir were used. The use of granulocyte colony-stimulating factor (G-CSF) was generally not recommended. Tooth brushing
with soft brush and fluoride toothpaste and mouth rinses with
chlorhexidine were performed twice daily under morphine coverage if needed. Patients were transferred to intensive care unit if
they needed pressure support or respiratory assistance.
All patients with neutropenic fever were treated with intravenous empiric antibiotic therapy. Neutropenic fever was defined as
ANC  0.5
109/L and axillary temperature 388C twice within 2 hours or 38.58C once. First line antibiotic therapy in Centre
A consisted of piperacillintazobactam 300 mg/kg/day in three
doses and gentamicin 6 mg/kg/day as one daily dose for at least
3 days. Metronidazole 3050 mg/kg/day in three doses was added
in case of persistent fever or abdominal symptoms. First line
antibiotic therapy in Centre B was meropenem, 60 mg/kg/day
in three doses, and gentamicin, 6 mg/kg/day as one daily dose,
for at least 3 days. In case of persistent fever ciprofloxacin,
30 mg/kg/day in three doses, and/or vancomycin, 40 mg/kg/day
in two doses, were added. Antibiotic therapy was adjusted according to blood cultures and the clinical response of the patient.
Empiric antifungal coverage consisting of liposomal amphotericin
B, 3 mg/kg/day as one daily dose, was mandated in case of
persistent fever, poor general condition or increased C-reactive
protein (CRP) despite 46 days of antibiotic therapy. Antibiotic
treatment was ceased when fever had been absent for more than
2 days, neutrophils were increasing, and CRP was decreasing
significantly. Antibiotics were continued for at least 7 days after
a positive blood culture. Patients were discharged in the absence
of infectious signs regardless of the neutrophil count.

1156

Johannsen et al.

Classification of Infectious Episodes

RESULTS

Each febrile episode requiring hospitalisation and empiric intravenous antibiotic therapy was counted as a separate infection
and categorized as a bloodstream infection (BSI), local infection,
or fever of unknown origin (FUO). Bloodstream infections were
defined as fever with at least one positive blood culture of bacteria
or fungi [18]. Multiple organisms from the same patient within
the same treatment block were considered separate infections,
except for polymicrobial bloodstream infections, which were defined as two or more microorganisms isolated from a single blood
culture [19]. Blood cultures with the same microorganism were
counted as different infections if they occurred more than 7 days
apart [2]. Sepsis was defined as Systemic Inflammatory Response
Syndrome (SIRS) caused by a bloodstream infection. SIRS was
diagnosed if the child fulfilled the following criteria for paediatric
sepsis: core temperature 38.58C and one of the following three
criteria: tachycardia defined as a mean heart rate 2 SD above
normal for age, bradycardia defined as a mean heart rate <10th
percentile for age or mean respiratory rate 2 SD above normal
for age [20]. Leukocyte count was disregarded in the SIRS definition since all patients were neutropenic following chemotherapy. Life-threatening infection was defined as a severe infection
causing admission to the intensive care unit. Clinical, radiological
or microbiological proven signs of an infection without a concurrent bloodstream infection defined a local infection [18]. FUO
was defined as no microbiologically documented infection as well
as no localized symptoms [3] except diarrhoea, mucositis, and
stomatitis, which were accepted as part of FUO. Febrile episodes
during cytarabine infusion without any pathogen identified were
considered cytarabine-related and excluded from the study.
Data were collected by review of each patients medical record. For every febrile episode the following data were listed: first
day of fever, number of days with fever, length of intravenous
antibiotic treatment, neutrophils at initiation of antibiotic treatment and the day of collecting a positive blood culture. Data on
mucositis were not available. Patient baseline data at diagnosis
were obtained from the NOPHO-AML database.

Patients and Therapy Courses

Data Analysis
Infections were registered after each course of chemotherapy.
The time period of risk was defined from the day chemotherapy
was initiated until the day before the next course of chemotherapy. After the last course of chemotherapy infections were registered until haematological recovery. Infections following FLAG
and GO-therapy were not included in this study due to the few
patient treated with FLAG (n 4) and the recently reported GO
data [17]. Patients who proceeded to SCT after initial therapy or
patients who relapsed were censored at the time of the event.
Descriptive statistics were reported as absolute frequencies
and percentages. Continuous variables such as days of fever,
intravenous antibiotic therapy and days to last documented absolute neutrophil count (ANC) < 0.5
109/L were described as
median values and interquartile ranges (IQR) and compared by
the MannWhitney U-test. Categorical variables were compared
by Pearsons chi-square test and Fishers exact test as appropriate.
All P values were two-sided and considered statistical significant
when <0.05. All statistical analyses were performed using IBM
SPSS Statistics, Version 20.
Pediatr Blood Cancer DOI 10.1002/pbc

We enrolled all 45 Danish children with a diagnosis of AML

from January 2004 to September 2011. Two patients died from
AML-related bleedings within the first week after the induction
course was initiated. The remaining 43 patients were included in
the final analyses. The clinical characteristics of the 43 children
are illustrated in Table II. The median age was 4.0 years (range:
014 years). Thirty-six patients received all six courses of chemotherapy. One patient relapsed after AIET. One patient died
following the third induction course (FLAG). Five patients proceeded to SCT, two after a third induction course (FLAG), another two after HA2E1, and one after HA3. The present study
included 236 courses of chemotherapy (Table III).

Febrile Episodes
Following the 236 courses of chemotherapy 268 infections
were registered. Most infections were classified as FUO (44%,
n 119). BSI constituted 37% (n 99) of the infections, but
comprised 48% (n 262) of the days of fever and 42%
(n 266) of the days of antibiotic therapy. Local infections constituted 19% (n 50) of the infections.
Fever occurred a median of 12 days after initiation of chemotherapy both for BSI (range: 132) and local infections (range: 0
28) and a median of 11 days after initiation of chemotherapy for
FUO (range: 237). The median number of days of fever was
3 days (range: 118 days) for BSI, compared with 1 day (range:
021 days) for FUO, and 2 days (range: 011 days) for local
infections. The median number of days of intravenous antibiotic
TABLE II. Clinical Characteristics of the 43 Children With
AML Included in the Study
Characteristics
Gender
Male
Female
Age (years)
01
29
1014
WBC (109/L)
09.9
1099
100
FAB classification
Mixed phenotype leukaemia
M0
M1
M2
M4
M5
M6
M7
Cytogenetics
t(8,21) or inversion (16)
Normal
Other
WBC, white blood cell; FAB, French-American-British.

No. (%)
22 (51)
21 (49)
16 (37)
19 (44)
8 (19)
11 (26)
27 (63)
5 (12)
1 (2)
1 (2)
3 (7)
10 (23)
7 (16)
15 (35)
2 (5)
4 (9)
8 (16)
8 (19)
27 (65)

Streptococcal Bacteraemia During AML Therapy

1157

TABLE III. Infectious Complications Following Each Course of Therapy in a Total of 236 Episodes

Patients admitted with at least one

infection, n
Infections
In total, n
BSI, n
BSI categorized as sepsis, n
Life-threatening, n
Days from start of chemotherapy to
last ANC < 0.5
109/L
before recovery
Median (range)
n
Days of fever
Median (range)
n
Days of antibiotic therapy
Median (range)
n
Patients who received G-CSF, n
Patients who received intravenous
antifungal therapy, n

AIET
(n 43)

AM
(n 42)

HA1M
(n 39)

HA2E1
(n 39)

HA3
(n 37)

HA2E2
(n 36)

42

39

36

37

31

34

62
17
5
2

46
14
9
1

45
22
11
4

44
18
10
3

34
16
8
2

37
12
8
1

27 (14141)
42

24 (1438)
42

28 (2048)
38

21 (1732)
39

21 (1329)
37

21 (1430)
35

4 (021)
59

1 (012)
44

2 (018)
45

2 (111)
44

1 (010)
34

2 (010)
36

9 (233)
61
3
20

7 (134)
45
2
3

10 (226)
45
4
11

7 (323)
44
2
8

6.5 (216)
34
1
6

7 (119)
37
1
4

BSI, bloodstream infection; ANC, absolute neutrophil count; G-CSF, granulocyte colony-stimulating factor. Antifungal therapy: amphotericin
B, fluconazole, voriconazole, caspofungin.

therapy for BSI was 9 days (range: 134 days), for FUO 7 days
(range: 233 days), and for local infections 6 days (range: 132
days).
The highest rate of infections occurred after AIET (Table III).
FUO compromised 56% of these infections. FUO constituted also
most of the infections after AM (Fig. 1). BSI peaked after first
consolidation therapy, and was the leading cause of infection after

HA1M. The number of BSI decreased through the consolidation

courses (Fig. 1). Neutropenia was present in 95% (n 264) of
the infectious episodes. The median number of days from the start
of chemotherapy course to last documented ANC < 0.5
109/L
before recovery and the median number of days of intravenous
antibiotic therapy correlated (Table III).

Bloodstream Infections
BSI caused 99 infectious episodes (Table IV). A single microorganism was isolated in 88 episodes and polybacteraemia in 11
episodes (Table IV). The frequency of BSI after each therapy
course varied from 33% to 56%. Fifty-two percentages of BSI
were categorized as sepsis (n 51, Table III). A total of 112
positive blood cultures were registered. Gram-positive bacteria
accounted for 65% of all bacteria isolated (Table IV). VGS and
coagulase-negative staphylococci were the most frequently isolated gram-positive bacteria. Pseudomonas species, Escherichia coli
and Klebsiella pneumonia constituted the most commonly isolated gram-negative bacteria. Only three fungaemia occurred, two by
Candida albicans and one by Candida parapsilosis. All three
were on fluconazole prophylaxis. We did not identify any significant difference between the infection rate or pattern between
Centres A and B.

Viridans Group Streptococci (VGS)

Fig. 1. Distribution of infections (BSI, FUO and local infections)

related to therapy course (the therapy course abbreviations are
explained in Table I); BSI, bloodstream infection; FUO, fever of
unknown origin.
Pediatr Blood Cancer DOI 10.1002/pbc

The dominating aetiology of BSI was bacteraemia with VGS

constituting 38% of all positive blood cultures (Table IV). VGS
were involved in 41% of infectious episodes categorized as a BSI
and in 45% of the episodes categorized as sepsis. Infections with
VGS resulted in more days of fever (median: 4, range: 018) and
intravenous antibiotic therapy (median: 10 days, range: 234)

1158

Johannsen et al.

TABLE IV. Number of Patients With One or More Bloodstream Infection After Each Course of Chemotherapy and the Type of
Microorganisms
AIET
n 43 No.

AM
n 42 No.

HA1M
n 39 No.

HA2E1
n 39 No.

HA3
n 37 No.

HA2E2
n 36 No.

In total
No.

Infection with a bloodstream infection

1 organism
2 organisms
3 organisms

17
13
4

14
12

22
22
0

18
16
2

16
14
2

12
11
1

99
88
9
2

Gram-positive
Bacillus spp.
Clostridium spp.
Coagulase-negative
Staphylococci
Coryneform spp.
Enterococcus faecium
Micrococcus spp.
Mycobacterium spp.
Streptococcus pneumonia
Staphylococcus aureus
Viridans group streptococci

14

12
1

18
2

13

10
1

71
4
1

2
1

2
6

3
2

1
1

12

9
2
4
4
1
1
2
43

Gram-negative
Acinetobactor spp.
Citrobactor spp.
Escherichia coli
Enterobactor cloacae
Flavobacterium spp.
Klebsiella pneumonia
Neisseria spp.
Pseudomonas aeruginosa
Stenotrophomonas malthophilia
Other pseudomonas spp.

1
1

1
1

1
1

1
1

7
1

1
1

8
1
1
1

3
1

3
1
1
2
1
1

38
2
1
9
3
1
7
5
3
4
3

Fungus
Candida albicans
Candida parapsilosis

3
2
1

3
2
1

No, number; spp, species. Viridans group streptococci were reported either: non-specifically; viridans group streptococci (n 20) or more
specifically; streptococcus mitis (n 11), streptococcus oralis (n 9), streptococcus salivarius (n 1), streptococcus sanguinis (n 2).

compared with other microorganisms (median 2 days of fever,

range: 115, and 8 days of antibiotic therapy, range: 129), however, the differences were not statistically significant.
The percentage of patients with VGS bacteraemia was
highest after HA2E1 (31%). No significant difference was
observed in the proportion of VGS isolated following courses
with low-dose cytarabine (induction; <0.5 g/m2 per dose) and
courses with high-dose cytarabine (consolidation; 1 g/m2 per
dose). VGS occurred in 18% of courses with low-dose cytarabine
versus 19% of courses with high-dose cytarabine (P 1.00;
Table IV).
Twenty-six patients experienced VGS bacteraemia. Fourteen
of these patients experienced more than one VGS bacteraemia.
Out of these 14 patients: one experienced two VGS bacteraemia
following the same course, nine experienced a second VGS bacteraemia in a subsequent course and another following three
consecutive courses. The last three children had a second VGS
infection in a non-consecutive course. This risk of developing a
second VGS bacteraemia seemed independent of the course after
which the first VGS developed. There were no associations between gender, age, white blood cell count (WBC), and cytogenetics and the risk of VGS infections.
Pediatr Blood Cancer DOI 10.1002/pbc

Antimicrobial Susceptibility of VGS

Table V shows the VGS strains susceptibility to six types of
antibiotics. Penicillin resistance was shown in 40% of VGS
strains. Ninety-four percentages of VGS were susceptible to
piperacillintazobactam. All VGS strains were intermediately susceptible to gentamicin and all VGS strains were vancomycin and
meropenem susceptible.
The ratio of penicillin resistance did not increase over time
when analysed for patients with more than one VGS strain isolated. No increase in the penicillin resistance was observed throughout the study period and no increase in the penicillin resistance
was observed throughout the courses of chemotherapy (the percentage of resistant VGS strains varied over the six courses from
67%, 22%, 44%, 33%, 40%, and 50%).

Life-Threatening Infections
Thirteen infectious episodes were categorized as life-threatening (Table III) of which nine were caused by bacteraemia. Polybacteraemia were the cause of two life-threatening infections: one
with VGS and Escherichia coli and the other with Klebsiella
pneumonia, Enterobactor cloacae and Pseudomonas putida.

Streptococcal Bacteraemia During AML Therapy

1159

TABLE V. Antimicrobial Susceptibility of Viridans Group Streptococci (VGS)

Antibiotic

Number of VGS strains tested

Susceptible %

Intermediate %

Resistant %

43
37
17
19
17
17

56
100
88
0
100
94

5
0
6
100
0
6

40
0
6
0
0
0

Penicillin
Vancomycin
Ceftriaxone
Gentamicin
Meropenem
Piperacillintazobactam

VGS bacteraemia caused five life-threatening infections and Stenotrophomonas malthophilia caused two. Four life-threatening
infections were categorized as FUO since no microorganism
was isolated. Two of these occurred after AIET. None of the
patients died due to infections.

DISCUSSION
This study evaluated infections in children with AML treated
according to the NOPHO-AML-2004 protocol. NOPHO AML
protocols include higher cumulated doses of cytarabine and
more chemotherapy series (n 6) than other protocols with the
potential increased risk of severe VGS infections. The study
population was homogeneous, only including children with
AML, receiving the same six courses of chemotherapy and the
same antifungal prophylaxis, but the study was limited by the low
number of patients.
Bacteraemia were the major contributor to morbidity and prolonged hospitalisation, causing the majority of life-threatening
infections, fever, and intravenous antibiotic therapy. A total of
112 positive blood cultures were registered after 236 courses of
chemotherapy. Gram-positive bacteria constituted 65% of positive
blood cultures, which correlates with previous reports [2,3]. There
may be an overestimate of the number of gram-positive organisms, since we were not able to distinguish between likely contaminants versus invasive infections. VGS and Coagulase negative
staphylococci were the most common gram-positive bacteria and
Pseudomonas spp., Escherichia coli and Klebsiella pneumonia the
most frequent gram-negative bacteria isolated in bloodstream
infections. This is consistent with previous reports [2,3]. VGS
were the most frequent types of bacteria isolated and associated
with a high frequency of complications which is also consistent
with previous reports [8,9].
High-dose cytarabine has been reported to be a risk factor of
VGS infections [9,12,13]. However, in this study we did not find
any significant association between the frequency of VGS bacteraemia and treatment with high-dose cytarabine. The second lowest rate of VGS-bacteraemia was found following the highest dose
of cytarabine (HA3: six doses of 3 g/m2). Other risk factors for
developing VGS bacteraemia include oropharyngeal mucositis,
severe neutropenia, the prophylactic use of sulfamethoxazole/trimethoprim and previous VGS infection [12,21,22]. Sulfamethoxazole/trimethoprim was used in this study as prophylaxis against
pneumocystis jiroveci, a drug believed to increase the frequency
of VGS infections by allowing the overgrowth of resistant grampositive organisms such as VGS [12,21,23,24]. Mucositis and oral
ulcers are common toxicities following high-dose chemotherapy
and are the possible origin of septicaemia. Intraoral ulcers are
often due to infection with herpes simplex virus [25]. Ringden
Pediatr Blood Cancer DOI 10.1002/pbc

et al. [26] observed a decrease in the incidence of VGS bacteraemia among the recipients of bone marrow transplants after the
introduction of acyclovir for prophylaxis of herpetic lesions.
Hence prophylaxis with acyclovir may be considered for patients
with AML during chemotherapy [25].
VGS are normally present in the gastrointestinal tract and
oropharyngeal mucosa. Several antineoplastic compounds induce
severe mucositis creating a potential entrance for invasive VGS
strains. As mentioned earlier, HA3 was the course with the second
lowest rate of VGS but HA3 was also the only course where
cytarabine was given as mono-chemotherapy, which may be
associated with less mucositis. Thus, the high frequency of
VGS infections seems to be primarily caused by the intensive
treatment causing severe neutropenia and oropharyngeal mucositis and maybe in addition the prophylactic use of sulfamethoxazole/trimethoprim. High-dose cytarabine did not play a major role
as risk factor for VGS infections.
Of the 43 VGS isolated 40% were penicillin resistant. Bruckner
et al. [10] found 37% of 38 isolates to be penicillin resistant, which
supports the high proportion of penicillin resistant VGS strains
among children with cancer. Nevertheless first line empiric antibiotic
therapy in both Centres A and B seemed appropriate, since all
VGS strains were intermediately susceptible to gentamicin and fully
susceptible to meropenem and susceptible to piperacillin-tazobactam
in 94%. All 37 tested strains were vancomycin susceptible.
In contrast to the high mortality from VGS infection in the
AML-BFM 93 trial [3], no patients died in the present study but
the high morbidity associated with VGS bacteraemia asks for
potential prophylactic measures. Vancomycin has been suggested
as prophylaxis against VGS [11] and Kurt et al. [27] reported an
almost complete eradication of VGS sepsis and more than 90%
reduction in bacterial sepsis when prophylactic intravenous vancomycin with oral ciprofloxacin or a cephalosporin were administered by home infusion during neutropenia. Brunet et al. [18]
reported VGS to account for only 5.1% of all pathogens isolated
by the use of prophylactic mouth washing with vancomycin in
children undergoing therapy of AML. However, prolonged administration of broad-spectrum antibiotics raises concern about the
development of resistant microbiological flora.
In conclusion, we found VGS to be the most commonly isolated pathogens causing the most severe infections and the majority of life-threatening infections, with a substantial proportion of
the strains being resistant to penicillin. The high rate of VGS
seemed independent of high-dose cytarabine but was more likely
caused by the prophylactic use of sulfamethoxazole/trimethoprim
and the intensive chemotherapy treatment leading to severe mucositis and neutropenia. Although the VGS infections were manageable in this protocol-setting, a high rate of morbidity and
mortality have been reported for children with VGS infections

1160

Johannsen et al.

[3], emphasizing the importance of early treatment or prevention

of VGS infection.

REFERENCES
1. Lie SO, Abrahamsson J, Clausen N, et al. Long-term results in children with aml: NOPHO-AML study
group-report of three consecutive trials. Leukemia 2005;19:20902100.
2. Sung L, Lange BJ, Gerbing RB, et al. Microbiologically documented infections and infection-related
mortality in children with acute myeloid leukemia. Blood 2007;110:35323539.
3. Lehrnbecher T, Varwig D, Kaiser J, et al. Infectious complications in pediatric acute myeloid leukemia:
Analysis of the prospective multi-institutional clinical trial aml-bfm 93. Leukemia 2004;18:7277.
4. Molgaard-Hansen L, Mottonen M, Glosli H, et al. Early and treatment-related deaths in childhood
acute myeloid leukaemia in the nordic countries: 19842003. Br J Haematol 2010;151:447459.
5. Viscoli C, Varnier O, Machetti M. Infections in patients with febrile neutropenia: Epidemiology,
microbiology, and risk stratification. Clin Infect Dis 2005;40:S240S245.
6. Ramphal R. Changes in the etiology of bacteremia in febrile neutropenic patients and the susceptibilities of the currently isolated pathogens. Clin Infect Dis 2004;39:S25S31.
7. Zinner SH. Changing epidemiology of infections in patients with neutropenia and cancer: Emphasis on
gram-positive and resistant bacteria. Clin Infect Dis 1999;29:490494.
8. Sung L, Buxton A, Gamis A, et al. Life-threatening and fatal infections in children with acute myeloid
leukemia: A report from the childrens oncology group. J Pediatr Hematol Oncol 2012;34:e30e35.
9. Gamis AS, Howells WB, DeSwarte-Wallace J, et al. Alpha hemolytic streptococcal infection during
intensive treatment for acute myeloid leukemia: A report from the childrens cancer group study CCG289. J Clin Oncol 2000;18:18451855.
10. Bruckner LB, Korones DN, Karnauchow T, et al. High incidence of penicillin resistance among alphahemolytic streptococci isolated from the blood of children with cancer. J Pediatr 2002;140:2026.
11. Bruckner L, Gigliotti F. Viridans group streptococcal infections among children with cancer and the
importance of emerging antibiotic resistance. Semin Pediatr Infect Dis 2006;17:153160.
12. Weisman SJ, Scoopo FJ, Johnson GM, et al. Septicemia in pediatric oncology patients: The significance of viridans streptococcal infections. J Clin Oncol 1990;8:453459.

Pediatr Blood Cancer DOI 10.1002/pbc

13. Rossetti F, Cesaro S, Putti MC, et al. High-dose cytosine arabinoside and viridans streptococcus sepsis
in children with leukemia. Pediatr Hematol Oncol 1995;12:387392.
14. Engelhard D, Elishoov H, Or R, et al. Cytosine arabinoside as a major risk factor for streptococcus
viridans septicemia following bone marrow transplantation: A 5-year prospective study. Bone Marrow
Transplant 1995;16:565570.
15. Paganini H, Staffolani V, Zubizarreta P, et al. Viridans streptococci bacteraemia in children with fever
and neutropenia: A casecontrol study of predisposing factors. Eur J Cancer 2003;39:12841289.
16. Abrahamsson J, Forestier E, Heldrup J, et al. Response-guided induction therapy in pediatric acute
myeloid leukemia with excellent remission rate. J Clin Oncol 2011;29:310315.
17. Hasle H, Abrahamsson J, Forestier E, et al. Gemtuzumab ozogamicin as postconsolidation therapy
does not prevent relapse in children with aml: Results from NOPHO-AML 2004. Blood 2012;120:978
984.
18. Brunet AS, Ploton C, Galambrun C, et al. Low incidence of sepsis due to viridans streptococci in a tenyear retrospective study of pediatric acute myeloid leukemia. Pediatr Blood Cancer 2006;47:765772.
19. Frommell GT, Todd JK. Polymicrobial bacteremia in pediatric patients. Am J Dis Child 1984;138:266
269.
20. Goldstein B, Giroir B, Randolph A. International pediatric sepsis consensus conference: Definitions for
sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005;6:28.
21. Bochud PY, Calandra T, Francioli P. Bacteremia due to viridans streptococci in neutropenic patients: A
review. Am J Med 1994;97:256264.
22. Reilly AF, Lange BJ. Infections with viridans group streptococci in children with cancer. Pediatr Blood
Cancer 2007;49:774780.
23. Elting LS, Bodey GP, Keefe BH. Septicemia and shock syndrome due to viridans streptococci: A case
control study of predisposing factors. Clin Infect Dis 1992;14:12011207.
24. Hammond SP, Baden LR. Antibiotic prophylaxis during chemotherapy-induced neutropenia for
patients with acute leukemia. Curr Hematol Malig Rep 2007;2:97103.
25. Bergmann OJ, Ellermann-Eriksen S, Mogensen SC, et al. Acyclovir given as prophylaxis against oral
ulcers in acute myeloid leukaemia: Randomised, double blind, placebo controlled trial. BMJ
1995;310:11691172.
26. Ringden O, Heimdahl A, Lonnqvist B, et al. Decreased incidence of viridans streptococcal septicaemia
in allogeneic bone marrow transplant recipients after the introduction of acyclovir. Lancet 1984;1:744.
27. Kurt B, Flynn P, Shenep JL, et al. Prophylactic antibiotics reduce morbidity due to septicemia during
intensive treatment for pediatric acute myeloid leukemia. Cancer 2008;113:376382.