1 epidemiology and antifungal susceptibility of bloodstream 1 fungal isolates in pediatric patients

22 Pages·2011·0.39 MB·English

JCM Accepts, published online ahead of print on 19 October 2011
J. Clin. Microbiol. doi:10.1128/JCM.05474-11
Copyright © 2011, American Society for Microbiology and/or the Listed Authors/Institutions. All Rights Reserved.
1

1 EPIDEMIOLOGY AND ANTIFUNGAL SUSCEPTIBILITY OF BLOODSTREAM
2 FUNGAL ISOLATES IN PEDIATRIC PATIENTS: A SPANISH MULTICENTER
3 PROSPECTIVE SURVEY
4
5 AUTHORS:
6 Javier Pemán, PhD, MD1; Emilia Cantón, PhD1; María José Linares-Sicilia,
7 PhD2; Eva María Roselló, MD3; Nuria Borrell, PhD4; María Teresa Ruiz-Pérez- D
o
8 de-Pipaon PharmD, PhD5; Jesús Guinea, PharmD, PhD6; Julio García, PhD7; w
n
9 Aurelio Porras, MD8; Ana María García-Tapia, PharmD9; Luisa Pérez-del- lo
a
10 Molino, MD10; Anabel Suárez, MD11; Julia Alcoba, PhD12; Inmaculada García- d
e
11 García, MD13. d
f
r
o
m
12

h
t
t
13 1Hospital Universitario La Fe, Valencia; 2Hospital Reina Sofía, Córdoba; p
:
/
/
14 3Hospital Valle Hebrón, Barcelona; 4Hospital Son Dureta, Palma de Mallorca; jc
m
15 5Hospital Virgen del Rocío, Seville; 6Hospital Gregorio Marañón, 7Madrid; .
a
s
16 Hospital La Paz, Madrid; 8Hospital Carlos Haya, Málaga; 9Hospital Universitario m
.
17 Puerta del Mar, Cádiz; 10Complejo Hospitalario, Santiago de Compostela; o
r
g
18 11Hospital Universitario Virgen de la Macarena, Seville; 12Hospital Universitario /
o
19 N. S. de la Candelaria, Tenerife; 13Hospital Clínico Universitario, Salamanca. n
A
p
r
20 il
1
2
21 CORRESPONDING AUTHOR: ,
2
0
1
22 Javier Pemán, PhD, MD 9

b
y
23 Servicio Microbiología, Hospital Universitario La Fe
g
u
24 Avda Campanar, 21, 46009 Valencia, Spain e
s
t
25 Tel: [phone]; Fax: [phone]
26 e-mail: [email protected]
27
28 No honorarium, grant, or other form of payment was given to any author to
29 produce the manuscript.
2

30
31 SHORT TITLE:
32 Epidemiology and antifungal susceptibility of fungemia isolates in pediatric
33 patients in Spain
34
D
o
35 KEY WORDS: w
n
lo
36 Fungemia, Candida, pediatric, epidemiology, neonates, in vitro susceptibility, a
d
e
37 antifungal agents, amphotericin B, fluconazole, itraconazole, posaconazole, d

f
38 voriconazole, anidulafungin, caspofungin, micafungin, flucytosine ro
m

39 h
t
t
p
:
40 ABBREVIATIONS: //
jc
m
41 CLSI: Clinical Laboratory Standards Institute .a
s
m
42 ICU: Intensive Care Unit .
o
r
g
/
43 SOT: solid organ transplant
o
n

A
44 HSCT: hematopoietic stem cell transplant
p
r
il
45 MIC: minimal inhibitory concentration 1
2
,

2
46 0
1
9
47 b
y

g
48 u
e
s
t
49
50
51
52
3

53 ABSTRACT
54 Data on fungemia epidemiology and antifungal susceptibility of isolates from
55 children are scarce, leading frequently to pediatric empiric treatment based on
56 available adult data. The present study was designed to update the
57 epidemiological, mycological and in vitro susceptibility data on fungal isolates
58 from children with fungemia in Spain.
D
o
w
59 All fungemia episodes were identified prospectively by blood culture over 13
n
60 months at 30 hospitals. Tests of susceptibility to amphotericin-B, flucytosine, lo
a
d
61 fluconazole, itraconazole, posaconazole, voriconazole, anidulafungin, e
d
62 caspofungin, and micafungin were performed at participant institutions by a f
r
o
63 microdilution colorimetric method. New species-specific clinical breakpoints for m

64 fluconazole, voriconazole and echinocandins were also applied. h
t
t
p
:
/
65 A total of 203 episodes of fungemia in 200 children were identified. A higher /jc
m
66 proportion of fungal isolates was from general wards than intensive care units
.
a
67 (ICU). Candida parapsilosis (46.8%) followed by Candida albicans (36.5%), s
m
68 Candida tropicalis (5.9%), Candida glabrata (3.9%), and Candida guilliermondii .o
r
g
69 (2.5%) were the leading species. C. parapsilosis was the predominant species
/

o
70 except in neonates. C. albicans was the most frequent in Neonatal ICU settings n

A
71 (51.9%). Intravascular catheter (79.3%), surgery (35%), prematurity (30%), and p
r
72 neutropenia (11%) were the most frequent predisposing factors. Most Candida il
1
2
73 isolates (95.1%) were susceptible to all antifungals. Applying the new species- ,

2
74 specific clinical breakpoints, all C. parapsilosis isolates were susceptible to 0
1
9
75 echinocandins except one, which was micafungin resistant.
b
y

76 This is the largest published series of fungemia episodes in the pediatric g
u
e
77 setting. C. parapsilosis is the most prevalent species in Spain, followed by C. s
t
78 albicans and C. tropicalis. Resistance to azole and echinocandin agents is
79 extremely rare among Candida species. Fluconazole-resistance rate in Spain
80 has decreased in the last ten years.
81
4

82 INTRODUCTION
83 The incidence of bloodstream fungal infection, in adults and pediatric patients,
84 has risen in the last decade as a result of a combination of several factors such
85 as: increased use of central venous catheters, extensive use of parenteral
86 nutrition, mucosa alteration or prolonged neutropenia due to more aggressive
87 antineoplastic treatments, ever more aggressive surgery and instrumentation
D
88 techniques and the widespread use of broad spectrum antibiotics (8, 9, 17, 18, o
w
n
89 24, 25). Currently, Candida spp. has become the fourth most frequent causal
lo
a
90 microorganism of nosocomial sepsis (17, 25). Furthermore, monitoring
d
e
91 programs have detected an increase in the prevalence of infections caused by d

f
92 non-Candida albicans (essentially Candida parapsilosis, Candida glabrata and ro
m
93 Candida krusei) and other yeast genera (1, 6, 20). Additionally, significant
h
t
94 regional differences have been reported in the distribution and pattern of tp
:
/
95 susceptibility to antifungal agents among the different species (21, 24). /jc
m
96 Unfortunately, data about epidemiology or antifungal susceptibility patterns in .a
s
m
97 pediatric patients with fungemia are scarce and, frequently, empiric treatment in
.
o
98 children with suspicion of invasive fungal infection has to be instituted r
g
/
99 extrapolating information from adult patients. The need to update the o
n
100 epidemiological and mycological profile in pediatric patients with fungemia was A
p
101 one of the aims of the FUNGEMYCA survey, developed prospectively in Spain r
il
1
102 during 2009. In the present descriptive study, we report the results of this
2
,
103 survey in the Spanish pediatric population along with the susceptibility patterns 2
0
104 of the fungal isolates recovered, comparing data with a previous study carried 1
9

105 out from 1997-1999 in Spain (14). b
y

g
u
106 METHODS e
s
t
107 Study design The FUNGEMYCA survey was a prospective, sequential,
108 hospital-population based study. Thirty Spanish institutions, widely distributed
109 throughout the country including the Canary and Baleares Islands, participated
110 in the study. Participating hospitals were required to collect sequentially and
111 identify the fungal isolates from blood cultures and to complete, for each
112 fungemia episode, a questionnaire about demographic information, clinical
5

113 signs of sepsis, and risk factors or predisposing diseases within the preceding
114 30 days. Approval for the study was obtained from the ethics committees of all
115 participating institutions. All institutions taking part in the study were tertiary
116 hospitals with pediatric departments.
117 Period of study The study was carried out over a 13-month period, from
118 January 2009 to February 2010.
D
o
w
119 Definitions An episode of fungemia was defined as the isolation of a yeast or
n
120 mold species from blood culture in a patient with temporally related clinical lo
a
d
121 signs and symptoms. In patients with more than 1 episode of fungemia, an e
d
122 episode was defined as a new case if it occurred more than 30 days after f
r
o
123 resolution of the previous episode. Outpatient-acquired fungemia was m

124 considered when the fungal etiologic agent was isolated in blood in the first 48 h
t
t
p
125 hours after hospital admission. Neonates were defined as < 1 month of age, :
/
/
126 infants from 1 to 12 months and children were between 1-15 years old. jc
m
.
a
127 Identification of organisms and antifungal susceptibility study. All yeast or s
m
128 mold species isolated from blood culture were identified at the participating .o
r
g
129 institutions by the routine methods in use at each laboratory. Isolates were
/

o
130 stored as suspensions in sterile water at ambient temperature for ulterior n

A
131 studies. Antifungal susceptibility testing was performed, in the first isolate from p
r
132 each fungemia episode, at the participating hospitals by the microdilution il
1
133 colorimetric Sensititre YeastOne® SYO-09 panel (TREK Diagnostic Systems, 2,

2
134 USA). This commercial method determines the minimal inhibitory concentration 0
1
9
135 (MIC) of nine antifungal agents: amphotericin B, flucytosine, fluconazole,
b
y
136 itraconazole, voriconazole, posaconazole, anidulafungin, caspofungin, and
g
u
137 micafungin. Breakpoints applied were those of Clinical Laboratory Standards e
s
138 Institute (CLSI) (4). Since no breakpoints have been published for t
139 posaconazole and amphotericin B, isolates inhibited by >2 mg/L and >1 mg/L,
140 respectively, were considered resistant to these drugs. The recently published
141 species-specific clinical breakpoints for fluconazole, voriconazole and
142 echinocandins were also applied (5, 15, 16, 19). Isolates of C. albicans,
143 Candida tropicalis and C. parapsilosis for which fluconazole MICs were ≤2mg/L
144 were categorized as susceptible and resistant at MICs > 4 mg/L. C. glabrata
6

145 was considered susceptible-dose dependent at MICs ≤32 mg/L and resistant at
146 MICs >32 mg/L. Voriconazole MICs ≤0.125 mg/L for C. albicans, C. tropicalis,
147 and C. parapsilosis were classified as susceptible, 0.25-0.5 mg/L as
148 intermediate and ≥1 mg/L as resistant. Anidulafungin, caspofungin and
149 micafungin MICs ≤0.25 mg/L for C. albicans, C. tropicalis, and C. krusei, were
150 classified as susceptible and >0.5 mg/L as resistant. For C. glabrata, and
151 anidulafungin or caspofungin MICs ≤0.12 mg/L were categorized as susceptible D
o
w
152 and MICs >0.25 mg/L as resistant, while for micafungin MICs ≤0.03 mg/L and
n
153 >0.12 mg/L were considered as susceptible and resistant, respectively. For C. lo
a
d
154 parapsilosis and Candida guilliermondii MICs ≤2 mg/L and >4 mg/L of the three e
d
155 echinocandins were classified as susceptible and resistant, respectively. f
r
o
m
156 Statistical analyses Data were analyzed with SPSS 10.0.7 (SPSS Inc,
h
t
157 Chicago, USA). Continuous variables were compared with Student’s t test and tp
:
/
158 categorical variables with the Chi-square or Fisher’s exact test. A P value <0.05 /jc
m
159 was considered significant.
.
a
s
m
160
.
o
r
g
161 RESULTS
/

o
n
162 A total of 203 episodes of fungemia in 200 patients <16 years of age were A
p
163 identified during the study period. Candida species accounted for 99% of the r
il
164 fungal isolates. Most fungemia occurred among males (60.6%) and patients 1
2
,
165 less than 1 year old (57.6%), and more among those hospitalized in general 2
0
166 wards (67.9%), including Hematology/Oncology departments, than in ICU 1
9

167 (18.7%) or NICU (13.3%) (Table 1). C. parapsilosis (46.8%) followed by C. b
y

168 albicans (36.5%), C. tropicalis (5.9%), C. glabrata (3.9%), C. guilliermondii g
u
e
169 (2.5%), Candida lusitaniae (2%), C. krusei (1%), Candida famata (0.5%), s
t
170 Trichosporon asahii (0.5%) and Rhodotorula glutinis (0.5%) were the species
171 causing fungemia during the study period. Mixed infections occurred in three
172 episodes (C. guilliermondii plus T. asahii; C. albicans plus C. lusitaniae; and C.
173 parapsilosis plus C. lusitaniae) all of them in male ICU patients more than 1
174 year of age. No Cryptococcus, Fusarium or Acremonium species were isolated
175 throughout the study. Causal agent of fungemia varied according to age,
7

176 gender or patients’ location in the hospital (Table 1) and their underlying
177 conditions (Table 2). C. parapsilosis was the predominant species in patients
178 over 1 month, while C. albicans was the predominant species in neonates,
179 52.8% vs. 31.1% in the group 1-12 months (P=0.02). C. glabrata was identified
180 more frequently from females than it was from males (7.5% vs. 1.6%, P=0.03)
181 and C. albicans was more common in the Neonatal ICU setting (51.9%) than in
182 Pediatric ICU (39.5%) or other hospital locations (32.6%). D
o
w
n
183 Presence of intravascular catheter (79.3%), surgery (35%), prematurity (30%),
lo
a
184 and neutropenia (11%) were the predisposing factors most frequently
d
e
185 associated with fungemia (Table 2). C. parapsilosis was the prevalent species d

f
186 related to all the predisposing factors analyzed, except for premature or ro
m
187 neutropenic children, in whom C. albicans was the most common species
h
t
188 isolated. No outpatient-acquired fungemia episodes were observed during the tp
:
/
189 study period. /jc
m
190 Table 3 summarizes the results of in vitro susceptibility testing of bloodstream .a
s
m
191 fungal isolates. Overall, the rate of resistance to amphotericin B and flucytosine
.
o
192 was very low; only 1 isolate (0.49%) of C. tropicalis was resistant (MIC 2 mg/L) r
g
/
193 to amphotericin B and one C. albicans was resistant to flucytosine (MIC 32 o
n
194 mg/L). Among Candida spp., only 4 isolates showed multi-resistance when A
p
195 applying the CLSI clinical breakpoints: one C. albicans, C. glabrata and C. r
il
1
196 tropicalis isolates resistant to all azole agents; and one C. tropicalis resistant to
2
,
197 amphotericin B and azoles. 2
0
1
9
198 Of all Candida species isolated, 95.1% were susceptible to the nine antifungal
b
y
199 agents tested (96% of C. albicans, 98% C. parapsilosis, 87.5% C. glabrata,
g
u
200 83.4% C. tropicalis, 100% of infrequent Candida species [C. lusitaniae, C. e
s
201 guilliermondii and C. famata]) including the two C. krusei isolates, intrinsically t
202 resistant to fluconazole. The resistance rate for Candida species and antifungal
203 agents range from 0% (caspofungin) to 2% (itraconazole). C. tropicalis was the
204 species most resistant to antifungals: two isolates (16.6%) were resistant to
205 fluconazole and itraconazole (Table 3).
8

206 Applying the new species-specific clinical breakpoints for fluconazole,
207 voriconazole and echinocandins (5, 15, 16, 19), the resistance rates were equal
208 to those obtained applying the CLSI breakpoints (4) except for C. albicans and
209 micafungin and fluconazole, where one isolate was categorized as resistant.
210 For voriconazole, one C. tropicalis isolate shifted from intermediate to resistant.
211 In contrast, all C. parapsilosis isolates were classified as susceptible with the
212 species-specific clinical breakpoints (Table 3). D
o
w
n
213
lo
a
d
214 DISCUSSION e
d

f
215 Studies on epidemiology of fungemia and susceptibility are more frequent in ro
m
216 adult populations and are based on large samples; thus, pediatricians often
h
t
217 have to rely on data from these studies when applying treatment. To our tp
:
/
218 knowledge this is one of the largest multicenter prospective series of fungemia /jc
m
219 reported in pediatric patients, it having a total of 200 patients from 30 Spanish
.
a
220 tertiary hospitals. Data were collected in 13 months and incorporate results of in s
m
221 vitro susceptibility to nine systemic antifungal agents, including those most .o
r
g
222 recently commercialized (posaconazole and micafungin). Furthermore, not only
/

o
223 have we applied the clinical breakpoints from CLSI but also the new species- n

A
224 specific clinical breakpoints for fluconazole, voriconazole, and echinocandins. p
r
il
225 As in adults, fungemia occurs more frequently in males (60.5%) (2, 3, 13). 1
2
,
226 Among pediatric patients, fungemia is more frequent in children over one year 2
0
227 of age (42.4%) and neonates (35.4%). In our study, a higher proportion (67.9%) 1
9

228 of fungemia episodes occurs in patients admitted to a general pediatric or b
y

229 surgical ward in contrast to other pediatric series where the ICU is the most g
u
e
230 prevalent hospital unit with fungemia (10, 26) . As in other studies, the most s
t
231 frequent predisposing factor associated with fungemia is indwelling intravenous
232 catheter (79.3%) in accordance with that reported by Stamos (87%) and Neu
233 (89%) (10, 23).
234 Overall, the predominant fungal species isolated in our country is C.
235 parapsilosis, which causes 46.8% of episodes. But species distribution varies
9

236 according to patients’ comorbidity, C. albicans being isolated most frequently in
237 neonates (52.8%), and in patients admitted in Neonatal ICU (51.9%).
238 When comparing these data with those from our previous study from 1997-1999
239 (14), no significant differences in global species distribution has been observed;
240 nevertheless, the percentage of C. parapsilosis isolation has decreased in all
241 age groups, mainly in patients less than 1 year old (53% vs. 42.7%), while the
D
242 prevalence of C. albicans has increased although without statistical o
w
n
243 significance.
lo
a
d
244 The overall species distribution observed in our national survey contrasts with e
d
245 other studies where C. albicans is the first species isolated in children followed f
r
o
246 by C. parapsilosis (2, 3, 23, 26). However, in the study published by Neu et al., m

247 C. parapsilosis is the predominant species regardless of the patients’ h
t
t
p
248 comorbidity or age groups (10). :
/
/
jc
m
249 In our study, polyfungal infections occurred in three cases (1.5%), all of them
.
a
250 involving two species. This rate of mixed infections is similar to those reported s
m
251 in other series (2, 3). Interestingly, in two of the four fungemia episodes .o
r
g
252 concerning C. lusitaniae, this species was isolated in combination with another
/

o
253 yeast. Additionally, no Cryptococcus, Fusarium or Acremonium species have n

A
254 been isolated as causal agents of fungemia. p
r
il
255 Among Candida species, 95.1% of all isolates are susceptible to the nine 1
2
,
256 antifungal agents tested. Despite the extensive use of fluconazole, the 2
0
257 susceptibility rate to this agent remains very high (99.4% of C. albicans plus C. 1
9

258 parapsilosis isolates), similar to that reported by other authors (3, 10), it even b
y

259 having increased since our last study ten years ago (14). Furthermore, as C. g
u
e
260 glabrata and C. krusei are very infrequent in the pediatric setting (2, 3, 10, 12, s
t
261 14, 22, 23, 26), fluconazole is still a reasonable option for fungemia treatment
262 before species identification, except in children with prior azole exposure as is
263 recommend by the latest guidelines published (7, 11).
264 Except intrinsically resistant species (T. asahii and R. glutinis), echinocandins
265 show a broad antifungal activity; only two C. parapsilosis and one C. tropicalis
266 isolates were resistant to micafungin or/and anidulafungin. To note, all C.
10

267 parapsilosis isolates were susceptible to the three echinocandins when applying
268 the new species-specific clinical breakpoints, only one isolate was resistant to
269 micafungin. Regarding amphotericin B and flucytosine, as in other series, both
270 agents present the lowest rate of resistance (0.5%). With respect to the newly
271 commercialized systemic antifungal agents, micafungin and posaconazole, both
272 show excellent in vitro activity against Candida species isolates.
D
273 Despite the increasing use of antifungal agents in the last decade, when o
w
n
274 comparing the present results with those obtained in our previous national study
lo
a
275 ten years ago (14) a decrease in percentage of resistant isolates of
d
e
276 amphotericin B, flucytosine, fluconazole, itraconazole and voriconazole is d

f
277 observed. One of the limitations of this study is the lack of data on antifungal ro
m
278 use or severity of illness; nevertheless, the large size of the sample gives
h
t
279 pediatricians valuable information about susceptibility when deciding on tp
:
/
280 appropriate treatment. /jc
m
281 This descriptive study is one of the largest multicenter series of fungemia .a
s
m
282 episodes in the pediatric setting. C. parapsilosis is the most frequently
.
o
283 implicated species in Spain, followed by C. albicans and C. tropicalis. The risk r
g
/
284 factors observed for fungemia are in accordance with those reported by other o
n
285 authors in different countries. In the Spanish pediatric population evaluated, A
p
286 resistance to azole and echinocandin agents is extremely rare among Candida r
il
1
287 species, confirming the utility of these agents for the empirical treatment of
2
,
288 fungemia in children. The percentage of fluconazole resistant isolates in the 2
0
289 pediatric population in Spain has decreased in the last ten years. This study 1
9

290 confirms the importance of epidemiological surveillance studies on fungemia for b
y

291 evaluating changes in species distribution and antimicrobial susceptibility g
u
e
292 patterns and for assessing the potential impact of new antifungal agents. s
t
293