Vous êtes sur la page 1sur 43

2nd ESCMID Conference on

Invasive Fungal Infections

y
r
a
r
b
i
L
e

Whats new in diagnostic modalities?


Roma, January 18, 2013

r
u
t
c
e r
L
o
Is the time ripe for MALDI-TOF
mass
spectrometry in the
e
h
t
n
i
u
l
a
n
identification of fungi?
O by
D
Maurizio Sanguinetti
I
M
C
Institute of
Microbiology
S
E

Universit Cattolica del Sacro Cuore - Roma

Mass spectrometry:
y
r
a
r
old method
b

i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

New
applications

2002 Tanaka Nobel prize in chemistry: with the proper


combination of matrix and laser wavelenght a protein can be
ionized
1991 first commercial apparatus

C
S
E

1980 developed by Karas & Hillenkamp and Tanaka

y
r
a
r
b
i
L
e

MALDI-TOF MS

C
S
E

100 cm

r
u
t
Matrix-Assisted
Laser
c
e
Desorption Ionization
r
L
(MALDI)
o
e
h
t
n
i
u
l
with
analysis
of
a
n
O by
D Time-Of-Flight
I
(TOF)

The advantage offered by the MALDI-TOF combination is the


capability to easily desorb and analyze positive as well as
negative ions from the same sample

Principle of operation of a MALDI-TOF


y
r
a
r
spectrometer for microorganism identification
b

i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

C
S
E

Time of flight

Molecular masses

Biomarkers

y
r
a
r
Cellular compounds detected
b
i
L
mostly ribosomal proteins or DNA-binding
proteins,
e
r
but also complex lipids and polysaccharides
u
t
Proteins detected
c
e hydrophilic,
r
extractable, soluble, moderately
stable,
L
o
e
h
and abundant
t
n
i
u
l
Determination ofnprotein mass
signal intensities
a
y stability, aminoacid
favored by O
abundance,
b
D (e.g.
I
composition
Arg and Lys)

M
C
S
E
Range of detection MALDI-TOF
2000-20000 Da

y
r
a
r
b
MALDI-TOF for the identification
and
i
L
classification of microorganisms
needs
e
r
u
t
dedicated software [e.g.,
BioTyper (Bruker
c
e
r
L
Daltonics Inc.), Saramiso
(AnagnosTec
e
h
t
n
GmbH), or Andromas
SAS] to enable
i
u
l
a
n
comparisons
O of bthey unknown protein with
D molecular
reference
masses. Ribosomal
I

M
proteins
are used normally as reference
C
S
Emolecular masses as they are most abundant
in the cells.

y
r
a
r
b
i
In contrast to bacteria, fungal cell
walls are
L
e
mainly composed of polysaccharides
r
u
t
c
(including chitin), but
proteins,
lipids,
e r
L
o
polyphosphates, and
inorganic
ions are also
e
h
t
n
i
u
l
present. Proteins
provide the most
a
n
y
O biomarkers
characteristics
available for the
b
D
I

analysis
of
intact organisms without
M
C
extraction,
separation, or amplification.
S
E

y
r
a
r
ib
Intact refers to microbial cellsLsuspended
edirectly onto a
in a solution and/or deposited
r
u
t
c
the sample holder. While
exposure
to
water,
e r
L
o
organic solvent, e
and/or strong
acid
in
the
h
t
n
u
li most
MALDI matrix n
lyses
vegetative bacteria,
a
y exposure of the fungus
a lysis step,Osuchbas
D
I

to a strong
organic
acid, is required to obtain
M
C
a MALDI spectrum.
S
E

YEASTS protocol ary


r
b
i
L
e
r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M
C
S
E
Extraction mandatory

Formic acid 70%


Acetonitrile 100%

EtOH 70%

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

C
S
E

From: Bizzini and Greub, CMI, 2010

y
r
a
r
b
i
L
e Morphology

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

Molecular
(ITS, -TUB, CAM, EF-1)
It could take 48 h
to identification

C
S
E

(colony & micromorphology)

SPECIES
IDENTIFICATION

MASS
SPECTROMETRY

1990

2000

Biochemical
metabolic capacities

y
r
a
r
MALDI-TOF MS applications
b
i
L
e
r
u
t
c
e
L
Clinical isolates identificationor
e
h
t
n
i
u
l
Direct identification
of
pathogens
in clinical
a
n
y
O
b
samples ID

M
C
Subtyping
S
E
Drug susceptibility testing

y
r
a
r
b
i
L
e

MALDI-TOF Identification of Yeasts: Which is the Reality?


Study

No. of isolates/species

Van Veen et al. 2010

61 isolates
12 species

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

% of isolates identified
85.2

Bizzini et al. 2010

24 isolates
4 species

Stevenson et al. 2010

194 isolates
23 species

87.1 (99)a

Bader et al. 2010

1192 isolates
36 species

97.6

Dhiman et al. 2011

138 isolates
14 species

92 .0 (96.3)a

Goyer et al., 2012

335 isolates
17 species

94.0

C
S
E

aUsing

100

a score threshold of 1.8

Reproducible and accurate, with low consumable costs and minimal preparation time
Several closely related species (e.g., Candida psilosis or Candida glabrata/bracarensis)
could be resolved by MALDI-TOF MS, but not by a biochemical approach

5 min of hands-on time per identification

$ 0.50 per sample

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

C
S
E

Cluster analysis of MALDI-TOF MS spectra of selected reference or


challenge isolates of C. neoformans and C. gattii

MALDI-TOF for characterization of Candida


palmioleophila, a previously overlooked patogen

y
r
a
r
b
i
L
e

C. palmioleophila has previously been

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

misidentified as C. famata or C. guilliermondii.

The susceptibility pattern for C.


palmioleophila is unique, with low
echinocandin MICs (range, 0.008 to 0.125
g/ml) and high fluconazole MICs (range, 8 to
>16 g/ml). Thus, correct identification of C.

palmioleophila is important.

C
S
E

Identification is possible yet laborious with


conventional techniques, whereas MALDITOF MS easily separates the related species.

Score-oriented dendrogram to cluster the MALDI-TOF mass spectra


obtained for all included isolates. All samples were named based on
molecular identification, illustrating the complete agreement between ITS
sequencing and the MALDI-TOF spectra.

From: Jensen RH and Arendrup MC, JCM, 2011

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

Identification of Yeast Species Directly from


Positive Blood Culture Media

y
r
a
r
b
i
L
e

Sepsityper specimen
processing

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

MALDI Biotyper system

Yeast-containing positive blood cultures


Tested (n = 42)

No. of specimens

Species identified

28

C. albicans

C. parapsilosis

C. tropicalis

C. neoformans

C
S
E

Limit of detection: 5.9 105 CFU


Starting material: 1 ml of the blood culture fluid
From specimen extraction to final result reporting: 1 hour

From. Yan et al. JCM 2011

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

In house protocol extraction:


detergent plus yeast protocol
extraction

C
S
E

Starting material: 6 ml of the blood


culture fluid

From specimen extraction to final


result: 25 min

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

MALDI-TOF Identification of Filamentous


y
r
a
r
Fungi: Which is the Reality?
b

i
L
e

r
u
t
c
In contrast to bacteria, relatively
few data are
e
r
L
o
available regarding the
identification
of
e
h
t
n
i
u
l
filamentous fungi
a
n
y
O
b
This due to
of standardized protein
Dthe lack
I

M
extraction
protocols for filamentous fungi and the
C
S
poor
E number of spectra currently included in the
database of commercially available devices

Filamentous fungi: state of the art


Study

No. of isolates/species

Lau et al. 2012

Several mold genera


421 isolates
76 genera, 152 species

y
r
a
r
b
i
L
e

% of isolates identified

88.9% (score >2.0)


93.2% (score >1.7)
Not identified isolates were not
included in the database

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

De Carolis et al. 2011

Coulibaly et al. 2011

Alanio et al. 2010

C
S
E

Marinach et al. 2009

Erhard et al. 2008

Aspergillus, Fusarium, and


Mucorales
103 isolates
29 species

96.8

Scedosporium
25 isolates
7 species

100

Aspergillus
140 isolates
28 species

98.6

Fusarium
62 isolates
9 species

92

Dermatophytes
20 isolates
5 species

100

Identification of filamentous fungi


y
r
a
r
remains difficult ib

L
e

r
u
t
c
e r
Biological complexity
L
o
e
h
t
n
i
u
l
Changes in classification
due
a
n
y
O
b
to the introduction
of nucleicD
I

M
acid based
techniques
C
S
E

Identification of species within the Aspergillus


y
r
species complex could drive appropriate
ra therapy

b
i
L
e

r
u
t
c
Existence of cryptic species
e r
L
o
e
h
t
implies differences in:
n
i
u
l
a
n
Clinical manifestations
O by
D
I

Prognosis
M
C
Antifungal
susceptibility profile
S
E

y
r
a
r
b
i
L
e

Superficial scraping of
mold colonies

r
u
H O ct
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
Unknown fungi
M
C
S
Generate
E
2

MALDI-TOF
profile
spectrum

Preparation onto a
MALDI target plate

ETOH

HCCA matrix
(-cyano-4-hydroxycinammic acid)

Sample
inactivation

From: De Carolis et al., CMI, 2012

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

C
S
E

y
r
a at different
Reference spectra
r
b
i
L
stages of maturation:
e
r
u
109
from young colonies
t
c
e 109 rfrom mature colonies
L
o
e
h
(diameter
> 3 cm)
t
n
i
u
l
a
n

O by 67 reference strains of
D Eurotiales
I
M
42 reference strains of

Mucorales and Hypocreales

From: De Carolis et al., CMI, 2012

Challenge of 103 clinical isolates


y
r
Identification by multilocus sequencing
a
Sequence analysis of the ITS1-5.8S-

r
u
t
gene regions was used as the
c
e r
reference identification method
L
o
e
h
85 isolates were unequivocally
t
n
i
u
l
identified to the species level
a
n
y
18 isolates producingO
ambiguous
b
D identified
I
results were initially
to the

genus level M
(Aspergillus section Flavi)
C
S
Further
E molecular analysis assigned
ITS2, -TUB, CAM, and/or EF-1

r
b
i
L
e

these isolates to the species

Aspergillus oryzae (17 isolates) and


Aspergillus flavus (1 isolate)

Challenge of 103 clinical isolates


y
r
Identification by MALDI a

r
b
i
MALDI-TOF MS identified, according to their designated
species,
L
91 of 94 clinical isolates (96.8%) of Aspergillus,
e Fusarium and
r
u
Mucorales
t
c
e withr correct results were
The log(-score) values of the 91Lisolates
o
all higher than 2.0, whereas e
three isolates
with a log(score) value
h
t
n
i
u
l
of <2.0 (1.817, 1.874 and 1.796, a
respectively)
could be identified
n
only to the genus level,
O butbhady concordant species designations
D the results
as compared Iwith
of the reference method

M
By contrast,
the clinical isolates that belonged to species not
C
S
included
in our database (9 isolates: 3 Alternaria alternata, 2
E
Scedosporium prolificans, 2 Curvularia, 1 Beauveria bassiana, 1
Cladosporium) had all log(score) values of <1.7, thereby
confirming the specificity of MALDI-TOF MS identification

Cluster analysis of MALDI-TOF spectra of selected


reference strains and challenge isolates (CH15, CH25 and
CH51) identified as Aspergillus section Flavi species

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

C
S
E

From: De Carolis et al., CMI, 2012

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

Fusarium solani species complex subtyping


y

r
a
r
b
i
L
FUSARIUM-ID
e

r
u
t
database
c
e r
SEQUENCE
L
o
e
h
TYPE
t
n
i
u
l
a
n
O by
D
I
Mass spectrometry-based
dendrogram VS MLST
M
C
The strains
were clustered in separate groups according to their
S
phylogenetic
E species designation

Each MSP is compared with the other in a matrix of cross-wise identification


values. The matrix is used to calculate the distance values for each pair
Based on the protein mass patterns, strains can be clustered hierarchically

y
r
MIC
a
r
determination
b
i
L
Microdilution methods
E-Test
Disk diffusion
e
r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M
C
New generation of
S
susceptibility tests
MASS
E
It could take 48 h
End-point

subjective

MEC

SUSCEPTIBILITY
DETERMINATION

SPECTROMETRY

MPCC, the minimal profile change concentration


New endpoint:

y
r
a
r
b
i
L
e

a value defined as the lowest drug concentration at which a mass spectrum profile
change can be detected

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

MPCC

C
S
E

Inoculum 106 yeast/ml


Turnaround time 15 h
Acid extraction
End-point reading
objective
Cost less than 1 euro

MPCC determinations were concordant with MIC BDM irrespective of the type of drug
resistance mechanism

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

y
r
a
r
b
i
L
e

C
S
E

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
Susceptible
isolate
Resistant isolate
M
C
S
E
0.90 > 0.76

0.63 < 0.89

Submitted to Anal. Chem

y
r
a
r
b
i
L
e

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

Performances of the of ms-AFST method for 62 clinical C. albicans isolates


according to the presence (R) or absence (S) of FKS1 hot-spot mutation

C
S
E

Submitted to Anal. Chem

y
r
MALDI-TOF intact cell MS (Pros)
a
r
b
i
L
Rapidity
e
r
u
t
Inexpensiveness in terms of
labor and
c
e
r
L
consumables
o
e
h
t
n
i
u
l
High discriminatory
power,
accuracy, and
a
n
y
O morphological
superiority over
analysis and
b
D
I

molecular
identification
M
C
Ability
to easily differentiate species that are
S
Emorphologically and phylogenetically similar to
each other

MALDI-TOF intact cell MS (Cons)ry

a
r
b
i
L
MALDI-TOF MS equipment is not cheap
e
r
u
Molecular diagnostic techniques
are still required in
t
c
e spectra
cases for which no reference
are present in
r
L
o
e
h
the MALDI-TOF MS n
databases
at the time of
t
i
u
l
a
n
analysis
y
O
b
Apart fromID
positive blood cultures, MALDI-TOF

cannot M
yet used directly on patient samples
C
S
Also,
E the system is not able to identify the presence
of several different pathogens in a sample

SPECIAL THANKS TO:

y
r
a
r
b
i
L
e

Istituto di Microbiologia, Universit Cattolica del Sacro Cuore, Roma, Italy


E. De Carolis
B. Posteraro
A. Vella
A. R. Florio
R. Torelli
L. Vaccaro
T. Spanu
G. Fadda

r
u
t
c
e r
L
o
e
h
t
n
i
u
l
a
n
O by
D
I
M

Innsbruck Medical University, Innsbruck, Austria


C. Lass-Flrl

Public Health Research Institute, New Jersey, USA


D. S. Perlin

Dipartimento di Biotecnologie Cellulari ed Ematologia,


Universit La Sapienza di Roma, Italy
C. Girmenia
C. Colozza

C
S
E

Dipartimento Sanit pubblica Microbiologia e Virologia,


Universit degli Studi di Milano, Italy
M. Cogliati
A. M. Tortorano