Utility of an immunocapture-agglutination test and an enzyme-linked immunosorbent assay test against cytosolic proteins from Brucella melitensis B115 in the diagnosis and follow-up of human acute brucellosis.

Abstract

The utility of an immunocapture-agglutination (Brucellacapt, Vircell SL, Granada, Spain) test and an enzyme-linked immunosorbent assay IgG, IgA, and IgM (ELISA-IgG, ELISA-IgA, ELISA-IgM) against cytosolic proteins from Brucella melitensis B115 (R) was compared with ELISA-IgG, ELISA-IgA, and ELISA-IgM against smooth lipopolysaccharide (S-LPS) from B. melitensis 16M (S), serum agglutination test (SAT), and Coombs test in the diagnosis and follow-up for 10 months of 51 patients with acute brucellosis. The sensitivities of ELISA tests against cytosolic proteins varied from 49.0 % for ELISA-IgG to 64.7% for ELISA-IgM and were lower than the sensitivities showed by ELISA S-LPS (from 88.2% to 92.2%), SAT (88.2%), Coombs (96.1%), and Brucellacapt (98.0%) tests. Specificity was over 93% in all cases. The evolutionary behavior of the SAT, Coombs, and Brucellacapt tests was similar. There was a decrease of between 20% and 40% in antibody titer in the 10th month of evolution after treatment. The evolutional curves of IgG, IgA, and IgM against cytosolic protein increased slightly till the eighth month. The specific IgM and IgA antibodies against protein fractions began to show a drop from the eighth month on, showing levels slightly lower than the initial sera values by the end of the 10th month. In this month, titers of specific IgG against proteins fractions remained higher than the titers showed by the initial sera.

Full text

Page 1

Utility of an immunocapture-agglutination test and an enzyme-linked
immunosorbent assay test against cytosolic proteins from Brucella
melitensis B115 in the diagnosis and follow-up of human acute brucellosis
Marı ´a A´ngeles Manteco ´na, Purificacio ´n Gutie ´rrezb, Marı ´a del Pilar Zarzosab, Ana I. Duen ˜asa,
Javier Solerac, Luis Ferna ´ndez-Lagod, Nieves Vizcaı ´nod, Ana Almaraza, Miguel Angel Bratosb,
Antonio Rodrı ´guez Torresb, Antonio Ordun ˜a-Domingoa,b,4
aUnidad de Investigacio ´n, Hospital Clı ´nico Universitario de Valladolid, 47005 Valladolid, Spain
bDepartamento Microbiologı ´a, Universidad de Valladolid, 47005 Valladolid, Spain
cMedicina Interna, Complejo Hospitalario de Albacete, 02006 Albacete, Spain
dDepartamento Microbiologı ´a y Gene ´tica Molecular, Edificio Departmental, Universidad de Salamanca, 37007 Salamanca, Spain
Received 19 August 2005; accepted 28 November 2005
Abstract
The utility of an immunocapture-agglutination (Brucellacapt, Vircell SL, Granada, Spain) test and an enzyme-linked immunosorbent
assay IgG, IgA, and IgM (ELISA-IgG, ELISA-IgA, ELISA-IgM) against cytosolic proteins from Brucella melitensis B115 (R) was compared
with ELISA-IgG, ELISA-IgA, and ELISA-IgM against smooth lipopolysaccharide (S-LPS) from B. melitensis 16M (S), serum agglutination
test (SAT), and Coombs test in the diagnosis and follow-up for 10 months of 51 patients with acute brucellosis. The sensitivities of ELISA
tests against cytosolic proteins varied from 49.0 % for ELISA-IgG to 64.7% for ELISA-IgM and were lower than the sensitivities showed by
ELISA S-LPS (from 88.2% to 92.2%), SAT (88.2%), Coombs (96.1%), and Brucellacapt (98.0%) tests. Specificity was over 93% in all
cases. The evolutionary behavior of the SAT, Coombs, and Brucellacapt tests was similar. There was a decrease of between 20% and 40% in
antibody titer in the 10th month of evolution after treatment. The evolutional curves of IgG, IgA, and IgM against cytosolic protein increased
slightly till the eighth month. The specific IgM and IgA antibodies against protein fractions began to show a drop from the eighth month on,
showing levels slightly lower than the initial sera values by the end of the 10th month. In this month, titers of specific IgG against proteins
fractions remained higher than the titers showed by the initial sera.
n 2006 Elsevier Inc. All rights reserved.
Keywords: Human Brucellosis; Serology; Cytosolic proteins; Serologic follow-up
1. Introduction
Brucellosis, produced by bacteria of the genus Brucella,
is a disease that affects both humans and animals. It is one
of the most widely spread zoonoses in the world and causes
serious problems (Boschiroli et al., 2001).
Clinically, human brucellosis is a systemic disease
characterized by great clinical polymorphism and an
undulating course, with a strong tendency to present
recurrences and evolve to a chronic form (Young, 2005).
Its polymorphism makes clinical diagnosis difficult, requir-
ing laboratory diagnosis to confirm suspicions of the disease.
Etiologic diagnosis of human brucellosis is based on
isolating Brucella in pathologic products, principally blood
(Young, 2005; Yagupsky, 1999). However, the percentage
of positive blood cultures depends on the clinical period
of the disease; it is low in chronic or evolved phases
(Gotuzzo et al., 1986). In addition, Brucella bacteria grow
slowly and their manipulation carries serious contamination
risks for laboratory personnel (Yagupsky et al., 2000). There-
fore, in most laboratories, human brucellosis diagnosis is
principally based on detection of antibodies against Brucel-
la. Among the most widely used serologic techniques are
the serum agglutination test (SAT) and the Coombs anti-
Brucella test, both of which primarily detect antibodies
0732-8893/$ – see front matter n 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.diagmicrobio.2005.11.003
T Corresponding author. Unidad de Investigacio ´n, Hospital Clı ´nico
Universitario de Valladolid, Avda Ramo ´n y Cajal 3, 47005 Valladolid,
Spain. Tel.: +34-983423063; fax: +34-983423066.
E-mail address: orduna@med.uva.es (A. Ordun ˜a-Domingo).
Diagnostic Microbiology and Infectious Disease 55 (2006) 27–35
www.elsevier.com/locate/diagmicrobio

Page 2

against smooth lipopolysaccharide (S-LPS). However, these
techniques present interpretation problems because the
antibody titers can remain elevated over long periods after
recovery from the disease (Ariza et al., 1992). This causes
difficulties in interpreting results in endemic zones, specif-
ically in distinguishing between a current infection and an
immune state from a previous infection. It also causes
difficulties in diagnosis of recurrences or reinfections
(Young, 2005). Similar problems appear in diagnosis of
chronic or prolonged evolution forms. Brucella also
presents cross-reactions with Gram-negative bacteria
(Clavijo et al., 2003; Chart and Jenkins, 1999; Cherwonog-
rodzky et al., 1989) such as Yersinia enterocolitica 0:9,
Vibrio cholerae, Escherichia coli O:157, Pseudomonas
maltophilia, or Francisella tularensis. Such reactions can
generate false positives in the serologic tests against
Brucella (Clavijo et al., 2003; Duen ˜as et al., 2000; Chart
and Jenkins, 1999; Colmenero et al., 1994; Dı ´az and
Moriyon, 1989).
Enzyme-linked immunosorbent assay (ELISA) is one
method that has been incorporated into serologic diagnosis
of brucellosis (Memish et al., 2002; Araj et al., 1988; Saz
et al., 1987). Among its advantages are speed and
automatization. Individualized studies of each class and
subclass of immunoglobulin can be performed as well, so
that the contribution of each of them in the distinct phases
and evolutional forms of the disease can be ascertained
(Ariza et al., 1992). ELISA also allows detection of
antibodies against different bacterial antigenic structures
such as corpuscular antigen (Araj et al., 1988; Saz et al.,
1987), S-LPS (Baldi et al., 1999; Cloeckaert et al., 1999;
Fernandez-Lago and Diaz, 1986), or protein antigens
(Boschiroli et al., 2001; Baldi et al., 1999; Goldbaum
et al., 1992; Araj and Kaufmann, 1989). Several authors
have reflected how interesting detecting antibodies against
protein antigens as an active infection marker can be, given
that these antibodies seem to become negative earlier than
those against LPS (Boschiroli et al., 2001; Baldi et al.,
1996a). However, in spite of these characteristics, there
have been few studies on brucellosis patients that focus on
the behavior of the immune response against Brucella
protein antigens.
On the other hand, a new technique of immunocapture-
agglutination test (Brucellacapt, Vircell SL, Granada,
Spain) that detects antibodies against whole Brucella
(mainly against S-LPS) has recently been studied. This
technique has shown good sensitivity and specificity
in the diagnosis of human brucellosis (Ordun ˜a et al.,
2000). However, its utility in the follow-up of the disease
is unknown.
The aim of this study is to evaluate the diagnostic
utility and the evolutional behavior of an immunocapture-
agglutination test (Brucellacapt) and an ELISA technique
against Brucella cytosolic proteins, comparing them with
the techniques normally used in brucellosis diagnosis (SAT,
Coombs test) and with ELISA tests against S-LPS.
2. Materials and methods
2.1. Clinical samples
The study was performed on 258 sera from 51 patients
diagnosed with acute brucellosis. All the patients presented
symptoms of brucellosis that was confirmed by laboratory
tests. The following criteria were used for including patients
in the study as acute brucellosis cases:
–Epidemiologic backgrounds existed for all patients
and they presented symptoms compatible with the
disease, plus one of the following:
Brucella was isolated from a pathologic patient
sample;
the first serum obtained had a SAT or Coombs anti-
Brucella test titer z1/160; or
a seroconversion or increase of 4 times the SAT titer
was observed.
–
–
–
From the 51 patients, 26 presented positive blood
cultures, 4 had seroconversion in the SAT test, and 4 patients
showed a 4-fold increase in titer in the SAT.
All the patients evolved favorably in less than 3 months
after specific treatment. No patients suffered recurrences in
the 12 months after ending treatment.
A blood sample was taken from each patient in the
consultation, in which the symptoms corresponding to
brucellosis were suspected (initial serum). Blood samples
at 2, 4, 6, 8, and 10 months after the start of treatment were
also studied (evolutional sera). None of the patients had
been diagnosed for brucellosis in the course of the year
before initial serum extraction.
To form a negative control group, 412 sera were taken
from healthy individuals randomly chosen from areas in
which brucellosis is an endemic disease (rural zones in the
autonomous region of Castilla y Leo ´n, Spain). All the serum
samples were distributed in aliquots and conserved for less
than 3 years at ?20 8C until use.
2.2. Methods
For each serum, both in the brucellosis patient group and
in the negative control group, the presence of antibodies
against Brucella spp. was determined by slide agglutination
test (Linear Chemicals, Barcelona, Spain), Coombs anti-
Brucella test (Linear Chemicals) (Hall and Manion, 1953),
Brucellacapt test (Vircell, Granada, Spain) (Ordun ˜a et al.,
2000), and ELISA-specific IgG, IgM, and IgA against
S-LPS of Brucella melitensis 16M (S). ELISA studies were
also performed to determine the presence of specific
antibodies of IgG, IgM, and IgA classes against the
cytosolic proteins of the B. melitensis B115 (R) strain
(Bhongbhibat et al., 1970). All the serum samples from each
patient were processed simultaneously in each test.
2.2.1. Obtaining S-LPS and Brucella protein antigen
S-LPS antigen was obtained from a soy tryptose broth
culture of B. melitensis 16M (S), according to the method
M.A. Manteco ´n et al. / Diagnostic Microbiology and Infectious Disease 55 (2006) 27–35
28

Page 3

described by Westphal and Jann (1965) (hot phenol
method), modified for Brucella (Redfearn, 1960). Briefly,
phenol-inactivated bacteria (B. melitensis 16M) were
harvested by centrifugation and washed twice with saline.
The bacteria sediment was resuspended in distilled water
and mixed with an equal volume of phenol at 66 8C. The
mixture was shaken vigorously for 15 min and then
centrifuged at 13000 ? g for 20 min a 4 8C. The LPS
contained in the phenol fraction was precipitated with
3 volumes of methanol with 1% methanol saturated with
sodium acetate at ?20 8C. The precipitant was dialyzed
against distilled water. The S-LPS was digested successively
with DNAse, RNAse, and proteinase K, and then obtained
by ultracentrifugation at 100000 ? g for 12 h at 4 8C.
Protein antigen (cytosolic protein) was obtained from a
soy tryptose broth culture of B. melitensis B115 (R),
according to the procedure described by Bhongbhibat et al.
(1970). The bacteria were collected by centrifugation,
saline-washed, and precipitated with 2 volumes of acetone
at ?20 8C. The bacteria were harvested by centrifugation,
dehydrated by 3 acetone washes, and vacuum dried. The
dried bacteria were suspended at 5% (w/v) in 2.5%
sodium chloride and shaken in a magnetic stirrer for
3 days at 4 8C. Then, the bacteria were centrifuged at
15000 rpm for 30 min. The supernatant was precipitated at
4 8C with 3 volumes of ice-cold ethyl alcohol. The
precipitate was collected by centrifugation at 15000 rpm
for 30 min, then dialyzed against distilled water and
lyophilized. A photograph of the silver-stained polyacryl-
amide gel electrophoresis (PAGE) of the preparations is
presented in Fig. 1.
To test for the absence of Brucella LPS in the protein
antigen sample, the limulus pyrogenicity test was performed
(Single Test Limulus Amebocyte Lysate, Pyrogent Plus,
Biowhittaker, Walkersville, ME).
2.2.2. SAT and Coombs anti-Brucella test
SAT and Coombs tests were performed in test tubes by
the 2-fold serial doubling dilutions method, with an initial
dilution of 1/20. Commercial corpuscular antigen of
B. abortus (Linear Chemicals) was used for both tests.
SAT reactions were read after 24-h incubation at 37 8C. The
highest serum dilution showing more than 50% agglutina-
tion was considered the agglutination titer. Coombs tests
(Hall and Manion, 1953) were performed on the SAT
reactions, after 3 washes with phosphate-buffered saline
(PBS; Oxoid, Basingstoke, Hamsphire, UK) (pH 7.2) by
centrifugation at 3000 ? g for 20 min. After the last wash
the sediment was resuspended in 1 mL of PBS and then
0.05 mL of previously standardized antihuman total anti-
immunoglobulin (Sanofi Pasteur, Marnes-la-Coquette,
France) was added to each test tube. The tubes were then
mixed and incubated for 24 h at 37 8C. Readings were
performed as for the SAT test.
2.2.3. Immunocapture-agglutination (Brucellacapt) test
The test was performed according to the manufacturer’s
instructions. The quantity of 50 AL of each serum dilution
was added to the wells of a microplate with a U-shaped
base coated with antihuman immunoglobulin. Then 50 AL
of the antigen suspension was added (colored B. melitensis,
killed by formaldehyde treatment). The plates were
covered with adhesive sheets and incubated for 24 h at
37 8C in a dark humid chamber. Plate readings were then
taken. Positive reactions show agglutination over the
bottom of the well. Negative reactions present a pellet
on the center of the bottom of the well.
2.2.4. ELISA test against B. melitensis B115 (R) protein
antigens and B. melitensis 16M (S) S-LPS
For the ELISA techniques, 100 AL of antigen solution
in PBS at the previously standardized concentration was
added to each well of microtiter plates. The plates were
incubated for 24 h at 4 8C. Once the antigen had been
fixed, a PBS wash cycle was performed and then blocked
with a PBS-T/BSA solution (PBS with 0.05% Tween 20
and V fraction bovine serum albumin [BSA; Sigma, St.
Louis, MO] at 1%) for an hour at room temperature, and
then washed with PBS-T (PBS with 0.05% Tween 20).
Next, each well received 100 AL of a serum diluted at
1/100 in PBS-T/BSA. All the problem sera were triple
assayed. Two positive control sera from patients with
confirmed brucellosis were also placed in each plate; these
sera had SAT titers of 1/1280 (high) and 1/160 (low). In
addition, each plate received 20 negative control sera from
the negative control group, chosen at random from the
412 healthy individuals in that group. All the plates
received the same control sera.
Fig. 1. Silver-stained PAGE gel from smooth-lipopolysaccharide
and cytosolic proteins from B. melitensis 16 M (LPS) and B. melitensis
B115 (CP).
M.A. Manteco ´n et al. / Diagnostic Microbiology and Infectious Disease 55 (2006) 27–35
29

Page 4

The plates were incubated in a humidity chamber for 2 h
at 37 8C. After 5 PBS-Twashes, 100 AL of human anti-IgG,
IgM, or IgA rabbit antibodies conjugated with horseradish
peroxidase (Dako, Cytomation, Cambs, UK) was added.
After a new 30-min incubation at 37 8C and further wash
cycles, the samples were developed with o-phenylenedi-
amine (Sigma). Positivity threshold was established in each
plate as the average plus 2 SD of the absorbance of
20 negative control sera group. Absorbance values above
this threshold level were considered as positive and those
below it as negative.
2.2.5. Evolutional studies
For the SAT, Coombs test, and Brucellacapt test evolu-
tional studies, the percentage of variation of the logarithm of
the inverse of the titer of each evolutional serum in relation to
the corresponding initial serum was calculated. A standard
curve was prepared with different dilutions of a mixture of
sera from patients strongly positive against Brucella to
determine the antibody levels of each serum against each
antigen in the ELISA tests. A value of 1000 U/mL was
assigned to this serum mixture. This standard curve was
included in all the ELISA plates, and the antibody concen-
tration of the sera in each plate was plotted against the
standard curve. Each point of the curves represents
the average of the percentages of variation of the
evolutional serum titers with respect to the result from the
initial sera. All the sera from the same patient were triple
assayed in the same test.
2.2.6. Statistical analysis
Sensitivity, specificity, and positive and negative likeli-
hood ratios (LRs) were calculated for the results from
ELISA techniques and for the SAT, Coombs, and Brucella
capt tests. Fifty-one initial sera from brucellosis patients
were taken as positive control group and 412 sera from
healthy individuals living in an area endemic for brucellosis
were taken as the negative control group to study the
diagnostic validity of the tests.
Sensitivity and specificity were obtained with the version
1.1 Bayesiano calculator (http://www.hsa.es/soft/bayes).
Ninety-five percent confidence intervals were calculated
using the Fleiss (1981) method.
Positive and negative LRs were calculated with the 1.0
Twobytwo program; 95% confidence intervals were found
through the methods of Koopman, (1984), Miettinem and
Murnimen (1985), and Gart and Nam (1988). Positive LR
[LR (+)] provides information on how many times more
probable it is to find a positive result for a patient with a
disease than for a healthy individual. The higher the LR (+)
level, the more useful the test is to confirm the disease.
Negative LR [LR (?)] lets us know how many times
more probable finding a negative result in a patient with
a disease is than finding it in a healthy individual. The lower
Table 1
Diagnostic validity of SAT, Coombs, and Brucellacapt tests
Sensitivity (95% CI)Specificity (95% CI)LR (+)a(95% CI)LR (?)a(95% CI)
0.11 (0.05-0.23)
0.19 (0.11-0.32)
0.35 (0.23-0.49)
0.53 (0.39-0.66)
0.03 (0.01-0.13)
0.05 (0.02-0.16)
0.13 (0.06-0.25)
0.25 (0.15-0.39)
0.02 (0.003-0.107)
0.02 (0.003-0.10)
0.07 (0.03-0.18)
0.11 (0.05-0.23)
SAT 1/40
1/80
1/160
1/320
1/40
1/80
1/160
1/320
1/40
1/80
1/160
1/320
88.2 (78.41-98.05)
80.4 (68.51-92.26)
64.7 (50.6-78.8)
47.1 (32.37-61.73)
96.1 (89.77-100)
94.1 (86.6-100)
86.2 (75.84-96.69)
74.5 (61.56-87.45)
98.0 (93.25-100)
98.0 (93.25-100)
92.1 (83.79-100)
88.2 (78.41-98.05)
99.5 (97.96-100)
99.5 (97.96-100)
99.5 (97.96-100)
99.7 (99.16-100)
98.3 (96.93-99.67)
99 (97.96-100)
99 (97.96-100)
99.5 (98.72-100)
95.8 (93.83-97.91)
95.8 (93.83-97.91)
98.5 (97.26-99.82)
98.7 (97.60-99.96)
181.7 (50.17-664.23)
165.6 (45.58-606.63)
133.3 (36.4-491.4)
193.8 (33.94-1114.66)
56.5 (27.6-116.61)
96.9 (38.01-249.38)
188.5 (40.5-349.5)
153.5 (42.14-563.4)
23.7 (15.03-37.87)
23.7 (15.03-37.86)
63.3 (29.21-138.21)
72.7 (31.21-170.7)
Coombs test
Brucellacapt
95% CI = 95% confidence interval.
aLR (+) values above 10 and LR (?) levels under 0.1 are considered clinically useful.
Table 2
Diagnostic validity of ELISA S-LPS and ELISA protein fractions from Brucella tests
Sensitivity (95% CI) Specificity (95% CI)LR (+)a(95% CI)LR (?)a(95% CI)
0.08 (0.03-0.19)
0.18 (0.1-0.32)
0.12 (0.06-0.25)
0.52 (0.38-0.65)
0.36 (0.24-0.50)
0.46 (0.32-0.59)
ELISA-IgG S-LPS
ELISA-IgM S-LPS
ELISA-IgA S-LPS
ELISA-IgG-CP
ELISA-IgM-CP
ELISA-IgA-CP
92.2 (83.79-100)
82.3 (70.90-93.79)
88.2 (78.41-98.05)
49.0 (34.31-63.72)
64.7 (50.6-78.8)
54.9 (40.26-69.53)
93.4 (90.93-95.95)
93.2 (90.65-95.75)
94.2 (91.79-96.5)
97.8 (96.28-99.34)
97.5 (95.96-99.18)
97.5 (96.28-99.34)
14.06 (9.72-20.43)
12.1 (8.29-17.68)
15.1 (10.18-22.6)
20.1 (10.38- 39.0)
26.6 (14.13-50.29)
22.6 (11.78-43.27)
CP = cytosolic protein.
aLR (+) values above 10 and LR (?) levels under 0.1 are considered clinically useful.
M.A. Manteco ´n et al. / Diagnostic Microbiology and Infectious Disease 55 (2006) 27–35
30

Page 5

a LR (?) figure, the more useful the test is in eliminating the
disease as a suspect. LR (+) values above 10 and LR (?)
levels under 0.1 are considered clinically useful (Jaeschke
et al., 1994).
3. Results
3.1. Diagnostic validity
All the initial patient sera were positive (1/20) in the
Coombs and Brucellacapt tests, whereas 4 sera had negative
SAT test results. These 4 patients seroconverted and showed
positive results, with titers z1/160 in the sera obtained
15 days after initial serum extraction. Within the 412 sera
from the negative control group, all the sera resulted
negative in the 3 tests, except for 7 sera (1.7%) that were
positive in the SAT test with a titer z1/40, and 2 of them
(0.5%) presented titers of 1/160. Using the Coombs test,
2 sera (0.5%) presented titers of 1/320, and with the
Brucellacapt test, 5 sera (1.2%) had titers of 1/320. All the
sera positive in SAT were also positive in Coombs test
and Brucellacapt.
Table 1 shows the sensitivity and specificity from each of
the techniques, as well as the positive and negative LRs.
Comparing the SAT, Coombs, and Brucellacapt test,
the lowest sensitivity was obtained by SAT. At a dilution
of 1/40 the SAT test showed a sensitivity of 88.2%, whereas
the Coombs and Brucellacapt tests presented a sensitivity of
96.1% and 98.0%, respectively, at the same dilution.
Specificity was over 95% in all cases. Likewise, LR (+)
was above 23.7 and LR (?) was lower than 0.53 for
all cases. However, only Coombs test and Brucellacapt
showed a LR (?) less than 0.1 and are thus considered
clinically useful.
The values for sensitivity, specificity, and LR (+) and LR
(?) with ELISA techniques are shown in Table 2. The
sensitivities obtained with ELISA tests against S-LPS were
similar to those obtained using the Coombs and Brucella
capt tests, with titers of 1/160 and 1/320 as respective
threshold levels for positivity. However, they were higher
than those obtained with ELISA against the cytosolic
protein antigen of B. melitensis B115. Among all the
ELISA tests, ELISA-IgG against S-LPS showed the greatest
sensitivity (92.2%); ELISA-IgG against the protein fraction
obtained a sensitivity of 49.0%. Among the patients with
positive blood culture, the most sensitive tests were Coombs
and Brucellacapt (100% sensitivity in both cases); the least
sensitive tests were ELISA-IgG, ELISA-IgA, and ELISA-
IgM against cytosolic proteins (57.6%, 61.5%, and 50%,
respectively) (Table 3).
Regarding specificity, the ELISA cytosolic protein test
result was higher than the 93% obtained by the ELISA-IgG
and ELISA-IgM against B. melitensis 16M S-LPS. From
the 412 negative control sera from healthy individuals,
27 (specificity, 93.2%) were positive with ELISA-IgG
against S-LPS and 28 (specificity, 93.4%) were positive in
the ELISA-IgM test.
LR (+) was above 9.4 in all cases, whereas LR (?) varied
between the 0.08 obtained with ELISA-IgG against S-LPS
and the 0.52 reached with ELISA-IgG against the protein
fraction of B. melitensis B115.
3.2. Antibody evolution
The evolution of the antibody titers determined by SAT,
Coombs, and Brucellacapt tests was studied, as well as that
of the different immunoglobulin classes determined by
ELISA against each of the antigens studied.
The evolutional behavior of the SAT, Coombs, and
Brucellacapt tests was similar. There was a decrease of
Table 3
Sensitivity of the serologic tests in brucellosis patients with positive blood
culture (n = 26)
Brucellosis patients with positive blood
culture and positive serology, n (%)
SAT
Coombs test
Brucellacapt
ELISA-IgG S-LPS
ELISA-IgM S-LPS
ELISA-IgA S-LPS
ELISA-IgG-CP
ELISA-IgM-CP
ELISA-IgA-CP
22 (84.6)
26 (100)
26 (100)
23 (88.4)
22 (84.6)
22 (84.6)
15 (57.6)
16 (61.5)
13 (50)
Fig. 2. Evolution of SAT, Coombs test, and Brucellacapt in patients with
acute brucellosis after the start of treatment. Percentage of variation with
respect to the initial serum expressed in percentage of the average of the
logarithms of the inverse of the titers.
Fig. 3. Evolution of antibodies against S-LPS detected by ELISA.
Percentage of variation with respect to the initial serum was expressed as
units per milliliter (see Materials and methods).
M.A. Manteco ´n et al. / Diagnostic Microbiology and Infectious Disease 55 (2006) 27–35
31