- Email: [email protected]
Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve Bordetella pertussis serodiagnosis in recently vaccinated subjects
Journal Pre-proof Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve Bordetella pertussis serodiagnosis in recently vaccinated subjects Lorenzo Subissi, Caroline Rodeghiero, Helena Martini, Amber Litzroth, Kris Huygen, Geert Leroux-Roels, Denis Piérard, Isabelle Desombere PII:
S1198-743X(19)30528-2
DOI:
https://doi.org/10.1016/j.cmi.2019.10.001
Reference:
CMI 1798
To appear in:
Clinical Microbiology and Infection
Received Date: 22 June 2019 Revised Date:
22 September 2019
Accepted Date: 1 October 2019
Please cite this article as: Subissi L, Rodeghiero C, Martini H, Litzroth A, Huygen K, Leroux-Roels G, Piérard D, Desombere I, Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve Bordetella pertussis serodiagnosis in recently vaccinated subjects, Clinical Microbiology and Infection, https://doi.org/10.1016/j.cmi.2019.10.001. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
1
Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve
2
Bordetella pertussis serodiagnosis in recently vaccinated subjects
3
Lorenzo Subissi1,2, Caroline Rodeghiero1, Helena Martini3, Amber Litzroth4, Kris Huygen1,
4
Geert Leroux-Roels5, Denis Piérard3, Isabelle Desombere1,*
5
Author affiliations
6
1
7
2
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
National Reference Centre (NRC) for Bordetella pertussis, Sciensano, Brussels, Belgium European Programme for Public Health Microbiology Training, European Centre for Disease Prevention and Control, Stockholm, Sweden 3 National Reference Centre (NRC) for Bordetella pertussis, Universitair Ziekenhuis Brussel, Belgium 4 Epidemiology of Infectious Diseases, Sciensano, Brussels, Belgium 5 Center for Vaccinology, Ghent University, Belgium
* corresponding author:
Isabelle Desombere National Reference Centre for Bordetella pertussis Immune Response Infectious Diseases in Humans Sciensano (Public Health Belgium) Rue Engelandstraat 642 B-1180 Brussels; Belgium Email: [email protected] Tel: +32 2 373 32 94 Fax: +32 2 373 33 67
25 26
Format: Original article
Abstract: 250 words
Manuscript: 2506 words
27
Keywords: Bordetella pertussis, serodiagnosis, serology, IgA, ACT, pertussis toxin, antibody,
28
vaccination
29
Running title: Anti-PT IgA and anti-ACT IgG to diagnose a Bordetella pertussis infection in
30
recently vaccinated subjects.
31
Abstract
32
Objectives. Quantifying IgG antibodies to pertussis toxin (PT) is the most specific and
33
sensitive method to serodiagnose a Bordetella pertussis infection. Since PT is a component
34
of acellular pertussis vaccines, anti-PT IgG is also induced by vaccination, precluding
35
pertussis serodiagnosis exclusively based on anti-PT IgG in recently vaccinated subjects.
36
Here, we aim to identify additional B. pertussis-specific serological markers that can
37
discriminate between infection and recent vaccination.
38
Methods. The clinical utility of measuring IgA directed to the vaccine antigen PT, and IgG
39
directed to non-vaccine antigens (Fim2/3, LPS, ACT, CatACT) was evaluated in 9 well-
40
characterized subject groups, aged 10-89 years (n=390). Serum anti-PT IgG levels (>125
41
IU/ml) served as an indicator for a recent B. pertussis infection. Comparing symptomatic
42
pertussis-infected (n=140) with recently vaccinated, non-infected (n=100) subjects, revealed
43
the optimal cut-off, accuracy, sensitivity and specificity for each single parameter.
44
Results. For pertussis diagnosis in recently vaccinated subjects, the measurement of anti-PT
45
IgA (cut-off 15 IU/mL) and anti-ACT IgG (cut-off 15 U/mL) resulted in accuracies of 95%
46
(91.5-97.1) and 87.5% (82.7-91.1), sensitivities of 92.9% (87.4-96.0) and 83.6% (76.5-88.8)
47
and specificities of 98% (93.0-99.4) and 93% (86.3-96.6), respectively. Comparing anti-PT IgA
48
levels between the youngest (10-19 years, n=38) and oldest (70-89 years, n=17) age groups
49
revealed an age-dependent increase of antibody levels in pertussis-infected subjects
50
(p<0.0001).
51
Conclusions. Reflex testing of anti-PT IgA and anti-ACT IgG improves pertussis serodiagnosis
52
in recently vaccinated symptomatic subjects with elevated anti-PT IgG levels. Furthermore,
53
both markers can discriminate between vaccination and recent infection in pertussis
54
serosurveillance studies. 2
55 56
Introduction
57
Testing for Bordetella pertussis, the etiological agent of whooping cough, is recommended
58
in patients with prolonged cough, which is often the only symptom of pertussis (1;2).
59
Unfortunately, pertussis diagnosis is challenging and does not exclusively rely on laboratory
60
testing. Because the sensitivity of culture and PCR (gold standard for diagnosis) decreases
61
over time, serological techniques are the most sensitive diagnostic method at a later stage
62
of the disease. For this, specific IgG antibodies against B. pertussis toxin (PT) are quantified
63
in serum or oral fluids. However, for a correct interpretation of serology, the knowledge of
64
the duration of symptoms, the age and vaccination status of the patient are essential. When
65
diagnosis cannot be clearly established with one serum sample, a second convalescent
66
sample is requested (1;2).
67
The Belgian National Reference Centre (NRC) follows the ECDC-guidelines and routine
68
serodiagnosis is performed by quantifying circulating anti-PT IgG, endowed with the
69
strongest diagnostic potential in older children, adolescents and adults(1;3-8). PT-induced
70
immune responses are specific for B. pertussis, whereas antibodies directed to other
71
virulence factors are less specific due to cross-reactivity with antigens present in other
72
Bordetella species and other microbial pathogens (1;6;9). PT antigen (in a toxoided form) is
73
also present in all licensed acellular vaccines (aP) in Europe and anti-PT IgG are thus induced
74
by vaccination (10). Since an increase in pertussis incidence has been reported worldwide
75
(11;12) and booster vaccinations with PT-containing aP vaccines are increasingly
76
recommended, the diagnostic power of anti-PT IgG can be hampered by possible
77
interference of previous vaccinations (13). Therefore, additional parameters have been
78
evaluated for their capacity to improve pertussis serodiagnosis in recently vaccinated 3
79
subjects. The clinical utility of measuring anti-PT IgA is still under debate (1;14;15). Interest
80
in measuring IgA, particularly in adolescents and adults, comes from the observation that
81
intramuscular vaccination may not induce IgA secretion, making it a unique infection marker
82
(14;16). However, several studies reported that IgA responses to B. pertussis infection tend
83
to increase with age (6;17-20). By consequence, no broadly accepted cut-off point is
84
available.
85
Antibody responses to B. pertussis antigens absent from aP vaccines have been examined,
86
including adenylate cyclase toxin (ACT), the catalytic domain of ACT (CatACT), lipo-
87
oligosaccharide (LPS) and fimbriae (FIM) (included in some vaccines). Compared to anti-PT
88
IgG, measuring antibodies to these antigens resulted in lower sensitivities and their
89
usefulness as unique markers in pertussis serodiagnosis is limited (6). However, their value
90
as additional markers for serodiagnosis in recently vaccinated subjects is unknown.
91
In here, we evaluated the clinical usefulness of IgA directed to the vaccine antigen PT, and
92
IgG directed to B. pertussis antigens FIM, LPS, ACT and CatACT to improve pertussis
93
diagnosis in symptomatic subjects with elevated anti-PT IgG levels and a history of recent
94
vaccination.
95 96
Methods
97
Study population
98
Study subjects were retrospectively selected from a Belgian serosurveillance study, a
99
Boostrix vaccination trial, and the Belgian NRC for Bordetella pertussis. In the
100
serosurveillance cross-sectional study, residual serum specimens were collected (2013-
101
2015) through voluntarily participating diagnostic laboratories (21). The Boostrix clinical trial
102
was conducted during the IMI-funded BioVacSafe project (2012-2018). Eligible subjects 4
103
were randomized based on duration since their last Diphtheria-Tetanus-Pertussis (DTP)-
104
vaccination and either vaccinated with Boostrix or placebo-treated. DTP-antibodies were
105
quantified in both studies using a bead-based multiplex assay (22). NRC-samples with
106
accompanying clinical data were selected from sera sent to the lab for routine pertussis
107
serodiagnosis (2017-2018) or PCR-diagnosis (2011-2017). Ethics and additional information
108
on study populations and selection of study subjects is available in Supplementary
109
Methods. Since serodiagnosis of pertussis is less indicated in younger children, subjects
110
under the age of 10 were excluded from all study groups.
111 112
Selection and grouping of study subjects
113
The characteristics of each study group are shown in Table 1. For symptomatic patients,
114
PCR-positivity (23) or serum anti-PT IgG levels exceeding 125 IU/ml four weeks after
115
coughing-onset served as an indicator for a recent B. pertussis infection (‘infected’). Subjects
116
not meeting these criteria were considered ‘non-infected’. Groups A-D included sera from
117
non-infected individuals, stratified by symptoms and vaccination status. Group A originated
118
from the serosurveillance study, groups A+ and B from the Boostrix trial. Individuals from
119
groups A+ and B had received a T, DT or DTPa-vaccine dose less than 5 years earlier and
120
were given a saline placebo (group A+) or a Boostrix dose (group B). Blood was taken 28
121
days later. Groups C and D consisted of recently (≤2 years, group C) and less recently (>2
122
years, group D) vaccinated patients with respiratory symptoms, but serodiagnosed as non-
123
infected (10;24;25). Groups E and F were infected individuals whose pertussis diagnosis was
124
confirmed by real-time PCR (23)(group E) or single-point serology (group F). Since our
125
primary goal is to diagnose infection in recently vaccinated subjects, individuals with
126
unknown (Group A) or no recent vaccination history (Group D) were excluded from the non5
127
infected group for comparison between infected and non-infected subjects. Groups G and H
128
included sera from individuals non-infected and infected, respectively, as based on anti-PT
129
IgG dual-point serology. Supplementary Figure 1 shows the individual anti-PT IgG levels of
130
all subjects and the characteristics of respiratory symptoms in symptomatic subjects. The
131
Belgian vaccination program and vaccines used are shown in Supplementary Figure 2.
132 133
Laboratory investigation
134
Quantification of anti-PT IgA was done using a commercial ELISA (SERION; ESR1201A)
135
calibrated with the WHO International standard 06/140 (NIBSC, UK). Results were expressed
136
as international units (IU)/ml. Commercial ELISAs from Alpha Diagnostic International (San
137
Antonio, USA) were used to quantify IgG responses to FIM2/3 (#960-370-FIG), LPS (#960-
138
410-LHG) and ACT (#960-510-AHG) following manufacturer instructions. Antibody
139
concentrations were calculated relative to internal calibrators and expressed as ELISA
140
Units/ml. The presence of IgG directed to CatACT was evaluated semi-quantitatively using
141
visual scoring in commercial immunoblot-strips (WE116G32, Virotech diagnostics GmbH,
142
Germany).
143 144
Statistical analysis and sample size
145
Differences between categorical variables were analyzed using Pearson’s Chi2 test, and
146
between continuous variables using the Mann-Whitney U-test. Sensitivity and specificity
147
were calculated from 2x2 contingency tables for a range of different cut-offs for each
148
studied antigen. Optimal cut-offs were determined using receiver operating characteristic
149
(ROC) analysis. Methods for sample size calculation are detailed in Supplementary
150
Methods. 6
151 152
Results
153
Specificity and age-dependency of anti-PT IgA, anti-Fim2/3 IgG, anti-LPS IgG, anti-ACT IgG
154
and anti-CatACT IgG responses in B. pertussis non-infected and infected subjects
155
Concentrations of antibodies against the antigens under study were measured in 210 sera
156
from not-recently infected individuals (groups A-D, ‘non-infected’, Figure 1, left panel) and
157
in 140 sera from individuals with an acute B. pertussis infection (groups E-F, ‘infected’,
158
Figure 1, right panel). In non-infected subjects (groups A-D), maximum antibody levels for
159
anti-PT IgA, anti-Fim2/3 IgG, anti-LPS IgG and anti-ACT IgG were 25, 6, 20 and 111 (I)U/ml,
160
respectively. In subjects who received a Boostrix dose four weeks prior to blood sampling
161
(Group B), anti-PT IgA levels never exceeded 15 IU/ml (GMT 6 IU/ml) and were comparable
162
to levels in the other non-infected subjects. Overall, anti-PT IgA and anti-ACT IgG
163
concentrations exceeded 15 U/ml in 4 and 18 subjects, respectively. The majority of
164
elevated anti-ACT IgG and anti-CatACT IgG levels were observed in group D. In infected
165
subjects (groups E-F), remarkably low anti-PT IgA levels were observed in three individuals.
166
Since diagnosis was based on single point serology, a certain degree of misdiagnosis cannot
167
be excluded, especially in the lower age groups.
168
For age-dependency, results from groups A-D and E-F were pooled and antibody
169
concentrations were stratified by age category (Figure 2). Comparing antibody levels
170
between the youngest (10-19 years, n=38) and oldest (70-89 years, n=17) age groups
171
revealed an age-dependent enhancement of anti-PT IgA levels in the infected populations
172
(p<0.0001).
173
7
174
Usefulness of IgA directed to the vaccine antigen PT, and IgG directed to non-vaccine
175
antigens to discriminate between infection and recent vaccination in B. pertussis-suspected
176
subjects.
177
With all data at hand, antibody levels were compared between infected (groups E+F, n=140)
178
and recently vaccinated, non-infected subjects (groups B+C, n=100, Figure 3A-E). Median
179
antibody levels for anti-PT IgA, anti-LPS IgG and anti-ACT IgG in non-infected versus infected
180
subjects were 6 and 42 IU/ml, 3.6 and 10.4 U/ml and 6 and 39 U/ml, respectively. For anti-
181
CatACT IgG, median score was 0 in non-infected and 2 in infected subjects. Differences
182
between groups were highly significant (p<0.0001) for all parameters, except for anti-
183
FIM2/3 IgG. Optimal cut-off values for IgA directed to PT and IgG directed to LPS, ACT and
184
CatACT were 15 IU/ml, 5 U/ml, 15 U/ml and 0.5 score, respectively (Supplementary Table
185
1). Excluding the youngest age-group (10-19 years old) for ROC-analysis resulted in the same
186
optimal cut-off levels.
187
The overall accuracy, specificity and sensitivity for each of the tested parameters is shown in
188
Figure 3F. For anti-PT IgA and anti-ACT IgG, a cut-off level of 15 U/mL resulted in an
189
accuracy of 95% and 87.5%, a specificity of 98% and 93%, and a sensitivity of 92.9% and
190
83.6%, respectively. Excluding 10-19 years old subjects, increased accuracy and sensitivity
191
for anti-PT IgA. Finally, sequential testing of anti-PT IgA and anti-ACT IgG, anti-LPS IgG or
192
anti-CatACT IgG resulted in an increased specificity, at the cost of sensitivity and accuracy
193
(Figure 4).
194 195
Usefulness of anti-PT IgA, anti-ACT IgG and anti-CatACT IgG to diagnose a pertussis infection
196
in the first of paired samples
8
197
At the NRC we seldom receive a second convalescent serum sample, hampering dual point
198
serology. We therefore evaluated whether anti-PT IgA, anti-ACT IgG or anti-CatACT IgG can
199
be used to predict infection at an early time point. The presence of these antibodies was
200
evaluated in two groups of 20 paired samples: non-infected and infected individuals (groups
201
G and H, Supplementary Figure 3A and 3B, respectively). Comparing first of paired samples
202
revealed that only anti-PT IgA is useful for the early prediction of a pertussis infection.
203
Whereas anti-PT IgA levels in all non-infected subjects were below 15 IU/ml (0/20, GMT 5
204
IU/ml), these levels exceeded 15 IU/ml in 11/20 first samples of infected subjects (GMT 17
205
IU/ml) (p<0.001). Measuring anti-PT IgA in the first of paired samples can therefore predict a
206
pertussis infection earlier than anti-PT IgG with 100% specificity but only 50% sensitivity
207
(Supplementary Table 2).
208 209
Discussion
210
Here we show that sequential testing of anti-PT IgA, and to a lesser extent anti-ACT IgG, is
211
useful to discriminate between infection and vaccination in recently vaccinated subjects
212
with elevated anti-PT IgG.
213
Anti-PT IgA are frequently induced after infection in children aged over 4 years, adolescents
214
and adults, but are a less reliable marker for B. pertussis infection than PT-specific IgG
215
(1;6;16;25). The major interest for using anti-PT IgA as additional marker in pertussis
216
serodiagnosis comes from the observations that these responses are only weakly induced by
217
aP-vaccination (1;6;26;27). Primary aP- or wP (whole cell pertussis)-vaccination in the first
218
year of life does not induce IgA antibodies, while boosting with aP at the age of 4 or 9 years
219
or in adolescents and adults, sometimes induces low levels of IgA (14;15;28;29). The main
220
drawbacks for using anti-PT IgA are its reported age-dependency and the lack of a 9
221
consensus cut-off value for recent infection. Increasing IgA-responses by age were reported
222
in infected as well as non-infected individuals (9;17;19). Our survey did not reveal any age-
223
dependent increase of anti-PT IgA responses in non-infected individuals, irrespective of their
224
vaccination status. However, in infected subjects, the previously reported low response in
225
younger individuals was confirmed and anti-PT IgA levels clearly tended to increase with
226
age. Increasing IgA levels in older subjects may reflect strong recall responses induced by
227
repeated B. pertussis exposure over time. An alternative explanation may be the
228
immunization background of the subjects. Since wP-priming was reported to induce higher
229
anti-PT IgA responses than aP-priming (30), the increased anti-PT IgA levels in infected
230
adults may be associated with the switch from wP to aP vaccines in Belgium in 2001. In the
231
present study, where vaccination history of subjects is age-dependent, around 20% of
232
subjects were primed with aP (subjects aged 10-13 years for group A, subjects aged 10-16
233
years for groups C-H, none for groups A+ and B). It can therefore be expected that age-
234
dependent IgA responses in infected subjects will decline with time since the proportion of
235
wP-primed subjects is progressively decreasing in Belgium. Since most countries in the
236
world are still using wP vaccines and type of childhood vaccination has an influence on
237
antibody responses following a B. pertussis infection later in life (30), the usefulness of anti-
238
PT IgA and anti-ACT IgG antibodies for diagnosis in these countries may be different. In the
239
present study, comparing anti-PT IgA responses in infected and non-infected vaccinated
240
subjects revealed an excellent specificity, sensitivity and accuracy for discriminating
241
infection from vaccination. The used infection cut-off of 15 IU/ml for anti-PT IgA is in line
242
with previously proposed cut-off values, ranging from 10 to 20 IU/ml (1;9;31). However,
243
since Belgium uses the Boostrix-vaccine for booster vaccinations, the findings of the study
10
244
only apply for countries where this vaccine is used. Other aP vaccines could possibly give
245
different immunological reactions both post-vaccination and at later infection.
246
The virulence factor ACT elicits long-lasting and high-titered anti-ACT IgG (32;33). While
247
these responses occur in adults and unvaccinated children with pertussis, they are low or
248
even absent in infected vaccinated children (34;35). Furthermore, cross-reactivity due to
249
similarities between the C-terminal region of ACT and other bacterial toxins, suggests that
250
intact ACT is probably not a suitable marker for pertussis serodiagnosis (34). Nevertheless,
251
the present study identifies anti-ACT IgG as a reliable marker to discriminate between
252
infection and vaccination. Watanabe et al. (5) evaluated the usefulness of IgG directed to
253
the Bordetella-specific N-terminal catalytic region of ACT and found it the second most
254
sensitive marker after PT. The present study confirmed that anti-CatACT IgG is a valuable
255
additional marker in serodiagnosis, although performing less well than anti-ACT IgG. This
256
can probably be explained by the drawbacks of the immunoblot technology used (semi-
257
quantitative, visual scoring). The clinical usefulness of Fim3 antibodies in adolescents and
258
adults was demonstrated by Oguchi et al. (36). Although the use of different antigens
259
(native/denaturated) in reported studies cannot be excluded, the present and other studies
260
(reviewed in (6)), did not reveal any added value for pertussis serodiagnosis.
261
Using paired samples revealed that measuring anti-PT IgA can be used for early
262
serodiagnosis, although with low sensitivity. Elevated anti-ACT and CatACT IgG levels were
263
often observed in the first of paired samples of infected as well as non-infected subjects,
264
indicating that these parameters are not useful for early serodiagnosis.
265
The strengths of the present study are the use of validated methods, purified (LPS) and
266
recombinant (PT, FIM, ACT, CatACT) antigens and the comparison of well-characterized
11
267
subject groups. However, the selection of groups based on vaccination-status, infection-
268
status and age may have introduced a sample selection bias.
269
The present results may lead to an improvement of B pertussis serodiagnosis in those cases
270
where anti-PT IgG levels are equivocal. Reflex testing of anti-PT IgA and anti-ACT IgG allows
271
to distinguish between vaccination and infection in recently vaccinated adults with
272
pertussis-like symptoms. Likewise, the use of both markers could help to discriminate
273
between vaccination and recent infection in pertussis anti-PT IgG serosurveillance studies
274
where individual vaccination data are often not available.
275 276
Conflict of interest
277
We declare no conflicts of interest.
278 279
Funding / Acknowledgments
280
This study was funded by Sciensano and the NRC Bordetella pertussis (RIZIV-INAMI). The
281
Boostrix clinical trial was part of the IMI-funded BIOVACSAFE project (EU funded). LS is
282
funded by an EUPHEM-fellowship (European Public Health Microbiology Training Program,
283
ECDC). DP and ID are members of EUPert LabNet.
284 285
Authors’ Contributions
286
ID and LS designed the study and wrote the manuscript. CR performed laboratory work. GLR
287
was a partner of the BIOVACSAFE consortium and principal investigator of the Boostrix
288
clinical trial. HM, AL, KH, GLR and DP provided intellectual input. All authors agreed with the
289
final version of the manuscript.
290 12
291
Reference List
292 293 294 295
(1) Guiso N, Berbers G, Fry NK, He Q, Riffelmann M, Wirsing von Konig CH. What to do and what not to do in serological diagnosis of pertussis: recommendations from EU reference laboratories. Eur J Clin Microbiol Infect Dis 2011 Mar;30(3):307-12.
296 297 298 299
(2) He Q, Barkoff AM, Mertsola J, Glismann S, Bacci S. High heterogeneity in methods used for the laboratory confirmation of pertussis diagnosis among European countries, 2010: integration of epidemiological and laboratory surveillance must include standardisation of methodologies and quality assurance. Euro Surveill 2012 Aug 9;17(32).
300 301 302 303
(3) de Melker HE, Versteegh FG, Conyn-Van Spaendonck MA, Elvers LH, Berbers GA, van Der ZA, et al. Specificity and sensitivity of high levels of immunoglobulin G antibodies against pertussis toxin in a single serum sample for diagnosis of infection with Bordetella pertussis. J Clin Microbiol 2000 Feb;38(2):800-6.
304 305 306
(4) Versteegh FG, Mertens PL, de Melker HE, Roord JJ, Schellekens JF, Teunis PF. Age-specific long-term course of IgG antibodies to pertussis toxin after symptomatic infection with Bordetella pertussis. Epidemiol Infect 2005 Aug;133(4):737-48.
307 308
(5) Watanabe M, Connelly B, Weiss AA. Characterization of serological responses to pertussis. Clin Vaccine Immunol 2006 Mar;13(3):341-8.
309 310
(6) van der Zee A, Schellekens JF, Mooi FR. Laboratory Diagnosis of Pertussis. Clin Microbiol Rev 2015 Oct;28(4):1005-26.
311 312 313 314
(7) Baughman AL, Bisgard KM, Edwards KM, Guris D, Decker MD, Holland K, et al. Establishment of diagnostic cutoff points for levels of serum antibodies to pertussis toxin, filamentous hemagglutinin, and fimbriae in adolescents and adults in the United States. Clin Diagn Lab Immunol 2004 Nov;11(6):1045-53.
315 316 317
(8) Markey K, Douglas-Bardsley A, Asokanathan C, Fry NK, Barkoff AM, Bacci S, et al. Improvement in serological diagnosis of pertussis by external quality assessment. J Med Microbiol 2019 Apr 16.
318 319 320
(9) May ML, Doi SA, King D, Evans J, Robson JM. Prospective evaluation of an Australian pertussis toxin IgG and IgA enzyme immunoassay. Clin Vaccine Immunol 2012 Feb;19(2):1907.
321 322 323
(10) Berbers GA, van de Wetering MS, van Gageldonk PG, Schellekens JF, Versteegh FG, Teunis PF. A novel method for evaluating natural and vaccine induced serological responses to Bordetella pertussis antigens. Vaccine 2013 Aug 12;31(36):3732-8.
324 325 326
(11) Tan T, Dalby T, Forsyth K, Halperin SA, Heininger U, Hozbor D, et al. Pertussis Across the Globe: Recent Epidemiologic Trends From 2000 to 2013. Pediatr Infect Dis J 2015 Sep;34(9):e222-e232.
327 328
(12) Celentano LP, Massari M, Paramatti D, Salmaso S, Tozzi AE. Resurgence of pertussis in Europe. Pediatr Infect Dis J 2005 Sep;24(9):761-5.
13
329 330
(13) Wirsing von Konig CH. Pertussis diagnostics: overview and impact of immunization. Expert Rev Vaccines 2014 Oct;13(10):1167-74.
331 332 333
(14) Hendrikx LH, Ozturk K, de Rond LG, de Greeff SC, Sanders EA, Berbers GA, et al. Serum IgA responses against pertussis proteins in infected and Dutch wP or aP vaccinated children: an additional role in pertussis diagnostics. PLoS One 2011;6(11):e27681.
334 335 336
(15) Le T, Cherry JD, Chang SJ, Knoll MD, Lee ML, Barenkamp S, et al. Immune responses and antibody decay after immunization of adolescents and adults with an acellular pertussis vaccine: the APERT Study. J Infect Dis 2004 Aug 1;190(3):535-44.
337 338
(16) Nagel J, Poot-Scholtens EJ. Serum IgA antibody to Bordetella pertussis as an indicator of infection. J Med Microbiol 1983 Nov;16(4):417-26.
339 340 341 342
(17) van der Zee A, Agterberg C, Peeters M, Mooi F, Schellekens J. A clinical validation of Bordetella pertussis and Bordetella parapertussis polymerase chain reaction: comparison with culture and serology using samples from patients with suspected whooping cough from a highly immunized population. J Infect Dis 1996 Jul;174(1):89-96.
343 344
(18) Prince HE, Lieberman JM, Cherry JD. Age-related differences in patterns of increased Bordetella pertussis antibodies. Clin Vaccine Immunol 2012 Apr;19(4):545-50.
345 346
(19) Deen JL, Mink CA, Cherry JD, Christenson PD, Pineda EF, Lewis K, et al. Household contact study of Bordetella pertussis infections. Clin Infect Dis 1995 Nov;21(5):1211-9.
347 348 349 350
(20) van T, I, Bonacic Marinovic AA, Kuipers B, van Gaans-van den Brink JA, Sanders EA, van Els CA. Impact of age and vaccination history on long-term serological responses after symptomatic B. pertussis infection, a high dimensional data analysis. Sci Rep 2017 Jan 16;7:40328.
351 352 353
(21) Litzroth A, Suin V, Wyndham-Thomas C, Quoilin S, Muyldermans G, Vanwolleghem T, et al. Low hepatitis C prevalence in Belgium: implications for treatment reimbursement and scale up. BMC Public Health 2019 Jan 8;19(1):39.
354 355 356
(22) Cabore RN, Pierard D, Huygen K. A Belgian Serosurveillance/Seroprevalence Study of Diphtheria, Tetanus and Pertussis Using a Luminex xMAP Technology-Based Pentaplex. Vaccines (Basel) 2016 May 10;4(2).
357 358 359
(23) Martini H, Detemmerman L, Soetens O, Yusuf E, Pierard D. Improving specificity of Bordetella pertussis detection using a four target real-time PCR. PLoS One 2017;12(4):e0175587.
360 361 362
(24) Weston W, Messier M, Friedland LR, Wu X, Howe B. Persistence of antibodies 3 years after booster vaccination of adults with combined acellular pertussis, diphtheria and tetanus toxoids vaccine. Vaccine 2011 Nov 3;29(47):8483-6.
363 364
(25) Heininger U, Cherry JD, Stehr K. Serologic response and antibody-titer decay in adults with pertussis. Clin Infect Dis 2004 Feb 15;38(4):591-4.
365 366 367 368
(26) Ward JI, Cherry JD, Chang SJ, Partridge S, Keitel W, Edwards K, et al. Bordetella Pertussis infections in vaccinated and unvaccinated adolescents and adults, as assessed in a national prospective randomized Acellular Pertussis Vaccine Trial (APERT). Clin Infect Dis 2006 Jul 15;43(2):151-7. 14
369 370 371
(27) Mertens PL, Stals FS, Steyerberg EW, Richardus JH. Sensitivity and specificity of single IgA and IgG antibody concentrations for early diagnosis of pertussis in adults: an evaluation for outbreak management in public health practice. BMC Infect Dis 2007 Jun 6;7:53.
372 373 374
(28) Littmann M, Hulsse C, Riffelmann M, Wirsing von Konig CH. Long-term immunogenicity of a single dose of acellular pertussis vaccine in paediatric health-care workers. Vaccine 2008 May 2;26(19):2344-9.
375 376 377
(29) Knuf M, Zepp F, Meyer C, Grzegowski E, Wolter J, Riffelmann M, et al. Immunogenicity of a single dose of reduced-antigen acellular pertussis vaccine in a non-vaccinated adolescent population. Vaccine 2006 Mar 15;24(12):2043-8.
378 379 380
(30) Raeven RHM, van der Maas L, Pennings JLA, Fuursted K, Jorgensen CS, van RE, et al. Antibody Specificity Following a Recent Bordetella pertussis Infection in Adolescence Is Correlated With the Pertussis Vaccine Received in Childhood. Front Immunol 2019;10:1364.
381 382
(31) May ML, Evans J, Holgate T, Doi SA, Ross P, Robson JM. Pertussis toxin IgA testing overdiagnoses recent pertussis infection. Pathology 2017 Dec;49(7):770-5.
383 384 385
(32) Carbonetti NH, Artamonova GV, Andreasen C, Bushar N. Pertussis toxin and adenylate cyclase toxin provide a one-two punch for establishment of Bordetella pertussis infection of the respiratory tract. Infect Immun 2005 May;73(5):2698-703.
386 387 388
(33) Eby JC, Gray MC, Warfel JM, Merkel TJ, Hewlett EL. Use of a Toxin Neutralization Assay To Characterize the Serologic Response to Adenylate Cyclase Toxin after Infection with Bordetella pertussis. Clin Vaccine Immunol 2017 Jan;24(1).
389 390 391
(34) Arciniega JL, Hewlett EL, Johnson FD, Deforest A, Wassilak SG, Onorato IM, et al. Human serologic response to envelope-associated proteins and adenylate cyclase toxin of Bordetella pertussis. J Infect Dis 1991 Jan;163(1):135-42.
392 393 394
(35) Cherry JD, Xing DX, Newland P, Patel K, Heininger U, Corbel MJ. Determination of serum antibody to Bordetella pertussis adenylate cyclase toxin in vaccinated and unvaccinated children and in children and adults with pertussis. Clin Infect Dis 2004 Feb 15;38(4):502-7.
395 396 397 398 399
(36) Oguchi K, Miyata A, Kazuyama Y, Noda A, Suzuki E, Watanabe M, et al. Detection of antibodies against fimbria type 3 (Fim3) is useful diagnostic assay for pertussis. J Infect Chemother 2015 Sep;21(9):639-46.
15
TABLE 1 Selection criteria and characteristics of study groups. origin samples
n
gender age range
pertussis infection
vaccination status
anti-PT IgG pertussis diagnosis based on
(M/F)
(years)
status
symptoms
GMT (IU/ml) 1
no disease, no symptoms, single point serology (anti-PT Ig G < 10 IU/ml)
28
no disease, no symptoms
121**
no disease, no symptoms
group A
serosurv.
50
18/32
10-89
non-INF
not applicable
group A+
clinical trial
10
5/5
18-43
non-INF
not applicable
group B
clinical trial
50
25/25
18-44
non-INF
not applicable
group C
NRC
50
14/36
12-82
non-INF
coughing 3-6w
recent vac ( ≤ 2y)
17
group D
NRC
50
21/29
10-83
non-INF
coughing 3-6w
no recent vac (> 2y)
11
group E
NRC
70
33/37
10-80
INF
coughing 1-6w
mixed
367
based on PCR positivity
group F
NRC
70
32/38
10-89
INF
coughing 1-6w
no recent vac (> 2y)
496
single point serology_ anti-PT IgG > 125 IU/ml
group G
NRC
20
7/13
11-84
non-INF
coughing 1-6w
mixed
group H
NRC
20
8/12
12-82
INF
coughing 1-6w
mixed
unknown booster1_< 5 y* placebo_1 m* booster1_< 5 y* booster2_1 m*
30 (t1) 29(t2) 32 (t1) 341(t2)
single point serology_w3: anti-PT IgG < 10 IU/ml or w4/5/6: anti-PT Ig < 125 IU/ml single point serology_w3: anti-PT IgG < 10 IU/ml or w4/5/6: anti-PT Ig < 125 IU/ml
dual point serology dual point serology
M, male; F, female; GMT, geometric mean titer; NRC, National Reference Centre; non-INF, non-infected with B. pertussis; INF, infected with B. pertussis; w, week; y, year; m, month; t1, time point 1; t2, time point 2; *last dose of vaccine containing T, TD, TDaP or Tdap was given <5 years before administration of Boostrix (or placebo) and blood was drawn 1 month after administration of Boostrix (or placebo); ** anti-FHA IgG (GMT 394 IU/ml) and anti-Prn IgG (GMT 620 IU/ml)
Supplementary TABLE 1 Performance to diagnose a B. pertussis infection for each parameter at different cut-off levels. anti-PT IgA
anti-FIM 2/3 IgG
anti-LPS IgG
anti-ACT IgG
anti-CatACT IgG
cut-off (IU/ml)
cut-off (U/ml)
cut-off (U/ml)
cut-off (U/ml)
cut-off (visual score)
> 10
> 15
> 20
>5
> 10
> 15
>5
> 10
> 15
> 10
> 15
> 20
0
> 0.5
>1
accuracy (%)
93.3
95.0
90.4
41.7
42.5
42.5
77.5
70.0
58.8
86.7
87.5
86.3
78.8
82.1
77.1
specificity (%)
87.0
98.0
98.0
98.0
100
100
69.0
96.0
99.0
82.0
93.0
98.0
58.0
89.0
100
sensitivity (%)
97.9
92.9
85.0
1.4
1.4
1.4
83.6
51.4
30.0
90.0
83.6
77.9
93.6
77.1
60.7
Supplementary TABLE 2 Performance of anti-PT IgA, anti-ACT IgG and anti-CatACT IgG to diagnose a B. pertussis infection in first samples of subjects diagnosed by dual point-serology.
cut-off
anti-PT IgA
anti-ACT IgG
anti-CatACT IgG
> 15 IU/ml
> 15 U/ml
score > 0.5
all subjects
adults only
all subjects
adults only
all subjects
adults only
accuracy, % (95% CI)
75.0 (59.8-85.8)
90 (77.0-96.0)
57.5 (42.2-71.5)
53.1 (36.5-69.1)
67.5 (52.0-79.9)
71.9 (54.6-84.4)
specificity, % (95% CI)
100 (83.9-100)
100 (80.6-100)
60.0 (38.7-68.1)
56.3 (33.1-76.9)
90.0 (69.9-97.2)
100 (80.6-100)
sensitivity, % (95% CI)
50.0 (29.9-70.1)
62.5 (38.6-81.5)
55.0 (34.2-74.2)
50.0 (28.0-72.0)
45.0 (25.8-65.8)
43.7 (23.1-66.8)
S1198-743X(19)30528-2
DOI:
https://doi.org/10.1016/j.cmi.2019.10.001
Reference:
CMI 1798
To appear in:
Clinical Microbiology and Infection
Received Date: 22 June 2019 Revised Date:
22 September 2019
Accepted Date: 1 October 2019
Please cite this article as: Subissi L, Rodeghiero C, Martini H, Litzroth A, Huygen K, Leroux-Roels G, Piérard D, Desombere I, Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve Bordetella pertussis serodiagnosis in recently vaccinated subjects, Clinical Microbiology and Infection, https://doi.org/10.1016/j.cmi.2019.10.001. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 European Society of Clinical Microbiology and Infectious Diseases. Published by Elsevier Ltd. All rights reserved.
1
Assessment of IgA anti-PT and IgG anti-ACT reflex testing to improve
2
Bordetella pertussis serodiagnosis in recently vaccinated subjects
3
Lorenzo Subissi1,2, Caroline Rodeghiero1, Helena Martini3, Amber Litzroth4, Kris Huygen1,
4
Geert Leroux-Roels5, Denis Piérard3, Isabelle Desombere1,*
5
Author affiliations
6
1
7
2
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
National Reference Centre (NRC) for Bordetella pertussis, Sciensano, Brussels, Belgium European Programme for Public Health Microbiology Training, European Centre for Disease Prevention and Control, Stockholm, Sweden 3 National Reference Centre (NRC) for Bordetella pertussis, Universitair Ziekenhuis Brussel, Belgium 4 Epidemiology of Infectious Diseases, Sciensano, Brussels, Belgium 5 Center for Vaccinology, Ghent University, Belgium
* corresponding author:
Isabelle Desombere National Reference Centre for Bordetella pertussis Immune Response Infectious Diseases in Humans Sciensano (Public Health Belgium) Rue Engelandstraat 642 B-1180 Brussels; Belgium Email: [email protected] Tel: +32 2 373 32 94 Fax: +32 2 373 33 67
25 26
Format: Original article
Abstract: 250 words
Manuscript: 2506 words
27
Keywords: Bordetella pertussis, serodiagnosis, serology, IgA, ACT, pertussis toxin, antibody,
28
vaccination
29
Running title: Anti-PT IgA and anti-ACT IgG to diagnose a Bordetella pertussis infection in
30
recently vaccinated subjects.
31
Abstract
32
Objectives. Quantifying IgG antibodies to pertussis toxin (PT) is the most specific and
33
sensitive method to serodiagnose a Bordetella pertussis infection. Since PT is a component
34
of acellular pertussis vaccines, anti-PT IgG is also induced by vaccination, precluding
35
pertussis serodiagnosis exclusively based on anti-PT IgG in recently vaccinated subjects.
36
Here, we aim to identify additional B. pertussis-specific serological markers that can
37
discriminate between infection and recent vaccination.
38
Methods. The clinical utility of measuring IgA directed to the vaccine antigen PT, and IgG
39
directed to non-vaccine antigens (Fim2/3, LPS, ACT, CatACT) was evaluated in 9 well-
40
characterized subject groups, aged 10-89 years (n=390). Serum anti-PT IgG levels (>125
41
IU/ml) served as an indicator for a recent B. pertussis infection. Comparing symptomatic
42
pertussis-infected (n=140) with recently vaccinated, non-infected (n=100) subjects, revealed
43
the optimal cut-off, accuracy, sensitivity and specificity for each single parameter.
44
Results. For pertussis diagnosis in recently vaccinated subjects, the measurement of anti-PT
45
IgA (cut-off 15 IU/mL) and anti-ACT IgG (cut-off 15 U/mL) resulted in accuracies of 95%
46
(91.5-97.1) and 87.5% (82.7-91.1), sensitivities of 92.9% (87.4-96.0) and 83.6% (76.5-88.8)
47
and specificities of 98% (93.0-99.4) and 93% (86.3-96.6), respectively. Comparing anti-PT IgA
48
levels between the youngest (10-19 years, n=38) and oldest (70-89 years, n=17) age groups
49
revealed an age-dependent increase of antibody levels in pertussis-infected subjects
50
(p<0.0001).
51
Conclusions. Reflex testing of anti-PT IgA and anti-ACT IgG improves pertussis serodiagnosis
52
in recently vaccinated symptomatic subjects with elevated anti-PT IgG levels. Furthermore,
53
both markers can discriminate between vaccination and recent infection in pertussis
54
serosurveillance studies. 2
55 56
Introduction
57
Testing for Bordetella pertussis, the etiological agent of whooping cough, is recommended
58
in patients with prolonged cough, which is often the only symptom of pertussis (1;2).
59
Unfortunately, pertussis diagnosis is challenging and does not exclusively rely on laboratory
60
testing. Because the sensitivity of culture and PCR (gold standard for diagnosis) decreases
61
over time, serological techniques are the most sensitive diagnostic method at a later stage
62
of the disease. For this, specific IgG antibodies against B. pertussis toxin (PT) are quantified
63
in serum or oral fluids. However, for a correct interpretation of serology, the knowledge of
64
the duration of symptoms, the age and vaccination status of the patient are essential. When
65
diagnosis cannot be clearly established with one serum sample, a second convalescent
66
sample is requested (1;2).
67
The Belgian National Reference Centre (NRC) follows the ECDC-guidelines and routine
68
serodiagnosis is performed by quantifying circulating anti-PT IgG, endowed with the
69
strongest diagnostic potential in older children, adolescents and adults(1;3-8). PT-induced
70
immune responses are specific for B. pertussis, whereas antibodies directed to other
71
virulence factors are less specific due to cross-reactivity with antigens present in other
72
Bordetella species and other microbial pathogens (1;6;9). PT antigen (in a toxoided form) is
73
also present in all licensed acellular vaccines (aP) in Europe and anti-PT IgG are thus induced
74
by vaccination (10). Since an increase in pertussis incidence has been reported worldwide
75
(11;12) and booster vaccinations with PT-containing aP vaccines are increasingly
76
recommended, the diagnostic power of anti-PT IgG can be hampered by possible
77
interference of previous vaccinations (13). Therefore, additional parameters have been
78
evaluated for their capacity to improve pertussis serodiagnosis in recently vaccinated 3
79
subjects. The clinical utility of measuring anti-PT IgA is still under debate (1;14;15). Interest
80
in measuring IgA, particularly in adolescents and adults, comes from the observation that
81
intramuscular vaccination may not induce IgA secretion, making it a unique infection marker
82
(14;16). However, several studies reported that IgA responses to B. pertussis infection tend
83
to increase with age (6;17-20). By consequence, no broadly accepted cut-off point is
84
available.
85
Antibody responses to B. pertussis antigens absent from aP vaccines have been examined,
86
including adenylate cyclase toxin (ACT), the catalytic domain of ACT (CatACT), lipo-
87
oligosaccharide (LPS) and fimbriae (FIM) (included in some vaccines). Compared to anti-PT
88
IgG, measuring antibodies to these antigens resulted in lower sensitivities and their
89
usefulness as unique markers in pertussis serodiagnosis is limited (6). However, their value
90
as additional markers for serodiagnosis in recently vaccinated subjects is unknown.
91
In here, we evaluated the clinical usefulness of IgA directed to the vaccine antigen PT, and
92
IgG directed to B. pertussis antigens FIM, LPS, ACT and CatACT to improve pertussis
93
diagnosis in symptomatic subjects with elevated anti-PT IgG levels and a history of recent
94
vaccination.
95 96
Methods
97
Study population
98
Study subjects were retrospectively selected from a Belgian serosurveillance study, a
99
Boostrix vaccination trial, and the Belgian NRC for Bordetella pertussis. In the
100
serosurveillance cross-sectional study, residual serum specimens were collected (2013-
101
2015) through voluntarily participating diagnostic laboratories (21). The Boostrix clinical trial
102
was conducted during the IMI-funded BioVacSafe project (2012-2018). Eligible subjects 4
103
were randomized based on duration since their last Diphtheria-Tetanus-Pertussis (DTP)-
104
vaccination and either vaccinated with Boostrix or placebo-treated. DTP-antibodies were
105
quantified in both studies using a bead-based multiplex assay (22). NRC-samples with
106
accompanying clinical data were selected from sera sent to the lab for routine pertussis
107
serodiagnosis (2017-2018) or PCR-diagnosis (2011-2017). Ethics and additional information
108
on study populations and selection of study subjects is available in Supplementary
109
Methods. Since serodiagnosis of pertussis is less indicated in younger children, subjects
110
under the age of 10 were excluded from all study groups.
111 112
Selection and grouping of study subjects
113
The characteristics of each study group are shown in Table 1. For symptomatic patients,
114
PCR-positivity (23) or serum anti-PT IgG levels exceeding 125 IU/ml four weeks after
115
coughing-onset served as an indicator for a recent B. pertussis infection (‘infected’). Subjects
116
not meeting these criteria were considered ‘non-infected’. Groups A-D included sera from
117
non-infected individuals, stratified by symptoms and vaccination status. Group A originated
118
from the serosurveillance study, groups A+ and B from the Boostrix trial. Individuals from
119
groups A+ and B had received a T, DT or DTPa-vaccine dose less than 5 years earlier and
120
were given a saline placebo (group A+) or a Boostrix dose (group B). Blood was taken 28
121
days later. Groups C and D consisted of recently (≤2 years, group C) and less recently (>2
122
years, group D) vaccinated patients with respiratory symptoms, but serodiagnosed as non-
123
infected (10;24;25). Groups E and F were infected individuals whose pertussis diagnosis was
124
confirmed by real-time PCR (23)(group E) or single-point serology (group F). Since our
125
primary goal is to diagnose infection in recently vaccinated subjects, individuals with
126
unknown (Group A) or no recent vaccination history (Group D) were excluded from the non5
127
infected group for comparison between infected and non-infected subjects. Groups G and H
128
included sera from individuals non-infected and infected, respectively, as based on anti-PT
129
IgG dual-point serology. Supplementary Figure 1 shows the individual anti-PT IgG levels of
130
all subjects and the characteristics of respiratory symptoms in symptomatic subjects. The
131
Belgian vaccination program and vaccines used are shown in Supplementary Figure 2.
132 133
Laboratory investigation
134
Quantification of anti-PT IgA was done using a commercial ELISA (SERION; ESR1201A)
135
calibrated with the WHO International standard 06/140 (NIBSC, UK). Results were expressed
136
as international units (IU)/ml. Commercial ELISAs from Alpha Diagnostic International (San
137
Antonio, USA) were used to quantify IgG responses to FIM2/3 (#960-370-FIG), LPS (#960-
138
410-LHG) and ACT (#960-510-AHG) following manufacturer instructions. Antibody
139
concentrations were calculated relative to internal calibrators and expressed as ELISA
140
Units/ml. The presence of IgG directed to CatACT was evaluated semi-quantitatively using
141
visual scoring in commercial immunoblot-strips (WE116G32, Virotech diagnostics GmbH,
142
Germany).
143 144
Statistical analysis and sample size
145
Differences between categorical variables were analyzed using Pearson’s Chi2 test, and
146
between continuous variables using the Mann-Whitney U-test. Sensitivity and specificity
147
were calculated from 2x2 contingency tables for a range of different cut-offs for each
148
studied antigen. Optimal cut-offs were determined using receiver operating characteristic
149
(ROC) analysis. Methods for sample size calculation are detailed in Supplementary
150
Methods. 6
151 152
Results
153
Specificity and age-dependency of anti-PT IgA, anti-Fim2/3 IgG, anti-LPS IgG, anti-ACT IgG
154
and anti-CatACT IgG responses in B. pertussis non-infected and infected subjects
155
Concentrations of antibodies against the antigens under study were measured in 210 sera
156
from not-recently infected individuals (groups A-D, ‘non-infected’, Figure 1, left panel) and
157
in 140 sera from individuals with an acute B. pertussis infection (groups E-F, ‘infected’,
158
Figure 1, right panel). In non-infected subjects (groups A-D), maximum antibody levels for
159
anti-PT IgA, anti-Fim2/3 IgG, anti-LPS IgG and anti-ACT IgG were 25, 6, 20 and 111 (I)U/ml,
160
respectively. In subjects who received a Boostrix dose four weeks prior to blood sampling
161
(Group B), anti-PT IgA levels never exceeded 15 IU/ml (GMT 6 IU/ml) and were comparable
162
to levels in the other non-infected subjects. Overall, anti-PT IgA and anti-ACT IgG
163
concentrations exceeded 15 U/ml in 4 and 18 subjects, respectively. The majority of
164
elevated anti-ACT IgG and anti-CatACT IgG levels were observed in group D. In infected
165
subjects (groups E-F), remarkably low anti-PT IgA levels were observed in three individuals.
166
Since diagnosis was based on single point serology, a certain degree of misdiagnosis cannot
167
be excluded, especially in the lower age groups.
168
For age-dependency, results from groups A-D and E-F were pooled and antibody
169
concentrations were stratified by age category (Figure 2). Comparing antibody levels
170
between the youngest (10-19 years, n=38) and oldest (70-89 years, n=17) age groups
171
revealed an age-dependent enhancement of anti-PT IgA levels in the infected populations
172
(p<0.0001).
173
7
174
Usefulness of IgA directed to the vaccine antigen PT, and IgG directed to non-vaccine
175
antigens to discriminate between infection and recent vaccination in B. pertussis-suspected
176
subjects.
177
With all data at hand, antibody levels were compared between infected (groups E+F, n=140)
178
and recently vaccinated, non-infected subjects (groups B+C, n=100, Figure 3A-E). Median
179
antibody levels for anti-PT IgA, anti-LPS IgG and anti-ACT IgG in non-infected versus infected
180
subjects were 6 and 42 IU/ml, 3.6 and 10.4 U/ml and 6 and 39 U/ml, respectively. For anti-
181
CatACT IgG, median score was 0 in non-infected and 2 in infected subjects. Differences
182
between groups were highly significant (p<0.0001) for all parameters, except for anti-
183
FIM2/3 IgG. Optimal cut-off values for IgA directed to PT and IgG directed to LPS, ACT and
184
CatACT were 15 IU/ml, 5 U/ml, 15 U/ml and 0.5 score, respectively (Supplementary Table
185
1). Excluding the youngest age-group (10-19 years old) for ROC-analysis resulted in the same
186
optimal cut-off levels.
187
The overall accuracy, specificity and sensitivity for each of the tested parameters is shown in
188
Figure 3F. For anti-PT IgA and anti-ACT IgG, a cut-off level of 15 U/mL resulted in an
189
accuracy of 95% and 87.5%, a specificity of 98% and 93%, and a sensitivity of 92.9% and
190
83.6%, respectively. Excluding 10-19 years old subjects, increased accuracy and sensitivity
191
for anti-PT IgA. Finally, sequential testing of anti-PT IgA and anti-ACT IgG, anti-LPS IgG or
192
anti-CatACT IgG resulted in an increased specificity, at the cost of sensitivity and accuracy
193
(Figure 4).
194 195
Usefulness of anti-PT IgA, anti-ACT IgG and anti-CatACT IgG to diagnose a pertussis infection
196
in the first of paired samples
8
197
At the NRC we seldom receive a second convalescent serum sample, hampering dual point
198
serology. We therefore evaluated whether anti-PT IgA, anti-ACT IgG or anti-CatACT IgG can
199
be used to predict infection at an early time point. The presence of these antibodies was
200
evaluated in two groups of 20 paired samples: non-infected and infected individuals (groups
201
G and H, Supplementary Figure 3A and 3B, respectively). Comparing first of paired samples
202
revealed that only anti-PT IgA is useful for the early prediction of a pertussis infection.
203
Whereas anti-PT IgA levels in all non-infected subjects were below 15 IU/ml (0/20, GMT 5
204
IU/ml), these levels exceeded 15 IU/ml in 11/20 first samples of infected subjects (GMT 17
205
IU/ml) (p<0.001). Measuring anti-PT IgA in the first of paired samples can therefore predict a
206
pertussis infection earlier than anti-PT IgG with 100% specificity but only 50% sensitivity
207
(Supplementary Table 2).
208 209
Discussion
210
Here we show that sequential testing of anti-PT IgA, and to a lesser extent anti-ACT IgG, is
211
useful to discriminate between infection and vaccination in recently vaccinated subjects
212
with elevated anti-PT IgG.
213
Anti-PT IgA are frequently induced after infection in children aged over 4 years, adolescents
214
and adults, but are a less reliable marker for B. pertussis infection than PT-specific IgG
215
(1;6;16;25). The major interest for using anti-PT IgA as additional marker in pertussis
216
serodiagnosis comes from the observations that these responses are only weakly induced by
217
aP-vaccination (1;6;26;27). Primary aP- or wP (whole cell pertussis)-vaccination in the first
218
year of life does not induce IgA antibodies, while boosting with aP at the age of 4 or 9 years
219
or in adolescents and adults, sometimes induces low levels of IgA (14;15;28;29). The main
220
drawbacks for using anti-PT IgA are its reported age-dependency and the lack of a 9
221
consensus cut-off value for recent infection. Increasing IgA-responses by age were reported
222
in infected as well as non-infected individuals (9;17;19). Our survey did not reveal any age-
223
dependent increase of anti-PT IgA responses in non-infected individuals, irrespective of their
224
vaccination status. However, in infected subjects, the previously reported low response in
225
younger individuals was confirmed and anti-PT IgA levels clearly tended to increase with
226
age. Increasing IgA levels in older subjects may reflect strong recall responses induced by
227
repeated B. pertussis exposure over time. An alternative explanation may be the
228
immunization background of the subjects. Since wP-priming was reported to induce higher
229
anti-PT IgA responses than aP-priming (30), the increased anti-PT IgA levels in infected
230
adults may be associated with the switch from wP to aP vaccines in Belgium in 2001. In the
231
present study, where vaccination history of subjects is age-dependent, around 20% of
232
subjects were primed with aP (subjects aged 10-13 years for group A, subjects aged 10-16
233
years for groups C-H, none for groups A+ and B). It can therefore be expected that age-
234
dependent IgA responses in infected subjects will decline with time since the proportion of
235
wP-primed subjects is progressively decreasing in Belgium. Since most countries in the
236
world are still using wP vaccines and type of childhood vaccination has an influence on
237
antibody responses following a B. pertussis infection later in life (30), the usefulness of anti-
238
PT IgA and anti-ACT IgG antibodies for diagnosis in these countries may be different. In the
239
present study, comparing anti-PT IgA responses in infected and non-infected vaccinated
240
subjects revealed an excellent specificity, sensitivity and accuracy for discriminating
241
infection from vaccination. The used infection cut-off of 15 IU/ml for anti-PT IgA is in line
242
with previously proposed cut-off values, ranging from 10 to 20 IU/ml (1;9;31). However,
243
since Belgium uses the Boostrix-vaccine for booster vaccinations, the findings of the study
10
244
only apply for countries where this vaccine is used. Other aP vaccines could possibly give
245
different immunological reactions both post-vaccination and at later infection.
246
The virulence factor ACT elicits long-lasting and high-titered anti-ACT IgG (32;33). While
247
these responses occur in adults and unvaccinated children with pertussis, they are low or
248
even absent in infected vaccinated children (34;35). Furthermore, cross-reactivity due to
249
similarities between the C-terminal region of ACT and other bacterial toxins, suggests that
250
intact ACT is probably not a suitable marker for pertussis serodiagnosis (34). Nevertheless,
251
the present study identifies anti-ACT IgG as a reliable marker to discriminate between
252
infection and vaccination. Watanabe et al. (5) evaluated the usefulness of IgG directed to
253
the Bordetella-specific N-terminal catalytic region of ACT and found it the second most
254
sensitive marker after PT. The present study confirmed that anti-CatACT IgG is a valuable
255
additional marker in serodiagnosis, although performing less well than anti-ACT IgG. This
256
can probably be explained by the drawbacks of the immunoblot technology used (semi-
257
quantitative, visual scoring). The clinical usefulness of Fim3 antibodies in adolescents and
258
adults was demonstrated by Oguchi et al. (36). Although the use of different antigens
259
(native/denaturated) in reported studies cannot be excluded, the present and other studies
260
(reviewed in (6)), did not reveal any added value for pertussis serodiagnosis.
261
Using paired samples revealed that measuring anti-PT IgA can be used for early
262
serodiagnosis, although with low sensitivity. Elevated anti-ACT and CatACT IgG levels were
263
often observed in the first of paired samples of infected as well as non-infected subjects,
264
indicating that these parameters are not useful for early serodiagnosis.
265
The strengths of the present study are the use of validated methods, purified (LPS) and
266
recombinant (PT, FIM, ACT, CatACT) antigens and the comparison of well-characterized
11
267
subject groups. However, the selection of groups based on vaccination-status, infection-
268
status and age may have introduced a sample selection bias.
269
The present results may lead to an improvement of B pertussis serodiagnosis in those cases
270
where anti-PT IgG levels are equivocal. Reflex testing of anti-PT IgA and anti-ACT IgG allows
271
to distinguish between vaccination and infection in recently vaccinated adults with
272
pertussis-like symptoms. Likewise, the use of both markers could help to discriminate
273
between vaccination and recent infection in pertussis anti-PT IgG serosurveillance studies
274
where individual vaccination data are often not available.
275 276
Conflict of interest
277
We declare no conflicts of interest.
278 279
Funding / Acknowledgments
280
This study was funded by Sciensano and the NRC Bordetella pertussis (RIZIV-INAMI). The
281
Boostrix clinical trial was part of the IMI-funded BIOVACSAFE project (EU funded). LS is
282
funded by an EUPHEM-fellowship (European Public Health Microbiology Training Program,
283
ECDC). DP and ID are members of EUPert LabNet.
284 285
Authors’ Contributions
286
ID and LS designed the study and wrote the manuscript. CR performed laboratory work. GLR
287
was a partner of the BIOVACSAFE consortium and principal investigator of the Boostrix
288
clinical trial. HM, AL, KH, GLR and DP provided intellectual input. All authors agreed with the
289
final version of the manuscript.
290 12
291
Reference List
292 293 294 295
(1) Guiso N, Berbers G, Fry NK, He Q, Riffelmann M, Wirsing von Konig CH. What to do and what not to do in serological diagnosis of pertussis: recommendations from EU reference laboratories. Eur J Clin Microbiol Infect Dis 2011 Mar;30(3):307-12.
296 297 298 299
(2) He Q, Barkoff AM, Mertsola J, Glismann S, Bacci S. High heterogeneity in methods used for the laboratory confirmation of pertussis diagnosis among European countries, 2010: integration of epidemiological and laboratory surveillance must include standardisation of methodologies and quality assurance. Euro Surveill 2012 Aug 9;17(32).
300 301 302 303
(3) de Melker HE, Versteegh FG, Conyn-Van Spaendonck MA, Elvers LH, Berbers GA, van Der ZA, et al. Specificity and sensitivity of high levels of immunoglobulin G antibodies against pertussis toxin in a single serum sample for diagnosis of infection with Bordetella pertussis. J Clin Microbiol 2000 Feb;38(2):800-6.
304 305 306
(4) Versteegh FG, Mertens PL, de Melker HE, Roord JJ, Schellekens JF, Teunis PF. Age-specific long-term course of IgG antibodies to pertussis toxin after symptomatic infection with Bordetella pertussis. Epidemiol Infect 2005 Aug;133(4):737-48.
307 308
(5) Watanabe M, Connelly B, Weiss AA. Characterization of serological responses to pertussis. Clin Vaccine Immunol 2006 Mar;13(3):341-8.
309 310
(6) van der Zee A, Schellekens JF, Mooi FR. Laboratory Diagnosis of Pertussis. Clin Microbiol Rev 2015 Oct;28(4):1005-26.
311 312 313 314
(7) Baughman AL, Bisgard KM, Edwards KM, Guris D, Decker MD, Holland K, et al. Establishment of diagnostic cutoff points for levels of serum antibodies to pertussis toxin, filamentous hemagglutinin, and fimbriae in adolescents and adults in the United States. Clin Diagn Lab Immunol 2004 Nov;11(6):1045-53.
315 316 317
(8) Markey K, Douglas-Bardsley A, Asokanathan C, Fry NK, Barkoff AM, Bacci S, et al. Improvement in serological diagnosis of pertussis by external quality assessment. J Med Microbiol 2019 Apr 16.
318 319 320
(9) May ML, Doi SA, King D, Evans J, Robson JM. Prospective evaluation of an Australian pertussis toxin IgG and IgA enzyme immunoassay. Clin Vaccine Immunol 2012 Feb;19(2):1907.
321 322 323
(10) Berbers GA, van de Wetering MS, van Gageldonk PG, Schellekens JF, Versteegh FG, Teunis PF. A novel method for evaluating natural and vaccine induced serological responses to Bordetella pertussis antigens. Vaccine 2013 Aug 12;31(36):3732-8.
324 325 326
(11) Tan T, Dalby T, Forsyth K, Halperin SA, Heininger U, Hozbor D, et al. Pertussis Across the Globe: Recent Epidemiologic Trends From 2000 to 2013. Pediatr Infect Dis J 2015 Sep;34(9):e222-e232.
327 328
(12) Celentano LP, Massari M, Paramatti D, Salmaso S, Tozzi AE. Resurgence of pertussis in Europe. Pediatr Infect Dis J 2005 Sep;24(9):761-5.
13
329 330
(13) Wirsing von Konig CH. Pertussis diagnostics: overview and impact of immunization. Expert Rev Vaccines 2014 Oct;13(10):1167-74.
331 332 333
(14) Hendrikx LH, Ozturk K, de Rond LG, de Greeff SC, Sanders EA, Berbers GA, et al. Serum IgA responses against pertussis proteins in infected and Dutch wP or aP vaccinated children: an additional role in pertussis diagnostics. PLoS One 2011;6(11):e27681.
334 335 336
(15) Le T, Cherry JD, Chang SJ, Knoll MD, Lee ML, Barenkamp S, et al. Immune responses and antibody decay after immunization of adolescents and adults with an acellular pertussis vaccine: the APERT Study. J Infect Dis 2004 Aug 1;190(3):535-44.
337 338
(16) Nagel J, Poot-Scholtens EJ. Serum IgA antibody to Bordetella pertussis as an indicator of infection. J Med Microbiol 1983 Nov;16(4):417-26.
339 340 341 342
(17) van der Zee A, Agterberg C, Peeters M, Mooi F, Schellekens J. A clinical validation of Bordetella pertussis and Bordetella parapertussis polymerase chain reaction: comparison with culture and serology using samples from patients with suspected whooping cough from a highly immunized population. J Infect Dis 1996 Jul;174(1):89-96.
343 344
(18) Prince HE, Lieberman JM, Cherry JD. Age-related differences in patterns of increased Bordetella pertussis antibodies. Clin Vaccine Immunol 2012 Apr;19(4):545-50.
345 346
(19) Deen JL, Mink CA, Cherry JD, Christenson PD, Pineda EF, Lewis K, et al. Household contact study of Bordetella pertussis infections. Clin Infect Dis 1995 Nov;21(5):1211-9.
347 348 349 350
(20) van T, I, Bonacic Marinovic AA, Kuipers B, van Gaans-van den Brink JA, Sanders EA, van Els CA. Impact of age and vaccination history on long-term serological responses after symptomatic B. pertussis infection, a high dimensional data analysis. Sci Rep 2017 Jan 16;7:40328.
351 352 353
(21) Litzroth A, Suin V, Wyndham-Thomas C, Quoilin S, Muyldermans G, Vanwolleghem T, et al. Low hepatitis C prevalence in Belgium: implications for treatment reimbursement and scale up. BMC Public Health 2019 Jan 8;19(1):39.
354 355 356
(22) Cabore RN, Pierard D, Huygen K. A Belgian Serosurveillance/Seroprevalence Study of Diphtheria, Tetanus and Pertussis Using a Luminex xMAP Technology-Based Pentaplex. Vaccines (Basel) 2016 May 10;4(2).
357 358 359
(23) Martini H, Detemmerman L, Soetens O, Yusuf E, Pierard D. Improving specificity of Bordetella pertussis detection using a four target real-time PCR. PLoS One 2017;12(4):e0175587.
360 361 362
(24) Weston W, Messier M, Friedland LR, Wu X, Howe B. Persistence of antibodies 3 years after booster vaccination of adults with combined acellular pertussis, diphtheria and tetanus toxoids vaccine. Vaccine 2011 Nov 3;29(47):8483-6.
363 364
(25) Heininger U, Cherry JD, Stehr K. Serologic response and antibody-titer decay in adults with pertussis. Clin Infect Dis 2004 Feb 15;38(4):591-4.
365 366 367 368
(26) Ward JI, Cherry JD, Chang SJ, Partridge S, Keitel W, Edwards K, et al. Bordetella Pertussis infections in vaccinated and unvaccinated adolescents and adults, as assessed in a national prospective randomized Acellular Pertussis Vaccine Trial (APERT). Clin Infect Dis 2006 Jul 15;43(2):151-7. 14
369 370 371
(27) Mertens PL, Stals FS, Steyerberg EW, Richardus JH. Sensitivity and specificity of single IgA and IgG antibody concentrations for early diagnosis of pertussis in adults: an evaluation for outbreak management in public health practice. BMC Infect Dis 2007 Jun 6;7:53.
372 373 374
(28) Littmann M, Hulsse C, Riffelmann M, Wirsing von Konig CH. Long-term immunogenicity of a single dose of acellular pertussis vaccine in paediatric health-care workers. Vaccine 2008 May 2;26(19):2344-9.
375 376 377
(29) Knuf M, Zepp F, Meyer C, Grzegowski E, Wolter J, Riffelmann M, et al. Immunogenicity of a single dose of reduced-antigen acellular pertussis vaccine in a non-vaccinated adolescent population. Vaccine 2006 Mar 15;24(12):2043-8.
378 379 380
(30) Raeven RHM, van der Maas L, Pennings JLA, Fuursted K, Jorgensen CS, van RE, et al. Antibody Specificity Following a Recent Bordetella pertussis Infection in Adolescence Is Correlated With the Pertussis Vaccine Received in Childhood. Front Immunol 2019;10:1364.
381 382
(31) May ML, Evans J, Holgate T, Doi SA, Ross P, Robson JM. Pertussis toxin IgA testing overdiagnoses recent pertussis infection. Pathology 2017 Dec;49(7):770-5.
383 384 385
(32) Carbonetti NH, Artamonova GV, Andreasen C, Bushar N. Pertussis toxin and adenylate cyclase toxin provide a one-two punch for establishment of Bordetella pertussis infection of the respiratory tract. Infect Immun 2005 May;73(5):2698-703.
386 387 388
(33) Eby JC, Gray MC, Warfel JM, Merkel TJ, Hewlett EL. Use of a Toxin Neutralization Assay To Characterize the Serologic Response to Adenylate Cyclase Toxin after Infection with Bordetella pertussis. Clin Vaccine Immunol 2017 Jan;24(1).
389 390 391
(34) Arciniega JL, Hewlett EL, Johnson FD, Deforest A, Wassilak SG, Onorato IM, et al. Human serologic response to envelope-associated proteins and adenylate cyclase toxin of Bordetella pertussis. J Infect Dis 1991 Jan;163(1):135-42.
392 393 394
(35) Cherry JD, Xing DX, Newland P, Patel K, Heininger U, Corbel MJ. Determination of serum antibody to Bordetella pertussis adenylate cyclase toxin in vaccinated and unvaccinated children and in children and adults with pertussis. Clin Infect Dis 2004 Feb 15;38(4):502-7.
395 396 397 398 399
(36) Oguchi K, Miyata A, Kazuyama Y, Noda A, Suzuki E, Watanabe M, et al. Detection of antibodies against fimbria type 3 (Fim3) is useful diagnostic assay for pertussis. J Infect Chemother 2015 Sep;21(9):639-46.
15
TABLE 1 Selection criteria and characteristics of study groups. origin samples
n
gender age range
pertussis infection
vaccination status
anti-PT IgG pertussis diagnosis based on
(M/F)
(years)
status
symptoms
GMT (IU/ml) 1
no disease, no symptoms, single point serology (anti-PT Ig G < 10 IU/ml)
28
no disease, no symptoms
121**
no disease, no symptoms
group A
serosurv.
50
18/32
10-89
non-INF
not applicable
group A+
clinical trial
10
5/5
18-43
non-INF
not applicable
group B
clinical trial
50
25/25
18-44
non-INF
not applicable
group C
NRC
50
14/36
12-82
non-INF
coughing 3-6w
recent vac ( ≤ 2y)
17
group D
NRC
50
21/29
10-83
non-INF
coughing 3-6w
no recent vac (> 2y)
11
group E
NRC
70
33/37
10-80
INF
coughing 1-6w
mixed
367
based on PCR positivity
group F
NRC
70
32/38
10-89
INF
coughing 1-6w
no recent vac (> 2y)
496
single point serology_ anti-PT IgG > 125 IU/ml
group G
NRC
20
7/13
11-84
non-INF
coughing 1-6w
mixed
group H
NRC
20
8/12
12-82
INF
coughing 1-6w
mixed
unknown booster1_< 5 y* placebo_1 m* booster1_< 5 y* booster2_1 m*
30 (t1) 29(t2) 32 (t1) 341(t2)
single point serology_w3: anti-PT IgG < 10 IU/ml or w4/5/6: anti-PT Ig < 125 IU/ml single point serology_w3: anti-PT IgG < 10 IU/ml or w4/5/6: anti-PT Ig < 125 IU/ml
dual point serology dual point serology
M, male; F, female; GMT, geometric mean titer; NRC, National Reference Centre; non-INF, non-infected with B. pertussis; INF, infected with B. pertussis; w, week; y, year; m, month; t1, time point 1; t2, time point 2; *last dose of vaccine containing T, TD, TDaP or Tdap was given <5 years before administration of Boostrix (or placebo) and blood was drawn 1 month after administration of Boostrix (or placebo); ** anti-FHA IgG (GMT 394 IU/ml) and anti-Prn IgG (GMT 620 IU/ml)
Supplementary TABLE 1 Performance to diagnose a B. pertussis infection for each parameter at different cut-off levels. anti-PT IgA
anti-FIM 2/3 IgG
anti-LPS IgG
anti-ACT IgG
anti-CatACT IgG
cut-off (IU/ml)
cut-off (U/ml)
cut-off (U/ml)
cut-off (U/ml)
cut-off (visual score)
> 10
> 15
> 20
>5
> 10
> 15
>5
> 10
> 15
> 10
> 15
> 20
0
> 0.5
>1
accuracy (%)
93.3
95.0
90.4
41.7
42.5
42.5
77.5
70.0
58.8
86.7
87.5
86.3
78.8
82.1
77.1
specificity (%)
87.0
98.0
98.0
98.0
100
100
69.0
96.0
99.0
82.0
93.0
98.0
58.0
89.0
100
sensitivity (%)
97.9
92.9
85.0
1.4
1.4
1.4
83.6
51.4
30.0
90.0
83.6
77.9
93.6
77.1
60.7
Supplementary TABLE 2 Performance of anti-PT IgA, anti-ACT IgG and anti-CatACT IgG to diagnose a B. pertussis infection in first samples of subjects diagnosed by dual point-serology.
cut-off
anti-PT IgA
anti-ACT IgG
anti-CatACT IgG
> 15 IU/ml
> 15 U/ml
score > 0.5
all subjects
adults only
all subjects
adults only
all subjects
adults only
accuracy, % (95% CI)
75.0 (59.8-85.8)
90 (77.0-96.0)
57.5 (42.2-71.5)
53.1 (36.5-69.1)
67.5 (52.0-79.9)
71.9 (54.6-84.4)
specificity, % (95% CI)
100 (83.9-100)
100 (80.6-100)
60.0 (38.7-68.1)
56.3 (33.1-76.9)
90.0 (69.9-97.2)
100 (80.6-100)
sensitivity, % (95% CI)
50.0 (29.9-70.1)
62.5 (38.6-81.5)
55.0 (34.2-74.2)
50.0 (28.0-72.0)
45.0 (25.8-65.8)
43.7 (23.1-66.8)