Development and evaluation of a multi-antigen peptide ELISA for the diagnosis of Chlamydia trachomatis-related infertility in women

Transcription

1 Journal of Medical Microbiology (2016), 65, DOI /jmm Development and evaluation of a multi-antigen peptide ELISA for the diagnosis of Chlamydia trachomatis-related infertility in women Shruti Menon, 1 Scott H. Stansfield, 1 Benignus Logan, 2 Jane S. Hocking, 3 Peter Timms, 4 Luk Rombauts, 5 John A. Allan 6,7 and Wilhelmina M. Huston 1,6,8 Correspondence Wilhelmina M. Huston wilhelmina.huston@uts.edu.au or willaonthego@gmail.com 1 Institute of Health and Biomedical Innovation, Queensland University of Technology, Kelvin Grove, Queensland 4059, Australia 2 Griffith University and UC Health Clinical School, The Wesley Hospital, Auchenflower, Queensland 4066, Australia 3 Centre for Epidemiology and Biostatistics, Melbourne School of Population Health, University of Melbourne, Carlton 3053, Australia 4 Faculty of Science, Health, Education and Engineering, University of the Sunshine Coast, Maroochydore, Queensland 4558, Australia 5 MIMR-PH Institute of Medical Research, Monash, Victoria, Australia 6 Wesley Medical Research, The Wesley Hospital, Auchenflower, Queensland 4066, Australia 7 UC Health Clinical School, The Wesley Hospital, Auchenflower, Queensland 4066, Australia 8 School of Life Sciences, Faculty of Science, University of Technology Sydney, Broadway, New South Wales 2007, Australia Received 31 May 2016 Accepted 8 July 2016 Chlamydia trachomatis results in tubal factor infertility in some women. Diagnosis of this tubal infertility is difficult and typically involves laparoscopy or hysterosalpingography to detect the tubal blockages. Numerous serological tests have been developed; however, they are presently not used for diagnosis without subsequent surgical investigation during the infertility investigation. This study aimed to develop a highly specific serological assay for chlamydial tubal factor infertility in women that could be used to recommend direct progression to in vitro fertilization (IVF) treatment for women who are positive. Women were recruited from a variety of settings including women seeking fertility treatment, sexual health and general practitioner (GP) consultations or antenatal care (n=259). The serological assay was developed using sera from a large group of women by using infertile microimmunofluorescence (MIF)-positive women with tubal damage as the positives compared to infertile or acute infection and/or fertile controls (atives). The new multi-peptide ELISA was highly specific for the detection of tubal factor infertility (P=0.011) compared to another ELISA (P=0.022) and MIF (P=0.099). The sensitivity of the assay should be improved before clinical utility. Potentially, a two-step testing protocol could be used during the initial infertility investigation, where MIF followed by a highly specific ELISA could be used to recommend direct progression to IVF for women who are positive. Abbreviations: HREC, Human Research Ethics Committee; HRP, horseradish peroxidase; IVF, in vitro fertilization; MIF, microimmunofluorescence; ROC, receiver operating characteristic; STI, sexually transmitted infection. Four supplementary tables are available with the online Supplementary Material ã 2016 Printed in Great Britain 915

2 S. Menon and others INTRODUCTION Chlamydia trachomatis is one of the leading causes of tubal infertility in women and is likely to be responsible for up to 45 % of tubal infertility cases (Price et al., 2012). This sexually transmitted infection (STI) is one of the most common bacterial STIs worldwide, with cases reported in the USA in 2014 (Sexually transmitted diseases surveillence, 2014). The infection progresses to pelvic inflammatory disease in approximately 9.5 % of cases (Oakeshott et al., 2010). While it is less clear how many infections progress to tubal infertility or ectopic pregnancy, the infection is clearly responsible for substantial reproductive morbidity worldwide [reviewed by Menon et al. (2015)]. The reproductive morbidity in women results from fallopian tube tissue damage, such as tubal adhesions, tubal occlusion and/or salpingitis. This tissue damage is generally detectable by investigations such as hysterosalpingography or laparoscopy (Swart et al., 1995). However, the tissue damage may not always be apparent as in a large study of 1279 women, of whom 70 were Chlamydia antibody positive but did not always have tubal damage detected and had 50 % lower pregnancy success rates until in vitro fertilization (IVF) treatment, where they had equivalent cumulative success rates (Keltz et al., 2013). This and other similar findings have resulted in the suggestion that Chlamydia antibody testing should be routinely implemented during the initial infertility investigation to identify women with chlamydial infertility who possibly should be investigated for tubal damage and/or are likely to need IVF treatment (Keltz et al., 2006; Coppus et al., 2011; Meikle et al., 1994; Mueller et al., 1991; Persson et al., 1999; Toye et al., 1993). In spite of this, chlamydial serology has not been universally implemented in fertility clinics, owing to both the varying accuracy of the assays and different protocols for using the serological result. The test results are most frequently used to identify women who can be recommended to undergo further tubal investigation (i.e. surgical investigation) or alternatively, they could be used as an indicator of likely tubal infertility to recommend direct progression to IVF treatment. A meta-analysis of the performance of various serological Chlamydia antibody tests for the prediction of tubal infertility reported that microimmunofluorescence (MIF) has the highest accuracy (compared to ELISA and immunofluorescence) but only a moderate ability to discriminate between women with and without tubal pathology (Broeze et al., 2011). This meta-analysis demonstrated that MIF has the lowest specificity, meaning up to 34 % of positives are not tubal pathology; thus, MIF is generally only used to identify women who require further tubal pathology investigation. The goal of this study was to develop a highly specific ELISA that detects women with C. trachomatis serology and tubal pathology that could be used to recommend progression to IVF treatment without further testing for tubal pathology. METHODS Participant recruitment. Four separate groups of women were included in this study: women with infertility with known tubal status (used to establish assay threshold). Women attending an IVF clinic were prospectively recruited. Only women whose medical record indicated that their tubal status had been investigated were eligible. Group 2 women with proven fertility. These women were recruited from antenatal clinics at a private tertiary care hospital. Women were eligible for participation if they were pregnant within less than a year of trying, with no assisted reproductive technologies used for this or any prior pregnancies. A history of seeking fertility treatment was an exclusion criterion. Group 3 women with a past or current C. trachomatis diagnosis. These women were prospectively recruited at sexual health and general practitioner (GP) clinics and at a university health practice. Women were eligible for inclusion if they had either a history of C. trachomatis infection (nucleic acid amplification test diagnosis from genital specimen) or were returning for antibiotic treatment after a recent positive test result for genital C. trachomatis infection (nucleic acid amplification test result). Women were excluded if they had evidence of PID (Pelvic Inflammatory Disease). Group 4 validation cohort of women with infertility. Women were both prospectively and retrospectively recruited from a single IVF clinic (a separate clinic to the site where the first infertile group were recruited). Women were eligible for inclusion if they were either attending to undergo IVF treatment or contacted as they were pregnant from recent IVF treatment. In each of these four group, women provided written informed consent, completed a brief questionnaire and provided a blood specimen for Chlamydia serology. In addition, women in the two infertility groups (groups 1 and 4) consented to ongoing monitoring of their IVF medical chart. Ethics approval was obtained from Monash Private Surgical Human Research Ethics Committee (HREC) (approval no ) (group 1); UC Health Human Research Ethics Committee (approval no. 1221) (group 2); Prince Charles Hospital HREC EC2809, Ipswich and West Moreton Health Services District 10-09, Gold Coast Hospital District , Cairns Sexual Health HREC/09/QCH/4 554 and QUT HREC (group 3); and UC Health HREC (approval no. 1314). ELISA method and statistical analysis. The ELISA method was as previously described by Stansfield et al. (2013). Briefly, the ELISA developed here based on the previous pilot study included 12-mer peptide sequences unique to immunodominant chlamydial antigens with a biotin motif to facilitate binding to streptavidin plates. Several of the peptides in the previous study were evaluated at a series of concentrations, formulations and combinations against the samples for this study to identify and further develop the best performing test. The best multiplewell assay here included three peptides (P11: ADTRGILVVAVE; P47: VFSSPPFSNKPP; P50: PVSFSGPTKGTIT) added to individual wells of streptavidin-coated plates (Pierce). The peptides were added at 0.25 µg in each well and coated overnight at 4 C; a commercial blocking buffer (Superblock, PBS blocking buffer with 0.1 % Tween-20; Pierce) was used for the coating and antibody steps, human serum was used at 1 : 200, anti-human IgG horseradish peroxidase (HRP) (Goat anti- Human IgG Secondary Antibody, HRP conjugate; Invitrogen) was used at 1 : as secondary and Sigma HRP colorimetric detection was used to detect the activity. Medac MOMP and Medac chsp60, combined to give the Medac Infertile (Medac), and MIF (Focus Diagnostics Microimmunofluorescence IgG) were conducted on every specimen in accordance with the manufacturer s instructions. Specimens that were unequivocal in any test after a minimum of two repeat tests were excluded from analysis for that assay. The multi-peptide ELISA developed during this study was established using analysis by receiver operating characteristic (ROC) curves in R for all conditions tested, including varied combinations of wells, concentrations of peptides, solutes and binding conditions. The training was controlled for greater than 95 % specificity. The threshold was established 916 Journal of Medical Microbiology 65

3 ELISA for Chlamydia tubal factor infertility Establish threshold table 1 Positive Negative Negative Negative tubal infertility MIF positive (n=11) MIF ative (n=80) Tubal infertility table 2 tubal infertility (n=44) no tubal factor (n=47) MIF tubal atives table 3 tubal infertility MIF positive (n=11) MIF ative (n=80) Antenatal fertile pregnant>1 year no ART (n=53) Group 2 Sexual health, GP/university Chlamydia past or present (n=112) Group 3 Validation table 4 Fertility clinic 2 tubal infertility (n=19) Group 4 Fertility clinic 2 other causes or unknown tubal status (n=54) Group 4 Fig. 1. Flow chart of study design. The figure shows a flow chart of the participant group who were recruited and the analysis conducted to develop, evaluate and validate the peptide assay. ART, assisted reproductive technologies. on the infertile cohort with known tubal infertility and MIF-positive status (Fig. 1). The optimal assay and peptides were then compared to tubal status or not within this cohort (group 1) and then against a much larger ative control cohort of pregnant and acute infection women (groups 2 and 3). The final evaluation was to test or validate the results on a separate infertile cohort (group 4). The sensitivity, specificity, positive predictive value, ative predictive value and Fisher s exact P value for the difference between assay-positive and true-positive grouping were analysed and provided within each results table. RESULTS A multi-peptide ELISA was more specific for chlamydial infertility but less sensitive than the existing tests A series of test development and evaluation steps was conducted (outlined in Fig. 1 flow chart). First, a threshold for the new multi-peptide ELISA was developed by defining the true-positive women as infertile women with both MIF C. trachomatis (CT)-positive serology and confirmed tubal infertility (n=11). The atives included 80 other participants who were also infertile but either had no tubal factor diagnosed or ative C. trachomatis serology (Table 1). MIF (used to define true positives) detected five participants (false positives) who did not have any tubal damage and were considered ative (94 % specificity). The Medac Infertile assay did not detect any false positives (100 % specificity) but only detected 5 of the 11 positives (45 % sensitivity). The leading new test from those trialled for this project was a combination of three peptides in separate wells that detected 3 of the 11 positives (27 % sensitivity) and 4 of the 82 atives (95 % specificity) in this first analysis. These absorbance thresholds were then used in all of the following evaluations. The thresholds were determined using ROC set for >95 % specificity (Fig. 2, each individual peptide ROC). The threshold for the highest performing assay (combined P11 or P47 or P50) had >0.5 absorbance for positive result. In the n=11 true-positive group of tubal infertile MIF-positive women, the average absorbance of the three [n=3 positives (P11/P47/ P50)] was with an SD of 0.547, and in the infertile and acute-infection-ative women, the average absorbance of the three was with an SD of Therefore, there is a clear distinction between positive and ative results in the test. We further assessed the distinction between the signal and background noise or signal in the ative assays by conducting a signal-to-noise calculation (mean of positive signal divided by SD of the noise is equal to 7.14). This indicates that the true-positive signal is considerably greater than the background noise in the assay. A difficulty commonly encountered with trying to improve the diagnostic performance of chlamydial serological assays is comparing against MIF results which have low specificity (possibly due to cross-reactivity) (Gijsen et al., 2001). This 917

4 S. Menon and others Table 1. Establishment of a threshold for the multi-peptide ELISA using MIF serology and tubal status to define chlamydial infertility Assay Absorbance threshold criteria* pos pos Sensitivity Specificity Positive predictive value Negative predictive value Odds ratio Fisher s exact P value (unadjusted) Adjusted odds ratio Medac Infertile ( ) 1 (0.93 1) 1 (0.36 1) 0.93 ( ) 0 < e+17 (0 Inf) MIF.CT (0.62 1) 0.95 ( ) 0.73 ( ) 1 (0.93 1) 0 < e+40 (0 Inf) MIF.CP ( ) 0.74 ( ) 0.26 ( ) 0.94 ( ) 5.08 ( ) (1.3 28) MIF.CS ( ) 0.99 (0.93 1) 0.8 ( ) 0.92 ( ) 44 ( ) ( ) P (0 0.41) 0.95 ( ) 0.2 ( ) 0.88 ( ) 1.85 ( ) ( ) P (0 0.41) 0.95 ( ) 0.2 ( ) 0.88 ( ) 1.85 ( ) ( ) P (0 0.41) 0.95 ( ) 0.2 ( ) 0.88 ( ) 1.85 ( ) ( ) or P47 or P50 and P47 and P50 and P50 P or P or P P and P and P P and P ( ) 0.95 ( ) 0.43 ( ) 0.9 ( ) 6.94 ( ) (0.6 36) ( ) 0.95 ( ) 0.33 ( ) 0.89 ( ) 4.11 ( ) ( ) ( ) 0.95 ( ) 0.33 ( ) 0.89 ( ) 4.11 ( ) ( ) CI, Confidence interval; MIF.CP, Microimmunofluorescence Chlamydia pneumoniae; MIF.CS, Microimmunofluorescence Chlamydia psittaci. *Analysis used to set the peptide assay threshold for all subsequent evaluations (group 1 participants only). Positives were group 1 (infertile women seeking IVF) with both confirmed tubal factor and MIF CT positivity (all participants from group 1) (n=11). Negatives are infertile women seeking IVF (can be tubal factor or not, can be MIF positive or not, but cannot be positive for both) (group 1) (n=80). 918 Journal of Medical Microbiology 65

5 ELISA for Chlamydia tubal factor infertility Sensitivity Specificity cross-reactivity was apparent in this study where MIF Chlamydia pneumoniae and MIF Chlamydia psittaci concordance was observed (Table 1). We next tested if the assays were significantly more likely to be positive in women with confirmed tubal factor infertility (regardless of MIF test result). This analysis showed that the Medac Infertile test and the new multi-peptide ELISA were highly specific for tubal factor infertility, whereas the specificity for MIF was lower at 89 % (Table 2). The combination multi-peptide assay produced a lower false-positive rate (0/47; P=0.011) for women with tubal infertility than MIF (5/47; P=0.099) and equivalent to Medac Infertile (0/46; P=0.022) (Table 2). In this analysis (Table 2), all tests appeared to have low sensitivity because not all tubal factor cases are caused by C. trachomatis. To confirm the specificity to chlamydial infertility, rather than just a history of chlamydial infection, we next compared the prediction of true positives (for this analysis, true positives were defined as MIF positive and confirmed tubal infertility; n=11) against infertile women, fertile women (group 2) and participants from sexual health and GP clinics (group 3). The peptide assays had the highest specificity of all tests [the combination assay only detecting 18 of the atives (93 % specificity), compared to 25 for Medac Infertile (89 % specificity) and 84 for MIF CT (66 % specificity)] (Table 3). C. trachomatis serological results show concordance P 11 P 47 P Fig. 2. ROC curve of the three top peptides that were combined to make the Peptide 11 ELISA combination assay (as outlined in Tables 1, 2 and 3). The figure shows each of the ROCs of the three peptide assays (key shown on figure). The tests were significantly more likely to be positive in the same women with significant concordance observed between MIF CT and Medac Infertile test (P<0.001) and Peptide 11 and Medac Infertile (P=0.038), and there was a trend towards concordance with MIF CT and Peptide 11 (P=0.058) (Table S1, available in the online Supplementary Material). Peptide 50 serology was significantly concordant with MIF C. pneumoniae positivity (P=0.009), even though in silico analysis was used to rule out peptides that showed similarity with C. pneumoniae in the original screen for these peptides (Stansfield et al., 2013). Live birth outcomes from the IVF treatment are the same for chlamydial sero-positive and sero-ative women The infertile participants were all either seeking to commence or already undergoing IVF treatment. It has previously been reported that C. trachomatis sero-positive women have the same cumulative IVF success as women with other causes of infertility (Keltz et al., 2013). For a chlamydial infertility assay to be effectively used during the infertility investigation, it is important that women positive in the assay have equivalent likelihood of IVF success to other infertile women. Therefore, we evaluated the IVF outcomes from this study, which demonstrated that there was no significant difference in live birth outcomes for Chlamydia sero-positive or -ative women (Table S2). Validation of the ELISA in a second fertility clinic site To validate the performance of the multiple antigen peptide ELISA, we recruited a second group of infertile women from a different fertility clinic (group 4). The tubal status of these 73 women had not always been confirmed. The serological assays were analysed by grouping the women into women with known tubal infertility (n=19) compared to all other causes (including unknown tubal status or confirmed no tubal infertility). A positive Chlamydia antibody test did not significantly correlate with tubal infertility status in this group using any of the tests. Only five women with tubal infertility were positive for C. trachomatis antibodies using MIF, compared to eight women with no or unknown tubal status. The lack of a significant correlation with tubal factor could be because eight Chlamydia sero-positive women who did not have known tubal status were in the unknown comparison group. Of these eight women, three were positive in both MIF CT and Medac Infertile, so it is likely that they had tubal damage and chlamydial infertility. None of the tubal infertile women were positive for C. trachomatis antibodies using the Medac Infertile test. None of the peptide assays developed during this study were positive for any of the women in the tubal pathology group, although some of the women who had either unknown or other causes of infertility were positive (6 of 49 women in that group). The results between Medac Infertile and MIF CT were significantly concordant in this group (Table S3). The IVF outcomes were monitored until closure of the study (several women were still undergoing treatment at the study closure time point), and the outcome of live birth was compared to 919

6 S. Menon and others Table 2. Testing the multi-peptide assays for the detection of tubal factor infertility Adjusted odds ratio Fisher s exact P value (unadjusted) Odds ratio Negative predictive value Positive predictive value Specificity Sensitivity pos pos* Assay Absorbance threshold criteria 1 (0.89 1) 1 (0.36 1) 0.55 ( ) e+07 (0 Inf) Medac Infertile ( ) ( ) 0.69 ( ) 0.56 ( ) 2.8 ( ) 0.89 ( ) MIF.CT ( ) ( ) 0.43 ( ) 0.49 ( ) 0.73 ( ) 0.66 ( ) MIF.CP ( ) (0.5 21) 0.71 ( ) 0.54 ( ) 2.88 ( ) 0.96 ( ) ( ) P or P or P or P47 or P50 1 (0.89 1) 1 (0.42 1) 0.55 ( ) e+07 (0 Inf) ( ) P and P and P and P47 and P50 *Positives were participants with confirmed tubal factor status and seeking IVF treatment (group 1) (n=44). Negatives were participants with confirmed no detected tubal factor and seeking IVF treatment (group 1) (n=47). The Medac assay had an unequivocal and repeat with new sample result repeatedly in two samples; therefore, for this assay alone, there are two results that could not be determined; however, these samples recorded reproducible results in all other assays tested. the serological assay result. There was no significant difference in the likelihood of achieving a live birth during the study time frame based on Medac Infertile and the peptide assay serological status (Table S4). DISCUSSION Chlamydial serology as a test to predict tubal infertility could be valuable during the initial infertility investigation for women. However, it is frequently not used and it is not listed in gynaecological guidelines in several countries. This lack of widespread implementation seems likely to be due to low specificity of the gold standard MIF and due to different perspectives on how the result can be best implemented into clinical protocol (Broeze et al., 2011). This has resulted in attempts to identify new antigens and develop new serological tests that could be used to specifically and sensitively identify chlamydial tubal infertility (Rodgers et al., 2010, 2011; Sanchez-Campillo et al., 1999; Forsbach- Birk et al., 2010; Wills et al., 2009). We have developed a highly specific serological test (a multiple-peptide ELISA) that had the most significant detection of tubal factor infertility (P=0.011; Table 2). However, the low sensitivity of the test (3 of 11 true positives detected) means that a ative result in this test does not exclude the possibility of chlamydial tubal infertility. The test developed here appeared to outperform the commercial ELISA tested [Medac Infertile, when tested by tubal status (Table 2, results)]. However, two samples (one from the tubal factor group and other from the nontubal factor group) were consistently unequivocal by the Medac test that was able to produce a result in all other samples and so were excluded from the table. This difference of one sample in each cohort if they had reacted the same as our new test would have meant our test has equivalent performance to the Medac ELISA. The test has been evaluated and then validated in only two separate clinics, and this is a possible limitation as further clinic sites should be tested. However, these results that are consistent with other international studies are likely generalizable to IVF clinics in developed world settings. A further limitation is that the set of samples used to establish the test threshold included only 11 true-positive samples. We have used the method outlined by Buderer (1996) to conduct a power analysis and confirmed, as with all other published studies conducting chlamydial serology, that it is not feasible to reach the ideal statistical power. In short, to confirm a specificity of 99 % specificity with 1 % precision with a 7 % disease prevalence, we would have required a sample size of n=409. This sample size has not been achieved in most published serology studies, and it is difficult for a preliminary assay development study, although it will ultimately be needed should an assay with higher sensitivity be developed. Nonetheless, the test performed very well in the development cohort for detection of chlamydial infertility (MIF and tubal infertile) (Table 1), tubal infertility in the fertility clinic group (Table 2) and chlamydial infertility in a larger 920 Journal of Medical Microbiology 65

7 ELISA for Chlamydia tubal factor infertility Table 3. Testing the performance of the multi-peptide ELISA against MIF-positive tubal infertile women compared to infertile, sexual health and fertile-ative cohorts Fisher s exact P value (unadjusted) Odds ratio Negative predictive value Positive predictive value Specificity Sensitivity pos Assay Absorbance threshold criteria pos* ( ) 0.97 ( ) 6.53 ( ) 0.89 ( ) Medac Infertile ( ) 0.12 ( ) 1 (0.97 1) 0 <0.001 MIF.CT (0.62 1) 0.66 ( ) ( ) 0.97 ( ) 2.2 ( ) 0.56 ( ) MIF.CP ( ) < ( ) 0.97 ( ) ( ) 0.98 (0.96 1) MIF.CS ( ) ( ) 0.97 ( ) 4.75 ( ) 0.93 ( ) ( ) P or P or P or P47 or P ( ) 0.96 ( ) 4.39 ( ) 0.95 ( ) ( ) P and P and P and P47 and P50 *Positives were defined as group 1 Infertile women with MIF CT positive and confirmed tubal factor status (n=11). Negatives were group 1 infertile not MIF CT positive or not tubal factor (n=80), acute infection (n=112) (group 2) or fertile currently pregnant (n=53) (group 3). evaluation that included fertile and chlamydial infection controls (Table 3). However, none of the chlamydial serological results were significantly associated with tubal infertility in the validation infertility clinic group (group 4). This result was inconsistent with most published studies, where chlamydial serology is typically significantly associated with tubal infertility (Akande et al., 2003; Mehanna et al., 1995; Claman et al., 1997; Land et al., 2003). However, it is possible that this is because the study included women with unknown tubal status in the comparison group, who would often be excluded in studies like this one. An alternative explanation could be that this clinic may represent a distinct or unusual demographic of patients, with some other aetiologies being more commonly responsible for tubal fertility. In spite of the lack of validation in the separate clinic, the data from the first fertility clinic and the high specificity of the test when tested against fertile and sexual health/gp controls are highly promising for this new multi-peptide ELISA to be developed into a specific test for chlamydial infertility in women. One possible method to effectively apply a highly specific test such as this is through a two-step diagnostic pathway that could be routinely implemented during the infertility investigation of women presenting at the fertility clinic setting. First, an MIF-positive result could be used to indicate that further tubal investigation is needed or tubal damage is possibly present that could mean IVF is needed if no other infertility aetiologies have been identified in the individual. Second, a highly specific ELISA, such as the multi-peptide test developed here, could be used to identify women who (if positive) are likely to have chlamydial tubal damage that could best be recommended to progress directly to IVF treatment. We demonstrated equivalent IVF success rates from both fertility clinic cohorts tested here for women who were positive in the serological tests compared to women who were ative. Therefore, we suggest that IVF treatment is a valid recommendation based on chlamydial serological results of women attending for fertility treatment. In summary, a new highly specific Chlamydia antibody test (multi-peptide ELISA) has been developed, that could be used in the fertility clinic setting to identify women with tubal infertility who will likely need IVF to conceive. ACKNOWLEDGEMENTS The authors want to thank and acknowledge all of the participants who kindly gave their time to contribute to the research study. The study nurses and research officers were extremely valuable for recruitment and data collection, especially Caroline Motteram, Deborah de Guingand and Helen Woodhouse. The authors wish to thank the research nurses, clinic nurses and staff that facilitated our study, consented participants and collected data (Monash IVF Clayton and Monash IVF Wesley). The authors also thank the staff and participants at the Sexual Health Clinics especially Dr John Patten, Dr John Dwyer, Dr Joseph Debattista and all the clinical staff of the Brisbane Sexual Health Clinic; Dr Stuart Aitken, Dr Maree O Sullivan, Brenda Henry and all the clinical staff of the Gold Coast Sexual Health; Dianne Farrell and all staff of the S.H.op 101; Dr Kuong Taing, Karen 921

8 S. Menon and others McGill and all the staff of the Clinic 87 Nambour Sexual Health; Dr Darren Russell and Collette Cashman and all the staff of the Cairns Sexual Health Clinic; and the QUT University Health Service Medical Centre staff especially Leonie O Keefe and the staff and medical students at the various antenatal clinics and practices that facilitated recruitment of pregnant participants. The collection and storage of some serological specimens were conducted in conjunction with the Wesley Medical Research Institute Tissue Bank. The study has been funded by the Wesley Medical Medical Research Institute grants awarded to J. A. A. and W. M. H., the Queensland Government National and International Research Alliance Project awarded to P. T. (sub-project leader W. H.), the Monash IVF Group Research and Education Foundation funding awarded to L. R. and W. H. and the QUT Bluebox Proof of Concept Funding. Several of the authors are named inventors on a patent that has been filed in Australia for this assay. The patent is owned by the QUT BLUEBOX and not by any of the authors. REFERENCES Akande, V. A., Hunt, L. P., Cahill, D. J., Caul, E. O., Ford, W. C. & Jenkins, J. M. (2003). Tubal damage in infertile women: prediction using chlamydia serology. Hum Reprod 18, Broeze, K. A., Opmeer, B. C., Coppus, S. F., Van Geloven, N., Alves, M. F., Anestad, G., Bhattacharya, S., Allan, J., Guerra- Infante, M. F. & other authors (2011). Chlamydia antibody testing and diagnosing tubal pathology in subfertile women: an individual patient data meta-analysis. Hum Reprod Update 17, Buderer, N. M. (1996). Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity. Acad Emerg Med 3, Claman, P., Honey, L., Peeling, R. W., Jessamine, P. & Toye, B. (1997). The presence of serum antibody to the chlamydial heat shock protein (CHSP60) as a diagnostic test for tubal factor infertility. Fertil Steril 67, Coppus, S. F., Land, J. A., Opmeer, B. C., Steures, P., Eijkemans, M. J., Hompes, P. G., Bossuyt, P. M., Van der Veen, F., Mol, B. W. & Van der Steeg, J. W. (2011). Chlamydia trachomatis IgG seropositivity is associated with lower natural conception rates in ovulatory subfertile women without visible tubal pathology. Hum Reprod 26, Forsbach-Birk, V., Simnacher, U., Pfrepper, K. I., Soutschek, E., Kiselev, A. O., Lampe, M. F., Meyer, T., Straube, E. & Essig, A. (2010). Identification and evaluation of a combination of chlamydial antigens to support the diagnosis of severe and invasive Chlamydia trachomatis infections. Clin Microbiol Infect 16, Gijsen, A. P., Land, J. A., Goossens, V. J., Leffers, P., Bruggeman, C. A. & Evers, J. L. (2001). Chlamydia pneumoniae and screening for tubal factor subfertility. Hum Reprod 16, Keltz, M. D., Gera, P. S. & Moustakis, M. (2006). Chlamydia serology screening in infertility patients. Fertil Steril 85, Keltz, M. D., Sauerbrun-Cutler, M. T., Durante, M. S., Moshier, E., Stein, D. E. & Gonzales, E. (2013). Positive Chlamydia trachomatis serology result in women seeking care for infertility is a ative prognosticator for intrauterine pregnancy. Sex Transm Dis 40, Land, J. A., Gijsen, A. P., Kessels, A. G., Slobbe, M. E. & Bruggeman, C. A. (2003). Performance of five serological chlamydia antibody tests in subfertile women. Hum Reprod 18, Mehanna, M. T., Rizk, M. A., Eweiss, N. Y., Ramadan, M., Zaki, S. A., Sadek, A., Chow, J. M. & Schachter, J. (1995). Chlamydial serology among patients with tubal factor infertility and ectopic pregnancy in Alexandria, Egypt. Sex Transm Dis 22, Meikle, S. F., Zhang, X., Marine, W. M., Calonge, B. N., Hamman, R. F. & Betz, G. (1994). Chlamydia trachomatis antibody titers and hysterosalpingography in predicting tubal disease in infertility patients. Fertil Steril 62, Menon, S., Timms, P., Allan, J. A., Alexander, K., Rombauts, L., Horner, P., Keltz, M., Hocking, J. & Huston, W. M. (2015). Human and pathogen factors associated with Chlamydia trachomatis-related infertility in women. Clin Microbiol Rev 28, Mueller, B. A., Luz-Jimenez, M., Darling, J. R., Moore, D. E., McKnight, B. & Weiss, N. S. (1991). Risk factors for tubal infertility; influence of history of prior pelvic inflammatory disease. Sex Transm Dis 19, Oakeshott, P., Kerry, S., Aghaizu, A., Atherton, H., Hay, S., Taylor- Robinson, D., Simms, I. & Hay, P. (2010). Randomised controlled trial of screening for Chlamydia trachomatis to prevent pelvic inflammatory disease: the POPI (prevention of pelvic infection) trial. BMJ 340, c1642. Persson, K., Osser, S., Birkelund, S., Christiansen, G. & Brade, H. (1999). Antibodies to Chlamydia trachomatis heat shock proteins in women with tubal factor infertility are associated with prior infection by C. trachomatis but not by C. pneumoniae. Hum Reprod 14, Price, M. J., Ades, A. E., Welton, N. J., Macleod, J., Turner, K., Simms, I. & Horner, P. J. (2012). How much tubal factor infertility is caused by Chlamydia? Estimates based on serological evidence corrected for sensitivity and specificity. Sex Transm Dis 39, Rodgers, A. K., Wang, J., Zhang, Y., Holden, A., Berryhill, B., Budrys, N. M., Schenken, R. S. & Zhong, G. (2010). Association of tubal factor infertility with elevated antibodies to Chlamydia trachomatis caseinolytic protease P. Am J Obstet Gynecol 203, 494 e7 94 e14. Rodgers, A. K., Budrys, N. M., Gong, S., Wang, J., Holden, A., Schenken, R. S. & Zhong, G. (2011). Genome-wide identification of Chlamydia trachomatis antigens associated with tubal factor infertility. Fertil Steril 96, Sanchez-Campillo, M., Bini, L., Comanducci, M., Raggiaschi, R., Marzocchi, B., Pallini, V. & Ratti, G. (1999). Identification of immunoreactive proteins of Chlamydia trachomatis by Western blot analysis of a twodimensional electrophoresis map with patient sera. Electrophoresis 20, Sexually transmitted diseases surveillence (2014). gov/std/stats14/. Accessed 30/11/2015. Stansfield, S. H., Patel, P., Debattista, J., Armitage, C. W., Cunningham, K., Timms, P., Allan, J., Mittal, A. & Huston, W. M. (2013). Proof of concept: a bioinformatic and serological screening method for identifying new peptide antigens for Chlamydia trachomatis related sequelae in women. Results Immunol 3, Swart, P., Mol, B. W., Van der Veen, F., Van Beurden, M., Redekop, W. K. & Bossuyt, P. M. (1995). The accuracy of hysterosalpingography in the diagnosis of tubal pathology: a meta-analysis. Fertil Steril 64, Toye, B., Laferriere, C., Claman, P., Jessamine, P. & Peeling, R. (1993). Association between antibody to the chlamydial heat-shock protein and tubal infertility. J Infect Dis 168, Wills, G. S., Horner, P. J., Reynolds, R., Johnson, A. M., Muir, D. A., Brown, D. W., Winston, A., Broadbent, A. J., Parker, D. & McClure, M. O. (2009). Pgp3 antibody enzyme-linked immunosorbent assay, a sensitive and specific assay for seroepidemiological analysis of Chlamydia trachomatis infection. Clin Vaccine Immunol 16, Journal of Medical Microbiology 65