Abstract
Background/Aim: The aim of this study was to investigate the association between serum IgG antibodies to C. trachomatis major outer membrane protein (MOMP), iron-binding periplasmic protein TroA, and high temperature requirement protein (HtrA), and the occurrence of epithelial ovarian cancer (EOC) and borderline ovarian tumors (BOTs).
Patients and Methods: The study included patients admitted to Oulu University Hospital due to suspicion of ovarian tumor between March 2008 and May 2018. Serum IgG antibody responses to recombinant C. trachomatis TroA, HtrA, and MOMP were analyzed using enzyme-linked immunoassay. Oulu University Hospital medical records were obtained and reviewed. Furthermore, we assessed median carbohydrate antigen 125 levels across different C. trachomatis serology groups in patients with benign, BOTs, and EOCs.
Results: The final study cohort consisted of 210 patients diagnosed with EOC, 46 with BOTs, and 175 with benign ovarian tumors. Among the 210 EOC cases, 40 occurred in C. trachomatis-seropositive patients. Of the 46 BOTs, 13 were found in the C. trachomatis-seropositive group. Logistic regression analysis revealed that C. trachomatis TroA IgG seropositivity was not associated with an increased risk of EOC (odds ratio=1.37, 95% confidence interval=0.72-2.70) but was linked to a significantly higher risk of BOTs (odds ratio=2.51, 95% confidence interval=1.02-5.98). In patients with BOTs, the median carbohydrate antigen 125 level was higher in the C. trachomatis-seropositive group than in the seronegative one (173 vs. 55 kU/l).
Conclusion: Our results suggest that the presence of serum IgG antibodies against C. trachomatis TroA is associated with an increased risk of BOT. No association was observed between positivity for any chlamydial immunological marker and EOC risk.
Introduction
Ovarian cancer ranks eighth worldwide in both incidence and mortality among women, with approximately 325,000 new cases and 207,000 deaths reported in 2022 (1). The most common subtype, epithelial ovarian cancer (EOC), accounts for over 90% of all ovarian cancer cases and includes five major histotypes: high- and low-grade serous, endometroid, clear-cell, and mucinous cancer (2, 3). The fallopian tubes have been identified as the site of origin for the most prevalent type, namely high-grade serous cancer (2-4), and emerging evidence suggests that serous borderline ovarian tumors (BOTs) also originate from this location (3, 5). BOTs are epithelial neoplasms that exhibit biological behavior intermediate between benign and malignant tumors (2, 6).
Chlamydia trachomatis is the most common bacterial cause of sexually transmitted infections (4), and a well-established contributor to pelvic inflammatory disease (PID) (4, 7). Several studies have suggested an association between PID and the development of BOTs (5, 8-11), especially serous BOTs, (5, 8, 9) and ovarian cancer more broadly (10-12). In addition, associations between C. trachomatis infection and an increased risk of both EOC (4, 13-16) and BOT (4, 14) have been reported. Chronic pelvic inflammation, such as that induced by persistent chlamydial infection, has also been implicated in the pathogenesis of EOC (17). C. trachomatis can cause subclinical and persistent infections (4, 18) and the resulting chronic inflammation may contribute to the observed links between chlamydial PID and EOC or BOTs (4).
Following C. trachomatis infection, serum immunoglobulin (IgG) antibodies against the chlamydial major outer membrane protein (MOMP) appear (19). IgG antibodies against C. trachomatis proteins such as iron-binding periplasmic protein TroA, and high temperature requirement protein (HtrA) have also been suggested as biomarkers of ascending (20) or persistent C. trachomatis infection leading to tubal scarring (20-22). TroA and HtrA are proteins expressed particularly during persistent chlamydial infection (23, 24). In vitro studies have indicated their upregulation under conditions that favor chlamydial persistence (25). IgG antibody responses to TroA and HtrA are more common in individuals with ascending or repeated chlamydial infections compared to healthy controls (20). Furthermore, patients with tubal factor infertility are more likely to have serum IgG antibodies against TroA and HtrA than subfertile patients with patent fallopian tubes (21). Previously, we found that patients with EOC who had IgG antibodies against TroA and HtrA demonstrated a better response to the first-line cancer treatment and had improved 3-year survival rates (26).
The present retrospective observational study aimed to investigate the association between serum IgG antibodies to C. trachomatis MOMP, TroA, and HtrA and the occurrence of EOC and BOT. Furthermore, we analyzed carbohydrate antigen 125 (CA125) levels in relation to C. trachomatis serology and determined optimal CA125 cut-off values to differentiate between benign, BOT, and EOC.
Patients and Methods
The study group included 550 patients admitted to the Oulu University Hospital due to suspicion of ovarian tumors between March 2008 and May 2018. Patients with other gynecological malignancies (n=39), cancers other than EOC (n=38), previous EOC (n=4), and unclear or unknown pathological diagnosis (n=9) were excluded. Also, duplicate records (n=13) and patients with invalid identity numbers (n=16) were excluded (Figure 1).
Flowchart of the study cohort.
Cancer data were collected from the medical records of Oulu University Hospital, including histology and stage of cancer, as well as CA125 values. Information on apparent confounders, such as age, gynecological history, and prior malignancies or premalignancies, was also collected from medical records. The patients provided informed consent, and serum samples for research purposes were collected before any cancer treatments. The serum samples were stored in the hospital’s laboratory at −20°C and transferred monthly to the research laboratory, where they were stored at −80°C until the required analysis.
IgG antibody responses to recombinant C. trachomatis TroA and HtrA were analyzed using an in-house enzyme-linked immunosorbent assay, which has been previously described in detail (20-22). The intra-assay coefficient of variation of TroA was 4.6% and of HtrA, 4.3%. The inter-assay coefficient of variation of the TroA assay was 12.3%-12.9% and of the HtrA assay, 7.6%-13.8%. C. trachomatis MOMP IgG antibodies were assayed with a commercial C. trachomatis IgG in-house enzyme-linked immunosorbent assay kit (Labsystems Diagnostics, Vantaa, Finland). Based on the manufacturer’s instructions, samples with signal/cutoff value <1 were interpreted as negative, ≥1.4 as positive, and 1.0≤ signal/cutoff value <1.4 as equivocal (27).
We analyzed the groups with serum IgG antibodies to MOMP, TroA, or HtrA separately. Additionally, we examined groups positive for all three markers and those positive for anyone.
The logistic regression was used to analyze the risk of EOC and BOT. Age and parity as confounding factors were applied to the model together with chlamydial TroA IgG, HtrA IgG or MOMP IgG data. Data were analyzed using R version 4.3.2. Logistic regression results were calculated using the function glm of the package stats. The package eulerr and its function euler were used to draw a Euler diagram. The package cutpointr and its function cutpointr were used to define the optimal cut-off points for CA125. The method used was the maximization of the sum of sensitivity and specificity.
The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Helsinki University Central Hospital, Laboratory Division (permit number: HUS/211/2020, date April 29, 2020). Serum sample collection and data collection from the hospital’s records were approved by the Regional Ethics Committee of the Northern Ostrobothnia Hospital District (53/2010, date September 15, 2010) and the National Supervisory Authority for Welfare and Health (1339/05.01.00.06/2009, date March 12, 2009). Informed consent was obtained from the research participants.
Results
The final study cohort included 210 patients diagnosed with EOC, 46 with BOT, and 175 with benign ovarian tumor (Figure 1). C. trachomatis serology was defined as positive when testing for any of the three markers (MOMP, TroA, or HtrA) was positive. Figure 2 illustrates the overlap between MOMP, TroA, and HtrA seropositivity. There were 40 EOC cases in C. trachomatis-seropositive patients and 170 cases in seronegative patients (Table I). Most EOCs were histologically high-grade serous in both the C. trachomatis-seropositive (63%) and C. trachomatis-seronegative (64%) groups. Furthermore, the distribution of other histological subtypes was similar in both groups (data not shown).
Euler diagram of distributions of IgG positivity for chlamydial iron-binding periplasmic protein TroA, high temperature requirement protein (HtrA), and major outer membrane protein (MOMP).
Baseline characteristics of the study population by ovarian tumor histology and Chlamydia trachomatis markers [serum immunoglobulin G (IgG) antibodies to chlamydial major outer membrane protein (MOMP), iron-binding periplasmic protein TroA, and high temperature requirement protein (HtrA)].
Altogether, there were 46 BOTs, of which 13 were in the C. trachomatis-seropositive group and 33 in the seronegative group (Table I). Most BOTs were the mucinous subtype in C. trachomatis-seronegative patients (n=21), but in seropositive patients, the most common subtype was serous BOT (n=8).
According to the logistic regression analysis, there was no difference in risk of EOC in the C. trachomatis TroA-seropositive group [odds ratio (OR)=1.37, 95% confidence interval (CI)=0.72-2.70] or in any other positive C. trachomatis immunological marker groups (Table II). Nor was there an altered risk of high-grade serous ovarian tumor type in groups with TroA IgG (OR=1.28, 95% CI=0.59-2.74) or in any group positive for other C. trachomatis immunological markers (data not shown).
Results of logistic regressions for epithelial ovarian cancer and borderline ovarian tumor risk. Age and parity, as confounding factors, were applied to the model, together with data on serum immunoglobulin G (IgG) antibodies to chlamydial iron-binding periplasmic protein TroA, and high temperature requirement protein (HtrA) and major outer membrane protein (MOMP).
Age increased the risk of EOC with C. trachomatis TroA seropositivity (OR=1.05, 95% CI=1.03-1.07), as well as the risk for EOC with HtrA or MOMP seropositivity (data not shown). Parity was not associated with the risk of EOC connected with C. trachomatis infection.
TroA seropositivity was associated with a higher risk for BOT (OR=2.51, 95% CI=1.02-5.98; Table II), but there was no association between BOT risk and HtrA or MOMP seropositivity. Age and parity were not associated with the risk of BOT.
The median values for CA125 tumor marker were 361 (range=8-30,016) kU/L for the EOC group, 57.5 (range=5-526) kU/l for the BOT group, and 31 (range=1-1,874) kU/l for the benign group (Table I). The median CA125 values did not substantially differ in subgroups defined by C. trachomatis seropositivity in the EOC and benign groups. However, in patients with BOT, the median CA125 was 173 kU/L in seropositive patients and 55 kU/l in seronegative patients (p=0.107). Regardless of C. trachomatis serology, the optimal cut-off for CA125 when comparing benign and malignant tumors was 122 kU/l, with a sensitivity of 0.73, a specificity of 0.88, and accuracy of 0.80.
Discussion
Our results suggest that the presence of serum C. trachomatis TroA IgG antibodies is associated with an increased risk of BOT. However, no association was observed between positivity for any chlamydial immunological marker and EOC risk. Notably, among patients with BOTs, the median CA125 level was higher in the C. trachomatis-seropositive group than in the seronegative group (median of 173 vs. 55 kU/l).
Previous studies have suggested a connection between C. trachomatis infection and EOC (13-16), but the association between C. trachomatis immunological markers and BOTs has been less extensively investigated. Previous studies (13-15) have defined C. trachomatis infection using serological markers in samples collected either before or after an ovarian cancer diagnosis. The markers used include antibodies against a highly immunogenic, plasmid-encoded protein 3 of C. trachomatis (Pgp3) (13-15) and MOMP (13, 15) proteins but not TroA or HtrA IgG antibodies. A large case–control study by Fortner et al. (14) reported that chlamydial seropositivity was associated with approximately twofold increased risk of invasive cancer, invasive serous EOC, and BOT. In our previous study, seropositivity to chlamydial TroA and HtrA among patients with EOC was associated with a better response to the first-line treatment and improved 3-year survival rates (26).
It has been suggested that both EOC and serous BOTs originate from the fallopian tube epithelium (3). Previous studies have also identified an association between PID and ovarian cancer (10, 11) and BOTs (5, 8-11), especially the serous subtype (5, 8, 9). Microbes, including C. trachomatis, can ascend from the cervix through the uterus to the fallopian tubes, leading to PID. Such infection can damage the tubal epithelium and lead to scarring, adhesions, and obstruction of the fallopian tubes (7). Seropositivity to chlamydial TroA and HtrA has been associated with tubal factor infertility (21, 22). Infections and chronic inflammation play a recognized role in carcinogenesis, and at least seven microbes, including human papillomavirus and hepatitis B virus, have been linked to cancer (4, 28). Elevated C-reactive protein levels have been associated with increased EOC risk, while the use of aspirin and nonsteroidal anti-inflammatory drugs have been indicated to lower this risk (4). Chronic inflammation and PID have been proposed as plausible mechanisms by which C. trachomatis may contribute to ovarian cancer (4). Moreover, C. trachomatis infection-induced DNA and cellular damage, along with the promotion of epithelial–mesenchymal transition in host cells, has been suggested as a possible pathogenic mechanism (4, 28).
Tumor marker CA125 is used to monitor ovarian cancer and has prognostic significance (29). Serum CA125 levels above 35 U/mL are often detected in patients with EOC, but the test is limited by high false-positive and -negative rates, making it unsuitable as a standalone diagnostic tool (30). The upper limit of the normal range is often considered 35 kU/l (30). We determined median CA125 values for patients with benign tumor, BOT, and EOC in relation to C. trachomatis serology. We also identified optimal CA125 cut-off values to distinguish between benign and malignant tumors, as well as between benign and malignant and BOT combined. The CA125 values varied widely within each group, and the optimal cut-off values were clearly higher than the conventional level of 35 U/ml. The optimal threshold was 122 kU/l for differentiating benign from malignant tumors, and 147 kU/l for differentiating benign from BOT/malignant tumors. New methods are being developed to improve the accuracy of CA125 testing in cancer screening, early diagnosis, and treatment planning (31). In our study, the median CA125 level in the BOT group was higher in C. trachomatis-seropositive patients than in seronegative patients. However, this difference did not reach statistical significance, possibly due to the small sample size. Given that TroA is expressed particularly during persistent C. trachomatis infection, infection might partially explain the observed elevated level. Previous studies have demonstrated that acute PID caused by C. trachomatis infection is correlated with both elevated CA125 levels and more severe PID compared to PID caused by other pathogens (32).
The major strength of this study was its single-center design, allowing for systematic data collection from a consistent source. Moreover, we utilized validated assays to determine C. trachomatis infection-related antibodies. The study limitations include the relatively small sample size, which may have limited the statistical power of our analysis. The limited number of healthy participants in the study was due to the recruitment strategy focused on a specific hospital setting, where healthy individuals were less likely to be present. In addition, we did not perform nucleic acid detection testing for C. trachomatis from urogenital samples; thus, acute chlamydial infections were not evaluated. In addition, potential confounding factors, such as genetic susceptibility, socioeconomic status, and lifestyle factors (e.g., diet and exercise), were not measured, although they can influence both infection risk and ovarian cancer development.
Conclusion
Our results suggest that the presence of serum C. trachomatis TroA IgG antibodies is associated with an increased risk of BOT. However, no significant association was observed between positivity for any chlamydial immunological marker and EOC risk. The results do indicate a potential infectious etiology underlying BOT development, although this remains uncertain.
Acknowledgements
The Authors thank Anu Kaitonen for technical assistance.
Footnotes
Authors’ Contributions
All the Authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship and contributed to the intellectual content of the study. HT, HH, TH, UP, MP and EU conceptualized and designed the study. HT and EU collected the data. MP supervised the laboratory work. HH managed the data set and performed the statistical analyses. All the Authors contributed to the interpretation of the results, as well as to the writing and editing of the manuscript. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests.
Funding
This study was supported by the Helsinki University Hospital Research Funding for Mirja Puolakkainen TYH2023329 and Y780024100), Oulu Duodecim Society, Thelma Mäkikyrö Foundation and Oulu University Hospital Research Funding (OYS EVO K77729) for Elina Urpilainen, the Finnish Medical Foundation for Tiina Holster and Jane and Aatos Erkko Foundation for Ulla Puistola.
Artificial Intelligence (AI) Disclosure
No artificial intelligence (AI) tools, including large language models or machine-learning software, were used in the preparation, analysis, or presentation of this manuscript.
- Received November 26, 2025.
- Revision received April 13, 2026.
- Accepted April 17, 2026.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.