CA-125 glycovariant assays enhance diagnostic sensitivity in the detection of epithelial ovarian cancer
-
Stefanos Moukas
,
Katri Kuningas
Abstract
Objectives
Ovarian cancer is the deadliest gynaecologic malignancy. Due to the lack of reliable biomarkers for the detection of the early disease, most patients are diagnosed at an advanced stage resulting in poor survival. We therefore aimed at establishing novel CA-125 glycovariant assays to improve the diagnostic sensitivity and specificity of ovarian cancer.
Methods
Blood samples of 184 patients with epithelial ovarian cancers (EOC), 127 benign ovarian tumors, and 115 healthy controls were measured using GLYVAR™ Ovarian I and II assays (Uniogen) and the conventional CA-125 protein assay (CanAg CA-125 EIA, Fujirebio).
Results
The two glycovariant assays differentiated benign and malignant ovarian masses with 88.0 % sensitivity at 99 % specificity, whereas CA-125 showed 72.8 % sensitivity. The improved performance was most evident in patients with borderline or moderately elevated CA-125 concentration at diagnosis, which is a challenging group for differential diagnostics. The CA-125 glycovariant assays showed 2.5 times higher sensitivity (33.3 % with CA-125 vs. 83.3 % with the CA-125 glycovariants) at 94 % specificity. CA-125 glycovariants corrected 82.4 % of false positive results given by CA-125 concentrations with the commonly used cutoff 35 U/mL. Importantly, the CA-125 glycovariant assays detected 63.6 % of early-stage serous carcinomas from benign and healthy controls with very high 99 % specificity, while CA-125 had a sensitivity of only 45.5 %, representing a 40 % increase.
Conclusions
This is the first study describing the clinical performance of GLYVAR Ovarian I and II assays in ovarian cancer diagnostics. The results indicate that the CA-125 glycovariant assays have remarkable potential to improve ovarian cancer diagnostics.
Introduction
With 69,472 new diagnoses and 46,232 deaths in 2022, ovarian cancer is the deadliest gynecologic malignancy in Europe [1]. It is responsible for 3.3 % of all malignant neoplasms and 5.2 % of all cancer deaths in Europe [2]. As characteristic early symptoms are missing, most patients are diagnosed at an advanced stage III–IV. In these cases, standard therapy consists of radical cytoreductive surgery and platinum-based chemotherapy. In recent years, the use of PARP inhibitors in BRCA-mutated and homologous repair deficient (HRD) tumors has improved the outcome of these patient groups significantly [3], 4]. However, despite optimal surgical and systemic therapy, relative 5-year survival remains around 40 % in European countries [5] and drops as low as ∼20 % in the FIGO stage IV [6], [7], [8], [9]. In stage I disease, relative 5-year survival is significantly better, around 90 %.
Thus, early detection of ovarian masses is an urgent clinical challenge, the resolution of which is the key to achieving satisfactory oncological outcomes. Moreover, benign ovarian lesions are a common finding, especially in premenopausal women. Since differentiation between malignant and benign masses is insufficient with the existing diagnostic tools including ultrasound and tumor markers, surgery is the only way to determine the dignity of an ovarian tumor. Thus, in order to avoid overtreatment and facilitate treatment planning, the establishment of reliable tools for the early detection of ovarian malignancies and the differentiation of benign and malignant ovarian masses represent an unmet medical need.
In this context, cancer-specific glycosylation can serve as a distinctive feature of cancer cells offering valuable opportunities for diagnostics and therapeutic targets [10]. During the cancer transformation process, the expression and function of glycosylation modification enzymes change, resulting in alterations in the glycosylation of their target proteins. Ovarian cancer is associated with several glycosylation changes in the MUC16/CA-125 protein including increased sialylation and core fucosylation of N-glycans, as well as truncated O-glycans and altered glycan branching [11], [12], [13], [14], [15], [16], [17]. Truncated O-glycans such as T, Tn and STn antigens are almost absent in benign ovarian tumors and normal ovarian tissues but are clearly expressed in most ovarian carcinomas and the serum of ovarian cancer patients [18].
To date, measuring glycostructures with conventional laboratory assays has been challenging due to the low affinity of binders such as lectins or anticarbohydrate antibodies [19]. Recently developed technology employed nanoparticle tracers to enhance binder affinity through the avidity effect, achieved by the high density of immobilized binders on the particles [20], [21], [22]. This technology also amplified the signal with tens of thousands of Eu3+-labels on each particle. Further, time-resolved fluorescence measurement was used to reduce background signals, further improving the signal-to-background ratios obtained with the assays.
The most widely used biomarker in therapy monitoring of ovarian malignancies is Cancer-Antigen (CA)-125. However, its use in the detection of ovarian malignancies is limited. So far, attempts at establishing screening methods using the combination of transvaginal ultrasound and tumor markers such as CA-125 have not been able to demonstrate benefits in terms of mortality reduction [23]. We, therefore, aimed at improving the performance of the biomarker CA-125 in the detection of epithelial ovarian cancer (EOC) and the discrimination of benign and malignant ovarian tumors. We assessed the performance of CA-125 glycovariant assays, GLYVARTM Ovarian I and II, which detect highly cancer-specific glycostructures on CA-125 protein and utilize the abovementioned nanoparticle traces, in a cohort of Finnish and German patients with benign and malignant ovarian masses.
Materials and methods
Study cohort
The study cohort included patients diagnosed at the University Hospital in Essen, Germany, and from three regional biobanks in Finland: Biobank of Eastern Finland, Finnish Clinical Biobank Tampere, and Auria Biobank (Table 1).
Description of the study cohort. Number of participants.
| Sample group | FinBB | Essen | Total | % |
|---|---|---|---|---|
| Healthy controls | 115 | 0 | 115 | 27.0 %a |
| Benign | 127 | 0 | 127 | 29.8 %a |
| Endometriosis, N80 | 76 | 0 | 76 | 59.8 %b |
| Neoplasms, D27 | 51 | 0 | 51 | 40.2 %b |
| EOC | 47 | 137 | 184 | 43.2 %a |
| HGSC, G3 | 31 | 95 | 126 | 68.5 %c |
| LGSC (G1-G2) | 4 | 42 | 46 | 25.0 %c |
| Endometrioid | 9 | 0 | 9 | 4.9 %c |
| Clear cell | 3 | 0 | 3 | 1.6 %c |
| Total | 289 | 137 | 426 | |
| Figo stage (EOC) of included | ||||
| I | 10 | 7 | 17 | 9.2 % |
| II | 3 | 2 | 5 | 2.7 % |
| III | 22 | 90 | 112 | 60.9 % |
| IV | 12 | 38 | 50 | 27.2 % |
| Age <50 years | ||||
| Controls | 33 | 0 | 33 | |
| Benign | 89 | 0 | 89 | |
| EOC | 10 | 21 | 31 | |
| Age ≥50 years | ||||
| Controls | 82 | 0 | 82 | |
| Benign | 38 | 0 | 38 | |
| EOC | 37 | 116 | 153 |
-
a% of all samples; b% of benign controls; c% of EOC. HGSC, high grade serous carcinoma; LGSC, low grade serous carcinoma.
The Essen cohort included serum and plasma samples from women with high or low grade serous ovarian carcinomas. Patients were diagnosed between 2003 and 2021. Clinical data were acquired from patient charts and the hospital’s clinical information system. The study was approved by the local Ethics Committee of the University of Duisburg-Essen (Essen 05–2870 and 17–7859). In total, 171 pairs of parallel serum and plasma samples were collected at different times of treatment course to evaluate whether the sample matrix affects the resulting biomarker concentrations. To evaluate the assays’ ability to diagnose EOC, only serum samples from untreated patients (n=137) were combined with the Finnish biobank cohort.
For the Finnish biobank cohort, serum samples with the ICD-10 diagnosis codes C56 for ovarian cancer, D27 for benign ovarian neoplasms, or N80 for endometriosis were included. The patients were diagnosed between 2013 and 2023. In the case of control samples, it was ensured that the patients had not had any of the above-mentioned diagnoses. Altogether, 39 women were excluded due to unspecified, non-epithelial, or mucinous type of ovarian cancer, borderline tumors, or unknown treatment status (Supplementary Table S1). The study subjects have given informed biobank consent for use of their specimens. The use of biobank samples in the study were evaluated by the scientific steering groups of Auria Biobank (Decision AB19-8768), Finnish Clinical Biobank Tampere (Decision BB2020-0104) and Biobank of Eastern Finland (Decision BB2020-0104) as well as the board of Finnish Biobank Cooperative – FinBB according to the Biobank Act (688/2012) of Finland. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and complied with relevant national regulations concerning the use of retrospective human specimens.
CA-125 glycovariant assay and conventional CA-125
Blood samples were measured using GLYVAR™ Ovarian I and II assays (Uniogen, Turku, Finland) and the conventional CA-125 protein assay (CanAg CA-125 EIA, Fujirebio Diagnostics, Mölndal, Sweden) according to the manufacturer’s instructions. The GLYVAR Ovarian I and II assays were measured in triplicates and CA-125 EIA as duplicates.
Both GLYVAR Ovarian I and II tests followed the same assay principle (Figure 1). First, biotinylated CA-125-specific capture antibody was added to the wells of a streptavidin-coated microtiter plate. After incubation and washing, calibrators, controls or samples were added, and the CA-125 protein present were bound by the immobilized capture antibody. After removal of unbound substances, nanoparticle conjugates specific to CA-125 glycovariants were added. After incubation, unbound conjugate was removed, and time-resolved fluorescence signal was measured at 615 nm from the well surface. All steps were performed at room temperature. The CA-125 glycovariant concentrations were calculated using the calibration curve fitted with “5 PL forced to origo weighted 1/x” regression in the My Assays software.
GLYVAR™ ovarian assay principle. While the conventional CA-125 immunoassay (EIA/ECLIA) detects the CA-125 protein core, the GLYVAR ovarian assay binds the CA-125 protein but detects cancer-specific glycovariants on the surface of the CA-125 protein core. GLYVAR™ ovarian I and II have a different binder coated on the nanoparticles and they detect separate glycovariants.
Statistical analysis
Initially, sample size calculation was conducted to ensure sufficient statistical power. Linear regression between paired serum and plasma samples was performed to evaluate whether both sample matrices can be used for the GLYVAR Ovarian I and II assays. The regression curves were visually inspected in various concentration scales to identify aberrations. The combination value of the GLYVAR Ovarian I and II assays with and without CA-125 were calculated using binary logistic regression for log2-transformed data.
Receiver operating characteristic (ROC) analysis with the area under curve (AUC) values and sensitivity at selected specificity were generated using a nonparametric method. Statistical analyses were not performed to compare AUCs or sensitivity values. Group comparisons were performed on log2-transformed data using one-way ANOVA with a post-hoc Tukey test. Violin plots of the biomarker concentrations were generated using DATAtab: Online Statistics Calculator (DATAtab e.U. Graz, Austria. URL https://datatab.net). A multivariate F-test was used to evaluate the effect of age groups (<50 years and ≥50 years) on the classification of diagnostic groups. p-value <0.05 was considered statistically significant in all analyses. Statistical analyses were carried out using IBM SPSS software (IBM Corp. Released 2016. IBM SPSS Statistics for Macintosh, Version 29.0. Armonk, NY: IBM Corp.).
Results
Cohort characteristics
The full combined population included 465 patients of which 39 were excluded (Supplemental Table S1). Thus, the study population consisted of 426 participants, of which 184 were diagnosed with EOC, 127 had benign ovarian neoplasms or endometriosis, and 115 were healthy controls (Table 1). The mean patient age was 55 years (range 19–91 years) for the full population, 57.7 years for unaffected, 49.4 years for benign neoplasms, 34.2 years for endometriosis and 63.1 years for EOC. EOC patients were significantly older (p<0.01) than women with benign conditions, as expected.
Applicability of CA-125 glycovariant assays on serum and plasma samples
The concentrations of CA-125 glycovariants were measured in 171 pairs of parallel serum and EDTA-plasma samples collected at different times of treatment course to evaluate the effect of sample matrix on the test results. Linear regression analysis between the two sample matrices demonstrated high correlation: adjusted R2 and slope values were 0.997 (p=2.9 × 10213) and 1.04 for GLYVAR Ovarian I, and 0.999 (p=2.9 × 10246) and 0.97 for GLYVAR Ovarian II (Supplemental Figure S2). Similarly, the CV% of the triplicate measures were alike in serum and plasma (Supplemental Table S3). These results indicate that both sample matrices are suitable for the glycovariant assay measurements. However, only serum samples were used in further analysis due to standardization and better availability.
CA-125 glycovariant and CA-125 protein concentrations in epithelial ovarian cancer and benign and healthy controls
The median concentrations of CA-125 glycovariants detected with GLYVAR Ovarian I and II assays and the conventional CA-125 in EOC, benign neoplasms, endometriosis and healthy controls are shown in Figure 2. The median concentrations of all biomarkers were highly elevated in EOC as compared to healthy controls (p<10−12). While CA-125 was clearly elevated in both benign neoplasms and endometriosis, the glycovariant markers were differently elevated in the two benign groups, with GLYVAR Ovarian I being elevated specifically in endometriosis and GLYVAR Ovarian II more in the benign neoplasms.
Median concentrations (25th and 75th quartiles) and violin plots of CA-125, GLYVAR ovarian I and II and their combinations in different diagnostic groups.
The largest EOC subtypes of the cohort were high-grade serous carcinomas (HGSC; n=126) and LGSC (n=46). We did not observe significant differences in the median concentration of CA-125 (499 U/mL and 450 U/mL; p=0.491) or GLYVAR Ovarian I (108 U/mL and 177 U/mL; p=0.356) between HGSC and LGSC. Instead, the concentration of GLYVAR Ovarian II, the combination of GLYVAR Ovarian I and II, and the combination of the two glycovariant tests with CA-125 were significantly higher in HGSC than in LGSC. The median concentrations of GLYVAR Ovarian II in HGSC and LGSC were 42 U/mL and 22 U/mL (p=0.022), respectively. The corresponding median values of the combination of GLYVAR Ovarian I and II were 0.748 and 0.714 (p=0.029), and the median values of the combination of all three biomarkers were 0.733 and 0.696 (p=0.002) in HGSC and LGSC.
Performance of CA-125 glycovariant assays in the differentiation of malignant and benign tumors
In this cohort, 85.3 % of patients had late-stage serous carcinomas, which typically show high concentrations of conventional CA-125. Single GLYVAR Ovarian I, GLYVAR Ovarian II, and CA-125 assays showed similar AUC values (0.965–0.966) (Figure 3), while the combination of GLYVAR Ovarian I and II performed somewhat better with the AUC of 0.975. Noteworthy, the glycovariant assays showed improved performance especially at very high 99 % specificity, where CA-125 showed the sensitivity of 72.8 % and single GLYVAR Ovarian I and II assays and their combinations had sensitivity of 80–90 %. The combination of GLYVAR Ovarian I and II with CA-125 reached 95 % sensitivity at 95 % specificity in the classification of EOC from benign and healthy controls.
ROC-curves, AUC values and sensitivity (SN) at high specificity (SP) in the classification of clinically relevant patient groups.
Patients with borderline or moderately elevated CA-125 concentrations
The greatest differences between CA-125 and the CA-125 glycovariant assays were observed in a clinically challenging group for differential diagnostics i.e., within patients who have borderline or moderately elevated CA-125 values (30–300 U/mL). The combination of GLYVAR Ovarian I and II resulted in an improved AUC value of 0.928 in comparison to 0.793 reached with the conventional CA-125 assay (Figure 3). Remarkably, the sensitivity at 94 % specificity was improved by 2.5 times from 33.3 % of CA-125 to 83.3 % with the combination of GLYVAR Ovarian I and II. Unfortunately, a reliable analysis of sensitivity at a higher specificity was not possible due to the small sample number.
Detection of early stage EOC
The conventional CA-125 assay detected 45.5 % of all early stage (FIGO stage I-II) ovarian carcinomas at high 99 % specificity, while the detection rates of GLYVAR Ovarian I and GLYVAR Ovarian II assays were 54.4 and 40.9 %, respectively (Figure 3). The performance of GLYVAR Ovarian I further improved to 63.6 % when the analysis was limited to early-stage serous carcinomas. In contrast, the detection rate of CA-125 remained at 45.5 % in this sample group. Thus, the improvement of the CA-125 glycovariant assays was 20 % in all early-stage carcinomas and 40 % in early-stage serous carcinomas in comparison to the conventional CA-125 assay. For the most lethal EOC subtype, high-grade serous carcinomas (HGSC), the cancer detection rate was even higher; 71.4 % of HGSC was detected with GLYVAR Ovarian I with an improvement of 66.4 % compared to 42.9 % detected with CA-125. However, with only seven women with early-stage HGSC included in the cohort, the power for the comparison remained at 0.772 for CA-125. In comparison, the power reached 0.999 for GLYVAR Ovarian I.
Assay performance in pre- and postmenopausal women
The AUC values of CA-125 and GLYVAR Ovarian assays did not show significant difference between the age groups although the difference for CA-125 was close to be statistically significant with a p-value of 0.058 (Table 2). Within women of ≥50 years of age, including the majority of the EOC patients, the glycovariant assays performed clearly better than CA-125. The AUC improved from 0.957 of CA-125 to 0.983 of GLYVAR Ovarian I, and the sensitivity at 98 % specificity improved from 71.2 to 93.5 %, correspondingly. Within women <50 years of age GLYVAR Ovarian and the biomarker combinations had similar AUC values (0.988 of CA-125, 0.978–0.991 of the glycovariant assays) but showed higher sensitivity at 98 % specificity (87.1–90.3 %, except 77.4 % for GLYVAR Ovarian II) as compared with CA-125 (80.6 %).
Performance of CA-125 and GLYVAR Ovarian in women <50 years (31 EOC and 89 benign controls) or ≥50 years of age (153 EOC and 38 benign controls).
| Assay | <50 years | ≥50 years | p-Value for AUC in age groups | ||||
|---|---|---|---|---|---|---|---|
| AUC | 95 % CI | SN at 98 % SP | AUC | 95 % CI | SN at 98 % SP | ||
| CA-125 | 0.988 | 0.973–1.002 | 80.6 % | 0.957 | 0.928–0.985 | 71.2 % | 0.058 |
| Glyvar ovarian I | 0.985 | 0.968–1.002 | 87.1 % | 0.983 | 0.968–0.997 | 91.5 % | 0.851 |
| Glyvar ovarian II | 0.978 | 0.953–1.003 | 77.4 % | 0.960 | 0.934–0.986 | 85.6 % | 0.321 |
| GLYVAR ovarian I+II | 0.989 | 0.975–1.003 | 87.1 % | 0.982 | 0.967–0.997 | 93.5 % | 0.514 |
| GLYVAR ovarian I+II + CA-125 | 0.991 | 0.980–1.002 | 90.3 % | 0.982 | 0.968–0.997 | 92.8 % | 0.335 |
When the diagnostic groups and age groups were included in a multivariate test, no difference was observed in the mean biomarker concentrations between the age groups of women within benign and malignant tumors except for GLYVAR Ovarian I, where the mean concentration was significantly lower in older women with benign tumors than in younger women with benign tumors (p=0.021). No difference was observed in GLYVAR Ovarian I concentrations between the age groups in the malignant tumors. Noteworthy, the patient groups were unbalanced with a low number of young patients with EOC and old patients in the control group, which causes uncertainty of the results.
Detecting EOC with CA-125 glycovariant concentrations
Finally, the ability of CA-125 glycovariants to detect occurrence of EOC was assessed. To this end, optimal cutoff concentrations for this cohort were determined to differentiate women with malignant ovarian tumors from those with benign ones, using Kolmokorov-Smirnov test. For CA-125, both the clinical cutoff and the optimal cutoff were used to appropriately compare the performances of the biomarkers. Positive likelihood ratio, i.e., the probability that a positive test result is correct, was 12.6 for CA-125 with an established clinical cutoff 35 U/mL and 20.4 with the optimal cutoff of 48 U/mL (Table 3). GLYVAR Ovarian II had similar positive likelihood ratio than CA-125, while GLYVAR Ovarian I and the biomarker combinations showed remarkably elevated positive likelihood ratios varying between 31.8 and 111.8. The negative likelihood ratios were comparable for all biomarkers studied.
Performance characteristics obtained using the optimal cutoff values calculated for the cohort. For CA-125, also an established cutoff 35 U/mL was used.
| Performance parameter | CA-125 | GLYVAR ovarian I | GLYVAR ovarian II | GLYVAR ovarian I+II | GLYVAR ovarian I+II + CA-125 | |
|---|---|---|---|---|---|---|
| Cutoff | 35 U/mL | 48 U/mL | 3.6 U/mL | 2.3 U/mL | 0.62 | 0.72 |
| True negative ratio (specificity) | 0.926 | 0.955 | 0.971 | 0.959 | 0.988 | 0.992 |
| True positive ratio (sensitivity) | 0.940 | 0.929 | 0.918 | 0.935 | 0.929 | 0.924 |
| Positive likelihood ratio | 12.6 | 20.4 | 31.8 | 22.6 | 75.0 | 111.8 |
| Negative likelihood ratio | 0.065 | 0.074 | 0.084 | 0.068 | 0.072 | 0.077 |
Discussion
Ovarian cancer diagnosis and the differentiation of malignant and benign ovarian masses is typically performed using physical examination, transvaginal ultrasound and the detection of an established biomarker CA-125. However, while CA-125 is known to perform well in late-stage EOC, it lacks sensitivity to detect early stage EOC and causes false positive results in various benign conditions [24]. It is meant to aid in diagnosis of ovarian cancer, not as a diagnostic biomarker alone. Therefore, there is a pressing need for more accurate biomarkers for the identification of ovarian cancer. Recently, liquid biopsy methods have become increasingly commonly used particularly in clinical trials. Despite their promise, these methods require sophisticated and costly equipment, limiting their accessibility. Thus, there is a demand for simple and cost-effective assays that utilize improved biomarkers.
In this study, we demonstrate that CA-125 glycovariants, measured with novel GLYVAR Ovarian I and II assays, enhance the diagnostic sensitivity of the detection of ovarian cancer. Their combination differentiated benign and malignant ovarian masses with 88 % sensitivity and 99 % specificity, whereas CA-125 showed 73 % sensitivity. The two glycovariant markers were differently expressed in women with benign ovarian tumors and endometriosis, and their combination led to the most improved performance. The addition of conventional CA-125 to the combination of GLYVAR Ovarian I and II did not significantly improve the diagnostic performance.
While all the studied markers often exhibit high concentrations in late-stage high-grade disease, a clinically challenging group for differential diagnostics includes patients with borderline or moderately elevated CA-125 concentration at diagnosis. We assessed the patients with CA-125 concentrations between 30 and 300 U/mL and found that both CA-125 glycovariants performed remarkably better than the conventional CA-125 in this group. The combination of the two CA-125 glycovariant assays showed 2.5 times higher sensitivity than the conventional CA-125 and corrected 82 % of false positive results given by CA-125 concentrations.
Identification of early-stage carcinomas is challenging but critical for improvement of patient survival. While the relative 5-year survival of patients in FIGO stages III and IV ranges between 20 and 40 %, the survival rate for stage I disease is 90 % [6], [7], [8], [9, 25]. Thus, the establishment of a highly sensitive marker, which could potentially be used for population-based screening programs, has the potential to reduce EOC mortality significantly. Moreover, ovarian masses are often diagnosed in routine transvaginal ultrasound. As a biopsy is generally not feasible due to the anatomic position and mobility of the ovaries as well as the risk of tumor cell spread, dignity can only be determined by (laparoscopic) surgery, often resulting in the removal of the suspicious organ. A reliable marker that can discriminate between benign and malignant masses with high sensitivity would be a massive improvement, facilitating the avoidance of over- and undertreatment. In this study, CA-125 detected 45 % of non-mucinous stage I-II carcinomas at 99 % specificity. The CA-125 glycovariant assays identified 20 % more early-stage (FIGO stage I-II) ovarian carcinomas, 40 % more early-stage serous tumors and 66 % more early-stage HGSC compared to conventional CA-125. Although the patient numbers were low, the results may indicate that measuring differentially glycosylated CA-125 holds remarkable potential for detection of early-stage ovarian carcinomas.
Previous studies have shown that elevated concentrations of specific CA-125 glycoforms (e.g., STn, Tn, ST) are associated with EOC and differentiate EOC from benign gynecological conditions or healthy controls [21], 22], 26], 27]. Improved diagnostic performance has been observed particularly in postmenopausal women and women with marginally elevated CA-125 concentrations. Additionally, high pre-treatment CA-125-STn concentrations have been associated with advanced disease stage carcinomas, high tumor load, and extensive peritoneal dissemination, predicting higher postoperative residual disease. Longitudinal analysis has demonstrated improved detection of EOC recurrence with CA-125-MGL and CA-125-STn.
A limitation of this study is that the biobank cohort included a limited number of benign and healthy controls as compared with the number of EOC patients in normal clinical presentation, which introduces spectrum bias. In addition, the EOC cohort included an exceptionally high proportion of low-grade serous carcinoma (25 %) and low proportion of clear cell and endometrioid subtypes (6.5 % altogether). However, the proportion of HGSC (68.5 % of EOCs) was typical. Thus, further studies are needed to validate the performance of CA-125 glycovariants using GLYVAR Ovarian I and II assays and to determine cut-off values in a large cohort representing the clinical presentation of women suspected of having ovarian cancer. A significantly larger number of patients are needed to reliably evaluate the CA-125 glycovariant assays’ ability to detect stage I EOC. An extended cohort could also provide information about the performance of CA-125 glycovariants in different EOC subtypes.
In conclusion, this is the first report describing performance of GLYVAR Ovarian I and II assays in the classification of benign and malignant ovarian masses. The study highlights the potential of detecting cancer-specific glycoforms of CA-125 in ovarian cancer diagnostics, especially in the early stage of the most common and lethal subtype HGSC and in cases where CA125 is falsely elevated.
Acknowledgments
The following biobanks are acknowledged for delivering samples to the study: Auria Biobank (https://www.auria.fi/biopankki/en/), Biobank of Eastern Finland (https://www.ita-suomenbiopankki.fi/en), and Finnish Clinical Biobank Tampere (https://www.pirha.fi/web/english/for-professionals/finnish-clinical-biobank-tampere). The authors thank all involved nurses and physicians from the Department of Gynecology and Obstetrics for their commitment in sampling and educating the patients. We acknowledge support by the Open Access Publication Fund of the University of Duisburg-Essen.
-
Research ethics: The study was approved by the local Ethics Committee of the University of Duisburg-Essen (Essen 17–7859). The use of Finnish biobank samples were evaluated by the scientific steering groups of Auria Biobank (Decision AB19-8768), Finnish Clinical Biobank Tampere (Decision BB2020-0104) and Biobank of Eastern Finland (Decision BB2020-0104) as well as the board of Finnish Biobank Cooperative – FinBB according to the Biobank Act (688/2012) of Finland. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and complied with all relevant national regulations concerning the use of retrospective human specimens.
-
Informed consent: Informed consent was obtained from all individuals included in this study.
-
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Use of Large Language Models, AI and Machine Learning Tools: None declared.
-
Conflict of interest: Katri Kuningas, Marjut Helle, Leena Kokko and Kaisa Huhtinen are employees of Uniogen Oy, Finland, which has developed the GLYVAR™ Ovarian I and II assays. All other authors state no conflict of interest.
-
Research funding: None declared.
-
Data availability: Not applicable.
References
1. Ferlay, J, Ervik, M, Lam, F, Laversanne, M, Colombet, M, Mery, L, et al.. Global cancer observatory: cancer today (version 1.1). Lyon, France: International Agency for Research on Cancer; 2024 [Online]. Available from: https://gco.iarc.who.int/today [Accessed 1 Apr 2025].Suche in Google Scholar
2. Siegel, RL, Giaquinto, AN, Jemal, A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12–49. Erratum in: CA Cancer J Clin. 2024;74:203.10.3322/caac.21820Suche in Google Scholar PubMed
3. Ray-Coquard, I, Leary, A, Pignata, S, Cropet, C, González-Martín, A, Marth, C, et al.. Olaparib plus bevacizumab first-line maintenance in ovarian cancer: final overall survival results from the PAOLA-1/ENGOT-ov25 trial. Ann Oncol 2023;34:681–92. https://doi.org/10.1016/j.annonc.2023.05.005.Suche in Google Scholar PubMed
4. Monk, BJ, Barretina-Ginesta, MP, Pothuri, B, Vergote, I, Graybill, W, Mirza, MR, et al.. Niraparib first-line maintenance therapy in patients with newly diagnosed advanced ovarian cancer: final overall survival results from the PRIMA/ENGOT-OV26/GOG-3012 trial. Ann Oncol 2024;35:981–92. https://doi.org/10.1016/j.annonc.2024.08.2241.Suche in Google Scholar PubMed PubMed Central
5. Allemani, C, Matsuda, T, Di Carlo, V, Harewood, R, Matz, M, Nikšić, M, CONCORD Working Group, et al.. Global surveillance of trends in cancer survival 2000-14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. Lancet 2018;391:1023–75. https://doi.org/10.1016/s0140-6736(17)33326-3.Suche in Google Scholar
6. Ronckers, C, Spix, C, Trübenbach, C, Katalinic, A, Christ, M, Cicero, A, et al.. (2023). Krebs in Deutschland für 2019/2020 (14. Ausgabe). Berlin: Zentrum für Krebsregisterdaten im Robert Koch-Institut (ZfKD) & Gesellschaft der epidemiologischen Krebsregister in Deutschland e. V. (GEKID); 2023.Suche in Google Scholar
7. Gynecological Association Finnish. Munasarjasyöpä [ovarian cancer]. Duodecim 2001;117:2318–32. Finnish.Suche in Google Scholar
8. Cancer Research UK. Ovarian cancer survival. Available from: https://www.cancerresearchuk.org/about-cancer/ovarian-cancer/survival [Accessed 2 Apr 2025].Suche in Google Scholar
9. Chen, T, Jansen, L, Gondos, A, Emrich, K, Holleczek, B, Katalinic, A, GEKID Cancer Survival Working Group, et al.. Survival of ovarian cancer patients in Germany in the early 21st century: a period analysis by age, histology, laterality, and stage. Eur J Cancer Prev 2013;22:59–67. https://doi.org/10.1097/cej.0b013e3283552e28.Suche in Google Scholar
10. Stowell, SR, Ju, T, Cummings, RD. Protein glycosylation in cancer. Annu Rev Pathol 2015;10:473–510. https://doi.org/10.1146/annurev-pathol-012414-040438.Suche in Google Scholar PubMed PubMed Central
11. Shetty, V, Hafner, J, Shah, P, Nickens, Z, Philip, R. Investigation of ovarian cancer associated sialylation changes in N-linked glycopeptides by quantitative proteomics. Clin Proteom 2012;9:10. https://doi.org/10.1186/1559-0275-9-10.Suche in Google Scholar PubMed PubMed Central
12. Saldova, R, Struwe, WB, Wynne, K, Elia, G, Duffy, MJ, Rudd, PM. Exploring the glycosylation of serum CA125. Int J Mol Sci 2013;14:15636–54. https://doi.org/10.3390/ijms140815636.Suche in Google Scholar PubMed PubMed Central
13. Wu, J, Xie, X, Liu, Y, He, J, Benitez, R, Buckanovich, RJ, et al.. Identification and confirmation of differentially expressed fucosylated glycoproteins in the serum of ovarian cancer patients using a lectin array and LC-MS/MS. J Proteome Res 2012;11:4541–52. https://doi.org/10.1021/pr300330z.Suche in Google Scholar PubMed
14. Kui Wong, N, Easton, RL, Panico, M, Sutton-Smith, M, Morrison, JC, Lattanzio, FA, et al.. Characterization of the oligosaccharides associated with the human ovarian tumor marker CA125. J Biol Chem 2003;278:28619–34. https://doi.org/10.1074/jbc.m302741200.Suche in Google Scholar
15. Liu, S, Tu, C, Zhang, H, Huang, H, Liu, Y, Wang, Y, et al.. Noninvasive serum N-glycans associated with ovarian cancer diagnosis and precancerous lesion prediction. J Ovarian Res 2024;17:26. https://doi.org/10.1186/s13048-024-01350-2.Suche in Google Scholar PubMed PubMed Central
16. Chen, K, Gentry-Maharaj, A, Burnell, M, Steentoft, C, Marcos-Silva, L, Mandel, U, et al.. Microarray Glycoprofiling of CA125 improves differential diagnosis of ovarian cancer. J Proteome Res 2013;12:1408–18. https://doi.org/10.1021/pr3010474.Suche in Google Scholar PubMed
17. Wu, J, Xie, X, Nie, S, Buckanovich, RJ, Lubman, DM. Altered expression of sialylated glycoproteins in ovarian cancer sera using lectin-based ELISA assay and quantitative glycoproteomics analysis. J Proteome Res 2013;12:3342–52. https://doi.org/10.1021/pr400169n.Suche in Google Scholar PubMed
18. Ricardo, S, Marcos-Silva, L, Pereira, D, Pinto, R, Almeida, R, Söderberg, O, et al.. Detection of glyco-mucin profiles improves specificity of MUC16 and MUC1 biomarkers in ovarian serous tumours. Mol Oncol 2015;9:503–12. https://doi.org/10.1016/j.molonc.2014.10.005.Suche in Google Scholar PubMed PubMed Central
19. Cummings, RD, Etzler, M, Hahn, MG, Darvill, A, Godula, K, Woods, RJ, et al., editors. Essentials of glycobiology [Internet], 4th ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2022. Chapter 48.Suche in Google Scholar
20. Soukka, T, Antonen, K, Härmä, H, Pelkkikangas, AM, Huhtinen, P, Lövgren, T. Highly sensitive immunoassay of free prostate-specific antigen in serum using europium(III) nanoparticle label technology. Clin Chim Acta 2003;328:45–58. https://doi.org/10.1016/s0009-8981(02)00376-5.Suche in Google Scholar PubMed
21. Gidwani, K, Huhtinen, K, Kekki, H, van Vliet, S, Hynninen, J, Koivuviita, N, et al.. A nanoparticle-lectin immunoassay improves discrimination of serum CA125 from malignant and benign sources. Clin Chem 2016;62:1390–400. https://doi.org/10.1373/clinchem.2016.257691.Suche in Google Scholar PubMed
22. Gidwani, K, Nadeem, N, Huhtinen, K, Kekki, H, Heinosalo, T, Hynninen, J, et al.. Europium nanoparticle-based sialyl-Tn monoclonal antibody discriminates epithelial ovarian cancer-associated CA125 from benign sources. J Appl Lab Med 2019;4:299–310. https://doi.org/10.1373/jalm.2018.028266.Suche in Google Scholar PubMed
23. Sideris, M, Menon, U, Manchanda, R. Screening and prevention of ovarian cancer. Med J Aust 2024;220:264–74. https://doi.org/10.5694/mja2.52227.Suche in Google Scholar PubMed PubMed Central
24. Charkhchi, P, Cybulski, C, Gronwald, J, Wong, FO, Narod, SA, Akbari, MR. CA-125 and ovarian cancer: a comprehensive review. Cancers 2020;12:3730. https://doi.org/10.3390/cancers12123730.Suche in Google Scholar PubMed PubMed Central
25. Surveillance, epidemiology, and end results (SEER) program (www.seer.cancer.gov) SEER*Stat Database: Populations - Total U.S. (1969–2022), released March 2024. Available from: https://seer.cancer.gov/statfacts/html/ovary.html [Accessed 2 Apr 2025].Suche in Google Scholar
26. Chen, K, Gentry-Maharaj, A, Burnell, M, Steentoft, C, Marcos-Silva, L, Mandel, U, et al.. Microarray glycoprofiling of CA125 improves differential diagnosis of ovarian cancer. J Proteome Res 2013;12:1408–18. https://doi.org/10.1021/pr3010474.Suche in Google Scholar
27. Wang, YS, Ren, SF, Jiang, W, Lu, JQ, Zhang, XY, Li, XP, et al.. CA125-Tn ELISA assay improves specificity of pre-operative diagnosis of ovarian cancer among patients with elevated serum CA125 levels. Ann Transl Med 2021;9:788. https://doi.org/10.21037/atm-20-8053.Suche in Google Scholar PubMed PubMed Central
Supplementary Material
This article contains supplementary material (https://doi.org/10.1515/cclm-2025-0561).
© 2025 the author(s), published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.
