The Journal of Molecular Diagnostics, Vol. 17, No. 3, May 2015
jmd.amjpathol.org
Panels of Cytokines and Other Secretory Proteins as Potential Biomarkers of Ovarian Endometriosis Vida Kocbek,* Katja Vouk,* Nick A. Bersinger,y Michael D. Mueller,y and Tea Lanisnik Rizner* From the Institute of Biochemistry,* Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia; and the Endometriosis Centre,y Department of Obstetrics and Gynaecology, Inselspital, University of Berne, Berne, Switzerland Accepted for publication January 28, 2015. Address correspondence to Tea Lanisnik Rizner, Ph.D., Institute of Biochemistry, Faculty of Medicine, University of Ljubljana, Vrazov trg 2, 1000 Ljubljana, Slovenia. E-mail: [email protected].
Endometriosis is a gynecologic disease that is characterized by nonspecific symptoms and invasive diagnostics. To date, there is no adequate noninvasive method for the diagnosis of endometriosis. Although more than 100 potential biomarkers have been investigated in blood and/or peritoneal fluid, none of these has proven useful in clinical practice. The aim to find a suitable panel of biomarkers that would allow noninvasive diagnosis thus remains of interest. We evaluated the concentrations of 16 cytokines and other secretory proteins in serum and peritoneal fluid of 58 women with ovarian endometriosis (cases) and 40 healthy women undergoing sterilization or patients with benign ovarian cysts (controls) using multiplexed double fluorescenceebased immunometric assay platform and enzyme-linked immunosorbent assay. Significantly higher concentrations of glycodelin-A were shown in serum, and significantly higher levels of glycodelin-A, IL-6, and IL-8, and lower levels of leptin were measured in the peritoneal fluid of cases versus controls. In serum, the best performance was shown by models that included the ratio of leptin/glycodelin-A and the ratio of ficolin 2/ glycodelin-A, whereas in the peritoneal fluid the best models included the ratio of biglycan/leptin, regulated on activation normal T-cell expressed and secreted/IL-6 and ficolin-2/glycodelin-A, and IL-8 per milligram of total protein, all in combination with age. The models using serum and peritoneal fluid distinguished between ovarian endometriosis patients and controls regardless of the menstrual cycle phase with relatively high sensitivity (72.5% to 84.2%), specificity (78.4% to 91.2%), and area under the curve (0.85 to 0.90). (J Mol Diagn 2015, 17: 325e334; http://dx.doi.org/10.1016/j.jmoldx.2015.01.006)
Endometriosis is a common progressive gynecologic disease that is characterized by the growth of endometrial tissue, containing epithelial glands and stroma, outside the uterus. Endometriosis is a very heterogeneous disease, or rather a group of diseases, of which three different types have been defined: ovarian endometriosis, peritoneal endometriosis, and deep infiltrating endometriosis.1 These commonly are classified into four revised American Fertility Society stages: minimal (class I), mild (class II), moderate (class III), and severe (class IV).2 Endometriosis affects 6% to 10% of women of reproductive age,3 and it results in subfertility and pain. Nonspecific symptoms also frequently are seen, such as chronic pelvic pain, dysmenorrhea, dyschezia, and dyspareunia. Nevertheless, it takes 7 years on average before a correct diagnosis is obtained from a laparoscopic procedure.4 Therefore, it would be highly convenient for the patient and clinician to have a Copyright ª 2015 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2015.01.006
tool that would allow an earlier diagnosis of endometriosis, and with the use of less-invasive methods. To date, there are no adequate noninvasive diagnostic methods available for the diagnosis of endometriosis, although transvaginal ultrasound and magnetic resonance imaging can help to detect some types of the disease, especially ovarian endometriomas and endometriotic nodules; however, these techniques are not suitable for smaller lesions.5,6 Additional methods thus are recommended, such as the determination of sensitive and specific biomarkers. More than 100 potential biomarkers have been investigated in Supported by the Slovene Human Resources Development and Scholarship Fund (V.K.) and grant J3-4135 from the Slovenian Research Agency (T.L.R.). Disclosures: A patent entitled “A novel method for the diagnosis of endometriosis” WO2013/178794 A1 is pending (K.V. and T.L.R.).
Kocbek et al Table 1
Characteristics of the Study Participants Control group
Characteristic
Healthy women
Benign cysts
All
Endometriosis group
Number
31
9
40
58
Age, means SD (range), years BMI, means SD (range), kg/m2 Stress score, means SD (range) Cycle phase, n (%) Proliferative Late proliferative/ early secretory Secretory Missing Endometriosis stage, n (%) Stages I þ II Stages III þ IV Endometriosis type, n (%) Ovarian Ovarian þ peritoneal Ovarian þ deep infiltrating Ovarian þ peritoneal þ deep infiltrating
39.1 3.9 (26e45) 38.1 9.7 (26e49) 38.9 5.6 (26e49) 32.7 5.7 (22e44) 25.7 4.3 (18.3e33.9) 22.5 3.4 (18.29e30.42) 24.8 4.3 (18.3e33.9) 21.8 3.5 (17.0e35.9) 4.1 2.5 (0e8) 3.7 2.1 (1e7) 4.0 2.4 (0e8) 5.0 2.9 (0e11)
NA
6 (19.4) 8 (25.8)
0 (0.0) 2 (22.2)
6 (15.0) 10 (25.0)
13 (22.4) 13 (22.4)
16 (51.6) 1 (3.2)
6 (66.7) 1 (11.1)
22 (55.0) 2 (5.0)
28 (48.3) 4 (6.9)
NA
NA 14 (24.1) 44 (75.9)
NA
NA
NA 15 28 9 6
(25.9) (48.3) (15.5) (10.3)
BMI, body mass index; NA, not applicable.
serum and/or peritoneal fluid, but none of these has been shown to be useful in clinical practice.7 The aim to find a suitable biomarker or panel of biomarkers that would allow early detection and lead to early treatment, or at least prevent the progression of this disease, thus remains of interest. The increased fertility and decreased pain would improve the quality of life of patients with endometriosis.7,8 Results from studies on biomarkers for endometriosis have been conflicting and contradictory. The reasons for this are many-fold, and result from the small numbers of patients included in the different studies; group heterogeneities regarding cycle phases, and endometriosis types and stages; control groups without or with nonendometriotic pelvic pathology; sample sources (serum or plasma); inclusion criteria; varying assay sensitivities; and/or inappropriate statistical analysis for biomarker panels.9 Moreover, some of the biomarkers investigated are more reliable for other diseases than for endometriosis [eg, cancer antigen 125 (CA-125) in epithelial ovarian cancer].10 The pathogenesis of endometriosis is still not understood fully, and although the most supported evidence is retrograde menstruation,11 other theories include coelomic metaplasia, Müllerian remnants, and endometrial transplantation.12 However, no single theory can explain all types of endometriosis. Retrograde menstruation occurs in the majority of women, which suggests that other factors need to be included. A negatively affected function of the intraperitoneal immune system probably is responsible for the survival, implantation, and proliferation of endometrial tissue at ectopic locations.13 Peritoneal endometriosis is thus a local pelvic inflammatory disease with an impaired immune response, in which different
cell types and their secreted products have been identified.14 Natural killer cells and peritoneal macrophages have important roles in the clearance of endometriotic cells, whereby decreased cytotoxicity of natural killer cells15 and increased numbers and activational status of peritoneal macrophages have been reported for women with endometriosis.16 Activated peritoneal macrophages secrete cytokines and growth factors, and, as a result, the concentrations of these molecules are increased in the peritoneal fluid and serum of women with endometriosis.12,17 Imbalanced levels of secreted products and inefficient clearance appears to contribute to the growth of endometriotic cells at different ectopic sites.18 Cytokines are proteins that have many different biological functions, among which there are proliferative and differentiative roles, with implications for cell motility, adhesion, chemotaxis, and morphogenesis, potentially promoting, or at least modulating, the pathogenesis of endometriosis.14,19 Cytokines have been studied as potential biomarkers by several groups.20 After a broad literature review, we decided to measure 11 cytokines: IL-6, IL-8, IL-15, IL-18; interferon geinduced protein; monocyte chemotactic protein 1; vascular endothelial growth factor; regulated on activation normal T-cell expressed and secreted (RANTES); growth regulated oncogene a (GRO-a); intercellular adhesion molecule 1; and vascular cell adhesion molecule 1. We combined these together with five other proteins that are involved in immune and inflammatory responses: C-reactive protein, secreted pattern-recognition lectin ficolin-2, endometrium-derived glycodelin, the adipokine hormone leptin, and recently identified potential biomarker biglycan. Although many studies have investigated the levels of most
326
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
New Diagnostic Models for Endometriosis Table 2
Concentrations of the Individual Biomarkers in Serum Concentrations
Statistics
Controls
Serum protein
Healthy women
CRP Ficolin-2 Leptin Glycodelin-A RANTES Biglycan IL-6 IL-8 IL-15 IP-10 MCP-1 VEGF IL-18 Gro-a ICAM-1 VCAM-1
2638 1466 17.7 6.8 39,494 12.90
jmd.amjpathol.org
Panels of Cytokines and Other Secretory Proteins as Potential Biomarkers of Ovarian Endometriosis Vida Kocbek,* Katja Vouk,* Nick A. Bersinger,y Michael D. Mueller,y and Tea Lanisnik Rizner* From the Institute of Biochemistry,* Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia; and the Endometriosis Centre,y Department of Obstetrics and Gynaecology, Inselspital, University of Berne, Berne, Switzerland Accepted for publication January 28, 2015. Address correspondence to Tea Lanisnik Rizner, Ph.D., Institute of Biochemistry, Faculty of Medicine, University of Ljubljana, Vrazov trg 2, 1000 Ljubljana, Slovenia. E-mail: [email protected].
Endometriosis is a gynecologic disease that is characterized by nonspecific symptoms and invasive diagnostics. To date, there is no adequate noninvasive method for the diagnosis of endometriosis. Although more than 100 potential biomarkers have been investigated in blood and/or peritoneal fluid, none of these has proven useful in clinical practice. The aim to find a suitable panel of biomarkers that would allow noninvasive diagnosis thus remains of interest. We evaluated the concentrations of 16 cytokines and other secretory proteins in serum and peritoneal fluid of 58 women with ovarian endometriosis (cases) and 40 healthy women undergoing sterilization or patients with benign ovarian cysts (controls) using multiplexed double fluorescenceebased immunometric assay platform and enzyme-linked immunosorbent assay. Significantly higher concentrations of glycodelin-A were shown in serum, and significantly higher levels of glycodelin-A, IL-6, and IL-8, and lower levels of leptin were measured in the peritoneal fluid of cases versus controls. In serum, the best performance was shown by models that included the ratio of leptin/glycodelin-A and the ratio of ficolin 2/ glycodelin-A, whereas in the peritoneal fluid the best models included the ratio of biglycan/leptin, regulated on activation normal T-cell expressed and secreted/IL-6 and ficolin-2/glycodelin-A, and IL-8 per milligram of total protein, all in combination with age. The models using serum and peritoneal fluid distinguished between ovarian endometriosis patients and controls regardless of the menstrual cycle phase with relatively high sensitivity (72.5% to 84.2%), specificity (78.4% to 91.2%), and area under the curve (0.85 to 0.90). (J Mol Diagn 2015, 17: 325e334; http://dx.doi.org/10.1016/j.jmoldx.2015.01.006)
Endometriosis is a common progressive gynecologic disease that is characterized by the growth of endometrial tissue, containing epithelial glands and stroma, outside the uterus. Endometriosis is a very heterogeneous disease, or rather a group of diseases, of which three different types have been defined: ovarian endometriosis, peritoneal endometriosis, and deep infiltrating endometriosis.1 These commonly are classified into four revised American Fertility Society stages: minimal (class I), mild (class II), moderate (class III), and severe (class IV).2 Endometriosis affects 6% to 10% of women of reproductive age,3 and it results in subfertility and pain. Nonspecific symptoms also frequently are seen, such as chronic pelvic pain, dysmenorrhea, dyschezia, and dyspareunia. Nevertheless, it takes 7 years on average before a correct diagnosis is obtained from a laparoscopic procedure.4 Therefore, it would be highly convenient for the patient and clinician to have a Copyright ª 2015 American Society for Investigative Pathology and the Association for Molecular Pathology. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jmoldx.2015.01.006
tool that would allow an earlier diagnosis of endometriosis, and with the use of less-invasive methods. To date, there are no adequate noninvasive diagnostic methods available for the diagnosis of endometriosis, although transvaginal ultrasound and magnetic resonance imaging can help to detect some types of the disease, especially ovarian endometriomas and endometriotic nodules; however, these techniques are not suitable for smaller lesions.5,6 Additional methods thus are recommended, such as the determination of sensitive and specific biomarkers. More than 100 potential biomarkers have been investigated in Supported by the Slovene Human Resources Development and Scholarship Fund (V.K.) and grant J3-4135 from the Slovenian Research Agency (T.L.R.). Disclosures: A patent entitled “A novel method for the diagnosis of endometriosis” WO2013/178794 A1 is pending (K.V. and T.L.R.).
Kocbek et al Table 1
Characteristics of the Study Participants Control group
Characteristic
Healthy women
Benign cysts
All
Endometriosis group
Number
31
9
40
58
Age, means SD (range), years BMI, means SD (range), kg/m2 Stress score, means SD (range) Cycle phase, n (%) Proliferative Late proliferative/ early secretory Secretory Missing Endometriosis stage, n (%) Stages I þ II Stages III þ IV Endometriosis type, n (%) Ovarian Ovarian þ peritoneal Ovarian þ deep infiltrating Ovarian þ peritoneal þ deep infiltrating
39.1 3.9 (26e45) 38.1 9.7 (26e49) 38.9 5.6 (26e49) 32.7 5.7 (22e44) 25.7 4.3 (18.3e33.9) 22.5 3.4 (18.29e30.42) 24.8 4.3 (18.3e33.9) 21.8 3.5 (17.0e35.9) 4.1 2.5 (0e8) 3.7 2.1 (1e7) 4.0 2.4 (0e8) 5.0 2.9 (0e11)
NA
6 (19.4) 8 (25.8)
0 (0.0) 2 (22.2)
6 (15.0) 10 (25.0)
13 (22.4) 13 (22.4)
16 (51.6) 1 (3.2)
6 (66.7) 1 (11.1)
22 (55.0) 2 (5.0)
28 (48.3) 4 (6.9)
NA
NA 14 (24.1) 44 (75.9)
NA
NA
NA 15 28 9 6
(25.9) (48.3) (15.5) (10.3)
BMI, body mass index; NA, not applicable.
serum and/or peritoneal fluid, but none of these has been shown to be useful in clinical practice.7 The aim to find a suitable biomarker or panel of biomarkers that would allow early detection and lead to early treatment, or at least prevent the progression of this disease, thus remains of interest. The increased fertility and decreased pain would improve the quality of life of patients with endometriosis.7,8 Results from studies on biomarkers for endometriosis have been conflicting and contradictory. The reasons for this are many-fold, and result from the small numbers of patients included in the different studies; group heterogeneities regarding cycle phases, and endometriosis types and stages; control groups without or with nonendometriotic pelvic pathology; sample sources (serum or plasma); inclusion criteria; varying assay sensitivities; and/or inappropriate statistical analysis for biomarker panels.9 Moreover, some of the biomarkers investigated are more reliable for other diseases than for endometriosis [eg, cancer antigen 125 (CA-125) in epithelial ovarian cancer].10 The pathogenesis of endometriosis is still not understood fully, and although the most supported evidence is retrograde menstruation,11 other theories include coelomic metaplasia, Müllerian remnants, and endometrial transplantation.12 However, no single theory can explain all types of endometriosis. Retrograde menstruation occurs in the majority of women, which suggests that other factors need to be included. A negatively affected function of the intraperitoneal immune system probably is responsible for the survival, implantation, and proliferation of endometrial tissue at ectopic locations.13 Peritoneal endometriosis is thus a local pelvic inflammatory disease with an impaired immune response, in which different
cell types and their secreted products have been identified.14 Natural killer cells and peritoneal macrophages have important roles in the clearance of endometriotic cells, whereby decreased cytotoxicity of natural killer cells15 and increased numbers and activational status of peritoneal macrophages have been reported for women with endometriosis.16 Activated peritoneal macrophages secrete cytokines and growth factors, and, as a result, the concentrations of these molecules are increased in the peritoneal fluid and serum of women with endometriosis.12,17 Imbalanced levels of secreted products and inefficient clearance appears to contribute to the growth of endometriotic cells at different ectopic sites.18 Cytokines are proteins that have many different biological functions, among which there are proliferative and differentiative roles, with implications for cell motility, adhesion, chemotaxis, and morphogenesis, potentially promoting, or at least modulating, the pathogenesis of endometriosis.14,19 Cytokines have been studied as potential biomarkers by several groups.20 After a broad literature review, we decided to measure 11 cytokines: IL-6, IL-8, IL-15, IL-18; interferon geinduced protein; monocyte chemotactic protein 1; vascular endothelial growth factor; regulated on activation normal T-cell expressed and secreted (RANTES); growth regulated oncogene a (GRO-a); intercellular adhesion molecule 1; and vascular cell adhesion molecule 1. We combined these together with five other proteins that are involved in immune and inflammatory responses: C-reactive protein, secreted pattern-recognition lectin ficolin-2, endometrium-derived glycodelin, the adipokine hormone leptin, and recently identified potential biomarker biglycan. Although many studies have investigated the levels of most
326
jmd.amjpathol.org
-
The Journal of Molecular Diagnostics
New Diagnostic Models for Endometriosis Table 2
Concentrations of the Individual Biomarkers in Serum Concentrations
Statistics
Controls
Serum protein
Healthy women
CRP Ficolin-2 Leptin Glycodelin-A RANTES Biglycan IL-6 IL-8 IL-15 IP-10 MCP-1 VEGF IL-18 Gro-a ICAM-1 VCAM-1
2638 1466 17.7 6.8 39,494 12.90
