- Email: [email protected]
or endometriosis
European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103 www.elsevier.com/locate/ejogrb
Peritoneal fluid cytokines and sICAM-1 in minimal endometriosis: search for discriminating factors between infertility and/or endometriosis Jana Skrzypczak a, Małgorzata Szczepan´ska a, Ewa Puk a, Marzena Kamieniczna b, Maciej Kurpisz b,* a
Clinic of Reproduction, University Medical School, ul. Polna 33, 60-533 Poznan´, Poland b Unit of Reproductive Biology, Institute of Human Genetics, Pol. Acad. Sci., ul. Strzeszyn´ska 32, 60-479 Poznan´, Poland Accepted 19 November 2004
Abstract Objective: To evaluate cytokine levels (IL-1b, TNF-a, IL-6, IL-8), soluble intercellular adhesion molecule-1 (sICAM-1) and number of macrophages in peritoneal fluid (PF) of women with no minimal endometriosis and associated (or not) infertility in order to discriminate between infertility and/or endometriosis. Study design: Cytokines and sICAM-1 were measured by using quantitative enzyme-linked immunosorbent assay (ELISA) while the macrophages were identified by May-Grunwald-Giemsa and non-specific esterase staining and presented as medians. The measurements were performed in 78 women belonging to four selected subgroups according to their endometriosis and/or infertility status. Statistical analysis was performed using Kruskal–Wallis non-parametric ANOVA test. Additionally, discriminant function analyses were performed. Results: We have found the most elevated macrophage numbers in the groups of women with endometriosis. IL-1b did not present any statistically significant values differentiating the analysed subgroups. IL-6 (110.0 pg/ml) and TNF-a exhibited the highest concentrations (statistically significant) in a group of fertile women with endometriosis. IL-8 clearly differentiated between the subgroups with infertility and sICAM-1 was statistically significantly elevated in the subgroups of infertile women. In the forward discriminant analysis, when subdividing the studied population according to its infertility status (we considered macrophages, IL-8 and IL-6 in order of probability values), we have found striking probability value of 92% for the correct classification into an infertile population. Conclusion: Out of the range of the analysed factors we have found only the sICAM-1 to be singled out between the standard discriminant analysis and the forward one. However, this important flagging molecule might be of considerable value for discrimination between different types of pathology at the level of immune effector function. The increased levels of TNF-a and IL-6 signified a group of fertile women with endometriosis; however only IL-6 presented a discriminating value in multifunctional analysis of examined subgroups. The analysed range of factors had a greater tendency to discriminate between infertility status rather than endometriosis. # 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Cytokines; Adhesion molecules; Endometriosis; Infertility
1. Introduction Endometriosis is considered to be one of the most common gynaecological disorders and has been associated with infertility [1]. * Corresponding author. Tel.: +48 61 8233 155; fax: +48 61 8233 235.
The relationship between minimal and medium stages of endometriosis with infertility is unclear although quite frequent observations of endometrial foci in women primarily diagnosed for persistent infertility may point out such an interdependence. In which way endometriosis may deteriorate the environment for conception has never been satisfactorily explained. Among the possible causes of
0301-2115/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2004.11.044
96
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
infertility in women with endometriosis, we should mention:
2. Material and methods 2.1. Patients
- endocrine and immune disorders as well as; - altered peritoneal fluid (PF) environment. Clearly the changes in the peritoneal fluid content, which are well demonstrated in women with endometriosis [2,3], may disturb a fertilization process on different levels. It is accepted that the presence of ectopic endometrium may be an inducing factor for inflammatory properties of peritoneal fluid. This, in turn, can be manifested by the evident and well established: (a) increase in PF volume, (b) elevated numbers and activity of macrophages, (c) elevated cytokine, prostanoid, adhesion molecules contents [4,5]. It should be emphasized that a significant amount of work on cytokine content in peritoneal fluid of women with endometriosis is what underlines the importance of these biological factors in perpetuation of the disease. Cytokines released by macrophages as IL-1b [6,7], IL-6 [8–10], TNF-a [11–14] have been extensively studied and found to be elevated when compared to peritoneal fluid of healthy women and seem to drive the inflammatory reaction, together with IL-8, in this topographical region [12,13,15,16]. Soluble intercellular adhesion molecule-1 is essential for various immunological reactions and may be a useful parameter in the diagnosis of many immune diseases. In endometriosis, sICAM-1 may allow implantation of endometrial cells in ectopic sites [5]. sICAM-1 levels in serum and peritoneal fluid of patients with endometriosis were elevated [17–19] and were released in high amounts by isolated endometriotic stromal cells [20]. However, the role of sICAM-1 as an indicator for the diagnosis of endometriosis and/or infertility has not been thoroughly investigated [17,21]. A link between infertility and endometriosis of minimal or medium stage is still not well explained. Further to that controversy [22], it was shown that fertility in non-human primates is comparable either during induced or without endometriosis. The rate of implantations after IVF (in women with or without endometriosis) has also been previously shown to be similar [23]. In the present study, we have therefore selected several subgroups of women: (a) infertile with endometriosis; (b) fertile with endometriosis; (c) infertile (no endometriosis); (d) fertile, control population (no endometriosis), screening them for a range of cytokines (IL-1b, TNF-a, IL-6, IL-8) and sICAM-1 molecule as well as defining the macrophage numbers in PF samples. A second goal was an attempt to assess the predictive values for all these parameters in order to discriminate between the different subgroups under study. For this reason, the levels of measured substances were statistically elaborated and the discriminant function analyses were performed in respect to each of the selected subgroups or dividing the studied populations according to their infertility status.
Seventy eight women of reproductive age (21–41) from the Department and Clinic of Reproduction of the Karol Marcinkowski University of Medical Sciences in Poznan´ were subjected to laparoscopy due to unexplained infertility (with normal parameters of ovulation, hormonal profile and patent fallopian tubes), pelvic pain or suspected endometriosis. Individuals with history of autoimmune reactions, neoplastic disorders or microbial infections were excluded from this study. All studied subjects had regular menstrual cycles (26–30 days) and were not taking any hormonal preparations, e.g. Danazol, gonadotropin-releasing hormone agonists or contraceptives for at least 3 months prior to laparoscopy. At the time of laparoscopy, performed in the luteal phase of sexual cycle, peritoneal fluid samples were collected from 78 individuals, out of whom 48 had an evidence of minimal endometriosis, while the remaining 30 were considered as individuals with no endometriotic symptoms. A stage of endometriosis was classified according to the Revised American Fertility Society classification [24]. The patient cohorts were divided into four following subgroups: 1. 2. 3. 4.
Endometriosis with associated infertility (n = 28); Endometriosis without infertility (n = 20); Unexplained infertility with no endometriosis (n = 20); Control group (n = 10).
All studied subjects were under 41 years of age. They had a mean (S.D.) age of: 27.6 1.2 years in the first group, 26.5 0.9 in the second group, 28.1 0.1 in the third group and 27.9 0.8 in the fourth group. Out of the 78 women included in this study, 48 (groups 1 and 3) were considered to be infertile (28 had minimal endometriosis and 20 had no laparoscopic evidence of endometriosis) and 30 (groups 2 and 4) were fertile (live birth during last 12 months). The indication for laparoscopy in the group of fertile women was pelvic pain syndrome. The minimal endometriosis was diagnosed at the time of laparoscopy in 20 fertile women (group 2). Thus, these women did not exhibit previous endometriotic history or infertility. In 10 women used as healthy controls, laparoscopic examination demonstrated normal status of pelvic organs. 2.2. Peritoneal fluid samples Peritoneal fluid was aspirated during laparoscopy having the patient in the Trendelenburg position. Samples were obtained via laparoscopic cannula introduced to the cul-desac and aspirated to the syringe before any further medical procedures were undertaken. The probe was introduced by
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
trocar, which was inserted over the pubical symphysis in the midline of the body. After macroscopic inspection (volume, clarity, colour), the obtained samples were centrifuged at 650 g for 10 min. Supernatant was divided into 200 ml aliquots and stored at 70 8C until use. Each sample was subjected to bacteriological analysis and the contaminated ones (12%) were excluded from this study.
97
test. A p-value of less than 0.05 was considered to be statistically significant. Regression analysis between studied parameters in each subgroup or total investigated population was performed as well as standard and forward discriminant function analyses [26].
3. Results 2.3. Bacteriological analysis 3.1. Macrophages Bacteriological analysis was performed using a solid phase quantitative method of assessment. Plates were seeded in aerobic and anaerobic conditions. After incubation in 37 8C, the number of microorganisms in 1 ml of sample was estimated by using Bio Merieux tests as following: (1) (2) (3) (4) (5) (6)
Staphylococci: ID32 Staph test; Streptococci: API20 Strep test; Bacteriaceae Gram-negative: K32E test; Corynebacteriaceae: API Coryne test; Anaerobic Bacteriaceae: ID 32C test; Mycoplasmaceae: Mycoplasmic IST test—allowing to define the species, their concentration as well as sensitivity to antibodies.
2.4. Measurement of cytokines and sICAM-1 by ELISA (enzyme-linked immunosorbent assay) The concentrations of cytokines, i.e. IL-1b, TNF-a, IL-6, IL-8 were measured using commercially available ELISA kits (Predicta-Genzyme Diagnostics, Germany) according to the instruction of manufacturer. Similar assays were performed to measure sICAM-1 levels. The concentrations of cytokines were expressed in pg/ml and for sICAM-1 at ng/ml, and the interassay and intraassay variation coefficients were <5 and <3%, respectively. Cytokine and sICAM-1 measurements for each patient were performed in duplicates. 2.5. Macrophage identification Macrophages were identified by May-Grunwald-Giemsa staining and reaction with non-specific esterase [25], and their concentrations were appropriately calculated (number of cells/ml of peritoneal fluid). Cells were counted in a Thoma hemocytometer. Cell viability test by 1% trypan blue exclusion was performed on the cells resuspended after the first spin and revealed 95% of viability in the cell processed samples.
Analysing the macrophage concentrations per milliliter of peritoneal fluid, we have found statistically significant differences ( p < 0.05) between subgroups of infertile women with endometriosis versus infertility alone (Fig. 1a). The highest macrophage numbers were observed in the studied subgroups with endometriosis. 3.2. Cytokine concentrations 3.2.1. Interleukin-1b The levels of this cytokine, on average, were relatively low in each of the examined subgroups and were not found to be statistically significant. Both those women with infertility and endometriosis as well as the infertile women exhibited median values slightly greater than 3 pg/ml (Fig. 1b). 3.2.2. Tumor necrosis factor (TNF-a) The highest and most significant levels of this cytokine were found in a group of fertile women with endometriosis, which statistically differentiated this subgroup from a group of women with endometriosis and infertility (Fig. 1c). The median values for all tested subgroups were 0 pg/ml and control subjects did not exhibit TNF-a in a detectable range. 3.2.3. Interleukin-6 The highest and statistically significant level of IL-6 (110.0 pg/ml) was found in a group of fertile women with endometriosis, thus statistically signifying this subgroup from the population of infertile women with endometriosis and women with unexplained infertility (Fig. 1d). 3.2.4. Interleukin-8 Although the highest level of this interleukin has been identified in a group of infertile women with endometriosis (16.0 pg/ml), interleukin-8 was found to be statistically significant for infertile women (Fig. 1e). 3.3. Soluble intercellular adhesion molecule (sICAM-1)
2.6. Statistical analysis The results are given as median with average deviation. Statistical analysis was performed using Kruskal–Wallis non-parametric ANOVA test. Differences between analysed subgroups were compared by Dunn Multiple Comparison
Interestingly, both infertile subgroups (with or without endometriosis) exhibited here almost equal levels of this soluble adhesion molecule, which differentiated them (in statistically significant way) from fertile women with endometriosis or fertile controls (Fig. 1f).
98
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
Fig. 1. (a) Macrophage concentration (thousands of cells/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. C; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. (b) IL-1b concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Kruskal–Wallis nonparamtric ANOVA test indicated the obtained data as insignificant. Percentage of data obtained below the sensitivity limit: A:35%; B:45%; C:30%; D:33%. (c) TNF-a concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. B; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
99
Fig. 1. (Continued ).
the sensitivity limit: A:93%; B:55%; C:30%; D:100%. (d) IL-6 concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. B; B vs. C; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:60%; B:30%; C:70%; D:70%. (e) IL-8 concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. C; p < 0.01. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis nonparametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:18%; B:35%; C:65%; D:50%. (f) sICAM1 concentration (ng/ml) in peritoneal fluid samples from infertile or fertile women with and without endometriosis A vs. B and B vs. C; p < 0.01. C vs. D; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:18%; B:45%; C:0%; D:50%.
100
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
Fig. 2. (a) Negative correlation between macrophages and sICAM-1 in a group of fertile women with endometriosis. (b) Positive correlation between macrophages and sICAM-1 in a group of infertile women with endometriosis (p = 0.0055).
3.4. Statistical analysis When calculating the regression curves, most strikingly the correlations were observed, between concentrations of macrophages and sICAM-1 (Fig. 2a and b). In this pair of variables (Fig. 2a) a negative correlation was observed between the concentration of macrophages and sICAM-1 in population of fertile women with endometriosis, however a positive correlation between the same pair of variables was found in a group of infertile women with endometriosis (Fig. 2b). It became clear that infertility status and not endometriosis was a critical factor for the mode of observed relationship. Among the six analysed parameters (macrophages, IL1b, TNF-a, IL-6, IL-8, sICAM-1) in the standard discriminant function analysis, the four variables (Fig. 3; sICAM-1, IL-6, macrophages, IL-8 in order of the probability values) appear to account for most of the variation in the dataset. This analysis provided reasonable significant probabilities to classify correctly at least three (out of four) selected to our study subgroups (i.e. with the
exception for a group of women with unexplained infertility (see Table 1). However, performing the forward stepwise discriminant analysis and subdividing the whole studied population according to its infertility status, we ended up with only three statistically significant (for classification) variables, i.e. macrophages, IL-8 and IL-6 (in order of the probability values). In such evaluation, we have obtained the striking 92% probability value of correct classification into infertile population (Table 2). It has to be noted that sICAM-1 was not included here ( p > 0.05) in spite of its highest statistical significance for the standard discriminant analysis ( p > 0.000009).
4. Discussion In the present study, we have evaluated several biological parameters of peritoneal fluid, as macrophage numbers, cytokine contents (IL-1b, TNF-a, IL-6, IL-8) and sICAM-1 levels. For a majority of the measured biological parameters
Fig. 3. Standard discriminant analysis for correct classification of four studied subgroups by using two variables (roots) with highest significance for discrimination (sICAM-1 and IL-6).
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
101
Table 1 Standard discriminant function analysis for correct classification of studied subgroups by using sICAM-I, IL-6, MØ and IL-8a Statistical discriminant analysis
Classification matrix (standard method) (rows: observed classifications columns: predicted classifications)
Group
Percent correct
G 1b (p = 0.35897)
G 2c (p = 0.25641)
G G G G
1 2 3 4
67.85714 65.00000 40.00000 60.00000
19 4 11 1
2 13 1 1
3 1 8 2
4 2 0 6
Total
58.97436
35
17
14
12
a b c d e
G 3d (p = 0.25641)
G 4e (p = 0.12821)
Rank of probability values used for standard discriminant function analysis: sICAM-1, p < 0.000009; MØ, p < 0.018; IL-6, p < 0.000095; IL-8, p < 0.05. Infertile with endometriosis. Fertile with endometriosis. Infertile (no endometriosis). Fertile (no endometriosis) controls.
Table 2 Classification of correct discrimination (forward analysis) between fertile and infertile populations by using MØ, IL-8 and IL-6 valuesa Statistical discriminant analysis
Classification matrix (standard method) (rows: observed classifications columns: predicted classifications)
Group
Percent correct
Fertile populations (p = 0.30189)
Infertile populations (p = 0.69811)
Fertile population Infertile population
50.00000 91.66666
15 4
15 44
Total
75.64103
19
59
a
Rank of probability values used for forward discriminat functional analysis: MØ, p < 0.0009; IL-8, p < 0.012; IL-6, p < 0.05.
in this study (macrophage concentration, TNF-a, IL-6; Fig. 1a, c and d), the obtained results supported previous observations indicating the elevated macrophage numbers and cytokine contents in peritoneal fluid of women with endometriosis as comparing to the control populations [2– 5,10,27]. However, in presented study, we have observed a significant difference between the concentration of macrophages in a group with minimal endometriosis and infertility versus infertile women without endometriosis (Fig. 1a, p < 0.05), while the increased levels of TNF-a and IL-6 signified fertile women with endometriosis (Fig. 1c and d). The more controversial situation has been observed for IL1b (Fig. 1b) since some authors reported elevated levels of this cytokine in peritoneal fluid of women with endometriosis [2,6,7,11]. A majority of these reports, however, can only really be accounted to one group of researchers. Additionally, in our previous observations, we have shown quite substantial reservation to the significance of IL-1 in endometriotic women, not excluding its pro-inflammatory action, albeit not just only to the endometriosis [28,29]. The observed high level of IL-8 in a group of infertile women with endometriosis (Fig. 1e) supported the previous reports on elevated (statistically significant) IL-8 levels in peritoneal fluid of women with endometriosis [11–13,16,30]. In the group of fertile women with endometriosis we did not observe similar high levels of IL-8 (Fig. 1e). The most striking observation revealed in our study, however, was the high and statistically significant level of sICAM-1 (Fig. 1f). ICAM-1 is one of the most intriguing
molecules of controversial significance. Undoubtedly, cellular ICAM-1 is involved in a range of biological actions, among the others helping at inter-cellular cooperation and antigenic presentation [31]. It can be at least dual explanation of elevated sICAM-1 values, only depending on its immune properties. Firstly, sICAM-1 molecule can be shed from the activated cells involved in immune reactions thus manifesting an extensive antigenic presentation in the system, e.g. increased mucosal concentrations of s-ICAM-1 were reported during intestinal inflammation [32]. Moreover, in inflammatory bowel disease, Crohn disease or ulcerative colitis [33,34] both sICAM-1 and IL-8 were concomitantly elevated, exactly as in a case in our observations of infertile women (Fig. 1e and f) To support a hypothesis of enhanced antigenic presentation during minimal endometriosis, we may further quote a positive correlation between peritoneal macrophages and sICAM-1 (Fig. 2b) found in a group of infertile women with endometriosis. It is intriguing, however, whether the second possibility could be also potentially valid, i.e. whether the shedded sICAM-1 can be an inhibitory molecule (negative feedback) suppressing the antigenic presentation but exhibiting a side effect resulting in diminution of cytotoxic effector response. It could be true that in such circumstances the negative sICAM-1 correlation with macrophages (Fig. 2a) in fertile women with minimal endometriosis could be a critical factor for perpetuation of this disease. The presence of IL-1b and TNF-a (in peritoneal fluid) can be a sign of immunoinflammatory reactions in the
102
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
pelvis, and not necessarily bound to endometriosis alone, hence they may be equally important in acute inflammatory reactions that can be involved at initiation of infertility (development of antigamete antibodies?). Such was a case in our previous observations [35] where antizona antibodies in follicular fluid were concomitantly identified with either IL1b or TNF-a. Another interesting conclusion stems from our calculations by using the discriminant function analysis tests. On the contrary to the moderate classification values obtained in the standard method of discrimination (not exceeding the 68% for correct classification of subgroups; see Table 1), the forward one brought a striking 92% correct classification to the group of infertile women when using obtained values of macrophages, IL-8 and IL-6 (Table 2). Interestingly, sICAM-1 which was of the highest probability value in a standard discriminant analysis (Fig. 3) to differentiate between the subgroup of infertile women with endometriosis from any other studied populations (Table 1) was totally left out in forward (stepwise) analysis for infertility discrimination (Table 2) due to its lack of significance in this context ( p > 0.05). Assuming that sICAM-1 seems to be of considerable value for flagging the disturbances of immune effector function, it should be further emphasized that elevated levels of soluble ICAM-1 molecule can compete and be a reason for disruption between cellular ICAM-1 and LFA-1 binding. It may thus weaken the antigenic presentation (10-fold) as well as the effective cytotoxic response [36,37]. On the other hand, it was previously reported that sICAM-1 levels may also be correlated with their release by granulosa luteal cells thus being a potential sign of ovarian function (or dysfunction) [38,39]. This would interestingly correspond to misleading sICAM-1 significance as for the altered immune local response. Overall, such overlapping biological phenomena may hinder interpretation of the obtained multifactorial analysis where sICAM-1 reached a high statistical significance in the standard discrimination ( p < 0.000009) and a lack of significance in the progressive one ( p > 0.05). Overall, our data point out an easier discrimination between the infertility status of the examined subjects than in respect to minimal endometriosis while taking into account selected to this study biological parameters. References [1] Strathy J, Molgaard C, Coulam C, Melton LJ. Endometriosis and infertility: a laparoscopic study of endometriosis among fertile and infertile women. Fertil Steril 1982;38:667–72. [2] Ho HN, Wu MY, Yang YS. Peritoneal cellular immunity and endometriosis. Am J Reprod Immunol 1997;38:400–12. [3] Lebovic DI, Mueller MD, Taylor RN. Immunobiology of endometriosis. Fertil Steril 2001;75:1–10. [4] Rier SE, Yeaman GR. Immune aspects of endometriosis: relevance of the uterine mucosal immune system. Semin Reprod Endocrinol 1997;15:209–20.
[5] Harada T, Iwabe T, Terakawa N. Role of cytokines in endometriosis. Fertil Steril 2001;76:1–10. [6] Fakih H, Baggett B, Holtz G, Tsang KY, Lee JC, Williamson HO. Interleukin-1: a possible role in the infertility associated with endometriosis. Fertil Steril 1987;47:213–7. [7] Kuo TM, Taketani Y, Shimizu T, Mizuno M. A possible mechanism of infertility in endometriosis – a role of interleukin 1 in peritoneal fluid. Asisst Reprod Technol/Androl 1990;1:24–31. [8] Rier SE, Parsons AK, Becker JL. Altered interleukin-6 production by peritoneal leukocytes from patients with endometriosis. Fertil Steril 1994;61:294–9. [9] Koyama N, Masuura K, Okamura H. Cytokines in the peritoneal fluid of patients with endometriosis. Int J Gynecol Obstet 1993;43:45–50. [10] Mahenke JL, Dawood Y, Huang JC. Vascular endothelial growth factor and interleukin-6 in peritoneal fluid of women with endometriosis. Fertil Steril 2002;73:166–70. [11] Rana N, Braun DP, House R, Gebel H, Rotman C, Dmowski WP. Basal and stimulated secretion of cytokines by peritoneal macrophages in women with endometriosis. Fertil Steril 1996;65:925–30. [12] Pizzo A, Salmeri FM, Ardita FV, Sofo V, Tripepi M, Marsico S. Behaviour of cytokine levels in serum and peritoneal fluid of women with endometriosis. Gynecol Obstet Invest 2002;54:82–7. [13] Harada T, Enatsu A, Mitsunari M, Nagano Y, Ito M, Tsudo T, et al. Role of cytokines in progression of endometriosis. Gynecol Obstet Invest 1999;47(Suppl. 1):34–9. [14] Bedaiwy MA, Falcone T, Sharma RK, Goldberg JM, Attaran M, Nelson DR, et al. Prediction of endometriosis with serum and peritoneal fluid markers: a prospective controlled trial. Hum Reprod 2002;17:426–31. [15] Ryan IP, Tseng JF, Schirock ED, Khorram O, Landers DV, Taylor RN. Interleukin-8 concentrations are elevated in peritoneal fluid of women with endometriosis. Fertil Steril 1995;63:929–32. [16] Iwabe T, Harada T, Tsudo T, Tanikawa M, Onohara Y, Terakawa N. Pathogenetic significance of increased levels of interleukin-8 in the peritoneal fluid of patients with endometriosis. Fertil Steril 1998;69: 924–30. [17] Somigliana E, Vigano P, Candiani M, Felicetta I, Di Blasio AM, Vignali M. Use of serum-soluble intercellular adhesion molecule-1 as a new marker of endometriosis. Fertil Steril 2002;77:1028–31. [18] Daniel Y, Geva E, Amit A, Eshed-Englender T, Baram A, Fait G, et al. Do soluble cell adhesion molecules play a role in endometriosis? Am J Reprod Immunol 2000;43:160–6. [19] Calhaz-Jorge C, Costa AP, Santos MC, Palma-Carlos ML. Soluble intercellular adhesion molecule 1 in the peritoneal fluid of patients with endometriosis correlates with the extension of peritoneal implants. Eur J Obstet Gynecol Reprod Biol 2003;106:170–4. [20] Vigano P, Gaffuri B, Somigliana E, Busacca M, Di Blasio AM, Vignali M. Expression of intercellular adhesion molecule (ICAM)-1 mRNA and protein is enhanced in endometriosis versus endometrial stromal cells in culture. Mol Hum Reprod 1998;4:1150–6. [21] Hammadeh ME, Fischer-Hammadeh C, Hoffmeister H, Huebner U, Georg T, Rosenbaum P, et al. Fibroblast growth factor (FGF), intracellular adhesion molecule (sICAM-1) level in serum and follicular fluid of infertile women with polycystic ovarian syndrome, endometriosis and tubal damage, and their effect on ICSI outcome. Am J Reprod Immunol 2003;50:124–30. [22] D’Hooghe TM, Bambra CS, Koninckx PR. Cycle fecundity in baboons of proven fertility with minimal endometriosis. Gynecol Obstet Invest 1994;37:63–5. [23] Badawy SZ, Cuenca V, Stitzel A, Jacobs RD, Tomar RH. Autoimmune phenomena in infertile patients with endometriosis. Obstet Gynecol 1984;63:271–5. [24] The American Fertility Society. Revised American Fertility Society classification of endometriosis. Fertil Steril 1985; 43: 351–2. [25] Hudson L, Hay FC, editors. Practical immunology. Oxford: Blackwell Scientific Publications, 1991. [26] Albert A. Discriminant analysis and medical diagnosis. Biocybernet Biomed Eng 1995;15:139–56.
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103 [27] Witz CA, Schenken RS. Pathogenesis. Semin Reprod Endokrinol 1997;15:199–208. [28] Skrzypczak J, Je˛drzejczak P, Kasprzak M, Puk E, Kurpisz M. Inflammatory cytokines in peritoneal fluid of women with endometriosis. Ann NY Acad Sci 1995;762:496–8. [29] Skrzypczak J, Puk E, Szymanowski K, Je˛drzejczak P, Kurpisz M. Peritoneal fluid cytokines (TNFa, IL-6 IL-1) in women with endometriosis and infertility. Arch Perinat Med 1996;1:77–81. [30] Arici A, Tazuke SI, Attar E, Kliman HJ, Olive DL. Interleukin-8 concentration in peritoneal fluid of patients with endometriosis and modulation of interleukin-8 expression in human mesothelial cells. Mol Hum Reprod 1996;2:40–5. [31] van de Stolpe A, van der Saag PT. Intercellular adhesion molecule-1. J Mol Med 1996;74:13–33. [32] Nielsen OH, Brynskov J, Vainer B. Increased mucosal concentrations of soluble intercellular adhesion molecule-1 (sICAM), sE-selectin and interleukin-8 in active ulcerative colitis. Dig Dis Sci 1996;41: 1780–5. [33] Brynskov J, Nielsen OH, Ahnfelt-Ronne I, Bendtzen K. Cytokines (immunoinflammatory hormones) and their natural regulation of inflammatory bowel disease (Crohn’ disease and ulcerative colitis): a review. Dig Dis 1994;12:290–304.
103
[34] Pedersen G, Brynskov J, Nielsen OH, Bendtzen K. Adhesion molecules in inflammatory and neoplastic intestinal diseases. Dig Dis 1995;13:322–36. [35] Skrzypczak J, Puk E, Szczepan´ska M. A possible effect of selected folicular fluid factors on fertilizing ability of the oocyte. Centr Eur J Immunol 1997;22:104–9. [36] Meyer DM, Dustin ML, Carron CP. Characterization of intercellular adhesion molecule-1 ectodomain (sICAM) as an inhibitor of lymphocyte function-associated molecule-1 interaction with ICAM-1. J Immunol 1995;155:3578–84. [37] Woska Jr JR, Morelock MM, Jeanfavre DD, Bormann BJ. Characterization of molecular interactions between intercellular adhesion molecule-1 and leukocyte function-associated antigen-1. J Immunol 1996;156:4680–5. [38] Vigano P, Gaffuri B, Ragni G, Di Blasio AM, Vignali M. Intercellular adhesion molecule-1 is expressed on human granulosa cells and mediates their binding to lymphoid cells. J Clin Endocrinol Met 1997;82:101–5. [39] Vigano P, Fusi F, Gaffuri B, Bonzi V, Ferrari A, Vignali M. Soluble intercellular adhesion molecule-1 in ovarian follicles: production by granulosa luteal cells and levels in follicular fluid. Fertil Steril 1998;69:774–9.
Peritoneal fluid cytokines and sICAM-1 in minimal endometriosis: search for discriminating factors between infertility and/or endometriosis Jana Skrzypczak a, Małgorzata Szczepan´ska a, Ewa Puk a, Marzena Kamieniczna b, Maciej Kurpisz b,* a
Clinic of Reproduction, University Medical School, ul. Polna 33, 60-533 Poznan´, Poland b Unit of Reproductive Biology, Institute of Human Genetics, Pol. Acad. Sci., ul. Strzeszyn´ska 32, 60-479 Poznan´, Poland Accepted 19 November 2004
Abstract Objective: To evaluate cytokine levels (IL-1b, TNF-a, IL-6, IL-8), soluble intercellular adhesion molecule-1 (sICAM-1) and number of macrophages in peritoneal fluid (PF) of women with no minimal endometriosis and associated (or not) infertility in order to discriminate between infertility and/or endometriosis. Study design: Cytokines and sICAM-1 were measured by using quantitative enzyme-linked immunosorbent assay (ELISA) while the macrophages were identified by May-Grunwald-Giemsa and non-specific esterase staining and presented as medians. The measurements were performed in 78 women belonging to four selected subgroups according to their endometriosis and/or infertility status. Statistical analysis was performed using Kruskal–Wallis non-parametric ANOVA test. Additionally, discriminant function analyses were performed. Results: We have found the most elevated macrophage numbers in the groups of women with endometriosis. IL-1b did not present any statistically significant values differentiating the analysed subgroups. IL-6 (110.0 pg/ml) and TNF-a exhibited the highest concentrations (statistically significant) in a group of fertile women with endometriosis. IL-8 clearly differentiated between the subgroups with infertility and sICAM-1 was statistically significantly elevated in the subgroups of infertile women. In the forward discriminant analysis, when subdividing the studied population according to its infertility status (we considered macrophages, IL-8 and IL-6 in order of probability values), we have found striking probability value of 92% for the correct classification into an infertile population. Conclusion: Out of the range of the analysed factors we have found only the sICAM-1 to be singled out between the standard discriminant analysis and the forward one. However, this important flagging molecule might be of considerable value for discrimination between different types of pathology at the level of immune effector function. The increased levels of TNF-a and IL-6 signified a group of fertile women with endometriosis; however only IL-6 presented a discriminating value in multifunctional analysis of examined subgroups. The analysed range of factors had a greater tendency to discriminate between infertility status rather than endometriosis. # 2005 Elsevier Ireland Ltd. All rights reserved. Keywords: Cytokines; Adhesion molecules; Endometriosis; Infertility
1. Introduction Endometriosis is considered to be one of the most common gynaecological disorders and has been associated with infertility [1]. * Corresponding author. Tel.: +48 61 8233 155; fax: +48 61 8233 235.
The relationship between minimal and medium stages of endometriosis with infertility is unclear although quite frequent observations of endometrial foci in women primarily diagnosed for persistent infertility may point out such an interdependence. In which way endometriosis may deteriorate the environment for conception has never been satisfactorily explained. Among the possible causes of
0301-2115/$ – see front matter # 2005 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2004.11.044
96
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
infertility in women with endometriosis, we should mention:
2. Material and methods 2.1. Patients
- endocrine and immune disorders as well as; - altered peritoneal fluid (PF) environment. Clearly the changes in the peritoneal fluid content, which are well demonstrated in women with endometriosis [2,3], may disturb a fertilization process on different levels. It is accepted that the presence of ectopic endometrium may be an inducing factor for inflammatory properties of peritoneal fluid. This, in turn, can be manifested by the evident and well established: (a) increase in PF volume, (b) elevated numbers and activity of macrophages, (c) elevated cytokine, prostanoid, adhesion molecules contents [4,5]. It should be emphasized that a significant amount of work on cytokine content in peritoneal fluid of women with endometriosis is what underlines the importance of these biological factors in perpetuation of the disease. Cytokines released by macrophages as IL-1b [6,7], IL-6 [8–10], TNF-a [11–14] have been extensively studied and found to be elevated when compared to peritoneal fluid of healthy women and seem to drive the inflammatory reaction, together with IL-8, in this topographical region [12,13,15,16]. Soluble intercellular adhesion molecule-1 is essential for various immunological reactions and may be a useful parameter in the diagnosis of many immune diseases. In endometriosis, sICAM-1 may allow implantation of endometrial cells in ectopic sites [5]. sICAM-1 levels in serum and peritoneal fluid of patients with endometriosis were elevated [17–19] and were released in high amounts by isolated endometriotic stromal cells [20]. However, the role of sICAM-1 as an indicator for the diagnosis of endometriosis and/or infertility has not been thoroughly investigated [17,21]. A link between infertility and endometriosis of minimal or medium stage is still not well explained. Further to that controversy [22], it was shown that fertility in non-human primates is comparable either during induced or without endometriosis. The rate of implantations after IVF (in women with or without endometriosis) has also been previously shown to be similar [23]. In the present study, we have therefore selected several subgroups of women: (a) infertile with endometriosis; (b) fertile with endometriosis; (c) infertile (no endometriosis); (d) fertile, control population (no endometriosis), screening them for a range of cytokines (IL-1b, TNF-a, IL-6, IL-8) and sICAM-1 molecule as well as defining the macrophage numbers in PF samples. A second goal was an attempt to assess the predictive values for all these parameters in order to discriminate between the different subgroups under study. For this reason, the levels of measured substances were statistically elaborated and the discriminant function analyses were performed in respect to each of the selected subgroups or dividing the studied populations according to their infertility status.
Seventy eight women of reproductive age (21–41) from the Department and Clinic of Reproduction of the Karol Marcinkowski University of Medical Sciences in Poznan´ were subjected to laparoscopy due to unexplained infertility (with normal parameters of ovulation, hormonal profile and patent fallopian tubes), pelvic pain or suspected endometriosis. Individuals with history of autoimmune reactions, neoplastic disorders or microbial infections were excluded from this study. All studied subjects had regular menstrual cycles (26–30 days) and were not taking any hormonal preparations, e.g. Danazol, gonadotropin-releasing hormone agonists or contraceptives for at least 3 months prior to laparoscopy. At the time of laparoscopy, performed in the luteal phase of sexual cycle, peritoneal fluid samples were collected from 78 individuals, out of whom 48 had an evidence of minimal endometriosis, while the remaining 30 were considered as individuals with no endometriotic symptoms. A stage of endometriosis was classified according to the Revised American Fertility Society classification [24]. The patient cohorts were divided into four following subgroups: 1. 2. 3. 4.
Endometriosis with associated infertility (n = 28); Endometriosis without infertility (n = 20); Unexplained infertility with no endometriosis (n = 20); Control group (n = 10).
All studied subjects were under 41 years of age. They had a mean (S.D.) age of: 27.6 1.2 years in the first group, 26.5 0.9 in the second group, 28.1 0.1 in the third group and 27.9 0.8 in the fourth group. Out of the 78 women included in this study, 48 (groups 1 and 3) were considered to be infertile (28 had minimal endometriosis and 20 had no laparoscopic evidence of endometriosis) and 30 (groups 2 and 4) were fertile (live birth during last 12 months). The indication for laparoscopy in the group of fertile women was pelvic pain syndrome. The minimal endometriosis was diagnosed at the time of laparoscopy in 20 fertile women (group 2). Thus, these women did not exhibit previous endometriotic history or infertility. In 10 women used as healthy controls, laparoscopic examination demonstrated normal status of pelvic organs. 2.2. Peritoneal fluid samples Peritoneal fluid was aspirated during laparoscopy having the patient in the Trendelenburg position. Samples were obtained via laparoscopic cannula introduced to the cul-desac and aspirated to the syringe before any further medical procedures were undertaken. The probe was introduced by
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
trocar, which was inserted over the pubical symphysis in the midline of the body. After macroscopic inspection (volume, clarity, colour), the obtained samples were centrifuged at 650 g for 10 min. Supernatant was divided into 200 ml aliquots and stored at 70 8C until use. Each sample was subjected to bacteriological analysis and the contaminated ones (12%) were excluded from this study.
97
test. A p-value of less than 0.05 was considered to be statistically significant. Regression analysis between studied parameters in each subgroup or total investigated population was performed as well as standard and forward discriminant function analyses [26].
3. Results 2.3. Bacteriological analysis 3.1. Macrophages Bacteriological analysis was performed using a solid phase quantitative method of assessment. Plates were seeded in aerobic and anaerobic conditions. After incubation in 37 8C, the number of microorganisms in 1 ml of sample was estimated by using Bio Merieux tests as following: (1) (2) (3) (4) (5) (6)
Staphylococci: ID32 Staph test; Streptococci: API20 Strep test; Bacteriaceae Gram-negative: K32E test; Corynebacteriaceae: API Coryne test; Anaerobic Bacteriaceae: ID 32C test; Mycoplasmaceae: Mycoplasmic IST test—allowing to define the species, their concentration as well as sensitivity to antibodies.
2.4. Measurement of cytokines and sICAM-1 by ELISA (enzyme-linked immunosorbent assay) The concentrations of cytokines, i.e. IL-1b, TNF-a, IL-6, IL-8 were measured using commercially available ELISA kits (Predicta-Genzyme Diagnostics, Germany) according to the instruction of manufacturer. Similar assays were performed to measure sICAM-1 levels. The concentrations of cytokines were expressed in pg/ml and for sICAM-1 at ng/ml, and the interassay and intraassay variation coefficients were <5 and <3%, respectively. Cytokine and sICAM-1 measurements for each patient were performed in duplicates. 2.5. Macrophage identification Macrophages were identified by May-Grunwald-Giemsa staining and reaction with non-specific esterase [25], and their concentrations were appropriately calculated (number of cells/ml of peritoneal fluid). Cells were counted in a Thoma hemocytometer. Cell viability test by 1% trypan blue exclusion was performed on the cells resuspended after the first spin and revealed 95% of viability in the cell processed samples.
Analysing the macrophage concentrations per milliliter of peritoneal fluid, we have found statistically significant differences ( p < 0.05) between subgroups of infertile women with endometriosis versus infertility alone (Fig. 1a). The highest macrophage numbers were observed in the studied subgroups with endometriosis. 3.2. Cytokine concentrations 3.2.1. Interleukin-1b The levels of this cytokine, on average, were relatively low in each of the examined subgroups and were not found to be statistically significant. Both those women with infertility and endometriosis as well as the infertile women exhibited median values slightly greater than 3 pg/ml (Fig. 1b). 3.2.2. Tumor necrosis factor (TNF-a) The highest and most significant levels of this cytokine were found in a group of fertile women with endometriosis, which statistically differentiated this subgroup from a group of women with endometriosis and infertility (Fig. 1c). The median values for all tested subgroups were 0 pg/ml and control subjects did not exhibit TNF-a in a detectable range. 3.2.3. Interleukin-6 The highest and statistically significant level of IL-6 (110.0 pg/ml) was found in a group of fertile women with endometriosis, thus statistically signifying this subgroup from the population of infertile women with endometriosis and women with unexplained infertility (Fig. 1d). 3.2.4. Interleukin-8 Although the highest level of this interleukin has been identified in a group of infertile women with endometriosis (16.0 pg/ml), interleukin-8 was found to be statistically significant for infertile women (Fig. 1e). 3.3. Soluble intercellular adhesion molecule (sICAM-1)
2.6. Statistical analysis The results are given as median with average deviation. Statistical analysis was performed using Kruskal–Wallis non-parametric ANOVA test. Differences between analysed subgroups were compared by Dunn Multiple Comparison
Interestingly, both infertile subgroups (with or without endometriosis) exhibited here almost equal levels of this soluble adhesion molecule, which differentiated them (in statistically significant way) from fertile women with endometriosis or fertile controls (Fig. 1f).
98
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
Fig. 1. (a) Macrophage concentration (thousands of cells/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. C; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. (b) IL-1b concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Kruskal–Wallis nonparamtric ANOVA test indicated the obtained data as insignificant. Percentage of data obtained below the sensitivity limit: A:35%; B:45%; C:30%; D:33%. (c) TNF-a concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. B; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
99
Fig. 1. (Continued ).
the sensitivity limit: A:93%; B:55%; C:30%; D:100%. (d) IL-6 concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. B; B vs. C; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:60%; B:30%; C:70%; D:70%. (e) IL-8 concentration (pg/ml) in peritoneal fluid samples from infertile and fertile women with or without endometriosis A vs. C; p < 0.01. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis nonparametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:18%; B:35%; C:65%; D:50%. (f) sICAM1 concentration (ng/ml) in peritoneal fluid samples from infertile or fertile women with and without endometriosis A vs. B and B vs. C; p < 0.01. C vs. D; p < 0.05. (A) Women with infertility and endometriosis; (B) fertile women with endometriosis; (C) women with infertility without endometriosis; (D) fertile women without endometriosis. Data were evaluated by Kruskal–Wallis non-parametric ANOVA test with Dunn’s multiple comparisons. Percentage of data obtained below the sensitivity limit: A:18%; B:45%; C:0%; D:50%.
100
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
Fig. 2. (a) Negative correlation between macrophages and sICAM-1 in a group of fertile women with endometriosis. (b) Positive correlation between macrophages and sICAM-1 in a group of infertile women with endometriosis (p = 0.0055).
3.4. Statistical analysis When calculating the regression curves, most strikingly the correlations were observed, between concentrations of macrophages and sICAM-1 (Fig. 2a and b). In this pair of variables (Fig. 2a) a negative correlation was observed between the concentration of macrophages and sICAM-1 in population of fertile women with endometriosis, however a positive correlation between the same pair of variables was found in a group of infertile women with endometriosis (Fig. 2b). It became clear that infertility status and not endometriosis was a critical factor for the mode of observed relationship. Among the six analysed parameters (macrophages, IL1b, TNF-a, IL-6, IL-8, sICAM-1) in the standard discriminant function analysis, the four variables (Fig. 3; sICAM-1, IL-6, macrophages, IL-8 in order of the probability values) appear to account for most of the variation in the dataset. This analysis provided reasonable significant probabilities to classify correctly at least three (out of four) selected to our study subgroups (i.e. with the
exception for a group of women with unexplained infertility (see Table 1). However, performing the forward stepwise discriminant analysis and subdividing the whole studied population according to its infertility status, we ended up with only three statistically significant (for classification) variables, i.e. macrophages, IL-8 and IL-6 (in order of the probability values). In such evaluation, we have obtained the striking 92% probability value of correct classification into infertile population (Table 2). It has to be noted that sICAM-1 was not included here ( p > 0.05) in spite of its highest statistical significance for the standard discriminant analysis ( p > 0.000009).
4. Discussion In the present study, we have evaluated several biological parameters of peritoneal fluid, as macrophage numbers, cytokine contents (IL-1b, TNF-a, IL-6, IL-8) and sICAM-1 levels. For a majority of the measured biological parameters
Fig. 3. Standard discriminant analysis for correct classification of four studied subgroups by using two variables (roots) with highest significance for discrimination (sICAM-1 and IL-6).
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
101
Table 1 Standard discriminant function analysis for correct classification of studied subgroups by using sICAM-I, IL-6, MØ and IL-8a Statistical discriminant analysis
Classification matrix (standard method) (rows: observed classifications columns: predicted classifications)
Group
Percent correct
G 1b (p = 0.35897)
G 2c (p = 0.25641)
G G G G
1 2 3 4
67.85714 65.00000 40.00000 60.00000
19 4 11 1
2 13 1 1
3 1 8 2
4 2 0 6
Total
58.97436
35
17
14
12
a b c d e
G 3d (p = 0.25641)
G 4e (p = 0.12821)
Rank of probability values used for standard discriminant function analysis: sICAM-1, p < 0.000009; MØ, p < 0.018; IL-6, p < 0.000095; IL-8, p < 0.05. Infertile with endometriosis. Fertile with endometriosis. Infertile (no endometriosis). Fertile (no endometriosis) controls.
Table 2 Classification of correct discrimination (forward analysis) between fertile and infertile populations by using MØ, IL-8 and IL-6 valuesa Statistical discriminant analysis
Classification matrix (standard method) (rows: observed classifications columns: predicted classifications)
Group
Percent correct
Fertile populations (p = 0.30189)
Infertile populations (p = 0.69811)
Fertile population Infertile population
50.00000 91.66666
15 4
15 44
Total
75.64103
19
59
a
Rank of probability values used for forward discriminat functional analysis: MØ, p < 0.0009; IL-8, p < 0.012; IL-6, p < 0.05.
in this study (macrophage concentration, TNF-a, IL-6; Fig. 1a, c and d), the obtained results supported previous observations indicating the elevated macrophage numbers and cytokine contents in peritoneal fluid of women with endometriosis as comparing to the control populations [2– 5,10,27]. However, in presented study, we have observed a significant difference between the concentration of macrophages in a group with minimal endometriosis and infertility versus infertile women without endometriosis (Fig. 1a, p < 0.05), while the increased levels of TNF-a and IL-6 signified fertile women with endometriosis (Fig. 1c and d). The more controversial situation has been observed for IL1b (Fig. 1b) since some authors reported elevated levels of this cytokine in peritoneal fluid of women with endometriosis [2,6,7,11]. A majority of these reports, however, can only really be accounted to one group of researchers. Additionally, in our previous observations, we have shown quite substantial reservation to the significance of IL-1 in endometriotic women, not excluding its pro-inflammatory action, albeit not just only to the endometriosis [28,29]. The observed high level of IL-8 in a group of infertile women with endometriosis (Fig. 1e) supported the previous reports on elevated (statistically significant) IL-8 levels in peritoneal fluid of women with endometriosis [11–13,16,30]. In the group of fertile women with endometriosis we did not observe similar high levels of IL-8 (Fig. 1e). The most striking observation revealed in our study, however, was the high and statistically significant level of sICAM-1 (Fig. 1f). ICAM-1 is one of the most intriguing
molecules of controversial significance. Undoubtedly, cellular ICAM-1 is involved in a range of biological actions, among the others helping at inter-cellular cooperation and antigenic presentation [31]. It can be at least dual explanation of elevated sICAM-1 values, only depending on its immune properties. Firstly, sICAM-1 molecule can be shed from the activated cells involved in immune reactions thus manifesting an extensive antigenic presentation in the system, e.g. increased mucosal concentrations of s-ICAM-1 were reported during intestinal inflammation [32]. Moreover, in inflammatory bowel disease, Crohn disease or ulcerative colitis [33,34] both sICAM-1 and IL-8 were concomitantly elevated, exactly as in a case in our observations of infertile women (Fig. 1e and f) To support a hypothesis of enhanced antigenic presentation during minimal endometriosis, we may further quote a positive correlation between peritoneal macrophages and sICAM-1 (Fig. 2b) found in a group of infertile women with endometriosis. It is intriguing, however, whether the second possibility could be also potentially valid, i.e. whether the shedded sICAM-1 can be an inhibitory molecule (negative feedback) suppressing the antigenic presentation but exhibiting a side effect resulting in diminution of cytotoxic effector response. It could be true that in such circumstances the negative sICAM-1 correlation with macrophages (Fig. 2a) in fertile women with minimal endometriosis could be a critical factor for perpetuation of this disease. The presence of IL-1b and TNF-a (in peritoneal fluid) can be a sign of immunoinflammatory reactions in the
102
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103
pelvis, and not necessarily bound to endometriosis alone, hence they may be equally important in acute inflammatory reactions that can be involved at initiation of infertility (development of antigamete antibodies?). Such was a case in our previous observations [35] where antizona antibodies in follicular fluid were concomitantly identified with either IL1b or TNF-a. Another interesting conclusion stems from our calculations by using the discriminant function analysis tests. On the contrary to the moderate classification values obtained in the standard method of discrimination (not exceeding the 68% for correct classification of subgroups; see Table 1), the forward one brought a striking 92% correct classification to the group of infertile women when using obtained values of macrophages, IL-8 and IL-6 (Table 2). Interestingly, sICAM-1 which was of the highest probability value in a standard discriminant analysis (Fig. 3) to differentiate between the subgroup of infertile women with endometriosis from any other studied populations (Table 1) was totally left out in forward (stepwise) analysis for infertility discrimination (Table 2) due to its lack of significance in this context ( p > 0.05). Assuming that sICAM-1 seems to be of considerable value for flagging the disturbances of immune effector function, it should be further emphasized that elevated levels of soluble ICAM-1 molecule can compete and be a reason for disruption between cellular ICAM-1 and LFA-1 binding. It may thus weaken the antigenic presentation (10-fold) as well as the effective cytotoxic response [36,37]. On the other hand, it was previously reported that sICAM-1 levels may also be correlated with their release by granulosa luteal cells thus being a potential sign of ovarian function (or dysfunction) [38,39]. This would interestingly correspond to misleading sICAM-1 significance as for the altered immune local response. Overall, such overlapping biological phenomena may hinder interpretation of the obtained multifactorial analysis where sICAM-1 reached a high statistical significance in the standard discrimination ( p < 0.000009) and a lack of significance in the progressive one ( p > 0.05). Overall, our data point out an easier discrimination between the infertility status of the examined subjects than in respect to minimal endometriosis while taking into account selected to this study biological parameters. References [1] Strathy J, Molgaard C, Coulam C, Melton LJ. Endometriosis and infertility: a laparoscopic study of endometriosis among fertile and infertile women. Fertil Steril 1982;38:667–72. [2] Ho HN, Wu MY, Yang YS. Peritoneal cellular immunity and endometriosis. Am J Reprod Immunol 1997;38:400–12. [3] Lebovic DI, Mueller MD, Taylor RN. Immunobiology of endometriosis. Fertil Steril 2001;75:1–10. [4] Rier SE, Yeaman GR. Immune aspects of endometriosis: relevance of the uterine mucosal immune system. Semin Reprod Endocrinol 1997;15:209–20.
[5] Harada T, Iwabe T, Terakawa N. Role of cytokines in endometriosis. Fertil Steril 2001;76:1–10. [6] Fakih H, Baggett B, Holtz G, Tsang KY, Lee JC, Williamson HO. Interleukin-1: a possible role in the infertility associated with endometriosis. Fertil Steril 1987;47:213–7. [7] Kuo TM, Taketani Y, Shimizu T, Mizuno M. A possible mechanism of infertility in endometriosis – a role of interleukin 1 in peritoneal fluid. Asisst Reprod Technol/Androl 1990;1:24–31. [8] Rier SE, Parsons AK, Becker JL. Altered interleukin-6 production by peritoneal leukocytes from patients with endometriosis. Fertil Steril 1994;61:294–9. [9] Koyama N, Masuura K, Okamura H. Cytokines in the peritoneal fluid of patients with endometriosis. Int J Gynecol Obstet 1993;43:45–50. [10] Mahenke JL, Dawood Y, Huang JC. Vascular endothelial growth factor and interleukin-6 in peritoneal fluid of women with endometriosis. Fertil Steril 2002;73:166–70. [11] Rana N, Braun DP, House R, Gebel H, Rotman C, Dmowski WP. Basal and stimulated secretion of cytokines by peritoneal macrophages in women with endometriosis. Fertil Steril 1996;65:925–30. [12] Pizzo A, Salmeri FM, Ardita FV, Sofo V, Tripepi M, Marsico S. Behaviour of cytokine levels in serum and peritoneal fluid of women with endometriosis. Gynecol Obstet Invest 2002;54:82–7. [13] Harada T, Enatsu A, Mitsunari M, Nagano Y, Ito M, Tsudo T, et al. Role of cytokines in progression of endometriosis. Gynecol Obstet Invest 1999;47(Suppl. 1):34–9. [14] Bedaiwy MA, Falcone T, Sharma RK, Goldberg JM, Attaran M, Nelson DR, et al. Prediction of endometriosis with serum and peritoneal fluid markers: a prospective controlled trial. Hum Reprod 2002;17:426–31. [15] Ryan IP, Tseng JF, Schirock ED, Khorram O, Landers DV, Taylor RN. Interleukin-8 concentrations are elevated in peritoneal fluid of women with endometriosis. Fertil Steril 1995;63:929–32. [16] Iwabe T, Harada T, Tsudo T, Tanikawa M, Onohara Y, Terakawa N. Pathogenetic significance of increased levels of interleukin-8 in the peritoneal fluid of patients with endometriosis. Fertil Steril 1998;69: 924–30. [17] Somigliana E, Vigano P, Candiani M, Felicetta I, Di Blasio AM, Vignali M. Use of serum-soluble intercellular adhesion molecule-1 as a new marker of endometriosis. Fertil Steril 2002;77:1028–31. [18] Daniel Y, Geva E, Amit A, Eshed-Englender T, Baram A, Fait G, et al. Do soluble cell adhesion molecules play a role in endometriosis? Am J Reprod Immunol 2000;43:160–6. [19] Calhaz-Jorge C, Costa AP, Santos MC, Palma-Carlos ML. Soluble intercellular adhesion molecule 1 in the peritoneal fluid of patients with endometriosis correlates with the extension of peritoneal implants. Eur J Obstet Gynecol Reprod Biol 2003;106:170–4. [20] Vigano P, Gaffuri B, Somigliana E, Busacca M, Di Blasio AM, Vignali M. Expression of intercellular adhesion molecule (ICAM)-1 mRNA and protein is enhanced in endometriosis versus endometrial stromal cells in culture. Mol Hum Reprod 1998;4:1150–6. [21] Hammadeh ME, Fischer-Hammadeh C, Hoffmeister H, Huebner U, Georg T, Rosenbaum P, et al. Fibroblast growth factor (FGF), intracellular adhesion molecule (sICAM-1) level in serum and follicular fluid of infertile women with polycystic ovarian syndrome, endometriosis and tubal damage, and their effect on ICSI outcome. Am J Reprod Immunol 2003;50:124–30. [22] D’Hooghe TM, Bambra CS, Koninckx PR. Cycle fecundity in baboons of proven fertility with minimal endometriosis. Gynecol Obstet Invest 1994;37:63–5. [23] Badawy SZ, Cuenca V, Stitzel A, Jacobs RD, Tomar RH. Autoimmune phenomena in infertile patients with endometriosis. Obstet Gynecol 1984;63:271–5. [24] The American Fertility Society. Revised American Fertility Society classification of endometriosis. Fertil Steril 1985; 43: 351–2. [25] Hudson L, Hay FC, editors. Practical immunology. Oxford: Blackwell Scientific Publications, 1991. [26] Albert A. Discriminant analysis and medical diagnosis. Biocybernet Biomed Eng 1995;15:139–56.
J. Skrzypczak et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 122 (2005) 95–103 [27] Witz CA, Schenken RS. Pathogenesis. Semin Reprod Endokrinol 1997;15:199–208. [28] Skrzypczak J, Je˛drzejczak P, Kasprzak M, Puk E, Kurpisz M. Inflammatory cytokines in peritoneal fluid of women with endometriosis. Ann NY Acad Sci 1995;762:496–8. [29] Skrzypczak J, Puk E, Szymanowski K, Je˛drzejczak P, Kurpisz M. Peritoneal fluid cytokines (TNFa, IL-6 IL-1) in women with endometriosis and infertility. Arch Perinat Med 1996;1:77–81. [30] Arici A, Tazuke SI, Attar E, Kliman HJ, Olive DL. Interleukin-8 concentration in peritoneal fluid of patients with endometriosis and modulation of interleukin-8 expression in human mesothelial cells. Mol Hum Reprod 1996;2:40–5. [31] van de Stolpe A, van der Saag PT. Intercellular adhesion molecule-1. J Mol Med 1996;74:13–33. [32] Nielsen OH, Brynskov J, Vainer B. Increased mucosal concentrations of soluble intercellular adhesion molecule-1 (sICAM), sE-selectin and interleukin-8 in active ulcerative colitis. Dig Dis Sci 1996;41: 1780–5. [33] Brynskov J, Nielsen OH, Ahnfelt-Ronne I, Bendtzen K. Cytokines (immunoinflammatory hormones) and their natural regulation of inflammatory bowel disease (Crohn’ disease and ulcerative colitis): a review. Dig Dis 1994;12:290–304.
103
[34] Pedersen G, Brynskov J, Nielsen OH, Bendtzen K. Adhesion molecules in inflammatory and neoplastic intestinal diseases. Dig Dis 1995;13:322–36. [35] Skrzypczak J, Puk E, Szczepan´ska M. A possible effect of selected folicular fluid factors on fertilizing ability of the oocyte. Centr Eur J Immunol 1997;22:104–9. [36] Meyer DM, Dustin ML, Carron CP. Characterization of intercellular adhesion molecule-1 ectodomain (sICAM) as an inhibitor of lymphocyte function-associated molecule-1 interaction with ICAM-1. J Immunol 1995;155:3578–84. [37] Woska Jr JR, Morelock MM, Jeanfavre DD, Bormann BJ. Characterization of molecular interactions between intercellular adhesion molecule-1 and leukocyte function-associated antigen-1. J Immunol 1996;156:4680–5. [38] Vigano P, Gaffuri B, Ragni G, Di Blasio AM, Vignali M. Intercellular adhesion molecule-1 is expressed on human granulosa cells and mediates their binding to lymphoid cells. J Clin Endocrinol Met 1997;82:101–5. [39] Vigano P, Fusi F, Gaffuri B, Bonzi V, Ferrari A, Vignali M. Soluble intercellular adhesion molecule-1 in ovarian follicles: production by granulosa luteal cells and levels in follicular fluid. Fertil Steril 1998;69:774–9.