- Email: [email protected]
kisspeptin-1 in polycystic ovary syndrome: a systematic review and meta-analysis with diagnostic test accuracy
685
RBMO VOLUME 39 ISSUE 4 2019
ARTICLE
Circulatory metastin/kisspeptin-1 in polycystic ovary syndrome: a systematic review and meta-analysis with diagnostic test accuracy BIOGRAPHY
Varikasuvu SR obtained his Doctorate in Medical Biochemistry from SVIMS University, India. He is now a member of Biochemistry Department at AIIMS, with research interests in disease biochemistry, redox homeostasis, endocrinology and diabetes, proteomics and bioinformatics, systematic reviews and meta-analyses. He is an expert editorial reviewer for specialty journals by ELSEVIER, Springer, Karger, WILEY, SAGE and Portland press. Seshadri Reddy Varikasuvu1,*, V Satya Prasad2, VC Vamshika3, MV Satyanarayana3, Jaipal Reddy Panga4 KEY MESSAGE
Circulating kisspeptin-1 (kiss-1) peptide levels in women with polycystic ovary syndrome (PCOS) were higher compared with healthy controls and showed significant associations with anti-Müllerian hormone and dehydroepiandrosterone levels. Overall diagnostic accuracy of KISS-1 for PCOS was good (area under the curve = 0.835), with a promising diagnostic odds ratio (13.71). ABSTRACT
Research question: A close association between Kisspeptin-1 (KISS-1) and reproductive physiology has been reported, but the results on circulatory KISS-1 are ambiguous in patients with polycystic ovary syndrome (PCOS). A systematic review and meta-analysis were conducted to evaluate the association between KISS-1 and PCOS, and to test its diagnostic test accuracy (DTA) through DTA meta-analysis. Design: Relevant studies were identified by searching PubMed and other databases in addition to manual searching of cross-references. Random-effects model was used to obtain standardized mean differences (SMD), pooled correlation coefficients and summary of DTA. Meta-regression and sub-group analyses were conducted to explore heterogeneity. The presence of publication bias was tested using funnel plot analysis. Results: This meta-analysis finally included 12 studies. Compared with controls, women with PCOS showed significantly increased circulatory KISS-1 levels (SMD = 0.47; P = 0.002). Meta-analysis of correlations showed positive associations between KISS-1 and anti-Müllerian hormone (AMH) (P = 0.03), testosterone (P < 0.001) and dehydroepiandrosterone (P = 0.004). The pooled diagnostic odds ratio and area under curve were 13.71 and 0.835, respectively. A one-study leave-out sensitivity analysis indicated that no single study had a significant influence on the overall outcome, suggesting the robustness of this meta-analysis. Conclusions: This meta-analysis showed significantly increased KISS-1 level in PCOS, and its association with AMH reflects its role in reproductive physiology. In our DTA meta-analysis, KISS-1 showed good accuracy for PCOS detection. Further large-scale studies are required to establish its validity. 1 Department
of Biochemistry, All India Institute of Medical Sciences (AIIMS), Bibinagar Telangana 508126, India of Anatomy, Maheshwara Medical College and Hospital, Hyderabad Telangana 502307, India of Obstetrics and Gynecology (MBBS Students), Maheshwara Medical College and Hospital, Hyderabad Telangana 502307, India 4 Institute for Systems Biology, Seattle WA, USA 2 Department
3 Department
© 2019 Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. *Corresponding author. E-mail address: [email protected] (S R Varikasuvu). https://doi.org/10.1016/j. rbmo.2019.04.018 1472-6483/© 2019 Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. Declaration: The authors report no financial or commercial conflicts of interest.
KEYWORDS
Kisspeptin-1 Meta-analysis Metastin Polycystic ovary syndrome
686
RBMO VOLUME 39 ISSUE 4 2019
INTRODUCTION
P
olycystic ovary syndrome (PCOS) is a common endocrinological disorder characterized by reproductive, endocrine and metabolic disturbances. It is the leading cause of female infertility, affecting 6–10% of reproductive-aged women (March et al., 2010). The clinical and biochemical signs of PCOS include polycystic ovaries, anovulation and hyperandrogenism (Rosenfield and Ehrmann, 2016). Although the aetiopathogenesis of PCOS remains unclear, the hypothalamic–pituitary gonadal (HPG) axis has been proposed to be involved, with observed disturbances in gonadotrophin secretion, increased LH levels and perturbed LH and FSH ratios (Panidis et al., 2005). Kisspeptin-1 (KISS-1), also known as metastin, is a 54-aminoacid peptide encoded by the Kiss-1 gene, and was first identified from the human placenta (Hori et al., 2001; Ohtaki et al., 2001). KISS-1 has been proposed to be important in the initiation of gonadotrophin-releasing hormone (GnRH) secretion from the hypothalamus at puberty and is also known to regulate LH secretion during the promotion of ovulation (Matsui et al., 2004; Navarro et al., 2005). The association between KISS-1 and PCOS, however, remains poorly understood. Available data on circulatory KISS-1 levels in PCOS have also been inconsistent (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Albalawi et al., 2018; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019). Furthermore, no previous meta-analysis has been conducted on the association between KISS-1 levels in patients with PCOS. Considering the important relationship between HPG axis and KISS-1, the aim of the present study was to conduct a systematic review and meta-analysis of circulatory KISS-1 levels and their association with biochemical, metabolic and hormonal parameters in PCOS. Diagnostic test accuracy (DTA) metaanalysis was also conducted for the utility of circulatory KISS-1 in PCOS.
MATERIALS AND METHODS The study was registered with the International Prospective Register of
Systematic Reviews (PROSPERO), number CRD42018112064. The criteria of preferred reporting items for systematic reviews and meta-analysis (PRISMA) were followed in conducting and reporting this systematic review and meta-analysis. Search strategy The NCBI PubMed database was searched using MESH and text word search strategies. The following search string was developed: (‘kisspeptins’[MeSH Terms] OR ‘kisspeptins’[All Fields] OR ‘kisspeptin’[All Fields]) OR (Kiss-1[All Fields] AND (‘peptides’[MeSH Terms] OR ‘peptides’[All Fields] OR ‘peptide’[All Fields])) OR (‘kisspeptins’[MeSH Terms] OR ‘kisspeptins’[All Fields] OR ‘metastin’[All Fields]) AND PCOS[All Fields]). The Web of Science, Google Scholar, Embase, Cochrane Library, Springer's AuthorMapper and Science Direct databases were also searched for articles that reported circulatory KISS-1 levels in women with PCOS. Furthermore, bibliographies of published articles were manually hand searched to identify additional studies. Two authors independently conducted literature searches, and any discrepancies were resolved with discussion. Efforts were made to obtain unpublished data, if any. When required, the corresponding authors of respective articles were E-mailed to obtain clarification. The literature search was first conducted on 17 November 2018, and all searches were conducted before 1 January 2019, with no timespan specified, with and without humans set as a limit. Study selection criteria The individual studies had to meet the following criteria to be included in this meta-analysis: peer-reviewed publications with case–control or observational study design; original data; study participants were human with description of PCOS and controls; studies in which circulating levels of KISS-1 in PCOS were compared with controls; description for KISS-1 estimation method provided; articles written in English with clear data; and studies diagnosing PCOS based on the revised 2003 Rotterdam criteria (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group, 2004). The exclusion criteria were as follows: studies with no control group; studies reported on diseases other than PCOS; KISS-1 levels were only shown by pictograms with no data; and reviews, commentaries and Letter to the
Editor article types. If the same group of authors published duplicate articles on the same study population, only a recently published article or article with all relevant information was included. Data extraction and quality assessment The following information was extracted from each of the included studies; first author names, country and year of publication, number of participants with PCOS and control participants, PCOS diagnosis criteria, means and SD of age and anthropometric, metabolic, hormonal parameters, KISS-1 levels, with its measurement method and units, correlation analysis data, receiver operating characteristic analysis data (accuracy measures: sensitivity, specificity and area under the curve values) and other study characteristics. A quality score evaluation for studies was carried out according to Newcastle– Ottawa Scale for case–control and observational studies (Li et al., 2008). Quality assessment of study scores range from 0 to 9 points and included three components: selection, comparability and exposure. Quality of studies included in the DTA analysis was reported using ‘Quality assessment of diagnostic accuracy studies’ (QUADAS-2) tool (Whiting et al., 2011). Statistical analysis Meta-analyses were conducted if three or more studies reported the circulatory KISS-1 concentrations in women with PCOS patients compared with controls. A separate subgroup meta-analysis was conducted based on body mass index (BMI) to report its effect on the overall outcome. We calculated the standardized mean difference (SMD) and its 95% confidence interval (CI) as a summary statistic for the difference of KISS-1 level between women with PCOS and control groups; normal weight women with PCOS and obese women with PCOS subgroups. Furthermore, a meta-analysis of the correlation coefficients of KISS-1 was conducted with various parameters. The effect size for SMD and pooled correlation coefficient values were presented as a Z-score. The Z-score P < 0.05 was considered statistically significant. The between-study heterogeneity was examined by Cochrane's Q statistic and expressed as percentages of I2. P < 0.05 or chi-squared statistic of over 50%
indicated a significant heterogeneity. A random-effects model was used to compute SMD. A meta-regression analysis was conducted to identify the potential source of heterogeneity using anthropometric, hormonal, metabolic and lipid parameters. Risk of publication bias was studied by funnel plot asymmetry with Begg's and Egger's tests. To avoid risk of bias, however, general search terms were chosen for a guaranteed retrieval for inclusion of articles reporting KISS-1 levels as a secondary outcome. In the case of P < 0.10, suggesting statistical significance, the ‘trim-and-fill method’ was used to correct that bias. To test the robustness of this meta-analysis, a one-study leaveout sensitivity analysis was conducted. All comparisons were two-tailed, and all analyses were conducted using the Review Manager Software version 5.3, which presents SMD as Hedges g; the difference between the two means divided by the pooled SD, with a correction for sample bias. The funnel plot for publication bias with Begg's and Egger's tests was conducted using Comprehensive meta-analysis, version 3. The MedCalc version 16.2.0 was used for the meta-analysis of correlations. In studies in which receiver operator characteristic (ROC) data were presented, the DTA meta-analysis was conducted to obtain summary ROC curve (SROC) with pooled sensitivity, specificity and diagnostic odds ratio (DOR) of KISS-1 for PCOS. Meta-Disc software, version1.4 was used to conduct the DTA analyses. The heterogeneity, caused by threshold and non-threshold effects, was assessed by Spearman correlation analysis and meta-regression, respectively. Moses linear model was used to obtain SROC curve and the pooled AUC value. the DerSimonian Laird method was used to obtain pooled sensitivity, specificity and DOR, with their respective 95% confidence intervals.
RESULTS Search results and study characteristics Initially, the literature search retrieved 2168 articles, of which, 2129 articles were excluded after reviewing for KISS-1 description in PCOS. After reviewing the full texts of the remaining 39 publications, 27 were excluded; by applying ‘humans’ filter (n = 20), gene expression/polymorphism studies (n = 3),
RBMO VOLUME 39 ISSUE 4 2019
reviews with no original data (n = 2), duplicates (n = 1) and no control group (n = 1). Finally, a total of 12 studies were included in this meta-analysis comparing circulatory KISS-1 levels between women with PCOS and control patients (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Albalawi et al., 2018; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019). Among these, four studies reported KISS-1 levels between patients with PCOS with normal weight and patients with PCOS who were overweight (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016). Correlations of KISS-1 have been reported with AMH (El-Shehawy and Safan, 2015; Ozay et al., 2016; Gorkem et al., 2017; Kaya et al., 2018), total-testosterone (Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017), LH (Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Ozay et al., 2016; Gorkem et al., 2017; Kaya et al., 2018), and other variables. In one study (El-Shehawy and Safan, 2015), the correlations between KISS-1 and AMH were reported separately in PCOS and control groups. Therefore, these correlation coefficients were considered as two observations. Three studies reported diagnostic accuracy measures and were included in the DTA metaanalysis (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018). The PRISMA flow diagram is presented in FIGURE 1 . The characteristics and the Newcastle– Ottawa Scale quality scores of included studies are presented in TABLE 1. With the obtained score range from 5 to 8, the overall quality of studies was medium to high. All of the included studies used Rotterdam criteria for PCOS diagnosis (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group, 2004). All of the included studies except two (Yilmaz et al., 2014; Albalawi et al., 2018) are age matched between PCOS and control groups. Matching of BMI was carried out in seven of the included studies (Chen et al., 2010; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Ozay et al., 2016; Nyagolova et al., 2016; Gorkem et al., 2017; Daghestani, 2018). The circulatory KISS-1 levels
687
were determined by enzyme-linked immunosorbent assay using serum in seven of the included studies (Yilmaz et al., 2014; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019), whereas the remaining five studies used plasma (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; El-Shehawy and Safan, 2015; Albalawi et al., 2018). Circulatory KISS-1 level in PCOS compared with controls As shown in FIGURE 2, women with PCOS had increased KISS-1 level compared with non-PCOS controls. With a significant between-study heterogeneity (I2 = 84%; P < 0.00001), the randomeffects model was applied to compute the pooled effect size. The pooled SMD and 95% confidence intervals were 0.47 (0.17 to 0.77). The overall effect size for SMD calculated as Z was 3.06 (P = 0.002). Sub-group analysis and metaregression A subgroup analyses of circulatory KISS-1 levels in PCOS compared with controls is shown in FIGURE 3A. The pooled SMD was not statistically significant in the subgroups of ‘age 24 years or younger’, ‘studies with no MBI-matching’, ‘obese BMI’, and ‘plasma’. This may suggest the possible source of heterogeneity among the included studies. The pooled SMDs were statistically significant in other subgroups of ‘age over 24 years’ (SMD = 0.42; Z = 3.41; P = 0.0006; I2 = 52%), ‘studies matched for BMI’ (SMD = 0.48; Z = 2.27; P = 0.02; I2 = 88%; P for I2 < 0.00001), ‘normoweight BMI’ (SMD = 0.93; Z = 2.08; P = 0.04; I2 = 88%; P for I2 = 0.0002), and ‘serum’ (SMD = 0.63; Z = 3.13; P = 0.002; I2 = 89%; P for I2 < 0.00001), indicating that circulatory KISS-1 levels were higher in women with PCOS compared with controls. When comparisons were made between obese women with PCOS and women with PCOS of normal weight for the KISS-1 level reported in four studies (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016), the statistical significance was not achieved (SMD = − 0.21; Z = 0.82; I2 = 77%; P = 0.004). The meta-regression analysis (TABLE 2) carried out with various covariates demonstrated that waist-to-hip
688
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 1 Study selection based on PRISMA criteria.
ratio (coefficient = 11.47; P = 0.03), free-testosterone (coefficient = 0.34; P = 0.001), AMH (coefficient = − 0.15; P = 0.04) and oestradiol (coefficient = − 0.01; P = 0.03) produced statistically significant regression coefficients, indicating their effect on the heterogeneity among included studies. Publication bias and sensitivity analysis No significant publication bias was detected among the included studies (FIGURE 4) according to the funnel plot asymmetry with Begg's correlation and Egger's regression tests. A one studyleave out sensitivity analysis showed that no single study had significantly influenced the combined overall outcome, and the combined SMDs obtained were stable and remained statistically significant, after omitting any particular study in sensitivity analysis (FIGURE 5). Two studies (Panidis et al., 2006; Yilmaz et al., 2014) have significantly influenced the overall result, whereas when the study by Panidis et al. (2006) was omitted,
the pooled SMD was increased from 0.47 (Z = 3.06; P = 0.002) to 0.56 (Z = 3.80; P = 0.0001) with only a 2% decrease in heterogeneity. When the study by Yilmaz et al. (2014) was omitted, the pooled SMD decreased from 0.47 (Z = 3.06; P = 0.002) to 0.36 (Z = 3.31; P = 0.0009), with decreased heterogeneity from 84% to 64%. Meta-analysis of correlation coefficients The results of correlation meta-analysis of KISS-1 with various anthropometric, metabolic and hormonal variables are presented in TABLE 3. The random-effects meta-analysis showed significant positive correlations between KISS-1 and AMH (pooled r = 0.25; P = 0.03), totaltestosterone (pooled r = 0.12; P < 0.001), DHEAS (pooled r = 0.13; P = 0.004), and statistically borderline positive correlations with LH (pooled r = 0.46; P = 0.05), free androgen index (pooled r = 0.25; P =0.06), and free testosterone (pooled r = 0.10; P = 0.06).
Diagnostic test accuracy meta-analysis Three studies (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018) reported the diagnostic accuracy measures of KISS-1 for PCOS. QUADAS-2 tool was used for quality assessment of these studies (TABLE 4). All the studies used enzymelinked immunosorbent assay for KISS-1 measurement, and the cut-off values used were not pre-specified. The results of DTA meta-analysis are shown in FIGURE 6. The pooled sensitivity was 0.67 (95% CI 0.58 to 0.75); specificity was 0.86 (95% CI 0.78 to 0.92), and the pooled DOR was 13.71 (95% CI 6.11 to 30.75) (FIGURE 6A–6c, respectively). The SROC curve for KISS-1, AUC and Q index are shown in FIGURE 6d. The pooled AUC obtained was 0.835 (Q* = 0.767). According to the threshold results, the b value was − 0.305, suggesting that the SROC curve was symmetric. The Spearman correlation coefficient obtained was 0.50, indicating no heterogeneity from the threshold effect.
689
RBMO VOLUME 39 ISSUE 4 2019
TABLE 1 CHARACTERISTICS OF INCLUDED STUDIES Study, year (country)
Study groups Control Age BMI
KISS-1
PCOS versus control groups
NOS
PCOS n
Age BMI
n
Method
Sample
Units
Matching/no- difference
Significant differences Age, BMI, LH
Albalawi et al., 26.7 2018 (Saudi Arabia)
36.6
30
29.4
46.0
28
ELISA
Plasma
pg/ml
FSH
Chen et al., 2010 18.6 (China)
20.11
20
17.8
21.4
19
ELISA
Plasma
fmol/ml
Age, BMI, FSH, T, SHBG, LH, T, FAI, DHEAS, INS, 2h-INS, FPG, 2h-GLU GLU/INS, HOMA
Daghestani, 2018 25.4 (Saudi Arabia)
22.1
44
24.0
22.8
44
ELISA
Serum
fmol/ml
Age, BMI, HC, FSH, leptin, ghrelin, vitamin D
LH, LH/FSH, E2, T, PG, SHBG, 7 WC, WHR, TC, TG, LDL, HDL, F-Ins, FPG, VEGF
El-Shehawy and Safan, 2015 (Kuwait)
27.5
22.1
18
24.4
23.4
23
ELISA
Plasma
pmol/l
Age, BMI, WC, FSH
LH, T, DHEAS, AMH
6
Gorkem et al., 2018 (Turkey)
27.9
26.0
57
27.7
26.3
60
ELISA
Serum
ng/ml
Age, BMI, WC, HC, WHR, FSH, ED, 7OHP
LH, T, DHEAS, AMH
7
Jeon et al., 2013 (Korea)
24.9
19.7
36
23.7
23.0
54
ELISA
Plasma
pmol/l
Age, FSH, SHBG, PROL LH, LH/FSH, BMI, WHR, GLU, INS, GLU/INS, HOMA, T, FAI, leptin, RBP4
7
Kaya et al., 2018 (Turkey)
30.2
26.4
27
28.7
28.6
29
ELISA
Serum
pg/ml
Age, LH, oestradiol, TSH, BMI, FSH, PROL, AFC AMH, gravidity, parity, infertility duration
6
Nyagolova et al., 26.6 2016 (Bulgaria)
24.8
42
24.9
25.5
87
ELISA
Serum
ng/ml
age, BMI, WC, FSH, PROL, TSH, FPG, INS
LH, LH/FSH, HOMA, T, SHBG, 8 FAI, GALP
Ozay et al., 2016 (Turkey)
24.4
23.4
150 23.9
24.3
250 ELISA?
Serum
ng/ml
age, BMI, BP, FAI, VFT, PROL, T, FPG, HDL, leptin, CRP, ET-1
LH, LH/FSH, FSH, WHR, INS, 8 HOMA, Ferriman–Galwey Score, CIMT, Hcy, T, DHEAS, SHBG, TC, TG, LDL, APO B/A1, AMH
Panidis et al., 2006 (Greece)
26.8
32.1
13
23.9
26.8
56
ELISA
Plasma
fmol/ml
Age, FSH, PROL, FPG
BMI, LH, T, FAI, 17OHP, SHBG, 7 INS, GLU/INS, HOMA
Umayal et al., 2019 (Sri Lanka)
33.8
25.2
110 24.6
26.8
55
ELISA
Serum
nmol/l
Age, WHR, FPG
T, BMI, MFG Score
6
Yilmaz et al., 2014 24.5 (Turkey)
23.1
66
24.5
83
ELISA
Serum
ng/ml
BMI, FSH, HOMA, TC, TG, LDL, HDL, E2, PROL, 17OHP, TSH
Age, LH, T, DHEAS, SHBG, FAI, Ferriman–Gallwey Score, FPG, INS
7
21.0
5 6
A1C, haemoglobin A1c; AFP, antral follicle count; AMH: anti-Müllerian hormone; APO A1, apoprotein A1; APO B. apoprotein B; BMI, body mass index; BP, mean arterial blood pressure; CRP, C-reactive protein; DHEAS, dehydroepiandrosterone; E2, oestradiol; ET-1, endothelin-1; ELISA, enzyme-linked immunosorbent assay; FPG, fasting plasma glucose; FAI, free androgenindex; GLU, glucose; GALP, galanin-like peptide; Hcy, homocysteine; HOMA, homeostatic model assessment-insulin resistance; INS, insulin; KISS-1, kiss 1 peptide/kisspeptin/metastin; LDL, low-density lipoproteins; n, number of controls/PCOS patients; NA/?, not available; NOS, Newcastle–Ottawa Scale; PCOS, polycystic ovary syndrome; PG, progesterone; PROL, prolactin; Rotterdam criteria, Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group-Revised 2003 consensus on diagnostic criteria and long-term health risks related to PCOS; RBP4, retinol-binding protein 4; SHBG, sex hormone binding globulin; T, testosterone-total/free; TC, total cholesterol; TG, triglycerides; TSH, thyroid stimulating hormone; VEGF, vascular endothelial growth factor; VFT, visceral fat thickness; WC, waist circumference; WHR, waist-hip ratio; 17OHP: 17-hydroxyprogesterone.
Furthermore, the DOR result suggested no heterogeneity from the non-threshold effect (chi-squared = 2.51).
DISCUSSION To the best of our knowledge, this is the first meta-analysis to compare KISS-1 levels in women with PCOS compared with control participants who did not have PCOS. Findings from this study indicate that PCOS is associated with a significant increase in circulatory KISS-1 levels compared with controls. The peptide KISS-1 has
been reported to play a critical role in the central control of hypothalamic– pituitary–gonadal axis and in the direct control of ovarian function, including follicular development, oocyte maturation, steroidogenesis, ovulation and metabolic control of fertility (Funes et al., 2003; Seminara et al., 2003; Navarro et al., 2004). KISS-1 functions through a G-protein-coupled receptor, kisspeptin receptor (KISS1R) to stimulate gonadotrophin releasing hormone release and subsequent secretion of FSH and LH in many mammals. Evidence also shows a direct
effect of KISS-1 on the gonadotrophs of the anterior pituitary gland, stimulating the release of LH and FSH (Hameed et al., 2006). Expression of both KISS-1 and KISS1R has been reported in many tissues, including liver, placenta, umbilical vein, pancreas, small intestine, central nervous system, aorta, adipose tissue, lymph nodes, peripheral blood lymphocytes, testis and female tract (Witchel and Tena-Sempere, 2013; Hu et al., 2018). Although evidence on the exact source of increased circulatory KISS-1 levels in PCOS is unclear, any of these tissues could be a source.
690
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 2 Circulatory kisspeptin-1 levels.
It has also been documented that locally produced kisspeptin/KISS1R directly participates in a series of physiological and pathological activities in the ovary. High plasma levels of kisspeptin have been reported in women with PCOS, reflective of its relationship to pathological conditions in the ovaries. Furthermore, animal models of PCOS have shown increased levels of KISS1 mRNA in peripheral tissues such as the ovaries and fat (Brown et al., 2012). Moreover, it has been reported that hypothalamic KISS-1 neurones relay gonadal steroid regulation on the HPG axis and hypothalamic KISS-1 mRNA levels are abnormal in women with PCOS. Dysregulation/alterations of the kisspeptin levels may affect the ovarian function, leading to generation of PCOS, reproductive pathology or female infertility (Witchel and Tena-Sempere, 2013; Nejad et al., 2017). Because of abnormal pulsatile GnRH secretion, increased LH pulsatility, decreased FSH secretion and perturbed LH–FSH ratios were reported in PCOS. As kisspeptin is known to be the upstream central controller for inducing GnRH and LH secretion, higher KISS-1 levels in PCOS were speculated. All these findings and the studies included in this meta-analysis support a potential link between PCOS and KISS-1 levels. In line with this, our meta-analysis demonstrated higher circulatory KISS-1 levels in women with PCOS. Our sub-group analysis revealed that the increase in circulatory KISS-1 level was more pronounced in women with PCOS (compared with controls) ‘aged over 24 years’, ‘BMI-matched studies’, ‘normo-weight BMI’, and in the subgroup of studies using ‘serum’ for KISS-1
estimation. The difference in KISS-1 level, however, did not reach statistical significance in the sub-groups of ‘age younger than 24 years’, ‘studies with no BMI-matching’, ‘obese-BMI’, and ‘plasma’ (FIGURE 3a). Also, when KISS-1 levels were compared between obese women with PCOS and normal weight women with PCOS, no significant difference was reported (FIGURE 3b). This indicates that the increase in KISS-1 level in PCOS is independent of BMI. In a recent study (Yilmaz et al., 2014), higher KISS-1 levels in PCOS were reported even after correcting for BMI. Also, the present meta-regression analysis shows BMI was not a significant source of heterogeneity, whereas waist-to-hip ratio, AMH, free thyroxine and oestradiol were a significant source of heterogeneity (TABLE 2). This needs further investigation, however, bearing in mind the limited and controversial data on the association of KISS-1 levels with BMI and androgen profile in PCOS (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Gorkem et al., 2017; Albalawi et al., 2018).
KISS-1 with testosterone and DHEAS, however, reflect the hyperandrogenism associated with PCOS (Yilmaz et al., 2014). Owing to the contradicting results reported previously (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016; Ozay et al., 2016; Albalawi et al., 2018), we also did not notice significant correlations between KISS-1 with BMI and insulin resistance in this meta-analysis. It has been suggested that free androgen levels increase as a result of insulin resistance, which in turn correlates to BMI in PCOS (Panidis et al., 2006).
Our results of correlation meta-analysis showed significant positive associations of KISS-1 with AMH, testosterone, DHEAS and a statistically borderline correlation with LH. These findings are in line with the previous evidence that KISS-1 upregulates AMH expression and promotes the maturation of large antral follicles (Lapatto et al., 2007). The suggested positive correlation between KISS-1 and LH substantiates previous speculation that KISS-1 plays a role in initiating GnRH secretion and the ability of KISS-1 in eliciting LH secretion in vivo (Gottsch et al., 2004; Navarro and Tena-Sempere, 2012). Correlations of
All three studies included in the DTA meta-analysis used the standard Rotterdam criteria for PCOS diagnosis and measured KISS-1 levels by enzymelinked immunosorbent assay method. The cut-off values obtained in these studies were by ROC analysis and were not pre-defined (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018). Our DTA meta-analysis (FIGURE 6 ) revealed that pooled sensitivity and specificity of KISS-1 for PCOS were 67% and 86%, respectively. It is noteworthy that, in our threshold analysis, we detected no significant source of heterogeneity from the threshold
The sensitivity analyses confirmed the validity and robustness of this metaanalysis. Omitting any single study at a time had no significant influence on the overall effect size. Moreover, no significant publication bias was detected by the funnel plot analysis (FIGURE 4) with Begg's correlation and Egger's regression tests. Therefore, we further aimed to study and report the diagnostic accuracy of increased KISS-1 levels in PCOS.
RBMO VOLUME 39 ISSUE 4 2019
A
s
r
(%)
et al et al et al et al et al
–
Chi-squared P
r
[–0.36 to 0.90] [0.36 to 1.23] [0.11 to 0.52] [–1.28 to –0.05] [1.30 to 2.05] [–0.18 to 1.18]
P
et al
[–0.18 to 0.86] [–0.42 to 0.42] [–0.44 to 0.80] [0.15 to 0.89] [0.62 to 1.77] [–0.08 to 0.66] [0.17 to 0.83] [0.18 to 0.65]
and et al et al et al et al Chi-squared P
P
et al
[–0.36 to 0.90] [–0.42 to 0.42] [–0.44 to 0.80] [0.15 to 0.89] [–0.08 to 0.66] [0.11 to 0.52] [1.30 to 2.05] [0.06 to 0.89]
and et al et al et al et al Chi-squared P
P
et al et al et al et al et al
–
Chi-squared P
[–0.18 to 0.86] [0.36 to 1.23] [0.62 to 1.77] [–1.28 to –0.05] [0.17 to 0.83] [–0.05 to 0.95]
P
[0.27 to 1.39] [–0.27 to 0.74] [1.22 to 2.24] [0.05 to 1.81]
et al
et al et al Chi-squared P
P
– –
et al et al et al Chi-squared P
P
[–0.62 to 0.45] [–1.59 to –0.22] [0.70 to 1.78] [–1.10 to 1.29]
[–0.42 to 0.42] [–0.15 to 0.89] [0.62 to 1.77] [–0.08 to 0.66] [–0.11 to 0.52] [0.17 to 0.82] [1.30 to 2.05] [0.23 to 1.02]
et al et al et al et al et al et al Chi-squared P
P
et al et al and et al et al
– Chi-squared P
[–0.18 to 0.86] [–0.36 to 0.90] [–0.44 to 0.80] [0.36 to 1.23] [–1.28 to –0.05] [–0.27 to 0.68]
P
c
B
s et al
(%)
et al et al et al Chi-squared P
r
r
– –
[–0.41 to 0.68] [–0.15 to 0.70] [–1.44 to 0.34] [–0.85 to 0.02]
–
[–0.72 to 0.29]
P
FIGURE 3 Sub-group analyses of circulatory kisspeptin-1 levels in (A) different subgroups for polycystic ovary syndrome (PCOS) versus controls;
and (B) in obese women with PCOS compared with normal weight women with PCOS.
691
692
RBMO VOLUME 39 ISSUE 4 2019
TABLE 2 RESULTS OF META-REGRESSION ANALYSIS Covariate (PCOS)
Meta-regression Coefficient (95% CI)
SE
P-Value
Sample size
0.00 (–0.005 to 0.005)
0.003
NS
Age
0.038 (–0.074 to 0.151)
0.057
NS
BMI
0.001 (–0.051 to 0.052)
0.026
NS
WC
0.029 (–0.006 to 0.065)
0.018
NS
WHR
11.473 (1.254 to 21.692)
5.214
0.03a
FAI
–0.087 (-0.235 to 0.060)
0.075
NS
Glucose
0.004 (–0.006 to 0.014)
0.005
NS
Insulin
–0.006 (–0.027 to 0.015)
0.011
NS
HOMA
0.627 (-0.281 to 1.536)
0.463
NS
LH
–0.013 (–0.144 to 0.119)
0.067
NS
FSH
–0.045 (–0.551 to 0.462)
0.258
NS
LH/FSH
–0.221 (–0.536 to 0.095)
0.161
NS
Free testosterone
0.341 (0.232 to 0.450)
0.056
0.001a
Total testosterone
–0.008 (–0.020 to 0.004)
0.006
NS
DHEAS
0.00 (–0.001 to 0.002)
0.001
NS
SHBG
0.006 (–0.028 to –0.040)
0.017
NS
Prolactin
–0.001 (–0.007 to 0.004)
0.003
NS
AMH
-0.150 (-0.295 to 0.006)
0.074
0.04a
Oestradiol
–0.011 (–0.020 to –0.001)
0.005
0.03a
17-OHP
2.179 (–9.779 to 14.137)
6.101
NS
AMH, anti-Müllerian hormone; BMI, body mass index; DHEAS, dehydroepiandrosterone; FAI, free androgen index; HOMA, homeostatic model assessment-insulin resistance; NS, non-significant; SE, standard error; SHBG, sex hormone binding globulin; WC, waist circumference; WHR, waist-hip ratio, 17-OHP, 17-hydroxyprogesterone. a P
< 0.05 is statistically significant.
effect. The overall accuracy of KISS-1 for the diagnosis of PCOS was good and favourable (AUC = 0.835). Our results of pooled DOR are promising (pooled DOR = 13.71), and indicated no significant heterogeneity among included studies (I2 = 20.3%). The results of DOR also showed no heterogeneity from the non-threshold effect. Although our results are promising, with no significant heterogeneity from the threshold and non-threshold effects, this DTA analysis is limited by the small number of included studies and relatively small sample size. Therefore, before drawing definitive conclusions, the results obtained in our DTA meta-analysis should be validated further with a larger sample size. Hence, large-scale studies are still needed to explore KISS-1 as a diagnostic biomarker in PCOS. Overall, the present study has some strengths and limitations. To the best of our knowledge, the present meta-analysis is the first of its kind to
systematically review and meta-analyse circulating KISS-1 levels in PCOS, and the protocol of this meta-analysis has been registered (PROSPERO CRD42018112064). Heterogeneity is one of the major concerns in any metaanalysis. Most of the studies included in this meta-analysis, however, are homogenous in PCOS diagnosis and KISS-1 measurement methods. Most of the included studies also have matched PCOS and controls groups for potential confounders, such as age and BMI. Although the heterogeneity has been addressed by applying random-effects model and by sub-group and metaregression analysis, clinical heterogeneity might exist among the included studies. Except for one study (Chen et al., 2010) in which adolescents were enrolled, all other studies recruited adults. Importantly, the funnel plot analysis with Begg's and Egger's tests detected no publication bias, and the sensitivity analysis confirmed the validity of our meta-analysis.
In conclusion, our meta-analysis suggests that circulatory KISS-1 levels increase significantly in women with PCOS compared with controls, and are associated with AMH, LH and androgen excess. The presence of PCOS seems to influence KISS-1 level, regardless of normal or obese BMI. Our diagnostic test accuracy meta-analysis suggests that KISS-1 as a useful biomarker for PCOS with good diagnostic accuracy. Recent research has proposed kisspeptin system as a therapeutic target (Witchel and Tena-Sempere, 2013), with studies providing evidence on peripherally administered KISS-1 induced ovulation in rats (Matsui et al., 2004), and triggering effects on egg maturation, fertilization and blastocyst implantation after singledose KISS-1 injection (Mumtaz et al., 2017). In view of these, the promising results obtained in this meta-analysis should be validated further in large-scale studies exploring KISS-1 as a diagnostic biomarker in PCOS, aiming the associated reproductive consequences.
693
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 4 Publication bias. CES, combined effect size.
Studies
Estimate (95% C.I.)
overall
0.47 [0.17 to 0.77]
− Albalawi et al., 2018 − Chen et al., 2010 − Daghestani 2018 − El−Shehawy and Safan 2015 − Gorkem et al., 2018 − Jeon et al., 2013 − Kaya et al., 2018 − Nyagolova et al., 2016 − Ozay et al., 2016 − Panidis et al., 2006 − Umayal et al., 2019 − Yilmaz et al., 2014
0.48 0.48 0.51 0.49 0.46 0.44 0.41 0.48 0.48 0.56 0.46 0.36
[0.16 [0.16 [0.20 [0.17 [0.13 [0.11 [0.10 [0.15 [0.13 [0.27 [0.13 [0.14
to to to to to to to to to to to to
0.80] 0.80] 0.83] 0.80] 0.79] 0.76] 0.71] 0.81] 0.83] 0.84] 0.80] 0.57] 0.2
0.3
0.4
0.5
0.6
Standardized Mean Difference FIGURE 5 A one-study leave-out sensitivity analysis.
0.7
0.8
694
RBMO VOLUME 39 ISSUE 4 2019
TABLE 3 META-ANALYSIS OF CORRELATIONS Correlation of KISS-1 with
Studies, n
Heterogeneity I2 (%)
Sample size
r
95% CI
Z
P-value
Age
7
68.86
299
0.02
–0.14 to 0.19
0.27
NS
BMI
7
29.57
628
0.02
–0.08 to 0.12
0.33
NS
WC
3
80.30
125
0.12
–0.31 to 0.50
0.53
NS
WHR
3
0.00
364
0.04
–0.06 to 0.14
0.81
NS
FBG
4
81.91
500
–0.08
–0.30 to 0.15
–0.68
NS
Insulin
5
79.15
562
0.07
–0.12 to 0.26
0.71
NS
HOMA-IR
5
71.87
562
0.07
–0.10 to 0.23
0.74
NS
FAI
6
89.78
609
0.25
–0.017 to 0.48
1.83
BS
AMH
5a
65.77
380
0.25
0.0245 to 0.452
2.16
0.03b
Total testosterone
7
37.36
633
0.12
0.088 to 0.30
3.51
<0.001b
Free testosterone
2
5.45
299
0.10
–0.006 to 0.19
1.83
BS
LH
8
97.15
650
0.46
–0.004 to 0.76
1.94
BS
FSH
7
29.60
588
0.01
–0.09 to 0.12
0.18
NS
LH/FSH
2
0.00
304
0.04
–0.07 to 0.15
0.67
NS
DHEAS
4
0.00
477
0.13
0.04 to 0.22
2.85
0.004b
SHBG
5
91.28
562
0.12
–0.17 to 0.41
0.81
NS
AMH, anti-Müllerian hormone; BMI, body mass index; BS, borderline significance (P = 0.05/0.06); DHEAS, dehydroepiandrosterone; FAI, free androgen index; FBG, fasting blood glucose; HOMA-IR, homeostatic model assessment-insulin resistance; NS, non-significant; SHBG, sex hormone binding globulin; WC, waist circumference, WHR, waist-hip ratio. a One b P
study (El-Shehawy and Safan, 2015) is counted twice as the correlations between KISS-1 and AMH were reported separately in PCOS and control groups.
< 0.05 is statistically significant.
TABLE 4 QUADAS-2 CRITICAL APPRAISAL Studies
Design
Patient Threshselectiona oldb
Blinded Valid Disease Verifica- Withindex test reference progres- tionf drawalg c d e results test sion
Risk of bias
RepreExtracta- Applisentative ble datai cability patient concern sampleh
El-Shehawy Case-control and Safan, study 2015
•
•
○
•
Moderate
•
•
Low
Kaya et al., Observa• 2018 tional cohort study
•
•
○
•
Moderate
•
•
Low
Yilmaz Prospective, • et al., 2014 case-control clinical study
•
•
○
•
Low
•
•
Low
a •
Consecutive order and well described in and exclusion criteria;
b •
Based on receiver operator characteristic analysis;
c •
Yes;
d •
Rotterdam criteria;
inappropriate exclusions. |
not reported.
index test was performed by biochemistry departments in papers involving both obstetrics and biochemistry authors;
e • Severe; f •
case–control with consecutive case selection;
pre-specified;
other lab findings;
no/not reported.
none/not reported.
mild to severe; not reported.|
All patients received both index and reference test. Reference test was the same for all patients;
selected patients received equal reference tests;
received different reference tests. g • h • i •
No loss to follow up;
loss to follow up, reasons given;
Patients with PCOS;
healthy control group;
2 × 2 table data extractable;
loss to follow up without reasons given/not reported. |
non-matching domain. |
levels of biomarkers reported, no 2 × 2 data extractable;
only correlation, no data on PCOS.
selected patients
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 6 The results of diagnostic test accuracy meta-analysis. (A) Sensitivity; (B) specificity; (C) diagnostic odds ratio; (D) summary receiver
operator characteristic curve for kisspeptin-1.
695
696
RBMO VOLUME 39 ISSUE 4 2019
ACKNOWLEDGEMENTS
REFERENCES
Dr Varikasuvu SR especially acknowledges Varikasuvu Sahasra Bhairavi for the time I could not give you during this work. We acknowledge Dr K Sreedhar (PhD in Cardiology, SVIMS University, Tirupati) for his help in preparing figures with correct resolution.
Albalawi, F.S., Daghestani, Maha H, Daghestani, Mazin H, Eldali, A., Warsy, A.S., rs4889 polymorphism in KISS1 gene, its effect on polycystic ovary syndrome development and anthropometric and hormonal parameters in Saudi women. 2018; 1–6. Brown, R.E., Wilkinson, D.A., Imran, S.A., Caraty, A., Wilkinson, M. Hypothalamic kiss1 mRNA and kisspeptin immunoreactivity are reduced in a rat model of polycystic ovary syndrome (PCOS). Brain Res. 2012; 1467: 1–9. https://doi. org/10.1016/J.BRAINRES.2012.05.049 Chen, X., Mo, Y., Li, L., Chen, Y., Li, Y., Yang, D. European Journal of Obstetrics & Gynecology and Reproductive Biology Increased plasma metastin levels in adolescent women with polycystic ovary syndrome. Eur. J. Obstet. Gynecol. 2010; 149: 72–76. https://doi. org/10.1016/j.ejogrb.2009.11.018 Daghestani, M.H. Evaluation of biochemical, endocrine, and metabolic biomarkers for the early diagnosis of polycystic ovary syndrome among non-obese Saudi women. Int. J. Gynecol. Obstet. 2018; 142: 162–169. https:// doi.org/10.1002/ijgo.12527 El-Shehawy, Y.M., Safan, M.A., Evaluation of Serum Anti-Mullerian Hormone (AMH) and Plasma Metastin Levels in Women with Polycystic Ovary Syndrome (PCOS). Med. J. Cairo Univ. 2015; 83: 399–406. Funes, S., Hedrick, J.A., Vassileva, G., Markowitz, L., Abbondanzo, S., Golovko, A., Yang, S., Monsma, F.J., Gustafson, E.L. The KiSS1 receptor GPR54 is essential for the development of the murine reproductive system. Biochem. Biophys. Res. Commun. 2003; 312: 1357–1363 Gorkem, U., Togrul, C., Arslan, E., Oruc, A.S, Duman, N.B. Is there a role for kisspeptin in pathogenesis of polycystic ovary syndrome? Gynecol. Endocrinol. 2018; 34: 157–160. https:// doi.org/10.1080 /09513590.2017.1379499 Gottsch, M.L., Cunningham, M.J., Smith, J.T., Popa, S.M., Acohido, B. V., Crowley, W.F., Seminara, S., Clifton, D.K., Steiner, R.A. A Role for Kisspeptins in the Regulation of Gonadotropin Secretion in the Mouse. Endocrinology 2004; 145: 4073–4077. https:// doi.org/10.1210/en.2004-0431 Hameed, S., Jayasena, C.N., Dhillo, W.S. 2006 Kisspeptin and fertility. https://doi.org/10.1677/ JOE-10-0265 Hori, A., Honda, S., Asada, M., Ohtaki, T., Oda, K., Watanabe, T., Shintani, Y., Yamada, T., Suenaga, M., Kitada, C., Onda, H., Kurokawa, T., Nishimura, O., Fujino, M. Metastin Suppresses the Motility and Growth of CHO Cells Transfected with Its Receptor. Biochem. Biophys. Res. Commun. 2001; 286: 958–963 https://doi.org/10.1006/BBRC.2001.5470. Hu, K., Zhao, H., Chang, H., Yu, Y., Qiao, J. Kisspeptin / Kisspeptin Receptor System in the Ovary. 2018; 8, 1–14. https://doi. org/10.3389/fendo.2017.00365 Jeon, Y.E., Lee, K.E., Jung, J.A., Yim, S.Y., Kim, H., Seo, S.K., Cho, S., Choi, Y.S., Lee, B.S. Kisspeptin, Leptin, and Retinol-Binding Protein 4 in Women with Polycystic Ovary Syndrome. Gynecol. Obstet. Invest. 2013; 75: 268–274. https://doi.org/10.1159/000350217 Kaya, C., Alay, İ., Babayeva, G., Gedikbaşı, A., Ertaş, S., Ekin, M., Yaşar, L. Serum Kisspeptin
levels in unexplained infertility, polycystic ovary syndrome, and male factor infertility. 2018; 3590. https://doi.org/10.1080/09513590. 2018.1519792 Lapatto, R., Pallais, J.C., Zhang, D., Chan, Y.-M., Mahan, A., Cerrato, F., Le, W.W., Hoffman, G.E., Seminara, S.B. Kiss1−/− Mice Exhibit More Variable Hypogonadism than Gpr54−/− Mice. Endocrinology 2007; 148: 4927–4936. https://doi.org/10.1210/en.2007-0078 Li, W., Ma, D., Liu, M., Liu, H., Feng, S., Hao, Z., Wu, B., Zhang, S. Association between Metabolic Syndrome and Risk of Stroke: A Meta-Analysis of Cohort Studies. Cerebrovasc. Dis. 2008; 25: 539–547. https:// doi.org/10.1159/000131672 March, W.A., Moore, V.M., Willson, K.J., Phillips, D.I.W., Norman, R.J., Davies, M.J. The prevalence of polycystic ovary syndrome in a community sample assessed under contrasting diagnostic criteria. Hum. Reprod. 2010; 25: 544–551. https://doi.org/10.1093/ humrep/dep399 Matsui, H., Takatsu, Y., Kumano, S., Matsumoto, H., Ohtaki, T. Peripheral administration of metastin induces marked gonadotropin release and ovulation in the rat. Biochem. Biophys. Res. Commun. 2004; 320: 383–388 https://doi.org/10.1016/J.BBRC.2004.05.185. Mumtaz, A., Khalid, A., Jamil, Z., Fatima, S.S., Arif, S., Rehman, R., Kisspeptin: A Potential Factor for Unexplained Infertility and Impaired Embryo Implantation. Int. J. Fertil. Steril. 2017; 11: 99–104. https://doi.org/10.22074/ijfs.2017.4957 Navarro, V.M., Castellano, J.M., FernándezFernández, R., Tovar, S., Roa, J., Mayen, A., Nogueiras, R., Vazquez, M.J., Barreiro, M.L., Magni, P., Aguilar, E., Dieguez, C., Pinilla, L., Tena-Sempere, M. Characterization of the Potent Luteinizing Hormone-Releasing Activity of KiSS-1 Peptide, the Natural Ligand of GPR54. Endocrinology 2005; 146: 156–163. https://doi.org/10.1210/en.2004-0836. Navarro, V.M., Fernández-Fernández, R., Castellano, J.M., Roa, J., Mayen, A., Barreiro, M.L., Gaytan, F., Aguilar, E., Pinilla, L., Dieguez, C., Tena-Sempere, M. Advanced vaginal opening and precocious activation of the reproductive axis by KiSS-1 peptide, the endogenous ligand of GPR54. J. Physiol. 2004; 561: 379–386. https://doi.org/10.1113/ jphysiol.2004.072298 Navarro, V.M., Tena-Sempere, M. Neuroendocrine control by kisspeptins: role in metabolic regulation of fertility. Nat. Rev. Endocrinol. 2012; 8: 40–53. https://doi. org/10.1038/nrendo.2011.147 Nejad, S.Z., Tehrani, F.R., Zadeh-vakili, A., The Role of Kisspeptin in Female Reproduction. 2017; 15, 1–11. https://doi.org/10.5812/ijem.44337. Review Nyagolova, P. V, Mitkov, M.D., Orbetzova, M.M., Terzieva, D.D. Kisspeptin And Galanin-like Peptide (GALP) Levels In Women With Polycystic Ovary Syndrome. 2016; 7–12. Ohtaki, T., Shintani, Y., Honda, S., Matsumoto, H., Hori, A., Kanehashi, K., Terao, Y., Kumano, S., Takatsu, Y., Masuda, Y., Ishibashi, Y., Watanabe, T., Asada, M., Yamada, T., Suenaga, M., Kitada, C., Usuki, S., Kurokawa, T., Onda, H., Nishimura, O., Fujino, M. Metastasis suppressor gene KiSS-1 encodes peptide ligand of a G-protein-coupled receptor. Nature 2001; 411: 613–617. https://doi. org/10.1038/35079135
Ozay, O.E., Ozay, A.C., Acar, B., Cagliyan, E., Seçil, M., Küme, T., Ozay, O.E., Ozay, A.C., Acar, B., Cagliyan, E., Sec, M. Role of kisspeptin in polycystic ovary syndrome. 2016; 3590. https://doi.org/10.3109/09513590.2016.1161019 Panidis, D., Farmakiotis, D., Rousso, D., Katsikis, I., Kourtis, A., Diamanti-Kandarakis, E. Serum luteinizing hormone levels are markedly increased and significantly correlated with Δ4-androstenedione levels in lean women with polycystic ovary syndrome. Fertil. Steril. 2005; 84: 538–540. https://doi.org/10.1016/j. fertnstert.2005.02.017 Panidis, D., Ph, D., Rousso, D., Koliakos, G. Plasma metastin levels are negatively correlated with insulin resistance and free androgens in women with polycystic ovary syndrome. 2006; 85, 0–5. https://doi.org/10.1016/j. fertnstert.2005.11.044 Rosenfield, R.L., Ehrmann, D.A. The Pathogenesis of Polycystic Ovary Syndrome (PCOS): The Hypothesis of PCOS as Functional Ovarian Hyperandrogenism Revisited. Endocr. Rev. 2016; 37: 467–520. https://doi.org/10.1210/ er.2015-1104
RBMO VOLUME 39 ISSUE 4 2019
Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group. Revised 2003 consensus on diagnostic criteria and longterm health risks related to polycystic ovary syndrome (PCOS). Hum. Reprod. 2004; 19: 41–47 Seminara, S.B., Messager, S., Chatzidaki, E.E., Thresher, R.R., Acierno, J.S., Shagoury, J.K., Bo-Abbas, Y., Kuohung, W., Schwinof, K.M., Hendrick, A.G., Zahn, D., Dixon, J., Kaiser, U.B., Slaugenhaupt, S.A., Gusella, J.F., O'Rahilly, S., Carlton, M.B.L., Crowley, W.F., Aparicio, S.A.J.R., Colledge, W.H. The GPR54 Gene as a Regulator of Puberty. N. Engl. J. Med 2003; 349: 1614–1627. https://doi.org/10.1056/ NEJMoa035322 Umayal, B., Jayakody, S., Chandrasekharan, N., S Wijesundera, W., Wijeyaratne, C. Polycystic ovary syndrome (PCOS) and kisspeptin – A Sri Lankan study. J. Postgrad. Med. 2019; 65: 18–23. https://doi.org/10.4103/jpgm. JPGM_683_17 Whiting, P.F., Rutjes, A.W.S., Westwood, M.E., Mallett, S., Deeks, J.J., Reitsma, J.B., Leeflang, M.M.G., Sterne, J.A.C., Bossuyt, P.M.M.
697
QUADAS-2 Group. QUADAS-2: A Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies. Ann. Intern. Med. 2011; 155: 529. https://doi.org/10.7326/0003-4819-155-8201110180-00009 Witchel, S.F., Tena-Sempere, M. The Kiss1 system and polycystic ovary syndrome: lessons from physiology and putative pathophysiologic implications. Fertil. Steril. 2013; 100: 12–22. https://doi.org/10.1016/j.fertnstert.2013.05.024. Yilmaz, S.A., Kerimoglu, O.S., Pekin, A.T., Incesu, F., Dogan, N.U., Celik, C., Unlu, A. European Journal of Obstetrics & Gynecology and Reproductive Biology Metastin levels in relation with hormonal and metabolic pro fi le in patients with polycystic ovary syndrome. Elsevier Irel. Ltd 2014; 180: 56–60. https://doi. org/10.1016/j.ejogrb.2014.06.004 Received 6 February 2019; received in revised form 10 April 2019; accepted 25 April 2019.
RBMO VOLUME 39 ISSUE 4 2019
ARTICLE
Circulatory metastin/kisspeptin-1 in polycystic ovary syndrome: a systematic review and meta-analysis with diagnostic test accuracy BIOGRAPHY
Varikasuvu SR obtained his Doctorate in Medical Biochemistry from SVIMS University, India. He is now a member of Biochemistry Department at AIIMS, with research interests in disease biochemistry, redox homeostasis, endocrinology and diabetes, proteomics and bioinformatics, systematic reviews and meta-analyses. He is an expert editorial reviewer for specialty journals by ELSEVIER, Springer, Karger, WILEY, SAGE and Portland press. Seshadri Reddy Varikasuvu1,*, V Satya Prasad2, VC Vamshika3, MV Satyanarayana3, Jaipal Reddy Panga4 KEY MESSAGE
Circulating kisspeptin-1 (kiss-1) peptide levels in women with polycystic ovary syndrome (PCOS) were higher compared with healthy controls and showed significant associations with anti-Müllerian hormone and dehydroepiandrosterone levels. Overall diagnostic accuracy of KISS-1 for PCOS was good (area under the curve = 0.835), with a promising diagnostic odds ratio (13.71). ABSTRACT
Research question: A close association between Kisspeptin-1 (KISS-1) and reproductive physiology has been reported, but the results on circulatory KISS-1 are ambiguous in patients with polycystic ovary syndrome (PCOS). A systematic review and meta-analysis were conducted to evaluate the association between KISS-1 and PCOS, and to test its diagnostic test accuracy (DTA) through DTA meta-analysis. Design: Relevant studies were identified by searching PubMed and other databases in addition to manual searching of cross-references. Random-effects model was used to obtain standardized mean differences (SMD), pooled correlation coefficients and summary of DTA. Meta-regression and sub-group analyses were conducted to explore heterogeneity. The presence of publication bias was tested using funnel plot analysis. Results: This meta-analysis finally included 12 studies. Compared with controls, women with PCOS showed significantly increased circulatory KISS-1 levels (SMD = 0.47; P = 0.002). Meta-analysis of correlations showed positive associations between KISS-1 and anti-Müllerian hormone (AMH) (P = 0.03), testosterone (P < 0.001) and dehydroepiandrosterone (P = 0.004). The pooled diagnostic odds ratio and area under curve were 13.71 and 0.835, respectively. A one-study leave-out sensitivity analysis indicated that no single study had a significant influence on the overall outcome, suggesting the robustness of this meta-analysis. Conclusions: This meta-analysis showed significantly increased KISS-1 level in PCOS, and its association with AMH reflects its role in reproductive physiology. In our DTA meta-analysis, KISS-1 showed good accuracy for PCOS detection. Further large-scale studies are required to establish its validity. 1 Department
of Biochemistry, All India Institute of Medical Sciences (AIIMS), Bibinagar Telangana 508126, India of Anatomy, Maheshwara Medical College and Hospital, Hyderabad Telangana 502307, India of Obstetrics and Gynecology (MBBS Students), Maheshwara Medical College and Hospital, Hyderabad Telangana 502307, India 4 Institute for Systems Biology, Seattle WA, USA 2 Department
3 Department
© 2019 Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. *Corresponding author. E-mail address: [email protected] (S R Varikasuvu). https://doi.org/10.1016/j. rbmo.2019.04.018 1472-6483/© 2019 Reproductive Healthcare Ltd. Published by Elsevier Ltd. All rights reserved. Declaration: The authors report no financial or commercial conflicts of interest.
KEYWORDS
Kisspeptin-1 Meta-analysis Metastin Polycystic ovary syndrome
686
RBMO VOLUME 39 ISSUE 4 2019
INTRODUCTION
P
olycystic ovary syndrome (PCOS) is a common endocrinological disorder characterized by reproductive, endocrine and metabolic disturbances. It is the leading cause of female infertility, affecting 6–10% of reproductive-aged women (March et al., 2010). The clinical and biochemical signs of PCOS include polycystic ovaries, anovulation and hyperandrogenism (Rosenfield and Ehrmann, 2016). Although the aetiopathogenesis of PCOS remains unclear, the hypothalamic–pituitary gonadal (HPG) axis has been proposed to be involved, with observed disturbances in gonadotrophin secretion, increased LH levels and perturbed LH and FSH ratios (Panidis et al., 2005). Kisspeptin-1 (KISS-1), also known as metastin, is a 54-aminoacid peptide encoded by the Kiss-1 gene, and was first identified from the human placenta (Hori et al., 2001; Ohtaki et al., 2001). KISS-1 has been proposed to be important in the initiation of gonadotrophin-releasing hormone (GnRH) secretion from the hypothalamus at puberty and is also known to regulate LH secretion during the promotion of ovulation (Matsui et al., 2004; Navarro et al., 2005). The association between KISS-1 and PCOS, however, remains poorly understood. Available data on circulatory KISS-1 levels in PCOS have also been inconsistent (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Albalawi et al., 2018; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019). Furthermore, no previous meta-analysis has been conducted on the association between KISS-1 levels in patients with PCOS. Considering the important relationship between HPG axis and KISS-1, the aim of the present study was to conduct a systematic review and meta-analysis of circulatory KISS-1 levels and their association with biochemical, metabolic and hormonal parameters in PCOS. Diagnostic test accuracy (DTA) metaanalysis was also conducted for the utility of circulatory KISS-1 in PCOS.
MATERIALS AND METHODS The study was registered with the International Prospective Register of
Systematic Reviews (PROSPERO), number CRD42018112064. The criteria of preferred reporting items for systematic reviews and meta-analysis (PRISMA) were followed in conducting and reporting this systematic review and meta-analysis. Search strategy The NCBI PubMed database was searched using MESH and text word search strategies. The following search string was developed: (‘kisspeptins’[MeSH Terms] OR ‘kisspeptins’[All Fields] OR ‘kisspeptin’[All Fields]) OR (Kiss-1[All Fields] AND (‘peptides’[MeSH Terms] OR ‘peptides’[All Fields] OR ‘peptide’[All Fields])) OR (‘kisspeptins’[MeSH Terms] OR ‘kisspeptins’[All Fields] OR ‘metastin’[All Fields]) AND PCOS[All Fields]). The Web of Science, Google Scholar, Embase, Cochrane Library, Springer's AuthorMapper and Science Direct databases were also searched for articles that reported circulatory KISS-1 levels in women with PCOS. Furthermore, bibliographies of published articles were manually hand searched to identify additional studies. Two authors independently conducted literature searches, and any discrepancies were resolved with discussion. Efforts were made to obtain unpublished data, if any. When required, the corresponding authors of respective articles were E-mailed to obtain clarification. The literature search was first conducted on 17 November 2018, and all searches were conducted before 1 January 2019, with no timespan specified, with and without humans set as a limit. Study selection criteria The individual studies had to meet the following criteria to be included in this meta-analysis: peer-reviewed publications with case–control or observational study design; original data; study participants were human with description of PCOS and controls; studies in which circulating levels of KISS-1 in PCOS were compared with controls; description for KISS-1 estimation method provided; articles written in English with clear data; and studies diagnosing PCOS based on the revised 2003 Rotterdam criteria (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group, 2004). The exclusion criteria were as follows: studies with no control group; studies reported on diseases other than PCOS; KISS-1 levels were only shown by pictograms with no data; and reviews, commentaries and Letter to the
Editor article types. If the same group of authors published duplicate articles on the same study population, only a recently published article or article with all relevant information was included. Data extraction and quality assessment The following information was extracted from each of the included studies; first author names, country and year of publication, number of participants with PCOS and control participants, PCOS diagnosis criteria, means and SD of age and anthropometric, metabolic, hormonal parameters, KISS-1 levels, with its measurement method and units, correlation analysis data, receiver operating characteristic analysis data (accuracy measures: sensitivity, specificity and area under the curve values) and other study characteristics. A quality score evaluation for studies was carried out according to Newcastle– Ottawa Scale for case–control and observational studies (Li et al., 2008). Quality assessment of study scores range from 0 to 9 points and included three components: selection, comparability and exposure. Quality of studies included in the DTA analysis was reported using ‘Quality assessment of diagnostic accuracy studies’ (QUADAS-2) tool (Whiting et al., 2011). Statistical analysis Meta-analyses were conducted if three or more studies reported the circulatory KISS-1 concentrations in women with PCOS patients compared with controls. A separate subgroup meta-analysis was conducted based on body mass index (BMI) to report its effect on the overall outcome. We calculated the standardized mean difference (SMD) and its 95% confidence interval (CI) as a summary statistic for the difference of KISS-1 level between women with PCOS and control groups; normal weight women with PCOS and obese women with PCOS subgroups. Furthermore, a meta-analysis of the correlation coefficients of KISS-1 was conducted with various parameters. The effect size for SMD and pooled correlation coefficient values were presented as a Z-score. The Z-score P < 0.05 was considered statistically significant. The between-study heterogeneity was examined by Cochrane's Q statistic and expressed as percentages of I2. P < 0.05 or chi-squared statistic of over 50%
indicated a significant heterogeneity. A random-effects model was used to compute SMD. A meta-regression analysis was conducted to identify the potential source of heterogeneity using anthropometric, hormonal, metabolic and lipid parameters. Risk of publication bias was studied by funnel plot asymmetry with Begg's and Egger's tests. To avoid risk of bias, however, general search terms were chosen for a guaranteed retrieval for inclusion of articles reporting KISS-1 levels as a secondary outcome. In the case of P < 0.10, suggesting statistical significance, the ‘trim-and-fill method’ was used to correct that bias. To test the robustness of this meta-analysis, a one-study leaveout sensitivity analysis was conducted. All comparisons were two-tailed, and all analyses were conducted using the Review Manager Software version 5.3, which presents SMD as Hedges g; the difference between the two means divided by the pooled SD, with a correction for sample bias. The funnel plot for publication bias with Begg's and Egger's tests was conducted using Comprehensive meta-analysis, version 3. The MedCalc version 16.2.0 was used for the meta-analysis of correlations. In studies in which receiver operator characteristic (ROC) data were presented, the DTA meta-analysis was conducted to obtain summary ROC curve (SROC) with pooled sensitivity, specificity and diagnostic odds ratio (DOR) of KISS-1 for PCOS. Meta-Disc software, version1.4 was used to conduct the DTA analyses. The heterogeneity, caused by threshold and non-threshold effects, was assessed by Spearman correlation analysis and meta-regression, respectively. Moses linear model was used to obtain SROC curve and the pooled AUC value. the DerSimonian Laird method was used to obtain pooled sensitivity, specificity and DOR, with their respective 95% confidence intervals.
RESULTS Search results and study characteristics Initially, the literature search retrieved 2168 articles, of which, 2129 articles were excluded after reviewing for KISS-1 description in PCOS. After reviewing the full texts of the remaining 39 publications, 27 were excluded; by applying ‘humans’ filter (n = 20), gene expression/polymorphism studies (n = 3),
RBMO VOLUME 39 ISSUE 4 2019
reviews with no original data (n = 2), duplicates (n = 1) and no control group (n = 1). Finally, a total of 12 studies were included in this meta-analysis comparing circulatory KISS-1 levels between women with PCOS and control patients (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Albalawi et al., 2018; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019). Among these, four studies reported KISS-1 levels between patients with PCOS with normal weight and patients with PCOS who were overweight (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016). Correlations of KISS-1 have been reported with AMH (El-Shehawy and Safan, 2015; Ozay et al., 2016; Gorkem et al., 2017; Kaya et al., 2018), total-testosterone (Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017), LH (Chen et al., 2010; Jeon et al., 2013; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Ozay et al., 2016; Gorkem et al., 2017; Kaya et al., 2018), and other variables. In one study (El-Shehawy and Safan, 2015), the correlations between KISS-1 and AMH were reported separately in PCOS and control groups. Therefore, these correlation coefficients were considered as two observations. Three studies reported diagnostic accuracy measures and were included in the DTA metaanalysis (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018). The PRISMA flow diagram is presented in FIGURE 1 . The characteristics and the Newcastle– Ottawa Scale quality scores of included studies are presented in TABLE 1. With the obtained score range from 5 to 8, the overall quality of studies was medium to high. All of the included studies used Rotterdam criteria for PCOS diagnosis (Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group, 2004). All of the included studies except two (Yilmaz et al., 2014; Albalawi et al., 2018) are age matched between PCOS and control groups. Matching of BMI was carried out in seven of the included studies (Chen et al., 2010; Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Ozay et al., 2016; Nyagolova et al., 2016; Gorkem et al., 2017; Daghestani, 2018). The circulatory KISS-1 levels
687
were determined by enzyme-linked immunosorbent assay using serum in seven of the included studies (Yilmaz et al., 2014; Nyagolova et al., 2016; Ozay et al., 2016; Gorkem et al., 2017; Daghestani, 2018; Kaya et al., 2018; Umayal et al., 2019), whereas the remaining five studies used plasma (Panidis et al., 2006; Chen et al., 2010; Jeon et al., 2013; El-Shehawy and Safan, 2015; Albalawi et al., 2018). Circulatory KISS-1 level in PCOS compared with controls As shown in FIGURE 2, women with PCOS had increased KISS-1 level compared with non-PCOS controls. With a significant between-study heterogeneity (I2 = 84%; P < 0.00001), the randomeffects model was applied to compute the pooled effect size. The pooled SMD and 95% confidence intervals were 0.47 (0.17 to 0.77). The overall effect size for SMD calculated as Z was 3.06 (P = 0.002). Sub-group analysis and metaregression A subgroup analyses of circulatory KISS-1 levels in PCOS compared with controls is shown in FIGURE 3A. The pooled SMD was not statistically significant in the subgroups of ‘age 24 years or younger’, ‘studies with no MBI-matching’, ‘obese BMI’, and ‘plasma’. This may suggest the possible source of heterogeneity among the included studies. The pooled SMDs were statistically significant in other subgroups of ‘age over 24 years’ (SMD = 0.42; Z = 3.41; P = 0.0006; I2 = 52%), ‘studies matched for BMI’ (SMD = 0.48; Z = 2.27; P = 0.02; I2 = 88%; P for I2 < 0.00001), ‘normoweight BMI’ (SMD = 0.93; Z = 2.08; P = 0.04; I2 = 88%; P for I2 = 0.0002), and ‘serum’ (SMD = 0.63; Z = 3.13; P = 0.002; I2 = 89%; P for I2 < 0.00001), indicating that circulatory KISS-1 levels were higher in women with PCOS compared with controls. When comparisons were made between obese women with PCOS and women with PCOS of normal weight for the KISS-1 level reported in four studies (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016), the statistical significance was not achieved (SMD = − 0.21; Z = 0.82; I2 = 77%; P = 0.004). The meta-regression analysis (TABLE 2) carried out with various covariates demonstrated that waist-to-hip
688
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 1 Study selection based on PRISMA criteria.
ratio (coefficient = 11.47; P = 0.03), free-testosterone (coefficient = 0.34; P = 0.001), AMH (coefficient = − 0.15; P = 0.04) and oestradiol (coefficient = − 0.01; P = 0.03) produced statistically significant regression coefficients, indicating their effect on the heterogeneity among included studies. Publication bias and sensitivity analysis No significant publication bias was detected among the included studies (FIGURE 4) according to the funnel plot asymmetry with Begg's correlation and Egger's regression tests. A one studyleave out sensitivity analysis showed that no single study had significantly influenced the combined overall outcome, and the combined SMDs obtained were stable and remained statistically significant, after omitting any particular study in sensitivity analysis (FIGURE 5). Two studies (Panidis et al., 2006; Yilmaz et al., 2014) have significantly influenced the overall result, whereas when the study by Panidis et al. (2006) was omitted,
the pooled SMD was increased from 0.47 (Z = 3.06; P = 0.002) to 0.56 (Z = 3.80; P = 0.0001) with only a 2% decrease in heterogeneity. When the study by Yilmaz et al. (2014) was omitted, the pooled SMD decreased from 0.47 (Z = 3.06; P = 0.002) to 0.36 (Z = 3.31; P = 0.0009), with decreased heterogeneity from 84% to 64%. Meta-analysis of correlation coefficients The results of correlation meta-analysis of KISS-1 with various anthropometric, metabolic and hormonal variables are presented in TABLE 3. The random-effects meta-analysis showed significant positive correlations between KISS-1 and AMH (pooled r = 0.25; P = 0.03), totaltestosterone (pooled r = 0.12; P < 0.001), DHEAS (pooled r = 0.13; P = 0.004), and statistically borderline positive correlations with LH (pooled r = 0.46; P = 0.05), free androgen index (pooled r = 0.25; P =0.06), and free testosterone (pooled r = 0.10; P = 0.06).
Diagnostic test accuracy meta-analysis Three studies (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018) reported the diagnostic accuracy measures of KISS-1 for PCOS. QUADAS-2 tool was used for quality assessment of these studies (TABLE 4). All the studies used enzymelinked immunosorbent assay for KISS-1 measurement, and the cut-off values used were not pre-specified. The results of DTA meta-analysis are shown in FIGURE 6. The pooled sensitivity was 0.67 (95% CI 0.58 to 0.75); specificity was 0.86 (95% CI 0.78 to 0.92), and the pooled DOR was 13.71 (95% CI 6.11 to 30.75) (FIGURE 6A–6c, respectively). The SROC curve for KISS-1, AUC and Q index are shown in FIGURE 6d. The pooled AUC obtained was 0.835 (Q* = 0.767). According to the threshold results, the b value was − 0.305, suggesting that the SROC curve was symmetric. The Spearman correlation coefficient obtained was 0.50, indicating no heterogeneity from the threshold effect.
689
RBMO VOLUME 39 ISSUE 4 2019
TABLE 1 CHARACTERISTICS OF INCLUDED STUDIES Study, year (country)
Study groups Control Age BMI
KISS-1
PCOS versus control groups
NOS
PCOS n
Age BMI
n
Method
Sample
Units
Matching/no- difference
Significant differences Age, BMI, LH
Albalawi et al., 26.7 2018 (Saudi Arabia)
36.6
30
29.4
46.0
28
ELISA
Plasma
pg/ml
FSH
Chen et al., 2010 18.6 (China)
20.11
20
17.8
21.4
19
ELISA
Plasma
fmol/ml
Age, BMI, FSH, T, SHBG, LH, T, FAI, DHEAS, INS, 2h-INS, FPG, 2h-GLU GLU/INS, HOMA
Daghestani, 2018 25.4 (Saudi Arabia)
22.1
44
24.0
22.8
44
ELISA
Serum
fmol/ml
Age, BMI, HC, FSH, leptin, ghrelin, vitamin D
LH, LH/FSH, E2, T, PG, SHBG, 7 WC, WHR, TC, TG, LDL, HDL, F-Ins, FPG, VEGF
El-Shehawy and Safan, 2015 (Kuwait)
27.5
22.1
18
24.4
23.4
23
ELISA
Plasma
pmol/l
Age, BMI, WC, FSH
LH, T, DHEAS, AMH
6
Gorkem et al., 2018 (Turkey)
27.9
26.0
57
27.7
26.3
60
ELISA
Serum
ng/ml
Age, BMI, WC, HC, WHR, FSH, ED, 7OHP
LH, T, DHEAS, AMH
7
Jeon et al., 2013 (Korea)
24.9
19.7
36
23.7
23.0
54
ELISA
Plasma
pmol/l
Age, FSH, SHBG, PROL LH, LH/FSH, BMI, WHR, GLU, INS, GLU/INS, HOMA, T, FAI, leptin, RBP4
7
Kaya et al., 2018 (Turkey)
30.2
26.4
27
28.7
28.6
29
ELISA
Serum
pg/ml
Age, LH, oestradiol, TSH, BMI, FSH, PROL, AFC AMH, gravidity, parity, infertility duration
6
Nyagolova et al., 26.6 2016 (Bulgaria)
24.8
42
24.9
25.5
87
ELISA
Serum
ng/ml
age, BMI, WC, FSH, PROL, TSH, FPG, INS
LH, LH/FSH, HOMA, T, SHBG, 8 FAI, GALP
Ozay et al., 2016 (Turkey)
24.4
23.4
150 23.9
24.3
250 ELISA?
Serum
ng/ml
age, BMI, BP, FAI, VFT, PROL, T, FPG, HDL, leptin, CRP, ET-1
LH, LH/FSH, FSH, WHR, INS, 8 HOMA, Ferriman–Galwey Score, CIMT, Hcy, T, DHEAS, SHBG, TC, TG, LDL, APO B/A1, AMH
Panidis et al., 2006 (Greece)
26.8
32.1
13
23.9
26.8
56
ELISA
Plasma
fmol/ml
Age, FSH, PROL, FPG
BMI, LH, T, FAI, 17OHP, SHBG, 7 INS, GLU/INS, HOMA
Umayal et al., 2019 (Sri Lanka)
33.8
25.2
110 24.6
26.8
55
ELISA
Serum
nmol/l
Age, WHR, FPG
T, BMI, MFG Score
6
Yilmaz et al., 2014 24.5 (Turkey)
23.1
66
24.5
83
ELISA
Serum
ng/ml
BMI, FSH, HOMA, TC, TG, LDL, HDL, E2, PROL, 17OHP, TSH
Age, LH, T, DHEAS, SHBG, FAI, Ferriman–Gallwey Score, FPG, INS
7
21.0
5 6
A1C, haemoglobin A1c; AFP, antral follicle count; AMH: anti-Müllerian hormone; APO A1, apoprotein A1; APO B. apoprotein B; BMI, body mass index; BP, mean arterial blood pressure; CRP, C-reactive protein; DHEAS, dehydroepiandrosterone; E2, oestradiol; ET-1, endothelin-1; ELISA, enzyme-linked immunosorbent assay; FPG, fasting plasma glucose; FAI, free androgenindex; GLU, glucose; GALP, galanin-like peptide; Hcy, homocysteine; HOMA, homeostatic model assessment-insulin resistance; INS, insulin; KISS-1, kiss 1 peptide/kisspeptin/metastin; LDL, low-density lipoproteins; n, number of controls/PCOS patients; NA/?, not available; NOS, Newcastle–Ottawa Scale; PCOS, polycystic ovary syndrome; PG, progesterone; PROL, prolactin; Rotterdam criteria, Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group-Revised 2003 consensus on diagnostic criteria and long-term health risks related to PCOS; RBP4, retinol-binding protein 4; SHBG, sex hormone binding globulin; T, testosterone-total/free; TC, total cholesterol; TG, triglycerides; TSH, thyroid stimulating hormone; VEGF, vascular endothelial growth factor; VFT, visceral fat thickness; WC, waist circumference; WHR, waist-hip ratio; 17OHP: 17-hydroxyprogesterone.
Furthermore, the DOR result suggested no heterogeneity from the non-threshold effect (chi-squared = 2.51).
DISCUSSION To the best of our knowledge, this is the first meta-analysis to compare KISS-1 levels in women with PCOS compared with control participants who did not have PCOS. Findings from this study indicate that PCOS is associated with a significant increase in circulatory KISS-1 levels compared with controls. The peptide KISS-1 has
been reported to play a critical role in the central control of hypothalamic– pituitary–gonadal axis and in the direct control of ovarian function, including follicular development, oocyte maturation, steroidogenesis, ovulation and metabolic control of fertility (Funes et al., 2003; Seminara et al., 2003; Navarro et al., 2004). KISS-1 functions through a G-protein-coupled receptor, kisspeptin receptor (KISS1R) to stimulate gonadotrophin releasing hormone release and subsequent secretion of FSH and LH in many mammals. Evidence also shows a direct
effect of KISS-1 on the gonadotrophs of the anterior pituitary gland, stimulating the release of LH and FSH (Hameed et al., 2006). Expression of both KISS-1 and KISS1R has been reported in many tissues, including liver, placenta, umbilical vein, pancreas, small intestine, central nervous system, aorta, adipose tissue, lymph nodes, peripheral blood lymphocytes, testis and female tract (Witchel and Tena-Sempere, 2013; Hu et al., 2018). Although evidence on the exact source of increased circulatory KISS-1 levels in PCOS is unclear, any of these tissues could be a source.
690
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 2 Circulatory kisspeptin-1 levels.
It has also been documented that locally produced kisspeptin/KISS1R directly participates in a series of physiological and pathological activities in the ovary. High plasma levels of kisspeptin have been reported in women with PCOS, reflective of its relationship to pathological conditions in the ovaries. Furthermore, animal models of PCOS have shown increased levels of KISS1 mRNA in peripheral tissues such as the ovaries and fat (Brown et al., 2012). Moreover, it has been reported that hypothalamic KISS-1 neurones relay gonadal steroid regulation on the HPG axis and hypothalamic KISS-1 mRNA levels are abnormal in women with PCOS. Dysregulation/alterations of the kisspeptin levels may affect the ovarian function, leading to generation of PCOS, reproductive pathology or female infertility (Witchel and Tena-Sempere, 2013; Nejad et al., 2017). Because of abnormal pulsatile GnRH secretion, increased LH pulsatility, decreased FSH secretion and perturbed LH–FSH ratios were reported in PCOS. As kisspeptin is known to be the upstream central controller for inducing GnRH and LH secretion, higher KISS-1 levels in PCOS were speculated. All these findings and the studies included in this meta-analysis support a potential link between PCOS and KISS-1 levels. In line with this, our meta-analysis demonstrated higher circulatory KISS-1 levels in women with PCOS. Our sub-group analysis revealed that the increase in circulatory KISS-1 level was more pronounced in women with PCOS (compared with controls) ‘aged over 24 years’, ‘BMI-matched studies’, ‘normo-weight BMI’, and in the subgroup of studies using ‘serum’ for KISS-1
estimation. The difference in KISS-1 level, however, did not reach statistical significance in the sub-groups of ‘age younger than 24 years’, ‘studies with no BMI-matching’, ‘obese-BMI’, and ‘plasma’ (FIGURE 3a). Also, when KISS-1 levels were compared between obese women with PCOS and normal weight women with PCOS, no significant difference was reported (FIGURE 3b). This indicates that the increase in KISS-1 level in PCOS is independent of BMI. In a recent study (Yilmaz et al., 2014), higher KISS-1 levels in PCOS were reported even after correcting for BMI. Also, the present meta-regression analysis shows BMI was not a significant source of heterogeneity, whereas waist-to-hip ratio, AMH, free thyroxine and oestradiol were a significant source of heterogeneity (TABLE 2). This needs further investigation, however, bearing in mind the limited and controversial data on the association of KISS-1 levels with BMI and androgen profile in PCOS (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Gorkem et al., 2017; Albalawi et al., 2018).
KISS-1 with testosterone and DHEAS, however, reflect the hyperandrogenism associated with PCOS (Yilmaz et al., 2014). Owing to the contradicting results reported previously (Panidis et al., 2006; Jeon et al., 2013; Yilmaz et al., 2014; Nyagolova et al., 2016; Ozay et al., 2016; Albalawi et al., 2018), we also did not notice significant correlations between KISS-1 with BMI and insulin resistance in this meta-analysis. It has been suggested that free androgen levels increase as a result of insulin resistance, which in turn correlates to BMI in PCOS (Panidis et al., 2006).
Our results of correlation meta-analysis showed significant positive associations of KISS-1 with AMH, testosterone, DHEAS and a statistically borderline correlation with LH. These findings are in line with the previous evidence that KISS-1 upregulates AMH expression and promotes the maturation of large antral follicles (Lapatto et al., 2007). The suggested positive correlation between KISS-1 and LH substantiates previous speculation that KISS-1 plays a role in initiating GnRH secretion and the ability of KISS-1 in eliciting LH secretion in vivo (Gottsch et al., 2004; Navarro and Tena-Sempere, 2012). Correlations of
All three studies included in the DTA meta-analysis used the standard Rotterdam criteria for PCOS diagnosis and measured KISS-1 levels by enzymelinked immunosorbent assay method. The cut-off values obtained in these studies were by ROC analysis and were not pre-defined (Yilmaz et al., 2014; El-Shehawy and Safan, 2015; Kaya et al., 2018). Our DTA meta-analysis (FIGURE 6 ) revealed that pooled sensitivity and specificity of KISS-1 for PCOS were 67% and 86%, respectively. It is noteworthy that, in our threshold analysis, we detected no significant source of heterogeneity from the threshold
The sensitivity analyses confirmed the validity and robustness of this metaanalysis. Omitting any single study at a time had no significant influence on the overall effect size. Moreover, no significant publication bias was detected by the funnel plot analysis (FIGURE 4) with Begg's correlation and Egger's regression tests. Therefore, we further aimed to study and report the diagnostic accuracy of increased KISS-1 levels in PCOS.
RBMO VOLUME 39 ISSUE 4 2019
A
s
r
(%)
et al et al et al et al et al
–
Chi-squared P
r
[–0.36 to 0.90] [0.36 to 1.23] [0.11 to 0.52] [–1.28 to –0.05] [1.30 to 2.05] [–0.18 to 1.18]
P
et al
[–0.18 to 0.86] [–0.42 to 0.42] [–0.44 to 0.80] [0.15 to 0.89] [0.62 to 1.77] [–0.08 to 0.66] [0.17 to 0.83] [0.18 to 0.65]
and et al et al et al et al Chi-squared P
P
et al
[–0.36 to 0.90] [–0.42 to 0.42] [–0.44 to 0.80] [0.15 to 0.89] [–0.08 to 0.66] [0.11 to 0.52] [1.30 to 2.05] [0.06 to 0.89]
and et al et al et al et al Chi-squared P
P
et al et al et al et al et al
–
Chi-squared P
[–0.18 to 0.86] [0.36 to 1.23] [0.62 to 1.77] [–1.28 to –0.05] [0.17 to 0.83] [–0.05 to 0.95]
P
[0.27 to 1.39] [–0.27 to 0.74] [1.22 to 2.24] [0.05 to 1.81]
et al
et al et al Chi-squared P
P
– –
et al et al et al Chi-squared P
P
[–0.62 to 0.45] [–1.59 to –0.22] [0.70 to 1.78] [–1.10 to 1.29]
[–0.42 to 0.42] [–0.15 to 0.89] [0.62 to 1.77] [–0.08 to 0.66] [–0.11 to 0.52] [0.17 to 0.82] [1.30 to 2.05] [0.23 to 1.02]
et al et al et al et al et al et al Chi-squared P
P
et al et al and et al et al
– Chi-squared P
[–0.18 to 0.86] [–0.36 to 0.90] [–0.44 to 0.80] [0.36 to 1.23] [–1.28 to –0.05] [–0.27 to 0.68]
P
c
B
s et al
(%)
et al et al et al Chi-squared P
r
r
– –
[–0.41 to 0.68] [–0.15 to 0.70] [–1.44 to 0.34] [–0.85 to 0.02]
–
[–0.72 to 0.29]
P
FIGURE 3 Sub-group analyses of circulatory kisspeptin-1 levels in (A) different subgroups for polycystic ovary syndrome (PCOS) versus controls;
and (B) in obese women with PCOS compared with normal weight women with PCOS.
691
692
RBMO VOLUME 39 ISSUE 4 2019
TABLE 2 RESULTS OF META-REGRESSION ANALYSIS Covariate (PCOS)
Meta-regression Coefficient (95% CI)
SE
P-Value
Sample size
0.00 (–0.005 to 0.005)
0.003
NS
Age
0.038 (–0.074 to 0.151)
0.057
NS
BMI
0.001 (–0.051 to 0.052)
0.026
NS
WC
0.029 (–0.006 to 0.065)
0.018
NS
WHR
11.473 (1.254 to 21.692)
5.214
0.03a
FAI
–0.087 (-0.235 to 0.060)
0.075
NS
Glucose
0.004 (–0.006 to 0.014)
0.005
NS
Insulin
–0.006 (–0.027 to 0.015)
0.011
NS
HOMA
0.627 (-0.281 to 1.536)
0.463
NS
LH
–0.013 (–0.144 to 0.119)
0.067
NS
FSH
–0.045 (–0.551 to 0.462)
0.258
NS
LH/FSH
–0.221 (–0.536 to 0.095)
0.161
NS
Free testosterone
0.341 (0.232 to 0.450)
0.056
0.001a
Total testosterone
–0.008 (–0.020 to 0.004)
0.006
NS
DHEAS
0.00 (–0.001 to 0.002)
0.001
NS
SHBG
0.006 (–0.028 to –0.040)
0.017
NS
Prolactin
–0.001 (–0.007 to 0.004)
0.003
NS
AMH
-0.150 (-0.295 to 0.006)
0.074
0.04a
Oestradiol
–0.011 (–0.020 to –0.001)
0.005
0.03a
17-OHP
2.179 (–9.779 to 14.137)
6.101
NS
AMH, anti-Müllerian hormone; BMI, body mass index; DHEAS, dehydroepiandrosterone; FAI, free androgen index; HOMA, homeostatic model assessment-insulin resistance; NS, non-significant; SE, standard error; SHBG, sex hormone binding globulin; WC, waist circumference; WHR, waist-hip ratio, 17-OHP, 17-hydroxyprogesterone. a P
< 0.05 is statistically significant.
effect. The overall accuracy of KISS-1 for the diagnosis of PCOS was good and favourable (AUC = 0.835). Our results of pooled DOR are promising (pooled DOR = 13.71), and indicated no significant heterogeneity among included studies (I2 = 20.3%). The results of DOR also showed no heterogeneity from the non-threshold effect. Although our results are promising, with no significant heterogeneity from the threshold and non-threshold effects, this DTA analysis is limited by the small number of included studies and relatively small sample size. Therefore, before drawing definitive conclusions, the results obtained in our DTA meta-analysis should be validated further with a larger sample size. Hence, large-scale studies are still needed to explore KISS-1 as a diagnostic biomarker in PCOS. Overall, the present study has some strengths and limitations. To the best of our knowledge, the present meta-analysis is the first of its kind to
systematically review and meta-analyse circulating KISS-1 levels in PCOS, and the protocol of this meta-analysis has been registered (PROSPERO CRD42018112064). Heterogeneity is one of the major concerns in any metaanalysis. Most of the studies included in this meta-analysis, however, are homogenous in PCOS diagnosis and KISS-1 measurement methods. Most of the included studies also have matched PCOS and controls groups for potential confounders, such as age and BMI. Although the heterogeneity has been addressed by applying random-effects model and by sub-group and metaregression analysis, clinical heterogeneity might exist among the included studies. Except for one study (Chen et al., 2010) in which adolescents were enrolled, all other studies recruited adults. Importantly, the funnel plot analysis with Begg's and Egger's tests detected no publication bias, and the sensitivity analysis confirmed the validity of our meta-analysis.
In conclusion, our meta-analysis suggests that circulatory KISS-1 levels increase significantly in women with PCOS compared with controls, and are associated with AMH, LH and androgen excess. The presence of PCOS seems to influence KISS-1 level, regardless of normal or obese BMI. Our diagnostic test accuracy meta-analysis suggests that KISS-1 as a useful biomarker for PCOS with good diagnostic accuracy. Recent research has proposed kisspeptin system as a therapeutic target (Witchel and Tena-Sempere, 2013), with studies providing evidence on peripherally administered KISS-1 induced ovulation in rats (Matsui et al., 2004), and triggering effects on egg maturation, fertilization and blastocyst implantation after singledose KISS-1 injection (Mumtaz et al., 2017). In view of these, the promising results obtained in this meta-analysis should be validated further in large-scale studies exploring KISS-1 as a diagnostic biomarker in PCOS, aiming the associated reproductive consequences.
693
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 4 Publication bias. CES, combined effect size.
Studies
Estimate (95% C.I.)
overall
0.47 [0.17 to 0.77]
− Albalawi et al., 2018 − Chen et al., 2010 − Daghestani 2018 − El−Shehawy and Safan 2015 − Gorkem et al., 2018 − Jeon et al., 2013 − Kaya et al., 2018 − Nyagolova et al., 2016 − Ozay et al., 2016 − Panidis et al., 2006 − Umayal et al., 2019 − Yilmaz et al., 2014
0.48 0.48 0.51 0.49 0.46 0.44 0.41 0.48 0.48 0.56 0.46 0.36
[0.16 [0.16 [0.20 [0.17 [0.13 [0.11 [0.10 [0.15 [0.13 [0.27 [0.13 [0.14
to to to to to to to to to to to to
0.80] 0.80] 0.83] 0.80] 0.79] 0.76] 0.71] 0.81] 0.83] 0.84] 0.80] 0.57] 0.2
0.3
0.4
0.5
0.6
Standardized Mean Difference FIGURE 5 A one-study leave-out sensitivity analysis.
0.7
0.8
694
RBMO VOLUME 39 ISSUE 4 2019
TABLE 3 META-ANALYSIS OF CORRELATIONS Correlation of KISS-1 with
Studies, n
Heterogeneity I2 (%)
Sample size
r
95% CI
Z
P-value
Age
7
68.86
299
0.02
–0.14 to 0.19
0.27
NS
BMI
7
29.57
628
0.02
–0.08 to 0.12
0.33
NS
WC
3
80.30
125
0.12
–0.31 to 0.50
0.53
NS
WHR
3
0.00
364
0.04
–0.06 to 0.14
0.81
NS
FBG
4
81.91
500
–0.08
–0.30 to 0.15
–0.68
NS
Insulin
5
79.15
562
0.07
–0.12 to 0.26
0.71
NS
HOMA-IR
5
71.87
562
0.07
–0.10 to 0.23
0.74
NS
FAI
6
89.78
609
0.25
–0.017 to 0.48
1.83
BS
AMH
5a
65.77
380
0.25
0.0245 to 0.452
2.16
0.03b
Total testosterone
7
37.36
633
0.12
0.088 to 0.30
3.51
<0.001b
Free testosterone
2
5.45
299
0.10
–0.006 to 0.19
1.83
BS
LH
8
97.15
650
0.46
–0.004 to 0.76
1.94
BS
FSH
7
29.60
588
0.01
–0.09 to 0.12
0.18
NS
LH/FSH
2
0.00
304
0.04
–0.07 to 0.15
0.67
NS
DHEAS
4
0.00
477
0.13
0.04 to 0.22
2.85
0.004b
SHBG
5
91.28
562
0.12
–0.17 to 0.41
0.81
NS
AMH, anti-Müllerian hormone; BMI, body mass index; BS, borderline significance (P = 0.05/0.06); DHEAS, dehydroepiandrosterone; FAI, free androgen index; FBG, fasting blood glucose; HOMA-IR, homeostatic model assessment-insulin resistance; NS, non-significant; SHBG, sex hormone binding globulin; WC, waist circumference, WHR, waist-hip ratio. a One b P
study (El-Shehawy and Safan, 2015) is counted twice as the correlations between KISS-1 and AMH were reported separately in PCOS and control groups.
< 0.05 is statistically significant.
TABLE 4 QUADAS-2 CRITICAL APPRAISAL Studies
Design
Patient Threshselectiona oldb
Blinded Valid Disease Verifica- Withindex test reference progres- tionf drawalg c d e results test sion
Risk of bias
RepreExtracta- Applisentative ble datai cability patient concern sampleh
El-Shehawy Case-control and Safan, study 2015
•
•
○
•
Moderate
•
•
Low
Kaya et al., Observa• 2018 tional cohort study
•
•
○
•
Moderate
•
•
Low
Yilmaz Prospective, • et al., 2014 case-control clinical study
•
•
○
•
Low
•
•
Low
a •
Consecutive order and well described in and exclusion criteria;
b •
Based on receiver operator characteristic analysis;
c •
Yes;
d •
Rotterdam criteria;
inappropriate exclusions. |
not reported.
index test was performed by biochemistry departments in papers involving both obstetrics and biochemistry authors;
e • Severe; f •
case–control with consecutive case selection;
pre-specified;
other lab findings;
no/not reported.
none/not reported.
mild to severe; not reported.|
All patients received both index and reference test. Reference test was the same for all patients;
selected patients received equal reference tests;
received different reference tests. g • h • i •
No loss to follow up;
loss to follow up, reasons given;
Patients with PCOS;
healthy control group;
2 × 2 table data extractable;
loss to follow up without reasons given/not reported. |
non-matching domain. |
levels of biomarkers reported, no 2 × 2 data extractable;
only correlation, no data on PCOS.
selected patients
RBMO VOLUME 39 ISSUE 4 2019
FIGURE 6 The results of diagnostic test accuracy meta-analysis. (A) Sensitivity; (B) specificity; (C) diagnostic odds ratio; (D) summary receiver
operator characteristic curve for kisspeptin-1.
695
696
RBMO VOLUME 39 ISSUE 4 2019
ACKNOWLEDGEMENTS
REFERENCES
Dr Varikasuvu SR especially acknowledges Varikasuvu Sahasra Bhairavi for the time I could not give you during this work. We acknowledge Dr K Sreedhar (PhD in Cardiology, SVIMS University, Tirupati) for his help in preparing figures with correct resolution.
Albalawi, F.S., Daghestani, Maha H, Daghestani, Mazin H, Eldali, A., Warsy, A.S., rs4889 polymorphism in KISS1 gene, its effect on polycystic ovary syndrome development and anthropometric and hormonal parameters in Saudi women. 2018; 1–6. Brown, R.E., Wilkinson, D.A., Imran, S.A., Caraty, A., Wilkinson, M. Hypothalamic kiss1 mRNA and kisspeptin immunoreactivity are reduced in a rat model of polycystic ovary syndrome (PCOS). Brain Res. 2012; 1467: 1–9. https://doi. org/10.1016/J.BRAINRES.2012.05.049 Chen, X., Mo, Y., Li, L., Chen, Y., Li, Y., Yang, D. European Journal of Obstetrics & Gynecology and Reproductive Biology Increased plasma metastin levels in adolescent women with polycystic ovary syndrome. Eur. J. Obstet. Gynecol. 2010; 149: 72–76. https://doi. org/10.1016/j.ejogrb.2009.11.018 Daghestani, M.H. Evaluation of biochemical, endocrine, and metabolic biomarkers for the early diagnosis of polycystic ovary syndrome among non-obese Saudi women. Int. J. Gynecol. Obstet. 2018; 142: 162–169. https:// doi.org/10.1002/ijgo.12527 El-Shehawy, Y.M., Safan, M.A., Evaluation of Serum Anti-Mullerian Hormone (AMH) and Plasma Metastin Levels in Women with Polycystic Ovary Syndrome (PCOS). Med. J. Cairo Univ. 2015; 83: 399–406. Funes, S., Hedrick, J.A., Vassileva, G., Markowitz, L., Abbondanzo, S., Golovko, A., Yang, S., Monsma, F.J., Gustafson, E.L. The KiSS1 receptor GPR54 is essential for the development of the murine reproductive system. Biochem. Biophys. Res. Commun. 2003; 312: 1357–1363 Gorkem, U., Togrul, C., Arslan, E., Oruc, A.S, Duman, N.B. Is there a role for kisspeptin in pathogenesis of polycystic ovary syndrome? Gynecol. Endocrinol. 2018; 34: 157–160. https:// doi.org/10.1080 /09513590.2017.1379499 Gottsch, M.L., Cunningham, M.J., Smith, J.T., Popa, S.M., Acohido, B. V., Crowley, W.F., Seminara, S., Clifton, D.K., Steiner, R.A. A Role for Kisspeptins in the Regulation of Gonadotropin Secretion in the Mouse. Endocrinology 2004; 145: 4073–4077. https:// doi.org/10.1210/en.2004-0431 Hameed, S., Jayasena, C.N., Dhillo, W.S. 2006 Kisspeptin and fertility. https://doi.org/10.1677/ JOE-10-0265 Hori, A., Honda, S., Asada, M., Ohtaki, T., Oda, K., Watanabe, T., Shintani, Y., Yamada, T., Suenaga, M., Kitada, C., Onda, H., Kurokawa, T., Nishimura, O., Fujino, M. Metastin Suppresses the Motility and Growth of CHO Cells Transfected with Its Receptor. Biochem. Biophys. Res. Commun. 2001; 286: 958–963 https://doi.org/10.1006/BBRC.2001.5470. Hu, K., Zhao, H., Chang, H., Yu, Y., Qiao, J. Kisspeptin / Kisspeptin Receptor System in the Ovary. 2018; 8, 1–14. https://doi. org/10.3389/fendo.2017.00365 Jeon, Y.E., Lee, K.E., Jung, J.A., Yim, S.Y., Kim, H., Seo, S.K., Cho, S., Choi, Y.S., Lee, B.S. Kisspeptin, Leptin, and Retinol-Binding Protein 4 in Women with Polycystic Ovary Syndrome. Gynecol. Obstet. Invest. 2013; 75: 268–274. https://doi.org/10.1159/000350217 Kaya, C., Alay, İ., Babayeva, G., Gedikbaşı, A., Ertaş, S., Ekin, M., Yaşar, L. Serum Kisspeptin
levels in unexplained infertility, polycystic ovary syndrome, and male factor infertility. 2018; 3590. https://doi.org/10.1080/09513590. 2018.1519792 Lapatto, R., Pallais, J.C., Zhang, D., Chan, Y.-M., Mahan, A., Cerrato, F., Le, W.W., Hoffman, G.E., Seminara, S.B. Kiss1−/− Mice Exhibit More Variable Hypogonadism than Gpr54−/− Mice. Endocrinology 2007; 148: 4927–4936. https://doi.org/10.1210/en.2007-0078 Li, W., Ma, D., Liu, M., Liu, H., Feng, S., Hao, Z., Wu, B., Zhang, S. Association between Metabolic Syndrome and Risk of Stroke: A Meta-Analysis of Cohort Studies. Cerebrovasc. Dis. 2008; 25: 539–547. https:// doi.org/10.1159/000131672 March, W.A., Moore, V.M., Willson, K.J., Phillips, D.I.W., Norman, R.J., Davies, M.J. The prevalence of polycystic ovary syndrome in a community sample assessed under contrasting diagnostic criteria. Hum. Reprod. 2010; 25: 544–551. https://doi.org/10.1093/ humrep/dep399 Matsui, H., Takatsu, Y., Kumano, S., Matsumoto, H., Ohtaki, T. Peripheral administration of metastin induces marked gonadotropin release and ovulation in the rat. Biochem. Biophys. Res. Commun. 2004; 320: 383–388 https://doi.org/10.1016/J.BBRC.2004.05.185. Mumtaz, A., Khalid, A., Jamil, Z., Fatima, S.S., Arif, S., Rehman, R., Kisspeptin: A Potential Factor for Unexplained Infertility and Impaired Embryo Implantation. Int. J. Fertil. Steril. 2017; 11: 99–104. https://doi.org/10.22074/ijfs.2017.4957 Navarro, V.M., Castellano, J.M., FernándezFernández, R., Tovar, S., Roa, J., Mayen, A., Nogueiras, R., Vazquez, M.J., Barreiro, M.L., Magni, P., Aguilar, E., Dieguez, C., Pinilla, L., Tena-Sempere, M. Characterization of the Potent Luteinizing Hormone-Releasing Activity of KiSS-1 Peptide, the Natural Ligand of GPR54. Endocrinology 2005; 146: 156–163. https://doi.org/10.1210/en.2004-0836. Navarro, V.M., Fernández-Fernández, R., Castellano, J.M., Roa, J., Mayen, A., Barreiro, M.L., Gaytan, F., Aguilar, E., Pinilla, L., Dieguez, C., Tena-Sempere, M. Advanced vaginal opening and precocious activation of the reproductive axis by KiSS-1 peptide, the endogenous ligand of GPR54. J. Physiol. 2004; 561: 379–386. https://doi.org/10.1113/ jphysiol.2004.072298 Navarro, V.M., Tena-Sempere, M. Neuroendocrine control by kisspeptins: role in metabolic regulation of fertility. Nat. Rev. Endocrinol. 2012; 8: 40–53. https://doi. org/10.1038/nrendo.2011.147 Nejad, S.Z., Tehrani, F.R., Zadeh-vakili, A., The Role of Kisspeptin in Female Reproduction. 2017; 15, 1–11. https://doi.org/10.5812/ijem.44337. Review Nyagolova, P. V, Mitkov, M.D., Orbetzova, M.M., Terzieva, D.D. Kisspeptin And Galanin-like Peptide (GALP) Levels In Women With Polycystic Ovary Syndrome. 2016; 7–12. Ohtaki, T., Shintani, Y., Honda, S., Matsumoto, H., Hori, A., Kanehashi, K., Terao, Y., Kumano, S., Takatsu, Y., Masuda, Y., Ishibashi, Y., Watanabe, T., Asada, M., Yamada, T., Suenaga, M., Kitada, C., Usuki, S., Kurokawa, T., Onda, H., Nishimura, O., Fujino, M. Metastasis suppressor gene KiSS-1 encodes peptide ligand of a G-protein-coupled receptor. Nature 2001; 411: 613–617. https://doi. org/10.1038/35079135
Ozay, O.E., Ozay, A.C., Acar, B., Cagliyan, E., Seçil, M., Küme, T., Ozay, O.E., Ozay, A.C., Acar, B., Cagliyan, E., Sec, M. Role of kisspeptin in polycystic ovary syndrome. 2016; 3590. https://doi.org/10.3109/09513590.2016.1161019 Panidis, D., Farmakiotis, D., Rousso, D., Katsikis, I., Kourtis, A., Diamanti-Kandarakis, E. Serum luteinizing hormone levels are markedly increased and significantly correlated with Δ4-androstenedione levels in lean women with polycystic ovary syndrome. Fertil. Steril. 2005; 84: 538–540. https://doi.org/10.1016/j. fertnstert.2005.02.017 Panidis, D., Ph, D., Rousso, D., Koliakos, G. Plasma metastin levels are negatively correlated with insulin resistance and free androgens in women with polycystic ovary syndrome. 2006; 85, 0–5. https://doi.org/10.1016/j. fertnstert.2005.11.044 Rosenfield, R.L., Ehrmann, D.A. The Pathogenesis of Polycystic Ovary Syndrome (PCOS): The Hypothesis of PCOS as Functional Ovarian Hyperandrogenism Revisited. Endocr. Rev. 2016; 37: 467–520. https://doi.org/10.1210/ er.2015-1104
RBMO VOLUME 39 ISSUE 4 2019
Rotterdam ESHRE/ASRM-Sponsored PCOS consensus workshop group. Revised 2003 consensus on diagnostic criteria and longterm health risks related to polycystic ovary syndrome (PCOS). Hum. Reprod. 2004; 19: 41–47 Seminara, S.B., Messager, S., Chatzidaki, E.E., Thresher, R.R., Acierno, J.S., Shagoury, J.K., Bo-Abbas, Y., Kuohung, W., Schwinof, K.M., Hendrick, A.G., Zahn, D., Dixon, J., Kaiser, U.B., Slaugenhaupt, S.A., Gusella, J.F., O'Rahilly, S., Carlton, M.B.L., Crowley, W.F., Aparicio, S.A.J.R., Colledge, W.H. The GPR54 Gene as a Regulator of Puberty. N. Engl. J. Med 2003; 349: 1614–1627. https://doi.org/10.1056/ NEJMoa035322 Umayal, B., Jayakody, S., Chandrasekharan, N., S Wijesundera, W., Wijeyaratne, C. Polycystic ovary syndrome (PCOS) and kisspeptin – A Sri Lankan study. J. Postgrad. Med. 2019; 65: 18–23. https://doi.org/10.4103/jpgm. JPGM_683_17 Whiting, P.F., Rutjes, A.W.S., Westwood, M.E., Mallett, S., Deeks, J.J., Reitsma, J.B., Leeflang, M.M.G., Sterne, J.A.C., Bossuyt, P.M.M.
697
QUADAS-2 Group. QUADAS-2: A Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies. Ann. Intern. Med. 2011; 155: 529. https://doi.org/10.7326/0003-4819-155-8201110180-00009 Witchel, S.F., Tena-Sempere, M. The Kiss1 system and polycystic ovary syndrome: lessons from physiology and putative pathophysiologic implications. Fertil. Steril. 2013; 100: 12–22. https://doi.org/10.1016/j.fertnstert.2013.05.024. Yilmaz, S.A., Kerimoglu, O.S., Pekin, A.T., Incesu, F., Dogan, N.U., Celik, C., Unlu, A. European Journal of Obstetrics & Gynecology and Reproductive Biology Metastin levels in relation with hormonal and metabolic pro fi le in patients with polycystic ovary syndrome. Elsevier Irel. Ltd 2014; 180: 56–60. https://doi. org/10.1016/j.ejogrb.2014.06.004 Received 6 February 2019; received in revised form 10 April 2019; accepted 25 April 2019.