This prospective, case control study was conducted at the Department of Gynecology and Obstetrics of a tertiary center.

The sample size of the study was calculated using the G*Power (G*Power Ver. 3.1.9.2, Franz Faul, Üniversität zu Kiel, Kiel, Germany) statistical package. The required sample sizes for α=0.05 and Zα=1.96, and β=0.10 and Zβ=−1.28 was calculated for each group. The study and control groups included 27 and 24 consecutive patients who underwent dilation and curettage at <10 weeks of gestation due to first trimester spontaneous pregnancy loss and for elective termination, respectively, between July 2017 and January 2018. Elective pregnancy termination is possible before gestational week 10 in Turkey. Therefore, we included first-trimester pregnancy losses only before gestational week 10.

Spontaneous pregnancy loss was diagnosed via transvaginal ultrasound and clinical assessment. We excluded patients with a retroplacental hematoma, active maternal infection, systemic lupus erythematosus, hypertension, diabetes, chronic renal insufficiency, alcohol consumption, a smoking habit (cigarettes), or a history of ≥2 spontaneous abortions. According to the guidelines of the Royal College of Obstetricians and Gynaecologists, recurrent pregnancy loss is defined as a total of ≥3 abortions [14]. We analyzed the karyotypes of all patients, and excluded those with abnormalities to unmask and avoid possible effects of chromosomal abnormalities on kisspeptin expression between the groups. Patients with vaginal bleeding and hematomas were not included due to low risk. As much as possible, we aimed to select only patients with healthy pregnancies for the control group.

Curettage specimens from 27 spontaneous pregnancy losses and 24 elective pregnancy terminations histopathologically diagnosed as “decidua showing the Arias-Stella reaction and chorionic villi” at the Faculty of Medicine, Department of Pathology between 2001 and 2003 were collected and archival hematoxylin and eosin stained sections were reevaluated. Kisspeptin was immunohistochemically assessed in tissue samples containing decidua and chorionic villi without areas of hemorrhage and necrosis.

Sections were incubated at 56°C for 24 hours, deparaffinized in xylene, then rehydrated using a descending alcohol series (100-70%). Endogenous peroxidase activity was blocked by immersing the sections for 5 minutes in methanol containing 0.5% hydrogen peroxide. Antigens were retrieved by microwave-heating the sections in trisodium citrate buffer (10 mM sodium citrate, pH 6.0) for 10 minutes. The sections were rinsed 3 times in de-ionized distilled water, immersed in 0.3% hydrogen peroxidase for 30 minutes to block endogenous peroxidase activity, then washed in PBS for 2-3 minutes. Non-specific background staining was minimized using Ultra V Block (Thermo Fisher Scientific Inc., Waltham, MA, USA) for 5 minutes. The sections were then incubated for 1 hour at room temperature with the mouse monoclonal primary antibody to KISS-1, 24-Q (Santa Cruz Biotechnology Inc., Dallas, TX, USA) diluted 1:100. This antibody was raised against a partial recombinant protein mapped within amino acids 46-145 of KISS-1 of human origin. After a wash with PBS for 5 minutes, Goat Anti-Polyvalent (Lab Vision Corp., Fremont, CA, USA) was applied, then the sections were washed in PBS. The sections were incubated for 15 minutes at room temperature with streptavidin peroxidase (Lab Vision Corp.). Immunoreactivity was visualized using the chromogen, DAB (Thermo Fisher Scientific Inc.). The sections were counterstained with Mayer’s hematoxylin solution and mounted. Staining specificity was confirmed by comparison with a positive control.

Two pathologists who were blinded to the pregnancy stage and outcomes analyzed all sections. Immunohistochemical staining was evaluated by scanning whole sections for antibodies using a BX51 light microscope (Olympus Optical Co., Ltd., Tokyo, Japan) at 4×, 10×, 20×, and 40× magnification. KISS-1 proteins were mainly located in the glandular epithelium, decidualized stromal cells, cytotrophoblasts, and syncytiotrophoblasts.

The results are expressed as ratios (%) of positivity based on staining intensity scored as 0 (negative), 1+ (low), 2+ (moderate), and 3+ (strong). Ratios (%) of stained cells were: 0 (<10%), 1 (10-25%), 2 (10-50%), 3 (51-80%), and 4 (>80%). The final score was calculated as the product of the ratio of stained cells and staining intensity, resulting in weak (0-2), moderate (3-6), and strong (8) KiSS-1 expression [15].

Variables were analyzed using SPSS 22.0 (IBM Corp., Armonk, NY, USA) and MedCalc 14 (MedCalc Software Ltd., Ostend, Belgium) software. The normality of data distribution was evaluated using Shapiro-Wilk tests. The homogeneity of variances was assessed using Levene tests. Pairs of independent groups were compared using independent samples t-tests with Bootstrap results, and Mann-Whitney U-tests with Monte Carlo results. Categorical variables were compared using Pearson χ² tests with exact results. Correlations between KISS-1 and age, number of births, parity, and abortus were assessed using Spearman rho tests. The sensitivity and specificity of relationships between classifications based on cutoff values calculated according to the number of abortions and KISS-1 variables for the 2 groups of patients, and the actual classification were analyzed using receiver operating characteristics curves. Diagnostic-dependent variables with explanatory variables were estimated using logistic regression analysis, K-Nearest Neighbor, and Decision Tree. However, the results were not significant and are not shown. Quantitative variables are expressed as mean±standard deviation, range (maximum-minimum), and medians. Categorical variables are expressed as number (%). The variables were analyzed at 95% confidence intervals, and P<0.05 was considered statistically significant.