Estradiol and Progesterone Levels are Related to Redox Status in the Follicular Fluid During in vitro Fertilization.

A total of 196 FF samples of women undergoing the ICSI procedure were collected and submitted to spectrophotometry analysis for their blood levels, and this analysis resulted in 163 blood-free FF samples enrolled for further analysis ( Fig. 1B ). The mean patient age was 36.25 ± 4.32 years (22-48), the mean height was 1.63 ± 0.06 meters (1.46-1.82), and the mean BMI was 23.31 ± 3.82 kg/m 2 (15.60-38.06). More than 70% of the patients were 31 to 40 years of age and had a normal BMI (18.5-24.9 kg/m 2 ). The 3 main causes of infertility diagnosis were unexplained infertility (30.06%; n = 49), male factor (25.77%; n = 42), and tubal factor (13.5%; n = 22). The distribution of women (n = 163) by age, BMI, and infertility diagnosis are illustrated in Fig. 2 . Main demographic parameters of the women enrolled in the study. (A.) Age (years); (B.) body mass index (BMI, kg/m 2 ); and (C.) infertility diagnosis. The FF estradiol level was positively correlated with the total FSH dose ( r = 0.165; P = 0.019; n = 158, Fig. 3A ), total LH dose ( r = 0.173; P = 0.018; n = 149, Fig. 3B ), and serum estradiol concentration on the day of hCG administration ( r = 0.482; P < 0.0001; n = 117, Fig. 3C ) and negatively correlated with the serum progesterone dose on the day of hCG administration ( r = − 0.183; P = 0.023; n = 119, Fig. 3D ). However, the FF progesterone level did not exhibit a correlation with the same parameters as observed for the FF estradiol level ( Fig. 4A ). The FF progesterone level was negatively correlated with the number of follicles punctured ( r = −0.246; P = 0.001; n = 157, Fig. 5A ), the number of oocytes retrieved ( r = −0.268; P < 0.0001; n = 157, Fig. 5B ), FF HPSC ( r = −0.294; P < 0.0001; n = 157, Fig. 5C ) and FF SOD ( r = −0.237; P = 0.002; n = 152, Fig. 5D ) and was positively associated with reduced thiol levels in FF ( r = 0.209; P = 0.004; n = 157, Fig. 5E ). No correlation was observed between the FF progesterone and NO 2 concentrations ( Fig. 5F ). Correlations between the estradiol level in FF with (A.) total follicle-stimulating hormone (FSH) used during the ovulation induction cycle; (B.) total luteinizing hormone (LH) used during the ovulation induction cycle; (C.) serum estradiol on the day of hCG administration; and (D.) serum progesterone on the day of hCG administration. Abbreviations: FF, follicular fluid; hCG, human chorionic gonadotropin; n, number of samples; p, P value; r, Spearman’s rho. Correlations between the progesterone level in FF with (A.) serum estradiol on the day of hCG administration and (B.) serum progesterone on the day of hCG administration. Abbreviations: FF, follicular fluid; hCG, human chorionic gonadotropin; n, number of samples; p, P value; r, Spearman’s rho. Correlations between the progesterone level in FF with (A.) number of follicles punctured; (B.) number of oocytes retrieved; (C.) hydrogen peroxide scavenging capacity (FF HPSC); (D.) superoxide dismutase (FF SOD); (E.) reduced thiol (FF thiol); and (F.) FF NO 2 . Abbreviations: FF, follicular fluid; FSH, follicle-stimulating hormone; hCG, human chorionic gonadotropin; n, number of samples; p, P value; r, Spearman’s rho. K-means clustering was used to detect possible differences in follicle and oocyte numbers related to oxidative stress biomarkers. The levels of progesterone and oxidative stress biomarkers were divided into 3 groups (low, medium, and high), and the numbers of follicles punctured and oocytes retrieved were plotted ( Figs. 6 , 7 ). The number of punctured follicles was higher in the high FF HPSC group than in the low group ( Fig. 6A ). No significant differences were observed related to SOD, reduced thiols, TBARS, and NO 2 ( Fig. 6B to E). Interestingly, the number of follicles punctured was higher in the low FF progesterone concentration group than in the medium group ( Fig. 6F ). Number of total follicles punctured after ovarian stimulation. (A.) Hydrogen peroxide scavenging capacity (HPSC); (B.) superoxide dismutase (SOD); (C.) thiol; (D.) thiobarbituric acid reactive substances (TBARS); (E.) NO 2 ; and (F.) Follicular fluid (FF) progesterone. Oxidative stress biomarkers and intrafollicular steroid hormones were stratified into 3 groups: low (●), medium (■), and high (▲). ** P < 0.01. TBARS results are expressed as ng malondialdehyde (MDA)/μL. Number of total oocytes retrieved after ovarian stimulation. (A.) Hydrogen peroxide scavenging capacity (HPSC); (B.) superoxide dismutase (SOD); (C.) thiol; (D.) thiobarbituric acid reactive substances (TBARS); (E.) NO 2 ; and (F.) Follicular fluid (FF) progesterone. Oxidative stress biomarkers and intrafollicular steroid hormones were stratified into 3 groups: low (●), medium (■) and high (▲). * P < 0.05, ** P < 0.01, *** P < 0.001. TBARS results are expressed as ng malondialdehyde (MDA)/μL. The number of retrieved oocytes was higher in the high FF HPSC group than in the low and medium groups ( Fig. 7A ). No significant differences were observed related to SOD, reduced thiols, TBARS, or NO 2 ( Fig. 7B to E). The number of oocytes was higher in the low FF progesterone concentration group than in the other groups (medium and high FF progesterone) ( Fig. 7F ). As we did not follow the exact oocyte from the studied FF to the embryo transfer process, it was not our aim to evaluate pregnancy rates. However, as shown in Table 3 , the women were divided into 2 groups according to whether they were pregnant (those who became pregnant after embryo fresh transfer) or nonpregnant (those who did not become pregnant after embryo fresh transfer). In our study, 127 women had the embryo transferred in the same cycle of the follicular puncture (41 women became pregnant and 86 women did not become pregnant. Pregnant women were associated with significantly higher FF estradiol levels than nonpregnant women ( P < 0.05). Embryo transfer occurred on days 3 or 5. Biochemical pregnancy was ascertained though positive β-hCG 15 days after embryo transfer, and clinical pregnancy was confirmed by ultrasound at 6 to 7 weeks of gestation. Data concerning all correlations of oxidative stress markers, antioxidant proteins, and hormone levels in FF versus clinical data (i.e., age, BMI, serum hormones on day of oocyte retrieval, numbers of follicles and oocytes retrieved) are presented in Tables 1 , 2 . Correlations Detected Between Hormone Levels and Clinical Data Bold values indicate statistically significant relationships. N = 114-163. Abbreviations: BMI, body mass index; E2, estradiol; FF, follicular fluid; FSH, follicle-stimulating hormone; hCG, human chorionic gonadotropin; LH, luteinizing hormone; P4, progesterone. Correlations Detected Between Oxidative Stress Biomarkers in Follicular Fluid and Hormone Levels Bold values indicate statistically significant relationships. N = 114-163. Abbreviations: E2, estradiol; FF, follicular fluid; FSH, follicle-stimulating hormone; hCG, human chorionic gonadotropin; HPSC, hydrogen peroxide scavenging capacity; LH, luteinizing hormone; P4, progesterone; SOD, superoxide dismutase; TBARS, thiobarbituric acid reactive substances.

Women undergoing ICSI treatment at Fertipraxis Clinic (Rio de Janeiro, RJ, Brazil), a human reproduction clinic, between September 2012 and February 2014 were enrolled in the study. This center for human reproduction is certified by the Brazilian Health Regulatory Agency (ANVISA) and the Latin American Network of Assisted Reproduction (REDLARA). The study was conducted with the approval of the Research Ethics Committee of the Maternity Hospital of Federal University of Rio de Janeiro and was registered with the Brazilian platform of research under the following numbers: CAAE: 02213812.4.0000.5275 and the number of purpose: 91.930; this study was approved on September 6, 2012. All participants provided written informed consent, and the procedures were in accordance with the Declaration of Helsinki. A total of 196 women undergoing ICSI with different infertility diagnoses were recruited to participate in the study, regardless of ethnicity, age, cause of infertility, or diagnosis of their male partner’s infertility. It is important to note that ICSI is the most utilized ART in Brazil regardless of the male infertility factor; thus, ICSI was the ART technique chosen for the study. Clinical data were obtained from medical records and included patient age, BMI, height, size of the first follicle punctured, number of follicles, oocytes retrieved, embryo quality, and pregnancy and other relevant data ( Fig. 1 and Table 3 ). The serum hormone levels were obtained from medical records at different times as described in Fig. 1A . Characteristics of Women Who Achieved Pregnancy (n = 41) Compared With Nonpregnant Women (n = 86) Mann-Whitney U test was used in analyses unless otherwise indicated. Grade 1 = equally sized blastomeres, no fragmentation; grade 2 = equally or unequally sized blastomeres with 50% fragmentation. Abbreviations: BMI, body mass index; FF, follicular fluid; HPSC, hydrogen peroxide scavenging capacity; MDA, malondialdehyde; SOD, superoxide dismutase; TBARS, thiobarbituric acid reactive substances; UFADEB, unfertilized, anomalous, or degenerate embryo. #Chi-square test (#) ##Unpaired t-test (##) *Significant difference ( P < 0.05) (A.) Schematic of the ovarian stimulation protocol. Two to 3 days after the start of menses, the ovarian stimulation protocol (FSH or FSH + LH) started based on the clinician criteria. Pituitary suppression began when the first follicle reached 14 mm in diameter. Recombinant hCG was given when at least one follicle reached 18 mm. Hormone serum levels were obtained before the ovarian stimulation and before hCG administration. Levels of estradiol, progesterone, HPSC, SOD, reduced thiol (Thiol), TBARS, and NO 2 were evaluated in the follicular fluid (FF). (B.) Flowchart demonstrating the recruitment of patients undergoing IVF/ICSI and FF sample admission or exclusion. (C.) Spectrum for blood contamination evaluation in FF samples. The spectrum 350-650 nm for blood contamination evaluation in FF samples shows different blood concentrations diluted in deionized water (1%, 0.5%, 0.25%, and 0.1%) from a blood sample. Note: 12.1 g/dL hemoglobin = 100% blood elements. Only deionized water (water); 1 FF sample without blood contamination, yellow line (177 FF); 1 FF sample with blood contamination, red line (74 FF). Abbreviations: HPSC, hydrogen peroxide scavenging capacity; IVF/ICSI, in vitro fertilization/intracytoplasmic sperm injection; SOD, superoxide dismutase; TBARS, thiobarbituric acid reactive substances. Controlled ovarian stimulation was performed according the clinical protocol as previously described and according to the specific clinical requirements of the patients [ 21 ]. Briefly, on the second to third day of menstruation, ovarian stimulation was initiated with synthetic follicle-stimulating hormone (FSH) alone (Gonal-F, Merck-Serono, Italy; or Bravelle, Ferring Pharmaceutical, Germany) or FSH and luteinizing hormone (LH) (Pergoveris, Merck-Serono, Italy; or Menopur, Ferring Pharmaceutical, Germany) treatments. FSH dosages varied from 150 to 300 IU/day, and LH dosages ranged from 75 to 300 IU/day. The gonadotropin-releasing hormone antagonist cetrorelix acetate (Cetrotide 0.25 mg, Merck-Serono, Italy) was administered to induce hypophysis suppression whenever the first follicle was ≥14 mm. When at least one follicle had reached 18 mm or at least 2 follicles had reached 16 mm (assessed by ultrasound), human chorionic gonadotropin (hCG) (Ovidrel 250 μg, Merck-Serono, Italy) was administered to mimic the LH peak. Thirty-five hours post-Ovidrel administration, the oocytes were retrieved, and FF was obtained during the follicular aspiration procedure. Briefly, follicular aspiration was undertaken using a transvaginal ultrasound probe (Medison SonoAce X8) and a 17G single lumen oocyte aspiration needle (Wallace) for ≥5 follicles with a diameter of at least 16 mm or a 17G double lumen oocyte aspiration needle (Swemed, Vitrolife) for <5 follicles with a 16 mm diameter connected to a vacuum system used for follicular aspiration; after oocyte retrieval by the embryologist, the FF was placed in a cryopreservation tube (Nunc, Thermo Scientific) and stored in a liquid nitrogen container. Approximately 4 to 6 hours after oocyte retrieval, the ICSI procedure was performed. The FF from each patient was collected following puncture of just one follicle with the largest diameter (>17 mm) to reduce blood contamination via rupture of small blood vessels during follicular aspiration ( Fig. 1 ). The FF sample for each patient was centrifuged at 12000 g for 15 minutes at 4°C, and the FF supernatant was sorted into aliquots for different assays to avoid possible losses associated with freezing and thawing cycles [ 22 ] and were stored at −80°C until the assay. The exclusion criteria were as follows: smoking, positive for hepatitis B virus (HBV), hepatitis C virus (HCV), human immunodeficiency virus (HIV), or thyroid dysfunction. Few studies have used a clear methodology for the exclusion of blood-contaminated FF samples that can alter the biochemical profile of the specimen [ 23 , 24 ]; most studies use visual inspection or do not mention the criterion used [ 10 , 13 , 14 , 18-20 , 25-40 ]. Here, the presence or absence of blood contamination was graded by visual inspection and spectrophotometric analysis. Although visual analysis can determine which FF samples need to be discarded due to a high degree of blood contamination [ 41 ], this methodology is subjective. Spectrophotometric analysis was performed to detect the presence of hemoglobin in FF samples ( Fig. 1C ). However, there is no consensus or definition on the acceptable limit of contamination with blood elements in the FF sample with the use of this technique [ 8 , 42 ]. Thus, a dilution curve with human blood samples (collected into EDTA tubes from one female healthy volunteer and measured using an automated colorimetric assay) with an initial concentration of 12.1 g/dL was performed. The oxygenated hemoglobin showed the typical maximal absorption at 415 nm and the double peak at 540 and 575 nm [ 43 ], so the absorbances of the blood and FF samples were measured in the range of 350 to 650 nm using a spectrophotometer (Spectramax Paradigm, Molecular Devices) and analyzed by Softmax Pro 6.2.1 software (Molecular Devices, Inc.). The spectrum was analyzed, and the absorbance values at 415 nm were used to determine the degree of contamination. All FF samples without culture medium contamination and with less than 0.035 g/dL of hemoglobin (less than 0.3% of blood elements), which showed a single absorbance peak of approximately 450 nm “uncontaminated fluid” (n = 163), were included for further analysis [ 42 ]. All experiments were performed in duplicate. The FF 17β-estradiol levels were determined using an ImmuChem Coated Tube 17β-Estradiol iodine 125 ( 125 I) RIA Kit (MPBio, cat# 07-238105, using the antibody MP Bio 07-238110, RRID:AB_2847903) [ 44 ], whereas the FF progesterone levels were determined using ImmuChem Double Antibody Progesterone 125 I RIA Kit (MPBio, cat#, 07-170102, using the antibody MP Bio 07-270110, RRID:AB_2847904) [ 45 ], according to the manufacturer’s instructions. For the measurement of 125 I, an automatic gamma counter (PerkinElmer Wizard 1470) was used. The FF samples were diluted 1:2000 v/v for 17β-estradiol and 1:3000 v/v for progesterone assays, both diluted in steroid-free gelatin phosphate buffer (0.1 M PBS pH 7.0, containing 5% BSA, 1% gelatin, 0.1% sodium azide). The results were expressed as pg/mL or ng/mL, respectively. Sensitivity of the 17β-estradiol assay was 10 pg/mL and intra-assay variation was 5.6%. Sensitivity of the progesterone assay was 0.35 ng/mL and intra-assay variation was 4.9%. All samples were measured in the same assay. All experiments were performed in duplicate. Malondialdehyde (MDA) [ 46 ] levels (a product of fatty acid peroxidation) were measured in FF by the thiobarbituric acid (TBA) assay [ 47 ]. Briefly, the mixture containing TBA reagent was prepared (10 µL, sodium dodecyl sulfate [SDS] 8.1%; 74 µL, trichloroacetic acid [TCA] 10%; 74 µL, TBA 0.8%, and 32 µL, distilled water) and mixed with 10 µL of each thawed FF sample. Then, the mixture was incubated in a water bath at 95°C for 1 hour and cooled on ice. TCA is used for protein precipitation, and TBA reacts with MDA to give a red compound, which can be read at 532 nm. Butanol was used to extract the red compound prior to spectrophotometer measurements (Asys UVM340, Bedfordshire, UK). The results were expressed as MDA ng/µL. Sensitivity was 9 μmol/mL. All samples were measured in the same assay. All experiments were performed in duplicate. The superoxide dismutase (SOD) activity was assayed using a colorimetric assay kit (SOD Assay kit, Cayman Chemical Company, catalog number 706002) following the manufacturer’s instructions. The absorbance was measured at 450 nm in a spectrophotometer (Asys UVM340, UK). One unit of SOD is defined as the amount of enzyme needed to exhibit 50% dismutation of the superoxide radical. The results were expressed as SOD activity units per mL (U/mL). The range of the assay was from 0.005 to 0.050 U SOD/mL. All samples were measured in the same assay. Intra-assay coefficient of variation was 3.2%. All experiments were performed in duplicate. The HPSC was evaluated in FF samples following adaptation of Aebi’s method [ 48 , 49 ]. Briefly, the FF samples were diluted (1:75) in Tris-HCL 0.1 mol/L pH 7.5 buffer. Then, 250 µL of 40 µM H 2 O 2 solution was added to 250 µL of diluted sample or the catalase standard curve standards, and all samples were incubated at 37°C for 30 minutes. Then, 50 µL of samples and standards were placed on 96-well microplates in duplicate and 100 µL of a solution containing 0.4 U/mL horseradish peroxidase (HRP), Tris-HCL 0.1 mol/L pH 7.5 buffer, and Ampliflu Red (Sigma-Aldrich). In this reaction, HRP uses Ampliflu Red as an electron donor during the reduction of H 2 O 2 to water, and fluorescent resorufin is formed. Resorufin has absorption and fluorescence emission maxima of approximately 571 nm and 585 nm, respectively, and it was measured with a spectrofluorometer (Victor X42030 Multilabel Reader, PerkinElmer) at 37°C. The results were expressed as arbitrary units of intensity (A.U.). The samples were measured in 3 different assays. All experiments were performed in duplicate. The concentration of nitrite (NO 2 ) in FF was measured through the detection of nitric oxide (NO) in the gas phase by the chemiluminescence of this reaction with ozone, producing energy in the form of light, which is proportional to the level of NO in the samples, using a Nitric Oxide Analyzer (NOA, Sievers 280i, General Electric). First, the calibration curve was constructed using a NaNO 2 solution (0.125, 0.25, 0.5, 1, and 2.5 µM) as the standard solution. The FF samples (100 µL) underwent deproteinization with 100% cold ethanol (200 µL), mixed and centrifuged at 1000 g for 10 minutes, and the supernatant (20 µL) was injected into a glass purge vessel, as previously described [ 50 ]. The FF NO 2 concentration was determined by interpolation of the NO 2 standard curve. The data were expressed in µM. Sensitivity of the assay was 1 picomole. All samples were measured in the same assay. All experiments were performed in duplicate. The reduced thiol group levels in FF were measured using Ellman’s reagent (5,5′-dithiobis-2-nitrobenzoic acid, DTNB). The Ellman method for assaying thiols is based on the reactions of thiols with DTNB [ 51 ]. For each FF sample, 1 blank tube and 2 sample tubes were prepared for reading. The blank tube contained FF sample (20 µL), Tris-EDTA buffer solution pH 8.0 (130 µL), and methanol (830 µL); the sample tubes were prepared with the same reagents as the blank tube plus 10 mM DTNB (10 µL). Then, the samples were mixed, incubated at 25°C for 15 minutes, and centrifuged at 3000 g for 15 minutes. The supernatant was used, and the absorbance was measured at 412 nm by a spectrophotometer (PerkinElmer LambdaBio+, UK). The results were expressed as μmol DTNB reduced/mL. All samples were measured in the same assay. All experiments were performed in duplicate. The total protein concentration of the FF samples was determined by a colorimetric method using a commercial kit (Pierce BCA Protein Assay Kit, Thermo Scientific) according to the manufacturer’s instructions and was measured using a spectrophotometer (Asys UVM340, Bedfordshire, UK). Data are presented as medians, means ± standard deviations. Normal distribution was verified in all data measured with the D´Agostino-Pearson test. Spearman correlation analysis was used for correlations. The Kruskal-Wallis test was used to analyze the differences between the groups. Student’s t test was used for normally distributed continuous variables and the Mann-Whitney U test for nonnormally distributed continuous variables. For qualitative characteristics, the chi-square test was used to determine a significant relationship between 2 nominal variables. To stratify populations with different redox status, each redox marker concentration was subjected to cluster analysis as used before [ 52 ]. Clustering using the K-Means algorithm was divided into 3 clusters: high, medium, and low, using MATLAB 7.5.0 software (MathWorks). Differences were considered to be significant if P < 0.05. Statistical analysis was performed using GraphPad Prism v.6 software (GraphPad Software, Inc) and SPSS v.24 software (IBM Corporation).

Instead of collecting the FF of different follicles, we aimed to focus on the analysis of the largest follicle and its correlation of oxidative stress biomarkers with hormonal parameters. This choice was based on the fact that the larger follicle would have the greater probability to achieve dominance and would be ovulated (normal reproduction process) or be chosen to follow the fertilization procedure. In addition, normally, the largest follicle has an increased capacity to produce hormones and compared with the physiology of smaller follicles, the physiology of the largest follicle would best correspond to the environment of the ovulated oocyte [ 40 ]. The choice of the first follicle is preferable to the pool of different follicles since we were able to individualize the microenvironment of each follicle. In our study, we were able to show that high levels of FF HPSC were positively associated with the highest total number of follicles ( Fig. 6A ) and oocytes punctured ( Fig. 7A ), which suggests a possible protective effect of estradiol and its relationship with a successful outcome. Additionally, the serum estradiol concentration was positively correlated with intrafollicular HPSC ( Tables 1 and 2 ). It has been shown that estradiol is able to protect granulosa cells from oxidative stress-induced apoptosis [ 53 , 54 ]. As HPSC is directly related to the levels of small antioxidants molecules in FF, our results suggest that estradiol stimulates follicular cells to release these molecules to the extracellular milieu. Since ovaries are the main site of estrogen synthesis in premenopausal women [ 55 ], it is plausible that serum estradiol levels positively correlate with the levels of this hormone in FF. As expected, we found a positive correlation between the serum estradiol concentration on the day of hCG administration and the intrafollicular estradiol level ( Fig. 3C ), suggesting that an increase in serum estradiol is accompanied by intrafollicular production of the same hormone, as observed by others [ 56 , 57 ]. We also found a higher concentration of estradiol in the FF of women who became pregnant following ART than in the FF of women for whom the ART treatments were not successful ( Table 3 ). In agreement with our observations, women who became pregnant on in vitro fertilization (IVF) cycles presented with higher levels of estradiol in the FF than those who did not become pregnant [58- 61 ]. However, in a recent study in which the authors evaluated the steroid levels by LC-MS/MS, they did not find any correlation between the FF estradiol levels and pregnancy [ 40 ]. We did not observe differences between pregnant and nonpregnant women in relation to serum estradiol concentrations on the day of hCG administration ( Table 3 ), which is in agreement with most studies [ 10 , 62-65 ], except one [ 66 ]. We also found a positive correlation between the FF estradiol levels and the total FSH dose or total LH dose used during an ovulation induction cycle ( Fig. 3A and 3B ). LH and FSH are both important for estradiol production since FSH induces the enzymatic machinery necessary for estradiol synthesis, and estradiol is not produced until LH stimulates the thecal cells to secrete androgen, which can be converted into estradiol by granulosa cells aromatase [ 67 ]. Sawada and Carlson (1966) demonstrated that superoxide (O 2 •− ) is involved in progesterone release once insertion of SOD into the corpus luteum of rats treated with physiological doses of LH inhibited progesterone secretion [ 68 ]. Moreover, plasma progesterone levels were significantly decreased in SOD1-deficient female mice, but plasma estradiol levels remained unaltered. It is important to highlight that SOD1 deficiency increases O 2 •− availability, which was also observed in the ovary [ 69 ]. Interestingly, in rat corpus luteum cells treated with low doses of LH, the insertion of SOD or catalase dose-dependently inhibited progesterone secretion. However, at high doses of LH, the electroporation of SOD or catalase produced the opposite response, with an increase in progesterone secretion [ 68 ]. Thus, progesterone secretion is likely to depend on the level of O 2 •− produced since a moderate elevation of O 2 •− level is associated with an increase in its secretion, whereas higher levels of O 2 •− appear to decrease or inhibit progesterone [ 68 , 70 ]. Thus, the negative correlation between serum or FF progesterone levels and SOD activity found in the present study reinforce the possible involvement of superoxide on progesterone secretion since lower SOD activity is related to higher levels of superoxide. Reduced thiol groups are present in many biological molecules, including glutathione. In its reduced form, glutathione is the main nonprotein thiol present in animal cells and protects them by removing reactive metabolites by their conjugation with reactive forms, such as H 2 O 2 [ 71 ]. HPSCs reflect the activity of small molecules that have the capacity to detoxify H 2 O 2 . We found a negative correlation between HPSC and progesterone in FF and a positive correlation between the progesterone and thiol levels in FF ( Fig. 5C and 5E ). The correlation between progesterone and reduced thiol groups suggests that progesterone production is decreased by ROS, since higher levels of reduced thiol groups reflects a lower ROS availability. Moreover, it was demonstrated that high concentrations of H 2 O 2 inhibit progesterone production in the culture of luteinized granulosa cells High concentrations of H 2 O 2 inhibit progesterone production in the culture of luteinized granulosa cells [ 72 , 73 ]. We believe that the levels of reduced thiol groups present in FF, together with the SOD and HPSC levels, create a protective effect against reactive metabolites and are involved in the progesterone production/inhibition process since it can detoxify H 2 O 2 . When analyzing the number of follicles and oocytes punctured, we observed a negative association with the FF progesterone levels ( Figs 5A , 5B , 6F , and 7F ). Interestingly, the serum progesterone levels on the day of hCG were positively correlated with the total number of follicles and oocytes punctured ( Tables 1 and 2 ). As with all studies, ours has limitations. First, we did not assess nutrition, physical activity, or psychological status at admission; second, we did not discriminate the analyses for infertility diagnosis or age, due to the low number of cases. One of the strengths of our study was the use of FF samples with very low levels of blood contamination. Certainly, the use of equipment with increased sensitivity in the detection of blood molecules would allow more accurate results [ 74 ]. However, such methods use expensive equipment and require well-trained professionals; therefore, these methods are not fast or inexpensive. It is important to say that a clear and reproducible method needs to be used to discriminate blood-contaminated FF samples to allow for a better comparison of information between different studies. In this context, our study described a rapid, inexpensive, and easily reproducible method to exclude contaminated samples. Regardless of the method used, the most important objective is to define the maximum acceptable amount of blood elements and/or blood present in the sample for a better comparison between different studies. In conclusion, the intrafollicular production of estradiol and progesterone are linked to the redox status of the follicular microenvironment, which reflects the number of oocytes obtained during the ICSI procedure. In fact, there must be adequate levels of hormones, oxidant molecules and antioxidants, and other factors within the follicle that result in a good quality oocyte. In a dynamic environment with hormonal variations throughout the day, understanding the role of each substance will help us to propose new strategies to achieve success in the treatment of infertile couples. Thus, additional studies are required to better understand the follicular microenvironment, contributing to improvement in the ICSI procedure and decreased costs for patients with infertility.