{"paper_id":"69a840bd-529a-4a1a-b711-99a86cb17ab3","body_text":"Infertility affects 15 to 20% of the population of reproductive age ( Vander Borght & Wyns, 2018 ). The Brazilian\nInstitute of Geography and Statistics (IBGE) counted the Brazilian population at 211\nmillion in June 2020. Of these, 52.97% are of reproductive age, accounting for 111,7\nmillion of the population. Thus, 16,765,000 people in Brazil are infertile ( IBGE, 2020 ).\nPart of this population will need in vitro fertilization (IVF), a treatment that\nstill has limited access to qualified professionals and is expensive ( Inhorn & Patrizio, 2015 ). Personalized care\nfor these patients during treatment is required. The greater the amount of\ninformation that increases the success rates, the better the public will be served,\nin addition to also benefiting the assistant physician in the clinical practice of\nIVF.\nA good embryo and endometrium, as well as adequate hormone levels, are needed to\nachieve pregnancy, with progesterone (P4) being a contributing factor. Produced by\nthe corpus luteum, P4 is essential in the second phase of the menstrual cycle ( Elgindy  et al ., 2018 ), which\nfacilitates embryonic fixation ( Drakopoulos\n et al ., 2019 ). P4 values depend on the period of the\ncycle in which they were collected ( Taylor\n et al ., 2019a ). In IVF, several corpora lutea are\nproduced, but ineffectively, making P4 supplementation necessary ( Fatemi, 2009 ).\nProlactin (PRL) plays a role in implantation, maintenance of pregnancy, and\nlactation. Unlike P4, it does not undergo variations in the menstrual cycle ( Taylor  et al ., 2019a ). The\nincrease in the serum concentrations of PRL can interfere with the proliferation of\ngranulosa cells, leading to a drop in P4, luteal phase insufficiency, and failure of\nembryo implantation ( Kamel  et al .,\n2018 ). In abortions, a decrease in PRL receptors in the endometrium can\nbe observed, suggesting its importance in maintaining pregnancy ( Kim  et al ., 2018 ).\nSome values of hormone dosages are established. On the human chorionic gonadotropin\n(hCG) day, values above 1.5ng/ml indicate early endometrial luteinization, allowing\nthe choice to freeze the embryos and transfer them in another cycle ( Panaino  et al ., 2017 ). In\nfrozen embryo transfer, P4 values greater than 10 ng/ml on Day 4 (D4) of embryonic\nevolution are described as a predictor of success ( Gaggiotti-Marre  et al ., 2019 ).\nNo parameters on the values of the serum concentrations of P4 and PRL at the time\nbetween embryo transfer and the performance of beta-hCG were mentioned in the\nliterature. No values for progesterone on D9 after oocyte retrieval, one week before\nthe pregnancy test, were established. Thus, the aim of this study is to evaluate\nwhether the serum dosage of P4 and PRL, performed on D9 after oocyte retrieval and\nbetween the embryo transfer and pregnancy test, is related to success in IVF and, if\nso, determine the cut-off points.\n\nThis observational, retrospective study analyzed the medical data of patients\nundergoing IVF with fresh embryo transfer in a private clinic in Curitiba,\nBrazil. The investigation was conducted in accordance with the Declaration of\nHelsinki and received a favorable opinion from the ethics committee of the\nHealth Sciences Sector of the Hospital de Clínicas of the Federal\nUniversity of Paraná (HC-UFPR) on October 2, 2018, under number\n2.932.466.\nThe fresh embryo transfer cycles performed between January 2013 and December 2019\nwere reviewed. The data collected were age, infertility factors, endometrial\nthickness on the hCG day, type of pituitary blockage, embryonic quality, number\nof embryos transferred, types and doses of P4 used in luteal phase support,\nserum P4 and PRL values collected on D9 after oocyte puncture, and pregnancy\nsymptoms.\nThe serum concentrations of P4 and PRL of the samples were determined by\nchemiluminescence with the support of the laboratory of the center where the\nresearch was conducted. All cases used the same methodology. P4 and PRL serum\nconcentrations were measured using Architech-Abbott ® , Beckman\nCoulter-Access ® , and Siemens ADVIA\nCentaur ®  equipment.\nCases of IVF cycles and fresh embryo transfer were included, with P4 and PRL\ncollection in the morning, nine days after oocyte retrieval. Cycles with own\noocytes and fresh embryo transfer; trilaminar endometrium between 07 and 13 mm;\ntransfer of at least one Lucinda Veeck Category A or B cleaved embryo ( Veeck, 1999 ) and/or good quality Gardner\nblastocysts ( Gardner & Schoolcraft,\n1999 ); easy embryo transfer; and the necessary serum dosages for the\nstudy were evaluated.\nCases with incomplete data for the purpose of this study, transfer of frozen\nmaterial, category C/D embryos or poor quality blastocysts, transfers with blood\nand/or mucus in the catheter, the use of Pozzi or obturator, and embryos\nretained in the catheter were excluded.\nThe data was analyzed by frequency and contingency tables. Fisher and chi-square\ntests were used for the comparison of nominal data, while an unpaired t-test was\nused for the comparison of numerical data.\nSignificant data was studied in ROC (receive operator characteristic) curves to\nestablish the relationship between the sensitivity and specificity of the\nquantitative diagnostic test.\nA univariate analysis of these parameters was performed in a logistic regression\nmodel. The multivariate analysis observed the relevant criteria among\nthemselves.\nThe Stata/SE v.14.1 programs (StataCorpLP, USA) and GraphPad Prism for Mac OS\nversion 8.4.3 (San Diego, California, USA) ( www.graphpad.com ) was\nused.\n\nA total of 675 IVF cycles with fresh embryo transfer were evaluated. In accordance\nwith the inclusion and exclusion criteria, 330 cycles performed in 293 patients were\nstudied.\nThe mean age of the patients was 35.5±4.1 years. The most common causes of\ninfertility were endometriosis in 24.2% of cases; male factor in 20.6%; and ovarian\nfailure and tubal factor in 15.5% and 15.8%, respectively, in addition to the\ncombinations of factors. The endometrial thickness assessed by ultrasound on the day\nof the trigger was 9.1±1.6 mm.\nPituitary blockage was performed with GnRH (gonadotropin-releasing hormone) analog in\n118 cycles (35.8%); GnRH antagonist in 185 cycles (56.1%); and no blockage in 27\ncycles (8.2%). P4 supplementation was performed intramuscularly (IM) with 50 mg/day\nin 73.9% of the cycles. Micronized natural progesterone (800 mg/day)\n(Utrogestan ® ) was used in 14.6% of the cases. In 11.5% of the\ncases, vaginal gel (90 mg) (Crinone ® ) was used.\nIn 42.2% of the cases, at least one embryo of six to eight cells was transferred,\nwith pattern A. Two to four quality A cells were found in 24.2% and in 6.4% of A\nembryos with 5, 9, 10, or 12 cells. Pattern B with six to eight cells occurred\nin 4.5% of the cycles, two to four cells at 5.8%, and five cells at 0.5%.\nBlastocysts were transferred in 16.4% of the cases. Of these, 88.8% were\nexpanded.\n49.4% of the transfers occurred on D3 of embryonic evolution and 34.2% on D2.\n16.4% of the transfer was in blastocysts (D5). Up to four embryos were\ntransferred per cycle, according to the patient’s age and following the guidance\nof the Federal Council of Medicine (CFM) in force at the time ( Brazil, 2017 ). In 17.6%, a single embryo was\ntransferred. Two embryos were implanted in the majority of cases (61.8%), three\nin 18.2% of cycles, and four embryos in 2.4% of cases.\nTable 1  shows the hormonal serum values\nbetween pregnant and non-pregnant women. A significant difference was observed\nin the P4 concentrations. Values above 32.1 ng/ml were associated with\npregnancy. Regarding the PRL, no significance was observed.\nProgesterone and prolactin in pregnant and non-pregnant women\n(n=330).\nNon-parametric Mann-Whitney test,  p <0.05\nA ROC curve was created, correlating progesterone and pregnancy ( Figure 1a ), with a sensitivity of 53.2% and\nspecificity of 61.4%. The cut-off point established was 30.5 ng/ml.\nFigure 1 A. Association ROC Curve between progesterone (P4) serum assessment\nvalues and pregnancy B.  ROC curve of association between\npregnancy and serum dosage of P4 in cycles with antagonist\nregardless of progesterone for luteal phase support  C.\n ROC curve of association between pregnancy and serum\ndosage of P4 in cycles with antagonist and injectable\nprogesterone use  D.  ROC curve of association\nbetween pregnancy and serum dosage of p4 in cycles without any\nblockage.\nA. Association ROC Curve between progesterone (P4) serum assessment\nvalues and pregnancy B.  ROC curve of association between\npregnancy and serum dosage of P4 in cycles with antagonist\nregardless of progesterone for luteal phase support  C.\n ROC curve of association between pregnancy and serum\ndosage of P4 in cycles with antagonist and injectable\nprogesterone use  D.  ROC curve of association\nbetween pregnancy and serum dosage of p4 in cycles without any\nblockage.\nDue to the statistical relevance of P4, the sample was evaluated by the type of\nblockage and supplementation used and whether pregnancy was achieved or not.\nTable 2  shows that no significant\ndifference in age, number of embryos transferred, endometrial thickness, or\nassociation between P4, PRL, and pregnancy was observed. The serum P4\nconcentrations did not show differences between pregnant and non-pregnant\nwomen according to the type of progesterone administered. As shown in  Table 3 , PRL was found to be associated\nwith pregnancy in users of Crinone ®  when the dose reached\n37.8 ng/ml\nProgesterone and prolactin between pregnant and non-pregnant women by\nthe type of pituitary blockage (n=330).\nNon-parametric Mann-Whitney test,  p <0.05\nProgesterone and prolactin according to the type of blockage and to\nthe progesterone used (n=330).\nNon-parametric Mann-Whitney test.  p <0.05\nTable 2  shows the association between\nage, progesterone dosage, and pregnancy in this group. Women aged 34.8 years\nhad more pregnancies. The median of P4 in pregnant women was 32.6 ng/ml and\n26.6 ng/ml in non-pregnant women.  Figure\n1b  shows the ROC curve, with a sensitivity of 56.0% and\nspecificity of 63.4% for pregnancy. The cut-off point was 30.5 ng/ml for\npregnancy.  Table 3  shows that\nantagonist and injectable progesterone were associated with pregnancy, when\nthe median of P4 was 37.8 ng/ml ( p =0.021). However, the PRL\nshowed no relevance. The ROC curve showed a sensitivity of 67.2% and a\nspecificity of 54.7% ( Figure 1c ) with a\ncut-off point of 30.5 ng/ml.\nTable 2  shows the association of P4\nand pregnancy, with a median of 38.7 ng/ml in pregnant women and 22.6 ng/ml\nin non-pregnant women. However, the PRL was not significant. The ROC curve\nshowed a sensitivity of 87.5% and a specificity of 73.7% ( Figure 1d ) and a cut-off point of 26.7\nng/ml.\nQuantitative data on age, endometrial thickness, and P4 and PRL between\npregnant and non-pregnant women ( Table\n4 ) entered the analysis. Age and P4 were significant, regardless\nof the type of blockage or supplementation ( p =0.001 and\n p =0.021, respectively). More pregnancies in women under\n34.7 years of age were observed, and the median P4 was over 31.9ng/ml.\nUnivariated analysis of quantitative and qualitative variables.\nLogistic Regression Model and Wald test.\n p <0.05\nThe qualitative variables were blockage, number of embryos transferred, and\ntype of P4 used. A correlation between agonist blocking and pregnancy was\nfound ( p =0.036).\nFor multivariate analysis, age, P4, and type of pituitary block, which had\nstatistical significance in the univariate analysis, were correlated with\npregnancy. The first two criteria showed statistical significance. Age\nshowed an odds ratio (OR) of 0.91 and a 95% confidence interval (CI) of\n0.86-0.97 ( p =0.002). For P4, we observed an OR of 1.01 and\na 95% CI of 1.0-1.02 ( p =0.027). As for pituitary blockage,\nno statistical significance was observed ( p =0.204 for\nblockade with antagonists and  p =0.087 for agonists).\n\nIn vitro  fertilization is a treatment with several particularities.\nIn addition to a good embryo, a favorable endometrial and hormonal environment is\nessential. Hormone dosages in IVF allow for adjustments in treatment or better\nexpectations in achieving pregnancy. According to the findings of this study, P4\ndosages greater than 32.1 ng/mL may be useful in predicting whether or not pregnancy\nis achieved. Further studies are needed to assess whether adjustments in\nprogesterone supplementation after embryo transfer would be effective in promoting\npregnancy.\nDespite the importance of P4 and PRL, only a few studies have investigated their\nserum concentrations after embryo placement. The choice of D9 after oocyte\nretrieval, between transfer and pregnancy test, was made to offer the support that\nIVF demands and maintain close contact with the patient undergoing treatment.\nElevated levels of P4 were observed in greater numbers in pregnant women at this\nstudy. In the literature, serum progesterone above 1.5 ng/ml on the day of the\noocyte trigger suggests early luteinization, indicating embryo freezing and transfer\nto another cycle to potentiate the result ( Panaino\n et al ., 2017 ). A P4 dosage of 10 ng/ml on D4 after\nthe trigger in frozen embryo transfer provides better pregnancy rates ( Gaggiotti-Marre  et al .,\n2020 ).\nThe mean age of patients in this study was 35 years, which was lower than that\nreported in Latin America (37.2 years) ( Zegers-Hochschild  et al ., 2022 ). Female age is one of\nthe main factors that influence fertility, as it is directly linked to oocyte\nquality, which declines between 25 and 30 years of age and increases over time\n( Vander Borght & Wyns, 2018 ).\nEndometriosis was found to be the main cause of infertility, which is present in 1/4\nof the cases, contrary to the 10% described in the literature ( de Ziegler  et al ., 2016 ). It was followed by\nmale factor and ovarian failure, the latter being present in 32% of IVF cases in the\nUnited States ( Pastore  et al .,\n2018 ). The difference observed may be due to the service where the\nresearch was conducted, which has gynecological surgeons with a focus on video\nlaparoscopy.\nEndometrial thickness was similar to that observed in the literature ( Kasius  et al ., 2014 ). The\nendometrium also needs a trilaminar appearance at the time of transfer to maximize\nthe chances of success.\nIn this study, half of the cycles occurred with antagonists, which is used for its\neffectiveness and practicality to start cycles ( Al-Inany  et al.,  2016 ). When studying P4 in IVF, one\nshould consider the type of pituitary blockage. Performed to inhibit the LH surge,\nit prevents ovulation before oocyte retrieval but interferes in the formation of P4\nby the corpus luteum. LDL-cholesterol, the raw material of this hormone, needs LH to\nenter the mitochondria and initiate P4 production. By inhibiting LH, formation is\nimpaired. Thus, we have several ineffective corpora lutea, leading to the need for\nsupplementation ( Taylor  et al .,\n2019b ). Agonist blockage acts on this axis for up to 10 days after the\nlast application. The antagonist, for 24 hours ( Fatemi, 2009 ). In unblocked cycles, increased ovarian steroid hormones\ndo this job.\nLuteal phase supplementation in this study was performed with IM P4, vaginal gel, or\nmicronized natural P4. At the center that hosted the study, IM was the preferred\nroute, which was used in 3/4 of the cases. It shows excellent absorption, but its\napplication is painful and requires assistance to be administered. The used dose of\nmicronized natural P4 requires the insertion of four vaginal eggs within 24 hours.\nHowever, it can ooze through the vulva and can be uncomfortable for the patient. The\nvaginal gel has better adherence to the vaginal fundus without this\ninconvenience.\nA quality cleaved embryos were obtained in 74% of the cycles. In 1/6 of the cases,\ngood blastocysts were found. The choice of treatments with good quality embryos was\nbased on the analysis of the cycles with the potential for a positive result, thus\nbeing able to interpret the data between good cycles, hormonal dosages, and positive\nbetas. Thus, the high rate was observed as part of the inclusion criteria. The\nbetter the embryonic quality, the higher the success rate; this data was also linked\nto age, as oocytes from young women tend to generate better embryos.\nApproximately half of the transfers were performed on D3 and only 14.2% on\nblastocysts. It was common in the institution to transfer on D2 and D3, a fact that\nhas changed in recent years with regard to blastocysts due to the improvement in\nembryonic cleavage rates. In almost half of the cycles, one to two embryos were\ntransferred, respecting the orientation of the CFM.\nWhen analyzing P4 and PRL collected on D9, regardless of supplementation, a higher\npregnancy rate was observed when P4 was above 32.1 ng/ml. Moreover, PRL showed no\nsignificance.\nRegarding the pituitary blockage, agonist cycles did not show significant differences\nregardless of the type of supplementation. For PRL, values of over 37.8 ng/ml in the\nagonist and Crinone ®  combination were associated with pregnancy.\nHowever, the result should be interpreted with caution due to the small number of\ncases.\nAgonist blockage is longer-lasting. Thus, the initial expectation was that high P4\nvalues would indicate better hormone replacement and would occur in pregnant women,\na fact that was not observed. In IM supplementation, the highest values were found\nin those who did not get pregnant. Using Utrogestan ® , the opposite\noccurred in pregnant women with higher P4 values. Further studies are needed to\nassess whether this is the best agonist supplementation.\nCycles with an antagonist were associated with age, P4, and pregnancy. Moreover, P4\nvalues collected on D9 above 32.6 ng/ml were associated with pregnancy. The\nantagonist and injectable progesterone combination showed more pregnancies if the\nvalues were above 37.8 ng/ml. The serum concentrations of PRL were not\nsignificant.\nIn blockage with antagonist, elevated serum P4 values were observed in pregnant women\nas expected, and the ROC curve cut-off value was 30.5 ng/ml. The values of IM\nsupplementation were also expected, as the antagonist blocks the axis for a shorter\ntime, continuing the function of the corpus luteum more quickly.\nCases without block and P4 above 38.7 ng/ml were associated with pregnancies. Despite\nthe blockage occurring due to high steroid levels and not due to pituitary action,\nthe high P4 was consistent with the initial expectation. However, the other analyses\nshowed no significant findings. The ROC curve showed its best performance with a\ncut-off point of 26.7 ng/ml. An assumption for the lower value of P4 in relation to\nthe other curves in the study may be related to a less effective block and normal\nfunctioning of the formed corpus luteum. Thus, one possibility is that lower doses\nof P4 would already reach the level necessary for embryo fixation. Agonists tend to\nimprove pregnancy rates in cycles compared with antagonists ( Lambalk  et al ., 2017 ), a combination widely\nused for endometriosis due to the embryonic quality obtained ( Xiao & Yu, 2021 ). However, no studies have compared\nagonists with non-blocking cycles.\nThe multivariate analysis considered age, P4, and block, assessing the significance\nof the first two. The data indicated that with each year of age, a 9% reduction was\nobserved in the chance of getting pregnant, in relation to the basal rate, which is\nconsistent with the findings of the literature ( Vander Borght & Wyns, 2018 ). As for P4, in the present study, for\neach ng/ml more than P4, an increase of 1% in the probability of getting pregnant\nwas observed.\nIn the present study, PRL showed no significant association with outcomes or\nvariables. Other authors have reported that increased serum concentrations of PRL\nlead to poorer quality embryos and miscarriages ( Kim\n et al ., 2018 ).\nHowever, this study has some limitations related to its retrospective design. It was\nnot confirmed whether progesterone supplementation was performed. It is known that\npatients undergoing IVF are highly motivated to use the medication correctly.\nAdditionally, all patients were treated by members of the clinical staff of the same\nteam and followed the same protocols, ensuring consistency in the treatment\nprocess.\nThe cut-off points found in this work are unprecedented for analysis of D9 after\noocyte capture. The purpose of this study is not to replace the beta-hCG test but to\nprovide subsidies that can help the attending physician follow up his patient in the\nweek before the pregnancy diagnosis. IVF treatments go along with great anxiety,\noften being exhausting from a psychological point of view. Thus, monitoring\ntreatments brings security to patients who receive assistance at each stage of the\njourney.\n\nThe measurement of serum P4 concentrations performed on D9 after oocyte retrieval was\nuseful in predicting success in the IVF treatment cycle. Serum P4 levels above 32.1\nng/ml collected on D9 after oocyte pick-up was associated with pregnancy. The value\nwas influenced by the type of pituitary blockage and the route of administration of\nthe supplement used.\nSerum concentrations of PRL did not show statistical significance in any of the\nscenarios evaluated.","source_license":"public-domain-us","license_restricted":false}