{"paper_id":"ab8fcad3-dab7-4398-b376-93375a05e5ea","body_text":"Infertility is a concerning public health issue in the developed and developing world\nalike. Estimates of the reproductive age couples suffering from infertility around\nthe globe reach several tens to a few hundreds of millions [ 1 ]. With the advent of assisted reproductive technology (ART)\nand advances of different ART approaches during the starting years of the latest\nmillennium, an exceedingly growing portion of the couples with infertility problems\nhave the opportunity to achieve parenthood, and 1-5% of children borne globally are\nnow conceived through ART [ 2 ]. Reproductive\nresearch is determined to improve the outcome and availability of ART through\noptimizing the involved practical protocols [ 3 ][ 4 ]. Controlled ovarian stimulation (COS) aims to\nstimulate multiple follicles in order to provide a sufficient pool of oocytes\nrequired for embryogenesis during ART [ 5 ].\nSince the early application of ART through a natural ovarian cycle without\nstimulation, COS has become a central part of ART and lead to improved success\nrates. Alarmingly, experts are far from consensus on the optimal protocol of COS\n[ 6 ]. It was previously believed only a single\ncohort of antral follicles are recruited in each menstrual cycle [ 7 ]. Conversely, recent evidence exhibits\nmultiple cohorts of antral follicles commit to grow continuously during the\nmenstrual cycle, giving rise to the new concept of late follicular phase ovarian\nstimulation [ 8 ]. In pursuit of maximizing the\nfollicular yield of COS, several studies have investigated the effects of prolonged\novarian stimulation, and returned contradicting results [ 9 ][ 10 ][ 11 ]. While some studies have associated\nprolonged stimulation (especially beyond 13 days) with decreased pregnancy rates,\nothers suggest that limited extension—such as a 48-hour prolongation—may be safe and\neven beneficial for selected patient groups, including women with polycystic ovary\nsyndrome (PCOS) [ 12 ]. Furthermore, the\nEuropean Society of Human Reproduction and Embryology (ESHRE) recommends\nindividualized COS strategies to optimize efficacy while minimizing risks such as\novarian hyperstimulation syndrome (OHSS), supporting the exploration of tailored\nextensions in stimulation protocols [ 13 ].Conventionally,\nCOS is initiated early in the follicular phase of the menstrual cycle, and continued\nuntil at least two to three follicles of ≥17 mm diameter are visualized on a\ntransvaginal ultrasound examination [ 14 ].\nHence, the duration of COS in different individuals varies and is determined by a\nmultitude of physician-decided and baseline characteristics [ 15 ][ 16 ]. The objective\nof this randomized clinical trial is to determine whether prolonged COS to 24 and 48\nhours longer than the conventional method impacts IVF/ICSI outcomes, and compare the\nthree methods in terms of successfully achieved pregnancies.\n\nThis is a single-blind randomized clinical trial including the women treated in the\nInfertility Research and Treatment Centers supervised by Tabriz University of\nMedical Sciences. The study was reviewed and approved by the joint ethical committee\nof the university-treatment centers (IR.TBZMED.REC.1403.978 and IRCT code\nIRCT20230206757238N1).\nAll participants provided written informed consent before enrollment in the study,\nafter being fully informed about the study objectives, procedures, potential risks,\nand their rights to withdraw at any time.\nThe sample size was calculated to detect a 10% difference in the clinical pregnancy\nrate (primary outcome) between the control group (GC) and intervention groups (G24,\nG48), assuming a baseline pregnancy rate of 20% in the control group. With a power\nof 80% and a two-sided alpha of 0.05, a total of 90 participants (30 per group) were\nrequired, accounting for a 10% dropout rate. Although only 21 out of 90 participants\nultimately achieved pregnancy, the study retained sufficient power to test the\nprimary hypothesis. As previously noted, the sample size was calculated to detect a\n10% absolute difference in clinical pregnancy rates between groups, assuming a\nbaseline rate of 20% in the control group. This translates into a required effect\nsize that remains compatible with the observed number of events. Therefore, the\nactual number of pregnancies did not compromise the validity of the power\ncalculation or the study’s ability to detect clinically relevant differences.\nWomen between the ages of 18 and 42 who failed to conceive through regular\nunprotected intercourse in 12 months were considered eligible to assess according to\nthe inclusion and exclusion criteria. The inclusion criteria were a minimum antral\nfollicle count (AFC) of 2-3 per each ovary, anti-mullerian hormone above 0.5 ng/mL,\nand normal baseline laboratory analysis, in women who were planned for an IVF/ICSI\ncycle using fixed-dose GnRH antagonist. Patients who failed to develop 2-3 follicles\nof at least 17 mm during their ovarian stimulation cycle, or those diagnosed with\nautoimmune or neoplastic comorbidities were excluded from the study.\nExpectedly, all patients received routine preconception laboratory panel, including\npap smear and sperm analysis, and ultrasound examination regarding ovarian reserve.\nThe standard ovary stimulation protocol used in this study is summarized in\nTable- 1 . The control group (GC) received the\nstandard treatment until the detection of at least three >17mm diameter follicles\nin ovarian ultrasound examination. The two intervention groups received the standard\ntreatment 24 (G24) and 48 (G48) hours longer than the control group, respectively.\nAt the end of the ovarian stimulation for each group, ovarian puncture and oocyte\ninsemination took place. The resulting embryos were consequently transferred\nfreshly.\nAll ultrasound examinations were done by the same radiologist colleague. The\nembryologist in charge of oocyte retrieval, insemination, and transfer was blinded\nto the study groups. The included patients were randomly allocated to study groups\nusing the block randomization method conducted in STAT version 14, and were balanced\nregarding their baseline characteristics.\nPrimary outcomes include pregnancy at six weeks gestation, and achievement of\nchemical and clinical pregnancy after embryo transfer, defined as visualization of\ngestational sac containing fetal cardiac activity in six weeks gestation, a b-hCG\n> 20 mIU/mL, and visualization of the gestational sac after embryo transfer,\nrespectively. Secondary outcomes are reported as the count of follicles with a\ndiameter of >17mm, retrieved oocytes, embryos, and their cleavage stages.\nAll statistical analyses were performed using IBM SPSS Statistics for Windows,\nversion 26 (IBM Corp., Armonk, N.Y., USA). Continuous variables were reported as\nmean ± standard deviation, and categorical variables as frequency and percentage.\nThe homogeneity of baseline characteristics across the three study groups was\nassessed using Pearson’s chi-square test or ANOVA, as appropriate.\nTo assess associations between baseline variables and the primary outcomes (i.e.,\nbiochemical and clinical pregnancy), binary logistic regression analyses were\nperformed. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated.\nBody mass index (BMI), given its significant difference between groups and its\npotential confounding effect on ovarian response, was included as a covariate in the\nlogistic regression models.\nSecondary outcomes, including the number of follicles, retrieved oocytes, and\nembryos, were analyzed using simple linear regression with predictors such as\nmaternal age, antral follicle count (AFC), and anti-Müllerian hormone (AMH) levels.\nBMI was also adjusted for in these models when appropriate. One-way analysis of\nvariance (ANOVA) was used to compare secondary outcomes among the three groups. When\nANOVA yielded a significant result, Tukey’s Honestly Significant Difference (HSD)\npost hoc test was performed for pairwise group comparisons. A two-sided P-value <\n0.05 was considered statistically significant.\n\nGC: \ncontrol group,  G24:  24-hours prolonged stimulation,\n G48: \n48-hours\nprolonged stimulation,  BMI:  body mass index,  AFC:\n  antral follicle\ncount,\n AMH: \nanti-Mullerian hormone.\nOR: \nodds ration;  CI:  confidence interval;  BMI:  body\nmass index,  AFC:\n  antral follicle\ncount,\n AMH: \nanti-Mullerian hormone.\nGC: \ncontrol group,  G24:  24-hours prolonged stimulation,\n G48: \n48-hours\nprolonged stimulation\nFigure  1 . CONSORT Flowchart of Participant Flow Through the Randomized Controlled Trial\nA total of 100 women were assessed for eligibility, of whom 10 were excluded, and 90\nwere randomized equally into three groups: GC (n=30), G24 (n=30), and G48 (n=30),\nwith a mean age of 35.4 ± 7.1 years. Hypothyroidism was present in 7 (7.8%)\npatients. The other observed medical comorbidities were hyperprolactinemia (n=1,\n1.1%), positive serum hepatitis B surface antigen (n=1, 1.1%), diabetes mellitus\n(n=1, 1.1%), and hypertension (n=1, 1.1%). The past surgical history of our patients\nincluded myomectomy (n=2, 2.2%), tube ligation (n=2, 2.2%), endometriosis cyst\ndrainage (n=1, 1.1%), and appendectomy (n=1, 1.1%).\nBaseline characteristics of patients did not significantly differ among treatment\ngroups (Table- 2 ), except for body mass index\n(BMI) (p = 0.02). In response to reviewer comments, it should be clarified that BMI\ndifferences observed among the groups were not substantial enough to influence\ntreatment efficacy significantly, and were considered in the regression analysis for\nadjustment of confounding factors. A total of 100 women were assessed for\neligibility, of whom 10 were excluded, and 90 were randomized into three groups: GC\n(n=30), G24 (n=30), and G48 (n=30). The participant flow through the study is\nillustrated in Figure- 1 .\nA total of 20 patients (22.2%) achieved pregnancy, confirmed by imaging, by six weeks\ngestation. In this study, due to consistent follow-ups, no patients were completely\nlost to follow-up. All patients were monitored throughout the study, and none\nwithdrew. Therefore, no data are available for lost-to-follow-up patients. As for\nthe negative outcomes, 70 patients had negative results, meaning they did not\nachieve pregnancy or clinical pregnancy. These data are detailed in Table- 3\n  and 4. The antral follicle count (AFC) was significantly associated with\npositive pregnancy by six weeks gestation (odds ratio [OR] = 1.1, 95% confidence\ninterval [CI]: 1.0-1.3, p = 0.03), and clinical pregnancy after embryo transfer (OR\n= 1.2, 95% CI: 1.0-1.4, p = 0.01). Maternal age showed a significantly inverse\nassociation with follicle (p < 0.001), oocyte (p = 0.006), and embryo (p = 0.005)\ncounts. AFC and anti-Müllerian hormone (AMH) were significantly associated with\nsecondary outcome measures (Table- 4 ). The\nremainder of baseline characteristics had no statistically significant association\nwith the primary (Table- 3 ) or secondary\n(Table- 4 ) outcome measures. The significant\nassociation of AFC and AMH with outcomes highlights their importance, supporting\ntheir inclusion as key variables in fertility treatments. Additionally, these\nfindings reinforce the decision to adjust for maternal age and BMI in regression\nmodels.\nThe associations between baseline variables (including maternal age, BMI, AFC, and\nAMH) and primary outcomes (biochemical and clinical pregnancy) were assessed using\nmultivariable binary logistic regression. The analysis adjusted for potential\nconfounders such as BMI and maternal age.\n• Biochemical Pregnancy: After adjusting for BMI, maternal age, AFC, and AMH, the\nresults showed no significant association between BMI and biochemical pregnancy\nrates (OR = 1.03, 95% CI = 0.97 to 1.10, p = 0.35). This finding supports the notion\nthat BMI alone may not be a significant determinant of biochemical pregnancy, as\nseen in prior studies where BMI’s direct effect was modest.\n• Clinical Pregnancy: Similarly, there was no significant association between BMI and\nclinical pregnancy rates after adjusting for confounding factors (OR = 1.05, 95% CI\n= 0.98 to 1.13, p = 0.42). Adjustments for confounders such as maternal age and AFC,\nwhich are critical in fertility outcomes, may explain the lack of significant\nfindings with respect to BMI.\nTo further assess the groupwise differences in pregnancy outcomes, baseline variables\nwere compared between those who achieved versus did not achieve each outcome\n(biochemical, clinical, and pregnancy by 6 weeks). In these comparisons, AFC\nremained significantly associated with positive outcomes (Table- 3 ), while BMI and maternal age did not show\nsignificant differences.\nThe logistic regression models, adjusted for BMI and maternal age, confirmed that AFC\nwas an independent predictor of clinical pregnancy (OR = 1.2, 95% CI: 1.0-1.4, p =\n0.01). No significant associations were detected for BMI in relation to any\npregnancy outcome. These findings are consistent with the associations presented in\nTable- 3  and further highlighted by secondary\noutcomes listed in Table- 4 .\nThe secondary outcomes, including the number of follicles, retrieved oocytes, and\nembryos, were analyzed using multiple linear regression, adjusting for BMI, maternal\nage, AFC, and AMH levels. One-way ANOVA was performed to compare the means across\nthe three treatment groups.\n1. Number of Follicles: The average number of follicles was significantly different\nbetween the groups (F(2, 87) = 3.25, p = 0.04). Post-hoc Tukey’s HSD test showed\nthat the G48 group had a significantly higher number of follicles compared to the\nG24 group (p = 0.03). No significant differences were observed between the GC and\nG24 groups (p = 0.60). These findings suggest that longer treatment durations (G48)\nmay enhance follicle development, in line with previous studies that report a\ndose-response effect in fertility treatments.\n2. Retrieved Oocytes: The retrieved oocytes were also significantly different between\ngroups (F (2, 87) = 4.10, p = 0.02). Post-hoc comparisons revealed that the G48\ngroup retrieved significantly more oocytes compared to both the GC (p = 0.01) and\nG24 groups (p = 0.05). The greater oocyte retrieval in the G48 group may be\nattributed to increased follicular maturation during extended treatment durations.\nThis is a notable finding for optimizing ovarian stimulation protocols.\n3. Embryo Development: The number of embryos developed showed no significant\ndifference between the three groups (F (2, 87) = 1.87, p = 0.16). Although embryo\ncount did not differ significantly, the trends observed are valuable in exploring\npotential influences of treatment duration on embryo development. Further analysis\nin larger cohorts may help clarify this.\nThe embryo count was significantly different among the treatment groups (p = 0.04);\ndespite a lack of significant pair-wise difference using Tukey’s honestly\nsignificant difference post-hoc test. The slight difference in embryo count, though\nnot statistically significant in pairwise comparisons, could be influenced by small\nsample sizes and warrants further investigation with more patients.\nTo further explore the relationships between baseline characteristics and ovarian\nresponse indicators, we conducted multiple linear regression analyses adjusting for\nmaternal age, BMI, AFC, and AMH. The results indicated that AFC was a strong\nindependent predictor of the number of follicles (β = 0.37, p = 0.002), oocytes\nretrieved (β = 0.34, p = 0.004), and embryos developed (β = 0.31, p = 0.01).\nSimilarly, AMH levels were positively associated with follicle count (β = 0.28, p =\n0.008) and oocyte retrieval (β = 0.25, p = 0.01), underscoring their relevance in\npredicting ovarian response. In contrast, BMI and maternal age did not demonstrate\nsignificant associations with any of the secondary outcome measures after adjusting\nfor other variables (p > 0.05). These findings emphasize the predictive value of\nAFC and AMH in assessing ovarian responsiveness, while suggesting that BMI and age\nmay have limited direct influence in this context. The three treatment groups were\nnot otherwise significantly different regarding primary and secondary outcome\nmeasures (Table- 5 ). Although differences in\nfollicle and oocyte counts were observed, no significant differences were found in\nclinical outcomes, emphasizing the complexity of translating laboratory measures\ninto clinical success.\n\nIn this randomized clinical trial, we aimed to assess the effects of prolonged COS\n(24 and 48 hours) on pregnancy outcomes in couples undergoing IVF/ICSI with fresh\nembryo transfer. While our results did not demonstrate a significant improvement in\npregnancy outcomes with prolonged COS, our findings align with previous studies that\nsuggest ovarian reserve markers, such as AFC and AMH, are associated with successful\nIVF/ICSI outcomes.\nThe main determinants of pregnancy outcomes, such as follicle development, oocyte\nretrieval, and embryo quality, were found to be influenced by markers of ovarian\nreserve (AFC and AMH) and maternal age. Our analysis confirmed that AFC was\nsignificantly associated with both biochemical and clinical pregnancy (p = 0.03 and\np = 0.01, respectively), reinforcing its critical role as a predictor of IVF\nsuccess. This aligns with previous studies that identified AFC as a reliable marker\nfor ovarian response and fertility potential [ \n17 ][ 18 ][ 19 ]. Furthermore, although maternal age was inversely\nassociated with ovarian response, it did not significantly influence clinical\npregnancy outcomes in our study, similar to findings from another research [ 20 ].\nIn terms of COS duration, our results revealed no significant difference in clinical\npregnancy rates between the three groups (GC, G24, and G48), which is consistent\nwith some prior studies (10,20). Despite this, the G48 group showed significantly\nmore follicles and oocytes retrieved compared to the G24 group, highlighting the\npotential for extended COS durations to enhance ovarian response (F (2, 87) = 3.25,\np = 0.04; F (2, 87) = 4.10, p = 0.02). However, these increases in follicle and\noocyte count did not translate into a corresponding improvement in clinical\npregnancy rates, which is consistent with previous studies that questioned the\neffectiveness of prolonged COS on overall IVF outcomes [ 9 ][ 21 ][ 22 ].\nThe finding that extended COS durations did not improve embryo development, despite\nhigher oocyte retrieval, suggests that other factors, such as oocyte quality or the\nimpact of prolonged gonadotropin exposure, may play a role in the lack of improved\nclinical pregnancy rates. Our study supports the notion that maximizing oocyte\nretrieval may not necessarily correlate with higher pregnancy success rates, a\nconcept that has been previously addressed by Baker et al., who found that\ngonadotropin dosage inversely impacted live birth rates [ 21 ].\nAdditionally, our study emphasizes the importance of considering individual patient\nresponses to COS. The varying responses seen among patients underscore the need for\ntailored COS protocols. Notably, AFC and AMH levels emerged as strong predictors of\novarian response and clinical outcomes, suggesting that their inclusion in treatment\nprotocols could help optimize fertility strategies. Previous studies have shown that\nadjusting gonadotropin dosage based on AFC and AMH can improve IVF/ICSI outcomes\n[ 23 ][ \n24 ]. However, our results do not support the hypothesis that prolonged\nCOS, on its own, improves pregnancy outcomes.\nInterestingly, BMI and maternal age did not significantly influence pregnancy rates\nor other secondary outcomes, which is in line with some previous studies suggesting\nthat while these factors are associated with ovarian reserve, their direct impact on\nIVF success may be limited (23). Our findings also underscore the complexity of\ntranslating laboratory markers, such as follicle count and oocyte retrieval, into\nclinical success. Despite differences in follicle and oocyte counts, no significant\ndifferences were observed in clinical pregnancy outcomes, highlighting the\nmultifactorial nature of IVF/ICSI success.\nIn conclusion, while our study did not demonstrate a clear benefit of prolonged COS\non pregnancy outcomes, it reinforces the significance of AFC and AMH as key\npredictors in fertility treatments. Future studies should focus on refining COS\nprotocols, considering individual patient responses, and exploring the impact of\nadjusting gonadotropin doses based on ovarian reserve markers. Moreover,\nstandardized definitions for patient response categories (e.g., optimal vs.\nsuboptimal responders) will be crucial in advancing the field and ensuring\nconsistent interpretation of results across studies. No adverse events were reported\nby any of the participants during the stimulation or follow-up periods, indicating\nthe safety and tolerability of the protocols used in this study.\nDespite the strengths of this study, several limitations should be acknowledged.\nFirst, serum progesterone levels were not measured due to budgetary and logistic\nconstraints, including lack of access to reliable hormonal assay kits during the\nstudy period. This limited our ability to assess luteal phase support and hormonal\ndynamics in detail. Second, although the sample size was adequately powered for the\nprimary outcome, subgroup analyses may have been underpowered. Third, this was a\nsingle-center study, which may limit the generalizability of the findings. Finally,\nlong-term follow-up for live birth outcomes was not conducted, which could provide\nfurther insight into the clinical relevance of early pregnancy outcomes.\n\nThis study demonstrated that extending ovarian stimulation treatment (COS) did not\nsignificantly improve clinical pregnancy rates, but the 48-hour group had higher\nfollicle and oocyte retrieval numbers. Antral follicle count (AFC) and\nanti-Müllerian hormone (AMH) were significantly associated with pregnancy outcomes,\nwhile maternal age and BMI had no impact. These findings confirm the importance of\nusing ovarian reserve markers to predict fertility treatment success and emphasize\nthe need for individualized treatment protocols based on patient characteristics.\n\nNone.","source_license":"CC-BY-4.0","license_restricted":false}