{"paper_id":"edf350a8-f049-4e4c-a205-36da88f7ca99","body_text":"Ovarian folliculogenesis is a complex process involving\ninteractions between the classical hypothalamus-\npituitary-ovarian axis and other intra- and\nextra-ovarian factors ( 1 ). Up to a certain point, follicular\ngrowth and development occur readily in the\npresence of normal basal concentrations of gonadotropin,\nmetabolic hormones and growth factors, but\nthe follicles eventually reach the end of their normal\nlife span under those basal conditions. At that time,\nonly the follicle exposed to specific additional signals\n(dominant follicle) will continue to grow until ovulation\nwhile the remainder (subordinate follicles) become\natretic and regress ( 2 - 4 ). This process, known as\nfollicle selection, involves a reduction in systemic follicle stimulation hormone (FSH) concentrations below\nthe concentration required by the smaller follicles\n( 4 ). This depressed FSH concentration is maintained\nby a negative feedback loop of protein produced by\nthe dominant follicle such as inhibin ( 5 ) and estradiol\n( 4 ).\nThe ability of the dominant follicle to continue\ngrowth under decreased FSH concentrations, while\nthe subordinate follicles regress, suggests that responsiveness\nto FSH or FSH dependence may be altered\nduring follicular development ( 4 ). Indeed, the\nlargest follicle acquires luteinizing hormone (LH) receptors\nor gene expression for LH receptors between\ntwo and four days after wave emergence ( 6 ,  7 ). Follicle\nselection involves a transient elevation in LH,\nwhich is required for the production of estradiol and\nfree insulin-like growth factor-I (IGF-I). IGF-I synergizes\nwith FSH to stimulate granulosa cell proliferation\nand steroidogenesis ( 8 ). The smaller follicles\nhave not reached a similar developmental stage and,\nbecause of their dependency on FSH, they become\nsusceptible to low FSH concentrations ( 9 ,  10 ).\nDuring controlled ovarian stimulation (COS), women\nare usually treated with an agonist or antagonist of\ngonadotropin-releasing hormone (GnRH) to block the\naction of the pituitary, and their ovaries are stimulated\nwith gonadotropins to induce the development and\nfinal maturation of multiple follicles ( 11 ). Therefore,\nthe increase in circulating levels of gonadotropins will\noverride the selection of a single dominant follicle and\nstimulate the development of multiple antral follicles\nwhose enclosed oocytes have the potential for fertilization\nand further development ( 12 ).\nIn monovular animal species, several intrafollicular\nevents occur before the beginning of diameter\ndeviation between the largest follicle and\nthe second largest follicle of the cohort. Therefore,\nalthough mature oocytes may be retrieved from\nmultiple follicles after COS, it is still a matter of\ndebate whether oocytes retrieved from small follicles,\nwhich have escaped from atresia under supraphysiologic\ndoses of gonadotropins, present\nthe same developmental competence as oocytes\nderived from larger follicles (12).\nThe present study compared the developmental\ncapacity of gametes retrieved from the largest follicle\nand the other small follicles of the cohort in COS\nintracytoplasmic sperm injection (ICSI) cycles.\n\nThis prospective study included a total of 1016 follicles\ncollected from 96 patients who underwent ICSI\ncycles in the Fertility-Assisted Fertilization Center,\nBrazil between January 2007 and December 2008.\nAfter follicular aspiration, we assigned the oocytes\nto two groups according to the diameter of the derived\nfollicle: i. oocytes derived from the leading\nfollicle of the cohort (large follicle, n=96) and ii.\noocytes derived from the smaller follicles of the\ncohort (small follicle, n=920). The fertilization and\npercentage of top-quality embryos were compared\nbetween groups. We assessed the effect of follicular\ndiameter on oocyte dimorphism.\nThe patients’ ages ranged from 22 to 43 years old\n(median ± SEM: 33.4 ± 0.42). All patients presented\nwith the following inclusion criteria: the presence of\nboth ovaries, a regular menstrual cycle, BMI lower\nthan 35 kg/m 2 , no ongoing infectious diseases, no uterine\npathology, basal FSH <14 IU/ml and basal E2 <70\npg/ml. All ejaculated semen used for ICSI presented\nmotile sperm concentrations above 5 × 10 6  sperm/ml.\nInfertility was defined as unexplained infertility\n(28/96: 29.1%), male infertility (27/96: 28.1%),\nmale- and female-associated factors (12/96: 12.5%),\nendometriosis (10/96: 10.4%), ovarian factors\n(11/96: 11.4%) and tubal obstructions (8/96: 8.3%).\nCOS was achieved by pituitary blockage using a\nGnRH antagonist (Cetrotide, Serono, Geneva, Switzerland)\nand by ovarian stimulation with recombinant-\nFSH (Gonal-F®, Serono, Geneva, Switzerland). The\npatients began daily recombinant-FSH treatment (225\nIU) from the third day of their menstrual cycles. The\nfirst ultrasound control and the E2 plasma dosage tests\nwere performed at the seventh cycle day. Depending\non the response of each patient that was determined by\nultrasound monitoring of the follicle size, we adjusted\nthe dose of recombinant-FSH. GnRH antagonist was\nadministered when the dominant follicle was 14 mm in\ndiameter. When at least three follicles reached 18 mm\nin diameter and serum estradiol level reached >600\npg/mL, recombinant human chorionic gonadotropin\n(r-hCG, Ovidrel™, Serono, Geneva, Switzerland) was administered to trigger final follicular maturation.\nOocytes were collected 35 hours after hCG administration\nby transvaginal ultrasound ovum pick-up.\nThe leading follicles (largest follicle of both ovaries)\nwere the first to be aspirated; smaller follicles\nwere subsequently aspirated in different tubes.\nBriefly, after retrieval, oocytes were incubated\nin culture medium (G-MOPS™-V1, Vitrolife,\nKungsbacka, Sweden) covered with mineral oil\n(Ovoil™, Vitrolife, Kungsbacka, Sweden) at 37ºC\nand 6% CO 2  for 5 hours, according to the previously\nestablished protocol (13). The oocyte retrieved from\nthe largest follicle was cultured in a different drop\nfrom the other oocytes. Cumulus cells were removed\nwith a 30 second exposure to Hepes-buffered medium\nthat contained 80 IU/mL hyaluronidase (Irvine\nScientific, Santa Ana, USA), after which coronal\ncells were manually removed with a finely drawn\nglass Pasteur pipette (Humagen Fertility Diagnostics,\nCharlottesville, VA, USA).The denuded oocytes\nwere then assessed for nuclear status by an inverted\nmicroscope. Oocytes that released the first polar body\nwere considered mature and used for ICSI.\nFor ICSI, oocytes were placed individually in 4\nµL droplets of buffered medium (G-Mops™-V1,\nVitrolife, Kungsbacka, Sweden). Sperm was placed\nin a central 4 µL droplet of polyvinylpyrrolidone\nsolution (PVP, Irvine Scientific, Santa Ana, USA)\nin a 50×40 mm glass culture dish (WillCo-dish®,\nNJ, USA) covered with warm mineral oil (Ovoil™,\nVitrolife, Kungsbacka, Sweden). Sperm injection\nwas carried out on the heated stage (37ºC) of an inverted\nmicroscope (Eclipse TE 300; Nikon®, Tokyo,\nJapan) 40 hours after hCG triggering.\nImmediately before sperm injection, we assessed\noocyte morphology by an inverted microscope and\nrecorded the following dysmorphisms: i. excessive\ncytoplasm granulation, ii. dark cytoplasm, iii. presence\nof vacuoles, iv. polar body fragmentation, v.\nperivitelline space dysmorphisms, vi. zona pellucida\ndysmorphisms, and vii. shape dysmorphisms.\nDuring the ICSI procedure, changes to membrane\nresistance to sperm injection were also recorded\nContinuous variables are given as means ± SEM,\nand proportions (%) are used for categorical variables.\nWe compared proportions by the Chi-square\nor Fisher's exact test, when the expected frequency\nwas five or less, and the results have been presented\nas proportions (%). To study the influence of the\nfollicular diameter (large or small) on oocyte morphology,\nbinary logistic regression models were conducted.\nThe results are expressed as odds ratios (OR),\n95% confidence intervals (CI) and p values. Results\nwere considered significant at the 5% critical level\n(p<0.05). Data analysis was carried out using Minitab\n(version 14), a statistical analysis program.\nWritten informed consent was obtained, in which patients\nagreed to share the outcomes of their cycles for research\npurposes. The study was approved by the Ethics\nCommittee of the Federal University of Sao Paulo.\n\nThe overall numbers of aspirated follicles were\n1016 and retrieved oocytes were 863, of which 604\nwere in the metaphase II (MII) stage, 95 in the metaphase\nI stage, 138 in prophase I stage and 26 were\ndegenerated.\nThe diameters of the larger follicles ranged from\n14 to 21 mm, and the diameters of the smaller follicles\nranged from 11.6 to 14.6 mm. The mean diameter\nof the leading follicle group (19.1 ± 2.1) was\nsignificantly higher than the mean diameter of the\nsmaller follicle group (13.0 ± 5.5, p<0.001).\nA significantly higher percentage of oocytes derived\nfrom the leading follicle were in the MII-stage\n(100% vs. 70.0%, p<0.001); however, no significant\ndifferences were observed regarding the percentage\nof degenerated oocytes between the large (6.25%)\nand small follicle (5.0%, p=0.550) groups.\nThere were 11.5% of the small follicles that were in the\nmetaphase I stage and 19.3% in the prophase I stage.\nRegression analysis demonstrated a nearly twofold\nincrease in the incidence of vacuoles in oocytes derived from the largest follicle of the cohort. We\nobserved a trend toward a higher chance of presenting\ndecreased membrane resistance to ICSI in\noocytes derived from the leading follicle ( Table 1 ).\nThere was no significant influence of the follicle\ndiameter in the presence of excessive cytoplasm\ngranulation, dark cytoplasm, perivitelline space\ndysmorphisms, polar body fragmentation, zona\npellucida dysmorphisms, shape dysmorphisms or\nincreased membrane resistance to ICSI ( Table 1 ).\nRegression analysis of the influence of the follicular\ndiameter on the incidence of oocyte defects\nOD;Odds ratio, CI; Confidence interval and OR; Refers to\nthe larger follicle diameter.\nThe fertilization rate (50.0% vs. 38.8%, p=0.038)\nand the percentage of top quality embryos (84.7% vs.\n76.4%, p=0.040) were significantly higher for oocytes\nderived from the largest follicle of the cohort. However,\nthe percentage of abnormal fertilized oocytes was\nequally distributed between the large follicle (15.0%)\nversus the small follicle (12.8%, p=0.550) groups.\nWhen embryo selection was performed without taking\ninto consideration the experimental group origin,\nwe observed that embryos derived from the largest follicle\n(53.5%) were more commonly selected for transfer\ncompared to the control group (33.8%, p<0.001).\n\nThe female gonad plays a key role in the differentiation\nand release of the mature oocyte for fertilization,\nembryo development and successful pregnancy.\nIn monovular species, following the recruitment of a\ncohort of follicles, one follicle is selected for dominance\nand continues to grow while growth of the\nothers is curtailed. Despite the critical importance of\nselection of the dominant follicle to ovarian function\nand fertility, why one follicle is selected from a group\nof similar follicles remains unknown (2).\nIn stimulated cycles, pharmacologic doses of gonadotropins\ncreate a supraphysiological hormonal\nenvironment that induces the growth of a cohort of\nfollicles, which, under natural conditions, would become\natretic and regress (2).\nHere, we evaluated the developmental competence\nof oocytes retrieved from the lead follicle compared\nto those retrieved from smaller follicles of the cohort\nin COS cycles. The data showed that oocytes derived\nfrom the largest follicle presented a higher rate of\nboth fertilization and top-quality embryos.\nThe most obvious sign that a follicle has been\nselected as dominant is a significant difference in\nsize compared to the largest subordinate follicle\n(4). However, it has been previously suggested that\nselection of the dominant follicle is a progressive\nprocess and that the initial stages of selection occur\nbefore there is a perceptible difference in size (2).\nA defining characteristic of the dominant follicle\nappears to be its greater capacity for estradiol production.\nPrevious studies in domestic animals have\nshown that as soon as the dominant follicle is detected,\nit has higher concentrations of estradiol in the\nfollicular fluid as soon as it becomes slightly larger\nthan the largest subordinate follicle (4, 14). Previous\nstudies have shown the presence of increased\nprotease activity of insulin-like growth factor binding\nproteins (IGFBP) and increased concentration\nof free IGF-1 (15) in dominant follicles. In addition,\nit has been demonstrated that granulosa cells of the\nlargest follicle acquire LH receptors shortly before\nthe dominant follicle can be detected (10).\nIn the present study, despite the exposure to increased\ndoses of exogenous gonadotropins, an increased\npercentage of immature oocytes were derived\nfrom smaller follicles rather than from larger follicles.\nIt has been previously suggested that follicles\ncontaining immature oocytes after the administration\nof large doses of hCG must lack sufficient blood supply\nto receive the ovulatory stimulus or have insufficient LH receptors to induce oocyte maturation in\nvivo (16), which is substantiated by the frequent nonexpansion\nof the corresponding cumuli (17).\nAn increase in vascularity would give the follicle an\nadvantage to receive a preferential supply of growth\nfactors, gonadotropins, steroid precursors and other\nnutrients required for its continued development. The\nrelationship between follicle vascularity and dominant\nfollicle detection has been studied directly by Doppler\nultrasonography in cattle (18, 19). Blood flow area\nbegins to differentially increase in the future dominant\nversus subordinate follicle about one day before\nthe beginning of diameter deviation. Differences in\nblood supply between follicles with different diameters\ncould also explain the decreased rate of fertilization\nand top-quality embryos observed in our study\nfor oocytes derived from smaller follicles. Rosen et\nal. (20) have previously demonstrated that the leading\nfollicle was most likely to have a mature oocyte\nwith increased fertilization and high quality embryo\ndevelopment capacity, however this study was performed\nin classic  in vitro  fertilization (IVF) cycles and\nthe oocyte morphology could not be evaluated. In the\npresent study we evaluated oocyte morphology immediately\nbefore ICSI; our data also suggest that although\noocytes derived from larger follicles presented\na higher developmental capacity, the incidence of vacuoles\nand decreased membrane resistance to ICSI was\nhigher in oocytes derived from the lead follicle.\nOocyte quality has been regarded as a variable\nthat influences the implantation potential of derived\nembryos (21-24). However, the predictive value of\ncriteria used in these studies is still controversial. In\nfact, previous studies on the developmental outcome\nof oocytes with cytoplasmic abnormalities have suggested\nthat cytoplasmic dysmorphisms are not related\nto fertilization or embryo quality (25, 26).\nVacuolization is probably the most apparent\nand dynamic cytoplasmic dysmorphism in human\noocytes. Vacuoles vary in size as well as in number\nand, according to Van Blerkom et al. (27), they are\nmembrane-bound cytoplasmic inclusions filled\nwith fluid that is virtually identical with perivitelline\nfluid. It is assumed that vacuoles arise either\nspontaneously or by fusion of preexisting vesicles\nderived from the smooth endoplasmic reticulum\nand/or Golgi apparatus (28). Vacuoles appear to\ndevelop rapidly (within several minutes) around\nextrusions of the first polar body (23).\nInterestingly, there is only one ICSI study (25)\nthat found a impaired fertilization rate in vacuolized\noocytes compared with vacuole-free MII gametes. All\nthe other papers either published a normal fertilization\nrate (29) or did not deal with vacuolization as a separate\ne feature (26, 30).\nAccording to Ebner et al. (23), during the ICSI,\nthree distinct types of oolemma responses can be\nobserved. Most of the injected oocytes show normal\nbreakage of the membrane. A second type of\nresponse called 'difficult breakage' is characterized\nby delayed penetration. The third type of membrane\nresponse, sudden breakage of the oolemma without\ncreation of a funnel may be observed. It has been described\nthat the sudden breakage of the oolemma or\ndecreased membrane resistance to ICSI is correlated\nwith decreased rates of survival and fertilization (26,\n29), however, once fertilization is achieved, apparently\nthe oocyte development is normal\nConversely, our findings suggest that both vacuole\nformation and decreased membrane resistance to\nICSI may occur in oocytes retrieved from over-aging\nfollicles; however, fertilization and development of\nthe embryo is not compromised by this feature.\nThe best way to evaluate oocyte quality is undoubtedly\nto evaluate the embryo implantation potential.\nIn the present study however, neither the pregnancy\nnor the implantation rate could be compared between\nthe groups, since in many cases embryos from both\ngroups were transferred for the same patient. This is\na limitation of the study, which could be avoided if\nexclusively elective single embryo transfers are performed.\nIn our study we have used oocyte quality,\nfertilization capacity and embryo development as\nvariables to evaluate oocyte competence.\n\nTogether with previous reports our data suggest\nthat intrafollicular mechanisms within the larger\nfollicle of the cohort may allow it to amplify the responsiveness\nto the exogenous gonadotropins, leading\nto the formation of more competent oocytes with\nhiher fertilization and developmental capacities.","source_license":"CC-BY-4.0","license_restricted":false}