Fertilization and early embryology: Cryopreservation of immature mouse oocytes

Mouse oocytes enclosed in cumulus cells were isolated from antralfollicles at the germinal vesicle (GV) stage. They were storedin straws at – 196°C by a conventional mouse embryofreezing method using dimethylsulphoxide (1.5 M) as the cryoprotectant.Overall survival assessed after removal of the cumulus cellswas 93% (299/320). A significantly greater proportion of freshoocytes remained arrested at the GV stage during culture (11versus 1%), but the rate of maturation to metaphase II was notsignificantly different between frozen and fresh oocytes (83versus 74%). The rate of fertilization in vitro was similarfor frozen and fresh oocytes matured in vitro (70 versus 81%)but significantly less than with mature ovulated oocytes (96%).Fertilization of frozen and fresh oocytes arrested after germinalvesicle breakdown was similar (77 versus 95%. No evidence ofparthenogenetic activation was found in the different groupsafter overnight incubation of metaphase II oocytes. Implantationwas similar for embryos derived from fresh and frozen GV-stageoocytes matured in vitro and mature ovulated oocytes, but theloss of embryos after implantation was significantly higherin the in-vitro matured groups (frozen, 40% and fresh, 46% versus24%). The overall survival of oocytes frozen at the GV stagewas 27%. This compares favourably with the estimated overallsurvival of mature oocytes cryopreserved by a similar procedure.We conclude that the increased post-implantation loss is dueto suboptimal conditions for maturation in vitro rather thanfreezing injury.     

Steroid Production by the cumulus: relationship to fertilization in vitro

Insemination media were Collected from 92 follicles of 14 patientsstimulated to progesterone and oestradiol in the inseminationdrops were assayed, corrected for carry–over from follicularfluid and volume and expressed as production per µg ofprotein in the cumulus. significantly higher progesteron productionper unit protein was associated with oocytes which fertilizedin vitro (P << 0.02). Oocytes fertilizing with subsequentfragmentation or degeneration showed progesterone levels significantlyhigher than oocytes fertilizing normaly (P << 0.05). Polyspermicoocytes ( n = 3 ) were associated with very high levels of progesteroneproduction but were not significantiy different due to the lownumbers. Oestradial production per unit protein was significantlygreater in oocytes which degenerated (P << 0.05). TheProtein content of cumuli whose oocytes fertilized appearedto be significantly lower than those which did not (P <<0.05). these results probably reflect the maturity of the folliclealthough direct actions of cumulus products upon gametes cannotbe ruled out.     

Fertilization and early embryology: Premature chromosome condensation of the sperm nucleus after intracytoplasmic sperm injection

A cytogenetic-cytological study was performed on unfertilizedhuman oocytes (first polar body visible) after intracytoplasmicsperm injection (ICSI) with respect to the rate of prematurelycondensed sperm chromosomes (G₁-PCC). Out of 163 prepared oocytesderived from 41 ICSI cycles, 133 (-82%) could be analysed successfully.A total of 60 oocytes (45.1%) showed metaphase II chromosomesin the haploid range along with an intact sperm head and 27oocytes (20.3%) were missing the sperm head, but two of themshowed an approximately diploid set of chromosomes; 38 oocytes(28.6%) exhibited the maternal metaphase II chromosomes as wellas G₁-PCC of the sperm nucleus showing a remarkable variationin the degree of condensation. Ten ICSI cycles (each followedby an embryo transfer) were characterized each by 2–3oocytes demonstrating G₁-PCC. It is concluded that the maincause of failed fertilization after ICSI is the failure of oocyteactivation. When the sperm nucleus is able to act with the chromosomecondensing factors and the oocyte does not become activated,this will lead to the induction of PCC. Absence of the spermhead might be due to injection or ejection of the spermatozoonin the perivitelline space except for two cases in which fertilizationmight have occurred. Finally, the observation of both a singlechromatin region (n = 6) or two chromatin regions (n = 2) indicatedoocyte activation which, however, was followed by developmentalarrest.     

Fertilization and early embryology: The effect of testosterone on the maturation and developmental capacity of murine oocytes in vitro

A dose-dependent inhibition of meiotic maturation and embryonicdevelopment was observed in both cumulus-enclosed and cumulus-denudedmurine oocytes following incubation in the presence of 10, 20,40, 60 and 80 µM testosterone for 18 h in vitro. Maturationto metaphase II was enhanced in cumulus-enclosed oocytes followingmaturation in the presence of human pre-ovulatory mural granulosacells. However, maturation of cumulus-denuded oocytes was enhancedonly when oocytes were cultured on a monolayer of human polycysticovarian granulosa cells. The presence of cumulus cells had asignificantly positive effect on both oocyte maturation (P =0.002) and embryonic development (P < 0.001). In addition,the presence of follicular cells during maturation improvedthe number of fertilized oocytes reaching the blastocyst stage.These data indicate that the exposure of immature murine oocytesto testosterone during maturation significantly reduces theirability to mature and undergo normal embryonic development.     

Partial zona dissection of human oocytes when failure of zona pellucida penetration is anticipated

Partial zona dissection (PZD) of human oocytes facilitates spermpenetration through mechanically made holes in the zona pellucida.Only 1 of 69 eggs was damaged when sucrose was used to shrinkthe ooplasm during micromanipulation. The fertilization rateof micromanipulated oocytes in 18 couples with male factor infertilitywas 68% (34/50), which compared favourably with inseminationof non-micromanipulated controls (21/45, 47%). PZD was advantageousin oligozoospermic patients, but not in cases of asthenozoospermia,combined semen problems or immunological infertility. Threetwin and two singleton pregnancies resulted following replacementof 23 micromanipulated and eight control embryos in 14 patients.No differences in embryo morphology and development rates werefound between the micromanipulated and control groups. The incidenceof polyspermy in couples with abnormal semen analyses was relativelylow (<20%) possibly due to partial activation of the oocytesfollowing exposure to sucrose. Polyspermy was high (57%) innormozoospermic patients with either immunological infertility(n= 3) or failure of fertilization in previous cycles (n= 4).In the three immunological patients, nine of 11 hyaluronidaseand sucrose-exposed control embryos fertilized and six implanted,possibly indicating that cumulus and corona cells are contributingfactors inhibiting fertilization in such cases.     

Differential gene expression in cumulus cells as a prognostic indicator of embryo viability: a microarray analysis

Besides the established selection criteria based on embryo morphologyand blastomere number, new parameters for embryo viability areneeded to improve the clinical outcome of IVF and more particularof elective single-embryo transfer. Genome-wide gene expressionin cumulus cells was studied, since these cells surround theoocyte inside the follicle and therefore possibly reflect oocytedevelopmental potential. Early cleavage (EC) was chosen as aparameter for embryo viability. Gene expression in cumulus cellsfrom eight oocytes resulting in an EC embryo (EC-CC; n = 8)and from eight oocytes resulting in a non-EC (NEC) embryo (NEC-CC;n = 8) was analysed using microarrays (n = 16). A total of 611genes were differentially expressed (P < 0.01), mainly involvedin cell cycle, angiogenesis, apoptosis, epidermal growth factor,fibroblast growth factor and platelet-derived growth factorsignalling, general vesicle transport and chemokine and cytokinesignalling. Of the 25 selected differentially expressed genesanalysed by quantitative real-time PCR 15 (60%) genes couldbe validated in the original samples. Of these 8 (53%) couldalso be validated in 24 (12-EC-CC and 12 NEC-CC) extra independentsamples. The most differentially expressed genes among thesewere CCND2, CXCR4, GPX3, CTNND1 DHCR7, DVL3, HSPB1 and TRIM28,which probably point to hypoxic conditions or a delayed oocytematuration in NEC-CC samples. This opens up perspectives fornew molecular embryo or oocyte selection parameters which mightalso be useful in countries where the selection has to be madeat the oocyte stage before fertilization instead of at the embryonicstage.     

Chromosome anomalies in human oocytes failing to fertilize after insemination in vitro

Three-hundred-and-two unfertilized oocytes left over from successfulin-vitro fertilization (IVF) attempts in 143 women (27–42years) on a follicular stimulating hormone-human menopausalgonadotrophin (FSH-HMG) stimulation regime were subjected tochromosome analysis. Ten oocytes were degenerated with no visiblechromosomes and 41 metaphases had chromosomes that were clumpedtogether which could not be interpreted either numerically orstructurally. Of the remaining oocytes, 76.6% (192/251) hada normal haploid complement (n = 23), 13% (33/251) were hypohaploid(n = 19–22), 8% (20/251) were hyperhaploid (n = 24–26),2% (5/251) were diploid (2n = 46) and 0.4% (1/251) had structuralrearrangements. The 21% aneuploidy was from 24 different patientsand hypohaploid sets had chromosomes missing mainly from theA, B, C, D and G groups while the hyperhaploid sets had extrachromosomes from A, B, D, G and E groups of the human karyotype.The mean age of patients showing aneuploid oocytes was 36.7years which was above the mean for the entire group. The aneuploidymay have been brought about by errors in oogenesis (anaphaselagging or non-disjunction) and may offer one explanation forfertilization failure and overall low pregnancy rates afterIVF.     

Fertilization and early embryology: Intracytoplasmic single sperm injection of 1-day-old unfertilized human oocytes

Although the average fertilization rate in most in-vitro fertilization(IVF) centres is 60–70%, there are cases of complete orvirtually complete fertilization failures. The aim of our workwas to study the fertilization and the subsequent cleavage characteristicsof 1-day-old human oocytes treated by intracytoplasmic singlesperm injection (ICSI) after failing to fertilize during thestandard IVF procedure. A total of 115 metaphase II 1-day-oldunfertilized oocytes were collected from 23 patients. No additionaltreatment was applied to the oocytes or to the semen sample.A single spermatozoon from the patient's husband was injectedinto the cytoplasm of each of these oocytes 21–33 h afterovum retrieval. Injected oocytes were observed at 16–18h and again 42–44 h after the ICSI procedure. Of the injectedoocytes, 92% (n = 106) were intact after ICSI, 38% (n = 44)had two distinct pronuclei and there was no difference in thefertilization rate of oocytes when andrological and non-andrologicalpatients were compared. Similarly, there was no difference inthe fertilization rate after ICSI where patients with acceptableor good (> 15%) fertilization after standard IVF were comparedto patients who had poor (<15%) fertilization after IVF.There was no significant difference in the sperm concentrationor in the progressive forward motility (a + b motility) in thesegroups except where a + b motility of andrological and non-adrologicalpatients was compared. The majority (84%) of the normally fertilizedoocytes cleaved and most (77%) of these embryos showed <20%fragmentation 2 days after the ICSI procedure. From this studyit can be concluded that 1-day-old metaphase II oocytes whichhave failed to be fertilized after standard IVF procedure canbe fertilized and cleave when ICSI is performed on them theday after oocyte retrieval.     

Fertilization and early embryology: Co-culture of human pronucleate oocytes with their cumulus cells

The aim of this prospective randomized work was to study thevalue of co-culturing human pronucleate oocytes with their cumuluscells. A total of 550 fertilized oocytes from 95 in-vitro fertilizationpatients were randomly divided into two groups on the day afterinsemination. Group A oocytes (n = 260) were left undisturbedwith their attached cumulus cells and group B oocytes (n = 290)were dissected from their cumulus cells. Both groups were incubatedand examined daily for 3 days. In group A, 78% (202/260) reachedthe 4-cell stage 48 h after retrieval compared to 69% (200/290)in group B. At 72 h after retrieval, 70% (141/202) had reachedthe 8-cell stage in group A compared to 56% (112/200) in groupB. The percentages of grade 1 embryos at 48 and 72 h after retrievalwere 70% (141/202) and 76% (107/141) in group A compared to50% (100/200) and 43% (48/112) in group B respectively. We concludedthat co-culture of human oocytes with their cumulus cells significantlydecreased their fragmentation and increased the number of embryosthat reached the 4-cell and 8-cell stages with regular blastomeres.The technique is simple and avoids the use of heterogeneouscells.     

Evaluation of the spindle apparatus of in-vitro matured human oocytes following cryopreservation

The present study was conducted to determine if the cryopreservationof immature human oocytes has a deleterious effect on the meioticspindle following maturation in vitro. Oocytes were obtainedin excess from in-vitro fertilization patients and divided intofour groups. Groups 1 (n = 98) and 2 (n = 80) consisted of immatureoocytes cryopreserved before or after maturation in vitro respectively.Groups 3 (n = 37) and 4 (n = 9) served as non-frozen controlsand included oocytes matured in vitro and in vivo respectively.The meiotic spindle was identified after incubation in anti-tubulinmonoclonal antibody (1 h, 37°C) and fluorescein-conjugatedgoat anti-mouse immunoglobulin G (IgG) (1 h, RT). Chromosomeswere counterstained with 4‘, 6’-diamidino-2-phenylindole.Following cryopreservation, group 1 oocytes demonstrated a 63%survival rate and 68% maturation rate in vitro. In all, 58%of the oocytes in group 2 survived the thaw. The number of oocyteswith normal spindles in group 1 (81.0%) was not significantlydifferent from control groups 3 (83.8%) and 4 (88.9%), whilethe number of group 2 oocytes with normal structures (43.5%)was significantly lower than groups 1 (P = 0.0004), 3 (P = 0.0002),and 4 (P = 0.025). These results suggest that cryopreservationof the prophase I human oocyte does not significantly increaseabnormalities in the resulting meiotic spindle.     

Fertilization and early embryology: Behaviour of spermatozoa in human oocytes displaying no or one pronucleus after intracytoplasmic sperm injection

The behaviour of sperm cells after intracytoplasmic sperm injection(ICSI) was investigated by analysing 192 unfertilized and 37one-pronuclear (1PN) oocytes following ICSI. Eighty-two unfertilizedoocytes were directly fixed whereas 110 were first parthenogeneticallyactivated by puromycin. In contrast to the findings in unfertilizedoocytes after in-vitro fertilization, most unfertilized oocytesafter ICSI (n = 76) contained evidence of the presence of spermatozoain the cytoplasm. Few oocytes (n = 6) contained prematurelycondensed sperm chromosomes (PCC), whereas the majority containedeither intact sperm heads (n = 31) or swollen sperm nuclei (n= 39) along with metaphase II chromosomes of the oocyte. Followingactivation by puromycin, swollen sperm nuclei and PCC were nolonger observed, whereas unchanged sperm heads persisted in12 oocytes displaying a single pronucleus. A non-decondensedsperm nucleus along with decondensed maternal chromatin werealso discovered in 32 out of 37 oocytes displaying a singlepronucleus after ICSI. The findings in unfertilized and 1PNoocytes after ICSI indicate that successful sperm injection,even followed by oocyte activation, is not sufficient to guaranteenormal fertilization. It seems that partial sperm membrane damageprior to injection is also required to ensure normal sperm decondensation.     

Gamete intra-Fallopian transfer: evaluation of 100 consecutive attempts

The results of 100 gamete intra-Fallopian transfer (GIFT) proceduresto treat persistent infertility are reported. Twenty-four pregnancieswere achieved, of these six aborted, two were extra-uterine,two stillbirths occurred and nine patients delivered 11 healthychildren (two sets of twins) and five pregnancies are progressingwell, including two sets of twins. Pregnancy rate in the differentgroups of patients was: 28% for idiopathic infertility (n =39), 13% for male infertility (n = 16), 22% for endometriosis(n = 27), and 29% in the presence of antisperm antibodies (n= 7). In our GUT procedure, we place three oocytes and 50 000to 100 000 motile spermatozoa per patient into one healthy tube,the remaining oocytes being inseminated and cultured in vitro.Of 502 oocytes recovered, 252 fertilized normally and 178 earlyembryos were frozen. The replacement of 41 frozen-thawed embryosresulted in five additional, ongoing pregnancies. The combinedtreatment by gamete intra-Fallopian transfer, in-vitro fertilizationand cryopreservation increases the chance of conception.     

Back muscle as a promising site for ovarian tissue transplantation, an animal model

BACKGROUND: The aim of this study was to evaluate the optimal transplantationsite for ovarian tissue fragments in murine hosts. We comparedthe transplantation to the back muscle (B) versus the kidneycapsule (K) in a mouse allograft model. METHODS: Hemi-ovaries from 12-day-old mice were allografted into B andK of bilaterally ovariectomized same strain recipients whichhad undergone gonadotrophin stimulation (n = 15). Graft survivalafter 27 days, angiogenesis and follicle development were scoredand compared to age-matched control ovaries (38-day old, n =5). The ability of oocytes to be fertilized was studied afterIVF, ICSI and embryos were transferred to recipient mothers.Anti-mouse CD 31+ antibody was used to evaluate neo-vascularizationin grafts. RESULTS: Primordial follicle survival was higher (P < 0.01) and vascularsupport was better (P < 0.01) in B- than in K-grafts. From34 oocytes retrieved from B-grafts (15 metaphase I, of which14 matured in vitro, and 19 collected at metaphase II), 18 morulaewere obtained. Transfer of 12 embryos obtained by ICSI led tothree live offspring, and transfer of six IVF embryos to anotherrecipient mother yielded four offspring, one of which was borndead and one showed placental anomalies. CONCLUSIONS: The back muscle is a promising site for ovarian allografts inmice. This is the first report of live offspring obtained afterback muscle grafting using both IVF and ICSI.     

Fertilization and early embrology: Atypical maturation of oocytes of strain I/LnJ mice

The maturation of strain I/LnJ oocytes was compared to oocytesof selected inbred strains. The time of germinal vesicle breakdown(GVB) of I/LnJ oocytes was greatly delayed compared to all otherstrains tested. In addition, 5% of the cumulus cell-enclosedoocytes isolated from antral follicles of I/LnJ mice failedto undergo GVB in vitro and 58% of the oocytes that underwentGVB failed to progress beyond metaphase I. Similar defects inthe progression of meiosis occurred when maturation was stimulatedin vivo by the administration of exogenous gonadotrophins. Whenin-vitro matured metaphase II oocytes were selected for in-vitrofertilization, similar percentages of I/LnJ oocytes underwentfertilization and cleavage to the 2-cell stage as oocytes fromanother inbred stain, C57BL/6J, and similar percentages of 2-cellstage oocytes completed the 2-cell stage to blastocyst transitionin vitro. However, unlike C57BL/6J oocytes, a much lower percentageof oocytes that matured in vivo in response to exogenous gonadotrophinsunderwent fertilization and cleavage to the 2-cell stage thanoocytes that underwent maturation in vitro. Likewise, lowerpercentages of 2-cell stage embryos derived from in-vivo maturedI/LnJ oocytes developed to blastocysts than embryos derivedfrom in-vitro matured oocytes. These results show than I/LnJoocytes are atypical in the progression of both nuclear andcytoplasmic maturation. These defects may account for the poorreproductive performance of I/LnJ mice. Thus, I/LnJ mice mightbe a useful model for studying infertility resulting from defectiveoocytes.     

Preimplantation adoption: establishing pregnancy using donated oocytes and spermatozoa

The experience of transferring embryos produced through in-vitrofertilization (IVF) utilizing donated oocytes and spermatozoais described. Recipients (n = 28; aged 38–59 years) receivedoral micronized oestradiol and i.m. progesterone and were synchronizedto donors undergoing ovarian stimulation. Reasons for selectingtherapy included advanced reproductive age (>42 years; n= 21) or hyper-gonadotrophic hypogonadism (n = 7), combinedwith severe male factor infertility in 23 couples. Five womenwere single and without partners. Oocytes were fertilized bycryopreserved spermatozoa designated for use by the recipient.Up to five embryos were transferred trans-cervically. Supernumeraryembryos were cryopreserved. A total of 36 aspirations produced15.6 ± 7.3 oocytes per retrieval. In 10/36 cycles (27.8%),embryos were available for cryopreservation. Using fresh embryos,the overall pregnancy rate was 38.9% (14/36), clinical pregnancyrate 33.3% (12/36), and ongoing/delivered pregnancy rate 30.6%(11/36). Three ongoing pregnancies were later established bytransferring cryopreserved embryos. Adjusting for these events,the per aspiration overall pregnancy rate per retrieval was47.2%, clinical pregnancy rate 41.7%, and ongoing/deliveredpregnancy rate 38.9%. Implantation rates per individual embryotransferred were 16.6% following fresh embryo transfer. A viablepregnancy was achieved by 14 of 28 women (50% cumulative pregnancyrate). We conclude that using donor oocytes and donor spermatozoais efficacious and allows couples of whom both members sufferfrom severe gamete abnormalities and single functionally agonadalwomen an effective means of achieving pregnancy.     

Fertilization and early embryology: Parthenogenetic activation of human oocytes following cryopreservation using 1, 2-propanediol

Fresh and aged human oocytes were cryopreserved using 1, 2-propanediol(PROH). After thawing, the oocytes were cultured for 20 h andexamined for parthenogenetic activation using light microscopyand an ultraviolet DNA stain. Control fresh or aged oocytesand oocytes exposed to PROH without cryopreservation were alsoexamined for activation. No control oocytes were observed toactivate spontaneously (n = 43) and parthenogenetic activationwas not induced by exposure to PROH alone (n = 26). In bothfresh and aged cryopreserved oocytes, 27 and 29% of the oocytesrespectively were activated, and these proportions were significantlyelevated compared with the controls (P < 0.01). Althougha similar rate of activation was observed for the cryopreservedfresh and aged oocytes, the form of parthenogenetic activationvaried between these two types of oocyte. A single pronucleuswas observed in 18% of the fresh and 5% of the aged cryopreservedoocytes. In contrast, the presence of two or more pronucleiwas observed in 0% of the fresh and 19% of the aged cryopreservedoocytes.     

Fertilization and early embryology: Consequences of non-extrusion of the first polar body and control of the sequential segregation of homologues and chromatids in mammalian oocytes

Absence of polar body formation, or premature chromatin condensation(PCC) in human oocytes can cause infertility. We studied in-vitromaturing mouse oocytes in order to identify risk factors forsuch conditions, and for the precocious segregation of homologuesor chromatids. Treatment with the actin-binding drug cytochalasinD (10 µg/ ml) arrested oocytes in metaphase I. Upon exposureto Ca²⁺-ionophores, anaphase I was triggered in the absenceof cytokinesis. Chiasmata resolved and homologues separatedinstantaneously. In some oocytes predivision of all chromatidsoccurred. Homologues or chromatids never separated even afterexposure to Ca²⁺-ionophores when microtubules were depolymerized,although bivalents could eventually decondense. Thus, in meiosisI checkpoints exist which ensure that homologue separation onlytakes place when a metaphase I spindle is present but cytokinesisand anaphase progression can be uncoupled. Cycloheximide induceda sequential separation of homologues in oocytes with intactmetaphase I spindle, resulting in metaphase II chromosomes andbivalents in individual cells as also found in some human oocytesof aged females. In oocytes which progressed to metaphase IIbut failed to extrude a first polar body, the two sets of chromosomeseventually aligned on a common spindle (‘diploid’metaphase II). PCC of one set was never observed. Ageing invitro of cytochalasin D-blocked metaphase I oocytes had no pronouncedeffect on chromosome segregation.     

Cytological studies of human zygotes exhibiting developmental arrest

Developmental arrest of 111 ( 5%) fertilized ova which had formedtwo pronuclei was observed during a 5-year period of an in-vitrofertilization programme. At least 30 zygotes demonstrated visiblepronuclei at 44–66 h after insemination, and 42 zygotesfragmented. Of the 107 prepared zygotes, 97 were informativeand revealed that developmental arrest occurred at differentstages of the cell cycle: from interphase (n = 48), to transitioninterphase-prophase (n = 20), to prophase (n = 11), and to metaphase(n = 13). The latter 13 zygotes were characterized by chromosomesets as follows: haploid (n = 2), diploid (n = 6), triploid(n = 1) and tetraploid (n = 4). Another five zygotes demonstrateddifferent numbers of metaphase chromosomes (between 10 and 40),as well as prematurely condensed chromosomes (PCC) as a resultof marked asynchrony in pronuclear morphogenesis. A total of18 zygotes exhibited asynchrony in the morphology of the twopronuclei. It is concluded that abnormal chromosome sets andpronuclear asynchrony might be causes for early developmentalarrest.     

Luteal phase start of low-dose FSH priming of follicles results in an efficient recovery, maturation and fertilization of immature human oocytes

In this prospective study we investigated whether the maturation and fertilization of immature oocytes can be improved by administration of recombinant follicle stimulating hormone (rFSH) starting in the late luteal phase in two groups of women: group 1 (n = 6) women with regular menstrual cycles; and group 2 (n = 6) women with irregular cycles and polycystic ovaries (PCO) on ultrasound examination. Low-dose (37.5 IU) rFSH was commenced 11 days after LH surge during a spontaneous menstrual cycle and on the ninth day of progesterone administration in an irregular cycle. Recombinant FSH was continued until the leading follicle was approximately 10 mm in diameter. The oocytes were retrieved after withdrawing rFSH for 2-5 days. In total, 136 oocytes were recovered (group 1, 67 oocytes; group 2, 69 oocytes). Nine of the oocytes from PCO women were atretic at retrieval. Oocytes complete with cumulus cells were cultured for 44 h in complex tissue culture medium supplemented with gonadotrophins and fetal calf serum. After maturation, the cumulus cells were removed and metaphase II oocytes were injected with spermatozoa. Respectively, the oocyte maturation and fertilization rates were 64 and 72% in group 1, and 78 and 57% in group 2 (not significant). After fertilization, the zygotes (group 1, n = 22; group 2, n = 11) and cleavage stage embryos (group 1, n = 9; group 2, n = 15) were frozen in propanediol. All women except one (11/12) had approximately five zygotes or cleaved embryos frozen. The viability of in-vitro matured frozen-thawed embryos was generally poorer than that (81%) seen after conventional intracytoplasmic sperm injection, with 61% survival in group 1 and 23% in group 2. Fifteen embryo transfers resulted in one miscarriage at 6 weeks gestation. The late luteal start of low-dose rFSH yielded a good number of immature oocytes in women with both regular and irregular cycles. Two out of three of these oocytes matured and fertilized. However, cryosurvival of the zygotes and cleaved embryos was unsatisfactory and thus cryopreservation of in-vitro matured embryos may not be an optimal procedure.