A large retrospective study on 1PN embryo transfer supports the need for updated harmonized best practice guidelines

Normal fertilization of human embryos is assessed by the visual identification of two pronuclei (2PN) 16–18 h post-insemination ( Working Group on the update of the ESHRE/ALPHA Istanbul Consensus et al. , 2025 ). However, in some instances, abnormal fertilization occurs when more or fewer than 2PNs are visualized, termed atypically pronucleated zygotes (APZ; Capalbo et al. , 2024 ). The formation of monopronuclear (1PN) zygotes occurs in 1.6–9% of fertilization events ( Staessen and Van Steirteghem, 1997 ; Destouni et al. , 2018 ), with these embryos widely considered atypical, and recommendations from the ESHRE Guideline Group on Good Practice in IVF indicating that transfer of embryos from 1PN zygotes is not recommended ( ESHRE Guideline Group on Good Practice in IVF Labs et al. , 2016 ). Several other societies offer similar guidance ( Cuevas Saiz et al. , 2018 ; Cimadomo et al. , 2021 ; Malhotra et al. , 2021 ). In spite of these guidelines, several studies have shown that a proportion of 1PN embryos are genetically normal and can form a healthy pregnancy ( Bradley et al. , 2017 ; Capalbo et al. , 2017 ; Si et al. , 2019 ; Chen et al. , 2020 ; Li et al. , 2020 ; Wang et al. , 2022 ; Tong et al. , 2023 ) with the most recent systematic review and meta-analysis showing that live birth rates of 1PN embryos were comparable with 2PN embryos following euploidy confirmation (relative risk (RR) 0.85, 95% CI 0.59–1.23) ( Lee et al. , 2026 ). This suggests that clinical recommendations for the use of 1PN embryos need to be reconsidered. Currently, clinics either transfer monopronuclear blastocysts (MPBs) without genetic testing or they require pre-implantation genetic testing for aneuploidy (PGT-A) and genotyping before use ( Capalbo et al. , 2024 ). The rationale for why clinics test 1PN embryos is to reduce the risk of hydatidiform molar pregnancy from uniparental inheritance ( Capalbo et al. , 2024 ). However, since PGT-A is invasive, expensive, and not all clinics have access to genetic testing, other means of assessing suitability for transfer are needed. Several morphologic features have been associated with normally fertilized 1PNs including size of the pronucleus ( Araki et al. , 2018 ; Kai et al. , 2018 ; Mateo et al. , 2020 ; Zhao et al. , 2022 ; Xiong et al. , 2024 ), number of nucleolar pronuclear bodies ( Xiong et al. , 2024 ), morphokinetics ( Mateo et al. , 2017 , 2020 ), and cleavage and blastocyst stage embryo morphology ( Fu et al. , 2022 ; Wang et al. , 2022 ). Identifying morphological features of normally fertilized MPBs may provide further confidence around embryo transfer without the need for genetic testing. Though many studies report aneuploidy rates for MPBs, the available literature on uniparental inheritance is limited. We previously reported one case of uniparental inheritance in an ICSI MPB and no cases among 15 IVF MPBs ( Bradley et al. , 2017 ). All other PGT-A studies of MBPs focus on ICSI-1PNs, and report biparental inheritance incidence of 40–100% ( Capalbo et al. , 2017 ; Destouni et al. , 2018 ; Xie et al. , 2018 ; Soler et al. , 2021 ; Yeshua et al. , 2025 ). Further studies are required to understand the incidence of biparental inheritance of both ICSI and IVF MPBs. The available evidence showing good transfer outcomes of IVF-MPBs coupled with our prior small dataset showing a high rate of biparental inheritance indicates more data are needed to establish the RR of transferring IVF-MPBs without genetic analysis. The aim of the study is to report the rate of uniparental inheritance and parental origin in IVF-MPBs and ICSI-MPBs and, combined with morphologic criteria, establish a framework for guiding the use of untested MPBs.

Table 1 describes baseline demographics and cycle characteristics of the 10 730 cycles that were found to have at least one 1PN present following fertilization check between January 2010 and February 2024. Figure 1 shows the project flow diagram and the numbers of 1PNs and 2PNs included in each level of analysis (i.e. embryo development, genetic testing, pregnancy determination, and maternal and neonatal outcomes). Cohort derivation and clinical outcomes by fertilization method . Blastocyst formation and utilization rates shown as proportion of total 1PNs fertilized. NAD rate as proportion of 1PNs undergoing PGT-A. Biparental/uniparental rates as proportion of 1PNs undergoing STR-based parental inheritance testing. Outcome rates as proportion of embryos transferred (all grades combined; see Table 4 for grade-stratified data). †Not genetically tested: patient preference or unsuccessful biopsy. PGT-A platforms and STR protocols evolved across the study window, see Supplementary Materials and Methods . 1PN: one pronuclear embryo; 2PN: two pronuclear embryo; NAD: no abnormalities detected; STR: short tandem repeat. This figure was generated with assistance from Claude (claude.ai, Anthropic, version Sonnet 4.6, accessed [April, 2026]), and subsequently reviewed and verified by the authors. Cycle characteristics. IVF, standard insemination; 2PN, two pronuclei; 1PN, monopronuclei. Continuous data are presented as median (range), proportional data are presented as number (percentage). Contains split insemination cycles, where cycles had a combination of both standard insemination and ICSI. Of the 10 730 cycles included, 49% were standard IVF cycles while the other 51% were ICSI cycles. Women were of moderate age (∼36 years—ranged 19–48 years) and generally of normal body mass index (∼23 kg/m 2 —range 15–55 kg/m 2 ), with men also of a moderate age (∼37 years—range 19–72.0 years). The median number of previous cycles was zero (range 0–20). The primary cause of infertility in cycles using standard IVF was unexplained (34%), while for ICSI it was male factor infertility (53%). The median number of oocytes collected was similar for both standard IVF and ICSI cycles (11 oocytes) ranging from 1 to 67 oocytes collected at oocyte pick up. The median number of oocytes inseminated was slightly lower in ICSI cycles (10 oocytes) versus standard IVF cycles (11 oocytes). Both the total incidence of 1PN embryos and within cycle incidence was similar for both standard IVF (16.9% and 23.6%, respectively) and ICSI (17.0% and 23.9%, respectively) cycles. Table 2 displays PGT-A and biparental inheritance results of 1PN and 2PN blastocysts. For IVF-MPBs, incidences of aneuploidy (37.3% vs 33.9%, respectively), monogenic or chromosomal rearrangements (62.5% vs 64.3%, respectively), and unknown outcomes (PGT-A 2.6% vs 2.8%; pre-implantation genetic testing for monogenetic (PGT-M)/structural rearrangements (SR) 12.5% vs 3.6%, respectively) were similar compared with 2PN blastocysts. ICSI-MPBs were more likely to be aneuploid (45.5% vs 31.5%, respectively) and less likely to be unaffected following PGT-M or PGT-SR testing (30.4% vs 40.7%, respectively) compared with 2PN blastocysts. Adjusting for patient ID and maternal age did not affect outcomes. Genetic outcomes of 1PN and 2PN embryos. PGT-A, pre-implantation genetic testing of aneuploidy; PGT-M/SR, pre-implantation genetic testing of monogenic conditions and/or structural rearrangements; 1PN, one pronuclear embryo; 2PN, two pronuclear embryo; IVF, standard insemination; NAD, No abnormalities detected; RR, relative risk. Data are expressed as number (percentage). Adjusted for maternal age and patient ID. Biparental inheritance was also conducted in a subset of 1PN embryos to reduce the risk of molar pregnancies. The presence of both maternal and paternal chromosomes was found in 440/454 IVF-MPBs (96.9%) versus 108/164 ICSI-MPBs (65.9%), a difference that was statistically significant ( P  < 0.001). Of those embryos that displayed uniparental inheritance, the majority (IVF 92.8%; ICSI 94.6%) were maternally derived. Maternal chromosome contribution was detected in 99.8% for IVF-MPBs and 98.2% for ICSI-MBPs (including biparental and maternally derived uniparental cases). Of the 1103 MPBs with a reportable PGT-A or PGT-M/SR result (778 IVF-MPBs; 325 ICSI-MPBs), STR-based biparental inheritance testing was performed on 618 embryos (454 IVF-MPBs; 164 ICSI-MPBs). The remaining 485 embryos did not proceed to STR testing for the following reasons: aneuploid PGT-A result, affected status for a monogenic or structural rearrangement condition (AFF), or absence of parental DNA blood spot cards required for STR extraction (DNC). STR testing was performed selectively on euploid or mosaic embryos, and on PGT-M/SR embryos not affected for the tested condition, where parental DNA was available. The reported biparental inheritance rates therefore reflect this clinically defined tested subset rather than all MPBs reaching biopsy. Table 3 describes blastocyst development and embryo utilization rates between 1PN and 2PN blastocysts. For both standard IVF and ICSI cycles, MPBs were less likely to develop on time and reach the blastocyst stage by Day 5 (IVF 19.3% vs 63.3%; ICSI 9.7% vs 60.6%, respectively). For MPBs that developed to a blastocyst on Day 5, more were of lower quality (IVF 53.2% vs 41.6%; ICSI 61.2% vs 40.5%, respectively). When assessing blastocyst utilization rates (embryos that were frozen and/or transferred on Day 5 or Day 6 of development), IVF-MPBs were less likely to be utilized compared with 2PN blastocysts and this was further reduced in ICSI cycles (IVF 13.0% vs 49.9%; ICSI 5.7% vs 47.4%). Adjusting for within-patient factors did not influence the above outcomes. Blastocyst development and embryo utilization of 1PN and 2PN embryos. 1PN, one pronuclear embryo; 2PN, two pronuclear embryo; IVF, standard insemination; RR, relative risk. Data are expressed as number (percentage). High-quality blastocyst include (3–6 AA, AB, BA), average-quality blastocyst include (3–6 BB and grade 1–2 early blasts), poor blastocyst quality include (3–6 CA, CB, CC, AC, BC, and grade 3 early blasts). Blastocyst utilization rate includes frozen (Day 5 or 6) for 1PN and frozen (Day 5 or Day 6) + fresh transfer for 2PNs. Adjusted for patient ID. To determine if we could differentiate uniparental versus biparental MPBs without the need for genetic testing, we further assessed key early morphological and morphokinetic events in those embryos generated from standard insemination, where biparental rates >96% ( Fig. 2 and Supplementary Table S1 ). While there were no changes seen in the PN circumference relative to total oolemma size ( P  > 0.05, Fig. 2A and B ) or PN circumference relative to zona pellucida circumference ( P  > 0.05, Fig. 2C – E ) or any embryo morphokinetic events between first cleavage until the early blastocyst stage ( P  > 0.05, Supplementary Table S2 ), we did find that biparental IVF-MPBs were more likely to have increased numbers of nucleoli present within their PN compared with uniparental IVF-MPBs ( P  = 0.008, Fig. 2F – H ), and this difference was more prominent in those embryos that went on to form an average- or low-quality blastocysts ( P  = 0.006 and P  = 0.01, respectively, Fig. 2G ). Morphological characteristics of uniparental and biparental IVF-1PN embryos. (A ) Pronuclear (PN) circumference relative to oolemma circumference, ( B ) Representative images of how PN circumference and oolemma circumference was measured, ( C ) PN circumference relative to inner zona pellucida (ZP) circumference, ( D ) PN circumference relative to outer ZP circumference, ( E ) Representative images of how ZP circumference was measured, ( F ) Number of nucleoli within PN of biparental and uniparental IVF-1PNs, ( G ) Number of nucleoli of IVF-1PNs by blastocyst grade, and ( H ) Representative images of the number of nucleoli present in biparental and uniparental IVF-1PNs. Numbers in (F) and (G) represent statistically significant P values of <0.05, all other graphs were not statistically significant. N=30 Biparental and N=14 Uniparental 1PN embryos. Table 4 displays the pregnancy and live birth rates following the transfer of genetically normal blastocysts. For standard IVF cycles, transferred MPBs were more likely to be of lower quality (14.8% vs 7.3%) compared with 2PN embryos. Biochemical (56.2% vs 65.1%, respectively), ongoing (44.8% vs 57.0%, respectively), and live birth (40.9% vs 50.4%, respectively) rates were lower compared with 2PN blastocysts. In the adjusted analysis (maternal age, embryo freeze day, and patient ID), these results were no longer significant. Given the difference in blastocyst quality at transfer between 1PN and 2PN blastocysts following standard IVF we further conducted a sub-analysis of pregnancy outcomes based on blastocyst quality ( Table 4 ). When results were categorized based on high, average, and low blastocyst quality, no differences were observed for biochemical, ongoing, or live birth rates in both the unadjusted and adjusted analysis between 1PN- and 2PN-derived blastocysts. Pregnancy and live birth rates following the transfer of genetically normal 1PN and 2PN blastocysts. 1PN, one pronuclear embryo; 2PN, two pronuclear embryo; IVF, standard insemination; RR, relative risk. Data are expressed as number (percentage). High-quality blastocyst includes (3–6 AA, AB, BA), average-quality blastocyst includes (3–6 BB and grade 1–2 early blasts), poor blastocyst quality includes (3–6 CA, CB, CC, AC, BC, and grade 3 early blasts). Adjusted for maternal age, embryo freeze day (Day 5 or 6), and patient ID. For ICSI cycles, no differences were seen in blastocyst grade at transfer or biochemical, ongoing, or live birth rates between 1PN- and 2PN-derived blastocysts. Sample size for ICSI-MPB transfers was low (n = 54). Table 5 presents the maternal and neonatal live birth outcomes following the transfer of genetically normal 1PN and 2PN blastocysts. For both standard IVF and ICSI cycles, there was no effect of 1PN-derived embryos on gestational age at birth, the number of term births, singleton pregnancies or proportion of male infants, infant birthweight or the prevalence of small or large for gestational age compared with 2PN-derived embryos. Similarly, the incidence of maternal or neonatal complications (4.5–9.9% vs 8.7–9.6%) was similar between 1PN- and 2PN-derived pregnancies, respectively. Maternal and neonatal outcomes following the transfer of genetically normal 1PN and 2PN blastocysts. 1PN, one pronuclear embryo; 2PN, two pronuclear embryo; IVF, standard insemination; RR, relative risk; SGA, small for gestational age; LGA, large for gestational age. Gestational length and singleton birthweight all are expressed as median (range), term singleton birthweights are expressed in grams (g) as mean ± SEM and proportional data are expressed as number (percentage). Maternal/neonatal complications include Amniotic fluid leak, hypertension, incompetent cervix, hemorrhage, cholestasis, gestational diabetes, pre-eclampsia, retained placenta, and inter uterine growth retardation. Adjusted for maternal age. Adjusted for maternal age and gestational length. Reduced denominator due to missing birth outcome data. Supplementary Table S3 describes the cycle rescue contribution of 1PN embryos. Split insemination cycles (n = 219) were excluded from this analysis, leaving 5129 IVF and 5230 ICSI cycles. In 161 IVF cycles (3.1%) and 186 ICSI cycles (3.6%), no 2PN fertilization occurred, leaving only 1PN embryos available; of these, 149 IVF (92.5%) and 178 ICSI cycles (95.7%) contained at least one utilizable 1PN embryo. A further 639 IVF cycles (12.5%) and 685 ICSI cycles (13.1%) produced 2PN embryos that did not reach a blastocyst; of these, 566 IVF (88.6%) and 642 ICSI cycles (93.7%) contained at least one utilizable 1PN embryo. These combined, 800 IVF cycles (15.6%) and 871 ICSI cycles (16.7%) had no utilizable 2PN embryo. In 715 IVF cycles (13.9%) and 820 ICSI cycles (15.7%), the 1PN embryo was the only utilizable embryo available. In the remaining 85 IVF cycles (1.7%) and 51 ICSI cycles (1.0%), neither 1PN nor 2PN embryos were utilized. Among cycles in which both 1PN and 2PN euploid embryos were transferred (IVF: n = 57; ICSI: n = 17), all 2PN transfers were biochemically negative in 11 IVF (19.3%) and 4 ICSI cycles (23.5%), with the subsequent 1PN transfer resulting in a biochemical pregnancy. A live birth following 1PN transfer, after failure of all preceding 2PN transfers, was recorded in 15 IVF (26.3%) and 2 ICSI cycles (11.8%).

Full methodological detail, including evolution of laboratory practices over the study period, PGT-A platform specifications, and short tandem repeat (STR)-based biparental inheritance testing procedures, is provided in the Supplementary Materials and Methods . Institutional approval for this negligible risk research project was obtained from Genea’s Scientific Advisory Committee (GSAC; reference number 030324-FSA-EMB) on March 11, 2024. The study utilized non-identifiable clinical data in accordance with the Australian National Health and Medical Research Council (NHMRC) National Statement on Ethical Conduct in Human Research (2023-updated 2025). This retrospective study included 10 730 cycles of standard IVF or ICSI where a 1PN embryo (N = 13 203) were detected at fertilization check (16–18 h post-insemination), between December 2010 and February 2024 at Genea Australian clinics (Bowen Hills, Canberra, Coffs Harbour, Illawarra, Launceston, Lismore, Liverpool, Melbourne, Newcastle, Northwest, Orange, RPAH, Sydney, and Wembley). Cycles utilizing donor gametes were excluded. Pregnancy outcomes were assessed in cryopreservation-warming cycles (FET) for the same embryos. The primary outcome was euploidy rate and biparental inheritance as determined by array-based comparative genome hybridization (CGH) or next-generation sequencing (NGS), both supplemented with STR genotyping as required. Secondary outcomes included blastocyst development and utilization, PN morphology and embryo morphokinetics, clinical pregnancy, live birth, and maternal and neonatal outcomes. Ovarian stimulation, oocyte collection, standard IVF and ICSI or a combination of both were carried out as described previously ( Roy et al. , 2014 ). Briefly, fertilization check occurred 16–18 h post-insemination and the number of PNs assessed. Oocytes showing a single pronucleus (1PN) were identified and cultured individually for tracking purposes, while 2PN oocytes were cultured either individually or in groups of up to five embryos. All embryos were cultured until the blastocyst stage utilizing either sequential media or single-step embryo culture media (Gems culture media, Genea Biomedx, Syndey, Australia). Blastocysts were graded internally on Days 5 and 6 using Genea’s standardized scoring system, which records expansion stage as alphabetic abbreviations and inner cell mass and trophectoderm quality as numeric grades. For reporting, all grades were translated to the equivalent Gardner nomenclature ( Gardner and Schoolcraft, 1999 ) as adopted by the revised Istanbul Consensus ( Working Group on the update of the ESHRE/ALPHA Istanbul Consensus et al. , 2025 ). High-quality blastocysts are defined as stage 3–6 grades AA, AB, BA; average-quality blastocysts 3–6 BB and grade 1–2 early blastocysts; and low blastocysts incorporated grades 3–6 CA, CB, CC, AC, BC and grade 3 early blastocysts. Embryos underwent assisted hatching on Day 3 between January 2010 and October 2016 and Day 4 after October 2016, with the use of a Zilos TK laser (Hamilton Thorne Biosciences, MA, USA) as previously described ( Bradley et al. , 2017 ). On the morning of Day 5, embryos were assessed for the presence of an expanding trophectoderm protruding from the zona opening. Blastocysts suitable for biopsy underwent trophectoderm biopsy as previously described ( McArthur et al. , 2005 ). Embryos not suitable for biopsy on Day 5 were cultured for an additional 6–24 h to determine suitability. Biopsied trophectoderm cells were immediately stored at −20°C for subsequent genetic analysis. Following biopsy, all blastocysts were allowed to recover for 1 h in culture, prior to vitrification using the Gems Vitrification Set (Genea Biomedx) and either the Gavi vitrification system (Genea Biomedx) or the open Cryotop device (Kitazato, Shizuoka, Japan) as detailed previously ( Roy et al. , 2014 ). Analysis of chromosome composition occurred as previously described ( Bradley et al. , 2017 ). Briefly, the SurePlex (Rubicon Genomics) whole-genome amplification (WGA) kit was used to amplify DNA from embryo biopsy tissue for genetic analyses following the manufacturer’s instructions. Comprehensive chromosome analysis was carried out by either array comparative genomic hybridization (array-CGH; 2010–2018) or NGS (2016–2024). Both platforms were available during an overlap period from 2016 to 2018, with platform selection based on laboratory workflow and patient testing requirements. For array-CGH, the Agilent 60K oligonucleotide microarray (Agilent Technologies, CA) or the Bluegnome 24 sure BAC array (Illumina, Cambridge) platforms were used. Log ratio calculations indicating chromosome dosage were interpreted with Genomic Workbench Standard Edition 7.0 (Agilent Technologies) or BlueFuse Multi 3.0 (Bluegnome), respectively. For NGS, the VeriSeq PGS Kit in combination with the MiSeq sequencer (Illumina) was used as per the manufacturer’s instructions. Alignment of sequencing data to the reference genome was carried out by MiSeq Control Software (v2.5), followed by analysis with BlueFuse Multi Analysis Software (Illumina) to provide a genomic profile. Throughout the study period, embryos were reported as: euploid (no abnormality detected; 46 chromosomes with normal copy number), aneuploid (whole chromosome gains or losses), mosaic (mixed cell populations with differing chromosome constitutions), abnormal ploidy, or no result (insufficient DNA or ambiguous profile). It should be noted that during 2010–2018, embryos with abnormal ploidy (i.e. triploidy –3PN) were typically classified within the aneuploid category rather than as a separate group. From 2018 to 2024, ploidy abnormalities were more routinely reported as a distinct category. Biparental genetic testing was conducted as previously described ( Bradley et al. , 2017 ). Briefly, STR loci spanning chromosomes 1, 2, 3, 6, 17, 21, 22, X, and Y were chosen as target regions to detect maternal and paternal DNA contribution to the embryos ( Supplementary Table S1 ). For each case, 8–12 markers were selected based on parental heterozygosity to maximize informativity across independent chromosomal loci. Maternal and paternal DNA controls, blanks, and an aliquot of the remaining WGA product from each embryo were used in separate amplifications for the relevant STR markers to track biparental inheritance. The polymerase chain reaction (PCR) reagent final concentrations were 0.2 mM dNTPs, 1.5 mM MgCl 2 , primer concentrations as determined during test validation (0.03–0.3 µM) and 2.5 U of Fast-Start Taq (Roche). The PCR amplification parameters used were 95 °C for 7 min, followed by 94 °C for 30 s and 60 °C for 1 min 30 s for 33–43 cycles, then 72 °C for 10 min. After amplification, diluted STR PCR products were subjected to capillary electrophoresis on an ABI 3130 or ABI 3500 Genetic Analyser (Life Technologies) and resulting electropherogram data analyzed using GeneMapper ID v4.0 or above analysis software (Life Technologies). Allele calls were made according to predefined criteria: a peak was accepted as a true allele if it met the minimum relative fluorescence unit threshold established during validation, fell within the expected size range for the locus, and was distinguishable from background noise and stutter artefacts. All results were independently reviewed by two experienced genetic scientists prior to reporting. Sample data were reviewed to confirm biparental inheritance by identifying inheritance of maternal and paternal alleles across multiple chromosomes, consistent with classic Mendelian segregation. Embryos were classified as biparental diploid where both maternal and paternal alleles were detected across ≥6 chromosomes. A pattern of exclusively maternal or exclusively paternal alleles across all informative markers on multiple chromosomes was classified as uniparental. Cases with mixed or contradictory signals were classified as inconclusive and excluded from transfer consideration. An inherent limitation of STR-based analysis on amplified material from small numbers of cells is the risk of allele dropout (ADO), whereby one allele at a heterozygous locus fails to amplify, potentially generating a false uniparental signal at that marker. The use of multiple independent STR markers distributed across a minimum of six chromosomes mitigates this risk: a genuine uniparental inheritance pattern produces a consistent single-parent signal across all chromosomes tested, whereas ADO at an isolated locus would be contradicted by biparental signals preserved at the remaining markers. Time-lapse videos were obtained from Genea’s video archive of all uniparental IVF-1PN embryos (N = 14) and IVF-1PN biparental embryos (N = 30) case matched for blastocyst quality prior to cryopreservation. To assess the area of the oolemma and pronucleus (PN), still images of each 1PN embryo were extracted immediately prior to pronuclear fading (PNf). Measurements were performed using the area tool in ImageJ (NIH, USA). Due to variation in image resolution across the dataset, the ratio of oolemma circumference to PN circumference and ratio of the zona pellucida circumference to PN circumference was calculated for normalization and used in subsequent analyses. Morphokinetic parameters were obtained using Geri ® Assess (Genea Biomedx) automated annotations where available. For embryos lacking automated annotation, manual annotation was performed by a single operator to ensure consistency. The developmental time points annotated included time of pronuclear fading (tPNf), and the time to reach 2-cell (t2), 3-cell (t3), 4-cell (t4), 5-cell (t5), 6-cell (t6), 8-cell (t8), and start of blastulation (tsb) ( Supplementary Table S2 ). These parameters were used for comparative morphokinetic analysis between IVF-1PN uniparental and biparental embryos. Chromosomally normal blastocysts (no detectable chromosomal abnormalities by CGH or NGS and biparental inheritance), were warmed with Gems Warming Set (Genea Biomedx) and transferred using embryo culture media (Gems culture media, Genea Biomedx) or EmbryoGlue transfer media (Vitrolife, Gothenburg, Sweden) previously described in Roy et al. (2014) . Blastocyst utilization rates include all frozen and transferred embryos as a percentage of available 1PN or 2PNs. A pregnancy test was performed 11 days following frozen embryo transfer (unless menstruation began prior to this date). A biochemical pregnancy was defined by rising serum βhCG concentrations >5 IU/l. An ongoing pregnancy was determined as the presence of at least one gestational sac, with fetal heart motion present at 6–8 weeks gestation on trans-vaginal ultrasound. Women were in the care of their treating IVF physician until confirmation of a viable pregnancy following the ultrasound at 6–8 weeks of gestation, then referred for primary obstetric care. Birth outcomes were supplied by the treating obstetrician as per the mandatory reporting to the Australian and New Zealand Assisted Reproduction Database. Only post-20 week births were considered and live birth was determined as the complete expulsion or extraction from its mother of a product of conception irrespective of the duration of the pregnancy, after such separation, breathes or shows any other evidence of life, such as beating of the heart, pulsation of the umbilical cord or definite movement of the voluntary muscles, whether or not the umbilical cord has been cut or the placenta is attached; each product of such a birth is considered live born. All data were analyzed using SPSS Statistics for Windows Version 26 (IBM Corp., USA). Data were checked for normality using one-sample Kolmogorov–Smirnov test. Continuous data were presented as median with upper and lower ranges (not normally distributed) or means ± SEM (normally distributed), while proportional data were presented as percentages and raw numbers. The association between exposure and outcome for blastocyst development, embryo utilization, pre-implantation genetic testing, pregnancy and live birth, term birth, singleton pregnancy, sex, and maternal/neonatal complications was analyzed by RR and expressed as RR with 95% CIs. For the adjusted analysis (i.e. adjusted for maternal age, embryo freeze day, patient ID, and gestational length), data were assessed by binomial logistic regression and expressed as β-estimates and 95% CIs. The proportion of embryos classified as biparental was compared between IVF-derived and ICSI-derived MPBs using a chi-squared test of independence. Singleton birthweight was assessed with a Kruskal–Wallis non-parametric test, while gestational length, and boy and girl birthweights were assessed by a general linear model. MPB morphology and embryo morphokinetics were assessed by a Mann–Whitney test. To examine the cycle rescue contribution of 1PN embryo inclusion, cycles were classified by utilizable embryo availability (defined as frozen or transferred on Day 5/6) for 2PN and 1PN embryos separately. Split insemination cycles (n = 219) were excluded as fertilization method could not be unambiguously assigned; full definitions and classifications are provided in Supplementary Table S3 . A P -value of <0.05 or if the 95% CIs did not cross 1.0 was classified as significant.

This study represents the largest retrospective assessment to date of clinical and maternal/neonatal outcomes following the utilization of MPBs. Our analysis demonstrates that in standard IVF cycles, the majority of single-pronuclear (1PN) zygotes that develop into usable blastocysts are diploid. In contrast, fewer ICSI-1PN zygotes develop into usable blastocysts, and among those that do, approximately one-third exhibit uniparental inheritance. After confirmation of euploidy status, both IVF- and ICSI-derived MPBs demonstrate live birth rates and maternal/neonatal outcomes comparable with those observed with euploid 2PN blastocysts. The principal strengths of our investigation include direct comparisons of 1PN versus 2PN blastocyst transfer outcomes within identical treatment cycles for both IVF and ICSI cycles, comprehensive analysis of more than 10 000 cycles containing 1PN embryos, detailed assessment of biparental genetic analysis outcomes on transferred MPBs, and thorough evaluation of corresponding morphokinetics on a representative subset of IVF-1PN embryos. The primary limitation of this research is its retrospective design, which introduces the potential for selection bias in the analysis. Currently, all pre-implantation genetic testing studies assessing the utility of 1PN embryos have focused on ICSI-derived MPBs. Biparental diploid rates in these embryos range from 43% to 100%, averaging 68% (111 of 164; Bradley et al. , 2017 ; Capalbo et al. , 2017 ; Destouni et al. , 2018 ; Xie et al. , 2018 ; Soler et al. , 2021 ; Tong et al. , 2023 ) similar to the 65.9% rate we report here. This relatively low rate of diploidy contrasts significantly with IVF-derived MPBs. While we previously found that all 15 IVF-derived MPBs (100%) were biparental ( Bradley et al. , 2017 ), 440 of 454 IVF-derived MPBs (96.9%) in the current study had biparental inheritance. Further highlighting that IVF-MPBs are almost always diploid. The difference between IVF- and ICSI-derived MPBs in uniparental inheritance is consistent with studies at the zygote stage, where only 30% of ICSI-derived 1PN zygotes were diploid compared with 87% of IVF-derived 1PN zygotes ( van der Heijden et al. , 2009 ). Early embryonic arrest is likely the main reason why diploidy rates increase considerably at the blastocyst stage for both ICSI-MPBs (66%) and IVF-MPBs (97%). Our results are consistent with reports that when biparental inheritance was confirmed, both IVF- and ICSI-MPBs show similar euploid rates to 2PN blastocysts ( Bradley et al. , 2017 ; Destouni et al. , 2018 ; Xie et al. , 2018 ; Soler et al. , 2021 ; Tong et al. , 2023 ). The parental origin of uniparental cases is important, as paternal diploidy carries a risk of forming hydatidiform molar pregnancies ( Maisenbacher et al. , 2019 ). Though few studies report parental origin of uniparental inheritance, all 16 cases from three studies showed exclusively maternal inheritance ( Destouni et al. , 2018 ; Xie et al. , 2018 ; Soler et al. , 2021 ). This is in contrasts to our findings where 7% (1/14) of IVF-derived MPBs and 5.5% (3/56) of ICSI-derived MPBs had paternally derived uniparental inheritance. A case report of a paternal ICSI-MPB ( Zhou et al. , 2021 ) suggests that the observed difference in our study relative to other reports is likely due to a sampling effect. The frequency of maternal versus paternal inheritance provides clues to how uniparental diploid MPBs form. Maternal disomy likely occurs through artificial or spontaneous activation of the oocyte that results in parthenogenesis and meiosis I non-disjunction ( Benn, 2021 ). This may occur after an oocyte is exposed to sperm during IVF, causing premature activation, or through the action of injecting a sperm at ICSI resulting in premature activation. Evidence for maternal disomy following ICSI is provided by a study in which oocytes that displayed a single pronucleus after ICSI contained a decondensed maternal chromatin alongside an undecondensed sperm nucleus ( Dozortsev et al. , 1994 ), suggesting that sperm decondensation failure can lead to maternal disomy in ICSI-activated oocytes. Incidence of maternal disomy is also strongly linked with advanced maternal age (>35 years) ( Kotzot, 2004 ). In our cohort, the median age of women with maternal disomy was 35 years (range 22–43 years), indicating a similar trend. Paternal disomy, while less common, is more likely to arise from a non-disjunction in maternal meiosis, resulting in a nullisomic oocyte, with postzygotic rescue by duplication of the paternal genome ( Benn, 2021 ). In IVF, biparental diploid 1PNs likely arise from sperm penetration occurring close to the meiotic spindle ( Wei et al. , 2022 ), whereas in ICSI, injection of sperm near the meiotic spindle can also cause a large, single pronucleus, as demonstrated in mouse oocytes ( Krukowska and Tarkowski, 2005 ). As illustrated in Fig. 3 , 1PNs can form through additional mechanisms including asynchronous pronuclear formation, early pronuclear fusion due to assessment of fertilization at a single time point, or premature pronuclear breakdown ( Egli et al. , 2011 ; Azevedo et al. , 2014 ; Mateo et al. , 2017 ). Proposed mechanisms by which 1PN embryos may form at static fertilization assessment at 16–18 h . All four mechanisms may produce a biparental diploid embryo ( Balakier et al ., 1993 ; Levron et al ., 1995 ; Payne et al ., 1997 ), though developmental outcomes are variable and proportional data per mechanism are not available from the current dataset. ( A ) Asynchronous pronuclear formation—following fertilization, one pronucleus forms before the other, resulting in the temporal appearance of a single pronucleus during the observational window. ( B ) Single pronucleus enclosing both genomes—following fertilization, both maternal and paternal genetic material is enclosed within a single pronucleus. ( C ) Early pronuclear fusion—both maternal and paternal pronuclei fuse early, forming a single pronucleus during the observational window. ( D ) Premature pronuclear breakdown—either the maternal or paternal pronucleus fades early, leaving the appearance of a single pronucleus during the observational window. These mechanisms are not mutually exclusive, their relative frequencies are not established, and the precise timing of individual events within the assessment window is not known for all mechanisms. Created in BioRender. McPherson, N. (2026) https://BioRender.com/2xl4mr6 Genetic classification of uniparental MPBs varies in the literature due to the unique nature of uniparental diploid genotypes and limitations of genetic platforms. It is unlikely that any MPBs are haploid. Using fluorescent in situ hybridization to visualize ploidy status, haploid zygotes undergo cell division but arrest at the cleavage stage, with no blastocysts displaying uniform haploidy ( Sandalinas et al. , 2001 ). This suggests haploid zygotes undergo endoreplication prior to the first cell division ( van der Heijden et al. , 2009 ), resulting in a uniparental diploid blastocyst ( Kratka et al. , 2023 ). Single-nucleotide polymorphism (SNP)- or STR-based genotyping cannot distinguish uniparental isodiploidy from haploidy. In a study of purported 2PN blastocysts, 0.42% were uniparental isodiploid/haploid, and 0.04% were uniparental heterodiploid ( Kratka et al. , 2023 ). The latter represent cases where chromatid crossover occurs, introducing a small amount of heterozygosity. Triploidy represents a further ploidy category warranting consideration. Neither array-CGH nor standard read-depth NGS platforms used in this study are capable of reliably detecting triploidy ( Marin et al. , 2018 ; Levy and Burnside, 2019 ; Park et al. , 2019 ), as both rely on relative copy number comparisons that normalize to zero when all chromosomes are uniformly affected. Throughout the earlier part of the study period (2010–2018), any ploidy abnormalities identifiable by other means were classified within the aneuploid category consistent with laboratory reporting practice at that time and are therefore not separately enumerated here. Additionally, given that triploidy is associated with impaired developmental competence ( Yao et al. , 2016 ), the low blastocyst formation rate observed in 1PN zygotes (13–19%) suggests that a proportion of these 1PN zygotes may have been triploid and underwent developmental arrest prior to reaching biopsy, further reducing their representation in the tested cohort. Compounding this, the STR-based biparental classification platform has not been validated against an orthogonal technology for parental origin calling in 1PN embryos, meaning triploid embryos that did reach biopsy may have been misclassified as biparental diploid or absorbed into unreported inconclusive STR outcomes. Taken together, the true contribution of triploidy to chromosomal abnormality in 1PN embryos is likely underestimated in this dataset, and this uncertainty should be considered when interpreting the reported biparental inheritance rates. Several professional associations advocate for genetic screening of MPBs given the risks associated with hydatidiform molar pregnancies. Case reports document complete or partial hydatidiform molar pregnancies following assisted reproductive technology, though morphologic fertilization checks reduce the rate compared to spontaneous conceptions (1:4000 vs 1:1000; Nickkho-Amiry et al. , 2019 ). Cases of complete or partial hydatidiform molar pregnancy have been reported following transfer of both IVF embryos and ICSI embryos ( Nickkho-Amiry et al. , 2019 ; Dzyubak et al. , 2020 ; Zhou et al. , 2021 ; Hosseini et al. , 2023 ; Mehendale et al. , 2024 ). Most were from transfers of 2PN embryos, with one involving transfer of a euploid ICSI-derived MPB ( Zhou et al. , 2021 ). The predominantly maternal origin of uniparental inheritance in IVF-MPBs has critical implications for risk assessment. Since complete hydatidiform mole requires paternal uniparental inheritance ( Maisenbacher et al. , 2019 ), the risk of paternal uniparental inheritance from an IVF-derived MPB in this study is ∼0.2%—which is low and similar to the background risk for spontaneous conception ( Nickkho-Amiry et al. , 2019 ). The risk of hydatidiform mole development from MPBs is directly correlated with the genomic origin of the single pronucleus. The significantly lower blastocyst formation rate observed in androgenetic (paternal-only) embryos compared to parthenogenetic (maternal-only) embryos ( Lagutina et al. , 2004 ; Colaco and Sakkas, 2018 ) further diminishes the probability that an MPB selected for clinical transfer would have paternal disomy. This low risk profile supports reconsideration of current universal genetic testing requirements for IVF-derived MPBs meeting specific morphological criteria and may inform individualized clinical decision-making and consent discussions in this subgroup, especially when 2PN embryos are unavailable. Our findings on developmental potential align with previously published research. On average, 15–30% of IVF 1PN zygotes form usable blastocysts compared with 6–40% of ICSI 1PN zygotes ( Bradley et al. , 2017 ; Capalbo et al. , 2017 ; Xie et al. , 2018 ; Mateo et al. , 2020 ; Soler et al. , 2021 ; Kemper et al. , 2023 ; Tong et al. , 2023 ; Lebovitz et al. , 2024 ). Previous comparative analysis studies of IVF- versus ICSI-1PN zygotes demonstrate lower developmental rates among ICSI-derived embryos ( Itoi et al. , 2015 ; Chen et al. , 2020 ; Tong et al. , 2023 ), with the average utilizable blastocyst formation rate below 10%. Our morphology data are mostly consistent with the literature on the important characteristics of viable MPBs. For IVF-derived 1PNs, the presence of a second polar body, multiple nucleoli (>5), large pronuclear size (>710 μm 2 or diameter >29 μm), and ability to make a moderate- to good-quality blastocyst on Day 5 are associated with normal diploidy and biparental inheritance ( Araki et al. , 2018 ; Kai et al. , 2018 ; Mateo et al. , 2020 ; Fu et al. , 2022 ; Wang et al. , 2022 ; Zhao et al. , 2022 ; Xiong et al. , 2024 ). Additionally, optimal pronuclear breakdown timing and completion of the first cell cycle within normal parameters correlate with improved developmental potential and genetic normalcy ( Fu et al. , 2022 ; Wang et al. , 2022 ). Although no significant difference in PN size was observed between biparental and uniparental 1PN embryos in our IVF subset, the retrospective nature of image analysis and variability in archived image quality may have limited our ability to detect subtle morphological differences. Nevertheless, PN size and timing parameters described in published studies remain potentially valuable stratification criteria, and prospective evaluation of these features alongside genetic testing outcomes would strengthen their incorporation into individualized risk assessment frameworks. Our study contributes to substantial literature demonstrating that 1PN blastocysts can successfully implant, establish pregnancy, and result in healthy live births. We provide evidence that maternal and neonatal outcomes of 1PN-derived euploid blastocysts are equivalent to those of 2PN euploid blastocysts, further confirming their clinical safety. Since our 2017 paper ( Bradley et al. , 2017 ), several additional publications have reported successful live births from 1PN zygotes ( Capalbo et al. , 2017 ; Si et al. , 2019 ; Chen et al. , 2020 ; Li et al. , 2020 ; Wang et al. , 2022 ; Tong et al. , 2023 ), supporting their clinical value. Beyond confirming clinical safety, the present study provides novel data on the cycle rescue contribution of 1PN embryo inclusion. In approximately one in six cycles, no utilizable 2PN embryo was available, with the 1PN embryo representing the sole option for utilization. For these patients, inclusion of 1PN embryos may negate the need for an additional stimulation cycle, and the associated physical, emotional, and financial burden this entails. This is particularly relevant in cycles where no 2PN fertilization occurred at all, leaving patients who would otherwise have no embryos available from that cycle. These findings highlight that the clinical value of 1PN embryo inclusion extends beyond aggregate outcome statistics and warrants consideration within individualized patient counseling, particularly where 2PN embryo availability is limited. Most published investigations and existing clinical guidelines regarding MPBs have focused on ICSI-MPBs, with notably limited data addressing IVF-MPBs. Consequently, current clinical recommendations for comprehensive genetic testing for all MPBs are primarily predicated on the elevated rates of uniparental inheritance observed in ICSI-MPBs. Our findings reveal a marked contrast in genetic outcomes between IVF- and ICSI-MPBs, with IVF-MPBs demonstrating significantly lower rates of uniparental paternal inheritance. This substantial difference in genetic profiles suggests that a uniform testing policy may not be appropriate for both groups, and supports the development of a risk-stratified approach that distinguishes IVF- from ICSI-derived MPBs when considering genetic testing requirements and informed consent discussions. Several clinics report transferring IVF-MPBs without genetic testing ( Itoi et al. , 2015 ; Si et al. , 2019 ; Li et al. , 2021 ; Lee et al. , 2026 ); however, numerous other clinics, including our own, employ a risk-mitigation approach that incorporates mandatory PGT-A and confirmation of biparental inheritance prior to embryo transfer. The substantial data presented in this investigation suggest that stringent morphological and genetic selection criteria can be systematically applied to stratify IVF-MPBs by risk profile, potentially identifying a subgroup in which the incremental benefit of mandatory genetic testing is low. Rather than advocating a universal change in practice, we propose that these criteria form the basis of an individualized risk assessment framework to guide clinician and patient decision-making, particularly in cycles where 2PN embryos are unavailable, with the goal of expanding the available embryo pool while maintaining appropriate safeguards against paternally derived uniparental diploidy. Several important technical limitations of our approach for detecting biparental inheritance warrant acknowledgment. First, while our multi-marker STR approach builds upon the methodology validated in Bradley et al. (2017) , we acknowledge that newer SNP-based technologies, including genome-wide SNP arrays and NGS platforms with integrated genotyping capabilities, offer enhanced resolution for detecting parental origin through assessment of B-allele frequencies and can more readily distinguish between diploid, haploid, and triploid embryos as well as detect genome-wide versus segmental uniparental disomy ( ESHRE PGT Consortium Steering Committee et al. , 2020 ; Handyside et al. , 2025 ). These platforms, which are increasingly being adopted for comprehensive PGT ( Tšuiko et al. , 2021 ; Xie et al. , 2022 ; Caroselli et al. , 2023 ), and more recent publications on integrated PGT-A/ploidy analysis ( Tšuiko et al. , 2021 ), provide thousands of informative markers across the genome rather than the 8–12 STR loci used in our approach. However, at the time this study was conducted (2010–2024), our laboratory utilized the STR-based methodology that was the established standard of care and had been validated for clinical PGT-M applications ( ESHRE PGT Consortium Steering Committee et al. , 2020 ). While SNP-based approaches offer theoretical advantages in resolution and the ability to detect more subtle genomic imbalances, our clinical outcomes, with no cases of hydatidiform molar pregnancy among 257 transferred embryos suggest that the multi-chromosome STR approach provided adequate discriminatory power for identifying genome-wide uniparental disomy, which is the primary clinical concern. Another limitation of the STR-based biparental inheritance platform is that it has not been validated against an orthogonal technology, such as SNP array, whole-genome sequencing, or independent re-biopsy confirmation, for the specific application of parental origin calling in 1PN-derived embryos. The concordance data available for this platform ( Bradley et al. , 2017 ) reflect internal technical consistency within the same WGA template and do not constitute cross-platform biological validation. While STR haplotyping represents the established methodology underpinning PGT-M practice ( ESHRE PGT Consortium Steering Committee et al. , 2020 ), its application to biparental classification in 1PN embryos is a distinct use case for which application-specific independent validation has not yet been performed. This limitation is acknowledged as a priority for future work; multi-center studies incorporating orthogonal validation methods would substantially strengthen confidence in the risk-stratification framework proposed here.

The findings presented advance our understanding of the developmental potential and genetic characteristics of MPBs and highlight a clinically meaningful distinction between IVF- and ICSI-derived embryos that current universal policies do not reflect. For IVF-derived MPBs with favorable morphological and developmental features, the risk profile observed in this largest cohort to date appears substantially lower than previously recognized and supports reconsideration of mandatory genetic testing requirements in this specific subgroup. We propose that these findings provide a foundation for individualized risk stratification and informed consent discussions particularly in cycles where 2PN embryos are unavailable rather than a basis for immediate universal practice change. Prospective multi-center studies with comprehensive long-term follow-up are needed to validate these observations, refine selection criteria, and provide the level of evidence required to underpin formal guideline revision.