Pregnancy and neonatal outcomes of borderline blastocysts: a novel evaluation model based on inner cell mass and trophectoderm parameters.

This study introduces a novel method for blastocyst assessment that is both applicability and safety, enhancing the consistency of evaluations among embryologists.

After exclusions, a total of 361 SETs derived from borderline expanded blastocysts in FET cycles were included in this study. This comprised 172 group A (grade B–B−), 79 group B (grade B–C), and 110 group C (grade CB/CB−). The data showed that no significant differences were found in the rates of maternal age, duration of infertility, BMI, basal FSH, etc. (Table  2 ). Table 2 Baseline characteristics of patients among groups A–C Group A Group B Group C P Cycle ( n ) 172 79 110 Age (years) 31.0 ± 4.0 31.4 ± 4.3 30.7 ± 4.4 0.447 Previous conception (%) 59.3 (102/172) 54.4 (43/79) 60.9 (67/110) 0.657 Duration of infertility (years) 4.3 ± 3.1 4.0 ± 2.7 4.3 ± 3.1 0.705 BMI (kg/m 2 ) 23.5 ± 3.7 23.1 ± 3.3 23.1 ± 3.4 0.576 Basal FSH (mIU/ml) 7.4 ± 2.5 7.5 ± 2.6 8.0 ± 2.8 0.261 AMH (ng/ml) 4.3 ± 3.6 4.1 ± 3.2 4.3 ± 2.9 0.814 Endometrial thickness on ET day (mm) 9.3 ± 2.0 9.6 ± 2.0 9.5 ± 1.6 0.643 Primary reasons for infertility (%)  Tubal factor 48.8 (84/172) 53.2 (42/79) 41.8 (46/110) 0.278  Polycystic ovary syndrome 7.6 (13/172) 6.3 (5/79) 14.5 (16/110) 0.083  Endometriosis 8.7 (15/172) 5.1 (4/79) 7.3 (8/110) 0.590  Male factor 18.6 (32/172) 17.7 (14/79) 28.2 (31/110) 0.108  Unexplained/others 16.3 (28/172) 17.7 (14/79) 8.2 (9/110) 0.095 Endometrial preparation (%) 0.717  Natural cycles 29.1 (50/172) 34.2 (27/79) 30.9 (34/110)  Hormone therapy cycles 73.7 (122/172) 73.2 (52/79) 70.6 (76/110)  IVF/ICSI ratio 4.9 (143/29) 5.1 (66/13) 3.4 (85/25) 0.402  Proportion of day-5 and day-6 blastocysts 1.5 (104/68) 0.8 (35/44) 1.2 (59/51) 0.055  Mean fertilization rate 74.9 ± 18.6 70.8 ± 21.0 69.2 ± 20.5 0.087  Mean blastulation rate 55.0 ± 18.7 54.4 ± 18.9 53.9 ± 18.3 0.899 BMI body mass index, AMH anti-Müllerian hormone, FSH follicle-stimulating hormone Baseline characteristics of patients among groups A–C BMI body mass index, AMH anti-Müllerian hormone, FSH follicle-stimulating hormone As shown in Table  3 , we found that the rates of biochemical pregnancy, clinical pregnancy, ongoing pregnancy, and live birth were significantly lower in the group C than those in the groups A and B ( P  < 0.05). No differences were found in terms of miscarriage rate among groups A, B, and C (26.7%, 25.0%, and 40.0%). Table 3 The pregnancy and neonatal outcomes of patients in groups A–C Group A Group B Group C Cycle ( n ) 172 79 110 Biochemical pregnancy rate (%) 40.1 (69/172) 32.9 (26/79) 17.3 (19/110)* Clinical pregnancy rate (%) 34.9 (60/172) 25.3 (20/79) 9.1 (10/110)* Miscarriage rate (%) 26.7 (16/60) 25.0 (5/20) 40.0 (4/10) Ongoing pregnancy rate (%) 26.7 (46/172) 19.0 (15/79) 6.4 (7/110)* Live birth rate (%) 25.6 (44/172) 19.0 (15/79) 5.5 (6/110)* Gestational age (GA) at birth (days) 270.2 ± 11.5 267.7 ± 22.3 269.8 ± 14.5 Number of pre-term deliveries ( 41 weeks GA) (%) 0 (0/41) 0 (0/15) 0 (0/6) Mean birth weight (g) 3471.0 ± 560.3 3235.7 ± 810.3 3608.3 ± 554.5 Birth weight (≥ 4000 g) (%) 17.1 (7/41) 6.7 (1/15) 16.7 (1/6) Birth weight (< 2500 g) (%) 0 (0/41) 6.7 (1/15) 0 (0/6) Malformation rate (%) 0 (0/41) 0 (0/15) 0 (0/6) * P  < 0.01; group C compared with groups A and B The pregnancy and neonatal outcomes of patients in groups A–C * P  < 0.01; group C compared with groups A and B Regarding the neonatal outcome, there were no differences in the gestational age at birth, numbers of pre-term and post-term deliveries, mean birth weight, high birth weight, and low birth weight among groups A, B, and C. Furthermore, the malformation rate was also no difference in the three groups. Additionally, we compared LBR based on the day 5 and day 6 formed blastocysts. As shown in Fig.  2 , the LBR of day 5 blastocysts were significantly higher than that of day 6 blastocysts in group A (29.8% vs 14.7%) and group C (10.2% vs 0%), there was no difference in group B (28.6% vs 11.4%) for blastocysts on day 5 and day 6. Fig. 2 Live birth rate from the day 5 and day 6, * P  < 0.05 Live birth rate from the day 5 and day 6, * P  < 0.05 The results of binary logistic regression analysis demonstrated that the following variables were found to be associated with the LBR (Table  4 ). Blastocysts transfer of grade C ICM embryos can obtain significantly lower LBR than blastocysts with ICM of grade B− (OR = 0.158; 95% CI 0.073–0.439). In addition, compared with day 5 blastocyst transfer, day 6 blastocyst can lead to a significant reduction in LBR (OR = 0.497; 95% CI 0.241–0.838). Table 4 Binomial logistic regression for likelihood of live birth based on inner cell mass (ICM) and trophectoderm (TE), days of blastocyst expansion, and maternal age (22–38 years) Odds ratio 95% CI P value ICM grade B− Ref ICM grade C 0.158 0.073–0.439 < 0.001 TE grade B− Ref TE grade C 0.876 0.417–1.805 0.883 Day 5 Ref Day 6 0.497 0.241–0.838 0.031 Maternal age 1.061 0.910–1.179 0.824 Binomial logistic regression for likelihood of live birth based on inner cell mass (ICM) and trophectoderm (TE), days of blastocyst expansion, and maternal age (22–38 years) Furthermore, we found that there are independently factors not associated with the LBR, maternal age (range of 22–38 years) (OR = 0.982; 95% CI 0.910–1.059) and TE grade (OR = 0.871; 95% CI 0.423–1.795). Table 5 displays the agreement between the four participating embryologists when assigning a grade to the ICM, TE of blastocyst and clinical decision-making using the modified blastocyst grading system. The kappa scores reflecting the level of agreement were good agreement for ICM (0.612, 95% CI 0.559–0.665) and TE (0.633, 95% CI 0.580–0.686). For clinical decision-making, the agreement achieved very good (0.814, 95% CI 0.730–0.898). Table 5 Kappa scores for the inter-observer consistency between embryologists according to ICM, TE morphology, and clinical utilization Morphology grading Kappa scores 95% CI ICM 0.612 0.559–0.665 TE 0.633 0.580–0.686 Decision-making 0.814 0.730–0.898 Kappa scores for the inter-observer consistency between embryologists according to ICM, TE morphology, and clinical utilization Overall, the kappa scores of all participating embryologists for intra-observer agreement of ICM and TE when regarding the same set of the images ranged from 0.648 (95% CI 0.515–0.781) to 0.791 (95% CI 0.685–0.897). Similarly, when regarding the clinical decision-making, all embryologists had a very good level of agreement (kappa scores ranged from 0.860 to 0.930) (Table  6 ). Table 6 Kappa scores for the intra-observer consistency of embryologists according to ICM, TE morphology, and clinical utilization Morphology grading Embryologist 1 Embryologist 2 Embryologist 3 Embryologist 4 ICM 0.740 (0.622–0.858) 0.791 (0.685–0.897) 0.682 (0.559–0.805) 0.726 (0.608–0.844) TE 0.768 (0.656–0.880) 0.784 (0.676–0.892) 0.648 (0.515–0.781) 0.717 (0.595–0.839) Decision-making 0.886 (0.790–0.982) 0.930 (0.852–1.008) 0.860 (0.752–0.968) 0.883 (0.783–0.983) Kappa scores for the intra-observer consistency of embryologists according to ICM, TE morphology, and clinical utilization 0.791 (0.685–0.897) 0.682 (0.559–0.805) 0.726 (0.608–0.844) 0.768 (0.656–0.880) 0.784 (0.676–0.892) 0.648 (0.515–0.781) 0.717 (0.595–0.839) 0.886 (0.790–0.982) 0.930 (0.852–1.008) 0.860 (0.752–0.968) 0.883 (0.783–0.983)

This was a retrospective cohort analysis of 364 frozen-thawed single-blastocyst transfer cycles performed during Jan 2019 and Dec 2023 at the Department of Reproductive Medicine Center in the 901st Hospital of the Joint Logistics Support Force of PLA. Three cycles were excluded due to embryo degeneration after thawing, 1 cycle in group B and 2 cycles in group C. All enrolled subjects were divided into three groups according to pre-frozen blastocyst morphology: group A ( n  = 172): grade B–B−, group B ( n  = 79): grade B–C, group C ( n  = 110), and grade CB/CB−. Patients who underwent standard IVF or ICSI were eligible for enrollment. The inclusion criteria were: (1) women age ≤ 38 years, (2) embryos survived and re-expansion after thawing. The exclusion criteria included: (1) multiple embryo transfer cycles, (2) embryos involving preimplantation genetic testing treatment, (3) embryos derived from vitrified-warmed oocytes, and (4) any factors which could influence embryos implantation, including untreated hydrosalpinx, intrauterine adhesion, submucosal fibroids or uterine polyps, etc. The controlled ovarian stimulation (COS) protocol and IVF/ intracytoplasmic sperm injection (ICSI) procedures in our center were described in previous studies [ 14 ]. Either a gonadotrophin-releasing hormone (GnRH) antagonist or a GnRH agonist treatment protocol was adopted. Oocytes were fertilized by conventional IVF or ICSI according to the standard insemination procedures on the day of oocyte retrieval. The fertilized oocytes were transferred to G1-plus media (Vitrolife, Sweden) overlaid with paraffin oil (Ovoil™, Vitrolife) and cultured until day 3 at 37 ℃, 6% CO2 and 5% O2 condition (Planer BT37, Origio, USA). Afterward, utilizable embryos were placed into G2-plus media (Vitrolife, Sweden) and incubated to blastocyst stage. Assessment of fertilization and embryo morphology was performed with an inverted microscope with Hoffman modulation contrast microscopy (Olympus IX71, Japan) at 250 × magnification. The oocytes with two pro-nuclei (2PN) and two polar bodies at 16–18 h after insemination were considered as normal fertilization. On day 3, assessment of embryos was performed according to Istanbul consensus. Grades A and B embryos were defined as good quality; grade C embryos were defined as poor quality, and grade D embryos were discarded. Grades A, B, and C embryos were considered as utilizable embryos. On days 5 or 6, blastocysts evaluation were performed according to modified scoring criteria, the new criteria partly include ICM morphometry and TE cell counting, and the detailed grading system as referred in Table  1 . As a general rule, blastocysts were assessed by at least two trained embryologists to reduce potential inter-observer variations in embryo grading. The degree of blastocoel expansion was assessed according to original Gardner criteria and was defined as follows: 1 = early blastocyst, (blastocoel  50% of the volume of the embryo), 4 = expanded blastocyst, (blastocoel larger than the embryo and thinning zona), 5 = hatching blastocyst, (TE is herniating through the zona), 6 = hatched blastocyst, (blastocyst completely escaped from the zona). For assessment of the ICM and TE cells, grade assignment depended on embryo characteristics were classified in different categories: A, B, B−, and C. The morphologically grade B–B−, B–C, and CB/CB− blastocysts were defined as borderline blastocysts. All blastocysts included in this study were derived from expansion stage 4 blastocysts, as refer Fig.  1 , 250× magnification. Table 1 Description of Gardner grading system and modified grading system Grade Gardner Modified Description Description ICM A Many cells and tightly packed Larger in diameter than 60 μm, symmetrical, tightly adhered together and compacted B Several cells and loosely grouped Larger in diameter than 60 μm, unsymmetrical, loosely grouped together B− Not described Obviously smaller in diameter than 60 μm and loose and faintly visible cells C Very few cells ICM are largely unobservable or degenerate TE A Many cells forming a cohesive epithelium Regular shape, more than 10 cohesive and consecutive cells B Few cells forming a loose epithelium About 10 loose cells, part of cells lack consecutiveness B− Not described Obviously less than 10 cells and more than 5 cells which appear irregular shape C Very few large cells Very few cells (less than 5 cells) or degenerate cells Fig. 1 Examples of blastocysts in the modified blastocyst grading system. 250× magnification. A 4AA; B 4BB; C 4B–B−; D 4BC; E 4B–C; F 4CB−; G 4CC. For TE assessment, cells were counted along the equatorial cross-section of the blastocyst circumference Description of Gardner grading system and modified grading system Examples of blastocysts in the modified blastocyst grading system. 250× magnification. A 4AA; B 4BB; C 4B–B−; D 4BC; E 4B–C; F 4CB−; G 4CC. For TE assessment, cells were counted along the equatorial cross-section of the blastocyst circumference The protocol for blastocyst vitrification and warming was performed according to the manufacturer’s instruction (Kitazato, Japan). Briefly, laser artificial shrinkage was performed before vitrification, the collapsed blastocyst was placed into equilibration solution (ES) containing 7.5% ethylene–glycol (EG) and 7.5% dimethyl-sulfoxide (DMSO) for 8–10 min, and transferred to vitrification solution (VS) with 15% EG, 15% DMSO, 0.5% sucrose for 1 min, and the blastocyst was then loaded on the Cryotop strip (Kitazato, Japan) and plunged into liquid nitrogen immediately. For the blastocyst thawing procedure, the Cryotop were immediately moved from liquid nitrogen into 37 °C thawing solution (TS) of 1 M sucrose for approximately 1 min, and then, the embryo was transferred in the diluent solution (DS) of 0.5 M sucrose for 3 min at room temperature, and embryo was then washed twice in the wash solution (WS) for 5 min, respectively. The embryos were then transferred to G2-plus culture medium for observation under an inverted microscope and recording by photography. Finally, the embryo was incubated for at least 2 h in G2-plus media before ET. All enrolled subjects underwent hormone replacement therapy (HRT) program or natural cycles (NC) to prepare endometrium of frozen embryo transfer (FET) [ 15 ]. Embryo was transferred under abdominal ultrasound guidance with a Wallace catheter. After transplantation, oral dydrogesterone 10 mg twice a day was administered to all the patients for luteal support. The pregnancy outcomes were evaluated as clinical rate and live birth rate, with following including rates of ongoing pregnancy and miscarriage. Serum human chorionic gonadotropin (hCG) was measured 14 days after the blastocyst transfer, biochemical pregnancy was confirmed by hCG concentrations were ≥ 50 IU/L. Clinical pregnancy was confirmed by fetal heartbeat or gestational sac 30 days after embryo transfer. Ectopic pregnancy was defined as a gestational sac that occurred outside of the uterine cavity. Miscarriage was defined by pregnancy loss after detection of fetal heartbeat on ultrasound, the patients with ectopic pregnancy were not accepted for calculating miscarriage rate. Ongoing pregnancy was defined as the presence of a fetal heartbeat past the 12 weeks of gestation. Live birth was defined as a live infant born after 24 weeks of gestation. The neonatal outcomes were evaluated as gestational age (GA) at birth, birth weight, and malformation. Four embryologists have at least six months experience using this modified blastocyst grading system. They were asked to view 90 still images of day 5 embryos. Still images collected from photographs before embryo transfer of previously cycles in our center. Each image received an identification code and was copied four times, and then distributed to the participating embryologists. Participants had to grade each embryo and make a clinical decision. They were kept in the blind about other embryologists’ assessments to minimize bias. For determination of intra-observer agreement, 6–8 weeks after the initial assessment, embryologists had to grade each embryo and make a clinical decision to the same set of still images. Before the second assessment, the order of the still images was manipulated randomly in an attempt to eliminate recall bias. All statistical calculations were performed using Statistical Package for the Social Sciences (SPSS) version 26.0 software (IBM Corp., Armonk, NY, USA). Quantitative variables were expressed as mean ± standard deviation (SD) compared by one-way analysis of variance (ANOVA) and followed by a post hoc test (Dunnett). Categorical variables were described as frequencies and percentages (%, n / N ) compared by the Chi-square test or Fisher’s exact test. A two-sided P value of 0.05 or less was considered statistically significant; meanwhile, odds ratios (OR) with 95% confidence intervals (CI) were calculated. A binary logistic regression was performed to identify potential confounding variables associated with live birth. Statistical significance was considered at P  < 0.05. For the determination of inter- and intra-observer agreement, the Fleiss–Kappa statistic was used to evaluate the extent of variability. The kappa values range from 1 to 0 represents complete agreement to no agreement among raters. For intermediate values, the level of agreement is 0.1–0.2: poor; 0.21–0.40: fair; 0.41–0.60: moderate; 0.61–0.80: good; and 0.81–1.00: very good.

Borderline blastocysts are blastocysts of low quality that fall in between available and discarded. Given the subjective nature of the assessment, embryologists are likely to disagree on whether to freeze or transfer borderline blastocysts. This is a modified grading system for the evaluation of blastocysts; in fact, it is mainly for the assessment of blastocysts with borderline morphology. In this study, we propose the modified blastocyst grading system to help embryologists make the right choice when only borderline blastocysts are available for transfer. This retrospective study investigates the grading of borderline blastocysts for transfer and the inter- and intra-observer agreement using the modified blastocyst grading system. The results suggest that borderline blastocysts have the potential for implantation and live birth without adverse effects on neonatal outcomes. Furthermore, inter- and intra-observer agreement between embryologists in the assessment of ICM, TE grades, and clinical decision-making ranges from good to very good. The Gardner grading system is currently the most extensively reported blastocyst grading system in the literature, and it appears to be the most clinically validated system by multiple clinics throughout the world, with numerous studies demonstrating that the blastocyst score is associated with clinical outcomes [ 16 – 18 ]. The grading method, on the other hand, intrinsically leads to higher inter- and intra-observer variability, because it relies on visual assessment of blastocyst morphological features and ambiguous descriptions of embryo shape [ 4 , 13 ]. For blastocyst transfer with morphology ranging from outstanding to good, this difference has little or no impact on clinical outcome [ 19 , 20 ]. However, not all patients receive good-quality blastocysts after IVF processes, particularly women with advanced age, low responders, or repeated poor-quality embryos. For these patients, selecting poor-quality or borderline embryos with the possibility for successful implantation is an essential aspect of IVF. When only questionable (including borderline morphology) blastocysts are available, embryologists are unsure whether to use or discard them. This is a question of both clinical value and ethical obligation. It is reasonable to assume that if embryologists select borderline embryos for transfer, this will increase the chance of successful implantation for patients, even if only slightly, and avoid the possibility of unnecessary repeat treatment, additional time and cost, and enormous psychological distress. In fact, too many embryos that have the potential for a successful pregnancy are currently discarded rather than cryopreserved [ 21 ]. However, undergoing non-selective transfer of low-grade embryos that fail to implant is also a financial and emotional burden for patients. With this in mind, low-grade embryos with implantation potential should be given more attention than embryos with excellent or good morphology. If only borderline embryos remain in the cycle, more information is required to determine which embryos should be transferred as a priority. Currently, contradictory results have been reported from several studies when low-grade embryos were transferred [ 9 , 22 – 24 ]. The outcomes of embryo selection vary, because embryologists disagree on the use of low-grade blastocysts as a result of blastocyst evaluation, which is far from standardized. According to Hammond et al., there is widespread disagreement regarding whether to transfer a blastocyst with a borderline shape, which is characterized by few or large TE cells and/or poorly differentiated ICM (kappa 0.30) [ 11 ]. Previous study indicated that when the embryologists graded ICM and TE for embryos with full blastulation, the overall agreement between the embryologists was only fair (0.349 and 0.397, respectively) [ 3 ]. The low level of agreement between TE and ICM scores could be due to bias from the greater subjectivity involved in assessing TE and ICM scores, making the results difficult to interpret. Therefore, it is necessary to accurately and easily quantify very detailed embryonic morphological parameters in the blastocyst grading system [ 11 ]. To reduce inter- and intra-observer variability in embryo grading, many modified blastocyst grading methods have been developed in recent years to assist in selecting and identifying the most appropriate blastocysts for transfer; such as automated blastocyst morphology grading of ICM and TE using time-lapse images [ 4 , 25 ], grades of ICM and TE assessment combined with degree of expansion [ 13 ], preimplantation genetic testing for aneuploidies (PGT-A) technology [ 26 ], and artificial intelligence (AI) [ 27 ]. However, embryologists' knowledge, methodologies, and specialized equipment will always influence embryo quality assessment, making it difficult to compare research and potentially limiting their applicability [ 28 ]. The data show that our blastocyst grading system could be easily implemented by an experienced embryologist to determine the morphological grade of the blastocyst (ICM morphometry and TE cell count). If some completely grown blastocysts cannot be identified by their ICM or TE cells, they can be viewed by rotating the blastocyst interface or decreasing the blastocoel cavity using a laser. Additionally, when four embryologists were asked to grade and decide the clinical decision-making of the identical batch of 90 still photographs, all embryologists demonstrated ‘good’ or ‘very good’ agreement among themselves. This suggests that our modified blastocyst grading method is reliable. Although only four embryologists took part in the study, they all had similar experience and backgrounds, which has been proved to be an effective method of reducing bias [ 3 , 29 , 30 ]. To minimize the impact of fresh cycles on pregnancy outcomes due to endometrial receptivity, we chose frozen cycles for embryo transfer. This approach proved that borderline blastocysts could obtain an acceptable birth rate. Furthermore, the association between blastocyst score and LBR suggests that ICM grade might be considered preferential to TE grade in borderline blastocysts. This is consistent with the recent literature reports, suggesting that ICM is superior to TE scoring in terms of live births [ 31 – 34 ]. However, scholars who hold the opposing viewpoint argue that because TE eventually develop into the placenta, continued ICM development in vivo will be impossible without adequate TE establishment for implantation [ 35 , 36 ]. Besides this, we found that the degeneration rate of borderline blastocysts after thawing was at standard level (0.8%, 3/364, the survival rate of the blastocyst after thawing in our laboratory is over 99%). However, the poor level of re-expansion of the blastocysts (13.9%) was higher than that of our laboratory data from good-grade blastocysts after thawing (< 10%, unpublished). The data showed that poor-quality blastocysts had a lower re-expansion rate after thawing than good-grade blastocysts [ 37 ]. Previous studies have shown that reduced cryo-survival rates with decreasing blastocyst quality [ 38 ], this could potentially influence the clinical outcome [ 39 ]. It should be highlighted that, whereas our analysis covered 361 single borderline embryo transfer sessions, the small sample size is still a limitation. There are two causes for this: first, single borderline embryo transfer is not a frequent technique in our facility; to optimize the patient's benefit, we normally choose two embryos (1 good plus 1 borderline or 2 borderlines) for transfer when there are more embryos. Second, many patients elect to end their current cycle and begin a new one if they only have one borderline embryo after being informed by workers about the low pregnancy rate associated with single borderline embryo transfers. Additionally, all embryos analyzed in this study were expansion stage 4 blastocysts. This selection reflects our laboratory's standard operating procedure (SOP), which requires blastocysts to reach at least expansion stage 4 to qualify for cryopreservation. Consequently, blastocysts at later stages (stage 5–6) were excluded from the statistical analysis due to inadequate sample size for robust subgroup comparisons. In conclusion, each IVF institution has its own set of standards for transferring or discarding borderline embryos; embryo grading is inherently subjective, and there is significant inter- and intra-observer variability. The findings of this study demonstrated the clinical outcomes of borderline blastocysts by assessing their ICM and TE, helping embryologists to make clinical decisions when selecting embryos with borderline morphology. In the future, we intend to collect additional data to properly determine the threshold of borderline blastocysts with minimal inter- and intra-observer variability, and then reassess their clinical value.

Evaluation of blastocyst morphology has always been an integral part of embryo selection for cryopreservation or embryo transfer. Blastocyst grading systems provide morphological criteria for different grades of ICM and TE. Currently, the morphological assessment of the blastocyst is still done according to the Gardner grading system [ 1 ] and has been shown to be associated with CPR and LBR [ 2 ]. However, the current blastocyst grading system does not provide the level of consistency required for optimal patient care. Although embryo grading systems ultimately determine whether an embryo is transferred or discarded, the process is highly subjective [ 3 , 4 ]. The reason for this lies in the description of embryonic grading and in the understanding and application of these criteria by individual embryologists. Grading systems continuous ranges of morphological appearance into preset groups (e.g., poor, medium, or good quality). The imprecise boundaries between categories can cause misunderstandings between neighboring grades [ 4 ]. For example, grades A, B, and C were described as ‘many cells’, ‘some/few cells’, and ‘very few cells’, respectively [ 5 ], which is known to be subjective and, in line with poorly distinguishable TE or ICM, would lead to disagreement between embryologists on the selection and grading of blastocyst [ 3 ]. Previous research has demonstrated good inter-observer agreement in clinical decision-making (to use or discard these embryos) for good-quality blastocysts [ 6 ]. However, when blastocysts have poor or borderline morphology, there is more disagreement among embryologists in making decisions [ 7 ]. Thus, there are conflicting conclusions in the literature regarding the value of morphological grading of poor-quality embryos in predicting clinical outcomes [ 8 – 11 ]. In general, ICM C grade blastocysts appear to have little clinical value, and hence, most in vitro fertilization (IVF) clinics and embryologists choose to discard them [ 9 , 12 ], but recent studies have reported that grade C blastocysts may still have the potential to be euploid and result in clinical pregnancy rate (CPR) and neonatal outcomes [ 8 , 10 ]. The inconsistency may be due to a lack of objectivity in the assessment process for low-quality embryos, which results in a high level of subjectivity and systematic error in morphological grading [ 3 , 7 , 13 ]. The aim of this study was to assess the efficacy and safety of blastocysts with borderline morphology evaluated in subgroups with a blastocyst grading system that measures the parameters of the ICM and TE (grades B–B−, B–C and CB/CB−).