Does quantity equal quality?-A morphokinetic assessment of embryos obtained from young women with decreased ovarian response to controlled ovarian stimulation

Purpose To assess oocyte quality in young patients with decreased ovarian response to controlled ovarian stimulation using time-lapse analysis.

A retrospective cohort study conducted at five medical centers between 2013 and 2017. The “decreased ovarian response” (DOR) group consisted of 241 women who underwent controlled ovarian stimulation with ≤ 5 retrieved oocytes and 519 cultured embryos. The “normal response” (NOR) group consisted of 667 women with ≥ 6 retrieved oocytes resulting in 3633 embryos. Data included annotation of morphokinetic events of embryos cultured in a time-lapse incubator from time of pronuclei appearance to time of starting blastocyst formation (tSB). Comparison was made between morphokinetic parameters of DOR and NOR patients with additional subgroup analysis according to the implantation status.

Implantation and clinical pregnancy rates were significantly higher in the NOR group compared with the DOR group (44.5% vs. 31.6% and 51.5% vs. 37.7%, respectively; p < 0.05). Embryos from the DOR group reached the morphokinetic milestones later than embryos obtained from NOR patients. In the DOR group, implanted embryos reached starting blastocyst formation (tSB) faster than embryos which failed to be implanted, however, manifested a protracted course compared with implanted embryos from the NOR group. In a multivariate analysis—decreased ovarian response, nulliparity, number of transferred embryos, and t4, and were predictive for implantation.

The quantitative decrease in ovarian response is associated with reduced oocyte quality, reflected by a slower developmental rate and lower implantation and pregnancy rates. Supplementary Information The online version contains supplementary material available at 10.1007/s10815-021-02113-4.

Low ovarian response, Poor ovarian response, Morphokinetics, Time-lapse monitoring system

Women presenting with low ovarian reserve are considered a challenging population in the in vitro fertilization (IVF) practice. The estimation of ovarian reserve can be performed by either pre-treatment evaluation using endocrine and sonographic markers [1] or by post-treatment assessment of the ovarian response to controlled ovarian stimulation (COS). An adequate response to ovarian stimulation, exhibited by a sufficient number of retrieved oocytes, is the cornerstone of IVF treatments and is believed to be one of the factors affecting the chances to conceive [2]. A failure to retrieve enough oocytes is considered a decreased response and is relative to patient age. The long-standing controversy regarding whether these women exhibit only a quantitative decrease in ovarian reserve or also impaired oocyte quality has not yet been settled. Some studies demonstrated an increased rate of low-grade embryonic morphology and pregnancy loss [3, 4] while others failed to establish such an association [5, 6]. It is well known that there is a physiologic decline in both oocyte quantity and quality attributed to aging [7]. However, it has not been demonstrated that the two are necessarily related when a premature decline in oocyte quantity is observed, suggesting perhaps a different pathophysiology responsible for this condition. While the aforementioned studies included populations of diverse patient ages, a recent publication [8] focused only on patients exhibiting decreased ovarian reserve or poor response to stimulation at young patient ages. This study demonstrated that blastulation, aneuploidy, and live birth rates per fertilized oocyte were equivalent to those obtained from women with normal ovarian reserve. Nonetheless, when assessing oocyte quality as represented by embryonic development and embryo transfer outcomes, a more comprehensive and detailed evaluation can be performed using time-lapse analysis. The predictive value of temporal developmental milestones regarding both embryo developmental capacity and clinical outcomes, manifested by the blastulation and implantation rates, was previously described by others [9–11]. In light of this association, it seems not only that TLM enables detailed morphologic and kinetic evaluation but also might facilitate the selection of embryos with the best reproductive potential. To the best of our knowledge, only scarce data have accrued concerning the morphokinetic evaluation of embryos obtained from poor responders, with only one case series of 53 women published recently [12]. Therefore, the aim of this retrospective study is to compare embryonic development by morphokinetic evaluation of oocytes obtained from women younger than 38 years with a decreased ovarian response, with patients with normal response.

Settings This retrospective study included data analysis of videos obtained from embryos cultured in nine time-lapse incubators (Embryoscope, Vitrolife) at five medical centers in Israel between January 2013 and December 2017. Each of the IVF laboratory video files and their corresponding metadata were collected from patient electronic records, including age, IVF indication, treatment protocol, and number of oocytes retrieved. Data regarding ovarian reserve assessment and gonadotropin dosage were not included in the analysis. Annotation of morphokinetic events during embryonic development recorded on the video files from the time-lapse incubators was performed by a trained embryologist. The study group, defined as the “decreased ovarian response” group (DOR) included patients younger than 38 years who underwent COS with an aspiration of 5 oocytes or less. The comparison group, referred to as the “normal ovarian response” group (NOR) was comprised of patients younger than 38 years who underwent COS with aspiration of 6 oocytes or more. All women included in the study were without known risk factors for genetic or acquired/induced ovarian insufficiency. Exclusion criteria were age 38 or older, infertility factors that might affect embryo quality, including endometriosis, polycystic ovarian syndrome, and preimplantation genetic diagnosis, and couples in which the males were diagnosed with sperm abnormalities. The study was approved by the Institutional Ethical Review Board of Hadassah Hebrew University Medical Center (IRB number HMO 558-14). Data were imported under the approval of the Institutional Review Board in each hospital. Controlled ovarian stimulation, embryo culture, and transfer COS protocol consisted of either GnRH agonist protocol (short or long) or an antagonist protocol. Intracytoplasmic sperm injection (ICSI) was used to fertilize mature oocytes, and the injected oocytes were placed in culture slides (EmbryoSlide, Unisense FertiliTech, Aarhus, Denmark) containing 12 micro-wells, each filled with 25-μl droplets of a single step Global medium or one-step medium “SAGE 1-step” (SAGE, Al-rad medical, Nes Ziona, Israel), covered with mineral oil to prevent evaporation. The oocytes were incubated in a time-lapse incubator—an EmbryoScopeTM system, at 37°C, 5.7% CO2, and 5% O2 (Unisense FertiliTech, Aarhus, Denmark). Fresh embryo transfer was performed on either day 3, 4, or day 5–6/blastocyst stage (according to patient and physician decision). Day 4 transfers (less than 10% of transfers) were regarded as cleavage-stage transfers unless start of blastocyst formation was observed, in which case the transfer was regarded as blastocyst transfer. Edwards-Wallace catheter (Classic Embryo Replacement Catheter; Smiths Medical, Hythe, Kent, U.K.) or the SIVF catheter (K-Jets-7019-SIVF; Cook IVF, Eight Miles Plains, Queensland, Australia) were used for the embryo transfer upon operator preference. Morphologic and morphokinetic parameter evaluation Recording of oocyte fertilization and embryo development events was performed using seven-frame z-stacks, 15 μm apart, recorded at 18- to 20-min intervals for up to six days of incubation. This allowed acquisition of continuous three-dimensional imaging of preimplantation embryo development and later analysis. Morphokinetic profiles were based on time series of these discrete events as defined previously [13]: pronuclei appearance (tPNa) and fading (tPNf), cleavage of N discrete cells (tN; N = 2 to 5 and 8), morula compaction (tM), and starting of blastulation (tSB). Additional dynamic preimplantation embryo development definitions included: second embryonic cell cycle (ECC2 = t4–t2), third embryonic cell cycle (ECC3 = t8–t4), the synchronous two blastomere divisions within the second cell cycle (S2 = t4–t3), and the synchronous four blastomere divisions within the third cell cycle (S3 = t8–t5). Morphology was additionally assessed by embryo fragmentation levels on day 3–66 h post insemination (0–10%, 10–20%, 20–50%). The morphokinetic parameters presented were calculated using tPNf as t0, in order to minimize the effect of shared early components. Direct cleavage and reverse cleavage embryos were excluded from the study due to their altered morphokinetic behavior. Implantation and clinical pregnancy data Data regarding the implantation of the transferred embryos are presented as Known Implantation Data (KID) and defined as one of two: (1) positive (KIDp), when the number of gestational sacs is equal to the number of transferred embryos; (2) negative (KIDn), when none of the embryos transferred implanted. KID unknown (KIDu) included cases in which implantation outcome was unknown, e.g., the number of implanted embryos was smaller than the number of transferred embryos. The existence of fetal heartbeat demonstrated by ultrasound measurement (week 6 of pregnancy or later) defined the cycle outcome as the implantation of a viable fetus and was labeled as clinical pregnancy (CP). Statistical analysis Quantitative variables were compared using a two-tailed Student’s t-test for the two independent groups. Association between two categorical parameters was tested by either chi-square test or the Fisher’s exact test as appropriate. Basic categorical variables and morphologic and several dichotomous morphokinetic parameters were similarly compared between the groups using the chi-square test. The correlation between each morphokinetic event and the rest of the morphokinetic events was tested using the Spearman’s correlation test. A correlation was considered significant when R (correlation coefficient) was > 0.7 with a p value of < 0.05. Variables that were statistically significant and associated with the dependent variable of KID positive embryos were included in the multivariate logistic regression model. The logistic regression model was used to simultaneously assess the effect of the morphokinetic variables and other clinical characteristics tested on dichotomous dependent variable—KID-positive embryos. A mixed logistic regressions with random effects of mother and the medical center was performed, using backward selection of parameters with entry and stay criteria of p-value of ≤ 0.1. Each independent variable was tested for significance and the odds ratio (95% confidence interval) was calculated for the entire population. A p-value of < .05 was considered statistically significant.

Study population The study population included 3633 embryos generated from oocytes retrieved during a COS cycle of 667 women younger than 38 years (mean age of 31.9 ± 3.8 years), defined as normal responders, and 519 embryos generated from COS cycles of 241 women younger than 38 years (mean age of 33.2 ± 3.5 years) demonstrating decreased ovarian response to ovarian stimulation. IVF treatments were conducted between January 2013 and December 2017 in the five medical centers participating in this study. The DOR group was significantly older than the NOR group, had a higher nulliparity rate, more previous COS cycles, and less embryos transferred and frozen (Table 1). Blastulation and fragmentation rates, however, were similar. Three hundred twenty-six and 1169 embryos were transferred at a fresh embryo transfer in the DOR and the NOR groups, respectively. Table 1. | Parameter | Decreased response | Normal response | p-value | |---|---|---|---| | N of embryos | 519 | 3633 | | | N of patients | 241 | 667 | | | Age at aspiration (years) | 33.2 ± 3.5 | 31.9 ± 3.8 | < 0.001 | | BMI | 25.6 ± 0.5 | 26.5 ± 0.5 | 0.514 | | Nulliparous | 294/519 (56.7%) | 1852/3633 (51.0%) | 0.016 | | N of previous treatment cycles | 2.5 ± 2.3 | 1.9 ± 1.4 | < 0.001 | | Mean no. of oocytes retrieved | 3.8 ± 1.4 | 15.2 ± 7.3 | < 0.001 | | N of fresh embryo transferred | 1.3 ± 0.7 | 1.1 ± 0.7 | < 0.001 | | N of frozen embryos | 0.7 ± 0.9 | 3.7 ± 4.9 | < 0.001 | | Blastulation rate | 115/147 (78.2%) | 1642/2004 (81.9%) | 0.263 | | Fragmentation (8 cells) | 0.182 | || | 0–10% | 297/312 (95.2%) | 2713/2797 (97.0%) | | | 10–20% | 14/312 (4.5%) | 81/2797 (2.9%) | | | 20–50% | 1/312 (0.3%) | 3/2797 (0.1%) | | | N of transferred embryos | 326 | 1169 | | | Stage at transfer | < 0.001 | || | Cleavage | 262/326 (80.4%) | 780/1169 (67.1%) | | | Blastocyst | 64/326 (19.6%) | 389/1169 (32.8%) | | | Implantation rate according to stage at transfer | ||| | Cleavage stage | 79/262 (30.2%) | 324/780 (41.5%) | 0.001 | | Blastocysta | 23/58 (39.7%) | 186/389 (47.8%) | 0.247 | | Total implantation rate | 102/320 (31.9%) | 510/1169 (43.6%) | < 0.001 | | Clinical pregnancy rate by transfer stage | ||| | Cleavage stage | 70/198 (35.3%) | 293/603 (48.6%) | 0.001 | | Blastocyst | 23/46 (50.0.%) | 183/322 (56.8%) | 0.384 | | Total clinical pregnancy rate | 93/244 (38.1%) | 476/925 (51.5%) | < 0.001 | Data presented as mean ± SD or n/N (%) COS, controlled ovarian stimulation; BMI, body mass index aImplantation data were not available for 6 blastocysts in the decreased response group Implantation rates and clinical pregnancy rates were significantly lower in the DOR group compared with the NOR groups (31.9% vs. 43.6%. and 38.1% vs. 51.5%, respectively; p < 0.001, Table 1), especially in cleavage-stage embryo transfer. Morphokinetic events In the entire cohort of embryos evaluated, embryos from the DOR group reached the morphokinetic milestone events of, t4, t8, tM, tSB, and ECC2 (t4–t2) significantly later than embryos from the NOR group (Table 2). Table 2. | Parameter | Decreased response (n (%)) | Normal response (n (%)) | p-value | |---|---|---|---| | t2 | 2.6 ± 1.0 (519) | 2.5 ± 1.2 (3633) | 0.291 | | t3 | 14.0 ± 4.8 (505 (97.3%)) | 13.2 ± 4.4 (3633 (100%)) | < 0.001 | | t4 | 16.1 ± 5.0 (503 (96.9%)) | 15.3 ± 5.0 (3581 (98.6%)) | <0.001 | | t5 | 26.9 ± 8.1 (413 (79.6%)) | 26.3 ± 7.9 (3441 (93.9%)) | 0.122 | | t8 | 37.0 ± 9.8 (348 (67.1%)) | 9.8 ± 9.3 (3209 (88.3%)) | 0.023 | | tM | 67.2 ± 11.6 (179 (34.5%)) | 63.4 ± 10.6 (2240 (61.7%)) | < 0.001 | | tSB | 79.4 ± 11.1 (119 (22.9%)) | 74.7 ± 9.3 (1708 (47.0%)) | < 0.001 | | ECC2 (t4–t2) | 13.6 ± 4.7 (503 (96.9%)) | 12.8 ± 4.7 (3581 (98.6%)) | < 0.001 | | ECC3 (t8–t4) | 21.2 ± 8.0 (348 (67.1%)) | 20.8 ± 9.1 (3208 (88.3%)) | 0.410 | | S2 (t4–t3) | 2.4 ± 4.3 (501 (96.5%)) | 2.1 ± 4.4 (3581 (98.6%)) | 0.294 | | S3 (t8–t5) | 10.7 ± 9.2 (347(66.9%)) | 9.8 ± 9.3 (3209 (88.3%)) | 0.060 | Data presented as mean ± SD (n of embryos (% of n/N) t(N), cleavage of N discrete cells; tM, time to morula; tSB, time to start of blastulation; ECC, embryonic cell cycle; S2, the synchronous two blastomere divisions within the second cell cycle; S3, the synchronous two blastomere divisions within the third cell cycle This trend was stronger in the comparison between transferred embryos from the DOR and NOR groups, in which nearly all morphokinetic events (except for t3 and t5) occurred significantly later in the DOR group (Table 3). Table 3. | Parameter | Decreased response (n (%)) | Normal response (n (%)) | p-value | |---|---|---|---| | t2 | 2.6 ± 1.1(326) | 2.4 ± 0.6(1219) | 0.037 | | t3 | 14.3 ± 4.3 (316 (96.9%)) | 13.9 ± 2.3 (1219 (100%)) | 0.133 | | t4 | 15.6 ± 4.1 (316 (96.9%)) | 15.0 ± 2.5 (1206 (98.9%)) | 0.008 | | t5 | 27.2 ± 6.8 (248 (76.1%)) | 27.1 ± 4.7 (1052 (86.3%)) | 0.918 | | t8 | 34.9 ± 7.6 (200 (61.3%)) | 32.7 ± 6.4 (979 (80.3%)) | < 0.001 | | tM | 65.2 ± 11.5 (77 (23.6%)) | 60.5 ± 8.2 (512 (42.0%)) | < 0.001 | | tSB | 78.0 ± 8.7 (60 (18.4%)) | 71.4 ± 7.3 (433 (35.5%)) | < 0.001 | | ECC2 (t4–t2) | 13.0 ± 3.5 (316 (96.9%)) | 12.5 ± 2.4 (1206 (98.9%)) | 0.016 | | ECC3 (t8–t4) | 19.7 ± 5.9 (200 (61.3%)) | 17.8 ± 5.2 (978 (80.2%)) | < 0.001 | | S2 (t4–t3) | 1.6 ± 3.2 (314 (96.3%)) | 1.1 ± 2.4 (1206 (98.9%)) | 0.012 | | S3 (t8–t5) | 8.0 ± 6.5 (199 (61.0%)) | 5.5 ± 5.5 (979 (80.3%)) | < 0.001 | | Decreased response KIDp (n (%)) | Normal response KIDp (n (%)) | p-value | | | t2 | 2.5 ± 0.7 (80) | 2.3 ± 0.4 (461) | 0.895 | | t3 | 14.5 ± 4.6 (80 (100%)) | 13.9 ± 1.3 (461(100%)) | 0.014 | | t4 | 15.1 ± 2.5 (80 (100%)) | 14.4 ± 1.4 (461(100%)) | 0.012 | | t5 | 27.8 ± 4.5 (61 (76.3%)) | 27.0 ± 3.5 (414 (89.8%)) | 0.002 | | t8 | 34.2 ± 6.2 (51 (63.8%)) | 31.1 ± 4.6 (401 (87.0%)) | 0.011 | | tM | 67.1 ± 9.0 (21 (26.3%)) | 59.1 ± 7.3 (224 (48.6%)) | 0.001 | | tSB | 73.9 ± 7.9 (17 (21.3%)) | 70.2 ± 6.3 (190 (41.2%)) | 0.236 | | ECC2 (t4–t2) | 12.6 ± 1.7 (80 (100%)) | 12.1 ± 1.3 (461 (100%)) | 0.013 | | ECC3 (t8–t4) | 19.4 ± 5.3 (51 (63.8%)) | 16.7 ± 3.8 (401 (87.0%)) | 0.061 | | S2 (t4–t3) | 1.1 ± 2.5 (79 (98.8%)) | 0.5 ± 0.7 (461 (100%)) | 0.221 | | S3 (t8–t5) | 6.8 ± 6.3 (50 (62.5%)) | 4.1 ± 3.9 (401 (87.0%)) | 0.471 | Data presented as mean ± SD (n of embryos (% of n/N) KIDp, known implantation data positive; t(N), cleavage of N discrete cells, ECC, embryonic cell cycle; S2, the synchronous two blastomere divisions within the second cell cycle; S3, the synchronous two blastomere divisions within the third cell cycle An analysis of implanted (KIDp) embryos from each group (NOR and DOR) demonstrated similar developmental rate in most of the early morphokinetic parameters, except for t8, which was earlier in the NOR group. Additional parameters assessing embryonal cell cycle (ECC3), synchronous divisions (S2 and S3), and timing of compaction (tM) similarly occurred earlier in the NOR embryos (Table 3). Furthermore, the rate of arresting embryos at each developmental stage was consistently higher in the DOR group compared with the NOR group for all comparisons (Tables 2 and 3). The subgroup of KID-positive and KID-negative embryos obtained from women with DOR was also analyzed (Supplementary Table 1), revealing that DOR- KIDp embryos reached the developmental milestone of tSB faster than the DOR-KIDn embryos, with a significant average difference of nearly 5.5 h in favor of the KIDp embryos (79.4±7.7 vs. 73.9±7.6 h, respectively; p = 0.024). Timing of S3 was also earlier in the KIDp embryos (5.7±5.1 vs. 8.8±7.0 h, respectively; p = 0.005). Figure 1 presents the comparison of tSB between the KIDp and KIDn embryos in the DOR group. Spearman correlation was performed separately in the DOR and NOR groups for transferred embryos and KIDp embryos, and showed a correlation of morphokinetic events, mainly with the first or second consecutive event, for both groups. Each morphokinetic event, except for tPNf (which correlated significantly with 3 parameters—t2, t3, and t4), was correlated significantly (R > 0.7, p <0.05) with the events from the same cell cycle or early consecutive cycle, but not with later morphokinetic events (Supplementary Tables 2-5). Two mixed multivariable logistic regression models, with the addition of random effect of patient, were evaluated for KIDp embryos in the entire study population (including the parameter of decreased ovarian response) and in the DOR population separately. These analyses included both morphokinetic parameters and clinical parameters (medical center, patient’s age, nulliparity, number of retrieved oocytes, cleavage stage or blastocyst stage, and number of transferred embryos). Four parameters were found to be independently associated with KIDp embryos in the entire cohort: nulliparity, DOR, t4, and number of transferred embryos (Table 4). In the regression model for KIDp embryos in the DOR population alone, a smaller number of transferred embryos and slightly later t5 were associated with pregnancy. Table 4. | OR | Lower 95% CI | Upper 95% CI | p-value | | |---|---|---|---|---| | Entire study population | |||| | Patient’s age | 0.97 | 0.94 | 1.01 | 0.094 | | Decreased ovarian response | 0.57 | 0.41 | 0.78 | < 0.001 | | Nulliparity | 0.71 | 0.54 | 0.93 | 0.014 | | N of transferred embryos | 0.26 | 0.20 | 0.33 | < 0.001 | | t4 | 0.90 | 0.85 | 0.96 | 0.001 | t4, cleavage of 4 discrete cells

Our results indicate decreased oocyte quality manifested by multi-staged delayed preimplantation development in the DOR group compared with normal responders. This delay in morphokinetic milestone events was demonstrated not only in the entire embryo population (including those who were discarded or frozen), but also in those which were considered suitable for transfer. We further demonstrated that embryos which failed to implant in the DOR group reached starting blastocyst formation significantly later than implanted embryos suggesting, as was previously published by Meseguar et al. [9], a potential utility of temporal embryonic developmental milestone events as predictive tools for implantation, before transfer in DOR patients. In order to assess whether implanted embryos from the DOR group have equal quality as implanted embryos from NOR patients, we further compared implanted embryos from each group. Like the aforementioned findings, this comparison revealed a protracted developmental rate in most of the morphokinetic milestones in the implanted DOR embryos and might imply that even the best embryos originating from DOR patients do not exhibit a developmental “catching up” to embryos from patients with normal ovarian response. Due to the continuous nature and relation of morphokinetic events to one another during embryonic development and the possibility that a major difference in a single early development stage affects consecutive parameters differences, the correlation between each morphokinetic parameter and the rest of the morphokinetic parameters was tested using the Spearman's correlation test. This analysis showed minimal effect of each event on later events, with a significant correlation mainly with the same cell cycle or early following cell cycle parameters, but not later events, for both the normal response and DOR groups. This result in a large cohort of embryos supports the hypothesis [14] that early developmental morphokinetic events should be evaluated separately from later events (e.g. tM, tSB). As we excluded all the treatments with male factor from this study, our observation of lower embryo quality in the DOR group, represented by slower developmental and implantation rates, should be in part the result of lower oocyte quality in young poor responders. This finding is supported by several hypotheses regarding the biological basis for the correlation between oocyte quantity and quality. One theory suggested that the change in the hormonal environment caused by a decreased ovarian pool may have harmful effects on oocyte genetic content [15, 16] and eventually on oocyte quality. Another theory postulated that oocyte selection might be impaired when the pool of available oocytes is decreased [17]. Most importantly, this insult to oocyte quality might have further consequences, manifested by decreased pregnancy rates, increased aneuploid pregnancies, and higher miscarriage rates [3, 15, 18, 19]. Despite the correlation between oocyte quantity and quality observed here, it is worth mentioning that while the clinical pregnancy rate was lower in the DOR group compared with the NOR group, it was not as far reduced as that seen in older women with the same low mean number of oocytes. . This observation suggests that even with the decrease in oocyte quality and quantity, age still remains one of the prominent factors associated with oocyte quality. By excluding women older the 38 years we intended to reduce this impact in order to better assess the contribution of the ovarian response to oocyte quality. This exclusion criteria might explain why age was not found to be predictive for implantation in the multivariate regression model performed. The discrepancy regarding oocyte quality in the DOR population between our results and the results published by Morin [8] and others [6, 20], which evaluated quality by embryo morphology, might be related to the method by which embryo quality was assessed. Our study, together with the case series published by Horjnik [12], are the only reports assessing embryo quality in woman with DOR with time-lapse technology. Time-lapse analysis enables monitoring of embryo development, hence facilitates a more detailed assessment of embryo viability compared with standard morphology, based on static assessment performance [21]. Therefore, this technology offers additive information regarding embryo features and quality. Although this study provides the largest analysis of embryonic morphokinetic behavior of oocytes obtained from women diagnosed with DOR, there are limitations derived from its retrospective nature. Though probably not clinically significant, patients in the DOR group were on average 1.3 years older compared with the patients in the normal ovarian response group (33.2 ± 3.5 vs. 31.9 ± 3.8). Since matching according to age might have reduced the sample size and hamper our ability to complete the statistical analysis, we further assessed this parameter in our multivariate logistic regression model which demonstrated no correlation to implantation. Nevertheless, age and other parameters such as number of previous IVF treatments, which differed between the groups, may act as confounders regarding the morphokinetic analyses. Another difference that might affect treatment outcomes is the higher rate of blastocyst culture and blastocyst transfer in women exhibiting normal response as compared with DOR patients. However, implantation and pregnancy rates in blastocyst transfers were similarly higher in the NOR vs. the DOR group. Limited by its retrospective design, we were not able to extract essential data regarding live birth rate, therefore chose implantation and pregnancy rates as our primary outcomes. Additionally, though accounted for in the logistic regression analyses, the study groups differed in some of their basic and IVF cycle-related characteristics, including woman’s age. At last, although this study focused on the DOR population, important information essential for applying and defining this group of patients was lacking, including ovarian reserve assessment, the maximal dose of gonadotropin obtained, and the overall dose of hormonal stimulation and previous IVF response. In summary, our data provide an insight into one of the largest DOR cohorts and demonstrates that the quantitative decline in ovarian reserve is combined with reduced oocyte quality and is reflected in slower developmental rate and lower implantation and pregnancy rates. These findings highlight another pathophysiology for decreased oocyte quality in DOR patients not related to age and might encourage further research concerning this unique population. Supplementary information Author contribution Natali Schachter-Safrai and Yoav Kan-Tor have contributed substantially to the conception and design of the study, analysis and interpretation of data, and drafting of the article. Gilad Karavani, Yuval Or, and Yoel Shufaro have contributed substantially to the acquisition, analysis and interpretation of data, and revision of the article. Iris Har-Vardi has contributed substantially to the conception and design of the study, analysis and interpretation of data, and drafting and revision of the article. Amnon Buxboim and Assaf Ben-Meir havve contributed substantially to the conception and design of the study, acquisition, analysis and interpretation of data, and drafting and revision of the article. All authors have approved the final version of the study Funding This study was financially supported by the Israel Innovation Authority – Kamin grant 55326. Data availability Available upon request. Declarations Conflict of interest The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Natali Schachter-Safrai and Yoav Kan-Tor are equal contribution first authors Amnon Buxboim and Assaf Ben-Meir are equal contribution last authors

- 1.Practice Committee of the American Society for Reproductive Medicine Testing and interpreting measures of ovarian reserve: a committee opinion. Fertil Steril. 2012;98:1407–1415. doi: 10.1016/j.fertnstert.2012.09.036. [DOI] [PubMed] [Google Scholar] - 2.Sunkara SK, Rittenberg V, Raine-Fenning N, Bhattacharya S, Zamora J, Coomarasamy A. Association between the number of eggs and live birth in IVF treatment: an analysis of 400 135 treatment cycles. Hum Reprod. 2011;26:1768–1774. doi: 10.1093/humrep/der106. [DOI] [PubMed] [Google Scholar] - 3.Haadsma ML, Groen H, Mooij TM, Burger CW, Broekmans FJ, Lambalk CB, et al. Miscarriage risk for IVF pregnancies in poor responders to ovarian hyperstimulation. Reprod BioMed Online. 2010;20:191–200. doi: 10.1016/j.rbmo.2009.11.005. [DOI] [PubMed] [Google Scholar] - 4.Tarasconi B, Tadros T, Ayoubi JM, Belloc S, de Ziegler D, Fanchin R. Serum antimüllerian hormone levels are independently related to miscarriage rates after in vitro fertilization-embryo transfer. Fertil Steril. 2017;108:518–524. doi: 10.1016/j.fertnstert.2017.07.001. [DOI] [PubMed] [Google Scholar] - 5.Ebner T, Sommergruber M, Moser M, Shebl O, Schreier-Lechner E, Tews G. Basal level of anti-Müllerian hormone is associated with oocyte quality in stimulated cycles. Hum Reprod. 2006;21:2022–2026. doi: 10.1093/humrep/del127. [DOI] [PubMed] [Google Scholar] - 6.Smeenk JM, Sweep FC, Zielhuis GA, Kremer JA, Thomas CM, Braat DD. Antimüllerian hormone predicts ovarian responsiveness, but not embryo quality or pregnancy, after in vitro fertilization or intracyoplasmic sperm injection. Fertil Steril. 2007;87:223–226. doi: 10.1016/j.fertnstert.2006.06.019. [DOI] [PubMed] [Google Scholar] - 7.te Velde ER, Pearson PL. The variability of female reproductive ageing. Hum Reprod Update. 2002;8:141–154. doi: 10.1093/humupd/8.2.141. [DOI] [PubMed] [Google Scholar] - 8.Morin SJ, Patounakis G, Juneau CR, Neal SA, Scott RT, Seli E. Diminished ovarian reserve and poor response to stimulation in patients <38 years old: a quantitative but not qualitative reduction in performance. Hum Reprod. 2018;33:1489–1498. doi: 10.1093/humrep/dey238. [DOI] [PubMed] [Google Scholar] - 9.Meseguer M, Herrero J, Tejera A, Hilligsøe KM, Ramsing NB, Remohí J. The use of morphokinetics as a predictor of embryo implantation. Hum Reprod. 2011;26:2658–2671. doi: 10.1093/humrep/der256. [DOI] [PubMed] [Google Scholar] - 10.Motato Y, de los Santos MJ, Escriba MJ, Ruiz BA, Remohí J, Meseguer M. Morphokinetic analysis and embryonic prediction for blastocyst formation through an integrated time-lapse system. Fertil Steril. 2016;105:376–84.e9. doi: 10.1016/j.fertnstert.2015.11.001. [DOI] [PubMed] [Google Scholar] - 11.Kirkegaard K, Kesmodel US, Hindkjær JJ, Ingerslev HJ. Time-lapse parameters as predictors of blastocyst development and pregnancy outcome in embryos from good prognosis patients: a prospective cohort study. Hum Reprod. 2013;28:2643–2651. doi: 10.1093/humrep/det300. [DOI] [PubMed] [Google Scholar] - 12.Hojnik N, Vlaisavljević V, Kovačič B. Morphokinetic Characteristics and Developmental Potential of In Vitro Cultured Embryos from Natural Cycles in Patients with Poor Ovarian Response. Biomed Res Int. 2016;2016:4286528. doi: 10.1155/2016/4286528. [DOI] [PMC free article] [PubMed] [Google Scholar] - 13.Ciray HN, Campbell A, Agerholm IE, Aguilar J, Chamayou S, Esbert M, Sayed S, Time-Lapse User Group Time-Lapse User Group. Proposed guidelines on the nomenclature and annotation of dynamic human embryo monitoring by a time-lapse user group. Hum Reprod. 2014;29:2650–2660. doi: 10.1093/humrep/deu278. [DOI] [PubMed] [Google Scholar] - 14.Bodri D, Sugimoto T, Serna JY, Kondo M, Kato R, Kawachiya S, et al. Influence of different oocyte insemination techniques on early and late morphokinetic parameters: retrospective analysis of 500 time-lapse monitored blastocysts. Fertil Steril. 2015;104:1175–81.e1-2. doi: 10.1016/j.fertnstert.2015.07.1164. [DOI] [PubMed] [Google Scholar] - 15.Nasseri A, Mukherjee T, Grifo JA, Noyes N, Krey L, Copperman AB. Elevated day 3 serum follicle stimulating hormone and/or estradiol may predict fetal aneuploidy. Fertil Steril. 1999;71:715–718. doi: 10.1016/s0015-0282(98)00525-1. [DOI] [PubMed] [Google Scholar] - 16.van Montfrans JM, Dorland M, Oosterhuis GJ, van Vugt JM, Rekers-Mombarg LT, Lambalk CB. Increased concentrations of follicle-stimulating hormone in mothers of children with Down's syndrome. Lancet. 1999;353:1853–1854. doi: 10.1016/s0140-6736(99)00936-8. [DOI] [PubMed] [Google Scholar] - 17.Warburton D. The effect of maternal age on the frequency of trisomy: change in meiosis or in utero selection? Prog Clin Biol Res. 1989;311:165–181. [PubMed] [Google Scholar] - 18.Saldeen P, Källen K, Sundström P. The probability of successful IVF outcome after poor ovarian response. Acta Obstet Gynecol Scand. 2007;86:457–461. doi: 10.1080/00016340701194948. [DOI] [PubMed] [Google Scholar] - 19.Hendriks DJ, te Velde ER, Looman CW, Bancsi LF, Broekmans FJ. Expected poor ovarian response in predicting cumulative pregnancy rates: a powerful tool. Reprod BioMed Online. 2008;17:727–736. doi: 10.1016/s1472-6483(10)60323-9. [DOI] [PubMed] [Google Scholar] - 20.Thum MY, Abdalla HI, Taylor D. Relationship between women's age and basal follicle-stimulating hormone levels with aneuploidy risk in in vitro fertilization treatment. Fertil Steril. 2008;90:315–321. doi: 10.1016/j.fertnstert.2007.06.063. [DOI] [PubMed] [Google Scholar] - 21.Fishel S, Campbell A, Foad F, Davies L, Best L, Davis N, Smith R, Duffy S, Wheat S, Montgomery S, Wachter A, Beccles A. Evolution of embryo selection for IVF from subjective morphology assessment to objective time-lapse algorithms improves chance of live birth. Reprod BioMed Online. 2020;40:61–70. doi: 10.1016/j.rbmo.2019.10.005. [DOI] [PubMed] [Google Scholar] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Data Availability Statement Available upon request.