Journal of Assisted Reproduction and Genetics最新文献

Purpose: Long non-coding RNAs (lncRNAs) are key regulators of transcriptional, epigenetic, and post-transcriptional processes involved in implantation, trophoblast function, and early placental development. Genetic polymorphisms within lncRNA loci may alter their expression or regulatory activity, potentially contributing to recurrent pregnancy loss (RPL). We aimed to evaluate the association between lncRNA gene polymorphisms and susceptibility to RPL.

Methods: A systematic review and meta-analysis were conducted according to PRISMA and MOOSE guidelines and prospectively registered in PROSPERO (CRD420261298804). A systematic search was performed in PubMed, Embase, Scopus, and Web of Science from database inception to January 2026, without language or date restrictions. Two reviewers independently screened studies, extracted data, and assessed risk of bias using ROBINS-E. Meta-analyses were performed using fixed- or random-effects models according to heterogeneity. Hardy-Weinberg equilibrium testing, sensitivity analyses, subgroup analyses by ancestry, and certainty of evidence assessment using the GRADE framework were conducted.

Results: Sixteen case-control studies involving 11,088 participants were included, comprising 5169 women with RPL and 6454 controls. Eight polymorphisms across four lncRNA genes were quantitatively synthesized. Significant associations with increased RPL susceptibility were observed for HOTAIR rs4759314 (AG vs AA: OR, 2.75 [95% CI, 2.04-3.71]; GG vs AA: OR, 2.16 [95% CI, 1.09-4.26]; G vs A: OR, 2.06 [95% CI, 1.48-2.88]) and HOTAIR rs920778 (CC vs TT: OR, 2.00 [95% CI, 1.28-3.11]; TC + CC vs TT: OR, 1.39 [95% CI, 1.09-1.76]; C vs T: OR, 1.38 [95% CI, 1.14-1.66]). A protective association was consistently identified for HOTTIP rs1859168 (CA vs AA: OR, 0.63 [95% CI, 0.45-0.89]; CC vs AA: OR, 0.33 [95% CI, 0.17-0.67]; C vs A: OR, 0.61 [95% CI, 0.45-0.82]). Sensitivity analysis further strengthened the evidence for HOTAIR rs1899663, with significant associations emerging for TT vs GG (OR, 2.59 [95% CI, 1.46-4.61]) and T vs G (OR, 1.41 [95% CI, 1.15-1.72]). Overall certainty of evidence for all pooled outcomes was rated as very low.

Conclusion: This meta-analysis provides the first quantitative evidence that inherited variation within lncRNA loci may contribute to susceptibility to RPL. Variants in HOTAIR, particularly rs4759314 and rs920778, were associated with increased RPL risk, whereas HOTTIP rs1859168 showed a protective association. However, the certainty of evidence remains very low, and these findings should be interpreted cautiously until confirmed by large, well-designed, multi-ancestry studies with functional validation.

Problem: Time-lapse imaging (TLI) enables longitudinal and non-invasive embryo monitoring in IVF, yet morphokinetic annotation remains labor-intensive and subject to inter-operator variability. Automated detection of developmental phase transitions is challenging due to subtle morphological changes, heterogeneous acquisition conditions, discrete and potentially irregular image sampling, focal-plane variability, and the need for temporally consistent predictions.

Aim: We aim to detect morphokinetic phase transitions in embryo TLI videos using a robust, clinically oriented framework designed to support human-in-the-loop annotation rather than replace clinical decision-making or commercial embryo selection systems.

Methods: We propose a spatio-temporal Neural Ordinary Differential Equation (Neural ODE) model for continuous-time latent representation learning between discretely acquired observations, coupled with a reference-based transition scoring mechanism and an online, one-class detection strategy. The method, named Reference-Based Neural ODE Change Detector (RB-NODE), is transition-agnostic and detects when the evolving latent state deviates from a prototype representation of the current developmental phase. Experiments are conducted on the public dataset of Gomez et al., a single-center TLI benchmark comprising 704 embryo videos acquired across seven focal planes. All reported results are computed over the union of test samples generated from the seven focal-plane views, with strict embryo-level data splitting to prevent leakage across focal planes.

Results: RB-NODE achieves an AUC of 0.988, an F1@frame of 0.975, a Det@5f of 0.873, a Det@10f of 0.956, and a mean absolute localization error of 2.64 frames. Although a ResNet18+LSTM baseline obtains a marginally higher Det@5f, RB-NODE provides the best AUC, F1@frame, Det@10f, and mean temporal error among the evaluated methods. Event-group analysis shows particularly strong performance on morula/blastocyst-related transitions, with Det@5f = 0.910, Det@10f = 0.978, and a mean temporal error of 1.90 frames. RB-NODE also remains compatible with online use, processing frames at approximately 171 FPS on an NVIDIA RTX 6000 Ada Generation GPU.

Conclusion: Continuous-time spatio-temporal modeling with Neural ODEs provides a promising and reproducible approach for robust morphokinetic transition detection in embryo TLI, offering practical support for standardized annotation workflows. The proposed framework is intended as an annotation-support tool, not as an autonomous embryo selection system. External validation across multiple IVF centers, TLI platforms, acquisition intervals, annotation protocols, and focal-plane configurations remains necessary before clinical deployment.

Purpose: The clinical application of sequencing guidelines often yields variants of uncertain significance (VUS), creating profound challenges in prenatal genetic counseling. This study demonstrates the clinical utility and real-world impact of dynamic, longitudinal variant re-classification on reproductive decision-making in a family affected by an atypical SEC24D-associated skeletal phenotype.

Methods: Over a 3-year period (2023-2026), a multidisciplinary diagnostic workflow was employed, integrating trio whole-exome sequencing, protein structural modeling, prenatal ultrasound, fetal magnetic resonance imaging, post-mortem histopathology, and preimplantation genetic testing for monogenic disorders (PGT-M). Variant pathogenicity was sequentially curated according to the American College of Medical Genetics and Genomics (ACMG) framework.

Results: A homozygous SEC24D variant (c.1942G > C, p.Gly648Arg) was identified in a proband presenting with isolated craniofacial ossification defects without classic long-bone fractures. Over 3 years, the variant was dynamically re-classified from VUS to likely pathogenic (LP), back to VUS, and to LP again, driven by institutional peer-review debates regarding the conservative application of computational structural criteria versus clinical intuition. This diagnostic instability directly dictated a cascade of critical reproductive events: the termination of a second pregnancy demonstrating recurrent fetal cranial defects confirmed by amniocentesis and post-mortem examination, followed by the deployment of PGT-M within a diagnostic "gray zone." Ultimately, the successful delivery of a healthy, non-carrier infant in June 2026 provided crucial family co-segregation data, successfully resolving the classification loop and securing a definitive likely pathogenic status.

Conclusions: Diagnostic uncertainty in prenatal genomics is a persistent clinical reality rather than a transient evidence gap. Longitudinal variant reinterpretation combined with transparent shared decision-making is vital to safely navigate borderline classifications and optimize reproductive trajectories.

Purpose: To develop a stage-structured, distribution-based prediction framework for in vitro fertilization (IVF) that generates full probability distributions at each treatment stage and enables conditional updating of downstream predictions when observed outcomes become known.

Methods: We conducted an observational modeling study using de-identified UK Human Fertilisation and Embryology Authority (HFEA) registry data (2017-2018; up to 101,217 model-ready cycles after stage-specific filtering) to model egg retrieval, maturation, and fertilization. Egg retrieval was modeled using a zero-inflated negative binomial specification. Downstream transitions such as blastocyst formation, euploidy, vitrification survival, and live birth after euploid transfer were modeled using stage-specific logistic regressions calibrated to published cohorts and national registry summaries (429,507 additional observations). Predictive performance of HFEA-derived models was evaluated on held-out HFEA test sets using point-prediction accuracy, train-test gaps, prediction-interval coverage, and calibration across predicted outcome strata.

Results: The framework propagates full probability distributions across sequential IVF stages rather than point estimates. Held-out HFEA validation showed minimal train-test degradation ( $R^2$ gaps under 0.007), with modest expected-count accuracy for egg retrieval and stronger expected-count accuracy for maturity and fertilization. Egg-retrieval prediction intervals showed near-nominal coverage (50%: 50.2%; 80%: 79.2%; 95%: 94.9%), and observed mean outcomes were close to predicted means across predicted-yield/rate strata. When observed stage outcomes were entered, downstream distributions updated appropriately, reducing uncertainty and preserving cycle-specific biological parameters in both single- and multi-cycle scenarios.

Conclusion(s): A sequential, distribution-based IVF prediction model with conditional updating provides uncertainty-quantified, stage-aware predictions that dynamically adapt to patient-specific outcomes, supporting more individualized counseling and treatment planning.

Maintaining physiological extracellular pH (pHe) is essential for oocyte viability and embryonic development, particularly during ICSI, when metaphase II oocytes have their intracellular pH (pHi) regulation mechanisms temporarily inactivated. HEPES- and MOPS-buffered handling media are routinely used in IVF laboratories, with decades of clinical use supporting their effectiveness and safety. Although concerns regarding their biological effects beyond pH buffering have occasionally emerged, most are derived from experimental settings and are not representative of clinical IVF procedures, more likely influenced by experimental conditions and media composition rather than by the zwitterions themselves. Nevertheless, while evidence from human oocytes remains extremely limited and preliminary, the improved embryological outcomes and transcriptomic profiles recently reported from bicarbonate-buffered handling media have renewed interest in exploring more physiological ICSI media formulations. While current evidence remains far from conclusive and insufficient to support immediate changes in clinical practice, emerging findings suggest that routine systems may not yet be fully optimized and could represent an opportunity for further improvement. Future evaluation of bicarbonate-buffered media for ICSI should include robust standardization of pH measurement methodologies together with optimization of handling conditions, including oil and dish characteristics, that influence pHe stability.

Purpose: To assess whether specific markers of oxidative stress and DNA damage are expressed at higher levels in primordial follicles of patients with endometriomas compared to age-matched controls.

Methods: A retrospective pilot cohort study was conducted using ovarian tissue from patients who underwent unilateral or bilateral salpingo-oophorectomy for symptomatic endometriomas, with controls undergoing an identical procedure for other benign gynecologic conditions. Immunohistochemistry (IHC) was used to assess percent positivity in primordial follicles for 4-hydroxy-2-nonenal (4-HNE) as a marker of oxidative stress; 8-Oxoguanine (8-Oxo) and phosphorylated H2AX (γH2AX) for DNA damage. Chi-square tests were used for categorical variables and Student t-tests for continuous variables, with p < 0.05 considered significant.

Results: Twenty-nine patients were included (8 controls, 21 with endometriomas). Age (39.4 ± 5.1 vs. 37.1 ± 4.6 years, p = 0.269), BMI (33.6 ± 10.6 vs. 29.5 ± 7.8 kg/m², p = 0.310) were similar between groups. Racial distribution, nulliparity, hormonal therapy use, and smoking status were also comparable. The average endometrioma size was 6.6 ± 2.8 cm. All molecular markers showed higher percent positivity in the endometrioma group compared to the control group. γH2AX demonstrated a statistically significant increase (63.6% vs. 50.1%, p = 0.033). 8-Oxo (67.5% vs. 55.8%, p = 0.070) and 4-HNE (65.1% vs. 61.8%, p = 0.168) were also elevated in the endometrioma group but did not reach statistical significance.

Conclusion: Patients with endometriomas demonstrated increased follicular DNA damage, with elevated γH2AX expression, supporting the notion that endometriomas may impair follicular quality through DNA damage mechanisms.

Similar live birth rates in frozen embryo transfer (FET) trials can be over-read as evidence that natural, modified natural, and programmed endometrial preparation protocols are biologically equivalent. This interpretation is not warranted when cancellation, delayed transfer, and rescue switching are built into care. Under these conditions, randomization compares starting strategies that unfold through managed clinical pathways, not uninterrupted protocols completed exactly as assigned. Using recent FET trials as practical examples, this Clinical Opinion argues for a narrower and more constructive interpretation: similar live birth supports pathway effectiveness only for the estimand that was actually analyzed. Clinicians should ask whether the reported effect reflects assignment to an initial strategy, completion of the first assigned protocol, or a hypothetical no-switch contrast. Cancellation and rescue switching should be treated as outcomes; pathway burden should be reported alongside live birth, and safety outcomes should use denominators aligned with the stage at which harms occur. These FET-specific reporting recommendations apply existing CONSORT, CONSORT Harms, and ICH E9(R1) principles to trials in which cancellation or switching can materially affect interpretation.

Purpose: To assess the extent to which expanded carrier screening (ECS) panels capture the genetic conditions for which preimplantation genetic testing for monogenic diseases (PGT-M) is performed.

Methods: A retrospective observational study conducted at a single university-affiliated PGT unit from 2015 to 2023. Women undergoing PGT-M for autosomal recessive or X-linked genetic conditions were included. Genetic conditions were cross-matched with five ECS panels (four commercial panels and one publicly funded ECS panel) and compared with a targeted, population-specific screening. Coverage was assessed at both the woman and gene levels using paired statistical comparisons.

Results: A total of 330 cases were included; seven women underwent PGT-M for two eligible conditions and were therefore counted twice. These cases involved 110 different genes with an uneven distribution: eight genes accounted for 53.2% of cases, while 71 genes were observed in a single case. At the individual-woman level, targeted, population-specific screening showed lower coverage (47.9%) than that of all ECS panels (71.8-91.5%), with overall differences statistically significant (P < 0.001). At the gene level, coverage ranged from 20.9 to 79.1% across panels (P < 0.001). The evaluated ECS panels included 271-787 genes, with the largest panel demonstrating higher case coverage than the other commercial panels.

Conclusions: Expanded carrier screening panels capture most autosomal recessive and X-linked genetic conditions for which PGT-M is performed. These findings support the clinical utility of ECS in identifying couples at risk of having children affected by such conditions.

Objective: To systematically evaluate the diagnostic accuracy and methodological quality of machine learning (ML) prediction models for pregnancy outcomes after assisted reproductive technology (ART).

Methods: PubMed, Embase, the Cochrane Library, IEEE Xplore, MEDLINE, ClinicalTrials.gov, CNKI, Wanfang, and VIP were searched from inception to July 2026. Eligible studies developed or validated ML models to predict clinical pregnancy or live birth after ART. For studies reporting complete 2 × 2 contingency data, pooled sensitivity, specificity, diagnostic odds ratio (DOR), and summary receiver operating characteristic (SROC) curves were estimated using random-effects diagnostic meta-analysis. Risk of bias was assessed with PROBAST.

Results: Twenty studies were included in the systematic review, of which 14 contributed to the diagnostic meta-analysis. Overall risk of bias was low in 1 study (5.0%), high in 8 studies (40.0%), and unclear in 11 studies (55.0%). The pooled sensitivity was 0.737 (95% CI, 0.662-0.799) and the pooled specificity was 0.789 (95% CI, 0.709-0.851), with substantial heterogeneity (I² = 97.7% and 99.0%, respectively). The pooled DOR was 10.49 (95% CI, 6.28-17.53), and the SROC curve indicated acceptable overall discrimination. Exploratory DOR subgroup analyses showed comparable performance for clinical pregnancy and live birth. No statistically robust subgroup difference was observed by algorithm type, center type, or validation status under a random-effects framework; study design showed a significant subgroup difference, but this estimate was driven by a single prospective study.

Conclusion: ML models show moderate diagnostic accuracy for predicting ART pregnancy outcomes, but the evidence base is limited by substantial heterogeneity and frequent high or unclear risk of bias. Future studies should follow TRIPOD + AI and PROBAST-aligned standards, report calibration and clinical utility, and prioritize prospective multi-center external validation before clinical implementation.

Systematic review registration: PROSPERO, CRD420251108846.