Development of a dynamic machine learning algorithm to predict clinical pregnancy and live birth rate with embryo morphokinetics.

Objective: To evaluate the feasibility of generating a center-specific embryo morphokinetic algorithm by time-lapse microscopy to predict clinical pregnancy rates. Design: A retrospective cohort analysis. Setting: Academic fertility clinic in a tertiary hospital setting. Patients: Patients who underwent in vitro fertilization with embryos that underwent EmbryoScope time-lapse microscopy and subsequent transfer between 2014 and 2018. Interventions: None. Main Outcome Measures: Clinical pregnancy. Results: A supervised, random forest learning algorithm from 367 embryos successfully predicted clinical pregnancy from a training set with overall 65% sensitivity and 74% positive predictive value, with an area under the curve of 0.7 for the test set. Similar results were achieved for live birth outcomes. For the secondary analysis, embryo growth morphokinetics were grouped into five clusters using unsupervised clustering. The clusters that had the fastest morphokinetics (time to blastocyst = 97 hours) had pregnancy rates of 54%, whereas a cluster that had the slowest morphokinetics (time to blastocyst = 122 hours) had a pregnancy rate of 71%, although the differences were not statistically significant (P=.356). Other clusters had pregnancy rates of 51%-60%. Conclusions: This study shows the feasibility of a clinic-specific, noninvasive embryo morphokinetic simple machine learning model to predict clinical pregnancy rates.

The importance of estradiol measurement in patients undergoing in vitro fertilization.

Successful outcomes of in vitro fertilization (IVF) are dependent in part on successful oocyte maturation and retrieval during a controlled ovarian stimulation process, which is guided by serial ultrasound and estradiol measurements. Yet, laboratory analysis of estradiol poses challenges due to the need for accuracy and specificity across concentrations that span multiple orders of magnitude. The Endocrine Society released a 2013 position statement that called for improvements in methods to analyze estradiol, and while some progress has been made in standardization and assay specificity, further work is needed to meet the needs of patients in both the IVF setting and in other clinical contexts. This review highlights the capabilities and challenges of current laboratory methods for the analysis of estradiol in the IVF setting, including automated immunoassays and liquid chromatography-tandem mass spectrometry, and discusses current efforts to improve the analytical sensitivity and standardization of estradiol assays. Clinical laboratorians should be aware of the limitations of current estradiol assays and select appropriate methods for the measurement of estradiol in their patient population.

Validation of the picoAMH assay on the Dynex DS2 platform.

OBJECTIVES: Anti-Müllerian hormone (AMH) is increasingly used as a biomarker of ovarian reserve in clinical practice, and is used both for management of fertility treatments and prediction of menopause. We sought to validate the newly FDA-approved Ansh Laboratories MenoCheck picoAMH ELISA on the Dynex-DS2 platform for clinical use in our obstetrics and gynecology center. DESIGN: Validation of the picoAMH ELISA on the Dynex-DS2 was performed according to CLSI guidelines. Intra- and inter-assay CV, assay linearity, and method comparison studies were carried out to verify assay precision and accuracy. The manufacturer's reference range was verified using 26 volunteer samples, and interference for hemolysis, lipemia, icterus, and biotin was evaluated. picoAMH results were additionally correlated with antral follicle count by ultrasound. RESULTS: Intra- and inter-assay CV of the picoAMH assay on the DS2 was ≤4% and the assay was linear between concentrations of 0.0067-16.24 ng/mL (0.048-116.0 pmol/L) AMH. Method comparison was performed with the manufacturer's laboratory and indicated good correlation, with Deming regression yielding slope of 0.928 and intercept of -0.0421. The assay displayed no significant interference from hemolysis (1000 mg/dL), lipemia (2000 mg/dL), conjugated bilirubin (66 mg/dL), or biotin (10,000 ng/mL). Measurement of AMH on the DS2 was also correlated with antral follicle count, with R = 0.7128. CONCLUSIONS: Our results indicate that the picoAMH ELISA on the DS2 has good analytical performance suitable for clinical use.

A Novel Ultrasound-Guided Technique for Hysteroscopic Adhesiolysis in High-Risk Patients.

Hysteroscopy is a common gynecologic surgical procedure. Certain diagnoses, notably intrauterine adhesions and cervical stenosis, make hysteroscopy more complicated because of an increased likelihood of complications. Three patients, 1 with cervical stenosis and 2 with Asherman syndrome, underwent ultrasound (US)-guided adhesiolysis. Access to the uterine cavity was obtained by either direct balloon-aided dilation or the US-guided Seldinger technique, followed by balloon-aided dilation to enter the endometrial cavity and disrupt intrauterine/intracervical adhesions. In this case series, we describe a novel approach of using US-guided balloon dilation to safely and effectively treat intrauterine adhesions and to decrease the risk of perforation.

Accuracy of Two Progesterone Immunoassays for Monitoring In Vitro Fertilization.

BACKGROUND: Progesterone concentrations are routinely monitored during in vitro fertilization cycles. Immunoassay-based platforms are used most often in this setting because they are simple to use and amenable to same-day sample collection and result-reporting. However, immunoassay methods are subject to variation in specificity between different assay manufacturers. In this study, a set of unexpectedly high progesterone concentrations led to a method comparison between two in-house immunoassay platforms relative to the reference method. METHODS: Progesterone was measured in 28 serum samples from women undergoing IVF cycles using the Siemens ADVIA Centaur Immunoassay system and the Abbott Architect i1000SR analyzer. A subset of these samples was selected for progesterone measurement by liquid chromatography-tandem mass spectrometry to define the accuracy of each immunoassay. RESULTS: The Siemens ADVIA Centaur immunoassay system overestimated progesterone concentrations by 19% and the Abbott Architect overestimated progesterone concentrations by 5%. CONCLUSIONS: The Abbott Architect progesterone immunoassay provides a more accurate measurement of serum progesterone than the Centaur immunoassay at concentrations relevant for monitoring in vitro fertilization populations.

Hormonal and Clinical Predictors for Post-egg Retrieval Pain in Women Undergoing Assisted Reproductive Technology Procedures.

OBJECTIVES: The intensity of post-egg retrieval pain is underestimated, with few studies examining postprocedural pain and predictors to identify women at risk for severe pain. We evaluated the influence of preprocedural hormonal levels, ovarian factors, and mechanical temporal summation (mTS) as predictors for post-egg retrieval pain in women undergoing in vitro fertilization. METHODS: Eighteen women scheduled for ultrasound-guided egg retrieval under standardized anesthesia and postprocedural analgesia were enrolled. Preprocedural mTS, questionnaires, clinical data related to anesthesia and the procedure itself, postprocedural pain scores, and pain medication for breakthrough pain were recorded. Statistical analysis included Pearson product-moment correlations, Mann-Whitney U tests, and multiple linear regressions. RESULTS: Average peak post-egg retrieval pain during the first 24 hours was 5.0±1.6 on a numerical response scale (0=no pain, 10=worst pain imaginable). Peak post-egg retrieval pain was correlated with basal antimullerian hormone (AMH) (r=0.549, P=0.018), preprocedural peak estradiol (r=0.582, P=0.011), total number of follicles (r=0.517, P=0.028), and number of retrieved eggs (r=0.510, P=0.031). Ovarian hyperstimulation syndrome (n=4) was associated with higher basal AMH (P=0.004), higher peak pain scores (P=0.049), but not with peak estradiol (P=0.13). The mTS did not correlate with peak postprocedural pain (r=0.266, P=0.286), or peak estradiol level (r=0.090, P=0.899). DISCUSSION: Peak post-egg retrieval pain intensity was higher than anticipated. Our results suggest that post-egg retrieval pain can be predicted by baseline AMH, high peak estradiol, and ovarian hyperstimulation syndrome. Further studies to evaluate intraprocedural and postprocedural pain in this population are needed, as well as clinical trials to assess postprocedural analgesia in women presenting with high hormonal levels.

Whole genome prediction for preimplantation genetic diagnosis.

BACKGROUND: Preimplantation genetic diagnosis (PGD) enables profiling of embryos for genetic disorders prior to implantation. The majority of PGD testing is restricted in the scope of variants assayed or by the availability of extended family members. While recent advances in single cell sequencing show promise, they remain limited by bias in DNA amplification and the rapid turnaround time (<36 h) required for fresh embryo transfer. Here, we describe and validate a method for inferring the inherited whole genome sequence of an embryo for preimplantation genetic diagnosis (PGD). METHODS: We combine haplotype-resolved, parental genome sequencing with rapid embryo genotyping to predict the whole genome sequence of a day-5 human embryo in a couple at risk of transmitting alpha-thalassemia. RESULTS: Inheritance was predicted at approximately 3 million paternally and/or maternally heterozygous sites with greater than 99% accuracy. Furthermore, we successfully phase and predict the transmission of an HBA1/HBA2 deletion from each parent. CONCLUSIONS: Our results suggest that preimplantation whole genome prediction may facilitate the comprehensive diagnosis of diseases with a known genetic basis in embryos.

A meta-analysis of the route of administration of luteal phase support in assisted reproductive technology: vaginal versus intramuscular progesterone.

OBJECTIVE: To perform an analysis of data with consideration for the current clinically accepted vaginal progesterone (P) or intramuscular (IM) P dosing regimens and the clinically relevant randomized clinical trials published during the time frame 1992 to 2008. DESIGN: Meta-analysis of progesterone luteal support in IVF cycles using odds ratios (OR) and 95% confidence intervals (CI). SETTING: Previously conducted randomized clinical trials meeting acceptance criteria. PATIENT(S): Infertility patients. INTERVENTION(S): Progesterone (50 mg) IM daily or 200 mg P-in-oil capsules three times a day vaginally or 90 mg P in bioadhesive gel daily vaginally. MAIN OUTCOME MEASURE(S): Clinical pregnancy, ongoing pregnancy, miscarriage. RESULT(S): This analysis showed a comparable effect between vaginal progesterone as an oil-in-capsule or as a bioadhesive gel and IM P administration on the endpoints of clinical pregnancy (OR = 0.91, 95% [CI 0.74, 1.13]) and ongoing pregnancy (OR = 0.94, 95% [CI 0.71, 1.26]). A nominally significantly lower rate of miscarriage was observed with vaginal P compared with IM P (OR = 0.54, 95% [CI 0.29, 1.02]). CONCLUSION(S): Administration of vaginal P is comparable to administration of IM P for luteal phase support in assisted reproductive technology.