Clinical outcomes following selection of human preimplantation embryos with time-lapse monitoring: a systematic review

BACKGROUND: Time-lapse monitoring (TLM) has emerged as a novel technology to perform semi-quantitative evaluation of embryo morphology and developmental kinetics in assisted reproduction. While this method has already been introduced into clinical practice in many laboratories, it is unclear whether it adds value to conventional morphology. Most studies only report blastocyst formation as the primary end-point. The aim of this systematic review is to provide a critical evaluation of the available studies that report clinical outcomes following embryo selection with TLM. METHODS: A literature search in MEDLINE, Cochrane CENTRAL and ISI Web of Knowledge Science Citation Index was performed to identify studies that assess the clinical utility of kinetic markers for non-invasive selection of human embryos with high implantation potential. Only studies published before 31 December 2013 in the English language that report rates of implantation, clinical pregnancy or live birth were included. RESULTS: Two hundred and fifty-one studies were identified by database search and reference list review; only 13 met eligibility criteria and were included in this analysis. The following morphokinetic parameters were assessed: pronuclear dynamics and morphology (n = 3), duration of first cytokinesis and reappearance of nuclei after cleavage (n = 3), time to various cleavage stages (n = 5), duration of various cleavage stages (n = 6), duration of cleavage cycles and mitotic synchronicity (n = 6), and time to morula, blastocyst and hatching (n = 4). Five studies used combined parameter grading to generate a cumulative score, and two studies retrospectively compared implantation rates following embryo selection by conventional morphology alone or with the addition of a hierarchal time-lapse classification. While several studies suggest higher implantation rates for fast-cleaving embryos and those with a timely duration (i.e. all time points within the defined ranges) of the 2-cell and 3-cell stages, no single morphokinetic parameter has been consistently shown to predict implantation potential. Furthermore, there was considerable disagreement regarding not only which parameters are useful, but also what constitutes normal and abnormal intervals for these measurements. CONCLUSIONS: While TLM has the potential to revolutionize clinical embryology, there are currently no high-quality data to support the clinical use of this technology for selection of human preimplantation embryos. Our recommendations for the adoption of this technique are thus limited by the available literature and the lack of robust prospective studies reporting clinical outcomes. Sparse, often incomplete and largely heterogeneous data suggest that TLM may be able to distinguish between high and low-implantation potential embryos. Only one study demonstrated significantly improved clinical pregnancy rates when embryos were selected by TLM in addition to conventional morphology. Prospective studies are currently underway and hopefully will clarify the role of TLM. As more data become available, it is of the utmost importance that groups using TLM share a common nomenclature for measured time points; herein, we have proposed a standardized system for describing any milestone along the preimplantation developmental timeline. Furthermore, future studies must publish complete datasets in an effort to define patient-specific algorithms with the clinically meaningful end-point of implantation, prior to routine adoption in the assisted reproduction technology laboratory. Until such evidence accumulates, selection of embryos by TLM should remain an experimental strategy subject to institutional review and approval.