How to implement time-lapse system in daily IVF practice

Since the birth of Louise Brown (Steptoe and Edward, 1978) the field of assisted reproduction underwent fundamental changes in fertilization techniques (Palermo et al., 1992), controlled ovarian hyperstimulation (Olivennes et al., 1998) or culture media composition (Gardner and Lane, 1998) to name but a few. But what remained constant over the years is that the developmental fate and the implantation potential of cultured embryos are estimated on the basis of their static morphological appearance, by scoring them once a day (Ebner et al., 2003).

The growing need for more detailed information on embryo development led to the refinement of the process of embryo selection. Tremendous achievements in image analysis technology made this scientific urge become reality, and, thus, paved the way for the development of a new generation of incubators providing digital video sequences of preimplantation development (Kirkegaard et al., 2012). It is fair to say that meanwhile these so-called time-lapse incubators or systems have taken over control of in vitro embryo culture and selection.

Since preimplantation embryo development is a dynamic event the introduction of time-lapse technology allows for continuous monitoring of embryo cleavage behavior through frequent multiple image acquisitions (Wright et al., 1990). A pleasant side effect of the literal 24-hour monitoring is the fact that viability of the embryos is less compromised due to more stable culture conditions (Oh et al., 2007). In addition, time-lapse systems were found to play an important role in quality assurance. It has been published that suboptimal media or disposables may have a negative impact on morphokinetics, while overall viability, e.g. blastocyst formation, is not affected (Wolff et al., 2013). Last but not least, current time-lapse systems offer a great degree of flexibility in terms of working schedule. In principle, annotation of morphokinetic features of embryos can be done from home office or whenever it is compatible with the working load of the IVF lab.

So there are definitely a lot of attractive reasons to invest in time-lapse incubators. The question is, however, how such techniques can be incorporated into routine IVF work without affecting overall lab performance? According to Harper et al. (2012) the introduction of new technologies into the domain of IVF should include preliminary work on animal models which of course is no option in human IVF laboratories. So one has to rely on smaller studies on human oocytes donated for research and preliminary prospective trials before the new technique can be considered safe and efficient (Harper et al., 2012), and as a consequence can be applied clinically.

The following paragraphs intend to illustrate how a medical device (e.g., time-lapse system) was implemented at the Kepler University in Linz, Austria, under the legal framework of the European Tissue Directives (2008) and strictly following Good Manufacturing Practice guidelines (ESHRE, 2015).

Positive reporting of time-lapse users and promising results published in peer-reviewed scientific journals encouraged us to acquire a modern time-lapse system. Our quality management representative then came up with a Design Qualification (DQ) in order to define specifications of the planned timelapse incubator. Based on software flexibility, video quality, incubation chamber type and costeffectiveness soon the decision was made to go for the MIRI^® TL (Esco Medical).

After the purchase the manufacturer installed the MIRI^® TL (with six incubations chambers) and adjusted actual and target values for temperature (chamber and heated lid) and CO₂. Esco Medical provided us with a certificate stating that the device was configured according to an approved installation checklist and that every physical aspect of the equipment (e.g., materials, dimensions, pressure ratings), software design specification (e.g., software, accessibility, processor speed) and instruments (operational parameters, accuracy, voltage, etc.) fulfilled the requirements (Installation Qualification, IQ).

Once the equipment has passed the IQ phase the operational requirements and consistency of the equipment had to be put to the test. A first series of temperature measurements confirmed that the surface of the six incubation chambers indeed had a temperature between 36.9 and 37.1°C which proved that the MIRI^® TL is a very stable incubator. In IVF culture, however, temperature within the volume of culture medium is more crucial. Therefore, we went for a second round of measurements and realized that in our culture medium the temperature was on average 0.2°C lower as compared to the surface of the chambers. After adaption of the temperature set point to 37.2 °C we had the optimal temperature in our culture dish.

The second crucial physical parameter to adjust is pH which is inversely proportional to CO₂. To begin with, we chose the same CO₂ concentration (6.5%) that we used in our routine bench-top incubator which proved to be reliable in the past. Using the external ports of the MIRI^® TL which allow for individual CO₂ measurement of all chambers we could demonstrate that the CO₂ content was 6.4% to 6.6% in the presence of 5% oxygen. This stable condition gave us exactly the medium pH we strived for (7.30-7.35) and which was well in the pH range recommended by the manufacturer of our culture medium in use.

Once we had proved stability and full functionality of the time-lapse system the whole set up had to be tested by bringing in viable cells into our time-lapse culture system (Performance Qualification, PQ). Since embryos are not allowed to be donated for research in Austria we used oocytes (giant oocytes, in vitro matures oocytes, oocytes with clusters of smooth endoplasmic reticulum) and zygotes (mono- and tripronuclear) usually considered for disposal. Soon we got some good quality blastocysts and, thus, were satisfied with the performance of the MIRI^® TL .Subsequently, the next step of our PQ was implemented which was by using sibling oocytes to compare developmental outcome in a routine (bench-top) and a time-lapse culture group. Over a period of one month we split oocytes in patients with more than six eggs followed by analysis of both rate of fertilization and blastocyst formation. According to these analyzed KPIs (ESHRE and Alpha, 2017) it turned out that both incubators performed well but the MIRI^® TL showed slightly increased rates of fertilization (82% vs. 79%) and blastulation (56% vs. 51%). This promising outcome led us to release the MIRI^® TL for clinical use in our lab.

Meanwhile we have approximately 200 live-births and never regretted our decision to work with timelapse technique. Much more, using annotation tools and deselection criteria (Ebner et al., 2017; 2017) we found that the time-to-pregnancy can be shortened (unpublished data) in our patient cohort.