Discussion:

Main findings:

The cumulative live birth rate is significantly higher with TLI incubators when compared with SC incubators. This appears to be driven by two factors. Firstly, there was an improvement in live birth rate per fresh embryo transfer in the TLI arm with no significant change in the number of embryos transferred or the multiple pregnancy rate. Secondly, there was a significant increase in the number of embryos available for cryopreservation in the TLI arm which translates into a trend towards a higher proportion of those with unsuccessful fresh embryo transfers having access to cryopreserved embryos in the TLI arm.

Strengths and limitations:

Strengths:

To our knowledge, this is the second study comparing cumulative live birth rate per oocyte retrieval between TLI and SC incubators and the first to show a significant difference. A criticism of TLI incubators in the ESHRE guideline was the paucity of cumulative live birth reporting in scientific literature which would help to assess the relative importance of the ‘undisturbed culture’ versus ‘embryo selection’. The data presented here aims to provide a better perspective on this underexplored aspect of TLI incubators. Additional strengths of the study include the single-centre nature which ensures that the laboratory conditions were uniform throughout the study period and the large sample size which provided statistical power. Furthermore, the baseline denominator was per oocyte retrieval which reduces selection bias due to failed embryo transfer.

Limitations:

A retrospective study does carry its limitations, primarily in terms of confounding factors. We attempted to adjust for this by an analysis for baseline characteristics and adjustment for factors such as age and number of oocytes retrieved but there is still a potential for unknown confounding factors that we have not adjusted for. As the findings are from a single unit, generalisability of these findings will be limited but the findings of this study may form the rational basis for approval for a well-powered randomised controlled trial. Findings from other centres are urgently needed to either confirm or refute these findings to determine whether change in laboratory practice to incorporate TLI would help in other fertility clinics.

Interpretation (in light of other evidence):

Comparison with previous literature:

The first studies comparing TLI and SC incubators were exploratory, had a small sample size and primarily focussed on the safety and comparability of TLI to SC incubators. The latest Cochrane review13 noted an absence of RCTs examining cumulative live birth rate and the ESHRE recommendations on good practice of TLI incubators 6 was able to identify only one study (previously published from our unit) on cumulative live birth rate in the available literature. In this previous study, we did note an improvement in live birth per fresh embryo transfer but failed to note a significant difference in cumulative live birth per oocyte retrieval4. However, in the period of the previous study, sequential culture media was used and so the opening of TLI incubator doors was required on day 3 for a change of the culture medium.

Biological plausibility of the findings:

The availability of morphokinetic data from TLI incubators led to the identification of ‘poor prognostic’ features which were developed into factors for exclusion. The suggested exclusion factors included unevenness at the two-cell stage, multinucleation at four-cell, direct first cleavage and irregular or reverse cleavage 3. Apart from true direct cleavage from one to three cells for which there appears to be reasonable evidence for a poor prognosis, the rest of the exclusion criteria do not appear to have a consistent prognostic effect between studies 14,15. Additionally, embryos may be capable of auto-correction, and an embryo which displays features of the exclusion criteria, but nonetheless develops into a top-quality blastocyst appears to have a better prognosis than embryos which arrest during their development 2. An alternate approach has been to not rely on exclusion criteria, but generate algorithms based on embryo morphokinetics to select the best embryo. Such algorithms have been generated through statistical analysis in combination with observation and expert opinion of embryologists and more recently using artificial intelligence for data analytics. Whilst these algorithms have shown promise in initial studies, they do not appear to be translatable across different embryology laboratories, which is a major limitation to widespread adoption 1. There is ongoing research on using artificial intelligence and deep learning to improve algorithms, with early attempts being made into introducing such algorithms into clinical practice 16,17.
The Cochrane review noted that use of algorithms did not appear to confer an additional benefit to TLI with conventional morphological assessment (OR for ongoing pregnancy rate 0.61, 95% CI 0.32 to 1.20, 1 RCT, N = 163) 13.
An increase in cumulative live birth rate with TLI with this group suggests that the undisturbed culture may contribute to this uplift. We have hypothesised in our previous paper on perinatal outcomes that the undisturbed culture in TLI may play a role in the impact of TLI systems. The choice of culture media may play a role, but successful embryo development also relies on a stable embryo culture environment.
Laboratory factors can impact on media efficacy and embryo development resulting in significantly different outcomes 18. A change in temperature can destabilise the cellular cytoskeleton (including the meiotic spindle) and might also affect embryo metabolism. Differences in embryo aneuploidy rates amongst egg donor programmes between different clinics 19 serves to illustrate the crucial factor of laboratory conditions (including maintaining a stable temperature) on the development of a competent embryo. A recent review listed several patient, clinical and laboratory factors which might affect the risk of embryo aneuploidy indicating that there is room for further research to improve embryo competence in IVF clinics20.