Abstract Study question Do the culture strategies, oocyte/embryo manipulations, and operators impact blastulation-rates per cohort of metaphase-II oocytes (BR per MII-oocytes) and/or live-birth-rates per single-embryo-transfer (LBR per SET)? Summary answer Undisturbed culture improved the BR per MII-oocytes, while culture drop volumes ≥80ml in the well-of-the-well system were associated with increased LBR per untested/euploid blastocyst SET. What is known already IVF lab performance is critical to achieve success in IVF. Yet, while a consensus exists on the main key-performance-indicators (KPIs), namely BR and LBR, and their inherent benchmark values, the strategies to fulfil them are still poorly standardized. The plasticity of human embryos along with the disparities in the patient populations might mask even largely different outcomes across clinics. Therefore, clear guidelines shall outline which procedures to standardize and carefully monitor in IVF laboratories. Here we investigated different embryo culture strategies, manipulations, and operators for their effect on BR and LBR per SET, adjusting the results for all main confounders. Study design, size, duration Retrospective analysis of 3705 cycles with ≥1 MII-oocyte and 2604 SETs (January2019-March2021). BR per MII-oocytes and LBR per SET were the main outcomes. Embryo-twinning was also assessed. The putative impact of culture strategies, manipulations, and operators’ expertise (0-5yr,6-11yr or > 12yr) was determined via regression analyses adjusted for possible confounders (autologous/donated oocytes, age, primary/secondary infertility and duration, sperm factor; for SET: also endometrial preparation, age at transfer, number of consecutive transfer, untested/euploid blastocyst, quality, and day). Participants/materials, setting, methods In donation cycles, only vitrified oocytes were used. In Preimplantation-Genetic-Testing (PGT) cycles, no assisted-hatching was performed and only euploid non-mosaic vitrified-warmed blastocysts were transferred. The culture strategies assessed were undisturbed (Embryoscope, Geri and Embryoscope-plus)/disturbed (KSystems), volume and approach (drops ≤30ml with single-culture or ≥ 80ml with well-of-the-well approaches), humidified/non-humidified atmosphere and media refreshed/not-refreshed. The manipulations assessed were oocyte/blastocyst vitrification (performed/not-performed), oocyte retrieval and denudation performed with oil overlay (yes/not), laser-assisted artificial-shrinkage (performed/not-performed), trophectoderm biopsy (performed/not-performed). Main results and the role of chance The only confounders on BR per MII-oocytes (overall:39.1±26.6%) were sperm factor and oocyte age. The linear regressions showed benefits for humidified atmosphere (unstandardized B-coefficient:+2.9%, p = 0.01), manipulations with oil overlay (+3.9%, p = 0.03) and drops≥80ml with a well-of-the-well approach (+4.3%, p < 0.01). However, only the positive effect of undisturbed incubators was significant when adjusting for confounders [41.7±27.8% (N = 1440 cycles) versus 37.5±25.7% in a disturbed incubator (N = 2265 cycles); unstandardized B-coefficient:+5.6%, 95%CI +3.9% to + 7.3%, standardized beta-coefficient:-0.103,p<0.01]. The main confounders on LBR per SET (overall: N = 1044/2604, 40.1%) were oocyte age, number of consecutive transfer, blastocyst chromosomal status (untested/euploid), quality and day. The univariate logistic regressions showed a benefit for undisturbed incubators (OR:1.3, 95%CI 1.1-1.5, p < 0.01), humidified atmosphere (OR:1.4, 95%CI 1.1-1.7, p < 0.01) and media refresh (OR:1.3, 95% 1.01-1.8, p = 0.05). However, only the use of drops ≥80ml with a well-of-the-well approach rather than drops ≤30ml with a single-culture approach involved an improvement when adjusting for confounders (untested SETs: N = 71/171, 41.5% versus N = 434/1284, 33.8%; euploid SETs: N = 170/316, 53.8% versus N = 369/833, 44.3%; multivariate-OR:1.33, 95%CI 1.12-1.58, adjusted p-value<0.01). This result was confirmed in a sub-analysis across only first patients’ SETs. Lastly, no feature under investigation was associated with embryo-twinning (overall: N = 23/1243, 1.9% and N = 14/1044, 1.3% per pregnancy and delivery, respectively). Limitations, reasons for caution Retrospective single center study. Only ICSI cycles and continuous culture media were assessed. Cleavage stage SETs were excluded. All operators had at least 3yr of experience. Perinatal and gestational outcomes were not evaluated. Wider implications of the findings Real-life data adjusted for confounders may unveil fluctuations in critical KPIs mainly imputable to culture strategies. An impact seldom derives from oocyte/embryo manipulations if experienced operators adopt validated protocols. An accurate interpretation of these evidence shall lead to properly designed studies with problem-solving/progress-building purposes, and guidelines to standardize culture practices. Trial registration number Not applicable
Read full abstract