Variability and implications of recurrent implantation failure definitions used in the scientific literature: a...

How is recurrent implantation failure (RIF) defined in published literature and what is the prognostic agreement of these definitions with recently introduced RIF criteria by ESHRE?

RIF definitions used in current clinical studies are highly variable and only a low proportion of published studies on RIF meet the ESHRE RIF diagnostic threshold.

RIF is a key cause of ART failure and growing focus of ART research. However, RIF remains poorly and inconsistently defined in published literature, thereby making the interpretation and clinical applicability of RIF research difficult and highly problematic.

The electronic databases EMBASE (Ovid), PubMed, Cochrane Central Register Of Controlled Trials (CENTRAL), Scopus, and Web of Science were systematically searched up to 30 June 2024 using the search terms 'recurrent implantation failure' and 'repeated implantation failure' for original peer-reviewed journal articles that included RIF patients.

The following data were manually extracted from eligible full-text articles: study methodology and characteristics, ART characteristics, and the RIF definition used. Extracted RIF definitions were analysed according to predetermined specifiers. The prognostic profile of these RIF definitions was compared with the 2023 ESHRE-recommended threshold for RIF diagnosis.

The literature search identified 9853 studies, of which 748 were eligible for inclusion. Of these 748 studies, 589 studies (78.7%) provided one RIF definition, 83 studies (11.1%) used two definitions, three studies (0.4%) provided three or more definitions while 73 studies (9.8%) did not provide a definition for RIF. Of the 838 RIF definitions retrieved, there were a total of 503 unique RIF definitions. The three most common specifiers used to define RIF were embryo morphological quality (n = 491, 58.6% of RIF definitions), number of transfer events (n = 439, 52.4%), and cumulative number of embryos transferred (n = 326, 38.9%). RIF was most frequently diagnosed as 'failure of ≥3 embryo transfer events' (n = 26) and 'failure of ≥3 stimulated cycles' (n = 22). The threshold for defining RIF based on the cumulative number of embryos transferred in total was significantly higher for cleavage-stage embryos compared to blastocysts (incidence rate ratio 2.15, P < 0.001). In most cases, the RIF definitions used did not meet the ESHRE-recommended RIF diagnostic threshold of >60% cumulative predicted chance of implantation.

This systematic review excluded abstracts and case-series. Several studies provided RIF definitions with limited detail or ambiguous terminology with potential for misclassification or misinterpretation.

There remains a high degree of variability and discrepancy between RIF definitions used in current clinical studies on RIF. Furthermore, the low proportion of studies meeting the ESHRE RIF diagnostic threshold casts doubts on whether the populations in these studies were truly RIF patients. As such, published research findings should be interpreted with caution. To enable wider clinical applicability of future research on the aetiology of and therapeutic interventions for RIF, it is imperative that a standardized RIF definition is meticulously implemented.

No specific external funding was sought or obtained for this study. All authors report no conflicts of interest with regard to this study.

This trial was registered in PROSPERO (CRD42022295349).