Where and when should natural killer cells be tested in women with repeated implantation failure?

• Natural killer (NK) cells are increased in idiopathic repeated implantation failure (RIF).
• This is found both in blood and in endometrium.
• The association between NK cells and implantation failure is stronger in endometrium.
• There is a positive correlation between blood and endometrial levels.
• In women with other types of RIF no differences are found.

The aim of this study was to identify the candidates for natural killer (NK) testing and to define the best methodology. For this purpose a prospective study was performed on 73 women with repeated implantation failure (RIF). RIF was considered to exist in patients not achieving clinical pregnancy after three transfers with at least one good-quality embryo. Idiopathic RIF was considered to exist in patients in whom thrombophilia, hysteroscopy and endometrial culture were normal, and no chromosomal factor was suspected. Thirty-two of the 73 patients were considered to have idiopathic RIF, and 17 fertile women with children were taken as controls. Immunohistochemical staining for endometrial CD56+ and blood CD56+ or CD16+ NK cells measured using flow cytometry were compared during the mid-luteal phase in both patients and controls. Seventeen out of the 32 patients with idiopathic RIF and only one of the controls had >250 CD56 cells per high power field 400× in endometrial biopsy (p < 0.001). The percentage of blood NK cells out of the total lymphocyte population was higher in women with idiopathic RIF (13.4 ± 1.2%; range, 2.63–29.01) than in controls (8.4 ± 0.7%; range, 5.72–13.28; p = 0.026). There was a positive correlation between blood and endometrial CD56 cells (ρ = 0.707; p < 0.001). No significant differences were found between patients with other types of RIF and controls. This study suggested that testing for NK cells might be useful in women with idiopathic RIF during the mid-luteal phase.