Marilen Benner

CHAPTER 5 116 abundance and high variation are challenging to detect through univariate analysis but could be revealed using the presented approach. As the differences in AUC of the different individual classifiers show, algorithms of multivariate approaches also present with varying power to classify a sample to its cohort. The ensemble machine learning strategy overcomes the bias associated with choosing a single machine learning algorithm. The power of combining classifiers, even in this low sample size setting, allowed to detect the discriminative value of low frequency cell types in MB that have a possible regulatory phenotype. CD8 + T cells positive for Ki67 (indicating proliferation) were identified as most classifying feature in MB samples. In healthy pregnancy, decidual CD8 + T cells are known to take longer than systemic cells to initiate proliferation upon stimulation, possibly through interaction of their coinhibitory molecules with the decidual microenvironment (36). Also involved in local regulation (19, 37-39), altered levels of Treg (overall cell population and HLA-DR + ), and CD24 Hi CD38 Hi B cells contributed to robust RPL classification, fitting with the theory that pregnancy is negatively affected when local mechanisms of tolerance fail (40, 41). This focus on immune regulatory subsets was not observed in PB, except for NKbright cells, which contribute to cohort classification in both PB and uterine samples. Of note, age contributed as risk factor in uterine, but not systemic, immunity. The observed differences in classifying leucocyte subsets, depending on the sample source, affirm that systemic and local immunity of reproduction demand independent consideration. Circulatory non-switched memory B cells contributed strongly to cohort classification of PB, which is in line with previously shown higher levels of circulatory non-switched memory B cells of RPL patients (13, 42). The presented data highlight that B cells deserve consideration in context of RPL. Memory B cells are altered in many autoimmune diseases such as systemic lupus erythematosus, systemic sclerosis, and antiphospholipid syndrome, known to be associated with poor reproductive outcome (43). B cell-related pregnancy complications might also result from incorrect activation through soluble factors that regulate B cell activity, as serum from women suffering from spontaneous abortion failed to induce normal levels of IL-10 production by B cells (44). Of note, we observed higher levels of CD27 + B cells in CMV seropositive RPL patients. It has previously been shown that CMV affects frequencies of CD27 + memory B cells in women, but not men (45). Thus, CMV status is important to include in similar approaches studying leucocytes due to its known effect of selectively expanding specific subsets (30, 33). RPL is a multifactorial condition (46). Small sample size is a limitation of the current study, as we cannot take into account subgroups of different in disease etiologies. Nevertheless, we were able to show that investigating a immunity by multivariate approach in RPL is worthwhile. The applied flow cytometry-based profiling of MB holds the possibility to detect locally involved rare cell types (47-50). Compared to the trending, high-dimensional approaches using mass spectrometry or single-cell sequencing, flow cytometry is relatively cheap to perform and thus