Patients Who Failed to Conceive Following an In Vitro Fertilization Cycle Can Be Clustered Into Different Failure Causes Using Gene Expression Hierarchical Analysis

The patient's response to the IVF stimulation protocol is highly variable and thus difficult to predict. When a cycle fails, there are often no apparent or obvious reasons to explain the failure. Having clues on what went wrong during stimulation could serve as a basis to improve and personalize the next stimulation protocol. This exploratory study aimed to investigate if it is possible to distinguish different failure causes or different follicular responses in a population of non-pregnant IVF patients. Using qRT-PCR, we analyzed a panel of genes indicative of different failure causes in patients who did not achieve pregnancy following an IVF cycle. Follicular cells samples from 135 women were analyzed. For each patient, a pool of follicular cells from all aspirated follicles was used as a sample which gives a global picture of the patient's ovary and not a specific picture of each follicle. We performed hierarchical clustering analysis to split the non-pregnant patients according to the gene expression pattern. The patient parameters and the gene expression levels were then compared between the groups resulting from the clustering. Hierarchical analysis showed that the population of non-pregnant IVF patients could be divided into three clusters. Gene expression was significantly different, and each cluster displayed a particular gene expression pattern. Follicular cells from patients in cluster 1 displayed a pattern of gene expression related to large incompetent follicles with a higher level of apoptosis (over matured). Gene expression in follicular cells from patients from cluster 2 was related to follicles not ready to ovulate (under mature); while follicular cells from patients in cluster 3 were characterized by an excess of inflammation with no visible symptoms. This study reinforces the idea that women often have different response to the same protocol and would benefit from more personalized treatments.