Menstrual fluid from patients with endometriosis could reveal changes in endometrial epithelial signatures that may also be detectable in whole menstrual fluid RNA.
In a pilot case-control study, researchers analyzed menstrual fluid from 5 patients with imaging- or surgery-confirmed endometriosis and 5 healthy controls. Samples collected during the heaviest day of menstruation underwent single-nucleus RNA sequencing and bulk RNA sequencing. The researchers compared cell-type composition, cell-specific gene expression, predicted cell-cell communication, and the ability of bulk RNA sequencing to reproduce findings from the single-nucleus analysis. They also evaluated candidate biomarkers in an independent external bulk RNA sequencing data set.
The researchers identified more than 23,000 high-quality nuclei representing endometrial epithelial, stromal, endothelial-like, and immune-cell populations. Although immune cells comprised most recovered cells, epithelial cells exhibited the greatest disease-associated transcriptional differences. The researchers identified 36 upregulated and 53 downregulated epithelial genes with greater than twofold expression changes, and 402 differentially expressed genes overall regardless of change. Stromal cells demonstrated fewer changes, whereas most immune-cell populations showed relatively limited differential gene expression.
Many of the epithelial genes were associated with biologic pathways previously implicated in endometriosis, including steroid hormone metabolism, extracellular matrix remodeling, WNT signaling, transforming growth factor–beta signaling, apoptosis, and mitogen-activated protein kinase signaling. The researchers also found altered predicted communication between endometrial and immune-cell populations. Compared with controls, the samples from patients with endometriosis showed reduced predicted epithelial-stromal interactions involving development and growth genes, including WNT2B, IGF2, and TGFB2, alongside increased predicted communication between dendritic cells and endometrial cells through the BMP signaling pathway.
To determine whether these epithelial signatures could be detected using a clinically practical approach, the researchers performed bulk RNA sequencing on whole menstrual fluid. Five genes—TIMP2, AKR1C2, DMBT1, FERMT1, and KCNK5—were differentially expressed in both the single-nucleus and bulk analyses. Evaluation in an independent external cohort supported four of the five candidates, with KCNK5 demonstrating the strongest overall evidence as a potential biomarker.
The researchers observed substantial variability in menstrual fluid cellular composition between the donors, and they found limited evidence that differences in cell-type abundance distinguished patients with endometriosis from controls. Instead, the disease signal was driven primarily by transcriptional changes within epithelial cells rather than by shifts in cellular composition.
The researchers acknowledged several limitations. This was a pilot study involving just 10 participants, limiting statistical power. All patients had deep infiltrating endometriosis, all included participants were nulliparous, and menstrual fluid composition varied considerably between the donors. In addition, more patients with endometriosis had used anti-inflammatory drugs, which may have influenced immune-related findings. The researchers said larger studies will be needed to validate the proposed biomarkers and evaluate their performance across additional endometriosis phenotypes before the findings can be translated into clinical practice.
"We identified a set of promising genes that may contribute to endometriosis pathophysiology understanding and have potential as diagnosis biomarkers in whole menstrual blood. Further assessment of their stability over time within each patient and in a larger cohort will confirm whether this represents a viable noninvasive strategy to reduce diagnostic delays," wrote lead study author Katie Leap, of the Institut Pasteur at the Université Paris Cité, and colleagues.
The study authors reported no conflicts of interest.
Source: BMC Medicine