Menstrual fluid could provide useful biological signals for a future endometriosis test. A French research team has identified candidate genes whose activity differed between small groups with and without the condition. It has not established a ready diagnostic test or shown how reliably such a test would identify people presenting with symptoms.
The interesting development is the specimen itself: material shed during a period contains cells from the uterine lining that researchers can examine without taking a surgical tissue sample.
Why menstrual fluid is different from an ordinary blood sample
Institut Pasteur’s explanation describes menstrual fluid as a window into cells originating in the endometrium, the lining of the uterus. It contains a mixture of cell types rather than only the circulating material collected from a vein.
That gives researchers a way to ask which cells carry a useful signal. It does not mean the sample directly shows every endometriosis lesion elsewhere in the body, or that collecting it is the same as having a diagnostic result.
The September institutional account describes research published on 30 June 2026. It is a later explanation of the same work, not independent confirmation or a newly available testing service.
What the small study found
The BMC Medicine paper analysed menstrual-fluid material from ten donors: five with endometriosis and five controls. The main comparison used single-nucleus RNA sequencing, a laboratory method that reads gene activity in individual cell nuclei.
The strongest differences appeared in epithelial cells, which line the uterine cavity, rather than simply across every immune cell in the sample. The investigators then examined whole-sample RNA, seeking signals that could remain detectable without analysing every nucleus separately.
They highlighted five candidate genes: TIMP2, AKR1C2, DMBT1, FERMT1 and KCNK5. A candidate biomarker is a measurable feature worth investigating. Identifying it is an earlier step than establishing a test with a fixed threshold and known performance.
There was an external check, with important qualifications
The researchers also examined an existing independent menstrual-fluid dataset. Some signals were encouraging, especially KCNK5, but the comparison was not uniformly consistent. TIMP2 shifted in the opposite overall direction, a difference the authors linked to the mixture of immune and other cells in the samples.
The paper also identifies limitations in that external dataset, including small size, different processing and repeated measurements. This is more evidence than a discovery comparison alone, but it is not independent prospective validation of a finished diagnostic assay.
That distinction matters. An extra dataset can help test whether a biological signal recurs. It does not automatically establish the proportion of symptomatic patients a proposed test would correctly identify or exclude.
The collection process is part of the test-development problem
The ten analysed donors were a selected subset whose samples met volume and RNA-quality requirements. Collection occurred on the second day of menstruation, and the material was processed promptly in the laboratory. All analysed endometriosis cases had deep infiltrating disease, sometimes alongside other conditions; the analysed donors had not given birth.
Those details limit how far the results can be generalized. A future test would need evidence across its intended population and realistic collection conditions, including whether signals remain stable between cycles and how sample handling affects the result. Large numbers of cell nuclei do not turn ten donors into a large patient study.
Research promise and present-day assessment
NHS information on endometriosis describes clinical assessment and further investigations that can include ultrasound, MRI and laparoscopy. The menstrual-fluid research does not replace that assessment, and it should not be read as evidence that a negative research signal rules the condition out.
The next useful milestone is a clearly specified assay tested in an appropriate, independent clinical population. Until then, the study’s value is in identifying where to look and which biological signals deserve further investigation—not in supplying a home-test accuracy percentage or a date when such testing will become routine.