2023-06-16 08:29:00
A saliva test developed in 2022 has seen its effectiveness confirmed by a clinical trial.
Last February, the French company Ziwig indicated that it had developed a saliva test allowing a early detection of endometriosis.
Today, new work, the results of which were published a few days ago in the New England Journal of Medicine confirm its effectiveness at more than 95%.
96% efficiency
The new clinical trial in question involved 200 patients with symptoms that seemed to suggest endometriosis. Aged 18 to 43, they had been recruited from five hospital departments and specialized centers.
Professor Horace Roman, surgeon who participated in the study, told Le Figaro :
This interim report already shows that the test correctly identifies 96% of endometriosis cases, with less than 5% of false negatives. This is an excellent result, on which I admit that I would not have bet a few years ago! This pathology, whose cause is still unknown, has very different faces depending on the woman.
The use of AI
And artificial intelligence has identified no less than 109 messenger microRNAs, their combined presence being linked to endometriosis.
According to Professor Roman, the diagnosis can therefore do without an exploratory laparoscopy, which requires general anesthesia.
Endometriosis: 1 in 10 women affected
This disease, of which it is still difficult to precisely trace the contours, concerns regarding 10% of women of childbearing age. Its diagnosis is complicated, largely confined to a clinical questionnaire; and often between 7 and more than 10 years following the first pain. Not to mention that an MRI is not able to flush out the lesions.
And even if the test costs more than 1,000 euros (it is on the market in a few countries), it does not locate the lesions, it remains a formidable diagnostic tool. In fact, management is improved because it reveals the proteins playing a role in the development of inflammation, and might promote genetic profiling of the pathology.
1686924928
#saliva #test #facilitate #screening
