The Puzzling Genetic Landscape of Endometriosis

- Written by Rachel Wolfson

Nora is a 29-year-old woman who you’ve been seeing in your clinic for many years. She returns to your office today discouraged and upset. She suffers from long, painful menstrual periods every month and has pelvic pain associated with intercourse and bowel movements, consistent with her diagnosis of deep infiltrating endometriosis, a common cause of pelvic pain in women. Today, she tells you that she is frustrated by her years of pain and asks, “What causes endometriosis anyway?”       

There are three anatomical subtypes of endometriosis: superficial peritoneal endometriosis, ovarian endometriotic cysts, and deep infiltrating endometriosis. Genetic predisposition is a known risk factor for endometriosis, and ovarian endometriosis is widely accepted as the direct precursor for clear-cell and endometrioid ovarian carcinomas. However, deep infiltrating endometriosis, the type that Nora suffers from, rarely undergoes malignant transformation. To date, studies of somatic mutations in endometriosis have only been conducted in patients with endometriosis and concurrent malignancy.

In this week’s issue of the NEJM, Anglesio et al. shed some light on the pathogenesis and somatic mutational landscape of endometriosis. They performed exome sequencing on 24 samples of deep infiltrating endometriosis (with matched control tissue) and found that 19 of these lesions (79%) harbored somatic mutations. Five of the 24 lesions (21%) harbored mutations in known cancer driver genes: ARIDA1, PIK3CA, KRAS, and PPP2R1A. The statistical probability that these mutations in cancer driver genes occurred by chance was very small.

These results raise a variety of interesting questions about the pathogenesis of endometriosis. First, given that it is unlikely that these mutations in the cancer-associated genes occurred by chance, they may play a role in the pathogenesis of this benign disease. Second, not all lesions harbored mutations, perhaps implying heterogeneity in the pathogenesis of deep infiltrating endometriosis. As Montgomery and Giudice note in an accompanying editorial, further genetic analysis is needed to address whether variability in clinical presentation can be described by genetic heterogeneity and lead to different drivers of disease patterns and progression.

For now, it’s hard to answer Nora’s question with certainty. You tell her that while deep infiltrating endometriosis lesions have some invasive properties and may carry mutations in cancer driver genes, these lesions are not malignant. In addition, you reassure her that future research will continue to shed light on the pathogenesis of this complex disease and its genetic drivers, leading to treatment innovations that someday might help patients like her.

Browse more From Pages to Practice »


 Rachel is a fifth-year MD/PhD student at Harvard Medical School. She is originally from Okemos, MI, and graduated from Stanford University in 2011. For her PhD research, she is studying the regulation of cell growth in response to nutrients.