Commentary: Endometriosis--epidemiologic considerations for a potentially 'high-risk' population

Endometriosis impacts quality of life and productivity. Women may require multiple admissions for surgery and prolonged pharmaceutical treatment with problematic side effects. In 2002, the annual costs of endometriosis in the USA were between $2 and $22 billion (assuming prevalence1⁄4 1 or 10%). Despite disease prevalence and morbidity burden, few epidemiological studies of endometriosis exist, with the predominant design being case–control, which is impaired by the difficulty of control group identification, as endometriosis can only be validly diagnosed via surgical visualization. Perhaps synergistically with steroid hormones, endometriosis may be dependent on immunological and inflammatory abnormalities. It is not clear as to what roles these play in influencing the volume of retrograde menstruation or promoting extra-uterine implantation and survival. Perhaps as a result of common pathogenesis or as a consequence of the endogenous response to the presence of endometrial plaques, studies suggest increased risk of autoimmune diseases among women with endometriosis as well as increased prevalence of vascular inflammation and abnormalities in fibrinolysis. The risk of several cancers may be greater among women with endometriosis—with the largest number of investigations focusing upon melanoma, ovarian and breast cancers. In this new paper, ‘Endometriosis risk in relation to naevi, freckles and skin sensitivity to sun exposure: the French E3N cohort’, Kvaskoff et al. expand their focus on the relation between endometriosis and melanoma risk to cutaneous factors that appear to be shared markers of risk. The authors go on to suggest candidate genes that may be associated with both melanoma risk and endometriosis. This potentially multi-systemic aetiology may limit the discovery of individual candidate genes associated with endometriosis, whereas control population variability and underpowered study has resulted in lack of replication for most observed associations. As genome-wide association studies of various cancers, cardiovascular, autoimmune diseases and endometriosis are underway, the next wave of genetic investigation will speed identification of shared candidate genes and haplotypes. The French E3N Cohort (E3N) adds to the known large population-based prospective cohorts that have collected endometriosis data. Previous publications include the Iowa Women’s Health Study—although surgical diagnosis data were not available; the California Teachers Study (CTS), with 5000 endometriosis cases reported at baseline and 229 incident surgical diagnoses through 2003; and the Nurses’ Health Study II (NHS2), in which 7000 cases were reported at baseline with 2527 incident laparoscopic diagnoses through 2005. Not only do these cohort studies provide extensive time-dependent covariate data, but they minimize the influence of undiagnosed cases that are diluted within the large comparison groups. However, within each of these cohorts, the majority of cases were prevalent at enrollment—indeed within E3N, this imbalance necessitated a nested case– control design. In addition, the surgical diagnosis of endometriosis has improved subsequent to the years of peak incidence within these cohorts and disease definitions have altered. This is particularly true for E3N, in which the primary diagnostic years include an era when diagnosis was often restricted to the observation of ‘powder-burn’ lesions or endometriomas. None of the cohorts can address risk factors or consequences of adolescent endometriosis. The cohorts also differ in the prevalence of concurrent infertility and the mode of surgical diagnosis with the CTS being primarily at hysterectomy and the NHS2 primarily laparoscopy—perhaps reflective of the younger Corresponding author. 181 Longwood Avenue, Boston, MA 02115, USA. E-mail: [email protected] 1 Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA. 2 Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA. 3 Department of Epidemiology, Harvard School of Public Health, Boston, MA, USA. Published by Oxford University Press on behalf of the International Epidemiological Association