Response: Endometriosis and malignancy—The intriguing relationship

Abstract

We appreciate and acknowledge the interest taken in our study “Endometriosis and malignancy: The intriguing relationship .1 We hereby address the concerns raised by Kvaskoff et al. regarding our paper. Kvaskoff et al. have suggested that it is incorrect to state that endometriosis is associated with an increased risk of gynecological cancer and that, due to lack of time varying analysis in our investigation, it cannot be concluded that increasing age, postmenopausal status, higher levels of CA125, larger endometriomas, and longstanding endometriosis are factors predictive of cancer risk in endometriosis patients. They have also suggested that in order to conclude that any factors are predictive of cancer risk in endometriosis patients, a longitudinal analysis is required in a cohort of patients with endometriosis followed up for cancer diagnosis over several years. We agree with the observations of Kvaskoff et al. As mentioned in our paper, the study was a crosssectional study with the aim of calculating the prevalence of coexistent malignancies (ovarian, uterine, and breast) with endometriosis, with a large enough sample size to provide clinically applicable and valid corelations. In agreement with Kvaskoff et al., we would also like to clarify that increasing age, postmenopausal status, higher levels of CA125, larger endometriomas, and longstanding endometriosis (>5 years as elicited from clinical database) only predicted an increased risk of coexistence rather than increased likelihood of a malignancy in the future (as this could only be established through a timevarying analysis). The conclusion of our study was based on a crosssectional analysis and should be interpreted accordingly. In concordance with the suggestions made, a longitudinal analysis in a cohort of patients with endometriosis, followed up for cancer diagnosis over several years, will be the most appropriate study design to prove or disprove the associations derived from our crosssectional study. However, due to the low prevalence of cancer in the cohort of patients, a longitudinal analysis which can provide temporal associations will require many years of follow up. Kvaskoff et al. have also called into question the risk of ovarian cancer in patients with endometriosis which was cited in the introduction of our study. In our study, we tried to ascertain the risk of ovarian, uterine, and breast cancer from as many relevant, recent, and reasonably large studies as possible. These included the metaanalysis by Kim et al.,2 and the most recent (2019) critical review of the literature on the association between endometriosis and gynecological cancers, performed by Kalaitzopoulos et al.3 A study by Mogensen et al.,4 one of the largest nationwide cohort studies on the subject, was also taken into account. Data on the risk of malignancy associated with endometriosis are variable, with an odds ratio/risk ratio of 2.0– 4.3.2,3,5 Many crosssectional studies have quoted a prevalence of coexistent malignancy with endometriosis, with a wide range from 2– 10%.57