An endometriosis classification, designed to be validated

Abstract

Several endometriosis classifications were proposed, based on the assumption that endometriosis is a progressive disease, and designed to score severity of visible lesions. In addition, several specific classifications, e.g., for deep endometriosis, were proposed. None of these classifications however, have been validated to be predictive for diagnosis, treatment prognosis, recurrence, progression or for the associated infertility or pain. The difficulties derive from the fact that pathophysiology and the natural history are still uncertain. A classification should avoid assumptions. It seems established beyond reasonable doubt that endometriosis presents as subtle, typical, cystic, and deep lesions and that severity of each lesion is related to size or volume. By pathology, these four lesions present as active, burnt-out, inactive, and active lesions, respectively. Besides this, there are many uncertainties. It is unclear whether endometriosis is one disease progressing ultimately into severe endometriosis or whether typical, cystic, and deep endometriosis represents three different diseases, each being an end stage. It is unclear whether endometriotic cells are different from endometrial cells or whether only the environment is different. It is unclear how adenomyosis, Müllerianosis, and peritoneal pockets should be considered. We therefore suggest a descriptive classification with the severity of Subtle, Typical, Cystic, Deep, Adenomyotic, and peritoneal pocket lesions, estimated by their area or volume. This classification should permit to evaluate the actual uncertainties in order to build subsequently a validated classification. The similarity of the classes for superficial and cystic lesions with the rAFS classification is considered an advantage. It is discussed why adhesions need not to be scored. In conclusion, a simple classification scoring separately severity of subtle, typical, cystic, deep, adenomyotic, and peritoneal pocket lesions is suggested. This will permit to confirm or reject statistically many of the actual uncertainties on endometriosis and to evaluate what the predictive power of the severity of each type of lesion is, both essential elements for a validated endometriosis classification.