Progesterone receptor status predicts aggressiveness of human endometriotic lesions in murine avatars.

Objective: To use murine avatars for studying human endometriotic lesion response to 2 different hormonal regimens to determine whether progesterone receptor (PR) can prospectively predict response to progestin-based therapy. Endometriosis is a chronic gynecologic disease afflicting 1-in-10 reproductive-age women; however response to medical therapy is highly variable because endometriotic lesions do not consistently respond to first-line progestin-based therapy. We have previously demonstrated in a retrospective study that PR status in lesions is correlated with response to progestins. Here, we hypothesize that a prospective approach using PR status to predict response to medical will allow clinicians to individualize effective, timely treatment for this debilitating disease.

Design: Patient-derived xenograft murine model.

Setting: The study was performed in an academic center and hospital research laboratory.

Animal(s): Eight-week old NOD/SCID mice undergoing transplantation of endometrioma lesions collected from women undergoing surgery for endometriosis.

Intervention(s): Daily subcutaneous injections with vehicle (dimethyl sulfoxide), medroxyprogesterone acetate (MPA), or gonadotropin-releasing hormone (GnRH) antagonist, cetrorelix, for 1 month.

Main outcome measure(s): Lesion size 1 month after treatment.

Result(s): Lesions with high PR demonstrated a robust response to MPA compared with lesions with low PR. The mean post-MPA treatment size in high-PR lesions was sixfold smaller than that in low-PR lesions. Low-PR lesions respond far more completely to GnRH antagonist than to MPA. Surprisingly, there were differences in response to GnRH antagonist between low- and high-PR lesions. High-PR lesions responded almost completely to GnRH antagonist with a 21-fold smaller posttreatment size on average than low-PR lesions.

Conclusion(s): The use of murine avatars to test clinical response is a novel approach in endometriosis. Hormonal suppression of disease is a cornerstone of therapy; however, response is not fully predictable. We have previously shown that women with low-PR lesions respond poorly to progestin-based therapy. Here, we prospectively validate our previous work using a mouse xenograft model, demonstrating that lesions with low-PR expression do not respond to progestin-based therapy; PR status predicted response to progestin-based therapy. Moreover, PR status identifies a more aggressive form of endometriosis that is not only progesterone resistant but is also less dependent on estradiol for growth. Our findings highlight the need to develop novel nonhormonal therapies aimed at targeting the more aggressive forms of endometriosis that do not rely on the usual hormonal signals.