Progesterone receptor expression is selectively reduced in adenomyosis among endometriosis-associated lesions: a cross-sectional immunohistochemical study of 118 specimens from 100 women

Abstract Background. Endometriosis and adenomyosis are both oestrogen-dependent and both are treated with progestins, yet response to progestin therapy is not uniform. Progesterone resistance is usually described as a property of endometriosis as a whole. Whether reduced progesterone receptor (PR) expression is shared by all endometriosis-associated lesions, or is concentrated in one lesion type, has not been established in a cohort with the lesion site recorded for every specimen. Methods. We analysed 118 surgical specimens from 100 consecutive women operated on for endometriosis-associated disease at one centre between November 2022 and May 2026. Oestrogen receptor (ER) and PR were quantified immunohistochemically using the modified H-score (0-300). Each specimen was assigned one lesion site: adenomyosis (uterus, n = 49), ovarian endometrioma (n = 60), deep infiltrating endometriosis (DIE, n = 4) or other (n = 5). Because 12 women contributed more than one specimen, the primary analysis was a linear mixed-effects model with patient as a random intercept, adjusted for age, pelvic adhesions and menopausal status. Three further analytical strategies and a within-patient paired comparison were prespecified as sensitivity analyses. Results. PR differed across lesion sites (Kruskal-Wallis p < 0.001), and the difference was confined to adenomyosis: median PR H-score 135.0 (IQR 60.0-185.0) versus 210.0 (150.0-255.0) in endometrioma, 220.0 (183.8-257.5) in DIE and 230.0 (185.0-260.0) in other lesions, the three non-uterine sites being mutually indistinguishable (all Holm-adjusted p = 1.00). Adenomyosis remained associated with lower PR after adjustment (beta = -75.7, 95% CI -102.5 to -48.9, p < 0.001) and under all four strategies. Among nine women contributing both an adenomyotic and a non-uterine specimen, PR was lower in the adenomyotic lesion in 9 of 9 (Wilcoxon p = 0.004). ER showed no comparable site effect (adjusted beta = -24.3, 95% CI -48.9 to 0.3, p = 0.053) and instead tracked age (beta = + 22.9 per SD, p = 0.004). Conclusions. Within endometriosis-associated disease, reduced PR expression is a property of the adenomyotic lesion rather than a generalised feature of the disease, and is dissociated from ER, which is governed by age. Lesion type may be relevant when progestin therapy is selected.