Are degenerative changes in a single endometrial biopsy representative for the endometrial state?

Abstract

Equine endometrial fibrosis is diagnosed by microscopy of an endometrial biopsy. It is however uncertain how representative a single biopsy is for the state of the entire endometrium. While previous studies have compared samples from multiple sites, further clarification is needed on specific degenerative changes that may vary in severity between sites. Histopathology was compared, with an emphasis on degenerative changes, in two full thickness specimens from the ventral base of the left and right uterine horns, collected post slaughter from cyclic mares (n=82). Hematoxylin and eosin-stained sections were blinded evaluated by the same evaluator for layers of periglandular fibroblasts (mild[0-3], moderate[4-10], severe[>10]), glandular nests per view (mild[0-2], moderate[3-4], severe[>4]), dilated glands per view (mild[0-3], moderate[4-6], severe[>6]), degree of glandular dilation (mild, moderate, severe), excessive lymphatic vessels (absent, present), and lymphatic lacunae (absent, present). In addition, each section was scored according to a modified Kenney and Doig classification. Statistical analysis included a multiple logistic regression model for fraction of mares with disagreement in Kenney and Doig category, a multiple generalized linear regression model for fraction of parameters with disagreement, a multiple logistic regression model for fraction of mares with disagreement within each parameter, and equivalence tests for agreement of a cumulative score ofdegenerative parameters. Possible effect of age, breed and estrous cycle stage on disagreement between paired uterine horns were included in the analysis. In 49% of mares, specimens from paired uterine horns were assigned the same Kenney and Doig category, and in 46% of mares, the specimens were assigned either of two adjacent categories. In 5% of mares, the specimens varied with two categories. Disagreement in Kenney and Doig category between paired uterine horns increased with each year of age (OR=1.1, p=0.012). Analysis of degenerative parameters showed a disagreement in at least two parameters in 39% of mares, and the number of parameters with disagreement increased with age (HR=1.04, p=0.005). Disagreement within each parameter ranged from 10% of mares (glandular nests), to 31% (periglandular fibroblasts) and 38% (excessive lymphatic vessels). Disagreement increased with age for parameters involving glands (OR=1.14-1.16, p<0.04). When tested for equivalence, the cumulative scores for paired uterine horns did not differ for mares younger than 20 years (p<0.001). In conclusion, although cumulative scores of degenerative changes were equal for paired samples in mares under 20 years of age, the changes were not uniformly distributed. This suggests that conclusions based on a single biopsy, particularly in mares over 20 years of age, should be interpreted with caution, and evaluation of two biopsies is recommended.