Background Uterine spiral artery (SpA) remodelling, characterised by loss of vascular smooth muscle (VSM), is essential for successful placentation; impaired SpA remodelling occurs in late miscarriage, pre-eclampsia and fetal growth restriction. Non-remodelled and completely remodelled SpA are easily identified histologically but identification of partially remodelled SpA is less defined; various stages have been proposed based on semi-quantitative scoring. The aim was to compare semi-quantitative scoring of VSM loss with quantification of VSM.

Methods Placental bed biopsies from women undergoing surgical pregnancy termination were immunostained to assess trophoblast (cytokeratin 7), endothelial cells (factor 8), myometrium and VSM (h-caldesmon, α smooth muscle actin). SpA VSM was scored using 4 categories: SM1 = intact but separated; SM2 = <50% lost; SM3 = 50–90% lost; SM4 = >90% lost. 20 SpA were independently scored by 2 individuals who showed >95% concordance. VSM was also quantified using a computerised pixel counting (using Adobe Photoshop).

Results VSM loss was scored in175 SpA (11–19 weeks gestation) in decidua, junctional zone and myometrium; SM1 = 47; SM2 = 24; SM3 = 35; SM4 = 69. The 4 categories of VSM loss correlated to the quantified VSM loss (analysis of variance P < 0.001); differences in VSM % based on pixel counts between groups were confirmed with t tests: mean, standard deviation: SM1 76.76, 12.12; SM2 62.46, 13.03; SM3 38.50, 16.51; SM4 11.28, 12.69; P < 0.001 all cases.

Conclusion Assessment of SpA remodelling may not require quantitative assessment since semi-quantitative scoring correlates well with quantified VSMC in partially remodelled SpA. This scoring technique provides an approach to assessment of uterine SpA remodelling in pregnancy pathology.