Sammy Hermans¹, Jacob Pilon¹, Dennis Eschweiler², Johannes Stegmaier², Carmen A. H. Severens – Rijvers³, Salwan Al-Nasiry⁴, Marc van Zandvoort^5,6, Dimitrios Kapsokalyvas^1,7

¹Department of Genetics and Cell biology, Maastricht University, Maastricht, The Netherlands.²Institute of Imaging and Computer Vision, RWTH Aachen University, Aachen, Germany³Pathology, Maastricht University Medical Centre, Maastricht (MUMC), The Netherlands.⁴Obstetrics and Gynaecology, GROW, Maastricht University Medical Centre (MUMC), Maastricht, The Netherlands.

⁵Department of Genetics and Cell biology, GROW, CARIM, MHeNS, Maastricht University, Maastricht, The Netherlands.

⁶Institute for Molecular Cardiovascular Research IMCAR, University Hospital RWTH Aachen, Aachen, Germany

⁷Interdisciplinary Centre for Clinical Research IZKF, University Hospital RWTH Aachen, Aachen, Germany

* Correspondence: Dimitrios Kapsokalyvas

Address: PO Box 616, 6200 MD Maastricht, The Netherlands

Telephone: N/A email: d.kapsokalyvas@maastrichtuniversity.nl

Short running title: Phenotyping pre-eclampsia based on 3D-imaging

Abstract

Objective:Pre-eclampsia is a severe placenta related complication of pregnancy and aetiological knowledge, with limited early diagnostic and therapeutic options. Phenotyping of native placental three-dimensional (3D) morphology offers a novel approach to improve our understanding of the functional and structural placental abnormalities underlying this clinical syndrome. The aim of this project was to develop a 3D placental imaging protocol using multiphoton microscopy (MPM) and demonstrate quantifiable imaging targets for phenotyping 3D features of pre-eclampsia.

Design: Exploratory pilot study.

Setting: Single centre, MUMC.

Sample: Formalin fixed placental biopsies from: term control (n=3), pre-eclampsia (n=3), preterm birth (n=2), 2^ndtrimester placenta (n=1), and intra-uterine growth restriction cases without pre-eclampsia (n=2).

Methods: Placental slabs were visualised with MPM. Collagen and cytoplasm (based on inherent signal), and fluorescently stained nuclei and blood vessels, enabled the visualization of villous tissue with subcellular resolution. Segmentation based on pixel classification, deep learning, and clustering algorithms were used to generate quantifiable features.

Main outcome measures: Trophoblast arrangement, 3D-villous tree structure, syncytial knots, fibrosis, and 3D-vascular networks were identified as imaging targets. Villous morphology, vascular fraction, vascular network (i.e., branchpoint density and diameter), nuclear density, and knot fraction were quantified to describe placental phenotypes.

Results: Pre-eclamptic placentas had disorganized trophoblast arrangement, decreased vascular fraction, and altered vessel diameters, compared to control placentas. The developed 3D-methodology indicated that placental vasculature, syncytial knotting, and villous growth are altered in pre-eclampsia.

Conclusion: Our preliminary data demonstrate the potential of the developed quantification method for phenotyping pre-eclampsia, to improve future disease stratification.

Keywords: Placenta, multiphoton microscopy, pre-eclampsia, placental syndromes, stratification.