With advances in fetal diagnosis and therapy, fetoscopy has become a good option to treat a series of life threatening diseases like twin-to-twin transfusion syndrome or severe congenital diaphragmatic hernia. The efficacy of fetoscopy for those types of disorders is not debatable; however, it comes with a daunting limitation: iatrogenic preterm premature rupture of the fetal membrane (iPPROM). The injury created by the instruments to access the amniotic cavity is the reason for fetal membrane rupture in almost one third of all interventions. The consequences of very early iPPROM are death or very preterm birth leading to physiological complications for the baby, and psychological burdens for the family. A solution preventing iPPROM is necessary to make fetoscopy a safer procedure. We first concentrated on developing a minimally invasive medical device to stabilize the fetal membrane defect caused by fetoscopic instruments. Our approach consists in placing bioadhesives at the wounded site from the inside of the amniotic cavity following the intervention. To do so, we designed and produced an umbrella-shape receptor whose mechanical properties enabled its crimping inside the accession catheter and ensured its automatic deployment inside the cavity. The umbrella is then pulled against the membrane and glued by gathering injectable bioadhesives. We demonstrated evidence of the functionality of the method by applying existing biocompatible glues on an ex vivo model replicating physiological conditions. While the application method is a requirement for a reliable delivery of injectable materials, questions remain about the nature of those materials. Restauration of the native tissue might provide functional advantages over the simple sealing of it. Hence, the development of platforms enabling the in vitro study of cell response to biological cues was the focus of the next investigations. First, we took advantage of the tunability of polyethylene glycol (PEG) synthetic hydrogels to locally bind growth factors (GFs) by affinity. We established a GF gradient in 3D and assessed multiple methods to quantify cell behaviour in response to those gradients. Next, we utilized a method to pre-fabricate hydrogel gradients in a full 96-well plate and established scripts to automate image acquisition, image processing and data analysis. This platform enabled a fast quantification of the influence of GF concentration on the recruitment of cells in large number of samples in parallel. We then reproduced a wound in a tissue model in vitro, injected PEG hydrogels carrying GF and quantified cell recruitment into the healing hydrogel. Finally, we used a proteomics approach to investigate the effect of bioactive signals on inducing a differential production of extracellular matrix (ECM) by the cells. This investigation was motivated by the critical role of ECM proteins in tissue functionality and thus by the importance of faithfully reconstituting the native structure of the tissue. To summarize, our efforts contributed to the advancement of fetal therapy by bringing innovative solutions in two distinct fields. On one side, by the development of a technology to apply materials at the site of defect; and on the other side, by the establishment of platforms allowing the exploration of cell behaviour. Combined, they carry a real potential to engineer methods for precise delivery of healing-inducing biomaterials and consequently to reduce the risks of iPPROM.