Ubiquitous computing is resulting in a proliferation of cyber-physical systems that host or manage valuable physical and digital assets. These assets can be harmed by malicious agents through both cyber-enabled or physically-enabled attacks, particularly ones that exploit the often ignored interplay between the cyber and physical world. The explicit representation of spatial topology is key to supporting adaptive security policies. In this paper we explore the use of Bigraphical Reactive Systems to model the topology of cyber and physical spaces and their dynamics. We utilise such models to perform speculative threat analysis through model checking to reason about the consequences of the evolution of topological configurations on the satisfaction of security requirements. We further propose an automatic planning technique to identify an adaptation strategy enacting security policies at runtime to prevent, circumvent, or mitigate possible security requirements violations. We evaluate our approach using a case study concerned with countering insider threats in a building automation system.