Self-driving medicine: closed-loop therapeutics for autonomous disease control

Abstract Closed-loop gene circuits are paving the way toward self-driving medicine: therapeutics that continuously sense physiological signals, compute responses, and autonomously tune therapeutic output. Yet clinical practice still relies largely on open-loop dosing and static interventions that are inherently blind to real-time fluctuations in disease dynamics. Since Paracelsus, medicine has recognized that dose governs both efficacy and toxicity, but closed-loop therapeutics now redefine dose as a continuously computed response to physiology rather than as a fixed schedule. Enabled by synthetic biology, feedback-controlled therapeutics can be built from modular sensors, processors, and effectors to maintain homeostasis around defined physiological setpoints, moving therapy beyond intermittent interventions toward continuous sensing and adaptive control.