Parkinson's disease (PD) is a neurodegenerative disease primarily characterized by severe motor symptoms that can be transiently relieved by medication (levodopa). While there is evidence of widespread alterations of the neuronal activities across the brain, reliable biomarkers are still lacking. To examine this question, we measured the resting state activities of 11 PD patients both with deep stimulation electrodes and with EEG electrodes placed in the motor areas, in order to provide simultaneous recording from both the subthalamic nucleus (STN) and the cortex. Then, we used a dynamical measure, the avalanche transition matrix (ATM) (Sorrentino et al. 2021), to study the propagation of aperiodic bursts (i.e. neuronal avalanches) between brain regions in PD. ATMs are matrices structured with regions in rows and columns. The ij-th edge is defined as the probability that region j would be active at time t+1, given that region i was active at time t (figure a). Then,we compare, in each patient, the ATMs in two conditions : ON-levodopa medication and OFF-levodopa medication. The patients showed significantly higher probability that an avalanche will travel between the STN and the motor cortex in the ON state (i.e. under medications; figure b), reflecting enhanced avalanche spread. Moreover, the probability of transition was more prominently increased for long-range connections as compared to the short-range ones during medication (figures b, c), particularly for the right STN. The right-left asymmetry is consistent with the asymmetrical (left side) clinical onset observed in this group of patients.