Development and Optimization of PET Modeling Methods for Imaging Addiction: Characterizing the Brain's Dopamine Signature of Cigarette Smoking

Development of Positron Emission Tomography (PET) imaging methods has greatly advanced research in addiction in a noninvasive manner and has promoted the development of therapies to treat addiction. PET studies in drug addiction mainly focus on imaging dopamine (DA) responses to drug challenges. Previous PET studies achieved success in imaging addictive substances such as amphetamine and methylphenidate, which induce relatively sustained DA responses with a large magnitude, while attempts to image smoking-induced DA release yielded highly inconsistent results. Unlike amphetamine or methylphenidate, nicotine induces a mild and transient DA event and thus violates the assumption of conventional PET modeling methods which consider the endogenous DA level to be non-varying and rely on the endpoint, binding potential (BP ND), estimated from time-invariant parameters. In fact, we previously demonstrated that use of conventional methods and BP ND could not reliably quantify short-lived changes in endogenous DA.;We have developed an enhanced model ("lp-ntPET"), in recent years to detect highly localized transient DA responses from PET data, which allows us to characterize the temporal patterns of short-lived DA responses to cigarette smoking in the striatum on a voxel-by-voxel basis. Subsequent to publication of the method, we used it to identify the brain's DA responses induced by cigarette smoking (while in the PET scanner). Thanks to our novel analysis method, we were able to identify a significant sex difference in the DA signature of cigarette smoking that could explain the differential efficacy of nicotine replacement therapy for men and women.;Going forward, we would like to guide the neuroimaging community in the use of this novel analysis tool to image addiction or to study psychiatric disorders by maximizing its sensitivity to detect neurotransmitter release induced by stimulus. We experienced several challenges from our previous study that need to be addressed to improve the sensitivity. One obstacle was getting sufficient radioactivity dose. Another challenge was the high noise level of voxel-by-voxel analysis. In addition, a common question we are asked by researchers from other groups who are considering using the 1p-ntPET approach in their studies is at what time point the drug challenge or task should be started during the scan and how long the scan duration should be. We conducted multiple simulation studies to address these challenges by optimizing the experimental designs and image processing strategies for [11C]raclopride PET that maximize the sensitivity of the 1p-ntPET method to small and short-lived dopamine responses given practical limitations on the scan duration and available radioactivity dose. The framework of our optimization process could be extended to 1p-ntPET analyses of other studies probing other neurotransmitter systems.;We also performed studies to address the question of temporal resolution of the 1p-ntPET modeling technique, i.e., the closest distance in time of two different DA responses that can be significantly distinguished. We presented the precision of DA timing parameters estimated from noisy simulated data using the 1p-ntPET approach. We also demonstrated the ability of this novel analysis tool to recover DA response profiles and to distinguish responses of different timing characteristics with the aid of a computer vision clustering algorithm in both simulated datasets and data collected from real human scans.