Surgery and Neuroscience of Addiction

This chapter includes an accompanying lecture presentation that has been prepared by the authors: .

Key Concepts

  • Addiction has been a major problem worldwide for decades. Alcoholism remains the most widely abused substance, with prescription opiates becoming as problematic in recent years as illegal drugs such as cocaine and heroin.

  • All drugs of abuse ultimately increase mesolimbic dopamine signaling to the nucleus accumbens (NAc) by different mechanisms.

  • The NAc is a popular anatomic target for focal interventions in animal models and in limited human trials. Other relevant brain regions that have been proposed as surgical targets include the lateral habenula, prefrontal cortex, and hippocampus.

  • Molecular mediators of addiction within relevant neurons include the dopamine master regulator DARPP-32, the unusually long-lived immediate-early gene variant deltaFosB, and the S100 family member p11, which regulates membrane presentation of neurotransmitter receptors.

  • Early reports with brain lesions have led to pilot studies and small series using NAc deep brain stimulation (DBS) to modulate reward signaling. Studies of heroin, cocaine, and alcohol addiction have shown promise, but there remains a need for larger, more definitive systematic human trials.

  • Experimental therapies showing promise in human brain disease, including gene therapies and focused ultrasonography, have also been considered for drug addiction.

The general lay view of drug addiction is that it most often represents a failure of individual will or resilience. Yet it has been clear for some time that addiction is a brain disorder that may be provoked by a variety of genetic, environmental, and social factors (e.g., polymorphisms in opioid receptors, poor quality of family bonds, high levels of drug availability) but ultimately results in dysfunction of relevant brain circuits, leading to dependence and an inability to resist the addictive substance. Although psychotherapies and certain drug therapies can be effective, many patients either have limited to no response, or they have substantial rates of relapse. Understanding the biology and anatomy of circuits underlying addiction can help lead to more targeted therapies, including surgical therapies, in the hope of developing major improvements in both safety and efficacy of treating drug addiction. In this chapter, we review the magnitude of the problem to be addressed, methodology used to identify and understand specific circuits that mediate addiction, anatomic and biologic pathways that may be targeted surgically, and clinical experiences with neurosurgery for addiction.

Epidemiology of Addiction

Addiction is a harmful, compulsive physical and/or psychological need for something that cannot be readily controlled and causes symptoms on withdrawal. For the purpose of this chapter, we will focus on substance abuse, primarily cocaine and opiates. Other substances that can be abused include alcohol, nicotine, and various prescription and nonprescription drugs. However, many behaviors and other activities can become addictive, including gambling and sexual behavior. In a 2019 survey, roughly 60% of the adult US population used some substance with abuse potential (e.g., alcohol, tobacco, drugs) recreationally within 1 month before responding. Of these, roughly 20 million adults are addicted to drugs or alcohol, with alcohol dependence affecting over 70% and illicit drug dependence reported in over 40% of respondents. The prescription opiate crisis has had a devastating impact on a broad range of communities, particularly in the United States, leading to a strong push to develop better pain therapies that avoid narcotics. From 1992 to 2012, the number of people reporting abuse of prescription opiates almost tripled from 4.9 million to 12.5 million, with roughly 1.7 million people developing a chronic opioid addiction. By 2015, long-term opioid dependence affected 2.4 million adults in the United States, but this decreased to 1.6 million by 2019, likely reflecting changes in prescription patterns and increased mechanisms for abuse prevention and monitoring resulting from greater awareness of the prescription opioid problem among health care workers, local governments, and the general public. Cocaine dependence affects over 2 million people, with roughly one-half of those meeting criteria for cocaine use disorder.

Despite the magnitude of the problem, only a fraction of those with addiction undergo treatment. Part of the reason for the small percentage of patients treated is the limited options for efficacious therapy in many cases. Psychotherapy and other psychosocial interventions, such as group therapies, are the most common methods for addressing drug abuse. These can help some patients, but without addressing the underlying biologic cause of addiction, very aggressive monitoring and oversight are necessary to prevent relapse, and even this is often insufficient. Opiate and heroin abuse have long been treated with methadone, which is a mu-opiate receptor agonist that does not have the same abuse or addiction potential as other drugs. Yet patients on methadone therapy usually must remain on treatment continuously for long periods or indefinitely, which can be difficult to maintain. Naltrexone is an opiate receptor antagonist also used to treat narcotic addictions. There is no drug specifically approved by the US Food and Drug Administration (FDA) for cocaine addiction, and therapies currently in use are mostly dopamine agonists attempting to modulate reward signals (see later) or GABA agonists/glutamate antagonists, which presumably dampen CNS responses to cocaine and thereby reduce the reward from continued abuse.

Clearly, the magnitude of the problem coupled with limitations in current therapies present the need for development of novel therapeutic approaches. Neurosurgeons are particularly adept at identifying and safely reaching specific brain targets to deliver therapies that can directly block dysfunctional circuits. Postulating what to deliver and where it should be deployed depends on understanding the specific brain circuitry mediating addictive behaviors and the molecular pathways that influence responses within these circuits.

Methods for Circuit Mapping in Animal Models of Addiction

Electrophysiology, histology, and behavioral analyses have long been the staple of preclinical addiction research, but in recent years techniques have been introduced to allow an unprecedented level of specificity to dissect neuronal circuits underlying substance abuse. A key development has been the identification of Cre as a site-specific recombinase allowing for deletion of sequences residing between two LoxP sites. Sequences placed between LoxP sites can then be used to disrupt and inactivate genes, which can then be removed in the presence of Cre to activate the target gene. A major advance has been the creation of so-called Cre-driver mouse lines, with Cre expression driven by highly specific genomic promoter sequences to restrict expression to particular cell types. The successful creation of the first Cre transgenic line led to a large-scale generation of Cre-driver lines and mice with floxed target genes and reporters. This permits restriction of gene expression to those cell types through either crossing Cre-driver mice with transgenic mice harboring an inactivated “floxed” gene or through injection of a viral vector containing an inactivated gene. Further development of double-floxed inverted open reading frame (DIO) and flip/excision (FLEX) constructs, which use two pairs of lox sites, have obtained more acceptance because of their less leaky nature, even under strong promoters. Combining this ability to target highly specific neuronal subtypes with powerful new methods for on-demand manipulation of neuronal activity, such as optogenetics and chemogenetics, has led to a much broader and more detailed understanding of the circuitry underlying addiction ( Fig. 125.1 ).

Figure 125.1, Technologies used for circuit mapping in addiction research.


Optogenetics comprises a combination of genetic and optical methods to activate or inhibit cell events in living tissue or fully awake animals. The control of biologic systems through light-sensitive recombinant proteins commonly requires three components: (1) microbial opsins, which transmit electrical depolarization on light exposure, (2) a vector that allows strong transgene expression in specific cell types, and (3) implanted probes with strong and precise light targeting in the desired brain region. In neuroscience, channelrhodopsin-2 (ChR2) is perhaps the most used to study causal relationships between neural circuits and phenotypes. ChR2 allows positively charged ions to enter the cell through the opsin pore in response to blue light causing depolarization, whereas halorhodospsin (NpHR) pumps Cl into the cell in response to yellow light, resulting in hyperpolarization and neuronal inhibition. Many new variations on this theme have been generated, with channels that transmit anions or cations with greater efficiency and variable kinetics in response to multiple wavelengths of light.


Chemogenetics refers to the interaction between engineered macromolecules in the cell membrane with previously unrecognized small molecules for manipulating neuron electrical activity. Most chemogenetic platforms use ligand-gated ion channels (LGICs) or G protein–coupled receptors (GPCRs). Designer Receptors Exclusively Activated by Designer Drugs (DREADD)-based chemogenetics is the most widely used approach. DREADDs selectively activate GPCRs by the inactive clozapine analogue clozapine- N -oxide (CNO). The interaction between CNO and the exogenous bioengineered protein triggers a GPCR-signaling cascade according to their class (Gq, Gi, Gs, Golf, and β - arrestin). The first Gq DREADD, named hM3Dq after the molecular manipulation of the human M3 muscarinic receptor, has been shown to depolarize neurons and enhance their excitability. Although the hM3Dq is subject to endogenous regulatory processes such as internalization, downregulation, and phosphorylation, its expression and function are stable for long periods of time. Gi-mediated signaling by the mutant receptors hM2Di and hM4Di are known to induce hyperpolarization and silencing of neuronal firing. More recently, another type of chimeric ion channel modified in the ligand-binding domain (LBD) has been reported. The so-called pharmacologically selective actuator modules (PSAMs) are channels that are sensitive to low doses of varenicline, and, similar to human muscarinic-like receptors, neuronal silencing and activation can be achieved with the PSAM4-GlyR and PSAM4-5HT3 constructs, respectively.


Magnetic-field–based stimulation (or magnetogenetics) has emerged as a new promising tool that could provide remote, reversible, fast, and implant-free modulation of neural circuits. Moreover, electromagnetic waves freely penetrate tissue to any depth. In this approach, thermosensitive and mechanosensitive ion channels of the TRP family are genetically encoded to couple a chimeric ferritin protein. Iron-loaded ferritin stimulated by the magnetic field then interacts with the TRPV1 channel to allow influx of positively or negatively charged ions into the neuron. Nevertheless, the physical and biologic properties as well as the possible mechanisms behind recent reports of gene-phenotype regulation with magnetogenetics is still an area of active debate.


Ultrasonic neuromodulation (UNM) is another noninvasive technique with great potential to target focused deep structures in the brain. Although the mechanisms of action are yet not well elucidated, ultrasonography has been capable of eliciting action potentials by opening voltage-gated ion channels. Similar to optogenetics and magnetogenetics, the sonogenetics concept was conceived as a combination of ultrasonic neuromodulation with targeted expression of mechanosensitive channels.

Animal Models of Addiction

Even though animal models of addiction, as in any other field, have limited translational predictive power, neuroanatomic and neurochemical similarities between rodents and humans in drug-intake behavior make the various models particularly attractive for testing hypotheses regarding human addiction. Furthermore, the animal models reviewed later in this chapter have been crucial to understanding the mechanisms involved in the reward system and to identify possible targets and therapeutic approaches.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here