Introduction

Tobacco use has become a worldwide epidemic, and cigarettes are the most common form of tobacco consumed. Smoking cigarettes constitutes the most sophisticated system of drug delivery (nicotine) to the brain, resulting in diseases that are among the most common and the deadliest and yet the most preventable of those involving human behavior: respiratory diseases, cardiovascular diseases, and cancer.¹ Tobacco use accounts for at least 30% of all cancer deaths in the United States, and smoking has been causally linked to 18 different cancers, including lung, head and neck, esophageal, pancreatic, bladder, kidney, cervical, endometrial, and gastric cancers and acute myeloid leukemia.² The scare of a diagnosis of cancer often impels smokers to quit, and although some people resume use of tobacco once their cancer is in remission, others manage to avoid a relapse.

The high recidivism rates could be attributed to comorbid conditions and to the neurobiology of tobacco addiction. Tobacco addiction is a complex phenomenon. Nicotine receptors are spread throughout most areas of the brain, and nicotine consumption activates the reward pathway, like all other substances of addiction.³ In addition, several other neurotransmitter systems are implicated in the process of establishing an addiction to nicotine.⁴ After years of tobacco consumption, complex neuronal involvements and adaptations develop, rendering the extinguishing of tobacco addiction a challenging task. However, pharmacologic and behavioral treatments have proven to be essential and highly effective in comparison with other medical treatments of chronic diseases.⁵ In recent years, individualized treatment has become the gold standard, with hopes that advancement in genetics will allow us to eventually match the best treatment to each individual profile.⁶ Unfortunately, multiple hurdles and obstacles remain in the way of extending basic treatment to all smokers and tobacco users, including those diagnosed with cancer and cancer survivors.⁷ In this chapter we summarize these aspects of tobacco use and their implication for patients with cancer.

Epidemiology and Tobacco Use in Patients with Cancer

Tobacco use, in particular smoking cigarettes, began an upward trend around the beginning of the twentieth century. This trend has been mostly attributed to James Buchanan “Buck” Duke, who introduced the mechanized manufacturing of cigarettes and used aggressive marketing techniques,⁸ in addition to other societal forces and circumstances such as the free cigarettes provided with daily rations to all enlisted American soldiers during World Wars I and II.⁹ That upward trend continued until a pivotal moment: the release of the landmark first Surgeon General’s report on smoking and health in 1964,¹⁰ at which point the evidence from more than 7000 scientific publications on harm from tobacco smoke was put together and received the endorsement of Surgeon General Luther Terry, the highest officer for public health within the United States government. That report had an enduring effect on public beliefs and attitudes about tobacco. As an example, a Gallup Survey conducted in 1958 found that only 44% of Americans believed smoking caused cancer, whereas 70% believed so by 1969, 5 years after the report was published. The belief that smoking causes cancer continued to rise, and the rate of belief reached 94% by 1990.¹¹ Furthermore, since 1964 a gradual and almost annual decrease in smoking rates has occurred in the United States, with the rate reaching a plateau of around 20% in the past 5 years. This lower rate nonetheless encompasses 60 million Americans who, almost 50 years after the landmark report, continue to use tobacco despite the clear knowledge of harm from its consumption. Other Surgeon General reports subsequently solidified the case against tobacco consumption, and to date 33 of those reports have been dedicated to tobacco, focusing on specific topics such as nicotine addiction, secondhand smoke, minorities, women, and youth.¹² Although these efforts have improved the public health, unfortunately, disparities in health care extend to tobacco consumption, as evidenced by higher smoking rates among American Indians/Alaskan Natives, persons with diagnosis of mental disorders, those who are under the poverty level of income, and the unemployed. It is of importance that level of education is inversely related to smoking rates.¹³

Although the trend for smoking among patients older than 45 years and who are recovering from cancer has decreased to about the same as for the general population, the prevalence among survivors in the younger age group is still high. Around 40.4% of the persons 18 to 44 years of age in the survivors cohort still smoke cigarettes compared with 24.7% in the general population of the same age range.¹⁴ Nevertheless, one potential confounding factor contributing to these statistics is the high mortality rate in patients with smoking-related tumors. Therefore smokers who would survive less lethal cancers are counted within the younger age group.¹⁵

Smoking cessation rates in the cancer population seem highly dependent on the disease (tumor) site and its relationship to smoking.¹⁶ As an example, around 80% of patients with lung cancer and 65% of patients with head and neck cancer were reported to stop smoking in the year after their treatment^17,18 compared with 31% of patients with bladder cancer.¹⁹ Other factors come into play in improving smoking cessation among patients with cancer, such as having a good prognosis, having spent a long time in the hospital (a smoke-free environment), and having treatment-related adverse effects. In particular, adverse effects of treatment such as nausea, difficulty breathing, facial surgery, or difficulty swallowing may make it harder or more difficult to smoke. Consequently, physicians should pay attention to the particular smoking-cessation pattern in patients with cancer. Typically, cancer survivors have an increased chance of a delayed relapse to smoking 1 to 6 months after treatment,²⁰ in contrast to persons in the general population, who generally relapse within the first week or so of cessation. Unfortunately, it has been reported that patients with lung cancer who have already abstained from smoking prior to their diagnosis tend to return to smoking once they are informed of their cancer.²¹ Many studies have been published on the benefits of smoking cessation in patients with cancer; many of the studies focus in particular on cancer treatment–related outcome (Table 24-1). Unfortunately, almost all published studies to that effect have methodologic limitations such as retrospective analyses, nonstandardized tobacco assessment, no biochemical confirmation, and no tracking of behavior after diagnosis. Addressing smoking behavior at the time of diagnosis and throughout treatment and recovery is thus essential for persons with cancer, constituting a continuum of potential teachable moments as described by McBride et al.²²

Like most drugs associated with abuse and dependence, nicotine use stimulates its receptors in the brain. These receptors are spread throughout most areas of the brain. Nicotine use leads to a rapid increase in dopamine release in the nucleus accumbens and the ventral tegmental area. This stimulation typically occurs within 10 seconds of smoking a cigarette.23,24 It has been established that natural rewards such as food consumption, social affiliation, and sexual activity, which are linked to survival of the individual or species, normally activate these two central areas of the reward pathway within the brain. Furthermore, the reward pathway projects from the nucleus accumbens and ventral tegmental area to the prefrontal cortex, the amygdala, and the olfactory tubercle (Figure 24-1), leading to a complex interaction between the reward, its saliency, and the ability to inhibit the intake of a drug. Several other brain systems (neurotransmitters and pathways) are thought to be involved in the process of addiction to a substance, whereas dopamine appears to be the final common neurotransmitter in the reward pathway. These other pathways are thought to be involved via neuroadaptation (the brain adjusting to the new conditions). The γ-aminobutyric acid, glutamate, serotonin, norepinephrine, cannabinoid, opioid, and cholinergic systems have all been shown to be involved in one way or another in the brain neuroadaptation process that results in addiction to nicotine and other substances.^27–27 However, it is important to note that these major neurotransmitter systems in the brain are in constant and dynamic interplay, possibly resulting in different types of neuroadaptations that lead to and maintain the addiction to a substance.⁴

Several recent genome-wide association (GWA) and candidate gene studies have revealed significant associations between smoking, nicotine dependence (ND), and lung cancer in a region of interest on the long arm of chromosome 15 (15q24/15q25.1), encompassing the nicotinic acetylcholine receptor CHRNA3-A5-B4 subunit cluster. Neuronal nicotinic acetylcholine receptors are pentameric ligand-gated channels composed of α and β subunits,³⁴ several of which contribute to nicotine-stimulated dopamine release in the striatum, a region associated with reward learning and vital to the development of substance dependence.³⁵ The CHRNA3-A5-B4 subunit cluster demonstrates extensive linkage disequilibrium, and the single nucleotide polymorphisms on that cluster are frequently correlated. Some of the statistically strongest associations with ND in persons of European or African ancestry have been obtained for the nonsynonymous CHRNA5 single nucleotide polymorphism rs16969968.^36–41 Other gene variants in this region (CHRNA3-A5-B4) identified by GWA and candidate gene studies as being significant contributors to the risk of smoking, ND, and lung cancer include the CHRNA3 rs1051730,^41–45 the CHRNA3 rs578776,^36,38 and the AGPHD1 rs8034191.^36,42,44 Finally, candidate gene studies have identified associations between ND and genetic markers associated with dopamine expression (ANKK1,^44,46–48 DRD2,^48,49 and SLC6A3^49–52), serotonin expression (SLC6A4),^55–55 and nicotine metabolism (CYP2A6).⁴⁵

An increasing amount of work has been focused on the investigation of genes that inform brain regions related to nicotine dependence. The CHRNA5 rs16969968 risk allele was found to be associated with decreased resting state connectivity strength in a dorsal anterior cingulate-ventral striatum/extended amygdala circuit associated with smoking.⁵⁶ A small functional magnetic resonance imaging study that investigated the role of CHRNA5 rs16969968 on brain reactivity to smoking cues in nicotine-dependent women found that women homozygous for the G allele had greater functional magnetic resonance imaging reactivity to smoking cues in the hippocampus and dorsal striatum, which are areas related to memory and habitual behavior.⁵⁷ Smokers identified as fast nicotine metabolizers, as indicated by the CYP2A6 gene, had greater brain responses to smoking cues in the amygdala, cingulate cortex, hippocampus, insula, and striatum compared with slow metabolizers.⁵⁸

Comorbid Conditions

The smoking rates among persons with no mental illness, past-month mental illness, and lifetime mental illness have been reported to be 22%, 34%, and 41%, respectively, which means that having a current or past mental disorder effectively doubles the likelihood of being a smoker. Furthermore, in a nationally representative sample, it was reported that smokers with a mental disorder in the past month consume 44% of all cigarettes smoked.⁸⁷ The evidence that smoking is more prevalent in and closely linked with several psychiatric comorbidities, including abuse of other substances, suggests a shared biological pathway between nicotine dependence and these psychiatric and substance use conditions. For example, several studies have demonstrated a positive correlation between smoking and alcohol use, abuse, or dependence, substance abuse or dependence, and other psychiatric disorders.^87–94 When looked at from the opposite direction, the prevalence of lifetime alcohol dependence and/or drug abuse with the adult smoker population is estimated to be between 23% to 30%.^95,96 Similarly, among current smokers (whether tobacco-dependent or nondependent), the lifetime rates of mood and anxiety disorders have been reported to range from 33% to 46% and from 12% to 26%, respectively.⁹⁵ Furthermore, among current smokers who are tobacco-dependent, even the 12-month prevalence of any mood or anxiety disorder is reported to be 22%, a relatively high proportion.⁹⁷ Adding support to the idea of a possible causal link between smoking and mental health disorders is the fact that smokers also have an elevated risk of a first onset of major depression, panic disorder, or generalized anxiety disorder.^98–102 Cognitive dysfunction has also been highly linked to smoking, as evidenced by the odds ratio comparing persons who have ever smoked with persons who have never smoked, and smoking is positively related to the number of attention deficit hyperactivity disorder (ADHD) symptoms. There seems to be an effect of early onset on the severity of ADHD or vice versa, because lifetime regular smoking has an inverse relationship between the number of ADHD symptoms and the age at onset. In addition, a positive relationship between the number of symptoms and the number of cigarettes smoked has been observed.¹⁰³ The aforementioned evidence of co-occurrence supports the importance of screening and treating mental health disorders among smokers, whether the relationship is causal or a simple correlation. Treating these disorders may increase the ability to focus on quitting smoking, or at least reduce the impact of negative affect or comorbid mental disorders on a person’s ability to quit using tobacco, and may have an impact on their resilience with regard to maintaining tobacco abstinence. Treating these disorders is of particular importance among patients with cancer and cancer survivors, because higher levels of confidence about the ability to quit smoking were found to be related to lower depression scores and lower tumor stage.¹⁰⁴

Patient Assessment

In the absence of specific biological markers, the diagnosis of nicotine addiction remains a clinical one. The Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR),¹⁰⁵ uses universal criteria for all substance dependence (traditionally referred to as addiction), including nicotine. Because of their universality, the DSM-IV-TR criteria are not particularly specific to nicotine and therefore do not capture many of the peculiar aspects of tobacco use and ND (addiction). This lack of specificity has led to the development of specific scales to quantify ND. Traditionally, the Fagerström Test of Nicotine Dependence (FTND) has been used, and more recently the Wisconsin Inventory of Nicotine Dependence Motives (WISDM) has become more accepted as a more comprehensive scale.

Despite their limitations, the DSM-IV criteria offer ease of use for clinicians because of their universality for all substances of dependence. According to DSM-IV-TR, the diagnosis is made when someone meets three or more of the seven criteria for at least a year. As previously mentioned, some criteria are more pertinent than others to nicotine. For instance, after prolonged smoking, the user develops nicotine tolerance and begins to exhibit withdrawal symptoms when nicotine is absent. That withdrawal can happen overnight with nicotine’s half-life in blood being 1 to 2 hours.¹⁰⁶ Having tolerance and/or withdrawal is referred to as physiological dependence in the DSM-IV-TR. Nicotine may also be responsible for three other criteria for dependence: loss of control over smoking (e.g., not being able to reduce or stop smoking, or smoking more than intended), compulsive smoking (e.g., smoking all day long and giving up important events or activities if there is no place to smoke), and continued smoking despite adverse consequences (e.g., heart attack, emphysema, or cancer). The newer DSM-5 criteria (planned to be released in 2013) are proposed with a similar universal applicability to all substances; however, they expand the “use despite consequences” criteria and incorporate “craving for a substance” as an important new dimension, and only two of those criteria would suffice to have the diagnosis of “substance use disorder,” a new nomenclature. The more criteria that are met, the more severe is the disorder, which will effectively replace the term “dependence,” which in past versions of the manual had been introduced to replace the term “addiction.”¹⁰⁷

The FTND (Table 24-2) is one of the most widely recognized and used scales for nicotine dependence. This tool, and in particular its first item (“How soon after you wake up do you smoke your first cigarette?”), measures physiological dependence reliably well (i.e., tolerance and withdrawal). However, it does not measure well the psychosocial and behavioral dimensions of ND.¹⁰⁸ A person with three points or more out of a possible maximum of 10 points on this scale is customarily considered to be ND.^109,110

The WISDM¹¹¹ is a more recently developed multidimensional scale of ND. This comprehensive scale includes measures of cognitive enhancement, negative reinforcement, positive reinforcement, automaticity, affiliative attachment, loss of control, behavioral choice/amelioration, craving, cue exposure/associative processes, social/environmental goals, taste/sensory processes, weight control, and tolerance. The WISDM has been validated and found to be reliable in measuring ND, and it has been reported to have internal consistency along a 13-factor model.¹¹²

At the Tobacco Treatment Program (TTP) at MD Anderson Cancer Center, we use breath carbon monoxide monitors to measure breath carbon monoxide as a strategy to provide patients with a concrete measurement of their smoking behavior. They can monitor it as it decreases, which we hope will motivate them and enhance their self-efficacy to proceed further until they quit smoking. We also utilize a battery of screening questionnaires that each patient is asked to complete at his or her first visit to help the clinician get a complete picture of each patient’s unique presentation. These tools aim to screen for the severity of ND and the most common concurrent conditions and include the following scales:

Treatment of Tobacco Use

Tobacco treatment among patients with cancer needs to be as intensive and tailored as possible. Hence if treatment programs are to be comprehensive and to make a difference, they ought to provide at a minimum behavioral modalities and medications.¹¹⁹ Comprehensive treatments by definition would need to address the biological, psychological, and social aspects of ND. Moreover, these approaches ought to be longitudinal, because tobacco addiction is a chronic disease that often is characterized by multiple relapses before reaching full remission.¹²⁰ The main components of a biopsychosocial treatment are behavioral therapies and social support in conjunction with medications. Among the various therapies, motivational interviewing, skill-building, problem-solving, and cognitive behavioral therapies have been empirically supported.⁵ Although the main first-line medications that are approved by the Food and Drug Administration (FDA) are NRTs, bupropion (Zyban), and varenicline (Chantix), the second-line medications clonidine and nortriptyline have been used off-label and tested in clinical trials with similar positive results.

Among several other medications that have been or are being researched, three have shown promising preliminary results but have faced different challenges: topiramate, rimonabant, and nicotine vaccines. Although a care provider may prescribe any available medications off-label, to date, none of the aforementioned three medications is FDA-approved for smoking cessation purposes, and only topiramate is available in the United States. However, their importance lies in their mechanisms of action, which may lead to improved or similar agents in the future. Topiramate, a voltage-dependent sodium channel blocker, augments γ-aminobutyric acid, antagonizes glutamate, and inhibits carbonic anhydrase activity; it has FDA label indication as an adjunct treatment for seizure disorders and migraine headaches.¹²¹ Topiramate is reported to have a beneficial effect in the treatment of alcohol dependence and ND but has several adverse effects that limit its use.¹²² Rimonabant is a selective CB1 endocannabinoid receptor antagonist reported to have beneficial effects for the treatment of obesity¹²³ and ND¹²⁴; it was shown to be relatively efficacious compared with placebo (odds ratio 1.6)¹²⁵ in large multinational trials. It was available in Europe from mid 2006 until January 2009; then the European Agency for the Evaluation of Medicinal Projects recommended its withdrawal because of neuropsychiatric adverse effects (i.e., depression and several reports of suicides).¹²⁶