Commentary on Johnstone et al.: Mechanisms underlying extended pre-quit varenicline treatment for smoking cessation

Currently available smoking cessation pharmacotherapies (nicotine replacement therapy, bupropion and varenicline) have limited effectiveness, and one strategy for improving on them has been to initiate treatment several weeks before the quit date. Not only does this increase efficacy [2-5], but pre-quit initiation of treatment also affords an opportunity to assess predictors of efficacy and adapt treatment before a target quit date. [6] This new research by Johnstone et al. [1] sought to identify behavioral mechanisms underlying the efficacy of extended pre-quit treatment with varenicline.

Consistent with the hypothesis that varenicline attenuates nicotine reinforcement, the authors found that pre-quit reductions in smoking rate occurred, which, along with decreased pre-quit craving and increased self-efficacy, mediated abstinence. Surprisingly, however, subjective rewarding or aversive effects of cigarettes were not mediators of abstinence. These results are amenable to two quite different interpretations.

One interpretation is that the reinforcement mechanisms maintaining cigarette use and dependence are dissociable from the evaluative conditioning that underlies ratings of subjective reward. That is, the smoking response/nicotine reinforcement outcome association can be degraded by varenicline without substantially changing the subjective evaluation of cigarettes. Indeed, evaluative conditioning has been found to be resistant to extinction and counterconditioning [7]. Previous studies have found that some subjective responses such as ‘craving reduction’ after smoking do not appear to be affected by varenicline [8]. A related concept that might apply is ‘automatic reinforcement’ [9], whereby a behavior becomes reinforcing in itself, apart from signaling a subsequent outcome such as nicotine reinforcement. The persistence of smoking cigarettes with little or no nicotine content [10, 11] is consistent with this general view that subjective ratings of reward can be partially dissociated from nicotine reinforcement. Neural correlates of this dissociation have been identified in the dorsal striatum, premotor cortex and other brain regions [12].

A second interpretation, however, is that the perceived rewarding value of cigarettes is dissociated from abstinence because individuals adjust their rate of smoking to compensate for the reduction in anticipated reward. A cigarette smoked after a period of abstinence, such as the first one of the day, is often rated more rewarding than others [13]. Therefore, reducing smoking frequency by spacing cigarettes farther apart would tend to offset the reward attenuation by varenicline and obscure a relationship between subjective reward and abstinence outcomes. Actual subjective reward could appear unchanged even if anticipated subjective reward were reduced and mediated reductions in smoking and subsequent abstinence. An analogy can be drawn to satiation with desirable foods, which may be subjectively rewarding and yet followed by a state of satiety in which anticipated reward is reduced and consumption is suppressed [14]. There is evidence that people who smoke can anticipate changes in reward value of cigarettes in response to pharmacologic manipulations. The Cigarette Evaluation Questionnaire that Johnstone et al. [1] used to assess subjective reward was developed in my laboratory in the 1990s [15], and in one experiment, we showed that study participants predicted reductions in cigarette reward when wearing a nicotine patch even before they smoked the first cigarette after patch application [16]. The hypothesis that anticipated reward was driving changes in smoking behavior could be tested in future studies by measuring anticipated reward in addition to actual perceived reward in people receiving varenicline treatment. Ideally, participants could be asked to rate the rewarding value of controlled smoke presentations after a controlled deprivation interval, to assess reward unconfounded by changes in smoking frequency. Johnstone et al. [1] attempted to measure cigarette reward in the laboratory, but used a choice procedure to assess smoking reward that allowed participants to vary the rate of smoke delivery, and therefore, raised the same question as to whether puffing rate was adjusted to maintain subjective reward value.

In the quest to develop more effective quit-smoking treatments, it will be important to continue this line of research to better understand the interplay between reinforcement, extinction, satiation and anticipated reward, to arrive at more effective treatments to help individuals overcome their addiction to cigarettes.

Research support for smoking cessation and tobacco harm reduction from Global Action to End Smoking, Philip Morris International, Swedish Match AB, Altria, Cabbacis, Nicotine BRST, Embera Neurotherapeutics, Otsuka Pharmaceutical and JUUL Labs; consulting with JT International, SA and consulting and patent purchase agreement with Philip Morris International (final payment 2014).