Effect of Extending the Duration of Prequit Treatment With Varenicline on Smoking Abstinence

Key Points Question Does extending the duration of prequit treatment with varenicline improve rates of smoking abstinence? Findings In this randomized clinical trial that included 320 adults, the percentage of bioverified continuous 4-week abstinence among participants at the end of treatment was not significantly greater among participants receiving extended (39%) compared with standard (36%) run-in varenicline. Meaning These findings show that among adult daily smokers, extending the duration of prequit varenicline treatment beyond the standard 1-week run-in period did not significantly improve continuous abstinence rates.

5][6][7] As a dual agonist and antagonist of nicotinic receptors, varenicline should decrease smoking even before a quit attempt; in learning theory terms, it should promote extinction of smoking behavior as a consequence of decreased reinforcement.Consistent with the hypothesis that varenicline reduces smoking reinforcement, varenicline has reduced nicotine self-administration in rats in a dose-dependent manner. 8,91][12][13][14] Thus, the standard 1-week run-in period for varenicline (during which the dose is titrated) is likely insufficient to maximize extinction.[17] These findings suggest that the efficacy of varenicline could be enhanced by preloading, 18,19 or increasing the duration of medication provided before the target quit date (TQD).The results of 2 preliminary randomized clinical trials 20,21 (RCTs) support this hypothesis.In both trials, extended run-in groups exhibited greater reductions in pre-TQD smoking (without accompanying increases in craving or withdrawal) and nominally higher rates of smoking abstinence at the end of treatment (EOT) relative to standard run-in conditions.Although the results were encouraging, both studies had critical limitations: they were small (100 and 60 participants), used lenient 21 or no 20 bioverification, and had only 3 months of follow-up.Results of a subsequent larger trial 22 (242 participants) suggesting an advantage of extended run-in varenicline treatment are difficult to interpret because participants were instructed to reduce pre-TQD smoking by more than 50%, resulting in considerable pre-TQD attrition, particularly in the standard run-in condition.
The primary aim of the present RCT was to rigorously test the hypothesis that extended run-in varenicline treatment produces better smoking abstinence rates than standard varenicline therapy.
The study used a substantially larger sample (N = 320) to provide adequate statistical power, used cotinine levels (which reflect past-week exposure to nicotine) to provide stringent bioverification of self-reported continuous abstinence, and followed up participants for 6 months after the TQD.
The study was also powered to evaluate candidate moderators, specifically sex.Although women generally have lower rates of long-term abstinence than men, 23,24 women exhibit a greater response to varenicline (relative to placebo) than men. 25In a pilot study, 21 only women showed greater pre-TQD reductions in smoking exposure and greater postquit abstinence rates with extended run-in varenicline treatment (relative to standard varenicline treatment).Replication of those preliminary findings could be of considerable practical importance for helping women quit smoking.

Methods
The trial protocol and statistical analysis plan for this RCT are provided in Supplement 1.The study was conducted in accordance with the ethical principles of the Declaration of Helsinki 26 and was approved by the institutional review boards of the University at Buffalo, Buffalo, New York (where data were collected) and the University of Toronto, Toronto, Ontario (where biological assays were performed).All participants provided written informed consent.The study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

Study Design
This RCT used a 2-group, balanced, randomized, double-blind, placebo-controlled, parallel-group design (a trial design schematic is provided in the trial protocol, section 5, in Supplement 1).Groups were distinguished by duration of varenicline treatment before the TQD (ie, run-in duration).The group with extended run-in received 4 weeks of varenicline before TQD.The group with standard run-in received 3 weeks of placebo, followed by the standard 1-week pre-TQD varenicline run-in treatment.Both groups received 11 weeks of post-TQD varenicline treatment and brief quit counseling.

Randomization
Participants were randomized by the statistician (C.R.C.) in small blocks (2:2) via prespecified tables implemented in REDCap. 27,28Randomization was stratified within self-reported sex (an a priori hypothesized moderator, men vs women) and race and ethnicity (non-Hispanic White vs all other racial and ethnic groups); both variables were collected per National Institutes of Health policy. 29

Recruitment and Participants
Adult smokers of combustible cigarettes who reported living within 50 miles from the University at Buffalo were recruited (beginning September 2017) through advertisements (radio, social media), recruitment databases (state Quitline, ResearchMatch.org), and a project website until the target sample size of 320 participants was enrolled.Inclusion criteria included smoking 10 or more cigarettes per day (CPD) for at least 6 months and an expired-air carbon monoxide level greater than 7 ppm at intake (to reduce exclusion of Black participants, the CPD criterion was reduced to Ն5 and the carbon monoxide criterion was eliminated in November 2019), motivation to quit smoking, 30 and 18 to 70 years of age.Exclusion criteria included use of other tobacco and/or nicotine products in the past 7 days or use of cessation medication in the past 14 days, lifetime diagnosis of schizophrenia or bipolar disorder, current use of antipsychotic medication, suicidal ideation in the past year, 31 current major depression, 32 moderate-to-severe risk of involvement with illicit or nonmedical prescription drug use, and pregnancy.For additional details, see the trial protocol, section 6, in Supplement 1.
Participant disposition is summarized in Figure 1, and demographic and smoking characteristics are presented in the Table .Participants were enrolled from October 2, 2017, to December 9, 2020.

Study Procedures
After an initial telephone screen, participants completed an intake visit (at which participants provided written informed consent), a baseline laboratory visit (discontinued March 2020 owing to the SARS-CoV-2 pandemic; data are reported elsewhere 33 ), and a baseline week of daily smartphone assessments.Participants made 6 clinic visits to the University of Buffalo-based research clinic at baseline (week 0), during the pre-TQD drug manipulation phase (weeks 1 and 3), on the TQD (week 4), and at weeks 6 and 8. Follow-up visits occurred at week 15 (EOT) and week 28 (6 months).All visits included assessment of vital signs; measurement of expired-air carbon monoxide level (Bedfont Inc); self-reports of smoking rate, craving, affect, withdrawal, and adverse events; and collection of a saliva sample for cotinine, trans-3′-hydroxycotinine (3HC), and varenicline assessment.Study medication and counseling (described below) were provided at each clinic visit.Participants were compensated as much as US $598 for completing study visits and measures.

Interventions Pharmacotherapy
Pfizer Inc provided varenicline and identically appearing placebo, which was dispensed at each clinic visit (except during the COVID-19 shutdown [April 1 to June 15, 2020], when dispensing occurred less frequently).Participants were initially dispensed a 1-week supply of tablets labeled 0.5 mg to be taken orally (1 morning tablet for days 1-3 and 1 morning and 1 evening tablet for days 4-7 consisting of varenicline for the extended run-in group and placebo for the standard run-in group).During weeks 2 and 3 of the medication manipulation phase, participants were given 1.0-mg tablets to be taken twice daily (varenicline for the extended run-in group and placebo for the standard run-in group).To maintain blinding and allow titration for participants assigned to the standard run-in group, during week 4 all participants were provided with 4 tablets/d (2 for morning and 2 for evening) marked 0.5 mg (varenicline for the extended run-in group; for the standard run-in group, 1 of 2 morning tablets was varenicline and, for days 4-7, 1 of 2 evening tablets was varenicline).
Beginning with the TQD, all participants were dispensed open-label 1.0-mg varenicline twice daily.

Counseling
Participants received brief (approximately 15 minutes) behavioral counseling per a treatment manual adapted from prior studies. 21,34Counseling focused on honing quit motivation; preventing or managing adverse effects, stress, and smoking triggers; and relapse prevention.To promote continued smoking and foster extinction, participants were instructed to smoke as usual during weeks 1 to 3. 21,35,36 Brief (approximately 5-minute) telephone check-ins occurred during weeks 5 and 11.

Outcome Measures
Our prespecified primary outcome was continuous abstinence at EOT (no self-reported cigarette smoking [not even a puff] during the final 4 weeks of treatment [weeks 12-15]), bioverified with EOT c Data were missing for 10 of 320 participants (3.1%).
g Scores range from 0 to 10, with higher scores indicating greater dependence on cigarettes.
i Data were missing for 93 of 320 participants (29.1%).This variable was added partway through the study.

Sample Size Calculation
A priori analysis used Monte Carlo simulations and abstinence rates from the pilot study 21 and focused on the primary outcome of continuous abstinence at EOT with 2-tailed α = .05and power of 0.90.With 320 participants, the study was powered to detect an increase in abstinence rate of at least 13% in the extended run-in group, compared with an expected 40% abstinence rate in the standard run-in group.With respect to sex moderation, the study was powered to detect a difference of increased abstinence of 32% in the extended run-in group for women compared with a decline in abstinence rate of 10% in extended run-in group for men, as previously observed. 21

Statistical Analysis
Data Significance level was 2-tailed α = .05.The primary outcome, continuous smoking abstinence during the final 4 weeks of treatment (bioverified at EOT), was analyzed based on intention-to-treat (N = 320) via logistic regression; participants with missing data were considered smokers (results of supplemental multiple imputation analyses are described below).Continuous abstinence at 6 months (a secondary outcome) was analyzed similarly.Percentage reduction in self-reported smoking rate (CPD) during the prequit period (a secondary outcome) was analyzed in a group × sex analysis of variance.

Baseline Characteristics
A total of 320 participants were randomized, including 179 women (55

Participants
As shown in Figure 1, retention was comparable between the extended and standard run-in groups

Secondary Outcomes Abstinence at 6 Months
The results for continuous abstinence at 6 months were similar to those obtained at EOT (Figure 3).
We observed no significant effects in run-in group (OR, Supplemental analyses of continuous abstinence at EOT and 6 months that incorporated covariates, considered moderators, and used multiple imputation for missingness all yielded similar results; there were no statistically significant main effects or interactions involving the run-in group  (eAppendix 2 in Supplement 2).The same was true in analyses of 7-day point prevalence abstinence at EOT and 6 months (eAppendix 2 in Supplement 2).

Percentage Reduction in Pre-TQD Smoking Rate
As hypothesized, although both groups self-reported reducing their smoking rate (in CPD) during the prequit period (Figure 4), the reduction was greater in the extended run-in group compared with

Unregistered Outcomes Craving and Withdrawal
Consistent with our hypothesis, craving at week 4 (TQD) was greater in the standard compared with the extended run-in group, a difference that was stronger among women than men.From weeks 4 to 8, withdrawal increased in the standard run-in group but declined in the extended run-in group, a pattern that tended to be stronger among women than men (details in eAppendix 4 in Supplement 2).

Adverse Events
Symptom reports were generally as expected for varenicline (details are provided in eAppendix 6 in Supplement 2).During the run-in manipulation period (weeks 1-3), nausea and abnormal dreams were more common among participants in the extended run-in group vs those in the standard run-in group; these differences were not significant once the standard run-in group began active varenicline therapy.Similar patterns were observed for multiple gastrointestinal tract symptoms, other sleep problems, and dizziness.Consistent with our hypotheses, the group difference in nausea was primarily driven by women.Serious adverse events were rare (3 in each run-in group) and deemed unexpected and unrelated to study medication (details in eAppendix 6 in Supplement 2).

Discussion
Previous large RCTs have demonstrated that both standard and extended run-in varenicline treatments enhance smoking abstinence relative to placebo treatment, 2,44,45 but these trials do not address the question of whether extended run-in varenicline treatment is better than standard varenicline treatment.From the perspective of extinction learning, a longer duration of pre-TQD varenicline treatment would allow numerous prequit opportunities for participants to experience a reduction of reinforcement while smoking, resulting in reduced prequit smoking exposure (without increasing craving or withdrawal) and improved postquit abstinence rates.
As hypothesized in the present RCT, the extended run-in group exhibited reduced smoking exposure during the prequit period (ie, lower self-reported smoking rate and cotinine levels) and attenuated peri-TQD craving and withdrawal relative to the standard run-in group.Rates of attrition were typical, albeit somewhat lower at 6 months for men in the extended run-in group.Overall, from a mechanism and process perspective, the trial results were largely as expected and supported the internal validity of the study.
However, rates of bioverified, self-reported continuous smoking abstinence at the 3-month (EOT; the primary outcome) and 6-month follow-ups were not significantly higher for extended run-in varenicline treatment.Thus, preliminary support for extended prequit varenicline treatment observed in a pair of pilot RCTs funded by Pfizer Inc, 20,21 was not substantiated in the present, much larger trial (320 participants vs 100 for Hajek et al 20 and 60 for Hawk et al 21 ).Although abstinence rates in the standard run-in condition were somewhat lower than estimated based on the pilot study, 21 they were similar to rates observed in recent large-scale varenicline trials. 43,46sed on the pilot RCT 21 and other evidence of sex differences in the neurobiology and behavioral pharmacology of smoking and of smoking cessation, [23][24][25]47 we hypothesized that the extended run-in varenicline treatment would outperform standard run-in treatment, specifically among women. Alhough the rates of abstinence were in the hypothesized direction, the interaction was not statistically significance at EOT or at 6 months.Moreover, there was no evidence that sex moderated the run-in group effect on pre-TQD changes in smoking behavior (CPD) or exposure (cotinine level).These data argue against the a priori hypothesis that pre-TQD varenicline treatment would extinguish smoking to a greater extent among women than men.33 Craving and withdrawal (unregistered but commonly reported outcomes in varenicline trials 1,20,21,48,49 ) were moderated by sex in a manner consistent with the hypothesized greater benefit for women.It is possible that extended run-in varenicline treatment potentiates, at least among women, the mechanisms of action recently reported 50 for standard varenicline (reduced craving and negative affect).However, in the present study, this did not translate into a significant enhancement of smoking abstinence.The present study may have been underpowered to detect the effect on dichotomous abstinence, because our a priori power analysis was based on a preliminary study 21 with a small sample size, limiting the precision of the estimate.Given that women generally have more difficulty maintaining long-term abstinence than men, 23 it may be worthwhile to test these hypotheses in the context of a larger, women-only multisite trial.

Limitations
In addition to the concern about statistical power just mentioned, additional limitations merit consideration.First, internal validity could have been compromised by the confounding of the duration of prequit varenicline treatment and total duration of varenicline treatment.That is, as in prior studies of medication preloading, 18,[20][21][22]35,36,51 the extended run-in group received a longer total duration of varenicline treatment (15 weeks) than did the standard run-in group (12 weeks).
However, because even doubling the duration of postquit varenicline treatment from 11 to 23 weeks did not enhance long-term abstinence in a large, recent trial, 52 any effects of extended prequit varenicline on abstinence-had they been statistically significant-were unlikely to be accounted for by the total duration of treatment.Second, although we followed recommendations to relax the exclusion of participants with conditions that are comorbid with smoking, including depression, schizophrenia, and problematic use of illicit and prescription drugs. 53Third, the a priori focus on a relatively short follow-up period (3 months) for the primary abstinence outcome, although common in varenicline trials 1,45,48,54,55 and supplemented by a 6-month follow-up, is also a limitation.

Conclusions
In this RCT, although extended run-in varenicline treatment reduced pre-TQD smoking rate to a greater extent than standard run-in varenicline treatment, it did not significantly enhance rates of smoking abstinence at the 3-or the 6-month follow-up.Thus, on average, adult daily smokers did not significantly benefit from extending the duration of prequit treatment with varenicline beyond the standard 1-week run-in period.

Figure 2 .
Figure 2. Bioverified Continuous Abstinence Rates for Each Run-in Group and Each Group × Sex Condition at End of Treatment

Figure 3 .
Figure 3. Bioverified Continuous Abstinence Rates for Each Run-in Group and Each Sex × Group Condition at 6-Month Follow-up

Figure 4 .
Figure 4. Percentage Reduction in Self-reported Smoking Rate for Each Run-in Group and Each Sex × Group Condition

Table .
Baseline Participant Characteristics for Each Run-in Group and Each Sex × Group Condition a Unless otherwise indicated, data are expressed as No. (%) of participants.To convert cotinine level to nmol/L, multiply by 5.675.bIncludes American Indian or Alaska Native, more than 1 race or ethnicity, and other race or ethnicity.