Psychopharmacology (2006) 187:476–485 DOI 10.1007/s00213-006-0449-5
ORIGINAL INVESTIGATION
Preferences among four combination nicotine treatments Nina G. Schneider & Margaret A. Koury & Chris Cortner & Richard E. Olmstead & Neil Hartman & Leonard Kleinman & Andrew Kim & Craig Chaya & David Leaf
Received: 28 February 2006 / Accepted: 20 May 2006 / Published online: 1 August 2006 # Springer-Verlag 2006
Abstract Rationale Acute nicotine replacement treatments (NRTs) are disliked or misused, leading to insufficient nicotine intake and poor outcome. Patches provide steady nicotine but are slow and passive. Combining systems may improve efficacy with acute NRTs tailored for compliance. Objective To test initial reactions to and use preferences among combinations of NRTs. Materials and methods Smokers (n=27) tested four combination NRTs in a 5-day crossover trial: 2/4-mg gum + 15mg patch (G/P), 2/4-mg lozenges + 15-mg patch (L/P), inhaler + 15-mg patch (I/P), and 10 mg + 15-mg patches (P/P). Subjects rated an NRT combination each day after 5– 6 h of use and ranked among the NRTs after testing all treatments. Results Double-patches (P/P) were ranked highest for “ease of use”, “safety”, and “use in public”. However, for “help to quit”, 70% preferred some form of acute-patch combination (A/P) compared to 30% choosing P/P. For “use under stress” (an immediate need), 93% preferred A/P
N. G. Schneider (*) : M. A. Koury : C. Cortner : R. E. Olmstead : N. Hartman : L. Kleinman : A. Kim : C. Chaya : D. Leaf Veteran Affairs Greater Los Angeles (VAGLA) Healthcare System, Building 210, 2nd Floor, 11301 Wilshire Boulevard, Los Angeles, CA 90073, USA e-mail:
[email protected] N. G. Schneider : R. E. Olmstead : N. Hartman Department of Psychiatry and Biobehavioral Sciences, University of California (UCLA), Los Angeles, CA 90024, USA L. Kleinman : D. Leaf David Geffen School of Medicine, University of California (UCLA), Los Angeles, CA 90024, USA
systems compared to 7% choosing P/P. L/P ranked lowest for “ease of use”, I/P and L/P were lowest on “safety”, and I/P ranked lowest for “use in public”. Expectations of NRTs changed with test experience for patches (better) and lozenges (worse). Conclusions In brief testing, all combinations were acceptable. P/P was favored for ease, safety, and public use, but a majority chose A/P systems for help in quitting and use under stress. Combined use is viable and needs to be made known and accessible to smokers. Keywords Nicotine replacement treatments . Combination treatments . Preferences . Nicotine gum . Nicotine inhaler . Nicotine lozenge . Nicotine patches . Nicotine dependence . Tailoring
Introduction Nicotine replacement treatments (NRTs) are well-established therapies for smoking (Henningfield et al. 2005; Silagy et al. 2004) but are still not used properly or sufficiently (Bansal et al. 2004). Long-term efficacy is low in clinical trials (10–30%; Foulds et al. 2004; Hajek et al. 1999; Shiffman et al. 2003; Silagy et al. 2004) and lower in practice (Hughes et al. 2003). In a meta-analysis of gum and patch use, Hughes et al. (2003) reported success rates of ∼7% at 6 months. Though easily accessed and ubiquitous, there is also low over-the-counter (OTC) use (Cummings and Hyland 2005). NRTs approved in the US include “acute” systems (gum, lozenges, nasal spray, and inhalers) and slower-acting patches (Henningfield et al. 2005); sublingual tablets are available in Europe (Glover et al. 2002; Schneider et al.
Psychopharmacology (2006) 187:476–485
2004). Development of new NRTs is ongoing (Cummings and Hyland 2005; Fagerstrom 2005) as are improvements to existing forms including a faster-acting gum (Niaura et al. 2005). Bupropion and, more recently, varenicline are non-nicotine treatments approved by the Food and Drug Administration (FDA) with other agents (e.g., rimonabant) and nicotine vaccines in development (Fagerstrom and Balfour 2005). NRTs currently dominate treatment and, with various routes available, there is an opportunity to tailor and/or combine treatment. Our focus is on two key NRT problems relevant to combined use and tailoring: (1) difficulties with single NRT and (2) dislikes associated with route (Schneider et al. 2004, 2005). Both can lead to insufficient nicotine intake, poor compliance, and failure. Single vs combination NRT treatment All acute NRTs require a high frequency of self-administration. Nicotine gum is used every 1–2 h initially and chewed for 20 min with similar requirements for lozenges, inhalers, and nasal spray. Underdosing is common (Bansal et al. 2004; Stapleton 1999) and can easily undermine the weaning process and success. Shiffman et al. (2002), in a study of lozenges, reported that increased use led to greater success. The insufficiency of single treatment (too little use) appears to be due to route dislikes (Schneider et al. 2004, 2005) and misunderstanding of how NRTs work (Bansal et al. 2004). Stapleton (1999) notes that underdosing reflects the inability of smokers to “master” NRT use and/or reactions to their side effects. Transdermal patches were a potential solution (used once a day); however, they are slow and passive and offer no means to manipulate nicotine for urges or “rescue” use. This can be essential as smokers are used to immediate effects and fine self-tuning of nicotine delivery with cigarettes. Combining patches with acute NRTs is a logical method for treating steady and ad lib nicotine needs (titration, crisis, and cued use) in cessation (Fagerstrom 1994; Schneider 1992). There is also less acute NRT use required in combinations. Several efficacy trials show that combination NRT can be superior to single treatment (Bohadana et al. 2000; Croghan et al. 2003; Kornitzer et al. 1995; Puska et al. 1995; Sutherland 1999) though Sweeney et al. (2001) note that the findings are not “robust”. In a recent study of multiple NRT use (Steinberg et al. 2006), the subjects had access to bupropion, patches, and four acute NRTs (gum, inhalers, lozenges, and nasal spray) to use in any combination of their choice. The researches showed greater success as the number of NRTs used increased (one to four or more treatments), particularly in the first month. In an experimental trial (Fagerstrom et al. 1993), combination NRT was shown to provide better relief. Over
477
3 days of abstinence (double-dummy, crossover trial), active patch + gum relieved withdrawal entirely vs either drug alone and in contrast to a double-placebo control. The issue of dosage is key as increased dosage per se could explain some findings. In Fagerstrom et al. (1993), cotinine concentrations and withdrawal relief increased linearly with more active drug. However, higher dosage patch testing has not shown much improvement (Stapleton 1999). While Shiffman et al. (2002) showed that more use of lozenges improved success, the frequency and constancy of use required may not be as viable as combining systems. It is likely that both sufficient dosage and access to ad lib NRT systems figure in combination treatment. Even if combining NRTs served solely to improve dosage and relief, that alone would be a valid reason for combining drugs. Combinations are currently recommended in guidelines for treatment in the USA (Fiore et al. 2000; Sachs 2005) but combinations are not approved by FDA and package inserts state that one NRT cannot be used with another. Thus, combined use will be unknown to smokers and discouraged in the USA. Physicians can prescribe “off label” but there is little intervention by physicians (Revell and Schroeder 2005). Nonetheless, the logic of combining systems is sound for improving success. Preference testing for single or combined treatment A corollary to individual vs combined use concerns is the issue of reactions to NRTs. NRT pharmacotherapy remains a trial-and-error process. Compliance can be easily undermined by strong dislikes associated with route of administration (sensory/ritual, ease, and embarrassment) (Schneider et al. 2004, 2005). One solution may lie in pre-testing acute NRTs and determining preferences among them. Medications could then be tailored for single or combined use to improve outcome. There are few preferences trials. Leischow et al. (1997) found that patches were preferred to 2-mg gum (in 2-day withdrawal testing) but the subjects used little gum. Fagerstrom et al. (1997) tested five NRTs (2-week crossover trial) and found (a) smoking reduction to be better with choice vs assigned drug and (b) inhalers to be least helpful. West et al. (2001) found no differences in 15-week outcome for choice vs assigned NRTs (four drugs) but this was based on video descriptions only (no actual drug testing); they note that the findings may be due to selection bias (the subjects knew the drugs to be tested) and that the subjects “came to like” any assigned treatment. In a crossover (n=48), we tested preferences among five NRTs each used ad lib for half a day (Schneider et al. 2004). The inhaler was preferred most (49%) followed by 4-mg gum (24%); 2-mg gum and 2-mg sublingual tablets were each chosen infrequently (10%) with a nasal spray
478
Psychopharmacology (2006) 187:476–485
chosen least (7%) due to side effects. In the trial, we knew little of how drugs were used, e.g., inhalers may have been puffed infrequently for easier use, which can reduce nicotine intake and result in insufficient relief during cessation. In a second trial (Schneider et al. 2005), subjects (n=16) tested three NRTs in one 5-h crossover session with monitored use of each NRT (vs ad lib). The subjects tested gum (2 and 4 mg), inhaler, and nasal spray. With enforcement of proper use (e.g., frequent inhaler puffing, slow chewing), gum was ranked over inhaler and nasal spray for ease of use, safety, and use in public. For a subset (n=9) trying two doses of each NRT, the inhaler ranked last on relief of withdrawal, use under stress, and choice to help quit. In this paper, we report on comparative acceptability and preferences among four combined NRTs. This was a crossover trial of 15-mg patches combined with: gum (2 and 4 mg), lozenges (2 and 4 mg), inhalers, and a 10-mg patch. Each combination was tested for half a day. We predicted that acute/patch NRTs would be preferred to double-patches as acute NRTs allow control of nicotine and address sensory/ritual needs. Among acute + patch conditions, we expected differences due to route of the acute NRT.
elements and in nicotine pharmacokinetics relevant for testing preferences. All NRTs used in this study are OTC except the inhaler. The three acute NRTs all work via oral absorption of nicotine; venous blood concentrations of nicotine range from one fourth to one half those of smoking depending on dosage of the drug (Schneider et al. 2001). In brief, nicotine gum (2 and 4 mg) is chewed slowly or chewed/parked in the cheek for 20 min. The systemic delivery for 2 and 4-mg gum is 1 and 2 mg (Benowitz 1993). About 25% of nicotine remain in the gum (Schneider et al. 2004). Nicotine lozenges (2 and 4 mg) are dissolved slowly in the mouth. A single lozenge fully dissolved can provide 25–27% more nicotine than gum (Shiffman et al. 2002) and takes an average of 15–35 min to dissolve (Choi et al. 2003). The nicotine inhaler consists of a holder into which is inserted a “plug” containing 10 mg nicotine (∼4 mg available) and 1 mg menthol (to reduce irritancy). It is “puffed” but not lit and users must puff frequently (∼80 puffs in 20 min) to extract nicotine. Shallow puffs are superior to deep puffs (fewer side effects) as little gets into the lungs (Schneider et al. 2001). Nicotine patches are placed on skin (dry, non-hairy area). Nicotrol 30 cm2 patches (15 mg nicotine, 16-h use) were used daily; a 20-cm2 (10 mg) patch was added for the double-patch condition.
Materials and methods
Design: crossover—four combined treatment conditions
Subjects
This was a four-condition, repeated-measures crossover design. Subjects were given baseline measures at entry and on a Monday (while smoking) to obtain baselines similar to the time of day NRTs were to be tested. On Tuesday–Friday, subjects tested each of four NRT combinations: 2- and 4-mg gum + 15-mg patch (G/P), 2- and 4-mg lozenges + 15-mg patch (L/P), inhalers + 15-mg patch and 10 + 15-mg patches (P/P). These were randomized to four orders (Latin Square) with a treatment appearing in each serial position only once (partial counterbalance). The subjects resumed smoking each afternoon in an attempt to create similar circumstances across days.
Smokers 25–55 years of age who smoked ≥15 cig/day and had carbon monoxide (CO) scores of ≥15 ppm were recruited via print/radio ads and postings on http://www. Craigslist.org. Exclusion included substance abuse, use of psychotropic medication, mental or acute medical conditions, cardiovascular disease, pregnancy/breast-feeding and mouth, throat, gastrointestinal and dermatological problems that preclude NRT use. Those with reactions to aspartame (in lozenges) or menthol (in inhaler) were also excluded. The subjects were paid US$150 for 1-week testing. Treatments
Measures NRTs “wean” smokers from nicotine by eliminating arterial “boli” and high blood nicotine concentrations of smoking (Henningfield et al. 1990; Russell and Feyerabend 1978). No NRT produces the carcinogens and gases of burned tobacco. We used three forms of acute NRT which act within 7–60 min (Choi et al. 2003; Schneider et al. 2001) and patches with nicotine concentrations which rise from the first hour to peak concentrations at ∼8 h (Benowitz 1993). Acute NRTs vary in sensory/ritual
Baseline measures Baseline assessments at entry included several one-time measures: medical/smoking/NRT use histories, Fagerstrom Test of Nicotine Dependence (FTND) (Heatherton et al. 1991) and ratings of six characteristics of smoking. Both entry and Monday baselines included two craving/withdrawal scales (Schneider 2004, 2005), heart rate, blood
Psychopharmacology (2006) 187:476–485
pressure, CO scores (Bedfont Smokerlyzer), and unstimulated saliva for cotinine estimates of nicotine intake (Schneider et al. 1997). Repeated and end-of-testing measures Vital signs, craving/withdrawal tests, CO testing, and saliva collection were repeated at each laboratory visit. The subjects filled out diaries on daily use of acute NRTs and side effects were recorded. Preference testing included ratings and rankings as described below. We also assessed changes in expectations at the end of testing. Data on craving and withdrawal will be reported in a separate paper. Preference ratings included seven-point Likert testing (1=“very definitely not” to 7=“very definitely”) of individual or combination NRTs for the following items: relief of withdrawal, relief of urges, easy to stay off cigarettes, liked administration, ease of use, enough nicotine, too much nicotine, might become dependent, still crave cigarettes, side effects bothersome, uncomfortable using in public, and would use to quit. Taste and tingling related to each NRT were also rated on a Likert scale. Preference rankings (forced choice) were taken after testing all combinations (one to four scale: 1=top choice, 4=last choice). Subject rankings of combined NRTs included the following items: relief of craving, relief of withdrawal, ease of use, safest to use, preferred use in public, use under stress, use >3 months, most nicotine, prevent from slipping, and final preference for help to quit. The subjects additionally ranked individual NRTs (either among the three acute systems or among all four systems) for some of the same variables plus best in morning, best after meals, and side effects. A “change in expectations” scale was given at the end of testing in which the subjects rated each NRT (better, same, or worse) for three variables: “ease of use”, “liking of drug”, and “help in quitting”.
Procedures The study was approved by VA/UCLA institutional review boards and performed in accord with ethical standards of the 1964 Declaration of Helsinki. The subjects came to the entry baseline between 3 and 6 P.M. and first provided informed consent. This was followed by tests for CO and vital signs, an electrocardiogram and a brief physical examination to determine eligibility. Saliva samples were collected and the subjects completed baseline measures. On Monday, while continuing to smoke, subjects came in for a second baseline (scales/CO scores/vitals/saliva) taken at the same time of day as the four NRTs to be tested (Tuesday–Friday). Subjects were given supplies for their first combination NRT to begin on Tuesday morning. Each
479
subsequent day, subjects were instructed in the use of the next day’s NRTs and of dosage limits. Gum was limited to six doses of 2- or 4-mg gums, and four doses of 2- or 4-mg lozenges were allowed. Subjects tested at least one of each dosage of gum/lozenges. There was a 750-puff limit for inhalers with use of up to six fresh nicotine “plugs”. Each test day, upon waking, subjects abstained from smoking and applied a 15-mg patch. Subjects also started their first acute NRT on waking (their choice of dosage for gum/lozenge). This allowed the subjects to discriminate acute effects before patch use. For double-patch (P/P), subjects used both upon waking. Subjects came to the lab after 1/2 day use of NRTs; craving/withdrawal, vitals, CO and side effects were assessed. Use diaries and saliva were collected. Subjects then rated NRTs on use variables. For gum and lozenges, ratings were obtained by dosage. Subjects resumed smoking postvisit to the lab. On Friday, after testing all combinations, subjects ranked preferences among NRTs on use variables, ranked overall choice to quit smoking, and rated changes in expectations. Subjects were thanked and paid. Statistical analyses Preference ratings were subjected to ANOVAs. Rankings were subjected to chi-square analyses.
Results Subjects Fifty-three subjects were enrolled. A total of 21 subjects dropped out: seven were no-shows and others stopped due to: travel problems (one), unrelated illness (four), family problems (one), protocol violations (three), and mild adverse events related to NRT [five: stomach ache, nausea, jaw irritation, sore throat, elevated heart rate (the heart rate was questionably related to drug)]. Five subjects did not test both dosages of gum or lozenges and were dropped from analyses. The final sample consisted of 27 subjects (20 male subjects/7 female subjects; 17 Caucasians/7 African–Americans/3 Asian–Americans). Table 1 shows the demographics of the sample.
Use of medications Average treatment time was figured from use of treatment (acute or patch) on waking to when all treatment ceased. The average use time was 5.99 h (SD=1.23 h). Table 2 shows the average number of times each acute NRT was self-administered.
480
Psychopharmacology (2006) 187:476–485
Table 1 Demographics Measure
Mean
SD
Range
Age Age started smoking Education in years Cigarettes/day FTND (0–10)a Time to first cigarette (0–3)b Baseline cotinine (nanograms per milliliter) Baseline CO scores (parts per million) Number of quit attempts
35.3 18.2 14.0 22.4 5.2 2.2 227.0
8.7 6.8 1.8 6.6 1.2 0.6 109.8
25–54 9–44 11–16 15–40 3–7 1–3 44–397
23.4 4.1
7.3 5.1
15–45 1–25
a
High score=more dependent. Range: 0 (least)–10 (most dependent) High score=more dependent. Range: 0=after 60 min, 1=within 31– 60 min, 2=within 6–30 min, 3=within 5 min
b
Slips and CO scores at visits Self-reported slips and CO scores served as partial checks on the 1/2 day of abstinence (we could not assess bedtime CO to determine the half-lives of CO for subjects at visits). Subjects reported slips as follows: five on P/P, three on G/P, three on I/P and one on L/P; a few of these slipped in more than one condition: two of the five P/P slipped on G/P; one of the five P/P was also the one who slipped on L/P. An ANOVA among CO scores from the Monday baselines and four NRT conditions (for all subjects including “slippers”) was significant (n=27, p
