Bupropion XL and SR Have Similar Effectiveness and Adverse Event Profiles When Used to Treat Smoking Among Patients at a Comprehensive Cancer Center

Abstract

Background and Objectives:

Bupropion extended release (XL; once-daily dosing) has equal efficacy with the sustained release (SR) formulation (twice-daily dosing) for treating depression, but no studies have compared the two formulations for the treatment of smoking. In a naturalistic open label study, we compared the effectiveness and the adverse event profiles of XL and SR in treating cancer patients for smoking.

Methods:

Cancer patients (N=648) were prescribed bupropion XL (n=454) or SR (n=194) alone or in combination with nicotine replacement therapy (NRT) for treating smoking from September 2006 to December 2017. We analyzed 7-day point prevalence abstinence at end-of-treatment (EOT; 3 months post-medication initiation) and evaluated for noninferiority. We also analyzed the adverse event profile differences between the medications.

Results:

There were no significant differences in abstinent rates at EOT between bupropion XL and SR when using intent-to-treat models, regardless of concomitant NRT. XL demonstrated noninferiority in treatment efficacy compared to SR when excluding those on combined treatment with NRT. Further, there were no significant differences in spontaneously reported adverse events between XL and SR.

Conclusions:

Our data did not reveal a difference between bupropion XL and SR formulations in terms of effectiveness or adverse event profiles among cancer patients prescribed bupropion alone or in combination with NRTs to quit smoking.

Scientific Significance:

In this first published direct comparison of their effectiveness and adverse event profiles, we found that bupropion XL is likely therapeutically equivalent to bupropion SR when treating smoking among cancer patients, and produces similar side effects.

INTRODUCTION

Bupropion is a non-selective inhibitor of dopamine and norepinephrine transporters and an antagonist of neuronal nicotinic acetylcholine receptors.¹ As an atypical antidepressant, bupropion has comparable efficacy to selective serotonin reuptake inhibitors (SSRIs) in treating individuals with Major Depressive Disorder.² However, bupropion does not share many of the adverse events associated with SSRIs, and it is often used as an adjunct to SSRIs to counterbalance several adverse events, including fatigue, sexual dysfunction (low libido), poor concentration, and weight gain.³ While initially developed as an antidepressant, bupropion’s potential as a treatment for smoking was later identified during post-marketing surveillance.⁴ It has since been recognized as a front line medication for smoking cessation, both as a single agent or in combination with other medications, including nicotine replacements or varenicline.^5–7

Bupropion is available in thrice-daily immediate release (IR), twice-daily sustained release (SR), and once-daily extended release (XL) dosing formulations that are bioequivalent.⁸ Both the XL and SR formulations have been shown to be efficacious when used to treat depression.³ However, patients using the XL formulation have been found to be more adherent to treatment compared with SR,⁹ likely due to XL’s once-daily regimen, lower price, and its pharmacokinetic profile (e.g., reduced plasma concentrations at bedtime), which may increase tolerability and decrease the chance of adverse events, in particular insomnia and nausea, compared to SR.¹⁰ While XL has become the most frequently prescribed bupropion formulation in the USA, to our knowledge there are no data published comparing the effectiveness and adverse event profiles of it with the SR formulation when used for either mood disorder treatment or smoking cessation.

In a naturalistic open label design, we compared the effectiveness and the adverse event profiles of two bupropion formulations (XL vs. SR), with and without nicotine replacement therapy (NRT), as prescribed to treat smoking among patients at a comprehensive cancer center. We hypothesized that the two formulations would be equivalent in terms of smoking cessation outcomes and adverse events.

METHOD

Participants

The present study included the 648 patients being treated or evaluated for cancer who were prescribed either bupropion XL or SR for smoking cessation out of the 6049 treated by The University of Texas MD Anderson Cancer Center Tobacco Treatment Program (TTP) from September 2006 to December 2017. Six additional patients died during smoking treatment and were excluded from these analyses, but the four patients who died after end-of-treatment (EOT) but before the end of adverse event monitoring are included. These 648 patients also included those who took NRT concomitantly with bupropion XL or SR. Our analyses were conducted as part of a chart review protocol, with waiver of informed consent, that was approved by the MD Anderson Institutional Review Board.

Procedures

Patients treated by the TTP were initially assessed during an in-person consultation that was 60 to 90 minutes in duration and, unless they refused it, they were typically provided with 10- to 12-weeks of pharmacotherapy either as monotherapy or in various combinations. Patients also received concurrent 6 to 8 sessions (30 to 45 minutes) of behavioral counseling, with most (over 95%) conducted by telephone. Treatment was free of charge and treatment plans were tailored to each patient’s smoking level, cessation history, preferences, and mental health needs following treatment pathways¹¹ consistent with National Comprehensive Cancer Network guidelines.¹² Noteworthy, for the first three years of the program (2006 – 2008), our medical staff mostly prescribed the SR formulation. When the XL formulation became available in generic formulations, our clinical staff shifted to primarily prescribing it for the convenience and cost reasons mentioned above, which accounts for the larger sample size for XL compared with SR.

Measures

At the initial consultation, we assessed for the symptoms or disorders of anxiety, panic, insomnia, and alcohol use with the Patient Health Questionnaire (PHQ),¹³ for major depression using the PHQ-9,¹⁴ for nicotine dependence severity using the Fagerström Test for Cigarette Dependence (FTCD; formerly the FTND),¹⁵ and for depressive symptoms using the Center for Epidemiologic Studies – Depression Scale (CES-D).¹⁶ The Alcohol Use Disorders Identification Test (AUDIT)¹⁷ was administered as an additional screening tool for alcohol-related problems. We used a timeline follow-back (TLFB) interview to capture patient-reported cigarettes per day (CPD) at consult and at follow-ups to calculate smoking abstinence.¹⁸ The primary outcome for this study was abstinence from combustible cigarettes, defined as 30-day point prevalence abstinence (30-day PPA; i.e., no smoking during the previous 30 days) at EOT (3 months post-medication initiation), obtained by self-report on the TLFB. In terms of safety, we examined the difference between the two bupropion formulations on adverse events (AEs; Common Terminology Criteria for Adverse Events [CTCAE], version 5.0)¹⁹ during treatment and out to 30 days after EOT.

Statistical Analysis

Baseline demographic differences were compared by formulation using chi-square and t-test analyses, with the Satterthwaite t-test approximation used for unequal variances. To examine whether bupropion XL and SR differed in abstinence outcomes, we analyzed 30-day PPA at EOT using chi-square analysis, with odds ratio greater than 1 indicating increased abstinence rates for the XL formulation compared to SR. We included all patients in these analyses using an intent-to-treat approach (i.e., missing smoking status was imputed as smoking) commonly used in the addiction literature,^20,21 because not all patients completed 3 months of smoking cessation treatment. We estimated all effects adjusting for the variables listed in Table 1, including age, race/ethnicity, PHQ-9 major depression status, CPD, and drinks per week, along with FTCD, AUDIT, and CES-D scores, and used multiple imputation averaging estimates across 10 imputed samples to account for missingness in these baseline predictors. For patients who had more than one treatment episode (i.e., one or more new consults separated by at least 12 months) in our clinic involving bupropion, we only included the first episode. We examined the impact of bupropion type on smoking abstinence outcome using three different models, including those who received bupropion XL or SR only, those who received bupropion and NRT, and both groups combined (i.e., all who received bupropion with or without NRT). For statistically nonsignificant comparisons, we ran tests of noninferiority to confirm whether no difference by formulation was likely (i.e., therapeutic equivalence). To compare the frequencies of adverse events by formulation, we conducted chi-square analyses.

Table 1.

Patient demographics, mental health, smoking behavior, and treatment characteristics.

Baseline Measure	Bupropion XL	Bupropion SR	Overall
	n (%)	n (%)	n (%)
Sample Size	454 (70.1%)	194 (29.9%)	648
Gender, Female	256 (56.4%)	92 (47.4%)*	350 (53.7%)
Race, White	350 (77.1%)	153 (79.9%)	503 (77.6%)
Race, Black	52 (11.4%)	21 (10.8%)	73 (11.3%)
Race, Other	52 (11.4%)	20 (10.3%)	72 (11.1%)
Depresseda	136 (30.0%)	73 (37.6%)	209 (32.2%)
Concurrent NRTb	265 (58.4%)	114 (58.8%)	379 (58.5%)
	Mean (SD)	Mean (SD)	Mean (SD)
Age	54.1 (12.0)	54.0 (11.5)	54.1 (11.8)
FTCD	3.9 (2.2)	4.6 (2.1)*	4.1 (2.2)
CPD	14.7 (9.8)	18.9 (14.0)*	15.8 (11.2)
AUDIT	4.9 (4.8)	4.1 (4.6)	4.8 (4.8)
Drinks Per Weekc	4.9 (8.6)	2.4 (5.2)*	4.2 (7.8)
CES-D	15.8 (11.5)	18.0 (12.7)*	16.4 (11.8)

Note:

^aMet Patient Health Questionnaire-9 criteria for Major Depressive Disorder or Other Depressive Syndrome.

^bReceived either the nicotine patch, episodic nicotine (e.g., gum, lozenge), or both during treatment, along with bupropion.

^cTotal drinks per week, in standard units. XL= extended release; SR=sustained release; NRT=nicotine replacement therapy; FTCD=Fagerström Test for Cigarette Dependence; CPD=cigarettes per day; AUDIT=Alcohol Use Disorders Identification Test; DPD=drinks per day; CES-D=Center for Epidemiologic Studies – Depression scale.

^*=significantly, p<.05.

RESULTS

Patient demographics

Baseline patient demographics, substance use, mental health, and treatment means and frequencies, by formulation, are shown in Table 1. A total of 648 patients were prescribed bupropion (454 were on XL and 194 on SR), and approximately 58% received concomitant NRT. The patients were largely White, averaging 54 years of age, with about a third meeting criteria for a depressive disorder on the PHQ-9. Those on XL reported significantly lower nicotine dependence (FTCD; t[559]=−2.91, p=0.0038), CPD (t[204.14]=−3.42, p=0.0008), and depression scores (CES-D; t[589]=−1.99, p=0.0469), but more drinks per week (t[461.17]=4.04, p<.0001), than those prescribed SR.

Abstinence Rates

The intent-to-treat 30-day PPA rates at EOT did not differ between those given bupropion XL (29.5%) and SR (30.9%; OR=0.86, 95% CI: 0.59, 1.26, p=0.438) for the pooled sample (i.e., with or without NRT). Likewise, the abstinence rates for those given only bupropion XL (34.9%) or SR (26.2%), without NRT, did not significantly differ (OR=1.52, 95% CI: 0.81, 2.83, p=0.188). The addition of a concomitant NRT did not make any significant difference in the odds ratios of abstinence between bupropion XL (25.7%) and SR (34.2%; OR=0.61, 95% CI: 0.37, 1.01, p=0.056).

Noninferiority Analysis

We further examined the therapeutic equivalence of bupropion XL to SR using noninferiority analysis. Our criterion for considering bupropion XL inferior was whether it demonstrated 50% or less of the efficacy of bupropion SR using the fixed-margin method, the criteria suggested by the FDA.²² To obtain the expected overall abstinence of bupropion SR, we used estimates of bupropion SR vs. placebo for abstinence by translating the relative risk (RR) ratio and the corresponding confidence intervals (RR = 1.64; CI: 1.52, 1.77) found in the 2020 Cochran review²³ to odds ratios (OR=1.78 (95% CI: 1.63–1.95). The lower limit of this CI, which is the closest limit to the null effect, was used to define the margin. This limit represents an OR of 0.61 (i.e., 1/1.63=0.61) of placebo compared to bupropion SR (i.e., a 39% decrease in odds). In this study, our margin for noninferiority consideration was 0.78 (50% of 0.61 on a logarithmic scale), meaning that for us to conclude noninferiority, the lower 95% CI limit of the bupropion XL vs. SR odds ratio must be > 0.78. Our analysis indicated that the odds ratio of bupropion XL vs. SR in the pooled sample (bupropion alone and bupropion plus NRT) was 0.86 (95% CI: 0.59–1.26). Since the lower limit of this CI was lower than the noninferiority margin of 0.78, noninferiority of bupropion XL to SR was not demonstrated. Similarly, noninferiority was not demonstrated for the comparison between bupropion treatments when NRT was also prescribed (OR=0.61, 95% CI: 0.37, 1.01). However, XL did demonstrate noninferiority in treatment efficacy compared to SR (OR=1.52, 95% CI: 0.81, 2.83) when excluding NRT (Figure 1).

Figure 1.

Analysis of noninferiority of bupropion XL to SR using odds ratios at the end-of-treatment. The noninferiority fixed margin (.78) is represented by the dotted line. Noninferiority was demonstrated only for the comparison without NRT.

Adverse Events

In terms of severe adverse events (SAEs), one patient taking bupropion XL reported mania and one person died between EOT and 30 days after EOT. Three patients taking SR died between EOT and 30 days after EOT. The rates of those reporting at least one AE did not statistically differ between those taking bupropion XL (30.3%) compared to those taking SR (34.9%; OR=0.81, 95% CI: 0.57, 1.16; p=0.25). Across bupropion types, the most common AEs included insomnia (5.75%), nausea (3.40%), anxiety (2.48%), headache (2.48%), and depression (2.35%). There were no statistically significant differences in the frequency of any specific AEs between the two formulations.

DISCUSSION

We found the once-daily bupropion XL dosing formulation to be therapeutically equivalent to the twice-daily SR formulation among patients receiving treatment for their smoking concurrently with behavioral counseling, in a naturalistic open-label study. There were no significant differences in intent-to-treat 30-day PPA rates at EOT between the formulations, regardless of whether the analyses included patients who were also prescribed NRTs. Equivalence was also demonstrated by noninferiority analysis when comparing the formulations without concomitant NRT. However, the noninferiority analyses were inconclusive for the sample in which NRT was combined with bupropion XL and SR and for the overall pooled sample (i.e., bupropion with and without NRT).

Our finding of therapeutic equivalence between bupropion XL and SR without concomitant NRT is consistent with the drugs’ pharmacokinetic profiles. Industry research reported that XL delivers an equivalent amount of bupropion, as measured by the area under the curve (AUC; i.e., blood plasma drug concentration as a function of time), but produces slightly lower peak serum concentration (C_max) than the SR formulation.¹⁰ XL also produces a longer time to peak serum concentration (T_max), which may increase its tolerability compared to SR (e.g., due to less insomnia).²⁴ While there have been concerns about the bio-equivalency of several XL generic formulation over the years,²⁵ recent work has found no differences between branded and three generic formulations in terms of pharmacokinetic profiles or efficacy in treating depression.²⁶

Despite the higher efficacy of varenicline or the combination of NRTs (patch plus an episodic such as gum or lozenge) compared to bupropion,²⁷ bupropion is still an effective treatment and has an important place in the armamentarium against tobacco use, especially for patients who either could not tolerate or did not respond to NRTs or varenicline. Bupropion therapy may be beneficial for addressing low energy,²⁸ difficulty with concentration,²⁹ concerns about reduced libido,³⁰ or for minimizing weight gain³¹ among those attempting to quit smoking. Additionally, as an atypical antidepressant, bupropion offers potential advantages because those with depressive disorders have a higher chance of being nicotine dependent, find it more difficult to quit smoking, and experience worse withdrawal symptoms compared with the general population.³² While bupropion has not been proven to provide added benefits for those who have mood disorders,³³ it has been found to produce better response for those who have a history of depression.³⁴

In terms of limitations, this was a non-randomized, unblinded naturalistic study, which precludes generalization beyond our sample. However, its strength was that it was conducted in in a real-world setting, among patients whose cancer diagnosis likely increased their motivation to quit compared to the typical “healthy” samples enrolled in clinical trials. Our findings warrant further replication, ideally in a double-blind placebo-controlled trial in another population setting to conclusively determine whether using XL formulation to treat smoking is equivalent in effectiveness to using the SR formulation. The lack of conclusive findings for two of the three tests of noninferiority we conducted could have been due to our limited sample size. A larger sample may have narrowed the ranges of our confidence intervals and thus produced more conclusive findings. Our findings could also be influenced by cohort effects, because XL was rarely prescribed in our clinic until generic formulations were available, at which time it became more prescribed than SR. There were differences in gender, nicotine dependence, cigarette consumption, alcohol consumption, and depression at baseline between the formulations which may have been exacerbated by a lack of treatment randomization. Other limitations were the lack of biochemical verification of smoking abstinence and the lack of treatment adherence data in our sample. This was primarily due to most patient sessions being conducted remotely, as only about 5% of the sessions were conducted in-person at our clinic. However, in a subsample of patients who did have in-person follow-up, we found concordance as high as 93% between patients’ self-report and measured expired carbon monoxide.³⁵ Finally, while XL has been found to produce greater treatment adherence,⁹ we did not have consistent adherence data and thus were unable to evaluate this in our sample.

Conclusions

We found no difference in smoking cessation effectiveness between the two bupropion long-acting formulations (SR vs. XL) as they produced comparable 30-day PPA rates in a sample of cancer patients who received concurrent behavioral counseling. Both formulations had similar adverse events profiles. However, these findings need to be replicated by additional randomized clinical trials in order for the XL formulation to be established as equivalent to the SR formulation and be generalized to other patient care settings and samples.