The Long-Term Effectiveness of the Anti-Obesity Medication Phentermine (LEAP) Trial: rationale, design, and baseline characteristics

Abstract

Background:

Current clinical practice guidelines support the long-term use of pharmacotherapy for obesity treatment. Historically, phentermine has been one of the most prescribed obesity medications (OMs), however, its long-term efficacy and safety have never been evaluated in a randomized trial as its market approval predates such requirements. Here we describe the design, rationale, and baseline characteristics for a 24-month, double-blind, randomized controlled trial evaluating the efficacy, cardiovascular risk, and safety of phentermine in adults with overweight or obesity.

Methods:

This multicenter trial will compare outcomes among participants randomized to phentermine 24 mg daily versus placebo for 24 months. All participants also receive an online lifestyle intervention. A total of 870 participants with body mass index of 27-44.9 kg/m² were randomized across six sites in North Carolina (2), Minnesota, Texas, and California (2). Primary outcomes are 24-month mean percent weight loss (efficacy), 24-month mean change in systolic blood pressure (cardiometabolic risk), and overall rates of adverse events and serious adverse events (safety). Secondary outcomes include changes in resting energy expenditure/resting metabolic rate, cardiac autonomic function measured using heart rate variability with electrocardiogram, and a self-reported measure of phentermine dependence.

Conclusions:

The safety and efficacy of long-term phentermine remains a pressing, unanswered question, particularly given its low cost and high availability when compared to newer OMs that are highly effective but often associated with significant costs. This study will impact clinical practice regardless of result – either providing evidence to support use of an available low-cost option or prioritizing the use of other OMs.

Trial Registration:

Clinicaltrials.gov, NCT05176626

1. INTRODUCTION

Over 40% of U.S. adults have obesity (body mass index (BMI) ≥30 kg/m²) [1], resulting in higher mortality risk and adding $150 billion annually to healthcare spending [2,3]. Lifestyle interventions focused on caloric reduction, increased physical activity, and behavioral counseling are considered first-line treatment for obesity [4]; however, most patients who engage in lifestyle modification alone have significant weight regain within 2 years [5,6]. Current practice guidelines recommend the long-term use of obesity medications (OMs) [7–9] as they can improve the proportion of people who achieve and sustain clinically significant weight loss.

OMs are indicated for adults with a BMI of 27-29.9 kg/m² with a weight-related complication or with a BMI ≥30 kg/m² [7–9]. Historically, the prescribing of OMs has been low – national cohorts consistently found only 1% of eligible adults were prescribed an OM [10,11]. Reasons for OM underuse have included concerns about adverse effects and prescribers’ frequent misperception of obesity as a lifestyle choice [12–14]. With the recent approval and popularity of new OMs, specifically incretin mimetics like semaglutide and tirzepatide, prescribing rates have increased, but these medications remain inaccessible to large portions of the population [15]. Limited insurance coverage and high out-of-pocket costs have exacerbated barriers to access for many patients who qualify for OMs [16,17].

Approved by the FDA in 1959, phentermine is a generic medicine [13] that historically accounted for most OM prescriptions [10,11]. Phentermine is a sympathomimetic amine that increases levels of norepinephrine (and to a lesser extent, dopamine and serotonin) in several hypothalamic nuclei, resulting in appetite inhibition [18]. Compared to new OMs, phentermine is an affordable option [7], typically costing less than $20 per month out-of-pocket. Few randomized controlled trials have tested the efficacy and safety of phentermine monotherapy beyond 3 months [13]. As a result, phentermine is only approved for short-term use, and its package labeling indicates that it should only be used as a short-term adjunct to lifestyle changes [19]. However, some clinical practice guidelines have recommended long-term phentermine use in patients without cardiovascular disease based on expert opinion and practice patterns [7–9]. Prescribers may be wary of phentermine due to concerns about cardiovascular risks including the potential to raise blood pressure and heart rate. These concerns have not been born out in large observational studies [20]; however, many prescribers remain hesitant to adopt long-term phentermine use given the lack of long-term efficacy and safety data [21]. Additional concerns center on the possible addictive potential of phentermine because of its structural similarities to amphetamine and its actions as a norepinephrine, dopamine, and serotonin-releasing agent. These concerns persist despite a paucity of evidence to support them in the existing literature [18,22].

The Long-Term Effectiveness of the Obesity Medication Phentermine (LEAP) trial was designed to address the unanswered question of whether long-term phentermine is an effective and safe obesity treatment. This manuscript describes the design, protocol, and baseline characteristics of this double-blind, placebo-controlled, randomized trial evaluating the efficacy, cardiovascular risk, and safety of phentermine among adults with overweight or obesity.

2. METHODS

2.1. Study Design

The LEAP trial is a 24-month Phase IV trial designed to test the effectiveness of phentermine combined with clinic visits and an online lifestyle intervention on body weight and blood pressure in adults who meet guideline-based criteria for OM therapy. The LEAP trial includes six clinical sites across the U.S.: two sites at Wake Forest University School of Medicine (WFUSM) in North Carolina, one site at the University of Texas Health Sciences Center in Texas, one site at HealthPartners Institute in Minnesota, and two sites at Kaiser Permanente Southern California.

2.1.1. Trial Organization and Oversight

Management is provided centrally by Clinical (CCC) and Data (DCC) Coordinating Centers housed at WFUSM. The CCC is responsible for protocol development, site communication, intervention implementation, regulatory coordination, and financial administration. The DCC is responsible for treatment allocation, receipt and processing of data, quality control, and statistical analyses. An unblinded Data and Safety Monitoring Board (DSMB) reviews and advises on study progress and monitors data on recruitment, quality, adherence/compliance, and outcomes.

The trial functions under the oversight of a single IRB at WFUSM who provided initial approval on February 7, 2022, and all other participating institutions entered into a reliance agreement. The trial was registered on clinicaltrials.gov [NCT05176626] and posted on January 4, 2022, before enrollment began in June 2022. The full study protocol, informed consent documents, and results will be published on clinicaltrials.gov upon study completion. All study procedures were performed in compliance with state laws and institutional guidelines and the privacy rights of human subjects have been observed.

2.2. Specific Aims

In an intent-to-treat fashion, with all participants provided an internet-based lifestyle intervention, the LEAP trial compares patients receiving phentermine monotherapy, provided as 8 mg tablets with a maximum tolerated dose up to 24 mg/day, or matching placebo for 24 months. The central weight-related efficacy hypothesis is that phentermine is associated with a 5% greater loss of initial body weight at 24 months compared to placebo. Based on prior research and observed differences in clinical trials of phentermine/topiramate combined and the hypothesized differences in body weight previously described, an additional hypothesis is that systolic blood pressure (SBP) is at least 3.6 mm Hg lower in the phentermine treatment arm at 24 months compared to placebo, based on the assumption that the benefits associated with sustained weight loss will counterbalance any acute increase in blood pressure that may be caused by phentermine.

2.3. Participants

Eligibility criteria (Table 1) were designed to identify a sample generally representative of the population of US adults with overweight or obesity for whom phentermine may be an appropriate long-term treatment option. An upper age limit of 70 years was selected because of the incidence of sarcopenia, osteopenia/osteoporosis, and frailty past that age [23,24]. Participation was limited to only English speakers as the scope and resources of the study prevented delivery of the lifestyle intervention and clinician counseling in other languages. Data from the electronic health record (EHR) and self-report were used to exclude persons with a history of cardiovascular disease, poorly controlled hypertension, cardiac arrhythmia, hyperthyroidism, pregnancy/breastfeeding, end-stage organ disease, or other conditions that may render weight loss or use of the study medication unsafe. The trial further excluded people who had recently gained or lost five percent or more of their body weight, those with a history of bariatric surgery, a prescription for any OM within the past year, or a prescription for phentermine or a phentermine-containing drug in the past two years.

Table 1.

Eligibility Criteria for the LEAP Trial

Inclusion Criteria
Demographics	English-speaking men and women 18-70 years of age who reside within reasonable geographic vicinity of a LEAP clinical site to make monthly travel for drug dispensing or clinic visits feasible
Body Mass Index (BMI)	30-44.9 kg/m2 or BMI 27-29.9 kg/m2 with weight related comorbidity defined as hypertension, prediabetes, type 2 diabetes mellitus, dyslipidemia, nonalcoholic fatty liver disease, obstructive sleep apnea, osteoarthritis, low back pain, or gastroesophageal reflux disease
Reproductive Status	For females of reproductive potential: use of effective contraception for at least 1 month prior to randomization and agreement to use such a method during study participation and for an additional 8 weeks after the end of study drug administration.
Internet Connectivity	Has a smartphone or other device with regular internet access
Willingness to comply with trial requirements	Potential participants must be interested in and willing to lose weight, able to take oral medication, and willing to adhere to the clinical visit schedule for the trial and lifestyle-based treatment regimen throughout the study as recommended by the study clinician
Exclusion Criteria
Weight loss/weight instability	Currently involved in a supervised program for weight loss; use of phentermine, phentermine-containing medication, or anti-obesity medication with similar mechanism of action to phentermine (e.g., phendimetrazine or diethylpropion) in the previous 24 months; use of any non-phentermine-containing medications prescribed for weight loss in previous 12 months or unstable dosing medications that could be used for weight loss; history of prior procedure for weight control including bariatric surgery, devices; documented or self-reported weight change (gain or loss) of more than 5% of current body weight in the past 3 months
History of cardiac abnormalities or cardiovascular, valvular, cerebrovascular, or peripheral arterial disease	Clinical history of coronary artery, cerebrovascular or peripheral arterial disease including myocardial infarction, unstable angina, revascularization, stroke/TIA, carotid intervention, claudication; congestive heart failure; history of any cardiac arrhythmia (including atrial arrhythmias (atrial fibrillation & atrial flutter), supraventricular tachycardia, ventricular arrhythmias, sudden cardiac arrest, severe conduction delays such as heart block, and anyone with a pacemaker or implanted cardiac defibrillator)
Uncontrolled or poorly controlled hypertension or elevated heart rate	Blood pressure > 149/94 mmHg or heart rate >110 bpm. Potential participants can be re-screened after control has been achieved.
Active/currently treated hyperthyroidism or poorly controlled / under-treated hypothyroidism	Self-reported treatment for hyperthyroidism or evidence of such in the electronic medical record; baseline laboratory value TSH >10 mIU/L
Pregnancy	Pregnancy, breast feeding, or planning pregnancy within 24 months; seeking or in active treatment for infertility.
Glaucoma	Clinical history of glaucoma or self-reported risk from ophthalmologist
History of heavy alcohol use or substance use disorder	Heavy alcohol use within the last 6 months (men: more than 4 drinks on any day or more than 14 drinks/week; women: more than 3 drinks on any day or more than 7 drinks/week); history of substance use disorder or active use of illicit substances within the last 12 months
Other chronic diseases or medical history which may make use of study drug unsafe	End-stage renal disease on dialysis or CKD class IV or higher (eGFR <30); cirrhosis or symptoms of liver failure in the last 2 years; severe pulmonary disease requiring supplemental oxygen; history of cancer other than non-melanoma skin cancer in the past 5 years; history of organ transplantation
Medication Usage	Use of a drug in the monoamine oxidase inhibitor class, currently or within the last 14 days; use of oral corticosteroids more than 5 days/month in the last 3 months; use of any stimulant medications in previous 12 months
Depression, anxiety, or other mental health concerns	Elevated depressive symptoms defined as a PHQ score greater than 7 during screening; uncontrolled anxiety symptoms defined as a score of 10 or higher on the GAD7; hospitalization for mental illness in the last 24 months; diagnosis of dementia or serious mental illness (e.g., schizophrenia, bipolar disorder, severe depression) Binge Eating Disorder, Bulimia or Anorexia Nervosa diagnosis or treatment within the last 2 years
Recent history of eating disorders	Binge Eating Disorder, Bulimia or Anorexia Nervosa diagnosis or treatment within the last 2 years
Other reasons	Conditions likely to interfere with participation and acceptance of randomized assignment, including the following: inability/unwillingness to give informed consent, known allergy or intolerance to phentermine or phentermine-containing medication; cessation of nicotine-containing products less than 6 months prior to baseline visit; plans to move outside the area in the next two years; unable to make changes to diet (e.g., severe food allergies or intolerances; medically necessary aspects of diet incompatible with intervention); already participating in another research study that includes lifestyle changes and/or study medication or has participated in such a study within the last 12 months

TSH = Thyroid Stimulating Hormone, CKD= chronic kidney disease, eGFR = estimated Glomerular Filtration Rate, PHQ = Patient Health Questionnaire; GAD7 = General Anxiety Disorder-

2.4. Recruitment

Using a computable phenotype query, lists of potentially eligible patients at each clinical site were identified through the EHR with all sites using software from Epic Systems Corporation (Verona, WI). Depending on the IT resources available at the site and local regulations, these lists were used to send messages through the patient messaging portal or by email. These messages invited potential participants to learn more about the study by watching a recruitment video and included a link to complete an eligibility self-screener in the site’s local Research Electronic Data Capture (REDCap) application [25,26].

2.5. Screening and Informed Consent

Individuals who completed the self-screener and remained eligible were asked to provide contact information for a comprehensive telephone screening interview. Informed consent was then administered electronically, and a 7 to 14-day behavioral run-in was initiated during which potential participants demonstrated facility with the technical components of the online intervention.

Following run-in, potential participants were invited for an in-person baseline visit where height, weight, and blood pressure were measured to verify eligibility. An electrocardiogram (ECG) was collected to identify potential signs of active ischemia, cardiac arrhythmia or other findings of immediate concern. Fasting blood samples were collected to evaluate kidney, liver, and thyroid function, along with assessing electrolytes and blood count. A pregnancy test was administered for individuals of childbearing potential. A clinician completed a physical examination and reviewed medical history and medications. Following review of all baseline and screening data, participants were scheduled for a randomization visit. This process is depicted in Figure 1.

Figure 1.

Flow of participation in LEAP Trial

2.6. Enrollment/Randomization

Before randomization, participants signed a behavioral contract to reinforce expectations related to study participation. Participants were then randomized by the study website to either study drug or placebo in a 1:1 ratio. Randomization was stratified by site using a permuted block design with random block sizes [27]. Participants, and all personnel at the clinical sites and CCC are blinded to treatment assignment. After randomization, all participants received lifestyle counseling from a study clinician, orientation to the online application and wireless scale, and were dispensed a 30-day supply of study drug with instructions for up-titration.

2.7. Intervention Procedures

Participants randomized to active drug in LEAP are provided with phentermine hydrochloride 8 mg scored tablets, a commercially available formulation of the drug procured from KVK Tech (Newtown, PA). This formulation was selected to minimize potential side effects and to maximize safety and tolerability. Participants randomized to placebo receive tablets manufactured by KVK Tech that consist of cellulose and corn starch with the same characteristics of the active drug, including size, shape, weight, and sensory perceptions. All participants were instructed to begin taking 8 mg daily and titrate upwards weekly during the first month to a maximum dose of 24 mg daily. Refills of study medication are dispensed at the site at 30 or 90-day intervals, depending on local/state policy for controlled substances.

In addition to study drug, all participants are provided with clinician-directed lifestyle change counseling, medical monitoring, and a 24-month subscription to the WW^™ online lifestyle intervention app and wireless scale. Visits with obesity medicine clinicians occur throughout the 24-month intervention period: monthly during months 1-3 as study drug is initiated and up-titrated and then every 3 months through 24 months for assistance with medication adherence, monitoring for side effects and adverse events, and lifestyle counseling.

2.8. Participant Retention Activities

All participants receive compensation for attending assessment visits, up to $250 over the course of the study. Telephone, email, and text reminders are sent to enhance attendance and rescheduling of visits. Participants also receive holiday and birthday cards from the study, along with certificates recognizing study anniversaries and completion. A centralized retention team has also assisted with visit rescheduling and follow-up with participants who face barriers to continued engagement.

2.9. Assessment

Assessment visits (Table 2) occur concurrently with the clinic schedule described above. Weight is measured in light clothing without shoes on a digital scale. Blood pressure is measured using a standardized protocol based on the SPRINT trial [28] with the Omron HEM907XL (Omron Healthcare, Kyoto, Japan). Following a 5-minute rest, 3 consecutive measurements obtained at 1-minute intervals are collected and averaged. All blood samples are collected following an 8-hour fast and analyzed at the regional LabCorp facility nearest the site. Hemoglobin A1c (HbA1c) and lipids are collected at baseline, 6, 12, and 24 months. HbA1c is analyzed using the Roche Tina-quant^® assay and lipids are assayed via enzymatic methods. Resting energy expenditure is measured via indirect calorimetry in a fasting state per the manufacturer’s recommended procedure using the ReeVue (KORR Medical Technologies, Salt Lake City, UT) at baseline, 6, and 24 months. Heart rate variability (HRV) is measured via ECG using the MAC 2000 (GE Healthcare Technologies, Inc., Chicago, IL) and assessed using two time-domain measures: standard deviation of all normal-to-normal R-R intervals (SDNN) and root mean square of successive differences between normal-to-normal R-R intervals (rMSSD) [29–31]. Potential dependence on study drug is measured using the Severity of Dependence Scale, a validated measure of psychological dependence used previously for illegal drugs and prescription drugs including phentermine [22, 32, 33].

Table 2.

Schedule of Data Collected in LEAP

Measure	Visit (Months)
	BL	RZ	1	2	3	6	9	12	15	18	21	24	25
Concomitant Medications	x	x	x	x	x	x	x	x	x	x	x	x
Survey Measures	x				x	x		x				x	x
Weight	x	x	x		x	x	x	x	x	x	x	x
Resting Metabolic Rate	x					x						x
Blood Pressure & Heart Rate	x	x	x		x	x	x	x	x	x	x	x
Waist Circumference	x		x			x		x				x
ASCVD Score	x				x	x		x				x
Baseline Safety Labs	x
HbA1c, Lipids	x					x		x				x
ECG	x				x	x		x				x
Adverse events		x	x	x	x	x	x	x	x	x	x	x	x

ASCVD = Atherosclerotic Cardiovascular Disease 10-year risk score, ECG = Electrocardiogram, HbA1c = Hemoglobin A1c

2.9.1. Primary Outcomes

This trial has co-primary outcomes– mean percent weight loss relative to baseline and SBP- at 24 months. Whenever possible, both measures will be obtained from in-person visits using the measurement protocols described above. However, because of the length and intensity of the trial, some participants may not be willing or able to come in for later study visits to all sites. This could result in informative censoring if participants with worse outcomes, or those who have greater barriers to attending in-person visits, are selectively lost to follow-up before 24 months. To ensure as complete follow-up as possible, with our primary safety and efficacy outcomes assessed on the maximum number of randomized participants, we will maintain the option of using additional avenues for gathering weight and blood pressure data. Specifically, we will explore the use of remotely collected weight from participants’ wireless scales as well as weight and blood pressure measured at home visits by trained coordinators for individuals who cannot physically come into their site. Additionally, for participants who otherwise are lost to follow-up, we will leverage weight and blood pressure data collected at primary care visits within participating systems’ EHRs, provided that the readings are from within the date range where a study visit should have taken place. When these outcome measures are assessed using one of these means, the method of assessment will be captured and included as a covariate in statistical models.

2.9.2. Secondary Outcomes

Secondary hypotheses for the trial compare the phentermine treatment group to placebo at 24 months for resting energy expenditure, HRV, and phentermine dependence. In addition, the primary outcomes of mean percent weight loss and mean SBP will be compared at 3, 6, 12, and 18 months of follow-up.

2.9.3. Other Outcomes

Other exploratory outcomes collected focus on drivers of energy balance and metabolism: a visual-analog scale of hunger [34] and self-reported physical activity using the Internal Physical Activity Questionnaire Short-form [35]; cardiometabolic risk factors including HbA1c and lipids, waist circumference [36], atherosclerotic cardiovascular disease (ASCVD) risk score using http://tools.acc.org/ASCVD-Risk-Estimator/, and survey measures on social support for lifestyle behaviors [37], weight-loss self efficacy [38], weight control strategies [39], food addiction [40], quality of life [41,42], and well-being [43].

2.10. Statistical Analysis

2.10.1. Primary Outcomes

The primary analyses of percent weight loss and SBP will use longitudinal models based on generalized least squares utilizing an autoregressive moving average correlation structure. Models will adjust for baseline (SBP only), clinic site, mode of data collection (research visit, home visit, EHR data), baseline age, sex, and race/ethnicity, given evidence of variability with respect to these factors for weight loss response and blood pressure control [44, 45]. Follow-up time will be modeled discretely, with percent weight loss and SBP measured at months 1, and 3 and then every three months thereafter throughout the total follow-up of 24 months. A multiplicative interaction term between treatment group and each post-randomization time point will be included in the model to allow evaluation of the primary inferential estimand, which will be a contrast reflecting the mean group difference at 24 months of follow-up. Of secondary interest will be mean contrasts estimating treatment group differences at 6, 12, and 18 months of follow-up. All primary comparisons will be conducted consistent with the intent to treat principle, and all tests of hypotheses and reported P values will be two-sided. We will utilize a Bonferroni-adjusted alpha level of 0.025 for the primary analyses of percent weight loss and SBP at 24 months.

2.10.2. Secondary Outcomes

Analyses for the secondary outcomes of resting energy expenditure and HRV measured via ECG will be based on analogous longitudinal models utilizing generalized least squares, including baseline values, clinical site, age, sex, and race/ethnicity as covariates. The primary inferential target will be mean contrasts at 24 months of follow-up, with secondary interest in treatment group difference at 3 (HRV only), 6 and 12 months of follow-up (resting energy expenditure is only measured at 6 and 24 months). For the patient reported Severity of Dependence scale, we will utilize generalized linear mixed models based on the beta-binomial distribution, which have been suggested by several groups to address the typical distributional features of patient reported outcomes [46,47]. To control for multiplicity amongst the secondary endpoints, we will utilize the Holm sequential procedure across the 24-month comparisons for these endpoints [48].

2.10.3. Trial Estimands, Sensitivity Analyses and Planned Subgroup Analyses

We will conduct sensitivity analyses targeting a trial product estimand that examine the effect of participants initiating other OMs during the trial [49]. These analyses will use the same longitudinal framework described for the primary analyses but will also include time-dependent treatment indicators for whether the participant is still receiving study medication and whether they are taking a different OM medication. We also plan to conduct to analyses of subgroups of interest including age group at baseline (<45 years, 45-59 years, ≥60 years), sex, race/ethnicity (White, Black, Hispanic, Other), baseline BMI (<35 versus ≥ 35 kg/m²), baseline blood pressure control (<130/80 mm Hg versus ≥130/80 mm Hg) [50], and history of diabetes mellitus at baseline.

2.10.4. Sample Size and Statistical Power

2.10.4.1. Power Calculation Assumptions and Estimates

We assumed a mean contrast estimating group differences at 24 months of follow-up, a Bonferroni-adjusted alpha level of 0.025, and cumulative loss to follow-up rates between 10% to 25% based on experiences in prior behavioral trials of weight loss interventions [51,52].

For weight change, we assumed that percent weight loss in the placebo group would be ≤3% at 24 months based on the CONQUER trial [53]. We estimated that, with 1000 participants we would have 94% power to detect a 1.75% mean difference in percent weight loss at 24 months (for example, 3.25% in the phentermine group versus 1.5% in the placebo group) assuming a 10% dropout rate, with power decreasing to 90% assuming a 25% dropout rate. With a minimum clinically meaningful difference between groups specified as 5% weight loss, however, our sample size for LEAP was primarily determined by the SBP outcome, as detailed below.

For SBP, we initially considered a range of scenarios for the trajectory of mean SBP in each treatment group, modeled after our previous observational EHR cohort estimates [20]. In that study, long-term continuous users of phentermine (defined as usage >365 days) had an estimated mean difference in SBP of 3.3 mm Hg (95% CI: 5.9 to 0.8 mm Hg) at 24 months. We also used data from the SPRINT trial [54] [55] to estimate the variance-covariance matrix of repeated SBP measurements. Based on these assumptions, we initially estimated 97% power to detect a 3.6 mm Hg mean difference assuming a 10% dropout rate, decreasing to 94% with a 25% rate. These estimates were intentionally conservative, using the baseline standard deviation (SD) estimate for SBP from SPRINT (~15 mm Hg). In follow-up, the SBP SD in SPRINT was about 15-20% lower (13.0 mm Hg at 12 months, 11.9 mm Hg at 24 months). However, for LEAP, participants could not have highly uncontrolled BP to be eligible for the trial (SBP<150/95 mm Hg and heart rate≤ 110 bpm with a baseline SD in LEAP = 11.8 mm Hg). As a result, the required sample size for detecting a 3.6 mm Hg difference is lower than originally anticipated.

In December 2023, NHLBI and the DSMB approved reducing the sample size from 1,000 to 900 participants. Even assuming a conservative follow-up SBP SD of 13 mm Hg at 24-months, this sample size would provide 95% power to detect a 3.3 mm Hg difference in SBP at 24-months assuming 10% dropout, or 91% power assuming 25% dropout. Similar calculations for percent weight loss indicate that 900 participants would still provide 97% power to detect a 2.25% mean difference in percent weight loss with 10% dropout, and 94% power for 25% dropout, still well below a clinically meaningful difference.

3. RESULTS

The LEAP trial was funded by a UG3/UH3 cooperative agreement from NHLBI in September 2021. Protocol development, drug procurement, and site training occurred in late 2021-early 2022. Screening and recruitment activities began in May 2022, and the first participant was randomized in June 2022. Enrollment was slower than initially anticipated and was finalized in October 2024 based upon a sponsor requirement to complete data collection for all participants by Fall 2026. Data analysis will commence in October 2026 with a goal of having primary results submitted for publication in December 2026.

In brief, 8498 potential participants completed self-screening surveys across the 6 sites, of whom 5884 were eligible for telephone screening. Over half of those (n=3303) completed telephone screening and 2159 were eligible to proceed to the run-in phase and medical record review. At the conclusion of the run-in phase, 1362 potential participants were eligible for a baseline visit and 975 of those potential participants were eligible for randomization. A total of 870 participants were ultimately randomized (see Figure 2). A separate manuscript is being developed to detail the recruitment methodology and screening results in LEAP.

Figure 2.

CONSORT Diagram

The baseline characteristics for the randomized sample are shown in Table 3. To compare the LEAP participants to those enrolled in comparable OM trials, as well as the general US population eligible for OM treatment, we present LEAP participant characteristics alongside those of participants in the CONQUER trial [53] and the National Health and Nutrition Examination Survey (NHANES). For the NHANES comparators, we used data from the August 2021-August 2023 dataset [56], restricted to those participants 18-70 years of age with a BMI between 27.0 and 44.9 kg/m² and not pregnant at the time of the exam.

Table 3.

Baseline Characteristics of LEAP and CONQUER trials and a nationally representative sample from NHANES

Characteristic	LEAP	CONQUER	NHANES
	N = 870	N = 2487	N = 113,473,986a
Age, years, mean (SD)	48.8 (12.1)	51.1 (10.4)	45.2 (14.4)
Sex at birth, N (%)
Male	236 (27.1%)	750 (30.2%)	58,539,974 (51.6%)
Female	634 (72.9%)	1,737 (69.8%)	54,934,012 (48.4%)
Race/Ethnicity, N (%)b
Hispanic	149 (17.2%)	-	22,931,444 (20.2%)
Non-Hispanic Asian	19 (2.2%)	-	4,721,440 (4.2%)
Non-Hispanic Black	153 (17.6%)	-	14,335,602 (12.6%)
Non-Hispanic White	526 (60.7%)	-	64,214,401 (56.6%)
Other race – including multi-racial	20 (2.3%)	-	7,271,099 (6.4%)
Unknown	3		0
Weight, lbs, mean (SD)	223.2 (37.6)	227.2 (39.5)	207.0 (36.2)
Body Mass Index, kg/m2, mean (IQD)	35.7 (4.1)	36.6 (4.5)	32.9 (4.6)
Waist circumference, cm, mean (SD)	113.2 (11.7)	113.2 (12.3)	107.9 (11.9)
Systolic Blood Pressure, mm Hg, mean (SD)	120.0 (11.8)	128.4 (13.5)	120.7 (15.9)
Diastolic Blood Pressure, mm Hg, mean (SD)	80.5 (7.8)	80.6 (9.1)	77.4 (10.7)
Heart rate (beats per min), mean (SD)	71.9 (9.9)	72.3 (10.0)	73.1 (11.9)
Total cholesterol, mg/dL, mean (SD)	194.0 (36.9)	204.2 (40.3)	193.2 (41.9)
LDL cholesterol, mg/dL, mean (SD)c	119.0 (31.6)	123.0 (34.8)	-
HDL cholesterol, mg/dL, mean (SD)	52.2 (13.6)	50.3 (15.5)	49.5 (12.4)
Triglycerides, mg/dL, mean (SD)c	126.4 (66.6)	159.4 (74.5)
Fasting glucose, mg/dL, mean (SD)	91.0 (18.5)	106.3 (22.3)	111.7 (38.4)
Glycated hemoglobin, %, mean (SD)	5.7 (0.5)	5.9 (0.8)	5.8 (1.2)

^aSample in NHANES restricted to participants 18-70 years of age with a BMI between 27.0 and 44.9 kg/m² and not pregnant at the time of the exam.

^bData on ethnicity not reported in CONQUER.

^cNot measured in NHANES.

CONQUER = Phase III Randomized, Double-Blind, Placebo Controlled Multicenter Study to Determine the Safety and Efficacy of VI-0521 in the Treatment of Obesity in Adults With Obesity-Related Co-Morbid Conditions, HDL = High-density lipoprotein, IQR = Interquartile Range, LDL = Low-density lipoprotein, LEAP = Long-Term Effectiveness of the Anti-Obesity Medication Phentermine, NHANES = National Health and Nutrition Examination Survey, SD = Standard Deviation

In LEAP, 634 (72.9%) randomized participants are female and 195 (22.4%) reported a race other than White. Additionally, 149 (17.1%) reported being of Hispanic or Latino ethnicity. The mean BMI was 35.7 kg/m² at baseline with a SD of 4.1 kg/m². When compared with an NHANES sample of adults who would qualify for an OM prescription, the LEAP population has a slightly lower percentage of Hispanic participants, a higher percentage of non-Hispanic Black participants, and a lower representation of males. The LEAP population is also slightly older. The percentage of females and population age are similar in CONQUER and LEAP. Although race and ethnicity data were reported differently for CONQUER, 86% of the trial population was designated as White, compared to just over 60% for LEAP. Blood pressure and cholesterol values are similar across populations, while fasting glucose is markedly lower in the LEAP population despite HbA1c being similar to the reference populations.

4. DISCUSSION

Although phentermine has been available for over 60 years and is frequently prescribed, evidence to support its long-term use is largely from observational data [18] that suggest safety and effectiveness in low-risk populations [20]. No randomized trial has tested phentermine monotherapy as a long-term treatment for adults with obesity; thus, the LEAP trial is being conducted to address this evidence gap.

Compared to prior OM trials, the LEAP population represents a step toward recruiting a sample that closely resembles the population for whom this treatment would be recommended. When compared to the recent STEP trials of semaglutide in the United States in adults without Type II diabetes, for example, the LEAP population includes more males and is more racially and ethnically diverse [57–60]. Compared to the SURMOUNT trials of tirzepatide, LEAP had a higher proportion of non-White participants and a higher share of older adults [61–63]. Some of the eligibility criteria implemented in LEAP may still reduce generalizability, specifically the inclusion of only English-speaking participants and those with internet access. The opt-in recruitment/screening process implemented in LEAP resulted in a study population highly motivated to lose weight. Given the double-blind, placebo-controlled design, maintaining high levels of adherence to medication and retention for data collection visits could prove challenging. Participants who feel dissatisfied with their own progress or suspect they are on placebo may seek out other weight management options, especially newer OMs. Importantly, the LEAP trial does not test all available doses of phentermine, and it is possible that higher doses might carry different levels of weight loss efficacy and potential for adverse effects. Additionally, this trial will not be able to comment on the comparative effectiveness of phentermine versus other medications for long-term use, the optimal timing of phentermine use (e.g. as a first line agent for weight loss or as a long-term maintenance tool), or whether certain subgroups of patients are more likely to respond to the drug. Depending on the results of LEAP, these could be fruitful areas for future investigations.

5. CONCLUSIONS

A focus on lifestyle interventions in preceding decades has done little to curb the increasing prevalence of obesity, suggesting that pharmacologic approaches must also be considered [18]. The arrival of incretin mimetics on the market and the subsequent media frenzy has made OM therapy sought after, but high costs [64], limitations in insurance coverage [65], patient and clinician misperceptions [66], and supply shortages for these medications [67] make it unlikely that they will be able to fully address the scope of the obesity epidemic. Should the long-term use of phentermine prove safe and effective, it could represent an affordable, accessible and effective alternative to newer, more expensive medications on the market, even when paid for directly by patients.

LEAP Supplementary Acknowledgements

LEAP Clinical Coordinating Center: Wake Forest University Health Sciences, Winston-Salem, NC: Kristina Henderson Lewis (MPI), Jamy Ard (MPI), Katy Martin-Fernandez (Co-I), Mara Vitolins (Co-I), Caroline Blackwell Young (Project Manager), Sally Eagleton (Central Retention), Ha Sprinkle (Business Manager), Claudine Curran (Grants Administrator)

LEAP Data Coordinating Center: Wake Forest University Health Sciences, Winston-Salem, NC: Nicholas Pajewski (PI), Byron Jaeger (Co-I), Catherine Price (Safety Officer), Emily Rives (Project Manager), KaShawna Guy (Project Manager), Courtney Simmons (Biostatistician), Brian White (Biostatistician), Darren Harris (Programmer), Brad Bouquio (Programmer)

LEAP ECG Reading Center: Epidemiological Cardiology Research Center (EPICARE), Winston Salem, NC: Elsayed Z. Soliman (PI), Oguz Akbilgic (Co-I), Yabing Li (Research Associate), Ibrahim Karabayir (Co-I), Takeki Suzuki (Co-I), Mohammed A. Mostafa (Research Scholar), Lisa Keasler (Project Manager), Semseddin Moldibi (Programmer), Stacey Belton (Grants Administrator)

LEAP Drug Distribution Center: VA Cooperative Studies Program Clinical Research Pharmacy Coordinating Center: Elliott Miller (PI), Barbara Del Curto (Project Manager)

Atrium Health Wake Forest Baptist Weight Management Center Clinical Sites, Winston-Salem and Greensboro, NC: Jessica Bartfield (Co-I), Terrika Stewart Simmons (Obesity Clinician), Elizabeth Flynt (Obesity Clinician), Sophia Ali (Obesity Fellow), Priyanka Nellori (Obesity Fellow), Priya Patel (Obesity Fellow), Kathleen Ruddiman (Obesity Fellow), Beatriz Ospino-Sanchez (Project Manager), Angelina Pack (Study Coordinator), Charlie Robinson (Study Coordinator), Kaleb Sizemore (Study Coordinator), Josh Weaver (Study Coordinator), Sarah Buff (Study Coordinator), Emma Jensen (Student Intern), Mai Soliman (Student Intern), Dylan Skelton (Student Intern)

HealthPartners Institute Clinical Site, Saint Louis Park, MN: Stephanie Hooker (PI), Karen Margolis (Co-I), Jennifer Vesely (Obesity Physician), Omar Fernandes (Obesity Clinician), Carol Fox (Obesity Clinician), Maureen Busch (Project Manager), Shelly Cook (Study Coordinator), Amy Norby (Study Coordinator), Karsen Aune (Research Specialist), Center for Evaluation and Survey Research (CESR; Recruitment)

Kaiser Permanente Southern California Clinical Sites, Downey Medical Center and West LA Medical Center, Los Angeles, CA: Deborah Young (PI), Amanda Velazquez (Obesity Physician), Melanie Dewar (Obesity Physician), Jocelyn Preston (Obesity Clinician), Judy Lee (Obesity Clinician), Portia Keough (Obesity Clinician), Marie Doan (Obesity Clinician), Galina Inzhakova (Project Manager), Melissa Preciado (Project Manager), Julliane Bacerdo (Study Coordinator), Samantha Quinones (Study Coordinator), Martha Cedeno (Study Coordinator), Kevin Mejia (Study Coordinator)

University of Texas Health Sciences Center Clinical Site, Houston, TX: Deborah B. Horn (PI), Thuyvan Hoang (Obesity Physician), Connie Klein (Obesity Clinician), Angielyn Rivera (Project Manager), Jasmine Coleman (Study Coordinator), Alexandrya Weston (Study Coordinator), Sofia Colon (study coordinator), Bernadette Tumaliuan (study coordinator), Saige McCann (study coordinator)

Johns Hopkins Division of General Internal Medicine, Baltimore, MD: Kimberly Gudzune (PI), Jeanne Clark (Co-I)

LEAP Data and Safety Monitoring Board: Steven R. Smith (Chair, 2021-2024), Daniel Lackland (Chair 2024-present), Cheryl Anderson, Jeffrey Szychowski, Kevin Thomas, Beverly Tchang