Use of Medication to Improve Weight Loss in Suboptimal Early Responders to Behavioral Treatment

Phase 4

Completed

Conditions: Obesity

Interventions: Behavioral: Behavioral Treatment
Drug: Placebo
Drug: Phentermine 15 MG

Registration Number: NCT03779048

Lead Sponsor: University of Pennsylvania

Brief Summary: This is a two-phase study. Phase 1 will evaluate obesity-related behavioral and biological characteristics as potential predictors of response to behavioral treatment (BT) for weight loss. Phase 2 is a double-blind, placebo-controlled, RCT to test whether adding weight loss medication to BT improves 24-week weight loss, as compared to BT with placebo, in subjects identified as having suboptimal early weight loss after 4 weeks of individual behavioral weight control. All participants, regardless of their early weight loss, will receive the same BT program of diet, physical activity, and behavior therapy for weight loss for an additional 24 weeks (28 total weeks of treatment).

Detailed Description: Subjects will be a total of 150 adults, aged 21-70 years, with a body mass index (BMI) of 31 kg/m2 or above (28 kg/m2 with an obesity-related comorbidity). In phase 1, eligible subjects will complete questionnaires and an in-person baseline assessment of obesity-related behavioral characteristics (satiety, hunger, the relative reinforcing value of food \[RRVfood\], and impulsivity \[delay discounting\]), neuropeptides, and gastric emptying. After this baseline assessment, participants will begin an initial 4-week behavioral treatment (BT) "run-in" delivered individually in 20-30 minute weekly sessions (delivered virtually).

The primary goal of phase 1 will be to evaluate baseline satiety, postprandial change in GLP-1, and gastric emptying as predictors of percent weight loss after 4 weeks of BT. We will also examine whether these variables predict categorization as a suboptimal early responder to BT (e.g., \<2.0% loss; co-primary outcome).

Secondary endpoints of phase 1 are percent weight loss from the start of the BT run-in (week -4) to randomization (week 0) and categorization as a suboptimal early responder, as predicted by additional behavioral characteristics (hunger as measured by VAS ratings, RRVfood as measured using a computer task, and impulsivity as measured using a delay discounting computer task) and neuropeptides (higher fasting ghrelin, lower fasting leptin, and lower postprandial changes in insulin and PYY).

In phase 2, suboptimal early responders (based on weight loss during the BT run-in) will be randomly assigned to 24 weeks of: 1) BT plus placebo (BT+P); or 2) BT plus medication (BT+M; phentermine 15.0 mg). Both treatment groups will continue to attend 20-30 minute individual BT sessions (delivered virtually), weekly for the first 12 weeks and every other week for the last 12 weeks (total of 18 visits). Both treatment groups will also take once daily study medication (placebo or phentermine 15.0 mg) for the duration of the intervention period. Early BT responders identified during the run-in will receive the same 24-week BT program, but will not receive study medication or be included in the randomized trial.

The assessments administered at baseline - questionnaires, including behavioral testing, blood draws, and measurements of body weight - will be repeated at randomization (week 0) and at week 24.

The primary endpoint of phase 2 is change in body weight (i.e., % reduction in initial weight), as measured from randomization to week 24, among suboptimal early responders assigned to BT+P vs. BT+M. A randomized sample size of 50 non-responders (25 per group), assuming a 20% attrition rate, will give us 81.5% power to detect between-treatment group differences at week 24 of 4.5% (effect size: d = 0.82).

Secondary endpoints of phase 2 will include change in body weight in kg from randomization to week 24, as well as the portion of suboptimal early responders who achieve a post-randomization loss of ≥ 5% and ≥ 10% of initial body weight. We will also examine differences between suboptimal early responders treated with BT+M vs. BT+P in changes in hunger, satiety, the reinforcing efficacy of food, and impulsivity between randomization and week 24. A comparison will also be made in percent weight loss from randomization to week 24 between suboptimal early responders treated with BT+M and early responders treated with BT alone.

If you are interested in participating in this study, information and a link to contact the research team can be found here: https://clinicalresearch.itmat.upenn.edu/3XOX/ or you can call us at the numbers listed below.

Recruitment & Eligibility

Status: COMPLETED

Sex: All

Target Recruitment: 147

Inclusion Criteria

BMI ≥ 31 kg/m² (or 28 kg/m2 with obesity-related comorbidity)
Age ≥ 21 years and ≤ 70 years
Eligible female patients will be:
- non-pregnant, evidenced by a negative urine pregnancy test
- non-lactating
- surgically sterile or postmenopausal, or they will agree to continue to use an accepted method of birth control during the study. Acceptable methods of birth control are: hormonal contraceptives; double barrier method (condom with spermicide or diaphragm with spermicide); intrauterine device; surgical sterility; abstinence; and/or postmenopausal status (defined as at least 2 years without menses).
Subjects must:
- have a primary care provider (PCP) who is responsible for providing routine care
- understand and be willing to comply with all study-related procedures and agree to participate in the study by giving written informed consent
- plan to remain in the Philadelphia area for the next 9 months or more

Exclusion Criteria

Pregnant or nursing, or plans to become pregnant in the next 9 months.
Uncontrolled hypertension (systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg)
Type 1 diabetes
Type 2 diabetes
A fasting blood glucose > 126 mg/dL (on second assessment after first elevated value)
History of cardiovascular disease (e.g., coronary artery disease, stroke, arrhythmias, congestive heart failure, or heart block greater than first degree
Clinically significant hepatic or renal disease
Hyperthyroidism
Other thyroid disease, not controlled
History of malignancy (except for non-melanoma skin cancer) in past 5 years
Narrow angle glaucoma
Presence or history of marked agitation
Current severe major depressive episode (BDI-II score ≥ 29), current active suicidal ideation, or history of suicide attempts within the past 5 years.
Any severity of thought or bipolar disorder, or bulimia nervosa.
Psychiatric hospitalization within the past 6 months
Self-reported alcohol or substance abuse within the past 6 months, including at-risk drinking (current consumption of ≥ 14 alcoholic drinks per week)
Past year history of drug abuse
Use in the past 2 weeks of monoamine oxidase inhibitors
Current use of serotonin-norepinephrine reuptake inhibitors (SNRIs; e.g. venlafaxine, duloxetine, desvenlafaxine, milnacipran, levomilnacipran).
Use in past 6 months of medications known to induce significant weight loss (i.e., prescription weight loss medications) or weight gain (e.g., chronic use of oral steroids, second generation antipsychotics)
Loss of ≥ 5% of initial body weight within the past 6 months
History of (or plans for) bariatric surgery (e.g., roux en y gastric bypass, sleeve gastrectomy, gastric banding), endoscopic intragastric balloon, or aspire assist.
Inability to walk 5 blocks comfortably or engage in some other form of aerobic activity (e.g., swimming)
Known or suspected allergy to sympathomimetic amines or related products
The receipt of any investigational drug within 6 months prior to this trial
Previous participation in this trial (e.g., randomized and failed to participate)
Changes to any chronic medication (type or dosage) within the past 3 months.
Any serious or unstable medical or psychological condition that, in the opinion of the investigator, would compromise the patient's safety or successful participation in the study

Other Therapy: Subjects will be expected to use medications (prescribed by their PCP) to control traditional cardiometabolic risk factors (e.g., hypertension, hypercholesterolemia, etc) and other co-morbid conditions, with the exception of medications listed above under "exclusions." In all cases, the subjects' PCP will be asked at the study's outset to keep medication does constant throughout the study, whenever possible. Subjects will be expected to have been on their medication regimen (including the dose) for 3 months prior to beginning the BT program.

To be eligible to participate in the randomized phase of the trial, subjects must also:

Complete at least 3 out of 4 treatment sessions during the 4-week BT run-in and attend a randomization visit. Attending an in-person makeup session within one week of a missed visit will count as having attended the run-in visit.
Lose < 2.0% of initial weight during the 4-week BT run-in.

Early BT responders who lose>=2% during the BT run-in will be offered the same 24-week BT program, but will not receive study medication or be included in the randomized trial.

Study & Design

Study Type: INTERVENTIONAL

Study Design: PARALLEL

Arm && Interventions

Group	Intervention	Description
Phase 2: Behavioral Treatment + Placebo	Behavioral Treatment	Participants with suboptimal early weight loss in the BT run-in will then be randomly assigned to 24 additional weeks of: 1) BT plus placebo (BT+P); or 2) BT plus medication (BT+M; phentermine 15.0 mg) in a double-blinded fashion. Both treatment groups will continue to attend individual BT sessions and will take a once daily study medication (placebo or phentermine 15.0 mg) for the duration of the intervention period. Early BT responders identified during the run-in will receive the same 24-week BT program, but will not receive study medication or be included in the randomized trial.
Phase 2: Behavioral Treatment + Placebo	Placebo	Participants with suboptimal early weight loss in the BT run-in will then be randomly assigned to 24 additional weeks of: 1) BT plus placebo (BT+P); or 2) BT plus medication (BT+M; phentermine 15.0 mg) in a double-blinded fashion. Both treatment groups will continue to attend individual BT sessions and will take a once daily study medication (placebo or phentermine 15.0 mg) for the duration of the intervention period. Early BT responders identified during the run-in will receive the same 24-week BT program, but will not receive study medication or be included in the randomized trial.
Phase 2: Behavioral Treatment + Medication	Behavioral Treatment	Participants with suboptimal early weight loss in the BT run-in will then be randomly assigned to 24 additional weeks of: 1) BT plus placebo (BT+P); or 2) BT plus medication (BT+M; phentermine 15.0 mg) in a double-blinded fashion. Both treatment groups will continue to attend individual BT sessions and will take a once daily study medication (placebo or phentermine 15.0 mg) for the duration of the intervention period. Early BT responders identified during the run-in will receive the same 24-week BT program, but will not receive study medication or be included in the randomized trial.
Phase 2: Behavioral Treatment + Medication	Phentermine 15 MG	Participants with suboptimal early weight loss in the BT run-in will then be randomly assigned to 24 additional weeks of: 1) BT plus placebo (BT+P); or 2) BT plus medication (BT+M; phentermine 15.0 mg) in a double-blinded fashion. Both treatment groups will continue to attend individual BT sessions and will take a once daily study medication (placebo or phentermine 15.0 mg) for the duration of the intervention period. Early BT responders identified during the run-in will receive the same 24-week BT program, but will not receive study medication or be included in the randomized trial.
Phase 1: 4-week Behavioral Treatment Run-in	Behavioral Treatment	All enrolled participants will complete an initial 4-week behavioral treatment (BT) run-in. This run-in will be used to identify early non-responders to BT, defined by a weight loss \<2% of initial weight after 4 weeks of BT. Early responders are those who lose \>=2%. Only early non-responders will then be enrolled in the randomized trial.

Primary Outcome Measures

Name	Time	Method
Phase 1: Percent Weight Loss	Week -4 (start of BT run-in) to week 0 (randomization)	Co-primary outcomes - phase 1
Phase 1: Number of Participants Who Are Categorized as Early Non-responders at Randomization (Week 0), Based on Percent Weight Loss	Week -4 (start of BT run-in) to week 0 (randomization)	Co-primary outcomes - phase 1
Phase 1: Baseline Satiety, as Measured by Visual Analogue Scales (Range 0-100 mm) During a Test Meal; Satiety Quotient = [(Post-preload Rating - Fasting Rating Before Preload)] / (Energy Content of Preload in kcal) x 100.	Baseline (week -5)	Primary predictor variable - phase 1 Appetite suppression was first calculated as the average of 100 mm VAS items: hunger (reverse score), satisfaction, fullness, and prospective consumption (reverse score), such that higher scores indicate more appetite suppression (less appetite) for each test meal rating (fasting, then every 10m for 60m). The satiety quotient was then calculated for each post-preload rating using the above formula (see measure title). More positive scores show increased satiety (more appetite suppression). The final analysis uses the 60-minute area under the curve (AUC) for the satiety quotient to predict phase 1 weight loss outcomes. Area under the curve is calculated using the trapezoidal rule to sum the area under each 10-minute interval. AUC = Σ i = 0 to i = 60 10min\*(x(i) + x(i-1))/2) where x is the satiety quotient value at time i. Higher scores indicate higher sustained satiety.
Phase 1: Baseline Postprandial Change in GLP-1 During a Test Meal	Baseline	Primary predictor variable - phase 1. Blood samples were drawn at time 0 (fasting) and 30- and 60-min postprandial samples after consumption of a test meal. Value presented below is the 60-minute incremental area under the curve (AUC) for GLP-1 in picomoles (pM). Area under the curve is calculated using the trapezoidal rule to sum the area under each 0.5-hour interval. AUC = Σ i = 0 to i = 1 0.5hr\*(x(i) + x(i-1))/2) where x is the GLP-1 value in pM at time i.
Phase 1: Baseline Gastric Emptying During a Test Meal (Acetaminophen Test)	Baseline	Primary predictor variable - phase 1 Gastric emptying was measured as the 60-minute area under the curve (AUC) for acetaminophen in micrograms per milliliter (ug/mL). Blood samples were obtained at time 0 (fasting/no acetaminophen - confirmatory) and 30 and 60-min after ingestion. Because acetaminophen is minimally absorbed by the stomach but quickly enters the bloodstream in the small intestine, gastric emptying is considered to be the primary factor influencing its appearance in the blood. Area under the curve is calculated using the trapezoidal rule to sum the area under each 0.5-hour interval. AUC = Σ i = 0 to i = 1 0.5hr\*(x(i) + x(i-1))/2) where x is the acetaminophen value in ug/mL at time i.
Phase 2: Percent Weight Loss	Week 0 (randomization) to week 24	Primary outcomes - phase 2 Percent change from randomization in body weight

Secondary Outcome Measures

Name	Time	Method
Phase 1: Baseline Postprandial Change in Insulin During a Test Meal	Baseline	Secondary predictor variable - phase 1 Blood samples were drawn at time 0 (fasting) and 30- and 60-min postprandial samples after consumption of a test meal. Incremental area under the curve in insulin measured in micro-international units per milliliter (ulU/mL). Area under the curve is calculated using the trapezoidal rule to sum the area under each 0.5-hour interval. AUC = Σ i = 0 to i = 1 0.5hr\*(x(i) + x(i-1))/2) where x is the insulin value in ulU/mL at time i.
Phase 1: Baseline Hunger, as Measured by Visual Analogue Scales (Range 0-100 mm, Higher = More Hunger) During a Test Meal	Baseline	Secondary predictor variable - phase 1 Hunger was rated before and at 10 min intervals after a test meal for 60 min. Data presented below are the 60-min area under the curve (AUC) for postprandial change in hunger at baseline.(more negative = more sustained reduction in hunger). Area under the curve is calculated using the trapezoidal rule to sum the area under each 10-minute interval. AUC = Σ i = 0 to i = 60 10min\*(x(i) + x(i-1))/2) where x is the hunger scale score at time i.
Phase 1: Baseline Relative Reinforcing Value of Food (Computer Task), Number of Food Reinforcer Points Earned	Baseline	Secondary predictor variable - phase 1 Subjects are allowed to work to earn points from a slot machine task at either of two computer stations, one of which provides points towards obtaining a serving of a preferred high-calorie food, and the other points towards a preferred low-calorie food. Points are earned on a progressive ratio scale that increases at fixed intervals. The primary outcome is the number of reinforcer points (servings) earned for the high energy density food, which is thought to reflect the subject's willingness to allocate time and effort to obtaining desired high-calorie foods. The minimum number of points that can be earned is 0; there is no specified maximum.
Phase 1: Baseline Delay Discounting (Computer Task), Area Under the Curve Representing the Ratio of Immediate Reward Size to Time Delay	Baseline	Secondary predictor variable - phase 1 Delay discounting is assessed via a computer program in which subjects are offered choices between small, immediate monetary rewards and larger, delayed rewards. The present delay discounting computer task used 7 time delays for $1000. The outcome is the area under the curve (AUC) representing the ratio of immediate reward size to time delay. AUCs were standardized to fall between 0 and 1 (Myerson et al., 2001) and were calculated for the plot of subjective values vs. delay, with lower values indicating greater discounting. Area under the curve is calculated using the trapezoidal rule to sum the area under each standardized time interval. AUC = Σ i = 0 to i = 1 proportion of max delay\*(x(i) + x(i-1))/2) where x is the proportion of the maximum price inflection value at time i.
Phase 1: Baseline Implicit Wanting, Reaction Time on Leeds Food Preference Questionnaire	Baseline	Secondary predictor variable - phase 1 Implicit wanting is measured by the Leeds Food Preference Questionnaire, a computer-based task using a forced choice paradigm for four categories: High-fat savory, high-fat sweet, low-fat savory, low-fat sweet. Reaction times are transformed to a standardized D-score that is then adjusted for the frequency of selection using a validated algorithm. Scores can range from -100 to 100 with more positive scores indicating a more rapid preference for one category over the other and more negative scores indicating the opposite.
Phase 1: Baseline Fasting Leptin	Baseline	Secondary predictor variable - phase 1 Unit: Picograms per milliliter (pg/mL) Blood samples were drawn at time 0 (fasting).
Phase 1: Baseline Fasting Ghrelin	Baseline	Secondary predictor variable - phase 1 Blood samples were drawn at time 0 (fasting). Active ghrelin. Unit: Picograms per milliliter (pg/mL)
Phase 1: Baseline Postprandial Change in Peptide YY During a Test Meal	Baseline	Secondary predictor variable - phase 1 Blood samples were drawn at time 0 (fasting) and 30- and 60-min postprandial samples after consumption of a test meal. Value presented below is the 60-minute incremental area under the curve (AUC) for PYY in picograms per milliliter (pgmL) at baseline. Area under the curve is calculated using the trapezoidal rule to sum the area under each 0.5-hour interval. AUC = Σ i = 0 to i = 1 0.5hr\*(x(i) + x(i-1))/2) where x is the PYY value in pg/mL at time i.
Phase 2: Weight Loss (kg)	Week 0 (randomization) to week 24	Secondary outcomes - phase 2
Phase 2: Number of Participants With a Weight Loss of 5% or Greater of Randomization Body Weight at Week 24	Week 0 (randomization) to week 24	Secondary outcomes - phase 2
Phase 2: Number of Participants With a Weight Loss of 10% or Greater of Randomization Body Weight at Week 24	Week 0 (randomization) to week 24	Secondary outcomes - phase 2
Phase 2: Change in Appetite Suppression, as Measured by Visual Analogue Scales (Range 0-100 mm) During a Test Meal	Week 0 (randomization) to week 24	Secondary outcomes - phase 2 Appetite suppression was first calculated as the average of 100 mm VAS items: hunger (reverse score), satisfaction, fullness, and prospective consumption (reverse score), such that higher scores indicate more appetite suppression (less appetite) for each test meal rating (fasting, then every 10m for 60m). The final analysis uses the 60-minute incremental area under the curve (AUC) for change in appetite suppression from fasting. Higher scores indicate higher sustained appetite suppression. Area under the curve is calculated using the trapezoidal rule to sum the area under each 10-minute interval. AUC = Σ i = 0 to i = 60 10min\*(x(i) + x(i-1))/2) where x is the appetite suppression value at time i.
Phase 2: Change in Hunger, as Measured by Visual Analogue Scales (Range 0-100 mm, Higher=More Hunger) During a Test Meal	Week 0 (randomization) to week 24	Secondary outcomes - phase 2 Hunger was rated before and at 10 min intervals after a test meal for 60 min. Data presented below are the 60-min incremental area under the curve (AUC) for postprandial change in hunger at randomization and week 24. More negative scores indicate greater sustained reductions in hunger from fasting. Area under the curve is calculated using the trapezoidal rule to sum the area under each 10-minute interval. AUC = Σ i = 0 to i = 60 10min\*(x(i) + x(i-1))/2) where x is the scale score at time i.
Phase 2: Change in Delay Discounting (Computer Task), Area Under the Curve Representing the Ratio of Immediate Reward Size to Time Delay	Week 0 (randomization) to week 24	Secondary outcomes - phase 2 Delay discounting is assessed via a computer program in which subjects are offered choices between small, immediate monetary rewards and larger, delayed rewards. The present delay discounting computer task used 7 time delays for $1000. The outcome is the area under the curve (AUC) representing the ratio of immediate reward size to time delay. AUCs were standardized to fall between 0 and 1 (Myerson et al., 2001) and were calculated for the plot of subjective values vs. delay, with lower values indicating greater discounting. Area under the curve is calculated using the trapezoidal rule to sum the area under each standardized time interval. AUC = Σ i = 0 to i = 1 proportion of max delay\*(x(i) + x(i-1))/2) where x is the proportion of the maximum price inflection value at time i.