Open-label pilot study of lisdexamfetamine for cocaine use disorder

Abstract

Background

Cocaine use disorder (CUD) is a substantial public health problem with no FDA-approved medication treatments. Psychostimulants have shown promise as pharmacotherapy for CUD. Lisdexamfetamine, a novel prodrug psychostimulant, is roughly 40–50% as potent as dextroamphetamine.

Objectives

To evaluate the safety, tolerability, and optimal dosing of lisdexamfetamine for treating CUD.

Methods

Open-label, 8-week trial of 17 CUD adults. Participants were titrated to the maximum tolerated dose of 140 mg over 2-week period and maintained for 4 weeks, followed by a two-week taper period. The primary outcome measures were the maximum daily dose achieved during the study period and tolerability as measured by medication-related study drop-out.

Results

Among the 16 participants with post-enrollment data, the mean dose of lisdexamfetamine achieved was 118.1 mg (standard deviation (SD) = 40.4), mean retention was 6.5 weeks (SD = 2.0), and no participants discontinued study medication due to adverse effects. Four participants had dose reductions due to adverse effects and continued in the trial. Six participants (37.5%) were abstinent for the last 3 weeks of their study participation. Mean dollars of cocaine spent per day significantly decreased from $19.72 at baseline to $7.57 during the last 3 weeks of study participation (t₁₅ = 3.60, p = .003). The mean percent of using days significantly decreased from 25% at baseline to 12% during the last 3 weeks of study participation (t₁₅ = 3.33, p = .005).

Conclusion:

The use of lisdexamfetamine for CUD in doses ranging to 140 mg daily was safe and generally well tolerated.

Introduction

Approximately 2% of the US population used cocaine in 2018, with approximately 1 million individuals meeting criteria for cocaine use disorder (1), yet there are no FDA-approved pharmacotherapies. Scores of double-blind, placebo-controlled pharmacotherapy clinical trials for CUD have been conducted (2–5), testing agents drawn from a wide variety of medication classes. Dopamine agonists have yielded the most promising results, but concerns regarding safety and the potential for misuse have hindered development (6,7). The need to provide treatment for the majority of individuals with CUD, for whom behavioral treatment alone is often inadequate, makes the development of safe and effective pharmacotherapy an important public health goal.

Amphetamine and cocaine have similar pharmacological and clinical characteristics; they differ mainly in onset of action and half-life. The primary action of amphetamine is to increase synaptic concentrations of monoamine neurotransmitters by promoting release (8), and as a less important mechanism, by blocking reuptake (9,10). Non-dopamine transporter-mediated mechanisms are also hypothesized to participate in amphetamine-induced dopamine release (11). Chronic amphetamine administration has been shown to reduce cocaine self-administration in rats (12) and rhesus monkeys [13–16]. Amphetamine has also been shown in human laboratory studies to reduce cocaine self-administration in humans (15,17,18).

Psychostimulants, including amphetamine analogs, methylphenidate, bupropion and modafinil, have been studied as substitution treatments for CUD, both in patients with (19–21) and without (22–31) co-occurring attention-deficit hyperactivity disorder (ADHD). The results of these studies have been mixed with regard to effects on cocaine use outcomes, with the most consistent therapeutic effects reported for amphetamine at robust doses (7,20,24,27,28,32,33).

Lisdexamfetamine mesylate is a novel prodrug approved by the FDA for ADHD and binge eating disorder in doses up to 70 mg per day. After absorption into the bloodstream, lisdexamfetamine (LDX) is metabolized in red blood cells by rate-limited, enzymatic hydrolysis to dextroamphetamine (DEX) and L-lysine (34). LDX is estimated to be approximately 50–60% less potent than immediate release (IR) dextroamphetamine (35), with a markedly delayed time of peak effect (36). When tested in rats, LDX in equivalent doses produced smaller, but more sustained, increases in striatal dopamine efflux than IR DEX, and substantially less locomotor activation (37). This pharmacokinetic effect may explain why in drug-experienced individuals, LDX evoked lower “drug liking” scores than IR DEX (36). In rhesus monkeys, LDX and DEX produced dose-dependent and complete substitution for the discriminative stimulus effects of cocaine (13). However, LDX had lower potency, a slower onset, and a longer duration of action than DEX, but had the same effect as DEX on reducing the choice of cocaine in rhesus monkeys.

A double-blind study of 43 individuals with CUD found that LDX 70 mg per day reduced craving, but not cocaine use, as compared to placebo (38). LDX was well tolerated in CUD patients, with no differences for blood pressure, heart rate, or body weight. While this study was under-powered to definitively determine efficacy, the reduction in craving and tolerability support the potential of LDX as a treatment for CUD. Since LDX is approximately 50–60% less potent than DEX, the equivalent medication dosing of this trial was lower than trials of DEX (up to 60 mg/day) (28,39) or mixed amphetamine salts (60–80 mg) (20,27,32). LDX has been studied in doses up to 250 mg per day in healthy volunteers (40) and in individuals with schizophrenia (41). LDX demonstrated linear pharmacokinetic parameters, with doses up to 150 mg daily well tolerated; doses higher than 150 mg were more likely to lead to elevated blood pressure or pulse, although no serious adverse events occurred in either study (40, 41).

Based upon the results of the Mooney et al. trial (38) using LDX 70 mg, and the scientific rationale for testing higher doses of LDX, we conducted an open-label trial of LDX at doses up to 140 mg in individuals with CUD to evaluate the safety, tolerability, and optimal dosing of LDX for treating CUD. The primary objective of the study was to determine the ideal target dose range and tolerability of LDX pharmacotherapy in CUD. The secondary objectives included the assessment of the effect of LDX on cocaine use as measured by reported days of use and confirmed by urine toxicology.

Methods

Study design

This study was conducted at the Substance Treatment and Research Service (STARS) of the Columbia University Medical Center and University of Minnesota, Department of Psychiatry from August 2011 through September 2012.

The primary outcome measure was the maximum total LDX dose achieved during the study period defined as the highest dose taken for a single day. Secondary outcome measures were: 1) tolerability, as measured by adverse effect report, need for dose reduction or discontinuation of study medication; 2) change in reported cocaine use (baseline compared to the last 3 weeks of study participation) as recorded by the timeline follow-back method; and 3) study retention.

Participants

The institutional review boards of the New York State Psychiatric Institute and University of Minnesota approved this research protocol and all participants provided informed consent prior to study enrollment. Participants were recruited by local advertising (internet, print, radio, television, and subway) or by clinical referrals in the New York City metropolitan area and Minneapolis area. At both sites, a total of 260 potential participants were screened.

Enrolled participants were 17 men and non-pregnant women between the ages of 18–60 who met DSM-IV criteria for current cocaine dependence. Sixteen participants had post-enrollment data for analysis, one participant did not report any cocaine use during the study and dropped out in 1^st week of the study. Participants were treatment seeking, used cocaine at least 4 days in the past month, were in general good health, were deemed capable of providing informed consent and complying with study procedures, and women of child-bearing age agreed to use a method of contraception with proven efficacy.

Exclusion criteria included: 1) meeting DSM-IV-TR criteria for bipolar disorder, schizophrenia, any psychotic disorder other than transient psychosis due to drug abuse, or current major depressive disorder; 2) any other current Axis I psychiatric disorder as defined by DSM-IV-TR that in the investigator’s judgment were unstable; 3) physiological dependence on any other drugs (excluding nicotine or cannabis) which require medical intervention; 4) current psychostimulant abuse or dependence (other than cocaine dependence); 5) current suicidal risk; 6) coronary vascular disease as indicated by history or suspected by abnormal ECG or history of cardiac symptoms; 7) unstable physical disorders which might make participation hazardous such as uncontrolled hypertension (SBP > 140, DBP > 90, or HR > 100), acute hepatitis or uncontrolled diabetes; 8) history of seizures, hyperthyroidism and/or glaucoma; 9) family history of sudden cardiac death; 10) history of allergic reaction to study medication; 11) pregnancy or nursing; 12) currently being prescribed psychotropic medication by another physician (other than sleep medication); 13) legally mandated to participate in substance abuse treatment program.

Treatment

Participants received LDX under open-label conditions using flexible-fixed dosing schedule with a target dose of 140 mg daily, with instructions to take the medication twice daily. Participants were titrated to the target or maximum tolerated dose over a 2-week period, then maintained on the highest tolerated dose for 4 weeks, followed by a two-week run-down period (Table 1). The two-week taper (weeks 7–8) was a gradual reduction from the week 6 dose achieved. The research psychiatrist made dose reductions for tolerability problems. All participants received medication management counseling using a structured compliance enhancement manual designed for pharmacotherapy trials in subjects with substance use disorders.

Table 1.

Open-label lisdexamfetamine dosing.

Open-Label Lisdexamfetamine Dosing
Study Week	1	1	1	1	2	2	2	2	3–6	7–8
Study Day	1–2	3–4	5–6	7	8–9	10–11	12–13	14	15–43	43–56
Total Daily Dose Lisdexamfetamine	20 mg	40 mg	60 mg	70 mg	80 mg	100 mg	120 mg	140 mg	140 mg	Taper

A structured pill count interview with a weekly financial incentive for medication bottle return was employed to enhance and measure medication compliance. A structured calendar-based interview covered the period since the last visit and accounted for every scheduled dose of medication. Participants were compensated 10 USD per week to return their medication bottles with any unused pills. Participants received 5 USD for each screening or study visit to offset travel costs. Participants who attended all screening and study visits and returned all medication bottles could earn a maximum of 215 USD for the entire study.

All patients had a weekly individual session with a research psychiatrist using a structured compliance enhancement manual designed for pharmacotherapy trials in subjects with substance use disorders (42). Sessions were approximately 30 minutes in duration and were focused on setting abstinence from cocaine use as a goal, participant compliance with study medications and procedures, and current functioning.

Study visits occurred twice weekly during the study period. At each visit, participants were evaluated by study staff for monitoring of vital signs, urine toxicology and medication compliance. Study exit criteria included: 1) medical or psychiatric emergency, cardiovascular, or ECG abnormalities; 2) cardiovascular instability defined as resting HR > 100 or resting BP at > 140/90 mm Hg for greater than 2 consecutive weeks or SBP > 160, DBP > 110, HR > 110 at any one visit; 3) chest pain, fainting or arrhythmias; 4) significant increase in drug use or drug-related problems as indicated by CGI score of 6 (much worse than baseline) or greater for 2 consecutive weeks; 5) development of serious psychiatric symptomatology as indicated by CGI score of 6 (much worse than baseline) or greater for two consecutive weeks; and 6) pregnancy. Vital signs (heart rate and blood pressure) and side effects were monitored at each visit. Urine toxicology was conducted twice weekly. ECG and pregnancy testing were performed every 4 weeks.

Measures

During the screening period, a comprehensive psychiatric and medical evaluation including history, physical and laboratory examination, and the Structured Clinical Interview (SCID) for DSM-IV Axis I disorders (43) interview were performed. Self-reported cocaine use data was collected using the TLFB method (44) for the 28 days preceding study entry and each day of the study period. Urine toxicology samples were tested for cocaine metabolites during screening, at study entry, and twice weekly during the study period. Adverse events were recorded weekly using the Systematic Assessment for Treatment Emergent Events (SAFTEE) (45).

Data analysis

Summary statistics of baseline demographics of the sample were computed using frequencies, means and standard deviations. The means of highest dose of LDX achieved, retention week, and cocaine use in dollars used per day and percent of using days were also computed. Paired t-tests were used to analyze whether cocaine use in dollars spent per day and percent of use per day changed from baseline to the last 3 weeks in the study. Side effects were summarized by creating indicators of whether subjects reported: any side effects during the trial, any dose reductions due to side effects, and any discontinuation of medication due to side effects. All analyses were conducted in SAS^® and all statistical tests were 2-tailed at a significance level of 5%.

Results

Seventeen participants were enrolled, with one participant lost to follow-up without any post-enrollment study data. The majority of the enrolled sample was male (88%, 15/17), African American (10/17, 59%) and not employed full-time (76%, 13/17). About half the sample were single (9/17, 53%) and completed some college or technical school or more (47%, 8/17). On average, participants were 43.8 years old (standard deviation (SD) = 8.4).

Table 2 displays the participant-wise summary statistics for the outcomes of interest in the trial. All study outcomes are for the active medication period (weeks 1–6) excluding the taper period. Among the 16 participants with post-enrollment data, the mean dose of LDX achieved was 118.1 mg (SD = 40.4), mean retention was 6.5 weeks (SD = 2.0), and no participants discontinued study medication due to adverse effects. Four participants (25%) had dose reductions due to adverse effects and continued in the trial; 1) reported insomnia and anxiety, 2) reported of decreased appetite, agitation, restlessness and dry mouth, 3) hypertension, and 4) tachycardia, hypertension, and agitation. Six participants (37.5%) were abstinent for the last 3 weeks of their study participation. Mean dollars of cocaine spent per day decreased significantly from 19.72 USD at baseline (28 days prior to study entry) to 7.57 USD during the last 3 weeks of study participation (t₁₅ = 3.60, p = .003). The mean percent of using days also decreased significantly from 25% at baseline (28 days prior to study entry) to 12% during the last 3 weeks of study participation (t₁₅ = 3.33, p = .005).

Table 2.

Participant outcomes.

ID #	During Last 3 Weeks in Study		Baseline		Highest Dose Achieved (mg)	Side Effects			Retention Week	Positive Urine Test
	Proportion of Use Days	Mean Dollars Spent per Day	Proportion of Use Days	Mean Dollars Spent per Day		Any side effects reported during trial	Any dose reduction due to side effects	Any d/c of medication due to side effects		Proportionduringthe entire trial	Proportionduringthe Last 3weeks
CU-2	0.00	0.00	0.36	15.54	140	yes	no	no	8	0.375	0.000
CU-4	0.00	0.00	0.18	9.29	140	yes	no	no	6	0.308	0.200
CU-6	0.00	0.00	0.18	42.86	60	yes	yes	no	8	0.100	0.000
CU-8	0.00	0.00	0.29	11.43	140	yes	no	no	8	0.214	0.000
CU-9	0.00	0.00	0.25	9.29	140	yes	yes	no	6	0.000	0.000
MN-2003	0.00	0.00	0.04	0.18	140	yes	no	no	8	1.000	1.000
CU-12	0.06	1.76	0.43	37.86	140	yes	no	no	5	1.000	1.000
MN-2001	0.06	1.18	0.25	11.43	140	yes	no	no	8	0.250	0.200
CU-11	0.10	1.67	0.14	5.00	60	yes	yes	no	8	0.467	0.667
CU-10	0.10	4.00	0.14	7.14	140	yes	no	no	4	0.143	0.500
CU-5	0.11	3.68	0.29	25.71	140	yes	no	no	5	0.455	0.667
MN-2005	0.20	3.33	0.26	17.83	140	no	no	no	8	0.938	1.000
CU-7	0.29	21.43	0.25	25.00	20	no	no	no	1	0.000	0.000
MN-2004	0.31	11.25	0.13	17.92	140	yes	yes	no	8	0.688	0.833
MN-2002	0.33	33.33	0.48	52.38	140	no	no	no	5	1.000	1.000
CU-1	0.38	39.52	0.32	26.61	70	yes	no	no	8	0.938	1.000
CU-3			0.39	78.57						1.000	1.000

^*ID CU-3 did not report any cocaine use during the study. ID CU-7 was only in the study for 1 week, summary measures for last 3 weeks in the study used all data available. CU = Columbia University site; MN = University of Minnesota site.

No participants were removed from the trial for safety exit criteria including cardiovascular complications. There were no serious adverse events.

Discussion

This 8-week outpatient pilot trial tested LDX in doses up to 140 mg under open-label conditions for the treatment of CUD. The objective of the study was to determine the ideal target dose range, tolerability, and feasibility of LDX as a pharmacotherapy for CUD. As most participants were able to tolerate the target dose of 140 mg for the duration of the study period, the tested dosing schedule, with a two-week titration, appears to be an appropriate dosing strategy. While most participants reported adverse effects, no participants discontinued study medication or exited the trial due to medication adverse effects. Out of the 16 participants with post-enrollment data, four (25%) participants required dose reductions. Of these four participants, three completed the entire trial, and one completed 75% of the study weeks, suggesting that a flexible-fixed dosing strategy for LDX can be effective in retaining participants who have adverse effects at a higher dose. In interpreting these results, it is important to consider the relative potency of LDX to dextroamphetamine. LDX is roughly 40–50% as potent as dextroamphetamine, which means that the dose of LDX tested in this study is roughly equivalent to 60 mg/day of dextroamphetamine. Participants in this study were using approximately 25% of days at baseline, which may limit the generalization of the tolerability the this LDX to more frequent users of cocaine.

Six (37.5%) of the 16 participants with post-enrollment data achieved abstinence for the last 3 weeks of the study. The mean dollars of cocaine used per using day and the mean days of use per week from baseline to last week of trial participation reduced for the study sample. The results of this open-label trial cannot be used to determine efficacy of LDX for treating CUD, but the improvement in cocaine use is promising. These results of this study are also consistent with the existing literature of psychostimulant treatment of CUD (7,46).

Concerns over the misuse potential for amphetamine are one of the main limitations inhibiting its development as a treatment for CUD. The unique properties of the prodrug LDX offer certain advantages over conventional amphetamine preparations with regards to potential for misuse. The need for hydrolysis of the prodrug LDX appears to prevent the rapid appearance of DEX in the blood that is seen following oral or intravenous administration of other amphetamine products. In rats, LDX did not serve as a positive reinforcer in cocaine-trained rats (47). Orally administered LDX in equivalent doses (100 mg to 40 mg) to IR DEX had attenuated responses to measures of abuse liability, while abuse-related liking scores of LDX at a dose corresponding to a 50% higher amphetamine base (LDX 150 mg) were similar to DEX 40 mg (36). Intranasal and oral administration of LDX results in comparable serum levels of dextroamphetamine, suggesting no pharmacokinetic advantage to intranasal use (48). Intravenous administration of LDX, in comparison to equivalent doses of intravenous dextroamphetamine, does not produce significant abuse-related drug liking scores (49). Non-medical use of prescription stimulant medication in the US is mostly for wakefulness and performance enhancement, medication is obtained from family/friends, and used by an oral route of administration (50). In the US population, the rate of non-medical use per 100,000 dispensed prescription of LDX is much lower (0.13) than Adderall (1.61) or Adderall XR (0.62) (50). Similar results are seen among substance use disorder treatment population, with LDX having lower rates of non-medical use as compared to other amphetamine preparations (51). Taken together, these data suggest that LDX is uniquely positioned as the amphetamine preparation to offer both a potential for therapeutic efficacy in CUD, while minimizing risk of misuse.

Another safety concern when considering the use of stimulant medication are the potential cardiovascular risks. The safety of psychostimulants for the treatment of ADHD in young and middle-aged individuals with ADHD is well-established (52). Therapeutic doses of LDX are not associated with cardiac function (53). While stimulant medications are generally safe and effective in the general population, concerns about cardiovascular safety need to be considered when combined with cocaine in individuals with CUD. Trials of stimulant medications for the treatment of CUD using standard screening methods have not encountered cardiovascular safety issues (3,20,32,54,55), suggesting that stimulant pharmacotherapy is generally well tolerated in a clinical trial population screened for cardiovascular disease. This study excluded participants at risk for cardiovascular adverse events and closely monitored for cardiovascular symptoms during the course of the trial. No participants were exited from the trial for cardiovascular complications.

This study has several limitations. The first limitation is that as an open-label trial with no comparison group we are unable to assess the effects of LDX as compared to placebo. Any improvement in cocaine use outcomes could be unrelated to medication treatment and a result of the nonspecific behavioral effects of trial participation. Another limitation is that we only tested a single target dose of LDX; it is possible that higher or lower doses of LDX would be more optimal for treating individuals with CUD.

The results of this study suggest that LDX in doses up to 140 mg was well tolerated in CUD patients and testing should progress to an adequately powered double-blind trial. Based on the results of clinical trials using DEX or mixed amphetamine salts for CUD in doses of 60–80 mg/day, evidence that LDX is well tolerated in healthy volunteers and individuals with schizophrenia in doses up to 150 mg/day, the literature that suggests that LDX is roughly 40–50% as potent as DEX (56), and the pilot study experience with testing a dose of LDX of 140 mg per day for CUD, we would propose that a dose of 140 mg a day would be appropriate for future studies testing LDX as a treatment for CUD.

No clearly effective pharmacotherapy for CUD has been developed, and one of the most promising lines of research is the use of psychostimulants as agonist replacement therapy (7), but concerns of misuse have limited its development. Preclinical and clinical trial evidence suggests that amphetamine is the most promising candidate psychostimulant medication for CUD. LDX, a unique prodrug amphetamine preparation, has preclinical and clinical evidence supporting its potential efficacy for CUD, while having lower risk of misuse than other amphetamine preparations. LDX possess the ideal characteristics for an agonist replacement agent; LDX has a gradual onset of action, no pharmacokinetic advantage of non-oral ingestion routes, and a long duration of action. Preclinical and clinical evidence supports the testing of LDX at doses of 140 mg per day for the treatment of CUD.