A Preclinical Evaluation of the Discriminative And

NP5101_proof ■ 12 June 2013 ■ 1/11

Neuropharmacology xxx (2013) 1e11

Contents lists available at SciVerse ScienceDirect

Neuropharmacology

journal homepage: www.elsevier.com/locate/neuropharm

1 56 2 A preclinical evaluation of the discriminative and reinforcing 57 3 58 4 properties of lisdexamfetamine in comparison to D-amfetamine, 59 5 ﬁ 60 6 methylphenidate and moda nil 61 7 62 a,* a a a b 8 Q5 David J. Heal , Niki W. Buckley , Jane Gosden , Nigel Slater , Charles P. France , 63 9 David Hackett c 64 10 65 a 11 RenaSci Ltd, BioCity Nottingham, Pennyfoot Street, Nottingham NG1 1GF, UK 66 b University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229-3900, USA 12 c Shire Pharmaceutical Development Ltd, Hampshire International Business Park, Chineham, Basingstoke RG24 8EP, Hampshire, UK 67 13 68 14 69 15 article info abstract 70 16 71

17 Article history: Lisdexamfetamine dimesylate, which consists of L-lysine covalently bound to D-amfetamine, is the first 72 18 Received 12 December 2012 prodrug for treating ADHD. Its metabolic conversion to yield D-amfetamine by rate-limited, enzymatic 73 19 Received in revised form hydrolysis is unusual because it is performed by peptidases associated with red blood cells. Other 74 3 April 2013 20 stimulants shown to be effective in managing ADHD include D-amfetamine, methylphenidate and 75 Accepted 13 May 2013 21 modafinil. All have the potential for misuse or recreational abuse. The discriminative and reinforcing 76 effects of these compounds were determined in rats using a 2-choice, D-amfetamine (0.5 mg/kg, i.p.)- 77 22 Keywords: cued drug-discrimination test, and by substitution for intravenous cocaine in self-administration. Lis- 23 Lisdexamfetamine 78 dexamfetamine (0.5e1.5 mg/kg [D-amfetamine base], p.o.) generalised to saline when tested 15 min 24 D-Amfetamine 79 post-dosing, but dose-dependently generalised to D-amfetamine at 60 min. At 120 min, its D-amfet- 25 Methylphenidate 80 fi amine-like effects were substantially diminished. At 15 min, methylphenidate (3.0e10 mg/kg, p.o.) and 26 Moda nil 81 Drug discrimination D-amfetamine (0.1e1.5 mg/kg, p.o.) dose-dependently generalised to the intraperitoneal D-amfetamine 27 Discriminative effects cue. Switching to the intraperitoneal route reduced the interval required for lisdexamfetamine to be 82 28 Self-administration recognised as D-amfetamine-like, but did not alter its potency. Switching to intraperitoneal injection 83 Reinforcer 29 increased the potency of methylphenidate and D-amfetamine by 3.4 and 2.2, respectively. Modafinil 84 30 (50e200 mg/kg, i.p.) generalised partially, but not fully, to D-amfetamine. Methylphenidate (0.1, 0.3, 85 31 1.0 mg/kg/injection, i.v.) maintained robust self-administration at the 2 highest doses. Neither lisdex- 86 32 amfetamine (0.05, 0.15 or 0.5 mg/kg/injection [D-amfetamine base], i.v.) nor modafinil (0.166, 0.498 or 87 33 1.66 mg/kg/injection, i.v.) served as reinforcers. The results reveal important differences between the 88 fi 34 pro les of these stimulants. Lisdexamfetamine did not serve as a positive reinforcer in cocaine-trained 89 fl 35 rats, and although it generalised fully to D-amfetamine, its discriminative effects were markedly in u- 90 enced by its unusual pharmacokinetics. 36 91 Ó 2013 Published by Elsevier Ltd. 37 92 38 93 39 94 40 95 41 1. Introduction 96 42 97 43 Attention deficit hyperactivity disorder (ADHD) is a childhood- 98 44 onset, psychiatric, cognitive and behavioural disorder that is 99 45 widely treated with the catecholaminergic stimulants, D-amfet- 100 amine and methylphenidate. These drugs are effective in managing 46 Abbreviations: ADHD, attention deficit hyperactivity disorder; AMF, amfet- 101 47 amine; C-II, Schedule 2 Controlled Drug; C-IV, Schedule 4 Controlled Drug; CD, the symptoms of approximately three quarters of children and 102 48 controlled drug; Cmax, maximum plasma drug concentration; DAT, dopamine re- adults (Spencer et al., 1996; Elia et al., 1999; Heal and Pierce, 2006; 103 49 uptake transporter; FR, fixed ratio; IACUC, Institutional Animal Care and Use Heal et al., 2009, 2012a; Buitelaar and Medori, 2010). Although 104 Committee; i.p., intraperitoneal; IR, immediate release; i.v., intravenous; PET, 50 these stimulants are undoubtedly effective, they have two major 105 positron emission tomography; PFC, prefrontal cortex; p.o., per os (oral); SAL, saline; 51 shortcomings. First, D-amfetamine and methylphenidate have 106 SR, sustained release; tmax, time to reach maximum plasma drug concentration. 52 * Corresponding author. Tel.: þ44 (0) 115 9124260; fax: þ44 (0) 115 9124263. relatively short half-lives that require the drugs to be administered 107 53 E-mail addresses: [email protected], [email protected] (D.J. Heal). several times a day, which makes them particularly unsuitable for 108 54 109 e Ó 55 0028-3908/$ see front matter 2013 Published by Elsevier Ltd. 110 http://dx.doi.org/10.1016/j.neuropharm.2013.05.021

Please cite this article in press as: Heal, D.J., et al., A preclinical evaluation of the discriminative and reinforcing properties of lisdexamfetamine in comparison to D-amfetamine, methylphenidate and modaﬁnil, Neuropharmacology (2013), http://dx.doi.org/10.1016/ j.neuropharm.2013.05.021 NP5101_proof ■ 12 June 2013 ■ 2/11

2 D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11

111 use by individuals whose disorder is characterised by inattention, cages with sawdust covered floors in a temperature and humidity controlled room. 176 112 distractibility and impulsivity. Second, when these catecholamin- Animals were maintained on 12 h:12 h lightedark cycle with free access to food and 177 tap water at all times when in their home cages. Rats were accustomed to these 113 ergic drugs are taken at doses above those recommended in the 178 conditions for 1 week before the start of training. 114 prescribing instructions and often by non-clinical routes, e.g. nasal For the self-administration study, 54 male, Sprague-Dawley rats (277e352 g at 179 115 insufflation (“snorting”) or intravenous injection, they have start of study) were purchased from Charles River UK, and 58 male, Sprague-Dawley 180 116 powerful psychostimulant and euphoriant properties which makes rats (277 ge342 g at start of study) from Harlan, USA. Rats were housed individually 181 117 them liable to diversion and recreational abuse. Both shortcomings in plastic cages containing rodent bedding and environmental enrichment on a 182 12 h:12 h lightedark cycle in a temperature and humidity controlled room. Animals 118 have to some extent been addressed by the development of long- were allowed to acclimatise to these conditions for at least 4 days before the study 183 119 acting formulations and by the use of novel delivery systems, e.g. commenced, during which time they underwent daily weighing and handling. Rats 184 120 osmotically controlled release or transdermal patches, that are also were allowed free access to tap water and standard rodent diet during the accli- 185 121 tamper deterrent (see reviews by Heal and Pierce, 2006; Heal et al., matisation period. After the acclimatisation period, food was restricted to 10 g/day 186 over 5 days, after which daily food intake was restricted to w90% of normal levels 122 2009, 2012a); nonetheless, all formulations of methylphenidate 187 (calculation based on the mean daily food intake during the acclimatisation period). 123 and D-amfetamine are classified as Schedule 2 Controlled Drugs Rats were given sufficient food to maintain age-appropriate growth. Body weights 188 124 (C-II) in the UK, USA and many other countries. were monitored and the amount of food given in home cages was altered when 189 Ò 125 Lisdexamfetamine dimesylate (Vyvanse ) is a relatively recent necessary. This regime was maintained throughout the remainder of the study, 190 except during the recovery period after surgery. 126 entry to the portfolio of ADHD medications. It is a D-amfetamine 191 In both studies, animals were tested in the light part of the lightedark cycle. 127 prodrug, which comprises the naturally occurring amino acid, L- 192 128 lysine, covalently bound to D-amfetamine via an amide linking 2.2. Drug-discrimination training and testing 193 129 group. Lisdexamfetamine is the first prodrug to have been 194 130 approved in the USA and Canada for the management of ADHD in D-Amfetamine-cued drug-discrimination testing in rats was based on the 195 fl 131 children (age 6e12), adolescents (age 13e17) and adults. It is method previously described by Heal et al. (1992). Brie y, female PVG rats were 196 trained to distinguish between D-amfetamine (0.5 mg/kg, i.p.) and saline (1 ml/kg, 197 132 currently undergoing evaluation for the treatment of ADHD in a i.p.) in a 2-choice lever-pressing task in response to a sweetened milk reward made 133 number of European countries. The metabolic route of conversion available on a FR-5 reward schedule (i.e. 5 lever-presses for 1 reward). Rats were 198 134 of lisdexamfetamine is unusual because after absorption into the randomly allocated one lever for D-amfetamine (0.5 mg/kg, i.p.) and the other for 199 135 bloodstream it is metabolised by red blood cells to yield D-amfet- saline. Once a rat had achieved approximately 60% correct lever-presses on most 200 trials, it began the test regime. 136 amine and the natural amino acid, L-lysine, by rate-limited, enzy- 201 On the test regime, rats were injected with drug cue or saline and then placed in 137 matic hydrolysis (Pennick, 2010). The prodrug is pharmacologically the test chamber. The treatments during testing were alternated to prevent rats 202 138 inert in vitro and lacks affinity for a wide range of molecular targets learning a particular sequence. On a test day, rats were not rewarded during the first 203 139 that mediate the effects of drugs of abuse (data on file, Shire 2.5 min of the session for presses on either lever and then rewarded on either lever 204 for the remaining 7.5 min of the session. 140 Pharmaceuticals). As a prodrug of D-amfetamine, lisdexamfetamine 205 The criterion for acceptable performance during testing was 75% correct lever- 141 fi 206 has been classi ed as a C-II in both the USA and UK. presses in response to the drug cue or saline in the initial 2.5 min of the 10 min test 142 Modafinil is an unusual stimulant with enigmatic pharmacology preceding a drug test and a mean of 75% correct lever-presses in 4 consecutive 207 143 (see reviews by Minzenberg and Carter, 2008; Heal et al., 2012a). drug cue and saline cue tests. When rats had achieved 4 correct saline and amfet- 208 144 Although its clinical development as a treatment for ADHD was amine test sessions they progressed to the test drugs, routes and time periods 209 145 terminated due to safety concerns, modafinil has been shown un- evaluated in this study. Test compounds were assessed in the same manner i.e. the 210 result for each rat was the percentage of responses on the amfetamine lever in the 146 equivocally to improve symptoms in children and adolescents with unrewarded 2.5 min of the test session. 211 147 ADHD in several, randomised, double-blind, placebo-controlled, Rats had to correctly complete one saline and D-amfetamine test and rein- 212 148 clinical trials (Biederman et al., 2006; Swanson et al., 2006; forcement session in a random order between each compound test. These sessions 213 149 Greenhill et al., 2006). Modafinil has a C-IV classification in the were repeated if a rat showed unacceptable performance in response to saline or D- 214 amfetamine. 150 USA, but it is not a CD in the UK. 215 Training of the rats with saline (i.p.) and D-amfetamine (0.5 mg/kg, i.p.) was 151 Thus, all of these stimulants have to a greater or lesser extent the performed 3e4 days each week, but test compounds were tested only once per 216 152 potential for misuse and/or recreational abuse. Drug-discrimination week. Prior to each rat being placed in a chamber, the levers and walls were 217 153 and self-administration studies are mandated by FDA and EMA for swabbed with 10% ethanol solution to prevent olfactory stimuli from the previous 218 fl ’ 154 all novel CNS-active drugs for use in man (Center for Drug rat in uencing the subsequent rat s lever choice (Extance and Goudie, 1981). 219 In test sessions where the operant responding after administration of a test 155 Evaluation and Research [CDER]/Food and Drug Administration compound was markedly suppressed, i.e. 50% decrease in operant responding 220 156 [FDA], 2010; Committee for Medicinal Products for Human Use compared to the mean number of responses in the previous 4 sessions made by the 221 157 [CHMP]/European Medicines Agency [EMA], 2006), and for this same rat when tested with the training cue, i.e. D-amfetamine 0.5 mg/kg, i.p., the test 222 158 reason, lisdexamfetamine and the other reference stimulants were was repeated 1 day later. If the result of 50% decrease in operant responding was 223 confirmed the repeat test, suppressed operant responding was taken as the exper- 159 224 tested in two established rodent models in laboratories where these imental outcome. On the other hand, if on repeat testing the rat showed an 160 protocols have been in use for more than 20 years and for which a acceptable level of operant responding, the percentage generalisation to D-amfet- 225 161 wealth of data and experience with other reference abused and amine was recorded and included in the analysis. When the dose of a test compound 226 162 non-abused drugs exists. In this study, we have explored the selected for testing produced 50% decrease in the operant responding for 50% of a 227 fi “ ” 163 discriminative effects of lisdexamfetamine in rats trained to group of rats, it was classi ed as behavioural disruption and testing at higher doses 228 was not performed. In these experiments, behaviourally disruptive doses of lis- 164 D 229 discriminate between -amfetamine and saline in a 2-choice lever- dexamfetamine and the reference comparators, methylphenidate and D-amfet- 165 pressing model, and its ability to serve as a positive reinforcer in rats amine, were not encountered. In the case of modafinil, only the highest 200 mg/kg, 230 166 trained to intravenously self-administer low-dose cocaine. In these i.p. dose of caused behavioural disruption. 231 167 experiments, the profile of lisdexamfetamine has been compared 232 168 with those of other stimulants that are effective ADHD medications, 2.3. Self-administration training and testing 233 fi 169 i.e. D-amfetamine, methylphenidate and moda nil. Training sessions were conducted on a FR-1 schedule of food reinforcement 234 170 (45 mg dustless pellets; F0021-B, Bilaney Consultants Ltd or PJAI-0045, Noyes Pre- 235 171 2. Methods cision Pellets, Research Diets Inc., New Brunswick, New Jersey, USA.). Operant 236 172 training sessions lasted for a maximum of 1.0 h, or finished once a rat had received 237 173 2.1. Animals and environment 50 food pellet rewards. Once rats had learnt to lever-press to receive 50 pellets in a 238 1.0 h session, the response requirement was increased to FR-2 and the left lever was 174 For the drug-discrimination study, 48 4-week old, female, PVG rats were ob- designated as the active lever. Thereafter only responses on the left lever resulted in 239 175 tained from Harlan UK. The animals were housed in groups of 4 in polypropylene the delivery of a reward. 240

D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11 3

241 Once operant responding for food was stable under the FR-2 schedule, a chronic from SigmaeAldrich and Johnson Matthey-Macfarlan Smith Ltd., cocaine hydro- 306 242 in-dwelling intravenous catheter was implanted into the jugular vein of each rat. chloride from SigmaeAldrich and modafinil base from Tocris Bioscience. 307 Surgery was conducted under aseptic conditions using isoflurane anaesthesia. The The vehicles used for lisdexamfetamine, D-amfetamine, methylphenidate and 243 308 catheter (11 cm IVSAp40 [3 French] silicone; Camcaths Ltd., Cambridge, UK or 10 cm cocaine hydrochloride were 0.9% saline for intraperitoneal (i.p.) or intravenous (i.v.) 244 CBAS-C30 [3 French] heparin-coated polyurethane; Instech Solomon, Plymouth injection and deionised water for oral (p.o.) administration. Modafinil was injected 309 245 Meeting, PA, USA) was implanted into the right jugular vein, secured to the vessel i.p. as a fine suspension in 1% methylcellulose in 0.9% saline and dissolved in 40% (2- 310 246 then tunnelled subcutaneously from the site of insertion to the midscapular region hydroxypropyl)-ß-cyclodextrin [w/v] in deionised water for i.v. injection. 311 247 where the access port exited. The wound was closed with sutures and dressed with Doses of D-amfetamine and methylphenidate are expressed as the mg/kg 312 an appropriate antiseptic spray and plastic dressing. In the case of PMINA-CBAS-C30 of base (correction factors: 1.36 and 1.16, respectively). To facilitate comparisons 248 313 access ports (7 mm high; Instech Solomon), rats were fitted with a mesh jacket between lisdexamfetamine and D-amfetamine, doses of both compounds are 249 (RJ02; Lomir Biomedical, Inc., Malone, New York, USA) to which the exteriorised port expressed as mg/kg of D-amfetamine base (correction factor: 3.37 for 314 250 and catheter were attached. lisdexamfetamine). 315 251 Catheters were filled with heparinised saline immediately post-surgery and after 316 252 every experimental session to maintain catheter patency. Catheter patency was 317 confirmed daily by drawing back and observing freely flowing blood in the catheter 2.6. Data presentation and statistical analysis 253 line. If blood could not be drawn, catheter patency was confirmed by observation of 318 254 immediate sedation upon intravenous injection of propofol or methohexital. If cath- In the drug-discrimination experiments, operant responding after administra- 319 255 eter patency failed during the study, the rat was terminated bya Schedule 1 procedure. tion of test compounds was calculated as the percentage generalisation to the D- 320 256 Animals were allowed to recover from surgery for at least 24 h, after which amfetamine cue for individual rats. 321 animals began the dose-finding study or the self-administration study. The purpose 257 of the dose-finding study was to identify the doses of test compound to be used in % generalisation to d amfetamine for compound C 322 258 the self-administration study. Rats were given a single i.v. injection of either vehicle ¼ number of d amfetamine lever 323 or test compound immediately prior to a 1.0 h operant training session, in which 259 presses in test of compound C=total lever 324 260 they were allowed to respond under a FR-2 schedule for food pellet rewards. During 325 presses in test of compound C 100 261 this session, rats were monitored for effects on lever-pressing or other clear evi- 326 dence of pharmacological activity. where total lever-presses ¼ D-amfetamine lever-presses þ saline lever-presses. 262 Cocaine was selected as the training drug for this study as it is the “gold stan- 327 ED values for generalisation to the D-amfetamine cue were calculated by non- 263 dard” stimulant reinforcer in preclinical self-administration studies. It is a powerful 50 328 linear regression using a logistic model with the minimum fixed at 0% and the reinforcer in humans, primates, rats and many other species. Cocaine is a well known 264 maximum fixed at 100%. Relative potencies of compounds administered orally to the 329 drug of abuse that is a Schedule II Controlled Drug in the USA and UK. D-Amfetamine 265 same compound administered intraperitoneally were calculated using a similar 330 will substitute as a reinforcer for cocaine in rats (Barrett et al., 2004; Liu et al., 2007) non-linear regression model, but forcing the slope to be equal for the two routes of 266 and its pharmacodynamic effect on dopamine efflux in the nucleus accumbens when 331 administration. The relative potency was the ratio of the 2 ED values. If the 95% 267 it is self-administered by rats is similar to that of cocaine (Di Ciano et al., 1995). 50 332 confidence interval for the relative potency did not include 1.0, it indicated a sig- 268 In the self-administration study, each rat was trained to self-administer multiple 333 nificant difference (P < 0.05) between the 2 ED values. injections of a low reinforcing dose of cocaine (0.32 mg/kg/injection, 1 ml/kg/in- 50 269 Results from the self-administration experiments were normally distributed 334 jection, i.v.) on a FR-2 schedule of reinforcement. After an initial non-contingent 270 with equal variance in all groups, and therefore, were considered to be appropriate 335 injection of cocaine (0.32 mg/kg, i.v.), the rats were allowed to lever-press for a for a parametric analysis. A separate statistical analysis was carried out for each dose 271 maximum of 20 injections (in addition to the non-contingent injection)/1.0 h ses- 336 of each test compound. Statistical analysis was performed using the mean of last 3 272 sion/day. After consistent and robust self-administration of cocaine had been 337 sessions for each animal of the test compounds, cocaine and saline. Analysis was by established, low dose cocaine was substituted by saline (1 ml/kg i.v.) to demonstrate 273 2-way analysis of variance with treatment and animal as factors, followed by the 338 extinction of self-administration behaviour. A test compound, i.e. lisdexamfetamine, 274 multiple t-test to compare the 3 treatments (cocaine, saline and test compound, i.e. 339 methylphenidate or modafinil, was then substituted into the paradigm. After lisdexamfetamine, methylphenidate or modafinil) to each other. Positive rein- 275 completing the evaluation of a test compound, the rats were again given access to 340 forcement or Non-reinforcement for the test compound (lisdexamfetamine) and 276 saline (1 ml/kg, i.v.) under a FR-2 schedule in daily 1.0 h sessions until lever-pressing 341 reference comparators (methylphenidate or modafinil) were compared statistically 277 was non-reinforced. Finally, lever-pressing for cocaine HCl (0.32 mg/kg/injection) 342 with the number of vehicle and cocaine injections/session. Positive reinforcement of was retested under a FR-2 schedule to ensure that rats would still self-administer a 278 test compound and reference comparators was defined where the mean number of 343 known drug of abuse with robust positive reinforcing effects. 279 injections/session was significantly greater than the mean number of vehicle in- 344 Stable responding for each rat was accepted when the number of injections/1 h jections and also 8 injections/session. Non-reinforcement of test compound and 280 session for each substance (cocaine, vehicle or test compound) did not vary by 345 reference comparators was defined where the mean number of injections/session 281 more than 20% of the mean of the 3 previous sessions, or where there was no 346 was 8 injections/session and also not different from the injections/session of obvious increasing or decreasing trend in self-administration. The maximum 282 vehicle. 347 283 number of 1 h test sessions employed for test compound or comparator compound 348 if stable responding was not observed was 10 sessions. If stable responding was not 284 349 achieved for an animal by this time, it was eliminated from the study. Positive 285 reinforcement with cocaine was defined as 3 consecutive test sessions where the 3. Results 350 286 mean number of cocaine injections per session was 15. Non-reinforcement with 351 fi 287 vehicle was de ned as 3 consecutive test sessions where the number of saline 3.1. Discriminative profiles of orally administered D-amfetamine 352 injections per session was 8. 288 and methylphenidate 353 289 354 2.4. Experimental locations 290 As shown in Fig. 1, the group of rats employed in this study were 355 291 The drug-discrimination experiments were performed in RenaSci’s laboratories highly proficient in distinguishing the discriminative effects of 356 292 at the University of Nottingham. Self-administration experiments to determine the intraperitoneal injection of 0.5 mg/kg D-amfetamine from saline. 357 reinforcing potential of lisdexamfetamine and methylphenidate were conducted at 293 When rats were given D-amfetamine (0.1e1.5 mg/kg) by the 358 the University of Texas Health Science Center in San Antonio. The bridging experi- 294 ment with methylphenidate and the self-administration experiments with mod- oral route, this stimulant was not recognised as D-amfetamine at 359 295 afinil were performed in RenaSci’s laboratories. the very low dose of 0.1 mg/kg (Fig. 1). Larger doses of the stim- 360 296 All in vivo experiments were performed in strict accordance with Home Office ulant were incrementally recognised as D-amfetamine with partial 361 Guidelines and licenced under the Animals (Scientific Procedures) Act 1986 or in 297 generalisation to the intraperitoneal D-amfetamine cue at doses of 362 accordance with the guidelines of the Institutional Animal Care and Use Committee 298 363 (IACUC), the University of Texas Health Science Center at San Antonio, and the Guide 0.25, 0.5 and 0.75 mg/kg, and full generalisation at a dose of 299 for Care and Use of Laboratory Animals of the Institute of Laboratory Animal Re- 1.5 mg/kg. 364 300 sources, National Research Council, Department of Health, Education and Welfare When tested 15 min after oral administration, methylphenidate 365 301 Publication No. (NIH)83-23, revised 1996. (3.0e10 mg/kg) dose-dependently generalised to the D-amfetamine 366 302 training cue (Fig. 1). Complete generalisation to the D-amfetamine 367 2.5. Drugs and formulation 303 cue was observed at the 10 mg/kg dose of methylphenidate. 368

304 Lisdexamfetamine was provided by Shire Pharmaceuticals. D-Amfetamine sul- None of the oral doses of D-amfetamine or methylphenidate 369 305 phate was purchased from SigmaeAldrich, dl-threo-methylphenidate hydrochloride inﬂuenced the operant response rates of the rats (Table 1). 370

4 D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11

371 d-Amfetamine Methylphenidate 436 372 100 100 437 373 Amfetamine 438 374 75 75 439 SD 375 SD 440 376 50 50 Partial generalisation 441 377 442 25 25 to d-AMF to

378 d-AMF to 443 379 Saline 444 Mean %Mean generalisation 0 %Mean generalisation 0 380 445 381 Sal AMF 0.1 1 10 Sal AMF 1 10 446 Dose (mg/kg po) 382 Dose (mg/kg po) 447 383 448 384 Saline d-Amfetamine ip d-Amfetamine po Saline d-Amfetamine ip Methylphenidate 449 385 450 Fig. 1. Discriminative effects of orally administered D-amfetamine and methylphenidate in rats trained to discriminate D-amfetamine. (0.5 mg/kg, i.p.) from saline. Results are mean 386 values for the percentage generalisation to the D-amfetamine (0.5 mg/kg, i.p.) cue SD. Saline (Sal) and D-amfetamine (0.5 mg/kg, i.p.) (AMF) data are the mean SD from 19 to 24 451 387 rats/group for the 4 test sessions preceding the first oral dose of D-amfetamine or methylphenidate. Groups of rats (n ¼ 6e11) were tested with oral dose of D-amfetamine or 452 388 methylphenidate 15 min after oral dosing. Generalisation was classified as: Amfetamine (75% responses on the D-amfetamine lever), Partial generalisation to D-amfetamine (26e 453 389 74% responses on the D-amfetamine lever) or Saline ( 25% responses on the D-amfetamine lever). 454 390 455 391 456 3.2. Discriminative profile of orally administered lisdexamfetamine partially generalised to D-amfetamine at the higher doses of 1.0 and 392 457 1.5 mg/kg, p.o. (Fig. 2). 393 458 The discriminative effects of orally administered lisdexamfet- None of the oral doses of lisdexamfetamine influenced operant 394 459 amine were investigated in rats trained to discriminate D-amfet- response rates of the rats (Table 1). 395 460 amine (0.5 mg/kg, i.p.) from saline using various time intervals 396 461 between compound dosing and testing. When the interval was 3.3. A comparison of the discriminative profiles of D-amfetamine, 397 462 15 min, cf the interval used for testing orally administered D- methylphenidate and lisdexamfetamine when given by the oral and 398 463 amfetamine and methylphenidate, lisdexamfetamine generalised intraperitoneal routes 399 464 to the saline cue at various pharmacologically active doses ranging 400 465 from 0.5 mg/kg to 1.5 mg/kg (Fig. 2). When the interval was Employing the usual 15 min interval between dosing and 401 466 extended to 60 min, lisdexamfetamine partially generalised to D- testing, intraperitoneal injection of D-amfetamine (0.1e0.5 mg/kg) 402 467 amfetamine at doses of 0.5, 0.75 and 1.0 mg/kg, p.o., and fully produced dose-dependent generalisation to D-amfetamine (Fig. 3). 403 468 generalised at 1.5 mg/kg, p.o. (Fig. 2). When the interval between Switching from the oral to the intraperitoneal route produced a 404 469 dosing and testing was further increased to 120 min, lisdexamfet- leftward shift in the doseeresponse curve (Fig. 3). The ED for 405 50 470 amine generalised to saline at doses of 0.5 and 0.75 mg/kg, p.o. and generalisation to the D-amfetamine training cue was 2.2-fold lower 406 471 407 472 408 Table 1 473 409 Operant response rates of rats after administration of lisdexamfetamine, D-amfetamine, methylphenidate, modafinil or vehicle. 474 410 475 Compound Dose Response rate (mean number of lever-presses/2.5 min S.D. [n value in parentheses]) 411 (mg/kg) 476 412 Oral administration Intraperitoneal injection 477 413 15 min 60 min 120 min 15 min 30 min 60 min 478

414 D-Amfetamine 0.5 31 16 [38] 479 415 Saline e 69 19 [38] 480 416 Lisdexamfetamine 0.5 79 15 [8] 57 14 [7] 75 25 [6] 77 20 [6] 481 417 0.75 88 29 [8] 79 41 [8] 88 30 [6] 82 37 [7] 482 1.0 74 34 [9] 95 37 [9] 76 20 [6] 55 20 [7] 418 1.5 100 26 [6] 57 30 [9] 76 46 [7] 57 34 [6] 483 419 D-Amfetamine 0.7 82 44 [8] 55 22 [6] 484 420 0.25 61 21 [8] 63 34 [6] 485 421 0.5 43 18 [9] 41 17 [10] 486 0.75 70 57 [11] e 422 1.5 40 20 [10] e 487 423 Methylphenidate 0.75 e 57 26 [6] 488 424 1.0 e 63 24 [6] 489 425 1.5 e 46 31 [8] 490 426 3.0 71 23 [6) 52 28 [6] 491 5.0 54 20 [7) e 427 10.0 54 31 [7) e 492 428 Modafinil 50 73 25 [6] e 493 429 100 59 26 [6] 55 14 [6] 494 430 150 95 55 [5/6] 94 38 [6] 495 200 e Dis 431 496 432 Results show the mean operant response rates (number of lever-presses/2.5 min S.D. [n value in parentheses]) when tested at various time-points after administration of 497 fi 433 lisdexamfetamine, D-amfetamine, methylphenidate, moda nil or vehicle. 498 e ¼ not tested. 434 Dis ¼ Behavioural disruption (50% of rats with 50% decrease of operant responding). 499 435 Saline and D-amfetamine (0.5 mg/kg ip) data are the mean of 4 sessions preceding the first dose of test compound for each rat. 500

D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11 5

501 Lisdexamfetamine (ED50 [mg/kg] with 95% CIs; Oral ED50 ¼ 0.86 [0.66; 1.14]; Intra- 566 502 peritoneal ED ¼ 0.83 [0.65; 1.07]). Relative potency with 95% 567 100 50 503 Amfetamine CIs ¼ 0.943 [0.656; 1.356] indicating no significant difference be- 568 504 75 tween the two ED50 values. 569 505 When comparing the influence of route of administration of 570 506 50 Partial generalisation lisdexamfetamine, D-amfetamine and methylphenidate on rates of 571 507 operant responding, no differences were observed between the oral 572 508 25 administration and intraperitoneal injection (Table 1). 573 509 574 0 Saline 510 3.4. Discriminative profile of modafinil determined after 575 511 intraperitoneal administration 576 512 Mean % generalisation to d-AMF 0.1 1 10 577 513 Dose of LDX (mg/kg po) Based on published descriptions of the time-course of mod- 578 514 afinil’s pharmacological effects after intraperitoneal injection (de 579 515 15 minutes after dosing 60 minutes after dosing Saint Hilaire et al., 2001; Paterson et al., 2010; Heal et al., 2012b), 580 516 581 120 minutes after dosing intervals of 30 min and 60 min between dosing and drug- 517 discrimination testing were selected. When tested 30 min after 582 518 Fig. 2. Discriminative effects of orally administered lisdexamfetamine in rats trained to dosing, all of the doses of modafinil, i.e. 50, 100 and 150 mg/kg, i.p., 583 519 discriminate D-amfetamine (0.5 mg/kg, i.p.) from saline. Results are mean values for partially generalised to D-amfetamine (Fig. 4). When tested at 584 520 the percentage generalisation to the D-amfetamine (0.5 mg/kg, i.p.) cue; n ¼ 6e9 rats/ fi 585 e 60 min, the pharmacological effect of moda nil was reduced 521 group. Groups of rats were tested with lisdexamfetamine (0.3 1.5 mg/kg, p.o.) 15, 60 fi 586 or 120 min after dosing in separate experiments. Generalisation was classified as: (Fig. 4). Moda nil (100 mg/kg, i.p.) produced only 28% generalisa- 522 587 Amfetamine, Partial generalisation to D-amfetamine or Saline (For definitions see Fig. 1 tion to amfetamine (2/6 Saline, 2/6 Partial Generalisation; 2/6 523 legend). SD values ranged between 26% and 192% of the mean. Amfetamine) compared with 45% generalisation to D-amfetamine 588 524 at 30 min (5/8 Saline; 2/8 Partial Generalisation; 1/8 Amfetamine). 589 525 fi 590 when D-amfetamine was administered intraperitoneally (ED50 [mg/ Moda nil differed from all of the other stimulants investigated by 526 kg] with 95% confidence intervals [95% CIs]); Oral ¼ 0.39 [0.31; virtue of the fact that irrespective of the dose tested, there were 591 527 0.49]; Intraperitoneal ¼ 0.17 [0.13; 0.23]). Relative potency with substantial inter-animal differences in the perception of its 592 528 95% CIs ¼ 0.419 [0.295, 0.595]. Because the 95% CIs do not include amfetamine-like discriminative effects. 593 529 fi 594 Q1 1.0, the 2 ED50 values were significantly (p < 0.05) different. Intraperitoneally injected doses of moda nil 150 mg/kg had 530 When tested 15 min after dosing, intraperitoneal injection of no effect the operant response rates of the rats (Table 1). 595 531 methylphenidate (0.75e3.0 mg/kg) dose-dependently generalised 596 532 597 to D-amfetamine (Fig. 3). Methylphenidate was 3.4-fold more 3.5. Reinforcing potential of methylphenidate, lisdexamfetamine 533 598 potent when administered by the intraperitoneal versus the oral and modafinil determined in rats using an intravenous self- 534 599 route (ED50 [mg/kg] with 95% CIs; Oral ¼ 4.26 mg/kg [3.75; 4.84]). administration model 535 Intraperitoneal ¼ 1.27 [0.98; 1.63]. Relative potency with 95% 600 536 601 CIs ¼ 0.299 [0.228, 0.394] indicating a significant difference The doses for methylphenidate (0.03, 0.1 and 0.3 mg/kg/injec- 537 602 (p < 0.05) between the 2 ED50 values. tion, i.v.) were selected from previously published scientific data 538 603 When the discriminative effects of lisdexamfetamine were (e.g. Nielsen et al., 1984; Marusich and Bardo, 2009; Burton et al., 539 604 determined 15 min after intraperitoneal injection, the prodrug 2010). An experiment was performed to select pharmacologically 540 605 dose-dependently generalised to D-amfetamine (Fig. 3). Partial active doses of lisdexamfetamine and modafinil to be tested in the 541 606 generalisation was observed at 0.75 and 1.0 mg/kg, i.p. with full intravenous self-administration model. Groups of rats which had 542 607 generalisation at 1.5 mg/kg, i.p. To determine the influence of dose been trained to lever-press consistently for food rewards under a 543 608 route on the potency of lisdexamfetamine, the intraperitoneal ED50 FR-2 schedule of reinforcement were given a single bolus injection 544 609 was compared against the oral ED50 obtained at the prodrug’s time of the test compounds and the rate of operant responding for food 545 610 of peak effect, i.e. with a 60 min interval between dosing and was monitored over the following 60 min or until 50 food pellet 546 611 testing. Switching from the oral to the intraperitoneal route of rewards had been collected. Lisdexamfetamine decreased the rate 547 612 administration did not increase the potency of lisdexamfetamine of operant responding by 20% and 28% at doses of 0.05 and 0.15 mg/ 548 613 549 614 125 125 Methylphenidate 550 Lisdexamfetamine 100 d-Amfetamine 615 100 100 551 Amfetamine 616 SD 75 552 75 75 617 553 50 618 50 50 554 Partial generalisation 619

555 d-AMF to 25 25 25 620

556 %Mean generalisation 621 0 0 Saline 557 0 622 558 0.1 1 10 0.01 0.1 1 10 0.1 1 10 100 623 559 Dose (mg/kg) Dose (mg/kg) Dose (mg/kg) 624 560 IP 15 minutes after dosing PO 15 minutes after dosing IP 15 minutes after dosing 625 561 PO 60 minutes after dosing IP 15 minutes after dosing PO 15 minutes after dosing 626 562 627 563 Fig. 3. Comparison of oral versus intraperitoneal potency of lisdexamfetamine, D-amfetamine and methylphenidate in rats trained to discriminate D-amfetamine (0.5 mg/kg, i.p.) 628 from saline. Results are mean values for the percentage generalisation to the D-amfetamine (0.5 mg/kg, i.p.) cue SD; n ¼ 6e11 rats/group. For lisdexamfetamine, its potency when 564 administered by the intraperitoneal route was compared against its oral potency determined at the prodrug’s time of peak effect, i.e. 60 min after dosing. Generalisation was 629 565 classiﬁed as: Amfetamine, Partial generalisation to D-amfetamine or Saline (For deﬁnitions see Fig. 1 legend). 630

6 D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11

631 positive reinforcer and it supported the same level of self- 696 125 Modafinil 632 administration responding in both laboratories. 697 ﬁ 633 100 Amfetamine Moda nil (0.166, 0.498 or 1.66 mg/kg/injection, i.v.) did not 698

634 SD maintain self-administration at levels signiﬁcantly above saline at 699 635 75 any dose (Fig. 5) showing that at these doses it did not serve as a 700 636 positive reinforcer in rats. When the data were assessed in indi- 701 50 Partial generalisation 637 vidual animals, 3/9 (33%) rats self-administered the high dose of 702

638 d-AMF to 25 modaﬁnil (1.66 mg/kg/injection, i.v.) at levels above saline. How- 703 639 ever, in each case, the rats showed highly variable intakes of 704

Mean % generalisation Saline 640 0 modafinil and all rats were tested for the maximum of 10 sessions 705 641 without achieving consistent responding on the drug. 706 10 100 1000 642 707 Dose (mg/kg ip) Lisdexamfetamine (0.05, 0.15 or 0.5 mg/kg/injection, i.v.) did not 643 maintain self-administration at levels significantly above saline at 708 644 IP 30 minutes after dosing any dose, (Fig. 5) showing that it did not serve as a positive rein- 709 645 IP 60 minutes after dosing forcer in rats. When the data for individual animals were analysed 710 646 2/10 (20%) and 1/9 (11%) rats self-administered the 0.15 mg/kg and 711 647 Fig. 4. Discriminative effects of modafinil determined in rats trained to discriminate 0.5 mg/kg doses of lisdexamfetamine at numerically greater levels 712 D-amfetamine (0.5 mg/kg, i.p.) from saline. Results are mean values for the percentage 648 than saline. 713 generalisation to the D-amfetamine (0.5 mg/kg, i.p.) cue SD; n ¼ 6e9 rats/group. 649 Groups of rats were tested 30 or 60 min after intraperitoneal dosing in separate ex- 714 650 periments. Generalisation was classified as: Amfetamine, Partial generalisation to 4. Discussion 715 651 D-amfetamine or Saline (For definitions see Fig. 1 legend). 716 652 Drug-discrimination and intravenous self-administration are 717 653 kg, i.v. (Table 2). Lisdexamfetamine showed a non-linear dosee well established models to compare the similarity of the discrimi- 718 654 response curve as it did not change the rate of operant responding native and positive reinforcing effects of novel centrally-acting 719 655 when administered at the lowest and highest doses of 0.015 and drugs to those of known substances of abuse (Johanson, 1990; 720 656 0.5 mg/kg, i.v. (Table 2). Based on these results, lisdexamfetamine Balster, 1991; Ator and Griffiths, 2003; Solinas et al., 2006). 721 657 doses of 0.05, 0.15 and 0.5 mg/kg/injection were selected for the The drug-discrimination procedure employing female, PVG rats 722 658 self-administration experiment. Modafinil (0.166, 0.498 and trained to discriminate D-amfetamine from saline has been exten- 723 659 1.66 mg/kg, i.v.) incrementally increased the rate of operant sively characterised with a range of monoaminergic drugs with and 724 660 responding across the 3 doses (Table 2), and therefore, these doses without liability for recreational abuse (Heal et al., 1992; Gosden 725 661 of modafinil were selected for the self-administration experiment. et al., 1996). In general, this model has good predictive validity 726 662 Cocaine (0.32 mg/kg/injection, i.v.) maintained high rates of self- for detecting stimulants with high to low levels of recreational 727 663 administration in all rats (p < 0.001 versus saline for all groups of abuse liability; however, some catecholamine reuptake inhibitors, 728 664 test compounds), whereas saline (i.v.) maintained only low rates of e.g. bupropion and nomifensine, show up as false positives (Heal 729 665 self-administration (8 injections/session) in all rats before and et al., 1992; Gosden et al., 1996). In this model, D-amfetamine and 730 666 after testing of lisdexamfetamine, methylphenidate or modafinil methylphenidate both dose-dependently generalised to the 731 667 (Fig. 5). discriminative cue elicited by intraperitoneal injection of D-amfet- 732 668 Although methylphenidate was not self-administered when amine. This finding is consistent with the cross-generalisation of 733 669 tested at 0.03 mg/kg/injection, doses of 0.1 and 0.3 mg/kg/injection the discriminative cues of these two stimulants as determined in 734 670 maintained levels of self-administration significantly above saline both rats (Witkin et al., 1991; Gosden et al., 1996; Craft and 735 671 (Fig. 5). Stratmann, 1996; Kollins et al., 2001; Stadler et al., 2001; Desai 736 672 To ensure that the data were consistent across both laboratories et al., 2010) and human subjects (Martin et al., 1971; Smith and 737 673 participating in this study, methylphenidate (0.1 mg/kg/injection, Davis, 1977; Heishman and Henningfield, 1991; Rush et al., 1998). 738 674 i.v.) was retested as a validation experiment. Cocaine (0.32 mg/kg/ As a prodrug, lisdexamfetamine is pharmacologically inactive 739 675 injection, i.v.) and saline respectively served as a positive reinforcer and its enzymatic conversion to yield the active moiety, D-amfet- 740 676 and non-reinforcer in this group of rats, and as shown in Table 3, amine, and the naturally occurring amino acid L-lysine is unusual 741 677 methylphenidate (0.1 mg/kg/injection, i.v.) was found to act as a because it is mediated by a rate-limited enzymatic hydrolysis that is 742 678 almost exclusively carried out by red blood cells (Pennick, 2010). The 743 679 metabolic route profoundly influences the pharmacokinetics of lis- 744 680 dexamfetamine’s active metabolite, which in turn influences its 745 681 Table 2 pharmacodynamic profile. In rats, plasma exposure to D-amfetamine 746 682 Effect of various doses of lisdexamfetamine and modafinil on FR-2 operant after administration of lisdexamfetamine was not different when 747 responding for food rewards. 683 compared with immediate release (IR) D-amfetamine, but the Cmax 748 684 Treatment n Mean number of % was 50% lower and the Tmax was doubled (Rowley et al., 2012). As a 749 685 lever-presses/ Baseline result of its unusual pharmacokinetics, lisdexamfetamine was 750 min SEM 686 shown to be markedly less stimulant than IR D-amfetamine when 751 687 Vehicle (Saline, i.v.) 4 39.0 2.4 e tested at equivalent doses in terms of D-amfetamine base and the 752 688 Lisdexamfetamine (0.015 mg/kg, i.v.) 4 39.0 2.4 100 time of maximum activation was substantially delayed (Rowley 753 689 Lisdexamfetamine (0.05 mg/kg, i.v.) 4 31.2 3.0 80 et al., 2012). However, this prodrug produced substantial motor 754 Lisdexamfetamine (0.15 mg/kg, i.v.) 4 28.2 4.8 72 690 Lisdexamfetamine (0.5 mg/kg, i.v.) 4 39.0 1.2 100 activation in rats by administering very high doses of lisdexamfet- 755 691 Vehicle (40% (2-hydroxypropyl)-b- 3 7.5 3.5 e amine (Rowley et al., 2012). Intracerebral microdialysis experiments 756 692 cyclodextrin, i.v.) performed in rats have revealed that lisdexamfetamine dose- 757 693 Modafinil (0.166 mg/kg, i.v.) 3 11.7 1.9 155 dependently increases the extraneuronal concentrations 758 Modafinil (0.498 mg/kg, i.v.) 3 14.7 3.7 195 694 of dopamine and noradrenaline in the prefrontal cortex (PFC) with 759 Modafinil (1.66 mg/kg, i.v.) 3 17.2 2.2 228 695 effects of equal magnitude on both catecholamine 760

D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11 7

761 Methylphenidate Modafinil Lisdexamfetamine 826 762 827 ### 763 20 20 20 828 764 829 18 18 18 765 *** 830 16 16 16 766 ### 831 767 14 14 14 832

768 12 12 12 833 769 834 10 10 10 770 835 8 8 8 771 *** *** 836 772 6 6 *** *** 6 837 773 *** 838 4 *** 4 4 *** 774 839 2 2 2 775 Mean number of injections/session ± SEM 840 776 0 0 0 841 777 Cocaine Vehicle 0.03 0.1 0.3 Cocaine Vehicle 0.166 0.498 1.66 Cocaine Vehicle 0.05 0.15 0.5 842 778 Dose of Methylphenidate Dose of Modafinil Dose of Lisdexamphetamine 843 (mg/kg/inj) (mg/kg/inj) 779 (mg/kg/inj) 844

780 Fig. 5. Evaluation of the possible reinforcing effects of various doses of methylphenidate, lisdexamfetamine and modafinil using a FR-2 schedule of drug reinforcement in rats 845 781 compared with cocaine (positive reinforcer) and saline (nonreinforcer). Results are the mean number of injections/session SEM for rats responding under a FR-2 schedule of 846 782 intravenous drug reinforcement obtained either (a) in the last 3 sessions for each condition where responding was stable (see Methods for criteria) or (b) over 10 test sessions where 847 783 responding was unstable. Results were analysed by multiple t-test using the average values from before and after test compound administration. Significantly different from saline: 848 ###p < 0.001. Significantly different from cocaine: ***p < 0.001.Cocaine and saline: closed symbols ¼ result obtained before evaluation of test compound; open symbols ¼ result 784 849 obtained after evaluation of test compound. Q4 785 850 786 851 787 neurotransmitters (Heal et al., 2012b). The D-amfetamine prodrug experiments performed in vitro and in vivo indicate that it has 852 788 also dose-dependently increased dopamine efflux in the striatum w5 mMaffinity for the human dopamine reuptake transporter 853 789 (Rowley et al., 2012; Heal et al., 2012b). Although this profile is (DAT) (Madras et al., 2006; Zolkowska et al., 2009). Despite this fact, 854 790 consistent with lisdexamfetamine producing its pharmacological Volkow et al. (2009) has reported that clinical doses of modafinil, 855 791 effects via D-amfetamine, clear differences between lisdexamfet- i.e. 200 and 400 mg, occupied w46% of DAT sites in the caudate, 856 792 amine and the IR formulations of D-amfetamine and methylpheni- w53% in the putamen and w60% in the nucleus accumbens in 857 793 date were observed in these experiments. In both microdialysis human subjects in vivo. Intracerebral microdialysis experiments 858 794 studies, the locomotor activity of the rats was monitored simulta- have revealed that modafinil increased the extracellular concen- 859 795 neously with dialysate collection and the results unequivocally trations of noradrenaline and dopamine in the PFC (de Saint Hilaire 860 796 demonstrated that lisdexamfetamine was different from either D- et al., 2001; Rowley et al., 2012), dopamine and 5-HT in the nucleus 861 797 amfetamine and methylphenidate by its ability to produce large and accumbens (Zolkowska et al., 2009), dopamine in the striatum 862 798 sustained increase in striatal dopamine efflux whilst producing only (Rowley et al., 2012), and noradrenaline and dopamine in the ros- 863 799 minimal behavioural activation (Rowley et al., 2012; Heal et al., tromedial hypothalamus (de Saint Hilaire et al., 2001). The finding 864 800 2012b). that modafinil can enhance the extracellular concentration of 865 801 Consistent with these findings, oral doses of lisdexamfetamine dopamine in the rat brain are supported by microdialysis de- 866 802 5.06 mg/kg (equivalent to 1.5 mg/kg of D-amfetamine base), were terminations of striatal dopamine in rhesus monkeys (Andersen 867 803 not recognised as amfetamine-like on the D-amfetamine-cued drug- et al., 2010) and positron emission tomography (PET) experi- 868 11 804 discrimination model when tested 15 min after dosing. When the ments showing the displacement of [ C]raclopride from striatal D2 869 805 interval between dosing and testing was increased to 60 min, the receptors by dopamine in human subjects (Volkow et al., 2009). 870 806 same doses of lisdexamfetamine dose-dependently generalised to Because of its weak potency and very poor solubility, the 871 807 the D-amfetamine cue. Extending the interval to 120 min diminished discriminative effects of modafinil in the D-amfetamine-cued drug- 872 808 the stimulant discriminative effect of lisdexamfetamine as shown by discrimination test were determined only after intraperitoneal in- 873 809 the result that the prodrug no longer generalized fully to the jection of the drug in suspension. In this model, modafinil partially 874 810 D-amfetamine cue. Together, these findings indicate that the generalised to D-amfetamine at doses ranging from 50 to 200 mg/kg 875 811 amfetamine-like discriminative effects of orally administered lis- indicating that it produces some D-amfetamine-like discriminative 876 812 dexamfetamine are delayed in onset and of relatively short duration. effects. However, this atypical stimulant never generalised fully to 877 813 Modafinil has an enigmatic pharmacological mechanism of ac- the D-amfetamine cue. In contrast to all of the other drugs tested, 878 814 tion (Minzenberg and Carter, 2008), but neurochemical substantial inter-animal variability in responding was present for 879 815 all of the doses tested revealing that modafinil was clearly recog- 880 816 nised as D-amfetamine-like by some rats, but different from D- 881 817 Table 3 amfetamine by others. 882 A comparison of the intravenous self-administration profile of cocaine, methyl- 818 phenidate and saline in the two participating laboratories. The only other comparison of the discriminative effects of 883 819 modafinil and D-amfetamine was performed by Dopheide et al. 884 820 Research facility Number of injections/session SEM (2007), who also observed that modafinil partially, but not fully, 885 821 Cocaine Saline Methylphenidate generalised to D-amfetamine with considerable inter-individual 886 822 (0.32 mg/kg/inj) (0.1 mg/kg/inj) variability in the responses of the rats. (Dopheide et al., 2007) 887 823 University of Texas 19.4 0.2 (n ¼ 8) 2.8 0.6 (n ¼ 8) 16.9 2.3 (n ¼ 8) also observed that modafinil partially generalised to the cocaine 888 824 RenaSci University 18.6 1.1 (n ¼ 4) 5.3 0.7 (n ¼ 4) 18.0 2.4 (n ¼ 4) discriminative cue in a separate group of rats trained to distinguish 889 825 of Nottingham between cocaine and saline. Partial or full substitution for cocaine 890

8 D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11

891 by modafinil has been observed in several other investigations The ability of lisdexamfetamine, methylphenidate and modafinil 956 892 performed in rats (Gold and Balster, 1996; Newman et al., 2010; to serve as positive reinforcers were compared in rats trained to 957 893 Paterson et al., 2010) and in rhesus monkeys (Gold and Balster, intravenously self-administer low dose cocaine. In self- 958 894 1996). It has previously been reported that D-amfetamine and administration experiments, this route is routinely employed 959 895 cocaine share a common discriminative cue in rats (Witkin et al., because it is the one that carries the greatest safety risks for rec- 960 896 1991; Gosden et al., 1996; Craft and Stratmann, 1996; Gold and reational drug abusers. Cocaine served as a robust positive rein- 961 897 Balster, 1996; Stadler et al., 2001). forcer in all of the rats with animals consistently taking >15 962 898 In drug-experienced human volunteers, who had been trained to injections/session. In contrast, saline, which was used as the pla- 963 899 discriminate between cocaine and placebo, modafinil generalised cebo control in these experiments, maintained only low levels of 964 900 partially (w60%) to the cocaine cue whilst methylphenidate pro- self-administration, i.e. generally <5 injections/session. Methyl- 965 901 duced high levels of generalisation to cocaine in most subjects (Rush phenidate served as a robust positive reinforcer at very low doses, 966 902 et al., 2002a). Modafinil also evoked subjective ratings of “high”, i.e. 0.1 and 0.3 mg/kg/injection and this finding agrees with 967 903 “take again”, “good effects” and moderate “drug liking” (Rush et al., numerous previous reports that this stimulant maintains high 968 904 2002a). Two other studies performed in drug-experienced humans levels of self-administration in rodents (Nielsen et al., 1984; Botly 969 905 also found that modafinil produced stimulant-like subjective effects et al., 2008; Marusich et al., 2010), and primates (Johanson and Q2 970 906 and had positive reinforcing properties (Jasinski, 2000; Stoops et al., Schuster, 1975; Bergman et al., 1989; Schindler et al., 2011). 971 907 2005). When all of the animal and human data are taken into Consistent with the preclinical findings, methylphenidate has been 972 908 consideration, theylead to the conclusion that modafinil differs from found to be a powerful reinforcer in many double-blind, placebo- 973 909 D-amfetamine, methylphenidate and lisdexamfetamine by virtue of controlled trials in human subjects (Smith and Davis, 1977; Kollins 974 910 eliciting a discriminative stimulus that is much weaker than that et al., 1998; Volkow et al., 1999; Jasinski, 2000; Rush and Baker, 975 911 evoked by these other stimulants. 2001; Stoops et al., 2003, 2004; Spencer et al., 2006). In an inter- 976 912 The rate at which stimulants and other substances of abuse enter esting study, Kollins et al. (1998) compared the subjective and 977 913 the brain is an important factor in determining their liability for reinforcing properties of a sustained release (SR) and IR formula- 978 914 recreational abuse (Volkow et al., 1996, 2003; Kollins et al., 1998; tion of methylphenidate in healthy volunteers and observed that 979 915 Gorelick, 1998; Cone, 1998). To maximise both the quantity and the stimulant and reinforcing effects of the former were attenuated 980 916 rate of entry of drugs into the brain, some abusers resort to non- and transient compared with the latter leading the authors to 981 917 clinical routes for self-administration including intravenous injec- conclude that the SR formulation posed a reduced risk for recrea- 982 918 tion, “snorting” and smoking. These approaches increase the chance tional abuse. To summarise, therefore, methylphenidate produces 983 919 of overdose and expose the individual to other serious hazards, e.g. both the subjective and reinforcing effects in humans that are 984 920 hepatitis or HIV infection from unclean needles. In this investiga- typical of stimulants like cocaine and D-amfetamine supporting the 985 921 tion, the influence of dosing route on potency was explored in the view that the preclinical results obtained in these drug- 986 922 rat drug-discrimination model by comparing doseeresponse curves discrimination and self-administration experiments have good 987 923 obtained when the stimulants were given orally and by intraperi- translational validity. 988 924 toneal injection. When given by the oral route, the ED50 values for Modafinil did not maintain self-administration at levels above 989 925 generalisation to the training cue revealed that D-amfetamine was saline when tested across a range of pharmacologically active doses 990 926 w10-fold more potent than methylphenidate. When the route was indicating that this atypical stimulant does not serve as a positive 991 927 switched to intraperitoneal injection, the potency of D-amfetamine reinforcer in cocaine-trained rats. Analogous to results obtained 992 928 increased 2.2-fold, but the potency of methylphenidate was from the drug-discrimination model, several of the rats showed 993 929 increased by 3.4-fold. When the finding that methylphenidate highly variable intakes of modafinil over the 10 sessions of the self- 994 930 maintained robust self-administration at a dose of 0.1 mg/kg is also administration experiment. In these individuals, the average 995 931 taken into consideration, it indicates that even greater increases in number of injections/session taken was numerically greater than 996 932 methylphenidate’s potency would occur when the intravenous the saline control value, which hints that modafinil may act as a 997 933 route is selected. In lisdexamfetamine’s case, switching from the weak reinforcer in a minority of rats. Although modafinil’s very 998 934 oral to the intraperitoneal route decreased the delay before the poor solubility limited the dose that could be evaluated in the rats, 999 935 prodrug’s amfetamine-like discriminative effects could be detected there was no suggestion in the results that higher doses of mod- 1000 936 by the rats, but intraperitoneal injection did not enhance the pro- afinil elicited greater levels of self-administration. That being said, 1001 937 drug’s potency. This lack of influence of the dose route on the po- it would be unwise to exclude the possibility that modafinil would 1002 938 tency of lisdexamfetamine extends to humans where the not serve as a positive reinforcer in rats under any circumstance. 1003 939 pharmacokinetics of lisdexamfetamine’s active metabolite, D- The results reported here are in general agreement with those of 1004 940 amfetamine, are not altered when nasally insufflated compared other determinations of modafinil’s reinforcing effect in rats. 1005 941 with the clinically relevant oral route (Ermer et al., 2011). Jasinski Modafinil failed to serve as a positive reinforcer in an intravenous 1006 942 and Krishnan performed trials in drug-experienced, human vol- self-administration procedure employing drug-naïve rats, and in 1007 943 unteers where the subjective and reinforcing effects of lisdex- addition, it did not induce place preference in rats (Deroche- 1008 944 amfetamine were evaluated when it was given orally (Jasinski and Gamonet et al., 2002). In contrast, D-amfetamine, which was 1009 945 Krishnan, 2009a) or by intravenous injection (Jasinski and employed as the positive control, produced a clear preference for 1010 946 Krishnan, 2009b). A recent post hoc analysis of results taken from the drug-associated compartment (Deroche-Gamonet et al., 2002). 1011 947 these two trials showed that the potency of lisdexamfetamine was On the other hand, Gold and Balster (1996) reported that modafinil 1012 948 unchanged in human subjects irrespective of whether the oral or unequivocally maintained intravenous self-administration at levels 1013 949 the intravenous route was employed (Heal et al., 2013). The expla- above saline in a group of 4 cocaine-trained rhesus monkeys. 1014 950 nation is the rate-limited enzymatic hydrolysis of lisdexamfetamine Additional evidence for similarity between the reinforcing prop- 1015 951 that is carried out by peptidases associated with red blood cells erties of modafinil and cocaine come from the findings that mod- 1016 952 dictates that the gradual and sustained liberation of D-amfetamine afinil reinstated both cocaine-conditioned place preference in rats 1017 953 cannot be accelerated by loading the blood-stream with the parent (Bernardi et al., 2009) and cocaine self-administration in rhesus 1018 954 compound. The outcome is lisdexamfetamine’s potency remains monkeys (Andersen et al., 2010). In drug-experienced volunteers, 1019 955 relatively constant irrespective of its route of administration. modafinil produced stimulant-like subjective effects and acted as a 1020

D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11 9

1021 reinforcer (Jasinski, 2000; Rush et al., 2002a). Stoops et al. (2005) effects were transient and this prodrug did not serve as a positive 1086 1022 reported that modafinil served as a reinforcer in human volun- reinforcer in rats. In all of the above respects, lisdexamfetamine was 1087 1023 teers performing a cognitive task, although not when they were clearly differentiated from IR methylphenidate. Although these 1088 1024 relaxing. However, the weakness of modafinil’s reinforcing effect preclinical experiments have explored lisdexamfetamine’s poten- 1089 1025 relative to cocaine was demonstrated by its failure to serve as a tial for recreational abuse, they do not address the issue of non- 1090 1026 positive reinforcer in stimulant abusers (Rush et al., 2002b; clinical misuse. 1091 1027 Vosburg et al., 2010) and its lack of clinical efficacy in the treat- 1092 1028 ment of cocaine dependence, showing that it has low reinforcing 1093 Acknowledgements 1029 effects of its own (Anderson et al., 2009). 1094 1030 Lisdexamfetamine did not maintain intravenous self- 1095 This study was conducted with financial support from Shire 1031 administration at levels greater than saline when tested across a 1096 Pharmaceutical Research Development, UK. The authors wish to 1032 range doses indicating that this prodrug does not serve as a positive 1097 thank Shire Pharmaceutical Research Development for financial 1033 reinforcer in cocaine-trained rats. As indicated by the reduced op- 1098 support in funding a portion of the writers’ time for preparation of 1034 erant responding for food rewards, the doses employed were 1099 this manuscript. 1035 pharmacologically active. Moreover, as the potency of this prodrug 1100 The authors wish to state that the opinions expressed in this 1036 is not materially influenced by its route of administration, a non- 1101 manuscript reflect their own views and do not necessarily reflect 1037 contingent injection of the highest dose, i.e. 0.5 mg/kg/injection, 1102 those of Shire Pharmaceuticals. 1038 which was employed to initiate the self-administration session, 1103 1039 would almost certainly have elicited partial generalisation to the 1104 1040 training cue if it had been tested in the D-amfetamine-cued drug- References 1105 1041 discrimination model. Although the reinforcing properties of lis- 1106 Aigner, T.G., Balster, R.L., 1979. Rapid substitution procedure for intravenous drug 1042 ’ D 1107 dexafetamine s metabolite, -amfetamine, were not determined in self-administration studies in rhesus monkeys. Pharmacol. Biochem. Behav. 10, 1043 this study, the latter has been consistently reported to maintain 105e112. 1108 1044 robust intravenous self-administration in rats (Yokel and Wise, Andersen, M.L., Kessler, E., Murnane, K.S., McClung, J.C., Tufik, S., Howell, L.L., 2010. 1109 fi 1045 1978; Di Ciano et al., 1995; Carroll and Lac, 1997; Crombag et al., Dopamine transporter-related effects of moda nil in rhesus monkeys. Psy- 1110 chopharmacology 210, 439e448. 1046 2008) and primates (Johanson et al., 1976; Aigner and Balster, Anderson, A.L., Reid, M.S., Li, S.-H., Holmes, T., Shemanski, L., Slee, A., Smith, E.V., 1111 1047 1979). Furthermore, D-amfetamine substituted for cocaine as a Kahn, R., Chiang, N., Vocci, F., Ciraulo, D., Dackis, C., Roache, J.D., Salloum, I.M., 1112 fi 1048 positive reinforcer in primates (Johanson et al., 1976) and also Somoza, E., Urschell III, H.C., Elkashef, A.M., 2009. Moda nil for the treatment of 1113 cocaine dependence. Drug Alcohol Depend. 104, 133. 1049 reinstated cocaine self-administration behaviour in rats (de Wit Ator, N.A., Griffiths, R.R., 2003. Principles of drug abuse liability assessment in 1114 1050 and Stewart, 1981; Suto et al., 2002). laboratory animals. Drug Alcohol Depend. 70, S55eS72. 1115 1051 In drug-experienced human volunteers, lisdexamfetamine was Balster, R.L., 1991. Drug abuse potential evaluation in animals. Br. J. Addict. 86, 1116 1549e1558. 1052 less potent than D-amfetamine (Jasinski and Krishnan, 2009a)in Barrett, A.C., Miller, J.R., Dohrmann, J.M., Caine, S.B., 2004. Effects of dopamine in- 1117 1053 producing amfetamine-like stimulant effects and there was a sub- direct agonists and selective D1-like and D2-like agonists and antagonists on 1118 1054 stantial delay in their appearance. However, consistent with the cocaine self-administration and food maintained responding in rats. Neuro- 1119 pharmacology 47 (Suppl. 1), 256e273. 1055 data from the rodent experiments (Rowley et al., 2012; Heal et al., Bergman, J., Madras, B.K., Johnson, S.E., Spealman, R.D., 1989. Effects of cocaine and 1120 1056 2012b; this study), lisdexamfetamine nonetheless has the ability related drugs in nonhuman primates. III. Self-administration by squirrel mon- 1121 1057 elicit stimulant pharmacological effects in man when given at high keys. J. Pharmacol. Exp. Ther. 251, 150e156. 1122 fi 1058 Bernardi, R.E., Lewis, J.R., Lattal, K.M., Berger, S.P., 2009. Moda nil reinstates a 1123 doses (Jasinski and Krishnan, 2009a). cocaine conditioned place preference following extinction in rats. Behav. Brain 1059 Predictions on whether a compound will be subjected to Res. 204, 250e253. 1124 1060 diversion and recreational abuse based on results from either Biederman, J., Swanson, J.M., Wigal, S.B., Boellner, S.W., Earl, C.Q., Lopez, F.A., 2006. 1125 fi 1061 preclinical experiments or trials in drug-experienced human vol- A comparison of once-daily and divided doses of moda nil in children 1126 attention-deficit/hyperactivity disorder: a randomized, double-blind and 1062 unteers is uncertain because many other factors come into play in placebo-controlled study. J. Clin. Psychiatry 67, 727e735. 1127 1063 the real world. Examples include how well the euphoriant and Botly, L.C., Burton, C.L., Rizos, Z., Fletcher, P.J., 2008. Characterization of methyl- 1128 1064 stimulant profile of the drug fits with the culture and context of phenidate self-administration and reinstatement in the rat. Psychopharmacol- 1129 ogy 199, 55e66. 1065 recreational abuse which varies widely from country to country, Buitelaar, J., Medori, R., 2010. Treating attention-deficit/hyperactivity disorder 1130 1066 the availability and cost of alternatives, its pharmacokinetic profile, beyond symptom control alone in children and adolescents: a review of the 1131 1067 and the ability to substantially increase the drug’s psychostimulant potential benefits of long-acting stimulants. Eur. Child Adolesc. Psychiatry 19, 1132 325e340. 1068 experience by employing non-clinical routes of self-administration. Burton, C.L., Nobrega, J.N., Fletcher, P.J., 2010. The effects of adolescent methyl- 1133 1069 In the case of modafinil, the preclinical and clinical data reveal that phenidate self-administration on responding for a conditioned reward, 1134 1070 its pharmacological profile is unlikely to make it attractive as a amfetamine-induced locomotor activity, and neuronal activation. Psychophar- 1135 macology 208, 455e468. 1071 recreational drug of abuse, but it does possess cognitive enhancing Carroll, M.E., Lac, S.T., 1997. Acquisition of i.v. amfetamine and cocaine self- 1136 1072 and wake-promoting properties that would support non-medical administration in rats as a function of dose. Psychopharmacology 129, 206e214. 1137 1073 misuse. In contrast, methylphenidate is a well established, stimu- Center for Drug Evaluation and Research (CDER)/Food and Drug Administration 1138 1074 (FDA), 2010. Guidance for Industry Assessment of Abuse Potential of Drugs 1139 lant euphoriant with powerful reinforcing properties that present a (Draft Guidance). http://www.fda.gov/downloads/Drugs/ 1075 substantial risk for recreational abuse. To some extent, this has GuidanceComplianceRegulatoryInformation/Guidances/ucm079655.pdf. 1140 1076 been counterbalanced by the development of long-acting prepa- Committee for Medicinal Products for Human Use (CHMP)/European Medicines 1141 1077 rations of methylphenidate that attenuate its stimulant potential Agency (EMA), 2006. Guideline on the Non-Clinical Investigation of the 1142 Dependence Potential of Medicinal Products. http://www.ema.europa.eu/docs/ 1078 (Kollins et al., 1998) and tamper-deterrent formulations and de- en_GB/document_library/Scientific_guideline/2009/09/WC500003360.pdf. 1143 1079 livery systems that increase the difficulty of recreational abuse by Cone, E.J., 1998. Recent discoveries in pharmacokinetics of drugs of abuse. Toxicol. 1144 e e 1080 snorting or intravenous injection. The unusual metabolic activation Lett. 102 103, 97 101. 1145 Craft, R.M., Stratmann, J.A., 1996. Discriminative stimulus effects of cocaine in fe- 1081 of lisdexamfetamine predicts that its recreational abuse liability male versus male rats. Drug Alcohol Depend. 42, 27e37. 1146 1082 will not be influenced by changing its route of administration Crombag, H.S., Ferrario, C.R., Robinson, T.E., 2008. The rate of intravenous cocaine or 1147 1083 (Jasinski and Krishnan, 2009a,b; Ermer et al., 2011; Heal et al., 2013; amfetamine delivery does not influence drug-taking and drug-seeking behavior 1148 in rats. Pharmacol. Biochem. Behav. 90, 797e804. 1084 this study). Furthermore, although lisdexamfetamine generalised de Saint Hilaire, Z., Orosco, M., Rouch, C., Blanc, G., Nicolaidis, S., 2001. Variations in 1149 1085 fully to D-amfetamine in the drug-discrimination test, its stimulant extracellular monoamines in the prefrontal cortex and medial hypothalamus 1150

10 D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11

1151 after modafinil administration: a microdialysis study in rats. Neuroreport 12, occupies dopamine and norepinephrine transporters in vivo and modulates 1216 e 1152 3533 3537. the transporters and trace amine activity in vitro. J. Pharmacol. Exp. Ther. 1217 de Wit, H., Stewart, J., 1981. Reinstatement of cocaine-reinforced responding in the 319, 561e569. 1153 rat. Psychopharmacology 75, 134e143. Martin, W.R., Sloan, J.W., Sapira, J.D., Jasinski, D.R., 1971. Physiologic, subjective, and 1218 1154 Deroche-Gamonet, V., Darnaudéry, M., Bruins-Slot, L., Piat, F., Le Moal, M., behavioral effects of amfetamine, methamfetamine, ephedrine, phenmetrazine, 1219 1155 Piazza, P.V., 2002. Study of the addictive potential of modafinil in naïve and and methylphenidate in man. Clin. Pharmacol. Ther. 12, 245e258. 1220 cocaine-experienced rats. Psychopharmacology 161, 387e395. Marusich, J.A., Bardo, M.T., 2009. Differences in impulsivity on a delay-discounting 1156 Desai, R.I., Paronis, C.A., Martin, J., Desai, R., Bergman, J., 2010. Monoaminergic task predict self-administration of a low unit dose of methylphenidate in rats. 1221 1157 psychomotor stimulants: discriminative stimulus effects and dopamine efflux. Behav. Pharmacol. 20, 447e454. 1222 1158 J. Pharmacol. Exp. Ther. 333, 834e843. Minzenberg, M.J., Carter, C.S., 2008. Modafinil: a review of neurochemical actions 1223 Di Ciano, P., Coury, A., Depoortere, R.Y., Egilmez, Y., Lane, J.D., Emmett- and effects on cognition. Neuropsychopharmacology 33, 1477e1502. 1159 Oglesby, M.W., Lepiane, F.G., Phillips, A.G., Blaha, C.D., 1995. Comparison of Newman, J.L., Negus, S.S., Lozama, A., Prisinzano, T.E., Mello, N.K., 2010. Behavioral 1224 1160 changes in extracellular dopamine concentrations in the nucleus accumbens evaluation of modafinil and the abuse-related effects of cocaine in rhesus 1225 1161 during intravenous self-administration of cocaine or d-amfetamine. Behav. monkeys. Exp. Clin. Psychopharmacol. 18, 395e408. 1226 e 1162 Pharmacol. 6, 311 322. Nielsen, J.A., Duda, N.J., Mokler, D.J., Moore, K.E., 1984. Self-administration of central 1227 Dopheide, M.M., Morgan, R.E., Rodvelt, K.R., Schachtman, T.R., Miller, D.K., 2007. stimulants by rats: a comparison of the effects of d-amfetamine, methylphe- 1163 Modafinil evokes striatal [3H]dopamine release and alters the subjective nidate and McNeil 4612. Pharmacol. Biochem. Behav. 20, 227e232. 1228 1164 properties of stimulants. Eur. J. Pharmacol. 568, 112e123. Paterson, N.E., Fedolak, A., Olivier, B., Hanania, T., Ghavami, A., Caldarone, B., 2010. 1229 fi 1165 Elia, J., Ambrosini, P.J., Rapoport, J.L., 1999. Treatment of attention-de cit hyperac- Psychostimulant-like discriminative stimulus and locomotor sensitization 1230 tivity disorder. N. Engl. J. Med. 11 (340), 780e788. properties of the wake-promoting agent modafinil in rodents. Pharmacol. 1166 Ermer, J.C., Dennis, K., Haffey, M.B., Doll, W.J., Sandefer, E.P., Buckwalter, M., Biochem. Behav. 95, 449e456. 1231 1167 Page, R.C., Diehl, B., Martin, P.T., 2011. Intranasal versus oral administration of Pennick, M., 2010. Absorption of lisdexamfetamine dimesylate and its enzymatic 1232 e 1168 lisdexamfetamine dimesylate: a randomized, open-label, two-period, crossover, conversion to d-amfetamine. Neuropsychiatr. Dis. Treat. 24, 317 327. 1233 single-dose, single-centre pharmacokinetic study in healthy adult men. Clin. Rowley, H.L., Kulkarni, R., Gosden, J., Brammer, R., Hackett, D., Heal, D.J., 2012. Lis- 1169 Drug Investig. 3, 357e370. dexamfetamine and immediate release d-amfetamine e differences in phar- 1234 1170 Extance, K., Goudie, A.J., 1981. Inter-animal olfactory cues in operant drug- macokinetic/pharmacodynamic relationships revealed by striatal microdialysis 1235 1171 discrimination procedures in rats. Psychopharmacology 73, 363e371. in freely-moving rats with simultaneous determination of plasma drug con- 1236 Gold, L.H., Balster, R.L., 1996. Evaluation of the cocaine-like discriminative stim- centrations and locomotor activity. Neuropharmacology 63, 1064e1074. 1172 ulus effects and reinforcing effects of modafinil. Psychopharmacology 126, Rush, C.R., Baker, R.W., 2001. Behavioural pharmacological similarities between 1237 1173 286e292. methylphenidate and cocaine in cocaine abusers. Exp. Clin. Psychopharmacol. 9, 1238 1174 Gorelick, D.A., 1998. The rate hypothesis and agonist substitution approaches to 59e73. 1239 cocaine abuse treatment. Adv. Pharmacol. 42, 995e997. Rush, C.R., Kelly, T.H., Hays, L.R., Wooten, A.F., 2002a. Discriminative-stimulus effects 1175 Gosden, J., Buckett, W.R., Heal, D.J., 1996. d-Amfetamine-cued drug-discrimination of modafinil in cocaine-trained humans. Drug Alcohol. Depend. 67, 311e322. 1240 1176 in rats: predictive value for detecting stimulant drugs of abuse. Br. J. Pharma- Rush, C.R., Kelly, T.H., Hays, L.R., Baker, R.W., Wooten, A.F., 2002b. Acute behavioural 1241 1177 col. 117, 169P. and physiological effects of modafinil in drug abusers. Behav. Pharmacol. 13, 1242 e 1178 Greenhill, L.L., Biederman, J., Boellner, S.W., Rugino, T.A., Sangal, R.B., Earl, C.Q., 105 115. 1243 Jiang, J.G., Swanson, J.M., 2006. A randomized, double-blind, placebo-controlled Rush, C.R., Kollins, S.H., Pazzaglia, P.J., 1998. Discriminative-stimulus and 1179 study of modafinil film-coated tablets in children and adolescents with participant-rated effects of methylphenidate, bupropion, and triazolam in d- 1244 1180 attention-deficit/hyperactivity disorder. J. Am. Acad. Child. Adolesc. Psychiatry amfetamine-trained humans. Exp. Clin. Psychopharmacol. 6, 32e44. 1245 e 1181 45, 503 511. Schindler, C.W., Gilman, J.P., Panlilio, L.V., McCann, D.J., Goldberg, S.R., 2011. Com- 1246 Heal, D.J., Cheetham, S.C., Smith, S.L., 2009. The neuropharmacology of ADHD drugs parison of the effects of methamfetamine, bupropion, and methylphenidate on 1182 in vivo: insights on efficacy and safety. Neuropharmacology 57, 608e618. the self-administration of methamfetamine by rhesus monkeys. Exp. Clin. 1247 1183 Heal, D.J., Frankland, A.T.J., Gosden, J., Hutchins, L.J., Prow, M.R., Luscombe, G.P., Psychopharmacol. 19, 1e10. 1248 1184 Buckett, W.R., 1992. A comparison of the effects of sibutramine hydrochloride, Smith, R.C., Davis, J.M., 1977. Comparative effects of d-amfetamine, l-amfetamine, 1249 bupropion and methamfetamine on dopaminergic function: evidence that and methylphenidate on mood in man. J. Psychopharmacol. 53, 1e12. 1185 dopamine is not a pharmacological target for sibutramine. Psychopharmacol- Solinas, M., Panlilio, L.V., Justinova, Z., Yasar, S., Goldberg, S.R., 2006. Using drug- 1250 1186 ogy 107, 303e309. discrimination techniques to study the abuse-related effects of psychoactive 1251 1187 Heal, D.J., Gosden, J., Smith, S.L., 2012a. ADHD: current and future therapeutics. Curr. drugs in rats. Nat. Protoc. 1, 1194e1206. 1252 Top. Behav. Neurosci. 9, 361e390. Spencer, T.J., Biederman, J., Ciccone, P.E., Madras, B.K., Dougherty, D.D., Bonab, A.A., 1188 Heal, D.J., Pierce, D.M., 2006. Methylphenidate and its isomers: their role in the Livni, E., Parasrampuria, D.A., Fischman, A.J., 2006. PET study examining phar- 1253 1189 treatment of attention-deficit hyperactivity disorder using a transdermal de- macokinetics, detection and likeability, and dopamine transporter receptor 1254 1190 livery system. CNS Drugs 20, 713e738. occupancy of short- and long-acting oral methylphenidate. Am. J. Psychiatry 1255 Heal, D., Rowley, H., Kulkarni, R., Hackett, D.A., 2012b. Microdialysis and behavioural 163, 387e395. 1191 comparison of lisdexamfetamine methylphenidate and modafinil in freely- Spencer, T., Biederman, J., Wilens, T., Harding, M., O’Donnell, D., Griffin, S., 1996. 1256 1192 moving rats. J. Psychopharmacol. 26, A73. Pharmacotherapy of attention-deficit hyperactivity disorder across the life cy- 1257 1193 Heal, D.J., Smith, S.L., Gosden, J., Nutt, D.J., 2013. Amfetamine, past and present e a cle. J. Am. Acad. Child Adolesc. Psychiatr. 35, 409e432. 1258 1194 Q3 pharmacological and clinical perspective. J. Psychopharmacol. (in press). Stadler, J.R., Caul, W.F., Barrett, R.J., 2001. Effects of training dose on amfetamine 1259 Heishman, S.J., Henningfield, J.E., 1991. Discriminative stimulus effects of d-amfet- drug discrimination: dose-response functions and generalization to cocaine. 1195 amine, methylphenidate, and diazepam in humans. Psychopharmacology 103, Pharmacol. Biochem. Behav. 70, 381e386. 1260 1196 436e442. Stoops, W.W., Glaser, P.E.A., Rush, C.R., 2003. Reinforcing, subject-rated, and phys- 1261 fi 1197 Jasinski, D.R., 2000. An evaluation of the abuse potential of moda nil using iological effects of intranasal methylphenidate in humans: a dose-response 1262 methylphenidate as a reference. J. Psychopharmacol. 14, 53e60. analysis. Drug Alcohol Depend. 71, 179e186. 1198 Jasinski, D.R., Krishnan, S., 2009a. Abuse liability and safety of oral lisdexamfet- Stoops, W.W., Glaser, P.E.A., Fillmore, M.T., Rush, C.R., 2004. Reinforcing, subject- 1263 1199 amine dimesylate in individuals with a history of stimulant abuse. rated, performance and physiological effects of methylphenidate and d- 1264 e e 1200 J. Psychopharmacol. 23, 419 427. amfetamine in stimulant abusing humans. J. Psychopharmacol. 18, 534 543. 1265 Jasinski, D.R., Krishnan, S., 2009b. Human pharmacology of intravenous lisdex- Stoops, W.W., Lile, J.A., Fillmore, M.T., Glaser, P.E.A., Rush, C.R., 2005. Reinforcing 1201 amfetamine dimesylate: abuse liability in adult stimulant abusers. effects of modafinil: influence of dose and behavioural demands following drug 1266 1202 J. Psychopharmacol. 23, 410e418. administration. Psychopharmacology 182, 186e193. 1267 1203 Johanson, C.E., 1990. The evaluation of the abuse liability of drugs. Drug Saf. 5, 46e57. Suto, N., Austin, J.D., Tanabe, L.M., Kramer, M.K., Wright, D.A., Vezina, P., 2002. 1268 Johanson, C.E., Balster, R.L., Bonese, K., 1976. Self-administration of psychomotor Previous exposure to VTA amfetamine enhances cocaine self-administration 1204 stimulant drugs: the effects of unlimited access. Pharmacol. Biochem. Behav. 4, under a progressive ratio schedule in a D1 dopamine receptor dependent 1269 1205 45e51. manner. Neuropsychopharmacology 27, 970e979. 1270 1206 Johanson, C.E., Schuster, C.R., 1975. A choice procedure fro drug reinforcers: Swanson, J.M., Greenhill, L.L., Lopez, F.A., Sedillo, A., Earl, C.Q., Jiang, J.G., 1271 cocaine and methylphenidate in the rhesus monkey. J. Pharmacol. Exp. Ther. Biederman, J., 2006. Modafinil film-coated tablets in children and adolescents 1207 193, 676e688. with attention-deficit/hyperactivity disorder: results of a randomized, double- 1272 1208 Kollins, S.H., Rush, C.R., Pazzaglia, P.J., Ali, J.A., 1998. Comparison of acute behav- blind, placebo-controlled, fixed-dose study followed by abrupt discontinua- 1273 1209 ioural effects of sustained-release and immediate-release methylphenidate. tion. J. Clin. Psychiatry 67, 137e147. 1274 e 1210 Exp. Clin. Psychopharmacol. 6, 367 374. Volkow, N.D., Wang, G.-J., Gatley, S.J., Fowler, J.S., Ding, D.-S., Logan, J., 1275 Kollins, S.H., MacDonald, E.K., Rush, C.R., 2001. Assessing the abuse potential of Hitzemann, R., Angrist, B., Lieberman, J., 1996. Temporal relationships between 1211 methylphenidate in nonhuman and human subjects: a review. Pharmacol. the pharmacokinetics of methylphenidate in the human brain and its behav- 1276 1212 Biochem. Behav. 68, 611e627. ioural and cardiovascular effects. Psychopharmacology 123, 26e33. 1277 1213 Liu, Y., Morgan, D., Roberts, D.C., 2007. Cross-sensitization of the reinforcing effects Volkow, N.D., Wang, G.-J., Fowler, J.S., Logan, J., Gatley, S.J., Wong, C., Hitzemann, R., 1278 of cocaine and amfetamine in rats. Psychopharmacology 195, 369e375. Pappas, M.R., 1999. Reinforcing effects of psychostimulants in humans are 1214 Madras, B.K., Xie, Z., Lin, Z., Jassen, A., Panas, H., Lynch, L., Johnson, R., Livni, E., associated with increases in brain dopamine and occupancy of D2 receptors. 1279 1215 Spencer, T.J., Bonab, A.A., Miller, G.M., Fischman, A.J., 2006. Modafinil J. Pharmacol. Exp. Ther. 291, 409e415. 1280

D.J. Heal et al. / Neuropharmacology xxx (2013) 1e11 11

1281 Volkow, N.D., Swanson, J.M., 2003. Variables that affect the clinical use and abuse of Witkin, J.M., Nichols, D.E., Terry, P., Katz, J.L., 1991. Behavioral effects of selective 1288 e 1282 methylphenidate in the treatment of ADHD. Am. J. Psychiatry 160, 1909 1918. dopaminergic compounds in rats discriminating cocaine injections. 1289 Volkow, N.D., Fowler, J.S., Logan, J., Alexoff, D., Zhu, W., Telang, F., Wang, G.-J., J. Pharmacol. Exp. Ther. 257, 706e713. 1283 Jayne, M., Hooker, J.M., Hubbard, B., Carter, P., Warner, D., King, P., Shea, C., Yokel, R.A., Wise, R.A., 1978. Amfetamine-type reinforcement by dopaminergic 1290 1284 Xu, Y., Nuench, L., Apelskog-Torres, K., 2009. Effects of modafinil on dopamine agonists in the rat. Psychopharmacology 58, 289e296. 1291 1285 and dopamine transporters in the male human brain. Clinical implications. Zolkowska, D., Jain, R., Rothman, R.B., Partilla, J.S., Roth, B.L., Setola, V., 1292 J. Am. Med. Assoc. 301, 1148e1154. Prisinzano, T.E., Baumann, M.H., 2009. Evidence for the involvement of dopa- 1286 Vosburg, S.K., Hart, C.L., Haney, M., Rubin, E., Foltin, R.W., 2010. Modafinil does not mine transporters in behavioural stimulant effects of modafinil. J. Pharmacol. 1293 1287 serve as a reinforcer in cocaine abusers. Drug Alcohol Depend. 106, 233e236. Exp. Ther. 329, 738e746. 1294