NEUROPSYCHOPHARMACOLOGY 1993-VOL. 9, NO. 1 77 Gamma-Hydroxybutyric Acid for Treatment of Opiate Withdrawal Syndrome Luigi Gallimberti, M.D., Mauro Cibin, M.D., Piero Pagnin, M.D., Roberta Sabbion, M.D., PierPaolo Pani, M.D., Roberto Pirastu, M.D., Santo Davide Ferrara, M.D., IIIdGian Luigi Gessa, M. D. IIa double-blind placebo-controlled trial, gamma­ suppressed and patients reported felling well. Urine tydroxybutyricacid (GHB) (25 mglkg orally) suppressed analysis failed to detect any presence of opiate aJ5I of the withdrawal symptomatology in 14 heroin metabolites. No withdrawal symptomatology recurred lidictsand 13 methadone-maintained subjects. The GHB after 8 days of treatment when GHB was suspended, and tftrt was prompt (within 15 minutes) and persisted for patients were challenged with an intravenous injection of IItuwt 2and 3 hours. Subsequently, the same patients 0.4 mg naloxone. The results indicate that GHB may be rraiwd GHB in an open study every 2 to 4 hours for useful in the management of opiate withdrawal. JIltfirst 2 days and 4 to 6 hours for the following 6 [Neuropsychopharmacology 9:77-81, 1993] �: most abstinence signs and symptoms remained lEY WORDS: Gamma-hydroxybutyric acid; Heroin; sidered to function either as a neurotransmitter or IdttluJdone; Opiates; Withdrawal syndrome; Opiate neuromodulator rather than as an incidental metabo­ Itptndence lite of gamma-aminobutyric acid (GABA) (Mandel et Gamma-hydroxybutyric acid (GHB), an endogenous al. 1987). mnstituent of the mammalian brain, is found in the Gamma-hydroxybutyric acid has been used as an llighest concentrations in the hypothalamus and basal intravenous hypnotic and anesthetic agent (Mamelak pn�ia (Snead and Morely 1981). Because there are et al. 1977) and in the treatment of narcolepsy. In the tpecihchigh-affinity binding sites for GHB in the cen­ latter condition, GHB is given orally at bedtime to im­ tralnervous system (CNS), and the compound is 10- prove nocturnal sleep quality, thus reducing cataplexy '*<I principally in synaptosomes-from which it is episodes during the day (Mamelak et al. 1986). Previ­ "ased in a Ca2+-dependent process-GHB is con- ous work in animals has shown that GHB both inhibits voluntary ethanol consumption in rats having a strong preference for ethanol and suppresses the ethanol withdrawal syndrome in rats physically dependent on ethanol (Fadda et al. 1989a, 1989b). From the Addiction Treatment Service (LG, RS) (SERT, USL 21), More recently, in a double-blind study we found Pidova; Neurotoxicology Unit, General Hospital of Dolo (MC), \'mice; MedicineThird Division, General Hospital of Padova (PP), that GHB, given orally in nonhypnotic doses, is highly Pidova; Centre of Behavioural and Forensic Toxicology, IML, effectivein suppressing the withdrawal symptomatol­ "'-'mitv of Padova (SDF), Padova; and Bernard B. Brodie ogy in alcoholics; the GHB effect being rapid and de­ D!putment of Neuroscience, University of Cagliari and Addiction lIIIIment Service (pPP, RP, GLG) (SERT, USL 21), Cagliari, Italy. void of negative side effects (Gallimberti et al. 1989). Address correspondence to Prof. Gian Luigi Gessa, Bernard B. By using GHB in the management of the withdrawal IIadie Department of Neuroscience, University of Cagliari, Via syndrome in a number of alcoholics who concomi­ f1m114, 1-09124Cagliari, Italy. llreived October 22, 1992; revised February 17, 1993; accepted tantly abused heroin, we observed that GHB not only IIIth 11, 1993. suppressed the alcohol withdrawal symptomatology 11993Americ an College of Neuropsychopharmacology Mtished by Elsevier Science Publishing Co., Inc. • Avenue of the Americas, New York, NY 10010 0893-133X/93/$6.00 78 L. Gallimberti et al. NEUROPSYCHOPHARMACOLOGY 1993-VOL. 9, NO.1 but, even more effectively, that of heroin. The present treatment conditions. Twenty-one items associated study was undertaken to clarify, in a double-blind con­ with the withdrawal syndrome were rated as present dition, whether GHB was effectivein suppressing the (1) or absent (0), according to Gold et al. (1978); the withdrawal syndrome in heroin- and methadone-de­ maximum score attainable was 21. Symptoms consid· pendent subjects. ered included: craving, nausea, anorexia, anxiety and restlessness, aching bones and muscles, insomnia or yen sleep, and hot and cold flashes. Signs considered SUBJECTS AND METHODS comprised: tremors, yawning, vomiting, diarrhea, perspiration, lacrimination, rhinorrhea, increased res· The subjects participating the study were 22 male piration rate and depth, goose flesh, mydriasis, spon· heroin users, with a mean age of 27.3 years (range 22 taneous orgasm, increased temperature, tachycardia, to 33 years), with a clear history of daily use of heroin and increased blood pressure. for over 3 years (3 to 6 years) claiming a high degree Withdrawal scoring was made every 30 minutes of opiate dependence; and 19 male subjects, with an for 3 hours prior to treatment (baseline period) and at age ranging from 24 to 31 years (mean 28 years) under­ various times afterwards, as indicated in the Results going a methadone maintenance treatment program Section. The experiment was randomized and double at the Addiction Treatment Service (SERT, USL 21), blind; subjects, nurse, and physician observer were Cagliari. unaware of the substance administered. Active medi· Methadone-maintained subjects were receiving a cation, consisting of GHB 17% solution in a black stabilized dose of 30 to 60 mg/day of methadone for at cherry syrup, and an identical placebo were provided least 6 months prior to hospitalization. All subjects by CT (San Remo, Italy). Subjects took GHB orally at were interested in discontinuing opiate consumption the dose of 25 mg/kg or the placebo syrup. Treatments and gave informed consent to the study. The present were given by a nurse who was not otherwise in· study required hospitalization for a period of 8 days. volved in the study. The Mann-Whitney U-test was Subjects were hospitalized on the morning of the day used to test differences between placebo and GHB following their last consumption of opiates. On admis­ treatment. A modifledWilcoxon test (Pratt's test) was sion each subject underwent a medical and psychiatric applied for within-patient comparison. examination, routine laboratory tests, urine screening analysis for opiate metabolites, amphetamine, co­ caine, benzodiazepines, barbiturates, cannabinoids, RESULTS and alcohol. The fIrst test revealed that all heroin-dependent All methadone- and heroin-dependent subjects showed subjects had urinary opiate metabolites. All subjects signs and symptoms of abstinence that increased duro under methadone maintenance presented methadone ing the 3 hours preceding treatment. Table 1 shows the metabolites, three of whom revealed also other opiate reported withdrawal scores obtained during the 30 metabolites. Small concentrations of alcohol and ben­ minutes preceding treatment and at various times af· zodiazepine metabolites were found in the urine of ter treatment. The administration of GHB (25 mg/kg) both groups of patients. After admission, methadone­ reduced most of the withdrawal signs and symptoms maintained subjects were left with no opiate adminis­ both in heroin- and methadone-dependent subjects. tration for 24 hours and the rating of withdrawal symp­ The GHB effect had a rapid onset but short duration. tomatology was started at 8 AM on the second day of Thus, the global withdrawal score was signilicantlyre­ hospitalization. Fifteen of the 22 heroin-dependent duced within 15 minutes and remained maximally subjects showed overt signs of abstinence on the morn­ suppressed between 30 minutes and 2 hours after ing of the day of admission; therefore, the rating of treatment. At the third hour, withdrawal scores withdrawal symptomatology for these subjects was tended to increase once again. Table 2 shows the effect made on the morningof the same day of admission. The of GHB on specifIc items of the opiate withdrawal other seven heroin-dependent subjects, who showed scale. It appears that GHB was effectivein reducing aD objective signs of abstinence after 6 PM on the day of the signs and symptoms of opiate withdrawal, exce� admission, received 1 or 2 intramuscular doses of 15 for diarrhea and insomnia, which were resistant to the mg of morphine to ease their discomfort (no morphine GHB effect. All patients on GHB referred to relief 01 was given after midnight) and their withdrawal symp­ subjective distress. Placebo had no signiflcant effect tomatology was scored on the second day of hospital­ on any of the withdrawal items; therefore, at the end ization starting at 8 AM as for methadone-maintained of the double-blind trial, these patients were assigned subjects. to a more conventional detoxifIcation treatment with Withdrawal symptomatology was evaluated by clonidine or methadone. one of the investigators (either RP or PPP) blind to the On the other hand, patients on GHB continued NEUROPSYCHOPHARMACOLOGY 1993-VOL. 9, NO. 1 GHB in Opiate DetoxifIcation 79 this treatment receiving the drug in an open study ev­ ery 2 to 4 hours on the frrst 2 days and every 4 to 6 hours the following 6 days, as indicated by the phar­ macokinetics of GHB (Lettieri and Fung 1979, Ferrara et al. 1992). Every day, withdrawal signs and symp­ toms were recorded before each dose. The withdrawal score remained reduced throughout the trial. On the eighth day, GHB administration was suspended, sub­ jects were observed during a period of 5 to 6 hours, .t:;) U -S:) u 0\ ...... <'")1'-. and then received an intravenous injection of 0.4 mg ""';N""';""'; of naloxone. No withdrawal signs and symptoms oc­ +I +I +1 +I .-;��o: curred before and after naloxone treatment (see Table lfl 0\ <'") lfl ...... ...... I). Urine analysis performed every other day during the trial failed to detect the presence of opiate metabo­ "'" "" '"' ............co N lites. Three methadone-dependent and two heroin­ cicrl""';""'; o dependent subjects reported transient dizziness or M +I +I +1 +I .... vertigo on the second and/or third day following the <'") 0 0\ ...... ....... .0 ...... \0 N ...... fIrstmorning dose of GHB; these symptoms were usu­ ally well tolerated.