Editors: Spanagel, Rainer; Mann, Karl F.
Title: Drugs for Relapse Prevention of Alcoholism, 1st Edition
Copyright 2005 Springer
> Table of Contents > Second generation opioidergic compounds: clinical data
Second generation opioidergic compounds: clinical data
John David Sinclair
POB 33, Department of Mental Health and Alcohol Research, National Public Health Institute (KTL), Helsinki 00251, Finland
Introduction
The only second generation opioidergic compound for which there is substantial clinical information is nalmefene. Nalmefene, C21H25NO3, with a molecular weight of 339.43, is a pure opioid antagonist with no abuse potential [1].
Like naltrexone and naloxone, nalmefene was originated by Jack Fishman and co-workers. The three antagonists have rather similar structures and properties. The molecular structure of nalmefene is shown in the Appendix for Chemical Structures. Nalmefene is an analog of naltrexone, while naltrexone is a congener of naloxone.
Nalmefene, like naltrexone and naloxone, is a relatively non-specific opioid antagonist, i.e., binding to mu, delta, and kappa opioid receptors. The binding abilities of the three antagonists [2] are shown in Figure 1. Nalmefene is similar to naltrexone in its binding of mu receptors, but somewhat more potent than naltrexone for kappa and delta binding. Both are more potent than naloxone in binding all three opioid receptor types. Thus, naltrexone has been characterized as being relatively selective for mu receptors, whereas nalmefene can be seen as a more universal opioid blocker. There is, however, currently no evidence that these three opioid antagonists differ in their basic actions once they reach the central nervous system.
Pharmacokinetics
Oral nalmefene is rapidly absorbed [3]. It has a mean elimination half-life of 10.7 h (range 7-15 h), which is substantially longer than for naloxone (1.1 h) and naltrexone (4 h) [4]. About 4% of nalmefene is excreted in the urine as unchanged nalmefene, and up to 60% is excreted as inactive glucuronide conjugates. Naltrexone, in contrast, is metabolized to an active compound, 6-beta-naltrexol (half-life: 12.9 h) which may be important for the efficacy of naltrexone in treating alcoholism [5].
P.126
Figure 1. Binding of nalmefene, naloxone, and naltrexone to the three types of opioid receptors. Derived from Michel et al. [2]. Shown are the reciprocals of the IC50S (nM), in order to represent stronger bindings as larger values. The specific binding compounds displaced were [3H]-dihydromorphine for mu receptors, [3H]-ethylketocyclazocine for kappa, and [3H]-D-ala-D-leu enkephalin for delta receptors. |
Oral dosing with nalmefene is subject to less variability than with naltrexone. Oral nalmefene has a bioavailability of 40-50% with relatively small differences between individuals [6]. In contrast, oral naltrexone is subject to high but variable amounts being removed by first-pass metabolism resulting in a bioavailability ranging from only 5% up to 40% [4, 7]. The 3- to 4-fold variation
P.127
Naloxone is metabolized so rapidly in the liver that the first-pass metabolism prevents it from being used orally; i.e., if naloxone is ingested, essentially all of it is metabolized on the first pass through the liver and none reaches the brain.
Nalmefene continues to block opiate effects longer than naltrexone and naloxone [10]. Oral doses of both 50 mg and 100 mg nalmefene blocked the effects of morphine for up to 48 h [11], whereas only a 100 mg naltrexone dose is still effective at 48 h [4]. Nalmefene also maintains occupancy of opioid receptors longer than naloxone does [12]. Renal insufficiency prolonged the half-life, but hemodialysis did not change the basic pharmacokinetics of nalmefene [13]. The kinetics of nalmefene are similar in children (ages 2 to 12 years) to those reported for adults [14]. The terminal half-life in children was 8.7 2.3 h.
Hepatic toxicity
Theoretically, effective usage of an opioid antagonist should not depend upon the ability to deliver a specific dose to the brain. The goal is to block all the opioid receptors (or perhaps all the mu opioid receptors); it should not matter if too much antagonist is used and some is left over. In practice, however, the naltrexone dose cannot be raised too high because of hepatic toxicity reported with 300 mg daily [15, 16]. Naltrexone is marketed with a black box warning of liver toxicity. Naltrexone is contraindicated for cirrhotic patients and should not be started in alcohol-dependent patients until after a blood test for liver cirrhosis.
In contrast, there is no evidence of nalmefene being detrimental to the liver [17]. No dose-dependent liver toxicity has been reported in over 1300 patients, including patients already having liver disorders [18]. Oral nalmefene was well tolerated in doses up to 300 mg [3]. Chronic administration is also well tolerated with no clinically significant adverse effects.
There has been discussion that the hepatic toxicity warnings for naltrexone are not warranted. Treatment of alcoholism with naltrexone usually results in reductions in the markers of liver damage, since alcohol itself is more toxic than naltrexone. In practice, only a small percentage of alcoholics are rejected initially for naltrexone administration because of cirrhosis. Nevertheless, if nalmefene turns out to be at least equivalent to naltrexone in efficacy for treating alcoholism, as seems likely, nalmefene without the toxicity would seem to be the preferred medication.
Blocking opioid effects
The Food and Drug Administration (FDA) has approved nalmefene, packaged for administration by injection, for complete or partial reversal of opioid drug
P.128
Hormonal effects
Like naltrexone and naloxone, acute nalmefene increases levels of testosterone and luteinizing hormone [19, 20]. In older impotent men it also increased follicle-stimulating hormone and cortisol levels [20]. Nalmefene caused more activation of the hypothalamic-pituitary-adrenal axis than naloxone [21].
Eating
Chronically injected nalmefene significantly suppressed caloric intake in the study on older impotent men by decreasing specifically their eating of highly palatable foods that are high in fat content [20]. The effect was replicated in a double-blind placebo-controlled clinical trial [22]. Nalmefene reduced caloric intake by 22%, without causing a significant change in hunger or satiety rating. The reduction was produced by a decrease in high fat and protein foods rated to be highly palatable. Nalmefene was given in grapefruit juice in doses of either 2.5 mg or 5.0 mg. Subjects were able to identify the higher dose from placebo but not the lower dose, whereas the effect on eating was seen with both doses. It might be noted that these are much lower than the oral doses found to be effective with alcoholism treatment.
Flushing and acetaldehyde
Nalmefene was found to inhibit the flushing reaction that some Orientals display after drinking alcohol [23]. The flushing reaction is caused by an accumulation of acetaldehyde resulting from a lower rate of aldehyde dehydrogenase activity. Flushers given 2 mg i.v. nalmefene plus alcohol had significantly less flushing (measured as facial skin temperature) than those given placebo plus alcohol. Nalmefene tended to decrease acetaldehyde levels found in flushers after drinking alcohol, although the effect did not reach significance. The fact, however, that nalmefene at the very least did not increase acetaldehyde levels, demonstrates that it is not acting like disulfiram (Antabuse ) and other sensitizing drugs that cause higher acetaldehyde levels and thus adverse reactions when alcohol is drunk.
P.129
Clinical trials of nalmefene in alcoholism treatment
Our pre-clinical experiments showed that when nalmefene was administered to rats before several sessions in which they had access to 10% ethanol solution as well as food and water, the alcohol intake decreased progressively with each session and remained significantly lower than baseline on subsequent sessions, when no nalmefene was administered [6, 24]. Nalmefene administered similarly but to rats without concomitant access to alcohol, however, did not reduce subsequent alcohol drinking but instead tended to increase it. These results were virtually the same as we observed in our naltrexone clinical trial: naltrexone given to patients instructed to control drinking was significantly better than placebo but naltrexone given with instructions to abstain completely, tended to be worse than placebo and was significantly worse than naltrexone with controlled drinking [25]. The findings are consistent with Wikler's hypothesis that opioid antagonists work by extinguishing opiate self-administration when the drug-taking response is made while the antagonist blocks reinforcement [26]. Nearly all of the naltrexone trials, not only with alcohol, but also with heroin and cocaine, have also supported this hypothesis (see review [24]), as have the results from the clinical trials with nalmefene.
A small double-blind placebo-controlled clinical trial of two doses of nalmefene was conducted by Barbara Mason and coworkers at the University of Miami [27, 28 and 29]. In order to examine the safety and efficacy of different doses, 21 patients were given either 10 mg (5 mg twice daily) or 40 mg (20 mg twice daily) of oral nalmefene or placebo (twice daily) for 12 weeks. The 40 mg nalmefene patients had significantly fewer relapses to heavy drinking than either of the other two groups. Relapse to heavy drinking was defined as five or more drinks per day, or 5 or more drinking days per week. They also had highly significant reductions from baseline in the mean number of abstinent days and in the mean number of drinks per drinking day. The later measure was also highly significant in the 10 mg nalmefene group but not in the control group. The 40 mg nalmefene had the greatest reduction in SGOT levels (validated marker for drinking), while the placebo group had the least reduction. Nalmefene had no significant effect on the number of days to first sampling of alcohol. There was a tendency for reduced depression, indicating nalmefene was not causing dysphoria. Nalmefene in both doses was well tolerated. The most common drug reactions were headache, insomnia, and nausea. Patients could choose psychosocial therapy as they wished. None of the 40 mg nalmefene patients were on any other therapy. One of the 10 mg nalmefene patients chose to attend group therapy meetings. Two of the placebo patients attended AA meetings and these were the only placebo patients who did not relapse to heavy drinking.
Mason's group subsequently tested 16 more alcoholics with nalmefene doses of either 20 mg (10 mg twice daily) or 80 mg (40 mg twice daily) [18] and compared the results with those previously tested with 10 mg or 40 mg [29]. The best results were obtained with the 80 mg dose. All of them remained in the
P.130
A larger double-blind placebo-controlled trial with doses of 20 mg and 80 mg was later reported [30, 31]. 105 outpatient alcoholics who had been abstinent for at least two weeks, were studied for 12 weeks, with Coping therapy (i.e., cognitive behavioral skills taught for coping with drinking situations) also being given to all subjects weekly. This is the protocol with which naltrexone has consistently been successful [24], and nalmefene with Coping therapy also was effective here. During the last 2 weeks, the 80 mg nalmefene group showed significantly fewer drinks per drinking day [30]. Otherwise, the results with the two nalmefene doses here were similar and thus were combined. The double-blind procedure was demonstrated to be effective by the inability of subjects to guess their group significantly better than chance. Nalmefene significantly reduced the rate of relapsing to heavy drinking (here, >4 drinks/day for men, >3 drinks/day for women). Although this difference between nalmefene and placebo was already significant in the first week, the difference between the groups expanded progressively during the first half of the study and thereafter remained essentially constant. Nalmefene was significantly better than placebo, not only in the percentage of patients relapsing but also in the number of relapses. Nalmefene tended (p = 0.06) to be better than placebo for the duration of relapses and for the number of drinks per drinking day. The latter measure decreased significantly from baseline in both groups. These results in turn were significantly correlated with final GGT levels, which decreased significantly over the 12 weeks. The group differences were still clearer in the subgroups of patients who had at least one drink during the entire study: 80% of those on placebo relapsed to heavy drinking while only 56% of those on nalmefene did (p < 0.03). The percentage of abstinent days did not differ significantly between nalmefene and placebo groups, with both groups increasing significantly from baseline. Both groups had significant reductions in craving but there was no significant difference between groups. Patients on nalmefene showed no medically serious adverse drug experiences and had high rates of medication compliance and treatment completion.
Newer clinical results
The 1999 study by Mason et al. [31] is the latest nalmefene trial to be published in a full-length paper. Several additional nalmefene trials have been conducted and reported at scientific conferences since then, but the published references for them are at best only abstracts. Consequently, they will be covered only briefly here, with the reader being cautioned that the results have not been subjected to peer review.
The clinical results with naltrexone have been shown to be highly dependent upon the protocol with which the medicine is given [24, 32]. This is best
P.131
The efficacy of nalmefene also appears to be dependent upon protocol. Mason and coworkers [31] had shown that nalmefene, like naltrexone, was effective when combined with Coping therapy. In 2002, Anton reported negative results from a multisite trial of nalmefene with Motivational Enhancement Therapy (MET) [36], a form of psychosocial therapy that had not previously been used in tests with opioid antagonists. The next year Anton reported results from a double trial comparing naltrexone with Coping to naltrexone with MET [37]. Naltrexone with Coping once again provided significant benefits, but naltrexone with MET, like nalmefene with MET, produced negative results. It seems likely that MET is similar to Supportive therapy and not suited for use with opioid antagonists.
Meanwhile, M kel reported results from a randomized double-blind placebo-controlled nalmefene study of 150 outpatients with impaired control over their drinking at 6 Finnish sites [38, 39]. The study used no structured psychosocial therapy. Nalmefene (10 mg or 40 mg) or placebo was taken once daily for 16 weeks. The smaller nalmefene dose produced only transient benefits, but the larger nalmefene dose was significantly better than placebo in the reduction of heavy drinking days. The increase in abstinence days and the decrease in mean weekly consumption were best in the 40 mg nalmefene group. Nalmefene was well tolerated with only minor adverse events, primarily at the beginning of treatment. It was concluded that nalmefene 40 mg once daily is safe and appears effective in the reduction of heavy alcohol consumption without structured psychosocial treatment.
Subsequently, a Phase III double-blind placebo-controlled nalmefene trial in Finland and the UK was reported [40]. Approximately 570 patients took nalmefene or placebo for 28 weeks only before alcohol drinking. Patients on nalmefene had a significantly greater reduction (nearly 50%) in heavy drinking days than did placebo patients. This was significant also in the Finnish patients separately but not in the British subgroup alone due to higher dropout rates. Significantly more nalmefene patients than controls were rated as much improved or very much improved . This was significant in both Finnish and British patients separately, too. No serious adverse effects were observed.
P.132
Finally, a double-blind placebo-controlled Phase II clinical trial with 200 subjects found that nalmefene was significantly better than placebo in the treatment of compulsive gambling [41]. Nalmefene reduces craving and thoughts about gambling to a level about half that in the placebo group.
Relapse: definition of terms
The results with nalmefene, and with naltrexone, emphasize the importance of precisely defining the word relapse . With traditional treatments, breaking abstinence by sampling alcohol often leads to a resumption of heavy drinking, so relapse could be used indiscriminately to mean both the first sip and also the return to heavy drinking. When opioid antagonists are used, however, there is a great difference between the two behaviors. The resumption of sampling has almost never been found to be affected by treatment with naltrexone or nalmefene [24], but resumption of heavy drinking has been the measure showing the greatest benefits. Probably for the same reason, naltrexone has been found to be helpful for alcoholics who were actively drinking at the onset of treatment but not for ones who had already been abstinent for weeks [42].
It would be advantageous if the word relapse were used exclusively for the resumption of heavy drinking, and if agreement could be reached as to what precisely constitutes heavy drinking. Unfortunately, we then need some other word for resumption of sampling and moderate consumption.
Conclusions
The clinical results show that nalmefene is safe and well tolerated by heavy drinkers and alcohol-dependent patients, regardless of whether they have been detoxified or are still actively drinking.
Nalmefene has inherent advantages over naltrexone in its dosing, duration of action, lack of active metabolite, and lack of liver toxicity.
Nalmefene can be effective, especially in reducing relapses to heavy drinking, in the treatment of alcoholism.
The efficacy of nalmefene in alcoholism treatment, like that of naltrexone, is dependent upon the protocol with which it is used:
Nalmefene is effective with Coping therapy.
Nalmefene is effective without any structured psychosocial therapy.
Nalmefene is not more effective than placebo with Motivational Enhancement Therapy and probably, like naltrexone, not with programs strongly supporting complete abstinence.
Nalmefene is effective also in treating compulsive gambling.
P.133
References
1 Fudala PJ, Heishman SJ, Henningfield JE, Johnson RE (1991) Human pharmacology and abuse potential of nalmefene. Clin Pharmacol Ther 49: 300-306
2 Michel ME, Bolgan G, Weisman B-A (1985) Binding of a new opiate antagonist, nalmefene, to rat brain membranes. Meth Find Exp Clin Pharmacol 7: 175-177
3 Dixon R, Gentile J, Hsu HB, Hsiao J, Howes J, Garg D, Weidler D (1987) Nalmefene: safety and kinetics after single and multiple oral doses of a new opioid antagonist. J Clin Pharmacol 27: 233-239
4 Dupont Pharma (1995) package insert for ReVia available at http://www.fda.gov/cder/ogd/rld/18932s14.pdf
5 McCaul ME, Wand GS, Sullivan J, Mummford G, Quigley J (1997) Beta-naltrexol level predicts alcohol relapse. Alcohol Clin Exp Res 21: 32A
6 Sinclair JD, Scheinin H, Lammintausta R (1992) Method for treating alcoholism with nalmefene. USA Patent 5,086,058 Feb. 4
7 Gonzalez JP, Brogden RN (1988) Naltrexone: A review of its pharmacodynamic and pharmacokinetic properties and therapeutic efficacy in the management of opioid dependence. Drugs 35: 192-213
8 Verebey K, De Pace A, Jukofsky D, Volabka JV, Mule SJ (1980) Quantitative determination of 2-hydroxy-3-methoxy-6 beta-naltrexol (HMN), naltrexone, and 6 beta-naltrexol in human plasma, red blood cells, saliva, and urine by liquid chromatography. J Anal Toxicol 4: 33-37
9 Bertolotti M, Ferrari A, Vitale G, Stefani M, Trenti T, Loria P, Carubbi F, Carulli N, Sternieri E (1997) Effect of liver cirrhosis on the systemic availability of naltrexone in humans. J Hepatol 27: 505-511
10 Gal TJ, DiFazio CA, Dixone R (1986) Prolonged blockage of opioid effects with oral nalmefene. Clin Pharmacol Ther 40: 537-542
11 Jones HE, Johnson RE, Fudala PJ, Henningfield JE, Heishman SJ (2000) Nalmefene: blockade of intravenous morphine challenge effects in opioid abusing humans. Drug Alcohol Depend 60: 29-37
12 Kim S, Wagner KN, Villemagne VL, Kao PF, Dannals RF, Ravert HT, Job T, Dixon RB, Civelek AC (1997) Longer occupancy of opioid receptors by nalmefene compared to naloxone as measured in vivo by a dual-detector system. J Nucl Med 38: 1726-1731
13 Matzke GR, Frye RF, Alexander AC, Reynolds R, Dixon R, Johnston J, Rault RM (1996) The effect of renal insufficiency and hemodialysis on the pharmacokinetics of nalmefene. J Clin Pharmacol 36: 144-151
14 Chumpa A, Kaplan RL, Burns MM, Shannon MW (2001) Nalmefene for elective reversal of procedural sedation in children. Am J Emerg Med 19: 545-548
15 Mitchell JE, Morley JE, Levine AS, Hatsukami D, Gannon M, Pfohl D (1987) High-dose naltrexone therapy and dietary counseling for obesity. Biol Psychiatry 22: 35-42
16 Morgan CJ, Kosten TR (1990) Potential toxicities of high doses of naltrexone in patients with appetitive disorders. In: LD Reid (ed.): Opioids, bulimia, and alcohol abuse and alcoholism. Springer-Verlag, New York, 261-273
17 Ohmeda (2000) Revex package insert, available at http://www.fda.gov/cder/foi/label/2000/20459S2lbl.pdf
18 Mason B (1996) Dosing issues in the pharmacotherapy of alcoholism. Alcohol Clin Exp Res 20 (Suppl): 10A-16A
19 Graves GR, Kennedy TG, Weick RF, Casper RF (1993) The effect of nalmefene on pulsatile secretion of luteinizing hormone and prolactin in men. Hum Reprod 8: 1598-1603
20 Billington CJ, Shafer RB, Morley JE (1990) Effects of opioid blockade with nalmefene in older impotent men. Life Sci 47: 799-805
21 Schluger JH, Ho A, Borg L, Porter M, Maniar S, Gunduz M, Perret G, King A, Kreek MJ (1998) Nalmefene causes greater hypothalamic-pituitary-adrenal axis activation than naloxone in normal volunteers: Implications for the treatment of alcoholism. Alcohol Clin Exp Res 22: 1430-1436
22 Yeomans MR, Wright P, Macleod HA, Critchley JAJH (1988) Effects of nalmefene on feeding in humans. Psychopharmacology 100: 426-432
23 Ho SB, DeMaster EG, Shafer RB, Levine AS, Morley JE, Go VLW, Allen JI (1988) Opiate antagonist nalmefene inhibits ethanol-induced flushing in Asians: A preliminary study. Alcohol Clin Exp Res 12: 705-712
P.134
24 Sinclair JD (2001) Evidence about the use of naltrexone and for different ways of using it in the treatment of alcoholism. Alcohol Alcoholism 36: 2-10
25 Hein l P, Alho H, Kiianmaa K, L nnqvist J, Kuoppasalmi K, Sinclair JD (2001) Targeted use of naltrexone without prior detoxification in the treatment of alcohol dependence: A factorial double-blind placebo-controlled trial. J Clin Psychopharmacol 21: 287-292
26 Wikler A (1976) The theoretical basis of narcotic addiction treatment with narcotic antagonists. In: D Julius, P Renault (eds): Narcotic Antagonists: Naltrexone. National Institute of Drug Abuse, Rockville, MD, 119-122
27 Mason BJ, Ritvo EC, Salvato F, Zimmer E, Goldberg G, Welch B (1993) Nalmefene modification of alcohol dependence: A pilot study. Proceedings of American Psychiatric Association Annual Meeting, San Francisco, CA, May 1993, 170 (abstract NR442)
28 Mason BJ, Ritvo EC, Salvato FR, Goldberg G (1994) Preliminary dose finding for nalmefene treatment of alcoholism. Alcohol Clin Exp Res 18: 464 (abstract 270)
29 Mason BJ, Ritvo EC, Morgan RO, Salvato FR, Goldberg G, Welch B, Mantero-Atienza E (1994) A double-blind, placebo-controlled pilot study to evaluate the efficacy and safety of oral Nalmefene HCL for alcohol dependence. Alcohol Clin Exp Res 18: 1162-1167
30 Litten RZ, Croop RS, Chick J, McCaul ME, Mason B, Sass H (1996) International update: New findings on promising medications. Alcohol Clin Exp Res 20: 216A-218A
31 Mason BJ, Salvato FR, Williams LD, Ritvo EC, Cutler RB (1999) A double-blind, placebo-controlled study of oral Nalmefene for alcohol dependence. Arch Gen Psychiatry 56: 719-725
32 Sinclair JD, Alho H, Shinderman M (2002) Treatment of alcoholism: Efficacious use of naltrexone. N Engl J Med 17: 1329-1331
33 O'Malley SS, Jaffe AJ, Chang G, Schottenfeld RS, Meyer RE, Rounsaville B (1992) Naltrexone and coping skills therapy for alcohol dependence. Arch Gen Psychiatry 49: 881-887
34 Balldin J, Berglund M, Borg S, M nsson M, Berndtsen P, Franck J, Gustafsson L, Halldin J, Hollstedt C, Nilsson L-H et al. (1997) A randomized 6 month double-blind placebo-controlled study of Naltrexone and coping skills education programme. Alcohol Alcoholism 32: 325
35 Schmitz JM, Stotts AL, Rhoades HM, Grabowski J (2001) Naltrexone and relapse prevention treatment for cocaine-dependent patients. Addict Behav 26: 167-180
36 Anton R (2002) Multisite study of Nalmefene combined with modified motivational enhancement therapy in the treatment of outpatient alcoholics Presented at the 25th Annual Scientific Meeting of the Research Society on Alcoholism, June 28-July 3, 2002, San Francisco, CA, USA
37 Anton RF, Moak DM, Latham PK, Myrick DL, Waid LR (2003) A double blind comparison of naltrexone combined with CBT or MET in the treatment of alcohol dependence. Presented at the 26th Annual Scientific Meeting of the Research Society on Alcoholism, June 21-25, 2003, Fort Lauderdale, Florida. Alcohol Clin Exp Res 27 (suppl): 191A (abstract S170)
38 M kel R, Kallio A, Karhuvaara S (2001) Nalmefene in the treatment of heavy drinking. Programme and Abstracts of the 2001 ISAM Meeting, Trieste, Italy, September 12-14 (no page numbers)
39 M kel R (2002) Multisite study of Nalmefene for the treatment of heavy alcohol drinkers with impaired control. Presented at the 25th Annual Scientific Meeting of the Research Society on Alcoholism, June 28-July 3, 2002, San Francisco, CA, USA
40 BioTie Therapies Corp., press release, April 24 (2003) Phase III clinical studies in alcoholism and alcohol abuse. Available at http://www.biotie.com/
41 BioTie Therapies Corp, press release, May 30 (2003) Results from a Phase II clinical study suggest Nalmefene effective in the treatment of pathological gambling. Available at http://www.biotie.com/
42 Killeen T, Brady K, Faldowski R, Gold P, Simpson K, Anton, R (2003) The efficacy of naltrexone in a community treatment program. Alcohol Clin Exp Res 27 (suppl): 146a (abstract 846)