- Email: [email protected]
Serotonin uptake inhibition by the monoamine oxidase inhibitor brofaromine
BIOL PSYCHIATRY 1993;33:373-379
373
Serotonin Uptake Inhibition by the Monoamine Oxidase Inhibitor Brofaromine P. C. Waldmeier, T. Graf, M. Germer, J. J. Feldtrauer, and H. Howald
The selective, reversible monoamine oxidase (MAO) A inhibitor brofaromine inhibits serotonin (5-HT) uptake in animal models in vitro and in vivo. We investigated whether such an effect can be demonstrated at clinical doses in humans by treating three groups of six volumeers with either placebo, 15 mg phenelzine three limes a day, or 75 mg brofaromine twice a day in a 2-week experiment. As an indirect, although relevant parameter, binding of 3H-paroxetine to the 5-HT uptake sites on blood plate[ets was assessed. Moreover, wholeblood 5-HT as a measure of platelet 5-HT, and serum hamovanillic acid (HVA) to tento~ve~ estimate MAO inhibition, were determined. Brofaromiae reduced 3H-paroxetine binding to platelets compared with placebo by 20%-25% throughc~ut the treatment period, significance being reached on the last treatment day. In contrast, Fhenelzine tended to increase 3H-paroxetine binding. Both drugs increased whole-blood 5-HT to approximately 140%-150%. Brofaromine moderately and on some days significantly decreased serum HVA, whereas phenelzine only tended to do so. Our results suggest that brofaromine at the clinically used dosage of 150 rag/day does indeed inhibit 5-HT uptake, as evidenced by measurements of ~H-paroxetine binding to platelets.
Key Words: 5-HT uptake, 3H-paroxetine binding, blood platelets, MAO inhibitor, brofaromine, volunteers
Introduc¢ion Brofaromine is a potent, l c v e l S l .O. l.e. ,. . .a n uJ ~pel~llIIS "~ ,tmottt, t of monoamine oxidase A (MAO-A), which is currently in phase III clinical trials for depression and severe anxiety states. It has long been known that the compound also possesses serotonin (5-HT) Ul~',ake-inhibitiag properties (Waldmeier and Baumann 1983); however, they were initially considered to be of minor importance because the
From the Research and Development Depar~r0ent: Pharmaceuticals Division, CibaAddress reprint requests to Peter Waldmeier, Ph.D., Research Department, Pharmaceuticals Division. K-125/607, Ciba-Geigy Ltd. CH.4002 Basel, Switzerland. Received April 25. 1992: revised October 1. 1992. © 1993 Society of Biological Psychiatry.
corresponding median effective dose value was approximately 10-30 times higher than that of MAO-A inhibition (Waldmeier and Baumann 1983; Waldmeier and St6cklin 1989). In rat brain synaptosomes in vitro, brofaromine is three ar~ five times less potent than imipramine and flucx etine, respectively, in inhibiting 5-HT uptake; however, when the 5-HT uptake inhibitory effects are assessed ex vivo (i.e~, in synaptosomes prepared from pretreated rats) or in vivo by means of antagonistic effects on the binding of 3H-cyanoimipramine to the 5-HT transporter in the rat brain, it is two to three times more potent than imipramine and four to five times less potent than fluoxetine (Waldmeier and Baumann 1983; Waldmeier and St6cklin 1989). The clinical daily doses of brofaromine and imipramine are similar [100-150 mg for brofarominc (Schiwy et al 0006-3223/93/$06.00
374
BIOLPSYCHIATRY
P.C. Waldmeier et al
1993;33:373-379
1989); 75-200 mg for imipramine (Langer and Heimann 1983)], and those of fluoxetine are at least five times lower (Physicians" Desk Reference, 45th ed., 1991). 5-HT uptake into platelets of patients treated with daily doses of 50-300 mg imipramine was clearly inhibited (Tuomisto et al 1979; Arora and Meltzer 1983). This may suggest that because brofaromine is more potent than imipramine as a 5-HT uptake inhibitor in animal models, such an effect could also occur at clinical doses in patients. A similar estimation can be made from ,~comparison of clinical doses of brofaromine and fluoxetine (100-150 versus 20-40 mg daily). Moreover, a component of its antidepressant action independent of MAO-A inhibition is indicated by the observation that the maximal MAO-A inhibitory effect seems to be reached with 50 mg in humans (Waldmeier et al 1983), whereas a dose-dependent increase in the antidepressant effect between 50 and 150 mg daily has been reported (Schiwy et al. 1989). Very recently, Celada et al (1992) reported a study of plasma 5-HT/5-hydroxyindoleacetic acid (5-HIAA) and platelet 5-HT levels in patients treated for 6 weeks with 45 mg phenelzine or t50 mg brofaromine daily. Changes in the ratio of plasma 5-HT/5-HIAA indicated that both drugs inhibited MAO-A to a similar extent at the doses used. Phenelzine caused an three-fold iiicreasc ~ piatelet 5-HT levels, whereas the increase after brofaromine was only approximately 30%. Because it has long been known that platelet 5-HT concentrations strongly decrease after repeated treatment with 5-HT uptake inhibitors, these investigators interpreted the intermediate result with brofaromine as an indication of concurrent inhibition of both processes. This study prompted us to investigate the potential 5HT uptake-inhibiting effect of brofaromine in humans at clinical dosage more directly. As pointed out previously (Waldmeier et al 1o90), an obvious approach to this end (i.e., the measurement of uptake of radiolabeled 5-HT into platelets) may yield e."roneous results in the case of an MAO-A inhibitor because elevated levels of circulating 5HT could interfere. This problem can be circumvented by measuring the binding of the 5-HT uptake inhibitor 3Hparoxetine to intact platelets in platelet-rich plasma (PRP). 3H-paroxetine was shown to bind to the neuronal and the platelet 5-HT transporter in a very similar fashion (Mellerup and Plenge 1986). Its displacement from these sites by drugs inhibiting 5-HT uptake was shown to accurately reflect their uptake inhibitory properties (Mellerup et al 1983). Neither the number of binding sites nor the Kd value differed in depressed patients and control subjects (D'Haenen et al 1988). In intact platelets in PRP, 3H-paroxetine bound with an apparent/i'd value of 1.3 nmol/L, which dropped to 0..23
nmoFL on removal of plasma proteins by sedimentation and resuspension of the platelets (Waldmeier et al 1990). This latter value is comparable to what has been reported in platelet membranes (D'Haenen et al 1988). The fact that the apparent/i'd is higher in PRP is a consequence of the high plasma protein binding of paroxetine (Kaye et al 1989), which lowers the concentration of the radioligand available for binding to the 5-HT transporter. Because only tracer concentrations of 3H-paroxetine are used in binding assays, this is not likely to affect comparisons of the extent of binding between individuals. 3H-paroxetine can therefore be regarded as a good tool for studying 5-HT uptake inhibition in human subjects. In consequence, we conducted a 2-week study in healthy male volunteers with brofaromine and phenelzine versus placebo and measured the binding of 3H-paroxetine to the 5-HT uptake sites on intact platelets in PRP. For comparison with the study of Celada et al (1992), whole-blood 5-HT was also measured. Moreover, in an attempt to obtain a parameter for MAO-A inhibition, serum homovanillic acid (HVA) was determined. HVA is a metabolite in dopamine catabolism, which is also controlled by MAO-A. The validity of serum HVA for the assessment of MAO-A inhibition is unclear. Because the similarity in the extent of MAO-A inhibition by 45 mg phenelzine and 150 mg brofaromine has previously been shown by means of the change in the plasma 5-HT/5-ttlAA ratio (Celada et al 1992), serum HVA constitutes more an exploratory than an essential part of the study.
Materials and Methods Subjects
Eighteen healthy male subjects (mean age 31.4 years; range 23-43) with an average weight of 77.0 kg (range 03-97) participated in the study after having given informed, written consent. They had to be cleared for participation in human pharmacological studies by an independent phy. . . . . . . . . . . ,,, ,_,ut-, ,~ue,,,~ Department UI till~ Ulllversity Hospital, Basel. The protocol was approved by an external ethics committee.
Drugs
The following medications were used in this study: (1) brofaromine (CGP 11305 A) in the form of film-coated tablets containing 25, 50, or 75 mg of active substance (2) brofaromine-matching placebo; (3) phenelzine in the form of film-coated tablets containing 15 mg of active substance; and (4) phenelzine-matching placebo.
Brofaromine and 5-HT Uptake Inhibition
BIOLPSYCHIATRY 1993;33:373-379
Table I. Scheme of Drug Administration Day
Morning
Noon
Evening
Treatment A (Brofaromine) I 2 3 4 :,-13 14
25 50 50 50 75 75
mg mg mg mg mg mg
l 2 3 4 5-13 14
Placebo Placebo Placebo Placebo 15 mg 15 mg
25 mg 50 mg 50 mg 5f~ nag 75 nag
Placebo Placebo Placebo Placebo Placebo
Treatment B (Phenelzine) Placebo Placebo Placebo Placebo 15 mg
Placebo Placebo Placebo Placebo 15 mg
Treatment C (Placebol 1-13 14
Placebo Placebo
Placebo
Placebo
Treatment Schedule Th, ......... ;nt,~ntton" wa¢ . to treat . the. subjects . with each d_rug for 10 days at the same dose level. Because it is not advisable to start treatment of volumeers with the full dose of 150 mg brofaromine, a gradual increase from 50 mg on the first day to reach the maximal dose on the fifth day was chosen. To keep the duration of treatment at the full dose the same with both test drugs, phenelzine administration was started only on day 5. The study was conducted in three periods, with six volunteers each. They were assigned to one of the three possible treatments (A, B, or C; Table I) such that in every period, two subjects each received treatrqent A, B, or C. Within the periods, assignment was at random. There were no significant differences among the treatment groups with respect to weight or age. At all application times (usually at 7 AM, 3 PM, and 10 PM), each subject took two tablets: either one tablet with active compound (brofaromine or pheneizh, e) and one tablet with placebo matching the active compound of the other treatment group or two placebo tablets matching brofaromine and phenclzine, respectively (double-dummy technique). Restrictions During the entire trial period, the subjects were not allowed to take any medication, including over-the-counter drugs. If intake of a drug was necessary for any reason, the subjects had to inform the investigator immediately and
375
had to note drug, dose, and time of intake. Driving or operation of machines requiring unimpaired alertness was not allowed during the whole treatment period. From the day before hhe first drug intake until 48 hr after the last dose, the subjects were also not allowed to consume alcohol. During the same period, they had to avoid strenuous physical activities (sports). The relative safety of brofaromine in cases of concomitant consumption of tyraminerich food has been confirmed in earlier studies; however, because of double-blinding with the phenelzine group, full dietary restrictions for food .=ich in tyramine had to be be maintained by all subjects involved in the trial.
Blood Sampling On day 0 (before the start of treatment) and on treatment days 5, 8, 12, and 14, 30 ml blood was collected in the supine position via venous punction, on average 83 -+ 3 rain (mean - SE) after the intake of the morning dose of the drugs, using the Vacutainer system for the following investigations: (1) 15 ml of heparinized blood for the determination of serum HVA; (2) 2 ml of ethylenediaminetetraacetic acid (EDTA) blood for platelet counting (Sysmex p!ate!et counter PL!00; TOA Medical Electronics, Japan) and for the determination of whole-blood 5-HT; and (3) 10 ml of EDTA blood for the preparation of PRP for the paroxetine binding assay, by centrifugation for 20 min at 1100 rpm using a Beckman centrifuge TJ-6. Platelets were also counted in PRP.
Determination of SH-Paroxetine Binding to Platelets 3H-paroxetine binding was performed as described previously (Waldmeier et al 1990). Briefly, 800 p.I PRP together with 1t30 ILl Tris(hydroxymethyi)aminomethane ('Iris) buffer (0.15 mol/L; pH 7.5) or fluvoxamine (IO 4 mol/L Jn Tris buffer and 50 Ixl "Iris buffer and 50 p.l 3H-paroxetine (2 x 10-8 tool/L; sp act 28.8 Ci/mmol; New England Nuclear, Boston, MA; final concentration in incubation medium 10-9 tool/L) were allowed to stand at room temperature for 2 hr; they were gently shaken occasionally. Thereafter, the samples were centrifuged at 2900 rpm (maximum speed, Beckman TJ-6) for 10 rain. The supernatant was decanted as completely and carefully as possible; the pellet was rinsed with 0.2 ml ice-cold Tris buffer and the rinsing liquid was decanted as completely as possible. After the addition of 500 ~1 of 0. i mol/L HC1 and immediate stirring on the vortex, the pellet was subjected to repeated aspiration and blowing out into/from an Eppendorf pipette; 0.4 ml of the resulting turbid solution was counted using 10 ml Irgascint. Duplicates were made.
376
DIOLPSYCHIATRY
P.C. Waldmeicr et al
1993;33:373-379
Determination of Whole-Blood 5-HT To 200 ttl blood in an Eppendorf tube, 20 Ixl internal standard (2 ixg/ml bufotenin in 0.001 mol/L HCI), 50 ttl ascorbic acid (1.5 mol/L in H20), and 50 ttl 3.4 mol/L perchloric acid were added. After vigorous shaking (vortex) and centrifugation (Sigma centrifuge, 15,000 rpm for 15 min at 4°C), the supernatant was recentrifuged under the same conditions, and 100 Itl of it was diluted to 600 pl with 0.001 mol/L HCI; 200 ttl of this solution was injected for analysis by high-performance liquid chromatography (HPLC) [column: ttBondapak Ct8 3.9 × 300 ram; mobile phase: 0.03 mol/L citric acid, 0.06 mol/L NaeHPO4, 3% butanol, and 1 g/L sodium octylsulfate; flow: 1.5 ml/min; fluorometric detector (Merck-Hitachi F1000) 280/340 nm, sensitivity setting 0.5, time constant 3 sec (modified from Arfigas et al 1985; Ortiz et ~ !988].
Determ&ation of Serum Concentrations of HVA Blood was centrifuged for 10 min at 1100 g and subsequently for 20 min at 4500 g and stored frozen at - 8 0 ° C for one night. After thawing, I ml serum plus 1 ml of 0.15 mol/Lchloroacetate buffer pH 2.9 plus internal standard [1000 ng vanillic acid) were allowed to stand in ice for I hr before centrifugation at 48,000 g for 20 rain at 4°C. This extract was stored frozen at - 8 0 ° C for one night. It was then thawed, recentrifuged at 48,000 g for 20 min at 4°C, and loaded onto 1 ml Bond Elut SAX strong anion-exchange disposable solid-phase columns (Analytichem Int., Harbor City, CA) prewashed with 1 ml methanol, followed by 1 ml of H20. After passing the extract, the column was washed with 1 ml It20 and eluted wit-, 1 ,'rd of 0.15 mol/L of chloroacetic acid buffer of pH 2.9; 100 lxl of the supernatants was injected into an HPLC system consisting of an Altex 110-A pump, a i~Bondapak C~s 300 × 3.9-mm (Waters Assoc., Milford, MA) column, an LC 22 temperature controller, an LC 23 A column heater (Bioanalytical Systems, West Lafayette, IN), and an ESA 5100-A coulometric detector fitted with a model 5010 analytical cell (Environmental Sciences Assoc., Bedford, MA). The mobile phase consisted of 0.15 mol/L chloroacetate buffer (pH 2.9) containing 0.03 mmol/L sodium octylsulfate and 0.2 mmol/L EDTA. The column was operated at 40°C and the flow rate was 1 ml/min. The potential of detector 1 was set at +0.25 V. Detector 2 was switched off. Typical retention times were 8.2 min for vanillic acid and 9.5 min for HVA.
Statistics The results of the primary parameters (3H-paroxetine binding to platelets, 5-HT concentrations in blood, serum concentrations of HVA) were submitted to an analysis of co-
8
i
- - 7'
I
J
p[me~ane
+
~
~
brofammlne
5' 4
1 p=0.07
0
5
"
~ p=O,03
8
12
14
day
Figure 1. Effectsof treatmentwith placebo,phenelzine, or brofaromine on 3H-paroxetinebinding to platelets in human volunteers. Groupsof six volunteerswere treatedwith placebo(14 days), 15 mg phenelzine three times a day (beginning on day 5 ~i~til day !3; one 15-mg dose only on day 14), or brofaromine [25 mg twice daily (BID) on day 1; 50 nag BID on days 2-4; 75 mg BID on days 5-13; one 75-rag dose only on day !4]. Blood was withdrawn by venipuncture on days 0 (before treatment), 5, 8, 12, and 14, on average 83 -+ 3 min after the morning dose. 3H-paroxetine binding to platelets was determined in platelet-rich plasma. Data are mean +- SEM of the adjusted values (see Statistics) in fmol/10 I° platelets.
variance using the pretreatment values as covariates to test the changes observed between the values recorded before and at the subsequent times of treatment and to compare these changes in the three treatment groups. Using the pretreatment values as covariates allows adjustment of the values so that they correspond to those one could have observed if all subjects had the same pretreatment value (equal to the general mean). Results
SH-Paroxetine Binding Assay 3H-paroxetine binding in the brofaromine group was 2 0 % 25% ,I... . .| I L days; •II...... .r~l ut~tgJ I•n t lk~ t ~ , [--I^~Ak^ J t a l . , l ~ o u ~ l U U l J U l l t-,, i l l .U.U.a.I.. I.I.I U statistical significance was reached on day 14. In the phenelzine group, in contrast, 3H-paroxetine binding was 10%20% higher than in the placebo group on all treatment days; however, this did not reach statistical significance (Figure 1). It may be noteworthy that in all three groups, 3H-paroxetine binding was higher on the first (control) day than at the later time points. This held true for each single volunteer whatever treatment he received thereafter. Because the study was carried out in three trials during approximately 3 months, methodological problems or factors, such as ultradian rhythms, are unlikely to be involved. On the other hand, anxiety or stress might conceivably have influenced results.
Brofaromine and 5-HT Uptake Inhibition
15--0-- pl~ebo 14J 13J
BIOLPSVCHt~TRY 1993;33:373-379
**
**
I~tobm~
12 j 11~ ~- 10~ tO
22 ¸
2o
[
!
0
5
377
pl~Nmel~ne
12
91
== a:
10
71
8
0
5
8 day
12
14
** p ~ o.ot
8 day
12
14
"* p = o o l
Figure 2. Effects of treatment with placebo, phenelzine, or brofaromine on whole-blood 5-HT levels of human volunteers. Groups of six volunteers were treated wit~ placebo, phenelzine, or brofaromine (for treatment scheme and details, see Methods or Figure 1). 5-HT was determined in whole blood as a measure of platelet 5-HT. Data are means - SEM of the adjusted values (see Statistics) in ng/107 platelets.
Figure 3. Effects of treatment with placebo, phenelzine, or brofaromine on serum HVA levels of human volunteers. Groups of six volunteers were treated with placebo, phenelzine, or brofaromine (for treatment scheme and details, see Methods or Figure I). HVA wa~ determined in serum. Data are mean ___ SEM of the adjusted values (see Statistics) in ng/ml serum.
Blood 5-HT Concentrations
tween the mean inhibitory concentration 0C5o) values of 5-HT uptake inhibitors to inhibit 5-HT uptake and 3Hparoxetine binding (compare data in Hyttel 1982 or Maitre et al 1982 versus Lhose by Mellerup and Plenge !98b and Mellerup et al 1983), one may infer that the reduction in 3H-paroxetine binding by brofaromine in our study reflects 5-HT uptake inhibition. The figure of 25% is almost certainly an underestimation of what occurs in the brain for two reasons. First, brofaromine inhibited 3H-paroxetine binding to partly purified human platelets with an ICso of 0.3 Ixmol/L (imipramine - 0 . 0 4 IxM) but is totally inactive at 1 i~mol/L in platelet-rich plasma, whereas the effect of imipramine did not differ in the twe preparations. This is a consequence of the extraordinarily strong binding of brofaromine to plasma proteins (->98%), whereby its affinity to the latter seems to be higher than that to the 5-HT transporter. This attenuating factor may be of lesser i,mporta~ce in the b:~i~. Second, brain concentrations of brofaromine are clearly higher than those in the blood; 6 min after intravenous administration of 3 mg/kg 14C-labeled material, (i.e., at a time when metabolism was insignificant), total radioactivity per unit weight was 16 times higher in the brain than in the blood (H. P. Gschwind, personal communication). For this reason, one might expect a more marked 5-HT uptake inhibition in the brain than in platelets. The increase in whole-blood 5-HT concentrations obtained in this study with hrofaromine are comparable to those reported by Celada et al (1992) in their 6-weeks study. The increase in phenelzine was similar to that of brofaromine in our study, whereas it was much more marked in the study of Celada et al (1992)• This is not a contra-
Whole-blood 5-HT concentrations, which accurately refleet platelet 5-HT concentrations because 99% of wholeblood 5-HT is contained in thrombocytes, remained relatively stable throughout the s~dy p e ~ d (diffcrence :~Animax = 20%) in the placebo group. In the brofaromine group, 5-HT concentrations increased more or less steadily to approximately 140%-150% of the corresponding placebo level, significance being reached at days 8, 12, and 14. With phenelzine, where treatment only began on day 5, a con'espondine, but delayed increase occurred, significance being reached on days i2 and 14 (Figure 2).
Serum HVA Concentrations Serum HVA levels showed marked interindividual and intraindividual variability. In the phenelzine group, a tendency toward a decrease versus placebo was noted, which, however, did not reach statistical significance. In the brofaromine group, a somewhat more marked decrease occurred, which did reach statistical significance on days 5 and 14. Nevertheless, the decreases in serum HVA by these two MAO-A inhibitors under our conditions did not seem to reasonably reflect the extent of MAO-A inhibition that they cause in humans (Figure 3).
Discussion The results of this study show that brofaromine inhibits 3H-paroxetine binding to human thrombocytes ex vivo by roughly 25%. Because there is very good correlation be-
378
BIOLPSYCHIATRY
P.C. Waldmeier et al
1993;33:373-379
diction because it is known that the extent of MAO-A inhibition by irreversible MAO-A inhibitors increases during repeated treatment owing to a cumulative inactivation of the enzyme (Waldmeier et al 1981). It is therefore plausible that in our study, whole-blood 5-HT concentrations did not reach the maximal values attainable at that dose of phenelzine because the treatment period was not long enough. Finally, the results of the measurements of serum HVA levels suggest that this parameter is not particularly suitable for assessing MAO-A inhibition in humans. Although brofaromine (but not phenelzine) did cause significant decreases at least at two time points, they did not seem quantitatively related to the extent of MAO-A inhibition by the two MAO-A (e.g., as determined by the ratio of 5-HT/5-HIAA in plasma) (Celada et al 1992). Our results are comparable to those reported by Berlin et al (1990) with moclobemide and toloxatone. The reason for the poor reduction in serum r i v A oy M~O-pt mmo~tors ts unclear~ it is unlikely that it is related to a proposed role for MAOB in dopamine deamination in humans because phenelzine as a nonselective inhibitor should then have caused a greater effect. On the other hand, the relatively slow turnover (half-life in plasma > 1 hr) and the high apparent distribution volume may play a role (Elchisak et al 1982). The principal conclusion from the present study is that brofaromine at the clinically used dosage of 150 mg/day
does indeed inhibit 5-HT uptake, as evidenced by ,,,~asurements of 3H-paroxetme binding to platelets. There :re reasons to believe that the extent of the effect o f brofaromine is greater in the brain. Because there is at present no experimental possibility available to study cerebral 5HT uptake in humans, we have to content ourselves with these data for the time being. The occurrence of both MAO-A inhibition and 5-HT uptake inhibition, although presumably at different levels, may have implications for the brofaromine therapeutic pro ~le because these properties are likely to be synergistic with respect to their impact on serotonergic transmission. Combinations of irreversible MAO-A inhibitors with 5-HT uptake inhibitors can produce severe adverse effects (see, e.g., Blackwell 1991). Although the occurrence of similar adverse effects might be expected with a drug possessing both properties, no such evidence has been obtained up to this point from clinical studies in approxim-".tc~y 1700 patients receiving brofaromine. This may be related to the specificity and reversibility of its interaction with MAO-A, which leaves safety valves open for the metabolism of excess 5-HT. In this context, it may be of interest that there is no evidence for the occurrence of such adverse effects when mociobemide, another reversible, preferential M A O - A inhibitor, is combined with 5-HT uptake inhibitors (Amrein et al 1992).
References Amrein R, Giintert TW, Dingemanse J, Lorscheid T, Stabl M, Schmid-Burgk W (1992): Interactions of moclobemide with concomitantly administered medication: evidence from pharmacological and clinical studies. Psychopharmacology 106:$24-$31. Arora RC, Meltzer HY ( 1983): Effects of amoxapine on serotonin uptake in human bleed platelets of depressed patients and ~rmal controls. Psychiatry Res 9:29-36. Artigas F, Sarrias MJ, Martinez E, Gelpi E ( 1985): Serotonin in body fluids: characterization of human plasmatic and cerebrospinal fluid pools by means of a new HPLC method. Life Sci 37:441-447. Berlin 1, Zimmer R, Thiede H-M, et al (1990): Comparison of the monoamine oxidase inhibiting properties of two reversible and selective monoamine oxidase-A inhibitors moclobemide and toloxatone, and assessment of their effect on psychometric performance in health:,' ~,ubjects. Br J Clin Pharraacol 30:805-816. Blackwell B (1991): Monoamine oxidase inhibitor interactions with other drugs. J Clin Psychopharmacol 11:55-58. Celada P, P6rez J, Alvarez E, Artigas F (1992): Monoamine oxidase inhibitors phenelzine and brofaromine increase nlasma serot,min and decrease 5-hydroxyindoleacetic acid in major depressive patients. Relationship to clinical improvement. J Clin Psychopharmacol 12:309-315. D'Haenen H, De Waele M, Leysen JE (1988): Platelet 3H-pa-
roxetine binding in depressed patients. Psychiatry Res 26:1117.
Elchisak MA, Polinsky RJ, Ebert MH, Kopin IJ (1982): Kinetics of homovanillic acid and determination of its production rate in humans. J Neurochem 38:380-385. Hyttei J (1982): Citalopram-pharmacologicalprofile of a specific serotorun uptake inhibitor with antidepressant activity. Pro~ Neurop ¢ychopharmacol Biol Psychiatry 6:277-295. Kaye CM, Haddock RE, Langley PF, et a! (1989): A review of the metabolism and pharmacokinetics of paroxetine in mrm. Acta Psychiatr Scand ( Suppl) 350:60-75. Langer G, Heimann H (1983): Psychopharmaka. Grundlagen und Therapie. Wien: Springer, p 667. Maitre L, Baumann PA, Jaekel J, Waldmeier PC (1982): 5-HT uptake inhibitors: psychopharmacological and neurobiochemical criteria of selectivity. In Ho BT, Schoolar JC, Usdin E (eds), Serotonin in Biological Psychiatry. New York: Raven Press, pp 229-246. Mellerup ET, Plenge P (1986): High affinity binding of [3H]paroxetine to rat neuronal membranes. Psychopharmacology 89:436-439. Mellerup ET, Plenge P, Engelstoft M (1983): High affinity binding of [3]paroxetine and [3H]imipramine to human platelet membranes. Eur J Pharmacol 96:303-309.
Brofaromine and 5-HT Uptake Inhibition
BIOLPSYCHIATRY
379
1993;33:373-379
G'rtiz 3, ,~figas F, Ge!pi E (198s): Serotoninergic status in human blood. Ltfe gel ,43:983-990. Schiwy W, Heath WR, Delini-Stula A (1989): Therapeutic and side-effect profile of a selective and reversible MAO-A inhibitor, brofaromine: results of dose-finding Irials in depressed patients. J Neural Transm (Suppl) 28:33-44. Tuomisto J, Tukiainen E, Ahlfors UG (1979): Decreased uptake of 5-hydroxytryptamine in blood platelets from patients with endogenous depression. Psychopharmacology 65:141-147. Waldmeier PC, Baumann PA (1983): Effects of CGP 11305 A, a new reversible and selective inhibitor of MAC) A, on biogenie amine levels and metabolism in the rat brain. Naunyn Schmiedebergs Arch Pharmacol 324:20-26. Waldmeier PC, St0cklin K (1989): The reversible MAO inhib-
itor, brofaromine, inhibits serotonin uptake in vivo. Fur J Pharmacol 169:197-204. Waldmeier PC, Felner AE, Maitre L (1981): Long-term effects of selective MAO inhibitors on MAO activity and amine metabolism. In Youdim MB, Paykel ES (eds), Monoamine Oxidase lnhibitors--The State of The Art. Chiche~ter" Wiley, pp 87-102. Waldrnei~r FC, Antonin KH, Feldtrauer JJ, et al (1983): Urinary excretion of O-methylated catecholamines, tyramine, and phenylethylamine in human volunteers treated with tranylcypromine and CGP ! 1305 A. EurJ Clin Pharmaco125:361368. Waldmeier PC, Kraetz J, Maitre L (1990): Newer aspects of the selective, reversible inhibitor of MAO-A, brofaromine. Clin Neuropharmacol i 3(suppl 2): 196-197.
373
Serotonin Uptake Inhibition by the Monoamine Oxidase Inhibitor Brofaromine P. C. Waldmeier, T. Graf, M. Germer, J. J. Feldtrauer, and H. Howald
The selective, reversible monoamine oxidase (MAO) A inhibitor brofaromine inhibits serotonin (5-HT) uptake in animal models in vitro and in vivo. We investigated whether such an effect can be demonstrated at clinical doses in humans by treating three groups of six volumeers with either placebo, 15 mg phenelzine three limes a day, or 75 mg brofaromine twice a day in a 2-week experiment. As an indirect, although relevant parameter, binding of 3H-paroxetine to the 5-HT uptake sites on blood plate[ets was assessed. Moreover, wholeblood 5-HT as a measure of platelet 5-HT, and serum hamovanillic acid (HVA) to tento~ve~ estimate MAO inhibition, were determined. Brofaromiae reduced 3H-paroxetine binding to platelets compared with placebo by 20%-25% throughc~ut the treatment period, significance being reached on the last treatment day. In contrast, Fhenelzine tended to increase 3H-paroxetine binding. Both drugs increased whole-blood 5-HT to approximately 140%-150%. Brofaromine moderately and on some days significantly decreased serum HVA, whereas phenelzine only tended to do so. Our results suggest that brofaromine at the clinically used dosage of 150 rag/day does indeed inhibit 5-HT uptake, as evidenced by measurements of ~H-paroxetine binding to platelets.
Key Words: 5-HT uptake, 3H-paroxetine binding, blood platelets, MAO inhibitor, brofaromine, volunteers
Introduc¢ion Brofaromine is a potent, l c v e l S l .O. l.e. ,. . .a n uJ ~pel~llIIS "~ ,tmottt, t of monoamine oxidase A (MAO-A), which is currently in phase III clinical trials for depression and severe anxiety states. It has long been known that the compound also possesses serotonin (5-HT) Ul~',ake-inhibitiag properties (Waldmeier and Baumann 1983); however, they were initially considered to be of minor importance because the
From the Research and Development Depar~r0ent: Pharmaceuticals Division, CibaAddress reprint requests to Peter Waldmeier, Ph.D., Research Department, Pharmaceuticals Division. K-125/607, Ciba-Geigy Ltd. CH.4002 Basel, Switzerland. Received April 25. 1992: revised October 1. 1992. © 1993 Society of Biological Psychiatry.
corresponding median effective dose value was approximately 10-30 times higher than that of MAO-A inhibition (Waldmeier and Baumann 1983; Waldmeier and St6cklin 1989). In rat brain synaptosomes in vitro, brofaromine is three ar~ five times less potent than imipramine and flucx etine, respectively, in inhibiting 5-HT uptake; however, when the 5-HT uptake inhibitory effects are assessed ex vivo (i.e~, in synaptosomes prepared from pretreated rats) or in vivo by means of antagonistic effects on the binding of 3H-cyanoimipramine to the 5-HT transporter in the rat brain, it is two to three times more potent than imipramine and four to five times less potent than fluoxetine (Waldmeier and Baumann 1983; Waldmeier and St6cklin 1989). The clinical daily doses of brofaromine and imipramine are similar [100-150 mg for brofarominc (Schiwy et al 0006-3223/93/$06.00
374
BIOLPSYCHIATRY
P.C. Waldmeier et al
1993;33:373-379
1989); 75-200 mg for imipramine (Langer and Heimann 1983)], and those of fluoxetine are at least five times lower (Physicians" Desk Reference, 45th ed., 1991). 5-HT uptake into platelets of patients treated with daily doses of 50-300 mg imipramine was clearly inhibited (Tuomisto et al 1979; Arora and Meltzer 1983). This may suggest that because brofaromine is more potent than imipramine as a 5-HT uptake inhibitor in animal models, such an effect could also occur at clinical doses in patients. A similar estimation can be made from ,~comparison of clinical doses of brofaromine and fluoxetine (100-150 versus 20-40 mg daily). Moreover, a component of its antidepressant action independent of MAO-A inhibition is indicated by the observation that the maximal MAO-A inhibitory effect seems to be reached with 50 mg in humans (Waldmeier et al 1983), whereas a dose-dependent increase in the antidepressant effect between 50 and 150 mg daily has been reported (Schiwy et al. 1989). Very recently, Celada et al (1992) reported a study of plasma 5-HT/5-hydroxyindoleacetic acid (5-HIAA) and platelet 5-HT levels in patients treated for 6 weeks with 45 mg phenelzine or t50 mg brofaromine daily. Changes in the ratio of plasma 5-HT/5-HIAA indicated that both drugs inhibited MAO-A to a similar extent at the doses used. Phenelzine caused an three-fold iiicreasc ~ piatelet 5-HT levels, whereas the increase after brofaromine was only approximately 30%. Because it has long been known that platelet 5-HT concentrations strongly decrease after repeated treatment with 5-HT uptake inhibitors, these investigators interpreted the intermediate result with brofaromine as an indication of concurrent inhibition of both processes. This study prompted us to investigate the potential 5HT uptake-inhibiting effect of brofaromine in humans at clinical dosage more directly. As pointed out previously (Waldmeier et al 1o90), an obvious approach to this end (i.e., the measurement of uptake of radiolabeled 5-HT into platelets) may yield e."roneous results in the case of an MAO-A inhibitor because elevated levels of circulating 5HT could interfere. This problem can be circumvented by measuring the binding of the 5-HT uptake inhibitor 3Hparoxetine to intact platelets in platelet-rich plasma (PRP). 3H-paroxetine was shown to bind to the neuronal and the platelet 5-HT transporter in a very similar fashion (Mellerup and Plenge 1986). Its displacement from these sites by drugs inhibiting 5-HT uptake was shown to accurately reflect their uptake inhibitory properties (Mellerup et al 1983). Neither the number of binding sites nor the Kd value differed in depressed patients and control subjects (D'Haenen et al 1988). In intact platelets in PRP, 3H-paroxetine bound with an apparent/i'd value of 1.3 nmol/L, which dropped to 0..23
nmoFL on removal of plasma proteins by sedimentation and resuspension of the platelets (Waldmeier et al 1990). This latter value is comparable to what has been reported in platelet membranes (D'Haenen et al 1988). The fact that the apparent/i'd is higher in PRP is a consequence of the high plasma protein binding of paroxetine (Kaye et al 1989), which lowers the concentration of the radioligand available for binding to the 5-HT transporter. Because only tracer concentrations of 3H-paroxetine are used in binding assays, this is not likely to affect comparisons of the extent of binding between individuals. 3H-paroxetine can therefore be regarded as a good tool for studying 5-HT uptake inhibition in human subjects. In consequence, we conducted a 2-week study in healthy male volunteers with brofaromine and phenelzine versus placebo and measured the binding of 3H-paroxetine to the 5-HT uptake sites on intact platelets in PRP. For comparison with the study of Celada et al (1992), whole-blood 5-HT was also measured. Moreover, in an attempt to obtain a parameter for MAO-A inhibition, serum homovanillic acid (HVA) was determined. HVA is a metabolite in dopamine catabolism, which is also controlled by MAO-A. The validity of serum HVA for the assessment of MAO-A inhibition is unclear. Because the similarity in the extent of MAO-A inhibition by 45 mg phenelzine and 150 mg brofaromine has previously been shown by means of the change in the plasma 5-HT/5-ttlAA ratio (Celada et al 1992), serum HVA constitutes more an exploratory than an essential part of the study.
Materials and Methods Subjects
Eighteen healthy male subjects (mean age 31.4 years; range 23-43) with an average weight of 77.0 kg (range 03-97) participated in the study after having given informed, written consent. They had to be cleared for participation in human pharmacological studies by an independent phy. . . . . . . . . . . ,,, ,_,ut-, ,~ue,,,~ Department UI till~ Ulllversity Hospital, Basel. The protocol was approved by an external ethics committee.
Drugs
The following medications were used in this study: (1) brofaromine (CGP 11305 A) in the form of film-coated tablets containing 25, 50, or 75 mg of active substance (2) brofaromine-matching placebo; (3) phenelzine in the form of film-coated tablets containing 15 mg of active substance; and (4) phenelzine-matching placebo.
Brofaromine and 5-HT Uptake Inhibition
BIOLPSYCHIATRY 1993;33:373-379
Table I. Scheme of Drug Administration Day
Morning
Noon
Evening
Treatment A (Brofaromine) I 2 3 4 :,-13 14
25 50 50 50 75 75
mg mg mg mg mg mg
l 2 3 4 5-13 14
Placebo Placebo Placebo Placebo 15 mg 15 mg
25 mg 50 mg 50 mg 5f~ nag 75 nag
Placebo Placebo Placebo Placebo Placebo
Treatment B (Phenelzine) Placebo Placebo Placebo Placebo 15 mg
Placebo Placebo Placebo Placebo 15 mg
Treatment C (Placebol 1-13 14
Placebo Placebo
Placebo
Placebo
Treatment Schedule Th, ......... ;nt,~ntton" wa¢ . to treat . the. subjects . with each d_rug for 10 days at the same dose level. Because it is not advisable to start treatment of volumeers with the full dose of 150 mg brofaromine, a gradual increase from 50 mg on the first day to reach the maximal dose on the fifth day was chosen. To keep the duration of treatment at the full dose the same with both test drugs, phenelzine administration was started only on day 5. The study was conducted in three periods, with six volunteers each. They were assigned to one of the three possible treatments (A, B, or C; Table I) such that in every period, two subjects each received treatrqent A, B, or C. Within the periods, assignment was at random. There were no significant differences among the treatment groups with respect to weight or age. At all application times (usually at 7 AM, 3 PM, and 10 PM), each subject took two tablets: either one tablet with active compound (brofaromine or pheneizh, e) and one tablet with placebo matching the active compound of the other treatment group or two placebo tablets matching brofaromine and phenclzine, respectively (double-dummy technique). Restrictions During the entire trial period, the subjects were not allowed to take any medication, including over-the-counter drugs. If intake of a drug was necessary for any reason, the subjects had to inform the investigator immediately and
375
had to note drug, dose, and time of intake. Driving or operation of machines requiring unimpaired alertness was not allowed during the whole treatment period. From the day before hhe first drug intake until 48 hr after the last dose, the subjects were also not allowed to consume alcohol. During the same period, they had to avoid strenuous physical activities (sports). The relative safety of brofaromine in cases of concomitant consumption of tyraminerich food has been confirmed in earlier studies; however, because of double-blinding with the phenelzine group, full dietary restrictions for food .=ich in tyramine had to be be maintained by all subjects involved in the trial.
Blood Sampling On day 0 (before the start of treatment) and on treatment days 5, 8, 12, and 14, 30 ml blood was collected in the supine position via venous punction, on average 83 -+ 3 rain (mean - SE) after the intake of the morning dose of the drugs, using the Vacutainer system for the following investigations: (1) 15 ml of heparinized blood for the determination of serum HVA; (2) 2 ml of ethylenediaminetetraacetic acid (EDTA) blood for platelet counting (Sysmex p!ate!et counter PL!00; TOA Medical Electronics, Japan) and for the determination of whole-blood 5-HT; and (3) 10 ml of EDTA blood for the preparation of PRP for the paroxetine binding assay, by centrifugation for 20 min at 1100 rpm using a Beckman centrifuge TJ-6. Platelets were also counted in PRP.
Determination of SH-Paroxetine Binding to Platelets 3H-paroxetine binding was performed as described previously (Waldmeier et al 1990). Briefly, 800 p.I PRP together with 1t30 ILl Tris(hydroxymethyi)aminomethane ('Iris) buffer (0.15 mol/L; pH 7.5) or fluvoxamine (IO 4 mol/L Jn Tris buffer and 50 Ixl "Iris buffer and 50 p.l 3H-paroxetine (2 x 10-8 tool/L; sp act 28.8 Ci/mmol; New England Nuclear, Boston, MA; final concentration in incubation medium 10-9 tool/L) were allowed to stand at room temperature for 2 hr; they were gently shaken occasionally. Thereafter, the samples were centrifuged at 2900 rpm (maximum speed, Beckman TJ-6) for 10 rain. The supernatant was decanted as completely and carefully as possible; the pellet was rinsed with 0.2 ml ice-cold Tris buffer and the rinsing liquid was decanted as completely as possible. After the addition of 500 ~1 of 0. i mol/L HC1 and immediate stirring on the vortex, the pellet was subjected to repeated aspiration and blowing out into/from an Eppendorf pipette; 0.4 ml of the resulting turbid solution was counted using 10 ml Irgascint. Duplicates were made.
376
DIOLPSYCHIATRY
P.C. Waldmeicr et al
1993;33:373-379
Determination of Whole-Blood 5-HT To 200 ttl blood in an Eppendorf tube, 20 Ixl internal standard (2 ixg/ml bufotenin in 0.001 mol/L HCI), 50 ttl ascorbic acid (1.5 mol/L in H20), and 50 ttl 3.4 mol/L perchloric acid were added. After vigorous shaking (vortex) and centrifugation (Sigma centrifuge, 15,000 rpm for 15 min at 4°C), the supernatant was recentrifuged under the same conditions, and 100 Itl of it was diluted to 600 pl with 0.001 mol/L HCI; 200 ttl of this solution was injected for analysis by high-performance liquid chromatography (HPLC) [column: ttBondapak Ct8 3.9 × 300 ram; mobile phase: 0.03 mol/L citric acid, 0.06 mol/L NaeHPO4, 3% butanol, and 1 g/L sodium octylsulfate; flow: 1.5 ml/min; fluorometric detector (Merck-Hitachi F1000) 280/340 nm, sensitivity setting 0.5, time constant 3 sec (modified from Arfigas et al 1985; Ortiz et ~ !988].
Determ&ation of Serum Concentrations of HVA Blood was centrifuged for 10 min at 1100 g and subsequently for 20 min at 4500 g and stored frozen at - 8 0 ° C for one night. After thawing, I ml serum plus 1 ml of 0.15 mol/Lchloroacetate buffer pH 2.9 plus internal standard [1000 ng vanillic acid) were allowed to stand in ice for I hr before centrifugation at 48,000 g for 20 rain at 4°C. This extract was stored frozen at - 8 0 ° C for one night. It was then thawed, recentrifuged at 48,000 g for 20 min at 4°C, and loaded onto 1 ml Bond Elut SAX strong anion-exchange disposable solid-phase columns (Analytichem Int., Harbor City, CA) prewashed with 1 ml methanol, followed by 1 ml of H20. After passing the extract, the column was washed with 1 ml It20 and eluted wit-, 1 ,'rd of 0.15 mol/L of chloroacetic acid buffer of pH 2.9; 100 lxl of the supernatants was injected into an HPLC system consisting of an Altex 110-A pump, a i~Bondapak C~s 300 × 3.9-mm (Waters Assoc., Milford, MA) column, an LC 22 temperature controller, an LC 23 A column heater (Bioanalytical Systems, West Lafayette, IN), and an ESA 5100-A coulometric detector fitted with a model 5010 analytical cell (Environmental Sciences Assoc., Bedford, MA). The mobile phase consisted of 0.15 mol/L chloroacetate buffer (pH 2.9) containing 0.03 mmol/L sodium octylsulfate and 0.2 mmol/L EDTA. The column was operated at 40°C and the flow rate was 1 ml/min. The potential of detector 1 was set at +0.25 V. Detector 2 was switched off. Typical retention times were 8.2 min for vanillic acid and 9.5 min for HVA.
Statistics The results of the primary parameters (3H-paroxetine binding to platelets, 5-HT concentrations in blood, serum concentrations of HVA) were submitted to an analysis of co-
8
i
- - 7'
I
J
p[me~ane
+
~
~
brofammlne
5' 4
1 p=0.07
0
5
"
~ p=O,03
8
12
14
day
Figure 1. Effectsof treatmentwith placebo,phenelzine, or brofaromine on 3H-paroxetinebinding to platelets in human volunteers. Groupsof six volunteerswere treatedwith placebo(14 days), 15 mg phenelzine three times a day (beginning on day 5 ~i~til day !3; one 15-mg dose only on day 14), or brofaromine [25 mg twice daily (BID) on day 1; 50 nag BID on days 2-4; 75 mg BID on days 5-13; one 75-rag dose only on day !4]. Blood was withdrawn by venipuncture on days 0 (before treatment), 5, 8, 12, and 14, on average 83 -+ 3 min after the morning dose. 3H-paroxetine binding to platelets was determined in platelet-rich plasma. Data are mean +- SEM of the adjusted values (see Statistics) in fmol/10 I° platelets.
variance using the pretreatment values as covariates to test the changes observed between the values recorded before and at the subsequent times of treatment and to compare these changes in the three treatment groups. Using the pretreatment values as covariates allows adjustment of the values so that they correspond to those one could have observed if all subjects had the same pretreatment value (equal to the general mean). Results
SH-Paroxetine Binding Assay 3H-paroxetine binding in the brofaromine group was 2 0 % 25% ,I... . .| I L days; •II...... .r~l ut~tgJ I•n t lk~ t ~ , [--I^~Ak^ J t a l . , l ~ o u ~ l U U l J U l l t-,, i l l .U.U.a.I.. I.I.I U statistical significance was reached on day 14. In the phenelzine group, in contrast, 3H-paroxetine binding was 10%20% higher than in the placebo group on all treatment days; however, this did not reach statistical significance (Figure 1). It may be noteworthy that in all three groups, 3H-paroxetine binding was higher on the first (control) day than at the later time points. This held true for each single volunteer whatever treatment he received thereafter. Because the study was carried out in three trials during approximately 3 months, methodological problems or factors, such as ultradian rhythms, are unlikely to be involved. On the other hand, anxiety or stress might conceivably have influenced results.
Brofaromine and 5-HT Uptake Inhibition
15--0-- pl~ebo 14J 13J
BIOLPSVCHt~TRY 1993;33:373-379
**
**
I~tobm~
12 j 11~ ~- 10~ tO
22 ¸
2o
[
!
0
5
377
pl~Nmel~ne
12
91
== a:
10
71
8
0
5
8 day
12
14
** p ~ o.ot
8 day
12
14
"* p = o o l
Figure 2. Effects of treatment with placebo, phenelzine, or brofaromine on whole-blood 5-HT levels of human volunteers. Groups of six volunteers were treated wit~ placebo, phenelzine, or brofaromine (for treatment scheme and details, see Methods or Figure 1). 5-HT was determined in whole blood as a measure of platelet 5-HT. Data are means - SEM of the adjusted values (see Statistics) in ng/107 platelets.
Figure 3. Effects of treatment with placebo, phenelzine, or brofaromine on serum HVA levels of human volunteers. Groups of six volunteers were treated with placebo, phenelzine, or brofaromine (for treatment scheme and details, see Methods or Figure I). HVA wa~ determined in serum. Data are mean ___ SEM of the adjusted values (see Statistics) in ng/ml serum.
Blood 5-HT Concentrations
tween the mean inhibitory concentration 0C5o) values of 5-HT uptake inhibitors to inhibit 5-HT uptake and 3Hparoxetine binding (compare data in Hyttel 1982 or Maitre et al 1982 versus Lhose by Mellerup and Plenge !98b and Mellerup et al 1983), one may infer that the reduction in 3H-paroxetine binding by brofaromine in our study reflects 5-HT uptake inhibition. The figure of 25% is almost certainly an underestimation of what occurs in the brain for two reasons. First, brofaromine inhibited 3H-paroxetine binding to partly purified human platelets with an ICso of 0.3 Ixmol/L (imipramine - 0 . 0 4 IxM) but is totally inactive at 1 i~mol/L in platelet-rich plasma, whereas the effect of imipramine did not differ in the twe preparations. This is a consequence of the extraordinarily strong binding of brofaromine to plasma proteins (->98%), whereby its affinity to the latter seems to be higher than that to the 5-HT transporter. This attenuating factor may be of lesser i,mporta~ce in the b:~i~. Second, brain concentrations of brofaromine are clearly higher than those in the blood; 6 min after intravenous administration of 3 mg/kg 14C-labeled material, (i.e., at a time when metabolism was insignificant), total radioactivity per unit weight was 16 times higher in the brain than in the blood (H. P. Gschwind, personal communication). For this reason, one might expect a more marked 5-HT uptake inhibition in the brain than in platelets. The increase in whole-blood 5-HT concentrations obtained in this study with hrofaromine are comparable to those reported by Celada et al (1992) in their 6-weeks study. The increase in phenelzine was similar to that of brofaromine in our study, whereas it was much more marked in the study of Celada et al (1992)• This is not a contra-
Whole-blood 5-HT concentrations, which accurately refleet platelet 5-HT concentrations because 99% of wholeblood 5-HT is contained in thrombocytes, remained relatively stable throughout the s~dy p e ~ d (diffcrence :~Animax = 20%) in the placebo group. In the brofaromine group, 5-HT concentrations increased more or less steadily to approximately 140%-150% of the corresponding placebo level, significance being reached at days 8, 12, and 14. With phenelzine, where treatment only began on day 5, a con'espondine, but delayed increase occurred, significance being reached on days i2 and 14 (Figure 2).
Serum HVA Concentrations Serum HVA levels showed marked interindividual and intraindividual variability. In the phenelzine group, a tendency toward a decrease versus placebo was noted, which, however, did not reach statistical significance. In the brofaromine group, a somewhat more marked decrease occurred, which did reach statistical significance on days 5 and 14. Nevertheless, the decreases in serum HVA by these two MAO-A inhibitors under our conditions did not seem to reasonably reflect the extent of MAO-A inhibition that they cause in humans (Figure 3).
Discussion The results of this study show that brofaromine inhibits 3H-paroxetine binding to human thrombocytes ex vivo by roughly 25%. Because there is very good correlation be-
378
BIOLPSYCHIATRY
P.C. Waldmeier et al
1993;33:373-379
diction because it is known that the extent of MAO-A inhibition by irreversible MAO-A inhibitors increases during repeated treatment owing to a cumulative inactivation of the enzyme (Waldmeier et al 1981). It is therefore plausible that in our study, whole-blood 5-HT concentrations did not reach the maximal values attainable at that dose of phenelzine because the treatment period was not long enough. Finally, the results of the measurements of serum HVA levels suggest that this parameter is not particularly suitable for assessing MAO-A inhibition in humans. Although brofaromine (but not phenelzine) did cause significant decreases at least at two time points, they did not seem quantitatively related to the extent of MAO-A inhibition by the two MAO-A (e.g., as determined by the ratio of 5-HT/5-HIAA in plasma) (Celada et al 1992). Our results are comparable to those reported by Berlin et al (1990) with moclobemide and toloxatone. The reason for the poor reduction in serum r i v A oy M~O-pt mmo~tors ts unclear~ it is unlikely that it is related to a proposed role for MAOB in dopamine deamination in humans because phenelzine as a nonselective inhibitor should then have caused a greater effect. On the other hand, the relatively slow turnover (half-life in plasma > 1 hr) and the high apparent distribution volume may play a role (Elchisak et al 1982). The principal conclusion from the present study is that brofaromine at the clinically used dosage of 150 mg/day
does indeed inhibit 5-HT uptake, as evidenced by ,,,~asurements of 3H-paroxetme binding to platelets. There :re reasons to believe that the extent of the effect o f brofaromine is greater in the brain. Because there is at present no experimental possibility available to study cerebral 5HT uptake in humans, we have to content ourselves with these data for the time being. The occurrence of both MAO-A inhibition and 5-HT uptake inhibition, although presumably at different levels, may have implications for the brofaromine therapeutic pro ~le because these properties are likely to be synergistic with respect to their impact on serotonergic transmission. Combinations of irreversible MAO-A inhibitors with 5-HT uptake inhibitors can produce severe adverse effects (see, e.g., Blackwell 1991). Although the occurrence of similar adverse effects might be expected with a drug possessing both properties, no such evidence has been obtained up to this point from clinical studies in approxim-".tc~y 1700 patients receiving brofaromine. This may be related to the specificity and reversibility of its interaction with MAO-A, which leaves safety valves open for the metabolism of excess 5-HT. In this context, it may be of interest that there is no evidence for the occurrence of such adverse effects when mociobemide, another reversible, preferential M A O - A inhibitor, is combined with 5-HT uptake inhibitors (Amrein et al 1992).
References Amrein R, Giintert TW, Dingemanse J, Lorscheid T, Stabl M, Schmid-Burgk W (1992): Interactions of moclobemide with concomitantly administered medication: evidence from pharmacological and clinical studies. Psychopharmacology 106:$24-$31. Arora RC, Meltzer HY ( 1983): Effects of amoxapine on serotonin uptake in human bleed platelets of depressed patients and ~rmal controls. Psychiatry Res 9:29-36. Artigas F, Sarrias MJ, Martinez E, Gelpi E ( 1985): Serotonin in body fluids: characterization of human plasmatic and cerebrospinal fluid pools by means of a new HPLC method. Life Sci 37:441-447. Berlin 1, Zimmer R, Thiede H-M, et al (1990): Comparison of the monoamine oxidase inhibiting properties of two reversible and selective monoamine oxidase-A inhibitors moclobemide and toloxatone, and assessment of their effect on psychometric performance in health:,' ~,ubjects. Br J Clin Pharraacol 30:805-816. Blackwell B (1991): Monoamine oxidase inhibitor interactions with other drugs. J Clin Psychopharmacol 11:55-58. Celada P, P6rez J, Alvarez E, Artigas F (1992): Monoamine oxidase inhibitors phenelzine and brofaromine increase nlasma serot,min and decrease 5-hydroxyindoleacetic acid in major depressive patients. Relationship to clinical improvement. J Clin Psychopharmacol 12:309-315. D'Haenen H, De Waele M, Leysen JE (1988): Platelet 3H-pa-
roxetine binding in depressed patients. Psychiatry Res 26:1117.
Elchisak MA, Polinsky RJ, Ebert MH, Kopin IJ (1982): Kinetics of homovanillic acid and determination of its production rate in humans. J Neurochem 38:380-385. Hyttei J (1982): Citalopram-pharmacologicalprofile of a specific serotorun uptake inhibitor with antidepressant activity. Pro~ Neurop ¢ychopharmacol Biol Psychiatry 6:277-295. Kaye CM, Haddock RE, Langley PF, et a! (1989): A review of the metabolism and pharmacokinetics of paroxetine in mrm. Acta Psychiatr Scand ( Suppl) 350:60-75. Langer G, Heimann H (1983): Psychopharmaka. Grundlagen und Therapie. Wien: Springer, p 667. Maitre L, Baumann PA, Jaekel J, Waldmeier PC (1982): 5-HT uptake inhibitors: psychopharmacological and neurobiochemical criteria of selectivity. In Ho BT, Schoolar JC, Usdin E (eds), Serotonin in Biological Psychiatry. New York: Raven Press, pp 229-246. Mellerup ET, Plenge P (1986): High affinity binding of [3H]paroxetine to rat neuronal membranes. Psychopharmacology 89:436-439. Mellerup ET, Plenge P, Engelstoft M (1983): High affinity binding of [3]paroxetine and [3H]imipramine to human platelet membranes. Eur J Pharmacol 96:303-309.
Brofaromine and 5-HT Uptake Inhibition
BIOLPSYCHIATRY
379
1993;33:373-379
G'rtiz 3, ,~figas F, Ge!pi E (198s): Serotoninergic status in human blood. Ltfe gel ,43:983-990. Schiwy W, Heath WR, Delini-Stula A (1989): Therapeutic and side-effect profile of a selective and reversible MAO-A inhibitor, brofaromine: results of dose-finding Irials in depressed patients. J Neural Transm (Suppl) 28:33-44. Tuomisto J, Tukiainen E, Ahlfors UG (1979): Decreased uptake of 5-hydroxytryptamine in blood platelets from patients with endogenous depression. Psychopharmacology 65:141-147. Waldmeier PC, Baumann PA (1983): Effects of CGP 11305 A, a new reversible and selective inhibitor of MAC) A, on biogenie amine levels and metabolism in the rat brain. Naunyn Schmiedebergs Arch Pharmacol 324:20-26. Waldmeier PC, St0cklin K (1989): The reversible MAO inhib-
itor, brofaromine, inhibits serotonin uptake in vivo. Fur J Pharmacol 169:197-204. Waldmeier PC, Felner AE, Maitre L (1981): Long-term effects of selective MAO inhibitors on MAO activity and amine metabolism. In Youdim MB, Paykel ES (eds), Monoamine Oxidase lnhibitors--The State of The Art. Chiche~ter" Wiley, pp 87-102. Waldrnei~r FC, Antonin KH, Feldtrauer JJ, et al (1983): Urinary excretion of O-methylated catecholamines, tyramine, and phenylethylamine in human volunteers treated with tranylcypromine and CGP ! 1305 A. EurJ Clin Pharmaco125:361368. Waldmeier PC, Kraetz J, Maitre L (1990): Newer aspects of the selective, reversible inhibitor of MAO-A, brofaromine. Clin Neuropharmacol i 3(suppl 2): 196-197.