Original Author: Paul Perry, Ph.D, BCPP
Latest Revisers: Paul Perry, Ph.D, BCPP, Brian C. Lund,
Pharm.D.
Creation Date: 1996
Peer Review Status: Internally Peer Reviewed
TRICYCLIC ANTIDEPRESSANTS
Absorption
The tricyclics (TCAs) are rapidly absorbed in the small intestine when taken orally with peak concentrations usually occurring 2 to 6 hours after the dose. An exception is protriptyline, as peak concentrations may not be reached for 12 hours after an oral dose. In the fasting state, absorption appears complete as suggested by nearly identical urinary recovery of nortriptyline metabolites following oral and intramuscular administration, and equal recovery of radioactivity in the urine following oral and intravenous doses of radiolabeled nortriptyline and imipramine. Concurrent ingestion of food apparently has no effect on bioavailability, peak concentration after oral administration, or time to peak concentration (Alexanderson 1969, DeVane and Jarecke 1992).
Despite the evidence for complete oral absorption, the systemic availability of tricyclics is low. Comparison of area under the concentration-time curve for oral and intravenous doses in the same subjects indicates an availability of drug between 30% and 70% as suggested in Table 1. This phenomenon is termed the "first pass effect" and is explained by the fact that all the absorbed drug in the portal vein must pass through the liver and small intestine where it can be metabolized before reaching the systemic circulation. This is compatible with earlier and higher metabolite concentrations from oral compared to parenteral administration. Differences in metabolite-parent drug ratios may assume importance in dosing as some metabolites formed during the first pass of tricyclics are active as antidepressants and will be discussed below (Alexanderson 1969, DeVane and Jarecke 1992).
Distribution
The tricyclics are highly lipophilic drugs that distribute widely throughout the body. This partially explains why they are so extensively distributed in the body. Postmortem toxicological examinations of brain and myocardial tissue from overdosed patients have revealed microgram per milliliter concentrations of tricyclics compared to the nanogram per milliliter concentrations normally observed in the blood. Thus it is not surprising that the primary target organs affected in overdoses are the heart and the brain. Although obesity has been incompletely studied as a specific factor affecting tricyclic pharmacokinetics, animal data suggest that adipose tissue is not a sink for tricyclic half-lives due to the larger volume of distribution in these patients.
Steady-state volumes of distribution may be as large as 60 L/kg. This parameter explains quite well why in tricyclic overdoses treatment with hemodialysis is an inefficient means of removing the offending tricyclic from the body. Although dialysis procedure can remove up to 75-95% of the tricyclic from the vascular compartment, this particular compartment contains less than 1% of the total amount of the tricyclic in the body.
Protein Binding
Distribution is closely related to protein binding. It was earlier assumed that tricyclics were highly bound to plasma proteins, principally albumin, with little intersubject variation. Recent reports have focused on the importance of binding to erythrocytes and proteins other than albumin in plasma. The binding of imipramine is higher in plasma of hyperlipoproteinemic patients and correlates well to plasma cholesterol and triglyceride concentrations. Alpha-1-acid glycoprotein (AAG) concentrations have been shown to negatively correlate to individual differences in the free fraction of imipramine in healthy volunteers. AGP concentrations may be increased in short and chronic inflammation, alcoholism, malignancy, myocardial infarction as well as hepatic and renal disease. The plasma protein binding of tricyclics may therefore vary widely in patients with these conditions, and has not been systematically studied. Also, tricyclic binding in depressed psychiatric patients has not been thoroughly studied in relation to the concentration of AAG (DeVane and Jarecke 1992).
Recently researchers have voiced the opinion that if the dynamic interaction between free and bound drug in relation to pharmacological activity is not considered, it may be impossible to discern a cause and effect relationship between drug and therapeutic response (Alexanderson 1969). It has also been postulated that much of the variability associated with the use of TCAs may actually be the result of changes in the level of free drug present within the system (Braithwaite 1980). This change in drug distribution would not be depicted in the routine TCA assay.
Preliminary studies done by Borga et al, stated that there was no significant variation in the binding of TCAs to various protein constituents, either intraindividually or interindividually (Borga et al 1969). A more recent finding has shown that this assumption may not be valid. Studies in which imipramine was studied revealed substantial variations in the fraction of free drug ranging from 5-23% in different patients (Glassman et al 1973). Alteration in the equilibrium between bound and free drug could significantly alter the clinical course of patients. However, the utility of measuring free TCA levels is debatable based on the findings of two clinical studies. Breyer-Pfaff et al (1982) found no relationship between free amitriptyline concentrations and therapeutic response. Likewise, Perry et al (1985) was unable to observe a correlation between free nortriptyline concentrations and response but did note that free concentrations above 10 ng/ml suggested a tendency toward nonresponse.
Metabolism and Elimination
The tricyclics are extensively metabolized. Renal clearance accounts for the elimination of less than 5% of the unchanged drug. The most important metabolic pathways are N-demethylation, and hydroxylation, followed by glucuronide coupling. The tricyclics appear to be metabolized by several hepatic enzyme systems, including the microsomal P-450 system. Some evidence exists for biliary excretion and reabsorption of imipramine, i.e., enterohepatic circulation. Clinically, this is significant since activated charcoal's adsorptive properties are obvious for single doses of TCAs for 6 to 9 hours after its administration (Devane and Jarecke 1992).
The metabolism of tricyclics appears to proceed by linear pharmacokinetics in the usual dosage range. Support for this statement comes from studies predicting steady-state concentrations of nortriptyline and imipramine from single doses. A recent study of imipramine metabolism in the perfused liver preparation suggests that saturation of desipramine metabolism can occur in rats, but at concentrations generally not achieved in humans. Evidence for non-linear metabolism in man is limited. Browne et al (1984) described a patient who presented with apparent dose-dependent kinetics for nortriptyline. Twelve-hour steady-state concentrations total plasma nortriptyline at doses of 10 mg qod, 10 mg/d, and 25 mg/d were 38, 86 and 647 ng/ml respectively. Half-lives for the 10 mg qod, 10 mg/d, and 25 mg/d doses were 30, 37, and 64 hours, respectively. In another case report where high doses of desipramine were used to treat a recalcitrant patient, a 14-fold increase in plasma concentration resulted from less than a 4-fold increase in dosage. This report is suggestive of metabolic saturation; however, it is unlikely that 500 mg doses, as was used to treat this patient, will be widely utilized in the absence of plasma concentration monitoring. Nevertheless, saturation of tricyclic metabolic pathways may occur in situations of inadvertent or intentional over-dosage, and long observation periods are recommended for these patients.
Active Metabolites
Pharmacologically active metabolites are important in a consideration of the tricyclics. Many of the demethylated and hydroxylated metabolites have been shown to accumulate to a significant degree and to be present in the CSF of patients. The hydroxylated metabolites of amitriptyline, nortriptyline, imipramine, and desipramine are presumed to have pharmacologic activity. Of these metabolites the data suggest that 10-hydroxynortriptyline and 2-hydroxydesipramine seem to be the most important metabolites. Both are capable of 1) inhibiting the reuptake of norepinephrine; 2) accumulating in the plasma; and 3) being present in the CSF. This finding may be of clinical significance.
Recently, three cases of sudden death due to desipramine were reported (Sudden death in children treated with a tricyclic antidepressant 1990). Up until this time there was only one other similar case reported in the literature (Saraf et al 1974). However, this patient, a 6 year old, 20.5 kg female was given an imipramine dose of 14.7 mg/kg/d as a single night time dose. Imipramine and desipramine plasma concentrations in all four of the children were reported to be therapeutic or subtherapeutic. Two of the children were being treated for ADHD, one for school phobia, and the fourth was undergoing therapy for an undisclosed cause. The children were between six and nine years of age.
Despite the suggestion that the imipramine plasma concentrations were within or below the suggested therapeutic range and therefore of no prognostic value in monitoring these children, animal data suggest the opposite. Jandhyala et al (1977) investigated the cardiovascular effects of several imipramine analogs (imipramine, 2-hydroxyimipramine, desipramine, 2-hydroxydesipramine, 3-chloroimipramine and 3-chloro-8-hydroxy imipramine) in dogs. They found that although all the analogs were capable of reducing cardiac output, the hydroxylated metabolites, especially 2-hydroxyimipramine, appeared to be most toxic to the myocardium. Based on these findings, clinical value of monitoring 2-hydroxyimipramine concentrations as means of estimating imipramine myocardial toxicity in children is currently being studied within the department.
|
|
||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Factors Influencing Plasma Concentration
The metabolism of tricyclics is remarkably variable among patients and is subject to perturbation by social and environmental influences as well as drugs. Factors known to influence plasma concentrations are listed below in Table 2.
Sampling
Absorption and tissue distribution of TCAs may take as long as 5-8 hours. Additionally, it is not known how much time is required for equilibrium to be reached between the receptor site and the TCA in the plasma. However, it is recommended that plasma sampling be carried out at approximately 12 hours after the last dose. Plasma and erythrocyte TCAs levels can vary significantly. Therefore plasma samples must be refrigerated or iced and centrifuged promptly to avoid hemolysis.
The plasma sample should be drawn after steady state, i.e., amount of drug ingested daily equals amount of drug excreted daily, has been reached. Thus, the time to draw the first meaningful sample is dependent upon the half-life of the individual TCA. It must be remembered that it takes five half-lives to reach steady state. The pharmacokinetic parameters of the tricyclics are summarized in Table 1.
Dosage Adjustment - multiple point method
Browne et al (1983) designed a prospective dosing protocol for nortriptyline. The method requires the administration of a 100 mg test dose of nortriptyline to each subject. Blood samples (10 cc) are drawn at 0 hr., 12 hr., 24 hr, and 36 hr. The plasma levels obtained are analyzed to determine the elimination rate constant and half-life. From these data R, and the daily accumulation factor, is computed. The kinetic data produced will then be applied to development of various dosing schedules based on predictions of CSS min TD in each individual patient. The accumulation factor R is calculated by equation 1.
where Ke is the elimination rate (h-1) and T = dosing interval (h) The steady state tricyclic level for a specific dose of drug is calculated:
CSSmin = D (CSSmin TD)/ TD (equation 2)
where CSSmin is the minimum serum nortriptyline level at steady state (ng/ml), D is the dose administered (mg), TD is the test dose (mg), and CSSmin TD, the minimum serum concentration produced by the test dose.
|
TABLE 2. Factors reported to effect cyclic antidepressant plasma concentration |
||
|
Factor or Drug |
|
Mechanism |
|
Acute ethanol ingestion |
|
Hepatic enzyme induction |
|
Barbiturates |
|
Hepatic enzyme induction |
|
Carbamazepine |
|
Hepatic enzyme induction |
|
Chloral Hydrate |
|
Hepatic enzyme induction |
|
Phenytoin |
|
Hepatic enzyme induction |
|
Smoking |
|
Stimulates P-450 liver enzyme system |
|
Trihexyphenidyl |
|
Interferes with intestinal absorption |
|
Acidic urine pH |
|
Reduces renal tubular reabsorption |
|
Aging |
|
Decreased metabolic clearance due to decreased liver perfusion |
|
Benzodiazepines |
|
|
|
L-triiodothyronine |
|
|
|
Methylphenidate |
|
Competitive enzyme inhibition of hydroxylation pathways |
|
Chloramphenicol |
|
Competitive enzyme inhibition |
|
Haloperidol |
|
Competitive enzyme inhibition |
|
Phenothiazines |
|
Mutual hepatic enzyme competition |
|
Cimetidine |
|
Hepatic enzyme inhibition of hydroxylation and demethylation P-450 pathway |
|
Disulfiram |
|
Hepatic enzyme inhibition |
|
Fluoxetine |
|
Hepatic enzyme inhibition |
|
Oral contraceptives |
|
Enzyme inhibition |
|
Alcoholic Liver disease |
|
Decreased metabolic capacity in advanced disease state |
|
Basic urine pH |
|
Increased urinary reabsorption (or) hepatic enzyme inhibition |
|
Smoking |
|
Increases free concentrations of TCAs |
A total of 23 patients were dosed according to the method. The correlation coefficient (r) of the observed nortriptyline plasma levels to the predicted nortriptyline levels for these patients is 0.94 (p < 0.001).
Dosage Adjustment - one point method
Perry et al (1984) compared two prospective pharmacokinetic dosing methods to predict steady state concentrations of nortriptyline. One method required multiple determinations of the nortriptyline plasma concentration to estimate the drug's steady-state concentration. The second method required a single nortriptyline concentration drawn at a fixed time, preferably 36 hours, following a 100 mg nortriptyline test dose. The 36-hour nortriptyline plasma concentrations (NTP 36h) were substituted into the straight-line equation of C = 17.2 + 3.74 (NTP 36h), were C is the average steady-state concentration for a 100 mg/day dose of nortriptyline. No differences were noted between the observed steady- state nortriptyline concentration of 121 ± 19 ng/ml, the 36-hr single-point prediction mean concentration of 121 ± 21 ng/ml or the multiple-point prediction mean concentration of 122 ± 19 ng/ml. The nortriptyline dosing nomogram in Figure 1 can be utilized to prospectively dose patients. The only data required to predict steady-state doses for maintenance doses between 50 and 150 mg/d using the nomogram is a plasma nortriptyline concentration measured 24 hours following the administration of a 100 mg test dose of nortriptyline. The only caveat regarding the use of the nomogram is that a dose of 150 mg/d should not be exceeded because of the potential for the relationship of nortriptyline plasma concentrations and maintenance doses not being linear at higher doses.
Figure 1. Nortriptyline single-point dosing nomogram.
Relationship of TCA plasma level to clinical response
Perry et al (1994) performed a meta-analysis of the TCA blood level literature using receiver operating characteristic (ROC) curves as the primary statistical tool. For nortriptyline a significant curvilinear relationship between therapeutic response and nortriptyline was observed that ranged from 53-148 ng/ml. The sensitivity and specificity for the nortriptyline therapeutic window were 78% and 61%, respectively. The response rate within the therapeutic range was 66% versus 26% outside the therapeutic range. For desipramine a significant linear relationship between therapeutic response and desipramine plasma concentration was observed. The threshold plasma concentration for therapeutic response was 116 ng/ml. The sensitivity and specificity for the desipramine threshold blood level were 81% and 59%, respectively. The response rate above the threshold level was 51% versus 15% below the therapeutic threshold concentration. For imipramine a significant curvilinear relationship between therapeutic response and total imipramine (imipramine + desipramine) plasma concentration was observed between 175-350 ng/ml. The sensitivity and specificity for the total imipramine plasma concentrations were 52% and 74%, respectively. The response rate within the therapeutic range was 67% versus 39% outside the therapeutic range. For amitriptyline a significant curvilinear relationship between therapeutic response and total amitriptyline (amitriptyline + nortriptyline) plasma concentration was observed between 93-140 ng/ml. The sensitivity and specificity for the total imipramine plasma concentrations were 37% and 80%, respectively. The response rate within the therapeutic range was 50% versus 30% outside the therapeutic range. A summary of these data are presented in tables 3 and 4.
|
Table 3. Summary of receiver operating characteristics curve analyses to determine therapeutic levels for TCAs and their sensitivity and specificity |
|||||
|
Drug |
|
|
|
|
|
|
nortriptyline |
|
|
|
|
|
|
desipramine |
|
|
|
|
|
|
imipramine |
|
|
|
|
|
|
amitriptyline |
|
|
|
|
|
|
Table 4. TCA response rates in and out of the suggested therapeutic concentrations. |
||
|
Drug |
|
|
|
nortriptyline |
|
|
|
desipramine |
|
|
|
imipramine |
|
|
|
amitriptyline |
|
|
SSRIs
Pharmacokinetic data for these antidepressants are presented in Table 1 (Devane 1992b).
The elimination half-life of fluoxetine ranges from 26-220 hours with a mean of 84 hours while the half-life of its active metabolite norfluoxetine ranges from 77-235 hours with a mean of 146 hours (DeVane et al 1992a). Therefore, 6 to 7 weeks may be required before steady-state serum concentrations are reached. Likewise the same period of time is required to washout the drug in nonresponding patients. The pharmacokinetic parameters of fluoxetine are not altered in patients with decreased renal function (e.g., Clcr <10 ml/min compared to >90 ml/min). The rate of elimination of fluoxetine is reduced in patients with alcohol-induced cirrhosis (Schenker et al 1988). Patients with liver disease should be treated with lower doses (e.g., 20 mg QOD) to minimize adverse effects. Plasma concentrations of fluoxetine and norfluoxetine did not differ significantly between depressed elderly, younger healthy controls, and younger depressed patients after receiving similar doses. Four fluoxetine studies have investigated the relationship between blood levels and therapeutic response (Kelly et al 1989; Beasley et al 1990b; Montgomery et al 1990; Tyrer et al 1990). Two of the studies were unable to find any relationship (Kelly et al 1989; Beasley et al 1990b). However, one observed a negative correlation between norfluoxetine concentrations and therapeutic response (Montgomery et al 1990) while another noted that patients with high ratios of fluoxetine to norfluoxetine were more likely to respond than patients with low ratios (Tyrer et al 1990).
The elimination half-life of paroxetine ranges between 7-37 hours. Like fluoxetine, paroxetine exhibits non-linear pharmacokinetics, i.e., doubling the dose may triple or quadruple the plasma concentration. Patients with severe renal impairment (Clcr < 30 mg/min) required dose reduction while hepatic dysfunction does not require a dosage adjustment (Grimsley and Jann 1992). Plasma paroxetine concentrations were measured in 94 depressed patients. Steady-state concentrations ranged from 1 to 190 ng/ml with 90% of the responders and nonresponders having plasma concentrations of < 110 ng/ml. No correlations were found between clinical response or adverse effects and the drug concentrations in the plasma (Tasker et al 1989).
The mean elimination half-life of sertraline is 26 hours while its less active (5-10 times) demethylated metabolite's mean half-life is 66 hours (DeVane 1992b). Unlike fluoxetine and paroxetine, sertraline exhibits linear pharmacokinetics, i.e., doubling the dose doubles the plasma concentration. Since less than 1% of the drug is excreted unchanged in the urine dose reduction is not necessary in patients with renal failure. Ingestion with food results in plasma concentrations that are 32% greater than concentrations observed in the fasting state (Grimsley and Jann 1992). No studies are currently available examining the relationship of sertraline blood levels to therapeutic response.
The mean elimination half-life of fluvoxamine in depressed patients is 23 hours. The drug does not have any active metabolites (DeVane 1992b). Most clinical trials have reported a lack of any correlation between plasma fluvoxamine concentration and therapeutic response. However, there is one unreplicated study that found that steady state fluvoxamine plasma concentrations between 160-200 ng/ml were more likely to be associated with a beneficial clinical response (Nathan et al 1990) .
There are a number of drug interactions associated with the ability of the SSRIs to inhibit the liver enzyme cytochrome P4502D6 which is responsible for the oxidation of numerous drugs that include TCAs, neuroleptics (e.g., clozapine or thioridazine), carbamazepine, metoprolol, and flecainide and encainide. Sertraline's inhibition of 2D6 is less than that of fluoxetine or paroxetine (von Moltke et al 1993). There are three cases of increased TCA adverse effects and serum levels reported following the administration of fluoxetine. In one patient, fluoxetine produced an increase in nortriptyline concentrations from 88 to 162 ng/ml (Vaughn 1988). In another patient, fluoxetine 20 mg/d raised the plasma desipramine level from 131 to 212 ng/ml while 40 mg/d raised the level from 131 to 419 ng/ml (Bell and Cole 1988). Although there is no justification for administering two antidepressants together, a clinical situation could occur in which a fluoxetine refractory patient requires being switched to a TCA. The relatively long half-life of fluoxetine and its metabolite may produce an exaggerated but transient rise in the plasma TCA concentration.
Bupropion
Bupropion is metabolized to three metabolites, a threo-amino alcohol, erythro-amino alcohol, and a hydroxylated compound. All appear in the CSF in concentrations greater than the parent drug. Additionally, the hydroxylated metabolite's half-life is twice as long a bupropion which is estimated to range from 10-21 hours. Thus despite the recommended twice daily dosing, the drug probably can be given once daily. It is estimated to possess approximately 50% of the activity of the parent drug. Despite the recommended twice daily dosing, the drug probably can be given once daily. The half-lives of the metabolites but not bupropion itself are increased in patients alcoholic liver disease.
Venlafaxine
Venlafaxine is rapidly absorbed after oral administration and metabolized extensively in the liver. Food has no effect on the absorption of this drug, or on the formation of the metabolite, ODV. The mean venlafaxine half life was 4 hours and 10 hours for ODV. The mean volume of distribution of venlafaxine for all doses was 6 L/kg, with the volume of distribution for ODV 5 L/kg. At doses of venlafaxine between 25 to 75 mg/d, linear kinetics were observed. However, an exponential rather than a linear increase in Cmax and AUC for the 150 mg dose were noted suggesting non-linear kinetics. As suggested by the drug's pKa of 9.4, venlafaxine and ODV are bound to albumins, a-amino glycoproteins, and lipoproteins. However, only 30% of the drug is protein bound. This low level of protein binding, independent of drug concentration, suggests that drug interactions associated with tissue binding are not expected with venlafaxine. Venlafaxine is extensively metabolized in the liver by a saturable process to two less active metabolites (N-desmethyl and N,O,-didesmethyl metabolite) and one active metabolite (O-desmethylvenlafaxine). ODV is 0.2 to 0.33 times as active as the parent compound. Venlafaxine and its metabolites are primarily renally cleared. The urinary excretion profile of 80 or 100 mg of parent drug in five healthy subjects consisted of 1-10% of unchanged venlafaxine, 30% of ODV and 20-60% as the other minor inactive metabolites. Accumulation date suggests the clearance of drug at high doses is lower than that found with single doses (1.9L/hrkg), indicating venlafaxine undergoes saturable metabolism at higher doses (Ellingrod and Perry 1995).
Nefazodone
The recommended starting dose of nefazodone is 100 mg twice daily. This dosage may be increased by 100-200 mg/day at one week intervals based on clinical response and ADRs. The maximum recommended dose of nefazodone is 600 mg/day given twice daily. In the elderly (>65 years old) the recommended starting dose of nefazodone is 100 mg/day (50 mg twice daily). These dosage recommendations should be followed for those patients with severe hepatic impairment since the half-life and AUC of both nefazodone and OH-NEF are approximately doubled. A curvilinear dose-response curve demonstrated that the optimum dose of nefazodone was between 300-500 mg/day. Other studies have suggested that elderly patients (> or = 65 years old) may respond to doses between 100-400 mg/day although this therapeutic range may be the same as recommended for healthy younger patients. In clinical trials the highest dose of nefazodone (600 mg/day) was safe but was associated with lower response rates than doses < 600 mg/day (Ellingrod and Perry in press).
Nefazodone is rapidly absorbed after oral administration with an absolute bioavailability of 20%. Food delays absorption by 20%. Nefazodone is widely distributed in the body tissues including the CNS. It is highly (>99%) protein bound which may cause clinical drug interactions with other medications that are highly protein bound. Nefazodone is extensively metabolized in the liver to three active metabolites, hydroxy-nefazodone (OH-NEF), triazole-dione and m-chlorophenylpiperazine (mCPP). At steady state, nefazodone's mean half-life is dependent on dose, varying from 2 hours at 100 mg/day and 4-5 hours at 600 mg/day. No adjustments in nefazodone dosing is required in patients with renal dysfunction. In patients with severe but not moderate hepatic dysfunction, caution is warranted due to nefazodone's altered kinetics (Ellingrod and Perry in press).
Dosing of MAOIs will in the future be monitored by estimating the degree of MAO enzyme inhibition. Two studies have been conducted that support the validity of this practice. Ravaris et. al (1976) treated 49 outpatients with symptoms of depression with anxiety with either phenelzine 60 mg/d, phenelzine 30 mg/d or placebo for six weeks. Phenelzine 60 mg/d was more effective than either phenelzine 30 mg/d or placebo according to the Hamilton Rating Scale for Depression (HRSD). The median MAO activity decrease in the phenelzine 60 mg/d was 87% (45-98%) versus 61% (31-91%) in the phenelzine 30 mg/d group. Davidson and Turnbull (1984) compared isocarboxazid doses of 30 mg/d and 50 mg/d for four weeks in 35 depressed (DSM III) subdivided as melancholic or nonmelancholic. The HRSD scores decreased significantly for the non-melancholic patients. The 50 mg/d dose resulted in significantly greater enzyme inhibition than did the 30 mg/d dose. Six 30 mg/d patients averaged 63% inhibition at week 4 while seven 50 mg/d patients averaged 89% at week 4. Thus it appears that the minimum effective doses of phenelzine and isocarboxazid are approximately 60 mg/d and 50 mg/d respectively. Both of the preceding authors generally agree that the minimum therapeutic trial of these two MAOIs is four weeks.
Grasso et al (1991) studied five adult patients, 33 to 61 years old who received tranylcypromine (TCP) for management of their affective illness. Each patient was given TCP 20 mg/d for at least one week. Blood levels of TCP were drawn at 0, 14, 20 and 24 hours following the last dose of the drug. Each patient, except for one nonresponding patient who was switched to ECT, was then restarted at 30 mg/d or 40 mg/d for at least one week. Blood samples were drawn at times similar to those for the first set of samples. Platelet MAO activity was measured. From these data, a statistically significant correlation (r2 = 0.60) between % MAO inhibition and TCP dose was found. This relationship was described by equation 3 and presented in figure 2.
% MAOI = 118.8 (log TCP dose) + 98.7 (Eq. 3)
Even though no patient in this study achieved a % MAO inhibition greater than 64.8%, the resulting logarithmic relationship indicates that 80% MAO inhibition can be achieved with a tranylcypromine dose of 0.7 mg/kg/d (p < or = 0.05).
ANTIDEPRESSANTS DRUG INTERACTIONS AND THE P450 ISOENZYMES
The cytochrome P450 oxidative hepatic isoenzymes are the most active family of drug metabolizing enzymes. Most psychotropic drugs with the exception of lithium are metabolized to some degree by one or more of the P450 enzymes. The are four cytochrome (CYP) P450 families. CYP1, CYP2, and CYP3 oxidize xenobiotics while CYP4 hydrolyzes long chain fatty acids. Each cytochrome family has subfamilies designated by an upper case letter. The major drug metabolizing subfamilies are CYP1A, CYP2C, CYP2D, and CYP3A. Individual CYP enzymes are designated by a number, e.g. CYP2D6. Additionally, genetic polymorphs may exist for some of the individual enzymes, e.g. CYP2D6 and CYP2C19 such that patients may be classified as super fast, fast or slow metabolizers (Gonzalez and Idle 1994). Table 5 below lists the various drugs metabolized by CYP450 isoenzymes, 2D6, 1A2, 3A4, and 2C19 as well as the drugs that inhibit these enzymes. Many of these potential drug interactions have not been studied. However, they are predictable from a theoretical basis. Thus although a theoretical drug interaction may be obvious according to Table 5, it may not be described in the chapter either because it is of minor significance or it has not been studied (Wrighton and Stevens 1992, DeVane 1994, Gonzalez and Idle 1994, Gelenberg 1995, Slaughter and Edwards 1995). Concurrently administered drugs metabolized by the same P450 ISOenzymes can demonstrate competitive inhibition of each others metabolism.
Figure 2. Logarithmic dose to % MAO inhibition correlation for tranylcypromine.
|
Table 5. Potential CYP450 Drug Interactions. |
|||
|
P450 Isoenzyme |
Metabolized by |
Inhibited by |
Induced by |
|
CYP2D6 |
antiarrhythmics encainide, flecainide, mexiletine, propafenone |
miscellaneous alfentanil, amiodarone, cimetidine, fentanil, propoxyphene, quinidine, yohimbine |
rifampin |
|
|
antipsychotics clozapine, haloperidol, perphenazine, reduced haloperidol, risperidone, thioridazine, |
|
|
|
|
beta-blockers bufuralol, metoprolol, propranolol, timolol |
|
|
|
|
opiates codeine, dextromethorphan, hydromorphone, methadone, oxycodone, tramadol |
|
|
|
|
TCAs amitriptyline, clomipramine, desipramine, imipramine, norclomipramine, nortriptyline, trimipramine |
|
|
|
|
SSRIs fluoxetine, norfluoxetine, paroxetine, sertraline |
|
|
|
|
miscellaneous antidepressants maprotiline, nefazodone, venlafaxine |
|
|
|
|
miscellaneous debrisoquine, methylenedioxy-methamphetamine (Ecstasy), ondansetron, perhexilene, phenformin, sparteine, tacrine, terfenadine, tropisetron, verapamil |
|
|
|
1A2 |
acetominophen, amitriptyline, antipyrine, caffeine, clomipramine, clozapine, enoxacin, haloperidol, imipramine, ondansetron, phenacetin, propranolol, tacrine, theophylline, R(-)warfarin, verapamil |
cimetidine, erythromycin, fluroquinolones, fluvoxamine, grapefruit juice, methoxsalen |
charbroiled meat, cruciferous vegetables (e.g., broccoli, cabbage, sprouts), omeprazole, tobacco |
|
3A4 |
antiarrhythmics amiodarone, lidocaine, propafenone, quinidine |
astemizole, cimetidine, ciprofloxacin, anhydroerythromycin, fluoxetine, fluvoxamine, grapefruit juice (naringenin), itraconazole, ketoconazole, nefazodone, sertraline, troleandomycin |
barbiturates, carbamazepine, glucocorticoids phenytoin, rifampin, rifabutin |
|
|
antidepressants sertraline, TCAs (amitriptyline, clomipramine, desipramine, imipramine, norclomipramine, nortriptyline, trimipramine), venlafaxine |
|
|
|
|
benzodiazepines alprazolam, diazepam, midazolam, triazolam |
|
|
|
|
calcium channel blockers diltiazem, felodipine, nifedipine, nimodipine, nisoldipine, verapamil |
|
|
|
|
nonsedating antihistamines astemizole, terfenadine |
|
|
|
|
miscellaneous acetaminophen, alfentanil, chloroquin, cocaine, cyclosporine, carbamazepine, cisapride, cyclophosphamide, codeine, cortisol, dapsone, dexamethasone, dextromethorphan, doxorubicin, erythromycin, ethinylestradiol, etoposide, fentanyl, ifosfamide, indinavir, lansoprazole, lidocaine, lomustine, losartin, lovastatin, omeprazole, ondansetron, paclitaxel, progresterone, quinine, ritonavir, simvastatin, troleandomycin, tamoxifen, taxol, testosterone, valproate, vincristine, vinblastine, vinolrebine, warfarin, |
|
|
|
2C8 |
arachidonic acid, paclitaxel, retinoic acid, warfarin |
|
|
|
2C9 |
cyclophosphamide, diclofenac, hexobarbital, ibuprofen, losartan, mefenamic acid, naproxen, phenytoin, piroxicam, tetrahyrocannabinol, tenoxicam, thiotepa, tolbutamide, TCAs, torsemide, S(-)warfarin, |
|
rifampin, barbiturates |
|
2C19 |
clomipramine, diazepam, hexobarbital, imipramine, lansoprazole, S-mephenytoin, mephobarbital, moclobemide, omeprazole, progesterone, proguanil, propranolol, warfarin, |
fluoxetine, fluvoxamine, fluconazole, sertraline |
|
|
2B6 |
cyclophosphamide |
|
|
|
2E1 |
acetominophen, chlorzoxazone, ethanol, enflurane, halothane |
disulfiram |
|
|
2A6 |
coumarin |
|
|
REFERENCES
Alexanderson B, Evans DAP, Sjouist F (1969). Steady state plasma levels of nortriptyline in twins: Influence of genetic factors and drug therapy. Br Med J 4:764-768.
Beasley CM Jr, Bosomworth JC, Wernicke JF (1990). Fluoxetine: relationships among dose, response, adverse events, and plasma concentrations in the treatment of depression. Psychopharmacol Bull 26:18-24.
Bell IR, Cole JO (1988). Fluoxetine induces elevation of desipramine level and exacerbation of geriatric nonpsychotic depression [letter]. J Clin Psychopharmacol 8:447-8.
Borga O, Azarnoff DL, Plymforshell G, et al (1969). Plasma protein binding of tricylic antidepressants in man. Biochem Pharmacol 18:2141-2142.
Braithwaite RA (1980). Plasma protein binding of tricyclic antidepressants. Post Grad Med J 56 (suppl):107-111.
Breyer-Pfaff U, Gaertner HJ, Kreuter F et al (1982). Antidepressive effect and pharmacokinetics of amitriptyline with consideration of unbound drug and 10-hydroxysortriptyline plasma levels. Psychopharmacology 76:240-44.
Browne J, Perry PJ, Alexander B, et al (1983). Pharmacokinetic protocol for predicting steady state nortriptyline plasma levels. J Clin Psychopharmacol 3:351-55.
Browne JL, Perry PJ, Taylor JW, et al (1984). Nortriptyline capacity-limited metabolism: A case report. J Clin Psychopharmacol 4:322-25.
Davidson J, Turnbull C (1984). The importance of dose in isocarboxazid therapy. J Clin Psychiatry 7(sec. 2):49-52.
Devane CL (1992b). Pharmacokinetics of the selective serotonin reuptake inhibitors. J Clin Psychiatry 53:13-20.
DeVane CL (1994). Pharmacokinetics of newer antidepressants. Am J Med 97[suppl 6A]:13S-23S.
DeVane CL, Jarecke CR (1992a). Cyclic antidepressants. In: Evans WE, Schoentag JJ, Jusko WJ, eds. Applied pharmacokinetics. Principles of therapeutic drug monitoring. San Francisco: Applied Therapeutics 33:1-34.
Ellingrod VL, Perry PJ (1995). Venlafaxine: a heterocyclic antidepressant. Am J Hosp Pharm
Gelenberg A (1995). The P450 family. Biol Therap Psychiatry 18:29-31.
Glassman AH, Hurwic MJ, Percel JM (1973). Plasma binding of imipramine and clinical outcome. Am J Psychiatry 130:1367-1369.
Gonzalez FJ, Idle JR (1994). Pharmacogenetic phenotyping and genotyping: Present status and future potential. Clin Pharmacokinet 26:59-70.
Grasso RA, Perry PJ, Sherman AD (1991). Relationship between tranylcypromine dose and MAO inhibition. DICP, Annals of Pharmacotherapy (letter) 25:99-100.
Grimsley SR, Jann MW (1992). Paroxetine, sertraline, and fluvoxamine: new selective serotonin reuptake inhibitors. Clin Pharm 11:930-57.
Jandhyala BS, Steenberg ML, Perel JM, et al (1977). Effects of several tricyclic antidepressants on the hemodynamics and myocardial contractility of the anesthetized dogs. Eur J Pharmacol 42:403-10.
Kelly MW, Perry PJ, Holstad SG, et al (1989). Serum fluoxetine and norfluoxetine concentrations and antidepressant response. Ther Drug Monit 11:165-70.
Montgomery SA, Baldwin D, Shah A, et al (1990). Plasma-level response relationships with fluoxetine and zimelidine. Clin Neuropharmacol 13(1,suppl):S71-5.
Nathan RS, Perel JM, Pollock BG, et al (1990). The role of neuropharmacologic selectivity in antidepressant action: fluvoxamine versus desipramine. J Clin Psychiatry 51:367-72.
Nelson JC, Jatlow PI, Quinlan DM et al (1982). Desipramine plasma concentration and antidepressant response. Arch Gen Psychiatry 39:1419-22.
Perry PJ, Arndt SV, Zeilman C (in press). Tricyclic antidepressant plasma concentrations: An estimate of their sensitivity and specificity as a predictor of response. J Clin Psychopharmcol.
Perry PJ, Browne JL, Alexander B, et al (1984). Two prospective dosing methods for nortriptyline Clin Pharmacokinet 9:555-63.
Perry PJ, Browne JL, Alexander B, et al (1985). Relationship of free nortriptyline levels to therapeutic response. Acta Psychiatr Scand 72:12-05.
Perry PJ, Zeilmann C, Arndt S (1994). Tricyclic antidepressant concentrations in plasma: An estimate of their sensitivity and specificity as a predictor of response. J Clin Psychopharmacol 14:230-40.
Piafsky KM, Borga O (1977). Plasma protein binding of basic drugs II. Importance of 1 acid glycoprotein to interindividual variation. Clin Pharmacol Ther 22:545-549.
Ravaris CL, Nies A, Robinson DS, et al (1976). A multiple-dose controlled study of phenelzine in depression-anxiety states. Arch Gen Psychiatry 33:347-50.
Saraf KR, Klein DF, Gittelman-Klein R, et al (1974). Imipramine side effects in children. Psychopharmacologia 37:265-74.
Schenker S, Bergstrom FR, Wolen RL, et al (1988). Fluoxetine disposition and elimination in cirrhosis. Clin Pharmacol Ther 44:353-9.
Slaughter RL, Edwards DJ (1995). Recent advances: The cytochrome P450 enzymes. Ann Pharmacother 29:619-24.
Sudden death in children treated with a tricyclic antidepressant (1990). Medical Letter 32:53.
Tasker TCG, Kaye CM, Zussman BD, et al (1989). Paroxetine plasma levels: lack of correlation with efficacy or adverse effects. Acta Psychiatr Scan 80(Suppl 350):152-5.
Tyrer SP, Marshall EP, Griffiths, HW (1990). The relationship between response to fluoxetine, plasma drug levels, imipramine binding to platelet membranes and whole-blood 5-HT. Prog Neuropsychopharmacol Biol Psychiatry 14:797-805.
Vaughan DA (1988). Interaction of fluoxetine with tricyclic antidepressants [letter]. Am J Psychiatry 145:1478
von Moltke L, Greenblatt D, Harmatz JS (1993). In vitro inhibition of desipramine oxidation by fluoxetine and norfluoxetine is greater than inhibition by sertraline or desmethylsertaline. Clin Pharmacol Ther 53:196.
Weder HJ, Bickel MH (1970). Interaction of drug with protein I. Binding of tricyclic thymoleptics to human and bovine plasma proteins. J Pharm 59:1505-1507.
Wrighton SA, Stevens JC (1992). The human hepatic cytochromes P450 involved in drug metabolism. Crit Rev Toxicol 22:1-21.