Tricyclic antidepressants for treatment-resistant depression

Mechanism of action

Individual TCAs bind with varying potencies to norepinephrine and serotonin transporters located in nerve terminals in the brain ( ; ). The TCAs also bind at nanomolar concentrations to various types of receptors and interact with a number of ion channels located throughout the central nervous system (CNS) and peripherally ( ). However, the blockade of norepinephrine and serotonin reuptake is thought to be the primary mechanism by which TCAs exert their antidepressive effects. That said, serotonin and norepinephrine transporter blockade occur acutely following exposure to TCAs, resulting in a rapid increase in the synaptic availability of these neurotransmitters. Clearly, these effects do not correlate in time with the 3–6 weeks that are often required for depressed patients who are treated with TCAs to show clinical improvement. The reason for this time lag in therapeutic response with chronic administration of tricyclic and other antidepressants is thought to be due to desensitization and downregulation of selected serotonin and adrenergic receptors in a manner that better approximates the lag time to therapeutic effect of conventional antidepressants (Baker et al., 1992; ). The link between these pharmacological activities and the delayed response to TCAs is theoretical and may not ultimately explain the therapeutic action for these drugs for treating depression.

Secondary- and tertiary-amine compounds

As mentioned earlier, TCAs are characterized structurally by their three, fused-ring structure and the amine on the attached side chains. The side chain nitrogen is either fully substituted (in this case, the hydrogen atoms have been replaced by two methyl groups) and called a tertiary amine; or has only one substituted hydrogen with a methyl group, making it a secondary amine ( Fig. 8.1 ).

The difference in chemical structure between tertiary- and secondary-amine TCAs appears to predict pharmacological properties that are of clinical relevance. For example, as noted earlier, TCAs bind with varying affinities to serotonin and norepinephrine transporters. In general, tertiary-amine TCAs display more potent serotonin transporter blockade than secondary-amine TCAs ( ; ; ), as illustrated in ( Table 8.1 ).

TCAs have additional pharmacodynamic properties that broadly differ between tertiary- and secondary-amine agents that are clinically relevant ( Table 8.2 ). For instance, TCAs bind with varying affinities to central and peripheral receptors that influence the propensity for various adverse effects ( ). This includes binding to histamine H1 receptors, adrenergic α-1 receptors, and muscarinic receptors, where they act as antagonists. Broadly, the H1 antagonist activity of TCAs is thought to contribute to their sedative properties and to their propensity to stimulate appetite, causing body weight gain ( ; ). Antagonist activity at α-1 receptors contributes to the risk of orthostatic hypotension, reflex tachycardia, sedation, sexual dysfunction (in addition to sexual dysfunction related to serotonin reuptake blockade), dizziness, and fall risk in the elderly ( ). Muscarinic cholinergic receptor antagonist effects of TCAs include blurry vision, dry mouth, constipation, urinary retention, increased heart rate (due to vagolytic effects), and cognitive dysfunction or confusion in susceptible patients. The risks for each of these types of side effects with particular TCAs correlate with their receptor binding affinities at these targets ( ; ; ; ; ).

Additional pharmacodynamic properties

TCAs have binding activities at various ion channels that are clinically important. In experimental models, TCAs demonstrate dose-dependent suppression of the delayed rectifier potassium current (IK _r ). Amitriptyline, imipramine, and doxepin have also been shown to block the human ether-a-go-go-related gene (hERG) potassium channel ( ; ). These mechanisms are associated with a slowing of cardiac conduction, and are thought to contribute to an increased risk of potentially fatal ventricular arrhythmias (including torsades de pointes ) in the setting of TCA toxicity ( ; ). A number of pharmacological actions in addition to norepinephrine reuptake blockade may contribute to the beneficial effects of TCAs for neuropathic pain syndromes, including inhibition of sodium channels, calcium channels, and glutamate NMDA receptors ( ).

Pharmacokinetics

In general, TCAs are rapidly absorbed from the gastrointestinal tract following oral administration, with peak concentrations in the blood reached within 2–8 h. In most cases, the presence of food does not affect absorption pharmacokinetics. Once absorbed, TCAs undergo first-pass metabolism and bind extensively to plasma proteins. Nevertheless, the impact of total and free concentrations of most TCAs on therapeutic response is similar enough that it is not necessary to measure the free fractions of TCAs ( ; ).

TCAs are metabolized in the liver by demethylation and hydroxylation ( ). Selected agents, such as imipramine and amitriptyline, are demethylated into active metabolites. Therefore, when interpreting plasma levels for imipramine, the total plasma concentration consists of the concentration of imipramine plus the concentration of desipramine, its monomethylated counterpart. Similarly, the total plasma concentration of amitriptyline consists of the combined concentrations of amitriptyline and nortriptyline. The tertiary-amine TCA, lofepramine (not available in many countries, including the United States and Canada), is structurally related to imipramine and, like imipramine, includes the secondary-amine TCA, desipramine, as a main metabolite ( ).

Several factors govern the metabolism and elimination of TCAs. Therefore, the elimination half-lives of TCAs vary widely, from as little as 6 h with doxepin to as much as 56 h with nortriptyline. As mentioned earlier and as shown in Table 8.2 , TCAs are metabolized by several cytochrome P450 enzymes, principally CYP2D6 and 2C19, with relatively minor contributions from other CYP450 isoenzymes including CYP1A2 and 3A4 ( ). Therefore, varying exposure to TCAs can be expected in people who are pharmacogenetic poor, intermediate, or ultrarapid metabolizers at these CYP450 isoenzymes. Similarly, wide variations in TCA exposure at a given daily dose can be expected in patients who take medications that inhibit or induce these hepatic isoenzymes ( ). Greater-than-expected plasma TCA concentrations also occur in people with varying degrees of hepatic compromise ( ). The importance of these pharmacokinetic factors rests in the possibility that, even at typical daily doses, elevated risk for severe side-effects and toxicity may occur when TCA metabolism is inhibited, or the risk of poor treatment response may increase when TCA metabolism is induced ( , ; ).

Only a small fraction of TCAs are eliminated in their active form in the urine. Therefore, major adjustments in the dose of most TCAs are not necessary in people with less than advanced-stage renal insufficiency. Even so, the renal clearance of TCAs is reduced as a function of normal aging, which is thought to confer a much higher risk of toxicity from TCAs in older adults, as compared to young people ( ). None of the TCAs are considered dialyzable, likely due to their lipophilicity, with tight protein binding, and large volumes of distribution. Drug-specific guidance for dosing TCAs in people in various stages of chronic kidney disease is provided below.

Depressive disorders, broadly defined

The efficacy of TCAs for depression in adults is well-established based on the results of numerous randomized trials, mainly involving people with major depressive disorder. At least three metaanalyses and one systematic review that supported the development of evidence-based guidelines for the treatment of depression in adults have shown clear advantages of TCAs over placebo in short-term trials and in studies lasting up to 12 months ( ; ; ; ). Pooled response rates from these reviews ranged from 45% to 60% for short-term studies (≤ 12 weeks). In a comprehensive review of 65 randomized trials, TCAs and placebo had similar dropout rates for any cause; however, the risk of dropout owing to adverse effects is significantly higher with TCAs than with placebo (8173 patients; RR 4.02, 95% CI 3.46–4.67) ( ).

For adults with depressive disorders, TCAs are as efficacious as SSRIs and venlafaxine ( ). Although it has been suggested that SSRIs are less-effective than TCAs for the acute treatment of major depression, a recent meta-analysis of 89 heterogeneous head-to-head trials (15,435 patients) lasting up to 12 weeks found no significant difference in the relative responder rate between SSRIs and TCAs (relative response rate (RRR) 1.03, 95% CI 0.97–1.09, number needed to treat (NNT) 72.6) ( ). The authors speculated that earlier findings of significantly greater efficacy of TCAs over SSRIs in acutely-depressed patients may have been due to lower placebo response rates from earlier trials, and not necessarily differential antidepressive efficacy. Dropout rates from lack of efficacy are similar between TCAs, SSRIs and SNRIs (< 10% for all antidepressant types); however, rates of treatment discontinuation due to adverse effects were higher with TCAs (20%) than with SSRIs (8%) or SNRIs (6%) ( ).

As reviewed elsewhere in this volume, the results of multiple metaanalyses suggest that irreversible monoamine oxidase inhibitors may have an efficacy advantage over TCAs for depression with atypical features ( ; ; ; ). However, TCAs may be particularly effective for melancholic depression, as reviewed immediately below.

Severe and melancholic depression

Major depression with melancholic features, or melancholic depression, is a depressive subtype characterized by prominent anhedonia, lack of emotional reactivity to pleasant stimuli, and marked psychomotor slowing ( ). A number of other clinical features may be present including a diurnal variation in symptom severity (worse in the morning), terminal insomnia, appetite suppression leading to weight loss, and excessive and ruminative guilt ( ; ). Melancholic depression has been associated with greater symptom severity, higher risk for psychotic features, and more adverse cognitive effects than other depressive subtypes ( ; ). As such, melancholic depression is considered to be an especially severe depressive subtype. The results of several studies suggest that melancholic and severely-depressed but nonmelancholic patients tend to respond more readily to antidepressants with dual effects on serotonergic and noradrenergic neurotransmission, including venlafaxine and selected TCAs ( ), although not all studies support these conclusions ( ; ). Nevertheless, in an early meta-analysis that compared the antidepressive effects of TCAs and SSRIs in depressed patients with high and low depression severity based on a median split of baseline depression rating scale scores, TCAs were more efficacious than SSRIs in the high-severity group and in analyses restricted to inpatient studies ( ). In a more recent meta-analysis of 25 trials comparing antidepressant response rates in people with melancholic (2597 patients) and nonmelancholic depression (5016 patients) who were treated with TCAs or serotonin reuptake inhibitors (including SSRIs and SNRIs), overall antidepressive response rates did not differ significantly between melancholic and nonmelancholic patients ( ). However, response rates were higher with TCAs than with serotonin reuptake inhibitors in both melancholic and nonmelancholic subjects.

Persistent depressive disorder and chronic depression

The efficacy of TCAs for persistent depressive disorder is supported by the results of four systematic reviews and metaanalyses. The first review supported the efficacy of antidepressants in general (TCAs included) for persistent depressive disorder ( ), and the second review documented similar efficacy for TCAs, SSRIs, and MAOIs—but higher dropout rates with TCAs than SSRIs—after 4–12 weeks of treatment ( ). The third review included a network meta-analysis of data from 45 trials totaling 5806 patients with persistent depressive disorder ( ). The analysis was restricted to trials that were at least 8 weeks in duration and had a sample size of at least 200 patients. Compared with placebo, imipramine was associated with a significantly higher response rate (13 trials, OR 4.53, 95% CI 2.80–7.34) and a higher but nonstatistically significant increase in the rate of all-cause dropout, a proxy for treatment acceptability (nine trials, OR 1.26, 95% CI 0.81–1.96). The clinical effects of other TCAs were not evaluated. Finally, a fourth meta-analysis of 34 studies (4769 patients with persistent depressive disorder) focused on comparative rates of adverse effects across several classes of antidepressants ( ). Both the odds of experiencing adverse events and treatment discontinuation due to adverse effects were significantly higher for TCAs than for placebo (for discontinuation, OR 3.98, 95% CI 2.54–6.21).

The clinical effects of TCAs for chronic major depression and major depression + persistent depressive disorder (double depression) have been investigated to a limited extent. Imipramine and sertraline were similarly effective (but sertraline was better-tolerated) in a 12-week randomized trial of 635 patients with chronic major depression or double depression ( ). In the second phase of the study, patients who responded poorly to imipramine ( n = 51) or sertraline ( n = 117) were crossed over to the alternative medication ( ). Response rates were higher for patients who switched from imipramine to sertraline than those who switched from sertraline to imipramine (60% vs 44%). Rates of early treatment discontinuation owing to adverse effects were lower with sertraline.

Long-term maintenance studies of TCAs for chronic depressive states are few. Desipramine was shown to be effective for maintaining symptomatic remission and preventing relapse in a cohort of patients with persistent depressive disorder or double depression for up to 20 weeks ( ). Patients who remained well were then randomized to a 24-month placebo-controlled maintenance trial. Relapse rates were higher with placebo (52%) than desipramine (15%) ( ). When the analysis was restricted to patients with persistent depressive disorder only, relapse rates were 46% with placebo and 0% with desipramine ( ).

Broad estimates of TCA efficacy

Based on a systematic review of randomized trials of reasonably good quality, clinicians can expect a positive treatment response in approximately 17%–49% of depressed patients after switching to TCAs following a poor response to at least one prior therapeutic antidepressant trial ( ). However, the broader portfolio of studies that investigated the clinical effects of TCAs for patients with a history of poor response to other antidepressants indicates an even wider variation in clinical effectiveness ( Table 8.3 ).

TCAs after failing to respond to a single antidepressant trial

Several studies have investigated the antidepressive efficacy of TCAs after failure to respond to a single antidepressant trial. To our knowledge, the first such study was a randomized cross-over trial that initially enrolled 122 patients with major depression ( ). Of these, 44 patients who responded poorly to initial double-blind treatment with imipramine, paroxetine, or placebo were crossed over to the alternative antidepressant (placebo failures were crossed over to paroxetine) and were followed for 6 weeks. Eleven of 15 patients (73%) who originally failed to respond to paroxetine eventually responded to imipramine, while 5 of the 10 patients (50%) who originally failed to respond to imipramine responded to paroxetine. Following cross-over, imipramine was associated with significantly greater reduction in depressive symptoms and clinical global state. Data on adverse effects were not reported.

In a larger 12-week randomized cross-over study, the clinical effects of imipramine and sertraline were compared in 168 patients with chronic depression who had previously failed to respond to a prospective trial of imipramine or sertraline ( ). There were no statistically significant differences in remission rates between patients who were switched to imipramine compared with those who were switched to sertraline (23% vs 32%). Treatment discontinuation owing to adverse effects was higher for patients who were switched to imipramine than those who were switched to sertraline (9% vs 0%).

In a 10-week, open, randomized switch study that enrolled 112 depressed patients with prior nonresponse to venlafaxine ( ), subjects were randomized to imipramine (dose adjusted to achieve a plasma level of 175–300 ng/mL) or mirtazapine augmentation of venlafaxine. Remission rates were significantly higher after switching to imipramine than with adjunctive mirtazapine (71% vs 39%). Completion rates in the study were very high, as only five patients randomized to imipramine and two subjects in the combination therapy group were unable to complete all study visits. Only three patients, in total, dropped out due to adverse effects.

TCAs after failing to respond to multiple antidepressant trials

Surprisingly few published studies have investigated the effects of TCAs in patients with higher levels of treatment resistance than those reviewed above. In an open single-arm trial, 92 patients with treatment-resistant depression with an average of 2.3 failed prior antidepressant trials of adequate design and duration were switched to nortriptyline and were followed for 6 weeks ( ), Blood levels of nortriptyline were drawn at week 2, and the dose was raised if blood levels were < 100 ng/mL. Approximately 40% of patients responded to nortriptyline after 6 weeks, although 35% of subjects dropped out of the study. At the end of follow-up, the mean dose and drug level of nortriptyline was 121.2 mg/day and 101.0 ng/mL, respectively. There was no significant difference between responders and nonresponders with respect to nortriptyline dose or drug level.

In a randomized trial, 235 patients with treatment-resistant major depression were treated with nortriptyline (mean dose 97 mg/day) or mirtazapine (mean dose 42 mg/day) for 14 weeks ( ). Individuals in this study had previously failed to remit following 12 weeks of citalopram monotherapy and a subsequent trial of either citalopram augmentation therapy (with bupropion or buspirone) or monotherapy with sertraline, venlafaxine extended-release, or bupropion sustained-release. Remission rates were low with both nortriptyline (19.8%) and mirtazapine (12.3%). There were no significant between-group differences in tolerability measures or adverse effects. Although dosing by nortriptyline drug level was allowed, it was also not required and was not done for the majority of nortriptyline-treated subjects.

A two-phase randomized, open trial compared the clinical effects of desipramine (minimum 150 mg/day) and citalopram (minimum 40 mg/day) in 189 patients who failed to respond to a previous therapeutic antidepressant trial ( ). After 4 weeks, patients who did not respond to their initially assigned treatment were randomized to the same antidepressant or the alternative drug for 4 additional weeks. In the second phase, continuation of the same antidepressant was associated with higher remission rates than switching. In the switching arms of the study, there were numerically higher rates of positive response after switching from citalopram to desipramine than switching from desipramine to citalopram (50.0% vs 37.5%). Completion rates across all four study arms in the second phase of the study were high (ranging from 78% to 94%); however, the sample sizes in each of the study arms were very small, ranging from just nine patients in the desipramine-to-citalopram switching arm to 23 subjects in the citalopram continuation arm, thus limiting conclusions that can be drawn from the reported differential response rates between groups.

A large retrospective chart review study compared the clinical outcomes of monotherapy with TCAs and monoamine oxidase inhibitors in 94 patients with treatment-resistant unipolar depression ( ). A little over half of the study subjects had received two or more therapeutic trials of TCAs and/or monoamine oxidase inhibitors on at least two separate occasions prior to study entry. Forty-seven of the 147 treatment outcome observations in this study involved TCAs that included desipramine (57%), clomipramine (15%), doxepin (13%), imipramine (11%), and protriptyline (4%). The results of analyses that examined the interaction between antidepressant treatment and the number of prior therapeutic antidepressant trials suggested that monoamine oxidase inhibitors were more effective than TCAs for people who had fewer prior trials; however, the differences in outcome between treatments diminished as the number of prior therapeutic trials increased.

Adverse effects