Introduction

Fiscal constraints in the health care system have led to closer scrutiny of medications with mechanisms of action similar to those of standard reference drugs. Often referred to informally as “me too” medications, these drugs may have similar efficacy to reference drugs but different adverse effects or drug interactions. Such differences may result in a preference for one agent over another in particular situations. Otherwise, the decision about which agent to use is most often based on price or the prescriber's familiarity with a product. This article reviews the literature on the antidepressants duloxetine and venlafaxine, both of which block the reuptake of serotonin (SE) and norepinephrine (NE). Duloxetine has been available in Canada since 2008 but was first marketed in the United States in 2004. Venlafaxine has been available in Canada since 1994.

Antidepressants that affect predominantly 2 neurotransmitter systems, such as duloxetine and venlafaxine, are referred to as dual-acting agents. Dual-acting drugs that block the reuptake of SE and NE are commonly referred to as serotonin–norepinephrine reuptake inhibitors (SNRIs). Tricyclic antidepressants (TCAs), in particular the secondary amines, affect both the SE and NE systems, but they also interact with a variety of other receptors, which predisposes them to cause more adverse effects than agents with greater selectivity. Other dual-acting antidepressants include bupropion, which inhibits the reuptake of NE and dopamine into presynaptic neurons, and mirtazapine, which influences NE and SE transmission via mechanisms different from that of the SNRIs.

In addition to duloxetine and venlafaxine, there are 2 other SNRI-like dual-acting antidepressants, namely milnacipran¹ and sibutramine.² Sibutramine is marketed in Canada and the United States for the treatment of obesity but not for major depressive disorder. Milnacipran is not available in Canada or the United States.

Pharmacology

Duloxetine and venlafaxine are SE and NE reuptake pump inhibitors with different affinities for the SE and NE receptor (Table 1).^3–6 Relative to venlafaxine, duloxetine has 106 times greater potency at the human SE transporter and 331 times greater potency at the human NE transporter.³ The higher affinity of duloxetine for these receptors correlates with the lower doses of this drug required for antidepressant effect, compared to venlafaxine. Both agents have relatively low affinity for the dopamine transporter.

TABLE 1. Inhibition of binding to monoamine transporters in vitro in humans

The SE:NE selectivity ratio of inhibition on the monoamine transporter is 9.4 for duloxetine and 30 for venlafaxine. Therefore, the differential affinity (favouring SE over NE transporters) of venlafaxine is considerably higher than that of duloxetine. Given this preferential affinity of venlafaxine, lower therapeutic doses would result in predominantly serotonergic activity, similar to that of a selective serotonin reuptake inhibitor (SSRI). Conversely, higher doses would be required to appreciably affect the NE transporter, so the drug's dual action would be clinically more relevant at higher therapeutic dosages.^7,8 Duloxetine has been touted as a balanced dual inhibitor of SE and NE reuptake.^4,9–11 However, one study showed that duloxetine at up to 60 mg/day did not alter NE uptake in healthy human subjects, whereas the same dose blocked the reuptake of SE.¹² A subsequent study suggesting that the drug does inhibit NE uptake used doses of 80–120 mg/day.¹³ An unresolved issue in both of these studies is whether the pressor response to tyramine was in fact a valid indicator of NE reuptake inhibition. Another report suggested that duloxetine's inhibition of NE reuptake may not become evident at doses below 60 mg/day.¹⁴ Although more uniform dual action across the dosage range might be expected with duloxetine, this theory requires further investigation.

Neither duloxetine nor venlafaxine has any significant affinity for postsynaptic adrenergic, muscarinic, histaminic-1, dopamine, SE, opiate, gamma-aminobutyric acid or substance P receptors.^3,4 Minimal activity at these receptor sites suggests that duloxetine and venlafaxine have a low propensity to cause adverse effects associated with antagonism at these receptors.

Summary of efficacy results from randomized, double-blind studies of duloxetine in depression

A total of 8 randomized, double-blind studies of duloxetine in depression^15–22 were identified in the literature (Table 2).

TABLE 2. Randomized placebo-controlled trials of duloxetine in adults with major depression with the HAMD17 as primary efficacy measure*†

One randomized, double-blind, parallel group study compared duloxetine 60 mg/day and extended-release venlafaxine 150 mg/day for 6 weeks, followed by an additional 6 weeks during which the dose could be increased (to 120 mg/day for duloxetine and 225 mg/day for venlafaxine).¹⁵ There were no significant differences between duloxetine and venlafaxine in terms of remission rates at the end point (6 or 12 weeks), as indicated by total scores on the 17-item Hamilton Depression rating scale (HAMD₁₇). However, the proportion of patients who completed the 12-week study was significantly higher in the venlafaxine group than the duloxetine group (74.5% vs 64.8%). The higher dropout rate with duloxetine was due to adverse effects, in particular, nausea and dizziness, which tended to occur earlier in the treatment period.

In 2 of the 6 studies that compared duloxetine 60–120 mg/day with placebo and/or paroxetine or fluoxetine, the rate of remission in depression was significantly higher among patients who received duloxetine than among those who received placebo, for analyses in which the last observation was carried forward.^17,19 In all of the studies that used duloxetine at a dose of 60 mg/day or more, reductions in HAMD₁₇ scores were greater with duloxetine than with placebo^16–21; however, 2 studies^20,21 did not report remission rates. No significant differences in efficacy were found between duloxetine and paroxetine or fluoxetine.

However, it should be noted that duloxetine at 40 mg/day was not significantly more efficacious than placebo.¹⁹ In one of the studies¹⁶ that used duloxetine at 120 mg/day, there was a significant difference favouring duloxetine over placebo in terms of rate of remission when the mixed-effects model for repeated measures was used, but not when the last observation was carried forward. In addition, one-third of the patients discontinued treatment before the end of the study. Most of the patients who stopped duloxetine did so because of adverse effects, whereas patients taking placebo dropped out because of lack of efficacy.

Three studies compared duloxetine and escitalopram.^22–24 In 2 of these studies, which used the Montgomery-Asberg Depression Rating Scale as the primary indicator of efficacy, escitalopram was significantly better than duloxetine in remission rates when the last observation was carried forward.^23,24 Both studies reported higher discontinuation rates with duloxetine than with escitalopram, but the difference was significant for only one of the studies (31% vs 13%²³ and 25% vs 22%²⁴). In the third study, which used the HAMD₁₇ as the primary measure of efficacy, there was no significant difference in remission rates.²² All of the studies used escitalopram at 10–20 mg/day and duloxetine at 60–120 mg/day.

Pharmacokinetics

The pharmacokinetic characteristics of duloxetine and venlafaxine vary in some respects (Table 3).^25–30

TABLE 3. Pharmacokinetic summary of duloxetine and venlafaxine

Duloxetine

Duloxetine is acid-labile and is therefore formulated with an enteric coating that begins to dissolve at pH 5.5.²⁵ Despite this coating, the drug is well absorbed, with peak plasma concentrations occurring about 6 hours after administration. Administration of duloxetine with food delays absorption by 2 hours and decreases the area under the curve (AUC) for concentration vs time by 10%. If this drug is taken in the evening rather than in the morning, absorption is delayed by about 3 hours and clearance is increased by about 33%.²⁵ Duloxetine is metabolized by cytochrome P450 enzymes (specifically CYP2D6 and CYP1A2) to numerous metabolites. Two of these metabolites, 5-hydroxy, 6-methoxy duloxetine and the 4-hydroxylated metabolite, have been identified as being active, but they are much less potent than the parent compound.²⁵ A detailed account of the various drugs involved in the cytochrome P450 systems is available elsewhere³¹ and Table 4 lists examples of common drug interactions with duloxetine and venlafaxine.

TABLE 4. Summary of drug interactions

Duloxetine is a moderate inhibitor of CYP2D6 but only a weak inhibitor of CYP1A2 in vivo.²⁶ Most (70%) of the metabolites of this drug are excreted in the urine, with about 20% being excreted in the feces.²⁵ The elimination half-life averages 12 hours (range 8–17 hours).²⁷

Venlafaxine

Venlafaxine is rapidly absorbed after oral administration and is unlike duloxetine in that administration with meals does not significantly delay absorption. Venlafaxine is widely distributed throughout the tissues, and has much lower protein binding in plasma than duloxetine (27% vs 96%). Venlafaxine is extensively metabolized by CYP2D6 to an active metabolite, O-desmethylvenlafaxine (ODV).²⁸ Therefore, plasma concentrations of venlafaxine may be increased by inhibitors of the CYP2D6 isoenzyme. The primary route of elimination for the major metabolites of venlafaxine is the kidneys. Venlafaxine and its active metabolite have elimination half-lives of about 4 and 11 hours, respectively. Steady-state concentrations of venlafaxine and ODV in plasma are attained within 3 days. Venlafaxine is a weak inhibitor of the CYP2D6 isoenzyme and does not contribute significantly to elevations of plasma concentrations of drugs metabolized by this isoenzyme.

Safety and adverse effects

The adverse effects of duloxetine and venlafaxine are similar and are associated with increased SE and NE concentrations in the synapse. To date, only one head-to-head study comparing venlafaxine and duloxetine has been conducted.¹⁵ Other studies have compared each of these drugs with placebo. In preparing this summary, we analyzed data on adverse effects from both types of study; for the placebo-controlled trials, we report only the adverse effects that occurred significantly more often with the drug than with placebo.

Gastrointestinal effects

The most frequent adverse effect is nausea, which is related to the effects of these drugs on SE levels. The incidence of nausea with either agent ranges from 20% to 44%, depending on the dose and dosage form. The extended-release formulation of venlafaxine has a lower incidence of nausea than the regular tablet.^32,33 Nausea occurred significantly more often with duloxetine than venlafaxine in the first 6 weeks of the head-to-head study (43.6% vs 35%).¹⁵ Venlafaxine was started at 75 mg/day and the dose was increased to 150 mg/day in the third week, which may have accounted for the lower reported frequency of adverse effects with this drug. Nausea, which typically declines with longer-term use, can usually be minimized by taking the medication with food.^32,34

Effects on sexual function

An indirect effect of increasing SE levels at the serotonin-2 receptor site is the occurrence of sexual adverse effects, which are associated with both duloxetine and venlafaxine, as well as with most other antidepressants that increase serotonergic transmission (except for mirtazapine).^35,36 Sexual adverse effects include decreased libido, anorgasmia and erectile dysfunction. In a study conducted in primary care clinics, the mean prevalence of sexual adverse effects among patients with depression who were receiving antidepressants was 36.9%.³⁷ The rate of sexual dysfunction is difficult to compare among different agents because some studies have relied on spontaneous reporting, whereas others have used rating scales such as the Arizona Sexual Experience Scale (ASEX) to identify this type of problem.³⁸ The incidence is expected to be lower in studies that rely on spontaneous reports. Another confounding factor is the greater rate of problems with sexual function among people with depression than among normal controls, even without antidepressant therapy. For example, loss of libido has been reported for 25%–75% of patients with unipolar depression.³⁹ Problems with arousal are also common among both men and women with depression and about 25% may experience difficulty with erection or lubrication. However, once symptoms of depression improve after treatment with antidepressants, most patients expect their sexual function to return to baseline.

In the study conducted in primary care clinics mentioned above, which enrolled more than 6000 patients, the prevalence of sexual dysfunction was similar for venlafaxine and SSRIs (as measured by the Changes in Sexual Functioning Questionnaire), ranging from 36% to 43%.³⁷

Pooled data from 6 studies of duloxetine revealed that for patients without sexual dysfunction at baseline, the incidence of sexual dysfunction during the acute phase of treatment was significantly higher among patients receiving duloxetine than among those receiving placebo (46.4% vs 28.8%).³² In an 8-month, randomized double-blind, placebo-controlled study of duloxetine and escitalopram, 114 of the 634 patients had no sexual dysfunction at baseline (based on the 14-item Changes in Sexual Functioning Questionnaire). Of these, 33% of the patients receiving duloxetine, 43.6% of those receiving escitalopram and 25% of those receiving placebo reported sexual dysfunction at the end of 8 months of treatment.⁴⁰

In summary, the incidence of sexual dysfunction is similar for these 2 drugs: from 33% to 46% with duloxetine therapy and from 36% to 43% with venlafaxine (based on pooled study data).

Anticholinergic effects

Drugs that antagonize the muscarinic receptor cause anticholinergic adverse effects, such as dry mouth, constipation, blurred vision and urinary retention. Duloxetine has a higher affinity for this type of receptor than venlafaxine, which accounts for the somewhat higher incidence of dry mouth and constipation with duloxetine (15% and 11%, respectively) than with venlafaxine (12% and 8%, respectively). Both agents caused blurred vision in about 4% of patients and neither caused urinary incontinence significantly greater than that observed with placebo.^32,33 Drugs with anticholinergic side effects can cause mydriasis, which is a hazard for patients with increased intraocular pressure and those at risk for angle-closure glaucoma. Because mydriasis may occur in association with both duloxetine and venlafaxine, these drugs should be used with caution in patients with controlled angle-closure glaucoma and should be avoided altogether in those with uncontrolled angle-closure glaucoma.

Neurological effects

Both duloxetine and venlafaxine are reported to cause insomnia more frequently than placebo, probably because of the increase in NE levels. The percentage of patients who experience insomnia increases with higher doses. In a sleep study of 10 patients with depression who were taking duloxetine, stage 3 sleep and rapid eye movement (REM) latency increased significantly, whereas REM sleep decreased significantly.⁴¹ About 10% of patients taking duloxetine (40–120 mg/day) and 18%–24% of those taking venlafaxine (75–375 mg/day) have reported problems with insomnia. Insomnia is relatively common among patients with depression and the rates of insomnia among those receiving placebo ranged from 6% to 11% in these studies.

The incidence of somnolence with duloxetine and extended-release venlafaxine has been reported as 7% and 17%, respectively.^32,33 The authors of a case series described excessive and disabling daytime yawning induced by duloxetine, although it is unclear whether this effect is directly related to somnolence.⁴² Dizziness is another common adverse effect with both agents and has been reported in 9% of patients receiving duloxetine and 19% of those receiving extended-release venlafaxine.^32,33 Blockade of alpha-1 receptors is a common drug-related cause of postural hypotension, which can in turn lead to dizziness and syncope. Pooled data on the safety and tolerability of duloxetine from 8 studies did not implicate a cardiovascular mechanism for dizziness.³² Because both duloxetine and venlafaxine have low affinities for alpha-1 receptors, the cause of dizziness may relate less to postural changes and more to changes of a neurological nature.

Duloxetine has not been directly evaluated in patients with convulsive disorders. According to the manufacturer's data for duloxetine, there has been one report of seizure in clinical trials involving subjects with major depressive disorder (n = 1139).⁴³ The same database included 2 reported cases of seizure in trials involving patients with neuropathic pain (n = 800). In this context, it should be noted that people with convulsive disorders were excluded from premarketing trials. There were no reports of seizures in premarketing trials involving extended-release venlafaxine (n = 2363), but there has been one reported case among individuals undergoing investigation for panic disorders (n = 1001).³³ While these data for venlafaxine show a low incidence of seizures, it should be noted that patients with a history of convulsive disorders were excluded from most of these trials. There have been some published reports of seizures resulting from venlafaxine overdose^44,45 and one report of seizures associated with therapeutic doses of venlafaxine.⁴⁶ Rapid dose titration, abrupt discontinuation, drug interactions leading to supratherapeutic levels of the SNRI and the potential of these drugs to cause hyponatremia may all contribute to SNRI-induced seizures. For both duloxetine and venlafaxine, use of the drug in patients with a history of seizures merits caution.

Cardiovascular effects

Because both duloxetine and venlafaxine increase NE levels, changes in hemodynamic function are to be expected. Duloxetine therapy is associated with a small increase in heart rate and blood pressure, but (in contrast to the situation for venlafaxine) the cardiovascular effects of duloxetine are not consistently related to dose and few patients experience clinically relevant elevation of blood pressure. An analysis of data from 8 double-blind, placebo-controlled trials involving 1139 patients taking duloxetine for 8–9 weeks found significant increases in heart rate and systolic blood pressure for duloxetine relative to placebo (1.6 vs 0.6 beats/min and 1.0 vs −1.2 mmHg, respectively); the difference in diastolic blood pressure was not significant (1.1 vs 0.3 mmHg).⁴⁷ Also not significant were differences between duloxetine and placebo in the incidence of sustained (for at least 3 consecutive visits) elevation in systolic and diastolic blood pressure. Similar results were found in a randomized double-blind placebo-controlled study involving 117 healthy women with normal cardiac function, 17–74 years of age, in whom dosages were increased, over a period of 16 days, from 60 mg twice daily to 200 mg twice daily. Changes in orthostatic blood pressure and pulse rate during dosing at 120 mg twice daily reached plateau values after 3–4 days and were generally not associated with subjectively reported orthostatic-related adverse events. Short-term dosing of up to 200 mg twice a day was not well tolerated but was not associated with severe, clinically important adverse events.⁴⁸ A recent case report detailed the worsening of symptoms of heart failure in a patient with previously stable cardiac function who was given, on separate occasions, prescriptions for duloxetine 60 mg daily and venlafaxine 75 mg twice daily.⁴⁹ In each instance, discontinuation of the drug led to resolution of the symptoms.

Venlafaxine at doses greater than 200 mg/day led to clinically significant increases in diastolic blood pressure (from 15 mmHg to ≥ 105 mmHg relative to baseline) in 5.5% of patients.^33,50 The incidence of sustained hypertension in patients taking venlafaxine at doses above 300 mg/day has been reported as 13%.³³ In the head-to-head study, significantly more venlafaxine-treated patients than duloxetine-treated patients experienced elevation of systolic blood pressure during the fixed-dosing period.¹⁵

With regard to electrocardiographic changes, QTc, PR and QRS intervals were not significantly altered, either clinically or statistically, with duloxetine 40–80 mg/day. However, at doses of 120 mg/day, duloxetine was associated with significantly greater decreases in PR and QRS intervals than placebo.⁴⁷ Zhang et al.,⁴⁸ however, found no relationship between change in QTc and plasma concentrations of duloxetine or its metabolites, even though average duloxetine concentrations were more than 5 times those achieved at therapeutic doses. Clinically significant conduction abnormalities and arrhythmias have not been reported with venlafaxine doses within the recommended range.⁵⁰

Wernicke et al.⁵¹ conducted a comprehensive review of the findings from 42 placebo-controlled studies of duloxetine to generate a cardiovascular safety profile.

Hepatic effects

The use of duloxetine has been associated with a risk of severe hepatic injury.⁵² Cases of hepatitis accompanied by abdominal pain, hepatomegaly and serum transaminase concentrations more than 20 times the upper limit of normal, with or without jaundice, have been reported during postmarketing surveillance. Elevation of serum transaminase concentrations has in some cases necessitated the discontinuation of duloxetine. Laboratory findings suggestive of severe hepatic injury, with evidence of cholestasis, were reported in 3 patients who received duloxetine in clinical studies.⁵³

Postmarketing reports have also highlighted elevated concentrations of serum transaminases (such as aspartate aminotransferase and alanine aminotransferase), bilirubin and alkaline phosphatase (ALP) in duloxetine-treated patients with chronic hepatic disease or cirrhosis.⁵³ The elevation of transaminases and bilirubin in the absence of obstruction is regarded as a potential indicator of hepatic injury. Conversely, elevated ALP is often a more sensitive indicator of obstruction. ALP levels may increase 4-fold or more because of biliary obstruction and may rise by a factor of 3 in various liver disorders, including hepatitis, cirrhosis and space-occupying lesions.⁵⁴ Three patients treated with duloxetine were reported to have concurrent elevation of transaminase, bilirubin and ALP, suggesting an obstructive hepatobiliary process.⁵⁵

Several case reports have cited liver toxicity in conjunction with venlafaxine therapy.^56–59 The hepatotoxic effects of venlafaxine (and of nefazodone, which has now been withdrawn from the Canadian market) have been described as an idiosyncratic reaction leading to hepatocellular damage.⁶⁰ A safety and tolerability analysis showed that 0.4% of venlafaxine-treated patients had elevation of aspartate aminotransferase to levels 3 times the upper normal limit (0.2% for placebo and 0.9% for active control groups; p-value not specified).⁶¹ Data on file for 9 patients with hepatic cirrhosis showed that the elimination half-lives for venlafaxine and ODV were prolonged by 30% and 60%, respectively, relative to normal subjects.³³ This increase in half-life correlated with a significant decrease in clearance of venlafaxine and ODV relative to patients without hepatic cirrhosis.

Dermatological effects

Although increased sweating is generally not troublesome, this effect may be seen with both venlafaxine and duloxetine. The reported incidence is 6% for duloxetine¹⁸ and from 7% to 19%, depending on the dose, with venlafaxine.³³ One case report described the onset of facial flushing 1–2 hours after placing half the contents of a 20 mg capsule of duloxetine on the tongue; this effect continued over 2 weeks.⁶² The flushing did not spread beyond the face, was not accompanied by itching and resolved 1 week after the drug was discontinued. Another case report described a generalized pruritic rash that occurred within 2 days in a female patient newly treated with duloxetine 60 mg/day.⁶³ Upon rechallenge with a 30 mg dose, the rash recurred within 2 hours. The symptoms resolved in 10 days with discontinuation of the duloxetine and treatment with a corticosteroid and an antihistamine.

Endocrine effects

Although SNRIs are generally not regarded as causing major changes in endocrine and reproductive function, cases of elevated prolactin resulting from both duloxetine and venlafaxine have been reported.^64,65 Both of these drugs, as well as the SSRIs, have been associated with the syndrome of inappropriate antidiuretic hormone secretion (SIADH), resulting in hyponatremia and, in some instances, seizures.^66,67 Older adults appear more likely to experience SIADH and monitoring of electrolytes is therefore recommended for this population. The symptoms of SIADH include confusion, syncope, fatigue, hallucinations, agitation, convulsions, delirium and ataxia.⁶⁸ Although the exact mechanism of antidepressant-induced SIADH remains unclear, serotonin-mediated effects on 5- HT₂ and 5-HT_1C receptors have been shown to release an antidiuretic hormone.^69,70 In addition, other factors, such as NE, acetylcholine and endorphin, may affect secretion of this hormone.^71,72

Metabolic effects

The metabolic parameters for duloxetine were not closely examined in the trials for major depressive disorder. Therefore, Hardy et al.⁷³ undertook a pooled analysis of data on diabetic peripheral neuropathic pain to better understand the effects of this drug on weight, glycemic control and lipid profile. In that study, patients with diabetic peripheral neuropathic pain who were treated with duloxetine had small increases in fasting plasma glucose in both short-term and long-term trials (0.50 and 0.67 mmol/L, respectively); however, the results were statistically significant only for long-term treatment. Changes in glycated hemoglobin (HbA_1C) were not statistically significant in short-term trials (12 weeks) but increased by 0.52% (p < 0.001 vs routine care) in long-term trials that involved a 52-week follow-up. Mean weight loss of 1 kg was observed for short-term duloxetine treatment (p < 0.001 vs placebo), but weight changes were not significantly different for long-term treatment. Changes in lipid profile with either short-term or long-term treatment were considered marginal and of uncertain significance. It should be noted, though, that the patients in this analysis did not reflect the general population and the trials were not designed to evaluate changes in metabolic parameters. In studies with populations not limited to people with diabetes, the effects of duloxetine on weight were variable, with one study²¹ showing no difference in weight change (relative to placebo) with doses of 60 mg/day and other studies^74,75 showing weight increases of 7% or more with doses of 80 or 120 mg/day. The mechanisms behind the effect of duloxetine on glucose homeostasis are not fully understood. Noradrenergic effects can promote hyperglycemia by inhibiting insulin secretion, stimulating gluconeogenesis and decreasing insulin sensitivity, and all of these factors are thought to be contributory.^76,77

Although venlafaxine therapy has been associated with weight loss, the extent of the change in weight can be variable from one patient to another. A pooled analysis found that weight loss was transient for some venlafaxine-treated patients and that long-term treatment (>6 months) was associated with a weight increase of 1–3 kg from baseline.⁶¹ A subsequent study comparing venlafaxine and fluoxetine showed no significant weight gain among subjects treated with either drug.⁷⁸ Placebo-controlled trials lasting 3 months or more showed clinically relevant increases in cholesterol for 5.3% of venlafaxine-treated patients but no such increases in the placebo group.³³ In a shorter 10-week trial, venlafaxine treatment was associated with an increase from baseline in total cholesterol (by 0.18 mmol/L, p < 0.01) and low-density lipoprotein cholesterol (by 0.12 mmol/L, p < 0.05), whereas the placebo group had no statistically significant changes.⁷⁹ Similarly, a 12-week trial showed mean increases in cholesterol (p ≤ 0.01) and lipids (p ≤ 0.05) among venlafaxine-treated patients and no increases in the placebo-treated patients.⁸⁰ Only limited data are available about the effects of venlafaxine on plasma glucose. A small study showed that a single dose of venlafaxine produced a dose-dependent increase in plasma cortisol after 2–3 hours.⁸¹ However, it was unclear whether this finding was correlated with clinically significant changes in glucose homeostasis. Given the potential for venlafaxine to increase blood pressure and cholesterol, monitoring of these parameters is especially appropriate for venlafaxine-treated patients with pre-existing cardiac conditions.

Suicidality

All antidepressants carry a warning of increased risk for behavioural changes, including an increased risk of suicidality, if administered to patients under 18 years of age. It is noteworthy that the development of suicidal ideation in adult patients treated with duloxetine was reported in a case series.⁸² Given the uncertain risks, patients treated with either of these 2 SNRIs should be monitored for suicidal behaviour or ideation, particularly during initial use and after dosage increases.

Toxicity

Venlafaxine overdose has been associated with sedation, sinus tachycardia, seizures, hypotension, hypertension, diaphoresis, hyponatremia and serotonin syndrome.^29,83 Disturbances in cardiac rhythm, including QTc prolongation leading to atrial fibrillation, have been reported.⁵⁰ Deaths due to overdose with venlafaxine alone or in combination with other agents, often associated with serotonin syndrome, have also been reported.^84,85 In 2008, Health Canada issued an advisory regarding overdose with venlafaxine.⁸⁶ This warning was based on retrospective studies reporting that the risk of death associated with venlafaxine overdose may be greater than that associated with SSRI overdose, but less than that associated with overdose with TCAs.^87–90 Although the reasons for these observations are not entirely clear, dual-acting anti-depressants do increase both SE and NE and the excess NE may lead to elevated blood pressure, arrhythmias and seizures.

Less information is available about the toxicity of duloxetine, simply because this drug has not been available for as long as venlafaxine. However, 4 cases of nonfatal acute ingestion of up to 1400 mg of duloxetine have been reported. In all 3 cases, the patients recovered without sequelae.²⁷ At this time, it is difficult to compare the toxicity profiles of these agents, given the discrepancy data available. It would be appropriate to exercise similar precautions for duloxetine as those for venlafaxine, in terms of safety concerns in overdose, since a class effect cannot be ruled out.

Drug interactions

The general principles of drug interactions involving duloxetine and venlafaxine are presented in Table 4.

Serotonergic effect

Like any psychotropic medication, duloxetine may interact with other centrally active drugs affecting SE, NE and dopamine. Given the similar mechanism of action and the potential for additive adverse effects, such as elevated blood pressure and tachycardia, duloxetine should not be co-administered with other SNRIs. Co-administration of duloxetine or venlafaxine with serotonergic agents such as SSRIs, triptans, lithium and tramadol will augment serotonin transmission, which may be synergistic but may also result in excessive serotonin, possibly leading to serotonin syndrome. Therefore, such regimens, although tolerated by many individuals, necessitate careful monitoring. Of greater concern is the combination of dual-acting agents with monoamine oxidase inhibitors (MAOIs). This type of regimen has the potential to cause both serotonin syndrome and hypertensive crisis. For this reason, neither duloxetine nor venlafaxine should be used concurrently with MAOIs.

Protein binding

Unlike venlafaxine, duloxetine is highly bound to proteins in human plasma, primarily to albumin and alpha-1-acid glycoprotein. If duloxetine is given in conjunction with another highly protein-bound drug, the 2 drugs may compete with each other for binding sites and one drug may displace the other from binding sites. This interaction would lead to higher levels of the unbound drug, causing both enhancement of the pharmacological effect and potential toxic effects. Although the body has compensatory mechanisms to buffer against such events, one case report cited such an interaction involving duloxetine and warfarin.⁹¹ In another case, which involved the synthetic coumarin anticoagulant acenocoumarol, the international normalized ratio increased significantly after a single 60 mg dose of duloxetine but returned to baseline 3 weeks after discontinuation of the drug.⁹² Such interactions are not expected with venlafaxine because of its low protein-binding profile.

Cardiovascular risks

Because venlafaxine causes dose-dependent increases in blood pressure, this drug should be used with caution in conjunction with other drugs that elevate blood pressure.⁵⁰ A meta-analysis of original data from over 3000 patients with depression revealed that the effect on supine diastolic blood pressure is clinically significant at high dosages.⁹³ Patients with hypertension who are treated with venlafaxine should therefore be monitored closely for changes in blood pressure during up-titration of venlafaxine dosage. Duloxetine appears to be the preferred SNRI for patients with hypertension. An analysis of the cardiovascular safety profile of duloxetine, based on the results of 42 trials, concluded that although this drug caused small increases in blood pressure, few patients experienced clinically relevant hypertension and there was no evidence suggesting that duloxetine-treated patients with pre-existing hypertension were more prone to persistent elevation in blood pressure than those receiving placebo.⁵¹ Unlike venlafaxine, which has not been reported to cause abnormalities in cardiac conduction, duloxetine was associated with decreases in PR and QRS intervals at a dose of 120 mg/day.⁴⁷ Therefore, caution and more vigilant monitoring are essential when higher doses of duloxetine are used with other drugs that may alter the electrocardiographic profile, such as certain antipsychotics (e.g., pimozide, ziprasidone).

Hepatotoxicity

In one study that investigated the effects of duloxetine on mental and motor functions that are also affected by alcohol, duloxetine did not add to impairment in these domains.⁹⁴ However, the use of duloxetine by patients who regularly consume large amounts of alcohol is not recommended because of additive hepatotoxic effects. Similarly, caution is required in the administration of duloxetine to patients who are receiving long-term, high-dose, acetaminophen therapy, again because of the risk of additive hepatotoxicity. Hepatotoxicity has also been described with venlafaxine and similar precautions may be warranted with this drug.

Metabolic interactions

Duloxetine is extensively metabolized by cytochrome P450 isoenzymes CYP1A2 and CYP2D6 before clearance, whereas venlafaxine is metabolized by CYP2D6 and, to a lesser degree, CYP3A4 (Table 4). Pharmacokinetic interactions are possible for each of these drugs when co-administered with inhibitors and inducers of these enzymes. For example, the bioavailability of duloxetine is one-third lower in smokers than nonsmokers because of induction of CYP1A2 metabolism. When duloxetine is used concurrently with fluvoxamine, a potent CYP1A2 inhibitor, there are substantial changes in the pharmacokinetic profile of duloxetine: a 3-fold increase in half-life, a 5-fold increase in bioavailability and a 2.5-fold increase in peak plasma concentration (C_max).⁹⁴

Venlafaxine does not appear to inhibit cytochrome P450 isoenzymes to any significant degree. In contrast, duloxetine, in addition to being a substrate for metabolism by CYP1A2 and CYP2D6, is also an inhibitor of both of these enzymes. Inhibition of CYP2D6 by duloxetine is dose dependent, yet inhibition of this enzyme by duloxetine is substantially less than occurs with fluoxetine or paroxetine and is similar to that observed with escitalopram and citalopram.^95,96 As a weak inhibitor of CYP1A2, duloxetine does not affect the kinetic profile of theophylline, a prototypical substrate of CYP1A2, even at doses of 120 mg/day.⁹⁴ Despite moderate inhibition of CYP2D6 and mild inhibitory effects on CYP1A2, the product monograph for duloxetine advises caution when co-administering duloxetine with TCAs. This is perhaps because the plasma levels of a TCA may be increased in the presence of duloxetine. Furthermore, TCAs have inherent risks for adverse effects on cardiac function. A rational approach would be to monitor carefully and adjust the dosage appropriately when duloxetine is co-prescribed with drugs that are metabolized by or that are inducers or substrates of CYP1A2 and CYP26D. A similar approach can be taken with venlafaxine when used with inducers and inhibitors of CYP2D6.

Dosage, administration and cost

Please refer to Table 5 for the costs of duloxetine and venlafaxine dosages.

TABLE 5. Unit cost of venlafaxine and duloxetine

Duloxetine

In Canada, duloxetine (Cymbalta) is currently available as extended-release capsules (30 or 60 mg) intended for once-daily dosing. Because the drug is acid labile, each capsule contains enteric-coated pellets of the drug; the enteric coating is designed to prevent degradation in the acidic environment of the stomach.

The dosage range in the majority of clinical trials has been between 40 and 120 mg/day. Although data suggest that the drug is generally well tolerated at doses above 60 mg daily, there is no evidence suggesting any additional antidepressant effect beyond this dose.⁴³

Nausea, dizziness and dry mouth are among the more common adverse effects.⁴³ In a study comparing the tolerability of duloxetine 30 and 60 mg in adult patients, treatment with 30 mg daily for 1 week followed by a dose increase to 60 mg daily was associated with significantly lower incidence of nausea than treatment initiated at the full dose of 60 mg daily.⁹⁷ If tolerability is a concern, a lower starting dose of 30 mg once daily can be considered.

Venlafaxine

Venlafaxine is also available in Canada in an extended-release capsule formulation (37.5 mg, 75 mg and 150 mg). Compared with immediate-release formulations, extended-release formulations are presumed to enhance patients' compliance with therapy and may be associated with a lower incidence of adverse effects because peak plasma concentration is lower (but sustained for a longer period). Venlafaxine has a wide dosage range, usually from 75 to 225 mg/day. Severe depression has been treated with doses of up to 375 mg a day. For adults, the recommended starting dose is 75 mg once daily with food, either in the morning or in the evening. If tolerability is a concern, a lower starting dose (37.5 mg once daily, for at least 4–7 days) may be considered before the dose is increased to 75 mg/day.³³

Geriatric populations (65 years of age and over)

Physiological and functional changes that occur with age can result in differences in efficacy and adverse event profiles between older and younger patient populations.⁹⁸ In a pharmacokinetic study of duloxetine involving a single 40 mg dose administered to 12 healthy older women (65–77 years of age) and 12 healthy adult women (32–50 years), investigators found no difference in C_max but noted a 24% increase in AUC and a 4.3-hour increase in the half-life in the older group.⁹⁹ None of these changes were statistically significant. Although the study revealed no overall difference in primary pharmacokinetic profiles between age groups, a larger sample size would have allowed more definitive conclusions. One study involving elderly patients with recurrent major depressive disorder compared duloxetine 60 mg daily (without prior titration) with placebo.¹⁰⁰ The discontinuation rate did not differ significantly between the duloxetine and placebo groups, but 8.3% of the patients treated with duloxetine required a dose reduction (to 30 mg daily).

The product monograph for duloxetine suggests that no adjustment in dosing is necessary for older adults, but recommends caution in treating this population.⁴³ Although many geriatric patients may tolerate an initial dose of 60 mg daily, it would be reasonable to start at a lower dose, particularly for frail elderly patients and those with comorbidities that would compromise elimination of the drug.

For venlafaxine, phase II and III clinical trials have shown no differences in efficacy and safety between geriatric and adult populations.³³ One study investigated the pharmacokinetic profiles of venlafaxine in 18 older (60–80 years) and 18 younger (21–44 years) adults and found that no dosage adjustments were necessary on the basis of age.¹⁰¹ Although no specific dosage adjustments are recommended for the geriatric population, a similar approach of conservative dosing with slow upward titration may be advisable.

Renal impairment

There are limited data pertaining to the effects of duloxetine in patients with end-stage renal disease. One study showed that patients with renal dysfunction who were given a single 60 mg dose of the drug had AUC and C_max values about 100% greater than people with normal renal function.²⁷ The same study showed that the AUC of 2 major metabolites were 7 and 9 times higher, respectively, in the group with renal impairment. There are no recommended dosage adjustments for patients with mild to moderate renal impairment, but duloxetine is not recommended for patients with creatinine clearance less than 30 mL/min or for patients undergoing dialysis.^43,102

The pharmacokinetics of venlafaxine and ODV were evaluated in an open-label study in which patients with various degrees of renal impairment were given a single 50 mg dose.¹⁰³ Compared with normal controls, the elimination half-life for both venlafaxine and ODV was prolonged among patients with renal impairment. The authors recommended a 50% dose reduction for patients with creatinine clearance less than 30 mL/min. Data on file indicate that patients with moderate to severe renal impairment (glomerular filtration rate 10–70 mL/min) who were treated with venlafaxine had a 50% increase in half-life and a 24% decrease in clearance compared to normal subjects.³³ The Canadian monograph for this drug also recommends a total daily dose reduction of 25%–50% for those with renal impairment and a 50% dose reduction for those undergoing hemodialysis, with the dose being given after completion of dialysis.³³ Given the variability in clearance of venlafaxine in patients with renal impairment, it would be reasonable to commence therapy at doses as low as 37.5 mg/day and titrate upward according to individual response and tolerability.

Hepatic impairment

Duloxetine has been associated with liver injury, significant elevation of liver enzymes and hepatitis. Six patients with moderate hepatic dysfunction had reduced plasma clearance, a 5-fold increase in AUC and a 3-fold increase in half-life compared with age- and sex-matched healthy controls.⁴³ This drug is not recommended for patients with hepatic insufficiency and should not be given to those with excessive alcohol consumption.

The pharmacokinetic profile of venlafaxine may be altered in patients with hepatic impairment, so dosage adjustment is recommended for this population. The recommended starting dose is 37.5 mg daily and the total daily dose should be reduced by 50% or more.³³

Conclusion

Duloxetine and venlafaxine have similar pharmacologic profiles and similar efficacy in the treatment of depression. However, their pharmacokinetic and adverse effect profiles differ. Hepatotoxicity has been well documented for duloxetine and has also been reported for venlafaxine. The information available at this time indicates that the most clinically important differences between these 2 drugs relate to blood pressure elevation, which is more likely with venlafaxine, and the degree of protein binding (and hence increased potential for drug interactions) with duloxetine. Given the paucity of information about duloxetine in overdose, further investigation comparing duloxetine and venlafaxine in terms of safety in overdose is warranted.