Antihistamines

Prolonged QT interval and ventricular dysrhythmias

The EIDOS and DoTS descriptions of this adverse reaction are shown in Figure 1 .

Several antihistamines can cause ventricular dysrhythmias of the torsade de pointes type [ ], first reported with astemizole [ ] and later with terfenadine [ ]. Astemizole and terfenadine both have a dose-dependent effect on cardiac repolarization and cause prolongation of the QT interval, which can lead to ventricular dysrhythmias (such as torsade de pointes), syncope, and cardiac arrest. Reported cases relate preponderantly to overdosage, especially in children [ ]. Terfenadine and astemizole have been described as having dysrhythmogenic actions, and deaths have been described [ , ]. The effects of some antihistamines on the QT interval are listed in Table 1 .

With a few exceptions, antihistamines are rapidly and completely absorbed after oral administration; peak plasma concentrations are reached after 1–4 hours and are highly variable, owing to differences in tissue distribution and metabolism [ ]. Many of the second-generation antihistamines (for example astemizole, ebastine, loratadine, and terfenadine) undergo extensive first-pass metabolism to pharmacologically active metabolites; as a common feature, the reaction is primarily supported by CYP3A4. Under normal circumstances this extensive metabolism leads to low or undetectable plasma concentrations of the parent drug. However, sometimes metabolism of the parent compound can be compromised. Accumulation of unmetabolized astemizole or terfenadine can result in blockade of cardiac potassium channels in the ventricular myocytes that regulate the duration of the action potential; consequent prolongation of the QT interval can result in potentially life-threatening ventricular tachycardia [ ].

Dysrhythmias can also occur with therapeutic doses of these and other antihistamines, if certain other susceptibility factors are present:

• impaired hepatic metabolism due to liver disease;

• simultaneous treatment with drugs that are inhibitors of the cytochrome P450 enzyme CYP3A4 (for example macrolide antibiotics, antifungal azoles, or grapefruit juice), leading to increased plasma concentrations thereby raising the risk of cardiotoxic effects [ ];

• pre-existing QT prolongation caused by congenital long QT syndrome, other heart disease, or treatment with antidysrhythmic drugs, such as class I antidysrhythmic drugs, amiodarone, or sotalol;

• electrolyte imbalance; in particular, hypokalemia predisposes to dysrhythmias.

Terfenadine is especially likely to cause torsade de pointes in patients in whom these susceptibility factors are present [ , ]. Ventricular dysrhythmias can also occur after overdosage of antihistamines that prolong the QT interval.

The mechanism responsible for dysrhythmias has been identified as blockade of HERG potassium channels [ ]. The dysrhythmogenic potential of antihistamines has been evaluated in vitro using cloned human potassium channels or guinea-pig heart muscle cells, and using an in vivo guinea-pig model. Studies in humans, including the assessment of drug interactions, are considered more reliable. Investigations in human volunteers have shown that there are no significant electrocardiographic changes with azelastine, cetirizine, fexofenadine, and loratadine even at several times the therapeutic doses, which shows that cardiotoxicity is not a class effect [ ]. Mizolastine also appears to cause no cardiac problems in humans [ ]. Large doses of ebastine have shown cardiac effects in guinea pigs, but QT prolongation has not occurred in human studies with up to three times therapeutic doses [ ]. Slight QT prolongation was seen on further increased doses to 100 mg/day and when subjects were given erythromycin or ketoconazole, but the effect was less than the effect of terfenadine and was not considered clinically relevant [ ]. The active metabolite of ebastine, carebastine, had no effect on the QT interval, even in large doses.

The absolute risk of antihistamine-induced dysrhythmias is low in the general population. In an epidemiological study using a general practice database, the crude incidence of ventricular dysrhythmias was 1.9 per 10 000 person-years, corresponding to a relative risk of 4.2 for all antihistamines compared with non-use. Astemizole presented the highest relative risk, whereas terfenadine was in the range of other non-sedating antihistamines. Older age was associated with greater risk. The absolute risk in this study was one case per 5300 person-years of use [ ].

In the USA, terfenadine was withdrawn from the market in 1998, and in other countries terfenadine has been moved from over-the-counter to prescription-only, with only 60 mg tablets available. The active metabolite of terfenadine, fexofenadine, is marketed as an alternative. For astemizole this option was not available, since the main metabolite (desmethylastemizole) is also cardiotoxic and has a half-life of 10 days; astemizole was therefore withdrawn from the market worldwide in June 1999.

Although it is widely believed that cardiotoxicity of antihistamines is limited to second-generation compounds, both hydroxyzine and diphenhydramine can block potassium channels. Caution should therefore be exercised in prescribing first-generation antihistamines for patients with a predisposition to cardiac dysrhythmias. For example, therapeutic doses of diphenhydramine caused prolongation of the QT interval in healthy volunteers and in patients undergoing angioplasty [ ], and one cannot exclude the possibility that first-generation drugs that modulate potassium channels may in some circumstances cause dysrhythmias [ ]. All antihistamines should be screened for cardiotoxicity, as some patients may be poor metabolizers or may be susceptible to plasma concentrations near to the usual therapeutic range. Useful information may be obtained from pharmacokinetic studies using potential inhibitors (see under Drug–Drug Interactions).

The single- and multiple-dose pharmacokinetics of ebastine (10 mg) have been determined in elderly and young healthy subjects using 24-hour Holter monitoring [ ]. There were no clinically relevant effects.

The incidence of ventricular dysrhythmias associated with non-sedating antihistamines (including cetirizine) has also been assessed using the UK-based General Practice Research Database [ ]. There were 18 cases over the period 1992–96. Astemizole was associated with the highest relative risk. The risk associated with terfenadine was no different from that with other non-sedating antihistamines, and there was no single case of ventricular dysrhythmia with the concomitant use of P450 inhibitors and terfenadine.

In a comparison of the dysrhythmogenic potential of a series of second-generation antihistamines, the antihistamines were given intravenously and electrocardiographic and cardiovascular parameters (blood pressure and heart rate) were measured. The lowest dose that produced significant prolongation of the QT _c interval was compared with the dose required to inhibit by 50% the peripheral bronchospasm elicited by histamine 10 micrograms/kg intravenously. Astemizole, ebastine, and terfenadine produced pronounced dose-dependent QT _c interval prolongation. In contrast, terfenadine carboxylate, norastemizole, and carebastine, the major metabolites of terfenadine, astemizole, and ebastine, and cetirizine had no effects [ ].

Antihistamines and drowsiness

Central nervous depression causing sedation is the most common adverse effect of the first-generation antihistamines [ ]. The sedative effect of antihistamines is evaluated using psychometric tests, tests of driving performance, and subjective scoring or visual analogue scales, but results from studies using healthy volunteers cannot necessarily be extrapolated to patients, one difficulty being that the treated disease can itself cause sedation [ ]. The drowsiness has been attributed to inhibition of histamine N -methyltransferase and to blockade of central histamine receptors, together with actions on other receptors, in particular 5-HT receptors [ , , , ]. Daytime drowsiness can be a problem, above all when driving or operating machinery. As with many other nervous system depressants, this effect may abate or disappear after several days of use, but co-medication with certain other agents or a short period of withdrawal of therapy may reactivate the sedative effect.

The signal characteristic of the second-generation antihistamines is their freedom from sedation [ , ]. The relative lack of sedative properties in the second-generation antihistamines has been ascribed to their relative hydrophilicity. Little is known about intracerebral concentrations of antihistamines and their metabolites, but positron emission tomography has shown that the first-generation antihistamine chlorphenamine occupied a larger fraction of brain histamine H ₁ receptors than terfenadine [ , ]. Differential affinity for, or different actions on, central and peripheral H ₁ receptors [ ] could also explain variations in sedative effect, but differences in receptor binding have only been shown for loratadine in vitro [ ].

The nervous system depressant effects of fexofenadine [ , ], loratadine [ ], and mizolastine [ ] appear to be no greater than those seen with placebo. However, the generality of the claim that second-generation antihistamines are free of sedative effects has been challenged [ ]. The issue is complicated by evidence that sedation in allergic disease (and subsequent impairment in performance and learning) can be a consequence of the condition itself, as opposed to being wholly due to antihistamines [ ]. This raises concerns about the purported risk-free sedation profiles of certain antihistamines, given that they are often based on objective studies in healthy volunteers [ ]. Another issue is the tendency of patients with allergies to self-medicate, titrating their antihistamine dosage upwards to achieve relief of symptoms; neurological impairment does in fact occur if the doses of cetirizine, loratadine, or mizolastine are increased sufficiently [ ]. Thus, it is more correct to describe the second-generation antihistamines as having minimal sedative effects when taken in recommended doses. A grouping of the antihistamines into those with marked, moderate, and very low sedative effect is possible. However, the dividing lines are not sharp and classification often depends on how many studies are taken into account, since results are not consistent.

The designs of protocols used in comparisons between sedative and non-sedative compounds have been questioned. They may not accurately reflect the clinical use of each drug, and the data may be misused in advice to prescribers, even though the reason a comparator was included was merely to provide an active control. Extrapolation of the results of cognitive studies in healthy volunteers to patients may be inappropriate, as a drug that is sedative in a healthy volunteer may well not be perceived to be sedative by a patient with allergic symptoms, although caution must be taken in relying on subjective assessments of performance and drowsiness. Patients with mild to moderate allergic rhinitis complain of sleep difficulties, and many who take a sedative can function reasonably well during the next day without further medication. Duration of treatment can also play a part, and certain investigations may have been too brief; in some patients, sedation is induced by the drug only after some weeks of treatment [ ]. Such discrepancies could explain why, despite all the investigations with positive results, some studies report sedation in 30% of the patients [ ].

It is likely that the controversy will be settled by accepting the relative merits of each drug, and that sedative drugs will continue to be prescribed, at least for overnight ingestion and for some skin conditions.

There is the special case of the use of antihistamines by individuals whose work may compromise their own safety or the safety of others, for example transport workers. Indeed, much of the support for second-generation drugs arises from safety considerations. Nevertheless, it is sometimes suggested that recommendations for the use of antihistamines by those involved in skilled activities should be based on studies of the patients themselves carrying out their day-to-day work, for example airline pilots with allergic rhinitis operating aircraft. This is an argument that lacks careful thought. In occupational medicine it is essential that controlled studies in healthy volunteers are used to establish whether an antihistamine has sedative properties, and then to choose the drug that is least likely to impair performance or cause drowsiness.

Observational studies : Antihistamines are effective and safe in preventing the symptoms of a mosquito bite; ebastine and loratadine did not cause sedation in such cases [ , ].

Comparative studies : The frequency of sedation due to acrivastine, cetirizine, fexofenadine, and loratadine has been investigated in four prescription-event monitoring studies in 43 363 patients in general practice in the UK [ ]. Prescriptions were obtained for each cohort in the immediate postmarketing period. Sedation and drowsiness were the main outcome measures. The odds ratios (adjusted for age and sex) for the incidences of sedation compared with loratadine were: 0.63 (95% CI = 0.36, 1.11) for fexofenadine, 2.79 (1.69, 4.58) for acrivastine, and 3.53 (2.07, 5.42) for cetirizine. There was no increased risk of accident or injury with any of the four drugs.

The effects of diphenhydramine, fexofenadine, and alcohol on driving performance have been studied in a randomized, placebo-controlled trial in the Iowa driving simulator [ ]. Participants had significantly better coherence after alcohol or fexofenadine than after diphenhydramine. Lane holding (steering instability and crossing the center line) was impaired after alcohol and diphenhydramine compared with fexofenadine. Mean response time to the blocking vehicle was slowest after alcohol (2.21 seconds) compared with fexofenadine (1.95 seconds). Self-reported drowsiness did not predict lack of coherence and was weakly associated with minimum following distance, steering instability, and left-lane excursion. In conclusion, the participants performed similarly when they took fexofenadine or placebo. After alcohol they performed the primary task well but not the secondary tasks, resulting in poorer driving performance. After diphenhydramine, driving performance was poorest, suggesting that diphenhydramine had a greater impact on driving than alcohol did. Drowsiness ratings were not a good predictor of impairment, suggesting that drivers cannot use drowsiness to indicate when they should not drive. Non-sedating antihistamines should therefore be preferred over sedating antihistamines in patients who drive [ ].

Mequitazine has a low propensity to cause drowsiness, comparable to that of cetirizine and loratadine; it therefore differs from truly sedative antihistamines, such as dexchlorpheniramine, which cause drowsiness and fatigue in patients with atopy to a degree that is measurably different from placebo [ ].

Although allergic rhinitis is not usually severe, it affects school learning performance and work productivity [ ]. The effects of loratadine and cetirizine on somnolence and motivation during the working day have been compared in 60 patients with allergic rhinitis in a parallel-group, double-blind study [ ]. Somnolence scores were similar in the two groups at baseline and at the time of dosing (0800 hours). However, cetirizine caused significantly more somnolence at 1000 hours, 1200 hours, and 1500 hours. The scores of motivation to perform activities were similar in the two groups at baseline and 0800 hours. The patients taking loratadine were relatively more motivated at 1000 hours, 1200 hours, and 1500 hours.

In a comparison of the effects over 7 days of a modified-release formulation of brompheniramine (12 mg bd) and loratadine (10 mg od), physicians’ and patients’ assessments were better for brompheniramine than for loratadine, but somnolence and dizziness were reported less often by those who took loratadine, although occurrences were claimed to be less frequent with brompheniramine as treatment continued [ ].

The authors of a report of a comparison of the effectiveness of ebastine (10 and 20 mg) and loratadine (10 mg) for perennial allergic rhinitis claimed that ebastine provided greater symptomatic relief than loratadine, but with a similar low incidence of central effects and headache [ ].

In a comparison of the incidence of drowsiness between cinnarizine (25 mg tds for 7 days, 25 mg bd for 15 days, and 25 mg daily for 15 days) and prochlorperazine (5 mg tds for 7 days, 5 mg bd for 15 days, and 5 mg od for 15 days), drowsiness was observed less often in those taking prochlorperazine [ ].

In a comparison of diphenhydramine, chlorphenamine, cetirizine, loratadine, and placebo in 15 healthy elderly subjects, there were no significant differences between the first- and second-generation antihistamines [ ]. In another study, even the first-generation sedative drug chlorphenamine failed to cause significant sedation in a group of children [ ].

In a comparison of astemizole, terfenadine, and triprolidine (positive control), only triprolidine caused reduced performance and motor incoordination [ ].

In a comparison of the effects of acrivastine, terfenadine, and diphenhydramine on driving performance, there was a dose-dependent effect of acrivastine, with severely affected driving in doses of 16 and 24 mg [ ]. Terfenadine in doses of 60–180 mg did not affect driving performance.

In a comparison of cetirizine and loratadine, cetirizine 10 mg had acute sedative effects and impaired driving performance [ ], whereas loratadine had no sedating potential; furthermore, there was an additive effect of alcohol and cetirizine but not alcohol and loratadine. However, in a study using a driving simulator cetirizine 10 mg did not affect driving ability [ ]. In other studies cetirizine 20 mg caused significant sedation, while in one study there was a dose-dependent sedative effect with 10 mg and 20 mg but not 5 mg [ ]. Pooling the available data [ ] shows that cetirizine is little more sedative than loratadine and terfenadine.

In a comparison of ebastine and triprolidine, only those taking triprolidine had impairment of several parameters of car driving performance [ ].

The effects of loratadine 10, 20, and 40 mg on tests of visuomotor coordination, dynamic visual acuity, short-term memory, digit symbol substitution, and subjective assessments of mood have been studied [ ]. Triprolidine was used as an active control and impaired performance on all the tasks presented. Loratadine 40 mg caused a significant impairment of the Digit Symbol Substitution Test and the Dynamic Visual Acuity Test, but the 10 and 20 mg doses were without effect. Loratadine did not affect objective sleepiness, as measured by Multiple Sleep Latency Test [ ]. In other studies of loratadine in the normal 10 mg dose the sedation rate was no different from placebo [ , ].

In a comparison of the initial and 5-day steady-state effects of loratadine, diphenhydramine, and placebo, using a number of psychometric tests, there was no detectable effect of loratadine compared with placebo, whereas diphenhydramine clearly reduced performance [ ].

Anticholinergic effects

The marked anticholinergic properties of the first-generation antihistamines can cause dryness of the oral and respiratory mucosae. Other antimuscarinic effects are less common, but nasal stuffiness, blurring of vision, urinary retention, and constipation can all occur.

Nervous system stimulation is less frequent than nervous system depression, but when it occurs it causes insomnia, irritability, and tremor; nightmares, and hallucinations. In overt intoxication, these effects may be related to anticholinergic effects. In an analysis of 113 200 admissions to a pediatric hospital there were only two patients with excitation, insomnia, visual hallucinations, and seizures, followed by coma [ ].

Antidopaminergic effects

Antidopaminergic effects of antihistamine drugs can cause extrapyramidal symptoms, including neuroleptic malignant syndrome [ , ]. Prolonged use of antihistamine-containing decongestants can cause facial dyskinesias, including blepharospasm, swallowing difficulties, and dysarthria. As patients with these effects have often been taking combination products containing antihistamines, proper evaluation of interactions is needed before final assessment is possible. As a dyskinesia can be unilateral, a neurological disorder should be excluded before thinking about an adverse effect [ ].

The prognosis of drug-induced parkinsonism has been discussed [ , ]. Negrotti and Calzetti [ ] considered that the results of Martí-Massó and Poza [ ] were over-optimistic. This they ascribed to uncertainty in the collection of their data, which may not have provided adequate evidence of the course of clinical recovery, although differences in cumulative dosages and concurrent use of other drugs could also have been involved. The differences in prognosis may be attributable to the fact that the patients studied by Martí-Massó and Poza [ ] were diagnosed earlier, were less severely affected, and had a good prognosis. Cinnarizine-induced extrapyramidal signs have tended to be associated with old age and prolonged treatment. However, cinnarizine-induced akathisia, parkinsonism, and depression have been reported in a 25-year-old patient after only 11 days of treatment [ ].