Claudio A. Naranjo

A method for estimating the probability of adverse drug reactions

The estimation of the probability that a drug caused an adverse clinical event is usually based on clinical judgment. Lack of a method for establishing causality generates large between‐raters and within‐raters variability in assessment. Using the conventional categories and definitions of definite, probable, possible, and doubtful adverse drug reactions (ADRs), the between‐raters agreement of two physicians and four pharmacists who independently assessed 63 randomly selected alleged ADRs was 38% to 63%, kappa (k, a chance‐corrected index of agreement) varied from 0.21 to 0.40, and the intraclass correlation coefficient of reliability (R[est]) was 0.49. Six (testing) and 22 wk (retesting) later the same observers independently reanalyzed the 63 cases by assigning a weighted score (ADR probability scale) to each of the components that must be considered in establishing causal associations between drug(s) and adverse events (e.g., temporal sequence). The cases were randomized to minimize the influence of learning. The event was assigned a probability category from the total score. The between‐raters reliability (range: percent agreement = 83% to 92%; κ = 0.69 to 0.86; r = 0.91 to 0.95; R(est) = 0.92) and within‐raters reliability (range: percent agreement = 80% to 97%; κ = 0.64 to 0.95; r = 0.91 to 0.98) improved (p < 0.001). The between‐raters reliability was maintained on retesting (range: r = 0.84 to 0.94; R(est) = 0.87). The between‐raters reliability of three attending physicians who independently assessed 28 other prospectively collected cases of alleged ADRs was very high (range: r = 0.76 to 0.87; R(est) = 0.80). It was also shown that the ADR probability scale has consensual, content, and concurrent validity. This systematic method offers a sensitive way to monitor ADRs and may be applicable to postmarketing drug surveillance.

Withdrawal reaction after long-term therapeutic use of benzodiazepines.

We conducted a double-blind, placebo-controlled trail in which 40 patients who had undergone long-term therapy with benzodiazepines were switched to placebo or to diazepam in a dose approximately equivalent to their usual dose of the benzodiazepine; the dose of diazepam was then tapered during an eight-week period. Patients were assessed clinically and psychologically and had weekly sessions of behavioral therapy. The subjects who received placebo had more symptoms, assessed their symptoms as more severe, and stopped taking the study drug at a higher rate than those receiving the tapering doses of diazepam. The subjects in the placebo group also had symptoms shortly after being switched to placebo, whereas those in the diazepam group had symptoms much later. Some withdrawal symptoms were distinct from those of anxiety (e.g., tinnitus, involuntary movement, and perceptual changes). Withdrawal symptoms occurred earlier in patients who had received short-acting benzodiazepines than in those who had received long-acting benzodiazepines. Symptoms gradually disappeared over a four-week period in both the placebo and the diazepam groups. Serial determination of plasma benzodiazepine concentrations was a useful way to assess compliance, treatment outcome, and relapse during withdrawal. We conclude that a clinically important, mild, but distinct withdrawal syndrome occurs after discontinuation of long-term therapeutic use of benzodiazepines.

The serotonin uptake inhibitor citalopram attenuates ethanol intake

No effective drug for decreasing ethanol intake is available for clinical use. Our previous studies showed that zimeldine decreased ethanol intake in rats and nondepressed alcohol abusers. However, zimeldine was withdrawn from the market because of serious toxicity. We tested citalopram, a selective serotonin uptake inhibitor, in 39 male nondepressed early‐stage problem drinkers (aged 19 to 61 years). Subjects were randomly allocated to receive either citalopram, 20 (n = 20) or 40 (n = 19) mg/day orally, or placebo in a double‐blind, crossover trial. Citalopram administration and ethanol intake were assessed by self‐report and objectively. Citalopram, 20 mg/day, did not show an effect. However, citalopram, 40 mg/day, decreased the number of drinks consumed (F1,17 = 5.27; P < 0.05) and increased the number of abstinent days (F1,17 = 13.18; P < 0.005). The effect is probably through modulation of the neurobiologic mechanisms regulating ethanol intake. Our results suggest a new pharmacologic approach to decrease ethanol intake.

Zimelidine‐induced variations in alcohol intake by nondepressed heavy drinkers

The effect of zimelidine, a specific serotonin‐reuptake inhibitor, on alcohol intake was tested in 13 healthy male, nondepressed heavy drinkers who were randomly allocated to receive zimelidine or placebo in a double‐blind, crossover experiment. There were five 2‐wk experimental periods (baseline, placebo 1 and 2, and zimelidine 1 and 2). Treatment was discontinued in three subjects due to a suspected adverse reaction and three other subjects dropped out. Thus, 13 subjects participated in at least two experimental drug periods and only 10 participated in all the periods. In the 13 subjects zimelidine increased the days of abstinence and decreased the daily number of drinks consumed, whereas in the 10 subjects only the number of days of abstinence increased. Subjects did not report aversive alcohol‐sensitizing reactions. Spielberger state‐anxiety test scores and depression scores (Montgomery/Asberg and Hamilton) were low at the beginning and throughout the study. Our data suggest that zimelidine modifies alcohol intake by a different mechanism than previously tested drugs, possibly by modulating the central neural mechanism that controls drinking of alcohol.

Fluoxetine differentially alters alcohol intake and other consummatory behaviors in problem drinkers

The effects of fluoxetine, a relatively selective long‐acting serotonin uptake inhibitor, on the consumption of alcoholic and nonalcoholic drinks, cigarette smoking, and body weight were assessed in 29 men who were early stage problem drinkers. After a 2‐week baseline, subjects were randomly assigned to receive 40 mg/day fluoxetine (n = 8), 60 mg/day fluoxetine (n = 11), or placebo (n = 10) for 4 weeks. Fluoxetine 60 mg/day decreased mean daily alcoholic drinks from (X̄ ± SEM) 8.3 ± 0.7 during baseline to 6.9 ± 0.7 and decreased total drinks per 14 days from 115.8 ± 9.3 to 96.5 ± 9.5 (p < 0.01; 17.3% decrease from baseline), with no significant increase in days of abstinence. Neither 40 mg/day fluoxetine nor placebo had effects on intake of alcohol. Fluoxetine 60 mg/day decreased total and mean daily alcoholic drinks compared with 40 mg/day fluoxetine (ANCOVA, both p < 0.02), but neither dose of fluoxetine was different from placebo. Compared with placebo, both 40 mg/day fluoxetine and 60 mg/day fluoxetine decreased the variability of the baseline to treatment changes in alcoholic drinks (both p < 0.05). Although no differences were detected between treatment groups, 60 mg/day fluoxetine increased mean daily nonalcoholic beverages from baseline (5.0 ± 0.4 to 5.6 ± 0.3, p < 0.01) and increased daily cigarettes smoked (from 25.1 ± 4.6 to 26.9 ± 4.5, p < 0.05), whereas no significant changes from baseline were observed with 40 mg/day fluoxetine or placebo. Body weight decreased with both 40 mg/day fluoxetine (from 75.7 ± 4.7 kg to 73.8 ± 4.6 kg, p < 0.01) and 60 mg/day fluoxetine (from 81.4 ± 2.6 kg to 79.2 ± 2.5 kg, p < 0.05) but not with placebo. Patterns of response varied, but decreases in alcohol consumption were not related to side effects, an alcohol‐sensitizing reaction, or changes in depression or anxiety. Our findings with men who are problem drinkers indicate that fluoxetine differentially alters consummatory behaviors. The reductions in alcohol intake and body weight are of clinical importance.

Diazepam loading: Simplified treatment of alcohol withdrawal

Alcohol withdrawal therapy can be simplified with a loading dose of diazepam, taking advantage of the kinetic tapering afforded by the drugs long t½s and its metabolites, and of the effectiveness of nonpharmacologic maneuvers. In a double‐blind trial, 50 inpatients in moderate to severe alcohol withdrawal received 20 mg oral diazepam and supportive care (n = 25) or placebo and supportive care (n = 25) every 2 hr until they were asymptomatic. Fifty‐six percent of patients responded to placebo within 5 ± 2.9 hr (mean ± SD), whereas 72% responded to initial diazepam within 6.3 ± 3.9 hr. Patients treated with diazepam had more rapid and greater improvement than those treated with placebo. Patients who did not respond to six doses of diazepam received further (unblinded) diazepam, 20 mg, every 1 to 2 hr. All patients who did not initially respond (n = 18) improved after more diazepam. Thus all patients who received diazepam (n = 36), during the experimental phase or subsequently, were effectively treated. There were no adverse effects. The median number of 20‐mg diazepam doses to treat alcohol withdrawal were three, given over a period of 7.6 hr (range = 1 to 12 and 0.33 to 45 hr). Complications occurred only in those who received placebo during the experimental phase, indicating that delay in therapy may be responsible for the appearance of complications in alcohol withdrawal.

Selective serotonin reuptake inhibitors and CNS drug interactions: A critical review of the evidence

SummaryThe potential for drug-drug interactions in psychiatric patients is very high as combination psychopharmacotherapy is used to treat comorbid psychiatric disorders, to treat the adverse effects of a medication, to augment a medication effect or to treat concomitant medical illnesses. Interactions can be pharmacodynamic or pharmacokinetic in nature. This paper focuses on the metabolic kinetic interactions between selective serotonin reuptake inhibitors (SSRIs) and other central nervous system (CNS) drugs. The evidence for and clinical significance of these interactions are reviewed, with special emphasis on antipsychotics, tricyclic anti-depressants and benzodiazepines.Many psychotropic medications have an affinity for the cytochrome P450 (CYP) enzymes which promote elimination by transforming lipid soluble substances into more polar compounds. SSRIs serve both as substrates and inhibitors of these enzymes. In vitro studies provide a screening method for evaluating drug affinities for substrates, inhibitors or inducers of CYP enzymes. Although in vitro data are important as a starting point for predicting these metabolic kinetic drug interactions, case reports and controlled experimental studies in humans are required to fully evaluate their clinical significance. Several factors must be considered when evaluating the clinical significance of a potential interaction including: (a) the nature of each drugs’ activity at an enzyme site (substrate, inhibitor or inducer); (b) the potency estimations for the inhibitor/inducer; (c) the concentration of the inhibitor/inducer at the enzyme site; (d) the saturability of the enzyme; (e) the extent of metabolism of the substrate through this enzyme (versus alternative metabolic routes); (f) the presence of active metabolites of the substrate; (g) the therapeutic window of the substrate; (h) the inherent enzyme activity of the individual, phenotyping/genotyping information; (i) the level of risk of the individual experiencing adverse effects (e.g. the elderly) and (j) from an epidemiological perspective, the probability of concurrent use.This paper systematically reviews both the in vitro and in vivo evidence for drug interactions between SSRIs and other CNS drugs. As potent inhibitors of CYP2D6, both paroxetine and fluoxetine have the potential to increase the plasma concentrations of antipsychotic medications metabolised through this enzyme, including perphenazine, haloperidol, thioridazine and risperidone in patients who are CYP2D6 extensive metabolisers. Controlled studies have demonstrated this for perphenazine with paroxetine and haloperidol with fluoxetine. Fluvoxamine, as a potent inhibitor of CYP1A2, can inhibit the metabolism of clozapine, resulting in higher plasma concentrations.Drug interactions between the SSRIs and tricyclic antidepressants (TCAs) can occur. Fluoxetine and paroxetine, as potent inhibitors of CYP2D6, can increase the plasma concentrations of secondary and tertiary tricyclic antidepressants. Sertraline and citalopram are less likely to have this effect. Fluvoxamine can increase the plasma concentrations of tertiary TCAs.Fluvoxamine inhibits, via CYP3A, CYP2C19 and CYP1A2, the metabolism of several benzodiazepines, including alprazolam, bromazepam and diazepam. Fluoxetine increases the plasma concentrations of alprazolam and diazepam by inhibiting CYP3A and CYP2C19, respectively. The clinical importance of the interaction with diazepam is attenuated by the presence of its active metabolite. Sertraline inhibits these enzymes only mildly to moderately at usual therapeutic doses. Therefore the potential for interactions is less; however, the in vivo evidence is minimal. Paroxetine and citalopram are unlikely to cause interactions with benzodiazepines.The evidence is conflicting for an interaction between carbamazepine and the SSRIs fluoxetine and fluvoxamine. These combinations should be used cautiously, and be accompanied by monitoring for adverse events and carbamazepine plasma concentrations. A lack of interaction between paroxetine or sertraline and carbamazepine has been documented.The SSRIs are not equivalent in their potential for drug interactions when combined with other CNS medications. Each combination must be assessed individually. Several factors must be considered when predicting the outcome of a potential interaction based on in vitro data (e.g. active metabolites and concentration ranges). In vivo studies are required to evaluate their clinical significance. Generally, sertraline and citalopram at the lower therapeutic dosage range appear to have less propensity for interactions. Anticipated pharmacokinetic interactions can usually be managed with careful monitoring and appropriate adjustments in dosage and titration.

A validation study of The Geriatric Depression Scale short form

The validity of the Geriatric Depression Scale (GDS) short form was assessed in a geriatric affective disorders outpatient clinic (N = 116). The GDS was highly correlated with the Montgomery Asberg Depression Rating Scale (MADRS), and with optimal cutoff scores of 5/6, demonstrated a sensitivity of 85% and a specificity of 74.0%. The GDS appears to be a useful, valid screening instrument in this population.

The efficacy, safety and tolerability of antidepressants in late life depression : a meta-analysis

BACKGROUND To determine the efficacy, safety and tolerability of antidepressants in depressed elderly patients. METHODS Search for randomized controlled double-blind studies evaluating atypical antidepressants (ATYPs), reversible inhibitors of monoamine oxidase-A, selective serotonin reuptake inhibitors (SSRIs) and tricyclic antidepressants in moderate/severe depressed patients > or = 60 years for > or = four weeks. The random effects model (single-arm; comparative) was used to aggregate efficacy, safety and dropout. RESULTS No difference in single-arm aggregation of outcomes for four antidepressant classes. Comparative analyses showed no statistical difference between outcomes, except SSRIs had a higher response rate than ATYPs. CONCLUSION Elderly show no differences in antidepressant class outcomes. LIMITATIONS Heterogeneity and lack of power. CLINICAL RELEVANCE There is little advantage for antidepressant classes over another in the aged.

Novel Pharmacological Interventions for Alcoholism

Improved methods for treating alcoholism are greatly needed, and the use of pharmacological agents holds great promise. Development was previously hindered by the lack of an accepted animal model. Great emphasis has been placed upon designated criteria for an animal model of alcoholism, but such face validity is unimportant for finding better treatments. What really is needed is predictive validity -the ability to predict which drugs will reduce alcohol drinking in alcoholics. The simple model of rats voluntarily selecting unflavored ethanol has been shown to possess good predictive validity (Sinclair, 1987) and is currently being used for screening drugs for potential usefulness in alcoholism treatment. The next step should be development of special animal models for screening drugs that act on specific components of alcohol drinking: e.g., blocking the deprivation-induced increase in craving or the accompanying arousal, correcting underlying disorders promoting drinking in some individuals, suppressing stimuli triggering drinking (e.g., withdrawal sensations or stimuli arising from the first drink after abstinence), satisfying the alcohol craving or blocking tolerance to satiety signals from alcohol. Particularly promising are drugs that block alcohol reinforcement and thus could be used to extinguish the alcohol-drinking response.