[Comparison of 6 different methods for lorazepam withdrawal. A controlled study, hydroxyzine versus placebo]

154 outpatients with generalized anxiety (DSM III-R criteria), followed by general practitioners, gave an informed written consent to participate in this multicenter, randomized, placebo controlled study previously approved by a legal ethic committee (CCPPRB). The patients had to be long term consumers (at least 3 months) of 2 mg daily of lorazepam and were withdrawn using transiently an antihistaminic anxiolytic (hydroxyzine or placebo TAD) according to 6 different procedures defining 6 parallel groups: hydroxyzine 50 mg, abrupt or progressive withdrawal; hydroxyzine 25 mg, abrupt or progressive withdrawal; placebo, abrupt or progressive withdrawal. Following this 4 week-period of withdrawal, the patients were without any treatment for a post-study follow up 2 month-period. Clinical evaluations for anxiety (HARS, Zung), sleep (Spiegel), BZD withdrawal syndrome (Tyrer), adverse reactions and clinical global impression (CGI) were performed at D0, D7, D14, D28, D35 and D88. Investigators opinion and patients attitude towards BZD were collected at D88.

Analysis of variance for quantitative variables and chi square test for qualitative or ordinal variables.

Whatever abrupt or progressive, with or without hydroxyzine support, using half or full dosage, lorazepam withdrawal proved to be feasible even after a long term BZD treatment (mean = 64 months +/- 60). GPs opinion (72% satisfied: D35; 78% satisfied: D88) is satisfying but patients attitude (at D88: 54% patients desired to be regiven a tranquilizer, 31% patients occasionally had a BZD and 22% formally demanded a prescription of BZD) is more questionable. Despite a high initial level of anxiety under lorazepam (HARS = 21 +/- 10 at D0), after a one-month period of withdrawal (under placebo or hydroxyzine) followed by a 2 month-period without any treatment, 75% patients were totally free of any drug and their level of anxiety was significantly decreased (D88: HARS = 12 +/- 9). Progressive withdrawal appeared preferable if compared to abrupt since the number of drop outs between D28 and D88 was less important and the procedure judged more favourably by the patients. Levels of anxiety significantly decreased in both the groups (progressive and abrupt) but sleep parameters and number of withdrawal symptoms between D7 and D28 were improved only in abrupt withdrawal group (p < 0.0001). Considering hydroxyzine, 2 patients dropped out between D0 and D28 in the group hydroxyzine 25 mg, 6 patients in the group 50 mg and 5 patients in the group placebo. Levels of anxiety (HARS et Zung) were significantly improved in hydroxyzine 50 mg group (p < 0.007) and in hydroxyzine 25 mg group (p < 0.012) but not in placebo group. Withdrawal symptoms (Tyrer) between D0 and D28 were improved only in hydroxyzine 50 mg group and the number of side effects was significantly improved in both the hydroxyzine (25 et 50 mg) groups but not in placebo group. However, no significant difference was found between the 3 groups. Daytime sleepiness is more frequent in hydroxyzine 50 mg group.

These results proved a significant improvement of anxiety, a decrease of side effects in both the groups treated with hydroxyzine and a reduction of withdrawal symptomatology in hydroxyzine 50 mg group. When a patient is engaged to be withdrawn from of a lorazepam long term treatment, it can therefore be proposed as a support a transient prescription of hydroxyzine 25 mg TAD to markedly anxious patients and of hydroxyzine 50 mg TAD to patients presenting a withdrawal symptomatology.