Predictors of therapeutic effects in amitriptyline treatment--1. Plasma drug levels.

OBJECTIVE

To analyze the spectrum of relationships between clinical effects of amitriptyline (At) treatment after 2 and 4 weeks (wks) and plasma levels of At, nortriptyline (Nt), At+Nt, demethylation rate of At, treatment modalities, age, and gender.

METHODS

Patients with major depression (ICD 10: F31-F33) and a HAMD-21 total score of 15-41 received At on a dosage schedule chosen by the doctor for at least 4 wks. Plasma drug levels were assessed at baseline and at wks 2 and 4.

RESULTS

Of the 58 patients enrolled in the study, 47 (15males, 32 females) were eligible for statistical analysis. An early response by wk 2 (decrease in HAMD-21 score of at least 50 % from baseline) was observed in 34.0 % of patients, and after 4 wks, the response rate was 63.8 % (males 86.6 %, females 53.1 %). There was a low, negative, and significant correlation between percent reduction in HAMD and steady state At concentration only at wk 2 (n = 47 r Sp. = -0.306 p < 0.05). However, the correlation was dependent on the degree of At demethylation and treatment modalities. A ratio of Nt/At >1 was observed in 23 patients; of these, 11 (47.8 %) were non-responders by wk 4. A low rate of demethylation (Nt/At </= 1) in 24 patients was associated with a lower frequency of non-response (6/24 = 25.0 %). In the latter subgroup, the 2-wk At plasma concentration varied between 20 and 240 ng/ml, and all patients with </= 110 ng At /ml were responders by wk 4. On the other hand, in the Nt/At >1 group, the non-responders were distributed in the whole range of observed At plasma concentrations (20-150 ng/ml). Only in patients with Nt/At </= 1 and constant doses from wk 1 to wk 4 (n = 14) was a strong, linear, and negative correlation between percent reduction in HAMD and plasma At, Nt, or At+Nt concentrations found for the 2-wk data (r Sp. = -0.754, -0.716, and -0.732, respectively), but not after 4 wks.

CONCLUSIONS

The frequently assumed dependence of clinical effects on plasma drug levels is not a simple function. There most likely exist different patterns of relationships with partially unknown backgrounds. The ratio Nt/At may become a discriminating variable in future studies on the nature of plasma level-therapeutic effect relationship in At treatment of major depression.