RESUMEN
In a prospective, open clinical study, the relationship between serum levels of amitriptyline (At) and nortriptyline (Nt) and the therapeutic effect after 6 weeks of treatment was investigated. Serum levels were measured by gas-liquid chromatography and the therapeutic effect was assessed by the Hamilton Depression Rating Scale (HAMD) and the Clinical Global Impression Scale (CGI). A number of 25 non-delusional, moderate to severely depressed inpatients were included. A therapeutic window relationship was detected by means of regression analysis (quadratic model). Low and high serum levels were associated with low therapeutic effect. In an intermediate range, the probability of good therapeutic effect was increased. This relationship reached significance for the serum levels of At (p < 0.05) and a trend for the sum of serum levels of At and Nt (p < 0.1). As expressed by the regression coefficient r2, about 25% to 35% of the variability of therapeutic effect was explained by serum levels. Dichotomized data sets according to limits of final values of HAMD and CGI as well as limits of a therapeutic window of 70 ng/ml and 200 ng/ml (sum of At and Nt) revealed significant differences by means of Fisher's exact test (p < 0.05). Furthermore, increased ratios of serum level of Nt per serum level of At were found to be associated with decreased therapeutic effect. Thus, the present data support the existence of a therapeutic window of serum levels of At in depression. Also taking into account other reports, this therapeutic window can be defined as being between about 70 and 220 ng/ml. The assay of serum levels of At can be used to lower the risk of unsatisfactory therapeutic outcome.