Objective: Recent research with several synthetic antidepressants indicates that early improvement during the initial weeks of treatment may be a highly sensitive predictor of therapeutic success in major depression. We investigated whether early improvement is sensitive and specific in predicting sustained response and non-response to antidepressant treatment with Hypericum extract WS(R) 5570/5572 and whether the results reported for synthetic antidepressants apply to these Hypericum extracts as well.
Methods: We analyzed original data of 3 double-blind, randomized trials including a total of 594 adult out-patients suffering from major depression according to DSM-IV criteria, who received well-characterized Hypericum extract preparations WS(R) 5570, WS(R) 5572, WS(R) 5573 or placebo for 6 weeks. The main outcome measure was the prediction of a sustained > or = 50 % decrease of the Hamilton Depression Scale (HAM-D) total score versus baseline ('sustained response') by the presence of > or =20 % HAM-D total score improvement after 1 and 2 weeks of treatment ('early improvement').
Results: For Hypericum extract, early improvement had a sensitivity of 87 % (95 % confidence interval [CI], 82-93 %) and a specificity of 54 % (95 % CI, 45-62 %) at day 14, and a sensitivity of 43 % (95 % CI, 35-51 %) and a specificity of 86 % (95 % CI, 80-92 %) at day 7 for predicting sustained response. After 2 weeks of treatment, 78 % (95 % CI, 69-87 %) of the patients who failed to improve did not show sustained response later during treatment.
Conclusion: A substantial fraction of the patients treated with Hypericum extracts WS(R) 5570/5572 showed a meaningful reduction of depressive symptoms during the first two weeks of treatment (early improvement), which was found to be a sensitive predictor of sustained response. The results determined for the herbal extracts were comparable to those for effective synthetic antidepressants investigated previously.