When the strictest criteria are used to sift through all the research literature on the pharmacological treatment of depression and dysthymia, only using large studies with low bias and meticulously designed trials, interestingly, only two studies since 2000 made the grade (with 90% certainty, using AI algorithm):

https://www.ncbi.nlm.nih.gov/pubmed/21154393

https://www.ncbi.nlm.nih.gov/pubmed/8792754

In sum, using sertraline or imipramine does work better than placebo.

Adding aripiprazole, olanzapine, quetiapine or risperidone to antidepressants does improve outcomes.

Quetiapine can treat depression alone.  Olanzapine can too.

Interesting indeed.