Prevalence and Management of TRD 17 J Clin Psychiatry 2007;68 (suppl 8) adequate course of treatment in a patient meeting criteria for major depressive disorder (MDD). In the past decade, almost all parts of this traditional definition have been sub- jected to scrutiny and have had empirical criteria applied to them. Inadequate response has been operationalized by applying specific criteria for treatment response or, alter- natively, remission. The magnitude of TRD increases sub- stantially if failure to achieve remission is used as the qual- ifying criterion. However, although remission is the gold standard outcome for antidepressant treatment, failure to achieve remission is not typically required for a patient to meet criteria for TRD. What constitutes an adequate course of treatment has been operationalized relatively recently by delineating a TRD staging system (Table 1). 1 Criteria for Stage 1 and 2 TRD require failure to respond, respectively, to 1 or 2 ad- equate antidepressant trials. Each trial must comprise anti- depressants of distinctly different classes. Stage 3 requires a third trial which must include a course of treatment with a tricyclic antidepressant, if this class was not used in Stage 1 or 2. Stage 4 TRD requires failure to respond to at least 4 different classes of antidepressants, one of which must be a monoamine oxidase inhibitor. Stage 5 TRD requires meet- ing all Stage 4 criteria, in addition to which a patient must have failed an adequate course of electroconvulsive therapy (ECT). Treatment-resistant depression typically is limited to pa- tients who meet criteria only for unipolar MDD. There is Prevalence and Management of Treatment-Resistant Depression Charles B. Nemeroff, M.D., Ph.D. Treatment-resistant depression (TRD) is a major public health problem in terms of its prevalence and in terms of individual suffering and cost to society. Best estimates indicate 12-month prevalence rates of ~3% for Stage 1 TRD (failure to respond to 1 adequate trial of an antidepressant) and ~2% for Stage 2 TRD (fail- ure to respond to 2 adequate trials). The current article provides a brief review of the definitions, prevalence, and various treatment options for TRD, including switching, augmentation, and combination therapies and use of nonpharmacologic treatments. Given the public health importance of TRD, the relative absence of adequately powered, double-blind trials is striking. (J Clin Psychiatry 2007;68[suppl 8]:17–25) From the Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, Ga. Based in part on a presentation from an expert consensus roundtable entitled “MAOIs Revisited: Their Role in Depressive Disorders,” held in Miami, Fla., on Jan. 29, 2006, and supported by an educational grant from Bristol-Myers Squibb Company. Financial disclosure for Dr. Nemeroff is listed at the end of the article. Corresponding author and reprints: Charles B. Nemeroff, M.D., Ph.D., Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, GA (cnemero@emory.edu). T he traditional definition of treatment-resistant depres- sion (TRD) requires an inadequate response to an no consensus as to whether patients with extensive and/or se- vere Axis I comorbidity should be categorized as having TRD. Additional research is needed to characterize the ex- tent to which TRD might be secondary to untreated comor- bid disorders such as anxiety syndromes or other Axis I dis- orders, Axis II diagnoses, or medical illness. More research is also needed to refine the TRD staging system, which, at this point, continues to be largely a pro- posed schema. Issues which need to be addressed include (1) Is response, remission, or complete recovery a more clinically useful outcome criterion? (2) Should each anti- depressant treatment have a different mechanism of action? (3) Should combined therapies or treatment augmentation be including in staging? (4) Should newer treatments (e.g., vagal nerve stimulation) be included in the staging schema? (5) How should Axis I comorbidity be handled? (6) Are all symptom response or remission criteria (e.g., using HAM-D ≤ 6) created equal? In other words, taking re- mission as an example, does the presence of single symp- toms rated as “severe” (such as insomnia or hopelessness/ suicidality) have the same prognostic significance as mul- tiple symptoms rated as “mild”? Another important question is whether Stage 4 or 5 TRD might not constitute a unique depressive subtype. 2 It would not be surprising if failure to respond to multiple adequate antidepressant trials, which typically target monoaminergic neurotransmitters, might define a unique pathophysiologic subgroup. Pharmacogenomic and functional imaging stud- ies are needed to clarify this issue. PREVALENCE OF TRD Given the lack of consensus criteria, it is perhaps not sur- prising that no agreed-upon estimates of the prevalence of TRD exist. It follows from the TRD staging schema, sum- marized above, that the prevalence of TRD is not a unitary phenomenon. Instead, different prevalence rates will be as- sociated with each TRD stage. Estimates of TRD prevalence also vary greatly depending on the treatment setting in which