Francisco A. Moreno, MD

Major depressive disorder (MDD) affects a large proportion of the world's population, but much still needs to be done to categorically improve the lives of people with this condition. Major depression is a syndrome defined by the presence of subjective symptoms and consequent suffering or dysfunction. This nonetiological categorization renders major depression a highly heterogeneous condition,

 which is reflected in the high variability of associated features, such as age of onset, symptom clustering, longitudinal course, comorbidities, relationship to life events, and psychological experiences. Most relevantly, the patterns of pharmacological treatment response are also inconsistent, and it is unclear whether differences in pathophysiology of the depressive condition or in the individual's interaction with antidepressant agents explain such variability in treatment response.

Depression and antidepressant effect: underlying mechanism

The role of neurotransmitters and neuromodulators in the pathophysiology of depression and the mechanism of action of antidepressant agents has been the subject of intensive study for many decades. Early neurobiology hypotheses, informed primarily by the pharmacological understanding of antidepressant mechanisms, postulated that depression was the result of a deficiency of monoamine neurotransmitters.

However, the evolved understanding of the dynamic functional balance of neuronal plasticity and apoptosis in response to stress,

 and the molecular events that follow agonist-receptor interactions

 led to more complicated models that explain both the pathophysiology of depression and the mechanisms underlying the effects of antidepressants.

Integration of information from imaging studies, health psychology, genetic epidemiology, and molecular psychiatry has made it clear that a full understanding of the biology of depressive disorders must incorporate knowledge of a broad range of neurochemicals, and  take into account the possible toxic effects of drugs, medical illness, stressful life events, and psychological risk features. We must also consider the role of multiple genetic influences that confer vulnerability and be able to explain how these effects converge to alter the structure and function of the brain's emotion and cognition circuits, culminating in the symptoms that define depressive disorders.

 Consequently, the treatment and recovery processes for MDD may be anticipated to be similarly variable and complicated.

Although organized psychiatry uniformly recognizes the need for treatment of major depression to remission in order to favor functional recovery and minimize risk of comorbidities, complications, and recurrence,

 the standard categorical outcome for clinical trials of antidepressants in most cases continues to be treatment response. Antidepressant response is often defined as a 50% decrease in the global score on depression rating scales.

Of course, many of the so-called responders truly fail to become categorically "undepressed." Meta-analytical studies of antidepressant clinical trials and recent large-scale collaborative studies for MDD suggest a response rate of 40% to 60% and a remission rate of 25% to 35%.13,14 An important question arises: when is an individual who remains depressed, in spite of adequate treatment, considered treatment-resistant?

Conceptualization of treatment-resistant depression as a phenotype

Many definitions of treatment-resistant depression (TRD) have been proposed. Some of them are simply categorical; others are based on numbers or types of treatments received and failed, offering a continuum of resistance.

 It is clear, however, that many people with depression are unlikely to respond to most conventional treatments when used at sufficiently high doses to anticipate therapeutic benefit. A number of questions remain:

Is it because certain patients have a pathophysiological mechanism different from those with "treatment- responding" depressions?

Do they have a more severe form of the same unique depressive condition for which the usual pharmacological "correction" is insufficient?

Do their pharmacokinetic profiles allow them to metabolize treatment agents at an exaggerated rate that renders the agent ineffective?

And, is it possible that some of these questions may be addressed with the use of genetic techniques?

Genetic basis for depression and treatment response

The field of genetics represents one of the most promising approaches to understanding the mechanisms underlying disease and behavior. Well-evidenced examples of this include genetic associations with personality traits, cognitive style, temperament, intellect, psychiatric diagnoses, and treatment outcomes. Major depression has been established as a heritable phenotype based in genetic epidemiology studies. However, similar to the examples listed above, its non-mendelian pattern of inheritance, also called "complex," exhibits effects that are difficult to predict and limits what inferences can be made about their function (ie, multiple susceptibility genes are involved, most of which have small effects and inconsistent clinical influence; these genes may also cause other disorders, and nongenetic depressions may also occur).

 The high complexity of depression genetics undoubtedly affects the patterns of illness course, severity, and drug response.

Depression and treatment resistance: role of environment

Although traditional genetic epidemiology studies use an approach that supports the role of genetic factors over environment in the determination of heritability, the interaction of gene and environment has long been observed. Recent evidence comes from genetic-association studies using candidate gene approaches. Caspi and colleagues

 reported an association between the presence of the short (

) allele of the serotonin transporter promoter gene polymorphism (5-HTTLPR) and an increased risk for depressive episodes, depressive symptoms, and suicidal ideation in early adulthood but only in those with a history of severe life stressors-a finding that has been replicated and extended.

-allele also magnifies the depressogenic effects of neuroticism

 on risk for depressive symptoms. Similarly, it has long been postulated that early life events, developmental issues, and ongoing environmental stress are important factors that affect outcome of treatment. For example, in a large collaborative study of treatment of chronic depression, Nemeroff and colleagues

 reported that patients with a history of childhood trauma experienced a decreased rate of antidepressant response to nefazodone but a greater response to treatment with cognitive-behavioral analysis system psychotherapy.

The reciprocal interplay of genetic and environmental factors takes place in the human mind and brain. Factors such as biological kindling and learned behaviors and reactions may significantly alter an individual's vulnerability to depressive recurrences, making it less closely related to genetic or external environmental events.

Genetic basis for pharmacokinetic differences

Individual variability in the effectiveness and tolerability of drugs is a well-known medical fact. For over half a century, specific enzymatic deficiencies that explain these differences have been identified. Advances in genomic medicine have provided evidence that the genes encoding for virtually all metabolizing enzymes have polymorphic variants that significantly affect clinical function. 

Figure 1 [Figure restricted. Please see print edition for content]

 provides a comprehensive summary of drug-metabolizing enzymes that exhibit gene polymorphisms with well-supported clinical consequences.

 Drug transporters such as G-glycoprotein also possess clinically relevant genetic variability that affects CNS passage through the blood- brain barrier.

 Testing for a number of pharmacokinetically relevant genes may soon become clinically meaningful.

Genetic basis for pharmacodynamic differences

Since all available antidepressants exert their effects through functional modification of monoamine neurotransmitters, common drug targets include synthetic or metabolic enzymes, neurotransmitter transporters, receptors, coupling proteins, and postsynaptic cellular factors. Genes encoding for these targets have, in many cases, clinically relevant polymorphisms. The first example of this was the association of the long (

) allele of the 5-HTTLPR with a better antidepressant response to the SSRI fluvoxamine.

 Since the appearance of this report, multiple studies using different antidepressants in various population samples have been conducted. This finding has been mostly replicated in studies in which SSRI drugs are used in patients of European descent.

Multiple studies in Asian populations have yielded largely inconsistent and contradictory results. In a recent study from Korea, depressed patients were randomly assigned to receive treatment with an SSRI (fluoxetine or sertraline) or nortriptyline, which is primarily a norepinephrine reuptake inhibitor (NRI). Genetic polymorphisms in the 2 common serotonin transporter gene polymorphisms (5-HTTLPR and intron 2 [STin2]) as well as the norepinephrine transporter gene (NET) were analyzed. Response to the SSRI was significantly associated with the 

-allele of 5-HTTLPR, in stark contrast to findings in European populations, again underscoring the ethnic/racial variability of findings not only in allele frequency distribution but also in association with clinical phenotypes.

The most striking finding of that study was that patients homozygous for the long (

) genotypeof STin2 had a 69% rate of response to SSRIscompared with only 9% for the other genotypes. Another very interesting finding was that patients carrying the GGpolymorphism of NET G1287A had a higher response rate to NRI than SSRI treatment(83% vs 59%), making this one of the first comparative reports that suggest the benefit of genotyping for treatment selection.

These findings are exciting and consistent with current antidepressant hypotheses; however, the large variability and conflicting results for the various racial groups impedes generalization of this knowledge and limits clinical application. Similar to the example described above, dozens of reports involving many pharmacodynamically related genes involving serotonin, norepinephrine, dopamine, G-proteins, neurotrophic factors, and hypothalamic-pituitary-adrenal systems, as well as circadian, immunological, and vascular systems, among others, have been analyzed. Mostly those yielding statistically significant associations have been reported, and many have subsequent inconsistent replications. (See Serretti and Olgiati

The genetic discrepancy between 2 diagnostically similar groups (although phenotypically different in antidepressant response profiles) may be detectable and provide insight into the biology of TRD. A recent pilot study from our group assessed 20 candidate genes selected for their relationship to synthesis, transport, recognition, or degradation of neurotransmitters, or their putative intracellular responses to receptor activation such as G-protein coupling, transcription, and neurotrophic factors (

Assessing for genetic association, we studied 3 subject groups. One included patients with a history of MDD who were capable of achieving remission. Another consisted of patients with current MDD who had not achieved or maintained appropriate antidepressant response after at least 2 trials of standard FDA-approved antidepressant drugs. These patients had well-documented TRD based on a clinical assessment, verification of past medical rec-ords, determination of adequacy of treatments with the standardized antidepressant history form, and had met criteria to be enrolled in a treatment study with vagus nerve stimulation. A third group of healthy volunteers who denied a personal or family history of major depression were used as genetic controls. False discovery rate methods were used to control for multiple comparisons given the large number of genes and genotypes tested.

In this largely European sample, the STin2 promoter was found to be significantly different between the patients who were depressed and controls, with the (

) genotype being overrepresented in the healthy control group. Important as well, the intron 3 polymorphism of the dopamine receptor-4 gene was significantly different between patients with TRD and treatment responders, with those homozygous for the 7 repeat allele being twice as likely to respond to an antidepressant. Interestingly, the gene explaining genetic vulnerability to depression did not explain treatment response. Furthermore, patients with TRD had a larger number of risk genotypes than treatment responders, who in turn had a greater number of risk genotypes than the healthy controls. This finding supports a model in which the additive small effects of multiple risk genes explain depression and treatment resistance (

Incorporating genetic findings with traditional therapeutic principles

When confronted with TRD a number of approaches have been suggested, and although there is limited evidence to support a specific algorithm for treatment, the following 4 principles (4 Ds of depression) are often proposed: diagnosis, dosage, duration, and drug.

Establishing an accurate diagnosis and identifying comorbidities and certain specific features that may predict a specific response currently facilitate treatment. It is possible that in the future, genetic approaches may help establish alternative phenotypes or genetically informed diagnosis with preferential response to certain treatments. For example, the 

-allele of the 5-HTTLPR has been associated with increased neuroticism, increased amygdala response to threats, and more depression and suicide outcome in the context of increased environmental stress. The same polymorphism is associated with decreased antidepressant response and greater intolerance to treatment in some populations of European origin. Although the clinical applicability of this knowledge remains limited because of the small effect conferred by a single gene, it is easy to anticipate that in the future we will be able to redefine conditions and select treatment approaches based not only on genetically informed phenotypes but also on genetic predictors of tolerability, safety, and efficacy.

What dosage will ensure the greatest likelihood of an optimal response in efficacy, tolerability, and safety? Genetic polymorphisms of drug-metabolizing enzymes are now receiving clinical attention given the commercial availability of microarrays for cytochrome P-450 genotyping. Availability, cost of genotyping, and interpretation of the findings are becoming more accessible to clinicians, although currently, establishing a patient's genetic metabolic profile is often reserved for cases of multiple drug intolerance or resistance. Understanding the role of genetics in drug metabolism, clinicians should remember that manufacturer-recommended dosages are seldom "one size fits all."

The traditional concept of optimization suggests that maximizing the dosage and duration of each trial may improve outcome, especially for patients who experience partial response. There have been a few reports of certain genetic polymorphisms associated with variable speed of onset of antidepressant response with certain drugs.

 Since we are not currently testing for all these potential factors, it may still be useful to remember that there may be a perfectly logical pharmacogenetic rationale to extend the initial trial period of an antidepressant in patients whose depression has not responded or remitted.

Selection of a drug must be based on  aspects of tolerability, safety, and efficacy. As the body of literature on antidepressant pharmacogenetics increases and more consistency is found, genetic information may also aid in the selection of drugs based on benefit and tolerability. For example, Murphy and colleagues

 reported a clinically and statistically significantly greater rate of paroxetine versus mirtazapine discontinuation in patients who were homozygous for the 

In some areas of medicine, genetics approaches already help improve prediction of individual treatment response, adverse effects, and optimization; facilitate identification of homogeneous populations; and improve understanding of mechanisms of illness and the nature of treatment response, all of which may lead to specifically designed therapies for some patients. In psychiatry, although many pertinent findings exist, the clinical application of this knowledge is still limited. TRD and the contributions from pharmacology or neurobiology that make an individual vulnerable to this phenomenon remain an enigma in spite of exciting recent developments. Large collaborative approaches such as STAR*D may help address the many questions that remain.

Diagnostic and Statistical Manual of Mental Disorders, 

4th ed. Text rev. Washington, DC: American Psychiatric Association; 2000.


 Bunney WE, Davis JM. Norepinephrine in depressive reactions. A review. 

Coppen A. The biochemistry of affective disorders. 

 Sapolsky RM. Glucocorticoids and hippocampal atrophy in neuropsychiatric disorders. 

 Hyman SE, Nestler EJ. Initiation and adaptation: a paradigm for understanding psychotropic drug action. 

 Warner-Schmidt JL, Duman RS. VEGF is an essential mediator of the neurogenic and behavioral actions of antidepressants. 

Duman RS. Depression: a Case of Neuronal Life and Death? 

Santarelli L, Saxe M, Gross C, et al. Requirement of hippocampal neurogenesis for the behavioral effects of antidepressants. 

Gould E, Tanapat P. Stress and hippocampal neurogenesis. 

 Nestler EJ, Duman RS. Relevance of intracellular signal transduction pathways to psychiatry. In: Dickstein LJ, Riba MB, Oldham JM, eds. 

American Psychiatric Press Review of Psychiatry, Volume 15. 

Washington, DC: American Psychiatric Press; 1996:279-309.


 Delgado PL, Moreno FA. Neurochemistry of Mood Disorders. In: Stein DJ, Kuper DJ, Schatzberg AF, eds. 

The American Psychiatric Publishing Textbook of Mood Disorders

. Washington, DC: American Psychiatric Publishing; 2006:101-116.


Rush AJ, Kraemer HC, Sackeim HA, et al. Report by the ACNP Task Force on Response and Remission in Major Depressive Disorder. 

 Fava M, Davidson KG. Definition and epidemiology of treatment-resistant depression

 Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report

Articles

Major depressive disorder (MDD) affects a large proportion of the world's population, but much still needs to be done to categorically improve the lives of people with this condition.

Francisco A. Moreno, MD

Articles

Treatment-Resistant Depression: The Role of Gene Factors