The Pharmacogenetics of Major Depression: Past,
Present, and Future
Gonzalo Laje and Francis J. McMahon
patients, maximizing efficacy while minimizing—or eliminat-
ing—adverse effects. This vision has been widely embraced out
of a real desire to better help patients. But we are now some
years into the pharmacogenetics enterprise, and we are begin-
ning to accumulate real-world data in juxtaposition to our
perhaps somewhat idealistic expectations. Some of the assump-
tions underlying the vision are beginning to be scrutinized more
closely. Can we really match medications with patients when the
range of available treatments is still quite small compared with
the range of clinical presentations? Can genetics really provide
the keys? How much of the difference between patients in
efficacy and adverse effects will actually be attributable to
differences in life experience and circumstances that we can
neither measure biologically nor address medically? The techno-
logical advances of human genetics over the past decade, along
with the recent availability of large treatment cohorts for study,
are beginning to answer these questions. Pharmacogenetics is
being transformed from a small field of promising leads into one
of the foundations of evidence-based medicine. Few specialties
stand to benefit as much as psychiatry, where all of our treatment
options are empirical and the matching of patients with treat-
ments is largely a question of clinical judgment.
The pursuit of genetic markers that can help predict response
to treatment or important adverse events has been the focus of
many investigations but so far has led to few clinical applications
(1). This may be changing, as specific genetic association find-
ings gain validity. Recently, the U.S. Food and Drug Administra-
tion added pretreatment genetic testing to the prescribing infor-
mation for the anticoagulant drug warfarin, which has a narrow
therapeutic index. About 40% of the variance in warfarin dose is
attributable to variation in cytochrome P450 and vitamin K
metabolism genes (2).
Major depression, which is predicted to be the second leading
cause of death and disability by the year 2020 (3), is an ideal
target for pharmacogenetic approaches (4). Although many
treatments exist, initial response rates are stuck at about 50%, and
only a minority of patients fully remit (5). Choosing the best
treatment for each patient is difficult, and clinical predictors of
differential response are scarce (6). Although adverse events are
uncommon in the age of selective serotonin reuptake inhibitors
(SSRIs), concern about some adverse events, such as treatment-
emergent suicidality, may reduce access to, and acceptability of,
treatment (7). In the face of all this, good genetic markers of
treatment outcome in major depression might have immediate
uch has been said and written in recent years about the
promise of pharmacogenetics. The vision is one of a
world where we will be able to match medications with
clinical relevance—yet in many ways, the search for such
markers has barely begun.
Here, we address some of the past history, ongoing research,
and near future of pharmacogenetic studies of major depression.
Rather than attempt a comprehensive review of the literature,
which has already been done well in some recent papers (4,8),
we aim to highlight some of the key issues facing the field and
sketch a vision for the future. Throughout, we use the term
pharmacogenetic rather than the grander pharmacogenomic but
consider both terms synonymous.
The Past: Small Samples and Few Genes
A MEDLINE search using the terms pharmacogenomics or
pharmacogenetics and depression in humans yields about 100
articles, almost half of which are labeled as reviews. This would
suggest that since 1968, the year of the first cited article, much
has been written but not as much new information has been
acquired. This characterization is not confined to the pharmaco-
genetics of major depression; it applies almost as well to the
pharmacogenetics of other fields. In retrospect, without many of
the necessary genetic tools and large samples to study, rapid
progress could hardly have been expected.
Faced with these limitations, past approaches have been
geared toward a limited number of markers in a small fraction of
the ?20,000 human genes. These include genes in the cyto-
chrome P450 system, a major pathway of drug metabolism, and
genes involved in the production, release, binding, and reuptake
of monoamines, especially the serotonin transporter (SERT).
Only the P450 genes have found a clinical application and this
has been so far quite limited (1), probably owing to the decline
in the use of tricyclic antidepressants, the information that comes
directly from a careful medication history, and the clinically small
effects of the P450 system on the most widely prescribed SSRIs.
The serotonin transporter and a functional polymorphism in
its promoter region (known as the linked polymorphic region
[LPR]) have received the most attention by researchers. The
serotonin transporter is undoubtedly the proximal target for the
SSRIs and the LPR seems to affect gene expression in important
ways (9), but consistent association with SSRI response has been
difficult to demonstrate. A recent meta-analysis of 15 published
studies concluded that there was evidence in favor of a significant
association of the L-allele with better response to SSRIs (10). This
result may reflect publication bias—since there is a lack of even
small studies that report an effect size below the pooled odds
ratio—but may be consistent with a true effect of the SERT LPR on
treatment outcome. Recent studies, reviewed below, suggest that
genetic variation in the SERT LPR has an impact on SSRI-related side
effects. This might explain some of the outcome findings.
The Present: Picking Up Speed
There have been two big developments in recent years that
have greatly accelerated progress in the field: the availability of
large, well-characterized samples and the advent of genomic
technologies of unprecedented power and efficiency. The Se-
quenced Treatment Alternatives to Relieve Depression (STAR*D)
From the Genetic Basis of Mood and Anxiety Disorders Unit, Mood and
Anxiety Program, National Institute of Mental Health, Bethesda, Mary-
Address reprint requests to Francis J. McMahon, M.D., 35 Convent Drive,
Room 1A202, Bethesda, MD 20892-3719; E-mail: firstname.lastname@example.org.
Received September 18, 2007; accepted September 19, 2007.
BIOL PSYCHIATRY 2007;62:1205–1207
© 2007 Society of Biological Psychiatry
study (11) has for the first time provided a sample of well-
characterized patients large enough to detect even modest
genetic effects. Although not designed specifically to answer
pharmacogenetic questions, STAR*D provides DNA from a clin-
ically representative cohort of about 2000 adults with major
depression, all treated with citalopram for at least 6 weeks and
evaluated prospectively for treatment response, remission, and
adverse events. Patients who failed to respond adequately to the
initial treatment were switched to a variety of other treatments.
The large size of the uniformly treated sample is crucial, since it
represents an order-of-magnitude increase over all previously
studied samples with major depression. Large samples are a key
element of robust genetic association findings (12).
Candidate gene association studies in the STAR*D sample
have already begun to uncover alleles associated with citalopram
response and adverse events such as treatment-emergent suicidal
ideation (13–17). The published effect sizes for the response-
associated alleles are modest, with odds ratios on the order of 1.5
to 1.7. The effect sizes for the adverse event associations appear
to be larger, but the true effect sizes cannot, of course, be
accurately estimated in a single sample. In any case, effect sizes
will need to be much larger if genetic markers are to help guide
treatment decisions. Such effect sizes might be achievable with
multimarker, multigene tests, but this remains to be seen.
The results in the large STAR*D sample have also begun to
clarify some of the questions concerning the role of the SERT
LPR: while there is no convincing evidence for association with
treatment response in the STAR*D sample (18), the data suggest
that a novel functional allele related to the traditional S-allele is
associated with perceived side-effect burden (13). This result
might have been mistaken for an association with treatment
response if side effects had not been considered, since patients
with a high side-effect burden may not be able to take a sufficient
dose of medication to achieve a good response.
The second big development has been the new technologies
that allow the screening of genome-wide sets of genetic markers.
These technologies have opened the door for genome-wide
association (GWA) studies that use large numbers of single
nucleotide polymorphism (SNP) markers (now 500,000 to
1,000,000) to screen the entire genome for alleles that influence
a trait of interest. For pharmacogenetics, this is a crucial advance,
since it transcends candidate gene studies, which are limited by
the genes chosen, and supplants linkage methods, which require
families and are impractical for most pharmacogenetic questions.
At least two GWA studies of antidepressant treatment
outcome are currently underway. Each has interrogated over
300,000 markers. The results of these studies will provide the
most comprehensive view to date of genetic influences on
antidepressant response. However, GWA studies are usually
underpowered to detect alleles of small effect. If antidepressant
treatment outcome is influenced by many genes, each of small
effect, which seems likely, then GWA studies may miss many of
them. Alternatively, such genes may be implicated by markers
with modest effects (and statistical significance), the importance
of which may only be recognized after several replications and
the demonstration of functional alleles. Additional large samples
of uniformly treated and longitudinally assessed patients will
thus be essential for validating the results of the GWA studies
The Future: Knocking at the Clinic Door
Since the recent publication of the first Encyclopedia of DNA
Elements (ENCODE) data (19), it has become clear that protein-
coding genes are only a very small part of the genome and that
most human DNA has significant biological activity: there may be
no “junk” DNA. Until we fully understand the mechanisms that
regulate gene function, we should maximize the results we
obtain from available samples. Near-future technology can add
significant amounts of information to what we already have. In
addition to ever denser genome-wide sets of SNP markers, such
as the million SNP chips that are already coming into use,
ultra-high throughput sequencing technology may make the
complete genome sequence a standard part of each patient’s
medical record. This will pose an enormous challenge to clini-
cians, who will need to interpret this large quantity of data for
patients and their families.
It is also likely that single nucleotide polymorphisms are not
the sole genetic regulators of antidepressant response. We will
need to interrogate other types of variation such as copy number
polymorphisms and epigenetic signatures. Accumulating evi-
dence demonstrates that copy number polymorphism is a com-
mon source of human genetic variation that may have major
effects on gene function (20). Epigenetic variation—heritable
changes in gene expression that are influenced by parent-of-
origin and environmental factors—is another previously unrec-
ognized but important source of human genetic variation. Early
findings suggest that antidepressant medications affect epige-
netic signatures, and agents that modify epigenetic signatures
can exert antidepressant effects (21).
In addition to novel genetic investigations, the future of
pharmacogenetics in mood disorders will depend heavily on the
availability of large, well-characterized samples. Along with
frequent, prospective assessment during treatment, these sam-
ples should ideally provide a detailed clinical and diagnostic
picture that reflects both current and prior episodes, along with
medication history, ethnic background, and psychosocial history.
The latter may be an important source of individual differences in
treatment outcome owing to temperament, resilience, and the
cumulative impact of adverse life events (22). Although direct
information on in vivo gene expression in the brain may be
beyond the reach of near-future technology, neuroimaging tech-
niques such as positron-emission tomography (PET) or magnetic
resonance spectroscopy (MRS) may provide valuable data on the
impact of antidepressant medications on brain chemistry in vivo.
In summary, the next decade should bring significant
progress for pharmacogenetics in psychiatry. This progress will
require a significant effort in sample collection, genetic assays,
and clinical validation. For the promise of clinically useful tools
to be realized, the field must produce clinically meaningful
findings and not just statistically significant associations. If we
succeed, the payoff will be big: more personalized care, more
effective medications, fewer adverse events, and a reduction in
the burden of mood disorders for both patients and society.
Supported by the Intramural Research Program of National
Institute of Mental Health (NIMH), National Institutes of Health
(NIH), U.S. Department of Health and Human Services (DHHS),
and by the National Alliance for Research on Schizophrenia and
Depression (NARSAD). The content of this publication does not
necessarily reflect the views or policies of the Department of
Health and Human Services, nor does mention of trade names,
commercial products, or organizations imply endorsement by
the U.S. Government.
The authors report no competing interests.
1206 BIOL PSYCHIATRY 2007;62:1205–1207
1. Swen JJ, Huizinga TW, Gelderblom H, de Vries EG, Assendelft WJ, Kirch- Download full-text
heiner J, et al. (2007): Translating pharmacogenomics: Challenges on
the road to the clinic. PLoS Med 4:e209.
Common VKORC1 and GGCX polymorphisms associated with warfarin
dose. Pharmacogenomics J 5:262–270.
3. Murray CJ, Lopez AD (1996): Evidence-based health policy—lessons
from the Global Burden of Disease Study. Science 274:740–743.
4. Perlis RH (2007): Pharmacogenetic studies of antidepressant response:
How far from the clinic? Psychiatr Clin North Am 30:125–138.
measurement-based care in STAR*D: Implications for clinical practice.
6. Nierenberg AA (2003): Predictors of response to antidepressants: General
7. Gibbons RD, Brown CH, Hur K, Marcus SM, Bhaumik DK, Erkens JA, et al.
(2007): Early evidence on the effects of regulators’ suicidality warnings
on SSRI prescriptions and suicide in children and adolescents. Am J
8. Binder EB, Holsboer F (2006): Pharmacogenomics and antidepressant
drugs. Ann Med 38:82–94.
nin transporter gene regulatory region. Science 274:1527–1531.
transporter gene promoter polymorphism (5-HTTLPR) association with
selective serotonin reuptake inhibitor efficacy in depressed patients.
(2004): Sequenced treatment alternatives to relieve depression
(STAR*D): Rationale and design. Control Clin Trials 25:119–142.
Genetic associations in large versus small studies: An empirical assess-
ment. Lancet 361:567–571.
(2007): Association between a functional serotonin transporter pro-
moter polymorphism and citalopram treatment in adult outpatients
with major depression. Arch Gen Psychiatry 64:783–792.
14. Laje G, Buervenich S, Manji H, Rush AJ, Wilson AF, McMahon FJ (2007):
Genetic markers of suicidal ideation emerging during antidepressant
(2006): Variation in the gene encoding the serotonin 2A receptor is
Association of GRIK4 with outcome of antidepressant treatment in the
pram and polymorphisms near cyclic adenosine monophosphate re-
sponse element binding protein in the STAR*D study. Arch Gen Psychi-
18. Kraft JB, Peters EJ, Slager SL, Jenkins GD, Reinalda MS, McGrath PJ,et al.
(2007): Analysis of association between the serotonin transporter and
antidepressant response in a large clinical sample. Biol Psychiatry 61:
19. Birney E, Stamatoyannopoulos JA, Dutta A, Guigo R, Gingeras TR, Mar-
in 1% of the human genome by the ENCODE pilot project. Nature
ity. Biol Psychiatry 62:1–3.
22. Nemeroff CB, Vale WW (2005): The neurobiology of depression: In-
roads to treatment and new drug discovery. J Clin Psychiatry
BIOL PSYCHIATRY 2007;62:1205–1207 1207