AUTHOR
Leigh Fylstra
Augmentation of
antidepressants
Chaiman’s Rounds, September 17, 2007
Reference: Papakostas GI. Shelton RC. Smith J. Fava M.
Augmentation of antidepressants with atypical antipsychotic
medications for treatment-resistant major depressive disorder: a
meta-analysis. Journal of
Clinical Psychiatry. 68(6):826-31, 2007 Jun.
Clinical
Question: According to the article, atypical antipsychotics
(AAP) are used as off label adjuvant treatment for
treatment-resistant depression (TRD), and only recently
double-blind RCTs have assessed the efficacy and tolerability of
augmentation with atypicals. The goal of the study is to
evaluate he RCTs by systematic review and meta-analysis in an
effort to determine whether the use of AAP is efficacious and
tolerable as adjuvant treatment in patients with TRD.
Description of intervention: addition of atypical
antipsychotic (Olanzapine, Risperidone, Quetiapine) or placebo
to patients with TRD who were already on a “standard”
anti-depressant: Fluoxetine, an “SSRI/SNRI,” or “various” other
anti-depressants.
Validity
Criteria for Systematic Review/Meta-Analysis:
-
Review
did explicitly address a sensible clinical question which
was focused and narrow in scope. However, criteria for
treatment resistant depression was not defined, as well as
doses or length of treatment of anti-depressants prior to
“resistance.”
-
Search
for relevant studies was detailed and exhaustive. EMBASE,
Cochrane database, articles presented at scientific
meetings, clinical trial results registries or direct
contact with makers of atypicals, and unpublished studies
were searched. However, ultimately only 3 peer-reviewed
studies from MEDLINE/pubmed and 7 from reports at scientific
meetings met criteria.
-
No
studies appeared to be omitted due to illegitimate reasons
-
Methodological quality of the primary studies is unknown.
No info on the evaluation of validity was discussed.
Therefore, we cannot determine the quality of the studies
used, which is a big limitation. Studies used may have been
pharma sponsored and 7 of the 10 were unpublished studies
not peer-reviewed
-
Author’s
assessments of the primary studies appeared reproducible, as
they used a pre-coded form to extract data, which is
standard and good practice. However, the number of raters
was not discussed, so the assumption is only one rater was
used, which gives more chance or random errors or systematic
bias.
-
Study
results were homogenous. Fig 1: CI’s consistently overlap
and do not differ significantly. No huge outliers in either
direction exists, so pooling the data is appropriate. Tests
of results showed no heterogeneity, but this study only
tested outcome measures: response & remission rates. Thus,
the results may be homogenous between the studies, but the
actual components of the studies may differ. No details
about differences are given concerning patients,
methodology, or interventions (doses of SSRI/SNRI or
Atypicals)
Relevant criteria for Systematic Review/Meta-analysis:
-
Data
Sources: 3 peer-reviewed studies and 7 articles from
scientific meetings
-
Study
Selection Criteria: Double-blind, randomized
placebo-controlled trials of atypical augmentation of pt’s
with TRD
-
Patient
population: Inclusion: Treatment-resistant major depressive
disorder in the acute phase only who were on a standard
anti-depressant. Exclusion: psychotic symptoms, dysthymic
disorder, SAD, substance abuse, bipolar disorder, Depression
d/t medical condition
-
Main
outcome measures: determined a priori
-
Primary: HAM-D or MADRAS scores (response rate = 50%
reduction, remission was different: HAM-D <8, MADRAS
<9-11
-
Secondary: overall discontinuation rate,
discontinuation d/t inefficacy, and discontinuation d/t
adverse effects
Main outcome results:
-
Primary:
-
Pooled Remission rates: 47.4% (atypicals) vs 22.3%
(placebo), RR 1.75 (95% CI = 1.36 to 2.24, p < .0001)
-
Pooled response rates: 57.2% (atypicals) vs 35.4%
(placebo), RR 1.35 (95% CI = 1.13 to 1.63, p = .001)
-
Secondary:
-
discontinuation overall: no difference
-
d/t
inefficacy: no difference
-
d/t
AE: lower among placebo treated (RR = 3.38, 95% CI =
1.98 to 5.76, p < .0001)
Conclusions: This systematic review/ meta-analysis
evaluates the efficacy and tolerability of augmentation with
atypicals to standard antidepressant medication for
treatment-resistant depression. According to the authors, the
adjunctive treatment appears to have potential efficacy but
limited tolerability. Results showed greater remission and
response with atypicals vs placebo, but also revealed a
three-fold greater discontinuation rate due to side effects with
atypicals (sedation, EPS, TD, weight gain/metabolic syndrome,
hyperprolactinemia); thus, the authors do not explicitly
recommend the strategy of atypical augmentation.
Multiple limitations of this review should raise caution when
drawing definitive conclusions from the results. The major
limitation is lack of analysis of methodology or validity
criteria of studies used; 7 of the 10 studies were from
scientific meetings, which were unpublished and likely not
peer-reviewed. Another problem is the lack of detail regarding
the specific SSRI/SNRIs used, doses, or length of treatment
prior to determining treatment resistance. In fact, specific
criteria for TRD was not explicitly stated. Doses of atypicals
also remained undefined. The definition of response was similar
in all studies, but some differed in definition of remission:
table 1 shows some studies considering remission as MADRAS <9,
while others defined remission as MADRAS <11. Additionally, the
longest study was only 12 weeks, so no data exists for further
duration of effects.
The
researchers acknowledged several other limitations. One cannot
generalize the results to all atypicals, because only olanzapine,
risperidone, and quetiapine were studied. No information is
known about aripiprazole or ziprasidone. Furthermore, no
information exists comparing atypical augmentation with
currently proven adjuvant treatments (Bupropion, Mirtazapine,
Mianserin, Buspirone, Liothyronine, S-adenosylmethionine, folic
acid, modafinil, lithium, or CBT).
Synopsis:
Despite the many limitation of this systematic review of 10 RCTs
evaluating AAP augmentation to TRD, it does appear that
atypicals may have efficacy as adjuvant treatment to SSRI/SNRIs
but tolerability is limited; therefore, definitive conclusions
of efficacy cannot be inferred form this study, and further
research is warranted to compare atypicals to current adjuvant
treatments.
Return
to top of the page
AUTHOR
Jennifer Holton
Chairman’s
Rounds, September 10, 2007
Citation:
Frye MA, Heinz G, Suppes T, McElroy SL, Keck PE, et al. “A
Placebo-Controlled Evaluation of Adjunctive Modafinil in the
Treatment of Bipolar Depression.” American Journal of
Psychiatry. 2007; 164: 1242-1249.
Clinical Question:
Is adjunctive Modafinil efficacious and safe in the acute
management of treatment-resistant bipolar depression?
Background
Info for Modafinil:
FDA-approved for improving wakefulness in sleep disorders.
Efficacious in the treatment of ADHD. May decrease cocaine use
in cocaine-dependent patients (w/ CBT). Primary outcome
measures have generally been negative in studies of unipolar
depression in which modafinil is used for antidepressant
augmentation/acceleration, but secondary measures of clinical
global improvement and rapid reduction in fatigue/sleepiness
have suggested some benefit.
Methods:
Design:
Double-blind,
randomized, 6 week, placebo-controlled trial
Setting:
Multi-site study; academic settings and community mental health
centers (4 sites – 3 in the US, 1 in Germany)
Patients/Population:
Enrollment:
120 screened; 30 excluded; 90 randomized; 85 analyzed (N=41 for
modafinil group, N=44 for placebo group)
Inclusion
Criteria:
18-65 yr olds;
Bipolar I or II depression (dx by SCID – kappa of 0.92);
Moderate symptom severity (IDS > 16); Inadequately responsive to
a mood stabilizer +/– concomitant antidepressant therapy
Exclusion
Criteria:
Concurrent use
of nefazodone or MAOI; Active suicidality; Active psychosis;
Current substance abuse/dependence; Unstable general medical
condition; Clinically relevant baseline lab abnormality; ECG
evidence of ischemia or ventricular hypertrophy; History of
stimulant-induced mania; Baseline sleep pattern of <6 hrs/night
Treatment
Protocol:
All
medications at stable doses for at least 2 wks prior to
randomization; held constant during the trial. Patients
randomized to Modafinil or Placebo at 100mg daily x 1 week, then
100mg BID for wks 2-6; in cases of early clinical response or
adverse effects, the dose could be reduced. Evaluations
performed at baseline and then weekly for the 6-week trial.
Outcomes:
Primary:
Baseline-to-endpoint change in IDS Score (30-item clinical
rating scale rating the severity of depression on 0-90 scale).
Secondary:
Clinical
response (50% reduction in IDS score at end of 6-week trial);
Remission (final IDS score <12); Baseline-to-endpoint change on
a subset of four IDS questions assessing fatigue and
energy-related symptoms; Treatment-emergent mania/hypomania (YMRS
>13); Mood destabilization (CGI-BP improvement score of 6 or 7);
Fatigue/Sleepiness (Fatigue Severity Scale; Epworth Sleepiness
Scale)
Therapy
Validity Criteria:
Randomization/Blinding:
Double-blind randomization
Intention to
treat:
Patients were
analyzed in the groups to which they were randomized as long as
they received at least 1 medication dose and had at least one
set of ratings after randomization. As such, data for 85 of the
90 randomized patients were analyzed.
Similar
groups:
Patients in
the treatment and control groups were similar at baseline with
the exception that more patients in the modafinil group were
receiving sedative-hypnotics such as clonazepam, lorazepam, and
zolpidem. There were also more patients with Bipolar I
diagnoses in the modafinil group, but this was not statistically
significant.
Equal treatment:
Aside from the intervention, groups were treated equally in that
all patients were on a mood stabilizer and were allowed to
receive ongoing adjunctive antidepressant therapy, but the mood
stabilizers and antidepressants were not standardized in terms
of dosing, class/type of medication, or duration of treatment.
Follow-up:
5 patients were lost to follow-up and not included in the analysis
(3 did not meet inclusion criteria but inadvertently received a
randomization number; 2 dropped out after the baseline
randomization but before the first treatment visit)
Main
Results:
Primary
Outcome:
Baseline-to-endpoint change in IDS Score was approximately 11 in
the Modafinil group and approximately 6 in the placebo group
(p=0.04; medium effect size=0.47). Concomitant antidepressant
therapy did not contribute to the difference in outcome between
groups.
Secondary
Outcomes:
Clinical
response (43.9% vs 22.7%, χ2=4.31, p=0.038) and
remission rates (39% vs 18%, χ2=4.55, p=0.033) were
significantly higher in the modafinil group than in the placebo
group. Baseline-to-endpoint change on the four-item
fatigue-and-energy subset of the IDS (p=0.01; medium effect
size=0.56) and the depression severity score from the CGI-BP
(p=0.005, medium effect size=0.63) were significantly reduced in
the modafinil group compared with the placebo group. No
significant differences between groups at endpoint in YMRS
score, treatment-emergent mania or hypomania, Epworth Sleepiness
Scale score, or Fatigue Severity Scale score.
Important Adverse Events:
1 patient in each group required hospitalization for mania; 1
patient in the modafinil group required hospitalization for
depression; 1 depression exacerbation in the placebo group;
minimal medication side effects
Comments:
Strengths of the study: well-designed; included community
mental health centers and academic centers; used a drug that had
not been studied before for this particular use. Weaknesses
of the study: unclear as to what ‘inadequate response’ to
mood stabilizer meant in this study; mood stabilizers and
adjunctive antidepressants were not standardized in terms of
dosing, type of medication, or duration of treatment; was only a
six-week study so did not address maintenance therapy; small N;
exclusion criteria included suicidality, active substance abuse,
active psychosis, and history of stimulant-induced mania; IDS
score as primary outcome (meaningfulness of outcomes; more
sensitive than similar scales – may have more false positives,
may exaggerate benefits of modafinil); lack of confirmatory
studies.
Bottom
Line:
This 6-week
double-blind RCT of 85 patients with Bipolar Depression
inadequately responsive to a mood stabilizer +/–concomitant
antidepressant therapy suggests that adjunctive modafinil at
doses of 100-200mg daily may improve depressive symptoms (medium
effect size) without contributing to mood destabilization and
with few adverse effects. The study size was small, however,
and the approximately 3 psychotropic medications that patients
were receiving were not standardized. Since modafinil showed
comparable efficacy to other available treatments that have been
well-studied, it may be prudent to await confirmatory studies
before starting patients on adjunctive modafinil.
Return
to top of the page
AUTHOR
Monica Slubicki, MD
Chairman’s Rounds, August 27,
2007
Citation:
Oberlander et al. Externalizing and Attentional Behaviors in
Children of Depressed Mothers Treated With a Selective Serotonin
Reuptake Inhibitor Antidepressant During Pregnancy. Arch Pediatr
Adolesc Med. 2007;161:22-29.
Clinical Question:
Is
prenatal SSRI exposure associated with increased externalizing
behaviors, increased activity level, and increased aggressive
behaviors in 4 to 5 year old children?
Terms:
PNA: Poor neonatal adaptation is a syndrome including respiratory distress,
irritability, jitteriness, and increased rate of admission to
special care nursery after birth. Mechanism not yet understood --
the symptoms may result either from SSRI withdrawal or a
type of serotoninergic syndrome.
Externalizing Behaviors:
A
grouping of behavior problems that are manifested in children’s
outward behavior and reflect the child negatively acting on the
external environment. In the research literature, these
externalizing disorders consist of disruptive, hyperactive, and
aggressive behaviors.
Background:
Maternal mental illness has been well established to be a risk
factor to the fetus and developing child and has been shown to
affect childhood attention, emotion, and arousal regulatory
disorders. The long lasting effects of maternal use of SSRI’s in
pregnancy remains under investigation. One study by these authors
found altered pain reactivity in neonates and increasingly difficult
temperament at age 3 months among infants exposed to SSRIs and
benzodiazepine. The authors hypothesized that these outcomes reflect
altered serotonin-mediated processes, and that these may persist and
may be expressed in the areas of attention, activity, impulsiveness,
defiance, compliance, problem solving, task persistence, and control
issues.
Study Design Type:
Prospective Cohort Study
Setting:
Reproductive Mental Health Program at British Columbia Women’s
Hospital, Vancouver. Academic Center?
Population/Patients: 46 SSRI and 24 control originally w/ 22 SSRI and 14 control in
follow-up. Originally, SSRI cases were mothers treated for mood
and/or anxiety disorders taking paroxetine, sertraline, fluoxetine
during pregnancy. Controls were non psychotropic using mothers with
no history of mental illness. Mother average age 32. In follow-up,
child age 4-5, over 50% in pre-school. 18% of SSRI and 50% control
group in daycare. Equal percentage (7-9%) reporting behavioral
diagnosis. Poor description of inclusion and exclusion criteria.
Exposures:
The SSRI group was exposed to either fluoxetine, paroxetine, or
sertraline in pregnancy. Some were also exposed to clonazepam. The
duration of exposure was variable. Dosing was variable (see Table
2). Previous article said majority were also breast fed but did not
give details. As part of evaluation, children were observed in
one-way mirror and videotaped. These results were coded by one
psychiatrist, blinded, and reliability was checked.
Outcomes:
Outcomes were determined a priori and were: (a) externalizing
behaviors of attention and aggression (questionnaires), (b) activity
and aggression levels (clinician observed). Previously collected
during labor and delivery were umbilical cord blood samples for
antidepressant levels.
Follow-up Period: Follow-up period was approximately 4+ years from prenatal
exposure (length of exposure was variable during the pregnancies)
Question Type: HARM
Validity Criteria:
“SSRI” and “control” groups differed also in maternal diagnosis of
mental illness. The outcomes were measured the same way in both
groups. The person doing the evaluation of childhood behavior was
blinded and was one single psychiatrist. Follow-up was not very
complete, with the high rate of drop-out, but demographic
characteristics of drop outs were similar to those staying in the
study. There was an extended period of time between exposure and
outcome. Dose-response gradients were not investigated, although
duration of exposure was investigated.
Main Results:
Regression Model:
To predict parental report of externalizing behavior: SSRI
exposure, maternal mood (significant), prenatal clonazepam
exposure, a history of PNA, and umbilical cord drug levels
Regression Model:
To predict laboratory-observed aggressive behavior: PSI
(stress) score, PNA. [overall model significant, neither
independently significant]
·
Externalizing behaviors did not differ between SSRI-exposed and
non-exposed groups
o
Relative Risk for score >60 of externalizing behavior in SSRI group
was 2.2
·
Current maternal mood symptoms were associated with increased
externalizing behaviors by maternal (not teacher) report
·
Maternal mood symptoms 2 months postpartum did not predict child
behavior or activity at 4 years of age
·
The
persistence score for child behavior was lower in the exposed group
·
Increased externalizing behaviors were associated with increased
SSRI umbilical cord drug levels, but levels only accounted for 11.2%
of behavioral outcomes when controlling for maternal depressed mood
·
A
history of PNA was associated with increased aggressiveness,
although neither this nor increased parental stress was a unique
(significant) predictor of this behavior
Conclusions:
Criticisms:
Control: should have been non medicated depressed cohort
followed more closely over time.
Better documentation on characteristics of the initial SSRI
and control subjects
Investigation into dose-dependent effects of SSRI’s on the
prenatal and developing brain
Behavioral data collection done over time with these children
and in many different contexts.
Improve follow up (only 50%)
Questions:
Who were the drop outs and could they have affected the results
discriminately?
Are there other factors that should have been added to the
regression models? (day care)
The SSRI treated mothers in this study could have suffered
from a type of bias, since they were the ones to identify their
children as having externalizing behaviors. The results could have
reflected the mother’s opinions (or a recall bias) rather than the
true behavior of the child.
Overall, the study
could not find an association between prenatal SSRI treatment and
prenatal depressed mood and externalizing behaviors in 4 year old
children. They were uncertain about the contribution of PNA in
predicting these behaviors, but did have a statistically significant
model.
Synopsis:
In this prospective, 4 year follow-up study of 70 children there was
no statistically significant difference in externalizing behaviors
between children exposed to SSRIs prenatally and those who were not
exposed. Current maternal mood was associated with increased
externalizing behavior. Increased aggressiveness was associated with
a history of poor neonatal adaptation, although this was not
statistically significant. The study’s control group differed from
the case group both in terms of exposure to SSRIs as well as other
important characteristics, such as state of maternal mental health,
that could have affected study results. Children could have been
evaluated more comprehensively over a longer time period.
Return
to top of the page
AUTHOR
Gregory Weiss
Childhood Sibling Relationships as a Predictor of Major Depression
in Adulthood
Chairman’s
Rounds, June 11, 2007
Clinical Question:
Are
childhood sibling relationships important predictors of depression
later in life?
Reference:
Waldinger,
RJ, Vaillant, GE, and Orav, EJ. 2007. Childhood Sibling
Relationships as a Predictor of Major Depression in Adulthood: A
30-year Prospective Study. AJP 164:6, June 2007.
Background:
Parental
relationships have long been thought to be the most important
emotional relationships in a child’s early years. Children typically
experience intense relationships with siblings, making them among
the most emotionally salient relationships in a child’s development.
While childhood adversity of various kinds, notably neglect/poor
care due to illness, death or divorce, has been implicated in later
life depression, the strength and quality of relationships in early
childhood has never been studied in a systematic, longitudinal way.
Methods:
Design/Setting -
Prospective, cohort study of 268 college aged white men who were
followed for 30 years.
Inclusion criteria –
college
sophomores between 1939 and 1942, perceived by their deans as
“likely to become successful adults.”
Exclusion criteria –no
physical or mental health problems
Population –
268 selected, 12
dropped out of study before finishing college, 8 killed in WWII, 19
dropped because their outcome data for depression was inconclusive
Protocol
– On entering the study, assessed by internists, psychiatrists,
psychologists and anthropologists. Completed questionnaires every 2
years up until the present. Interviewed by study staff at ages 25,
30 and 50. Independent raters assessed quality of relationship with
siblings and parents and rated each a 0, 1, or 2. Ratings added
together for a 1-5 scale for each relationship. An extensive social
history was taken and death of a parent in childhood or a family
history of depression was given a “yes” or “no” rating.
Outcomes -
All data were
assessed by a clinical research psychiatrist and rated for presence
or absence of depression. Criteria used were any 3 of the following:
1) self-report of 2+ weeks of depression, 2) clinical diagnosis of
depression given by a non-study clinician, 3) received
antidepressant medication, 4) anergia or decreased concentration for
2+ weeks, 5) neurovegetative signs or depression, 6) attempted or
completed suicide, 7) sustained anhedonia or, 8) psychiatric
hospitalization for reasons other than alcohol abuse.
Analysis -
Chi-square trend tests to determine relationship of
depression to variable, t-tests to test for confounders due to mean
birth order and mean number of siblings. Binary regression with
terms chosen that were significant in the univariate analysis (with
quality of parenting added).
Validity:
1) Comparison groups similar other than exposure of interest? –
yes. All upper class educated white men enrolled in a university
and chosen based on good physical and mental health..
2) Exposures and outcomes measured similarly in both groups? –
yes.
3) Follow-up sufficiently long and complete? yes. Only
20 drop-outs after 30 years. Some missing data on drugs and alcohol
(see Table 2).
4)
Temporal relationship corrent? – yes. Exposure preceded
outcome.
5) Groups treated equally ? – yes.
6)
Dose-response gradient? – yes. Relationships grouped into 3
levels with outcome checked for all 3.
Results:
(See
Table 2)
-Poorer relationships with siblings prior to age 20 and a family
history of depression independently predicted both the occurrence of
major depression and the frequency of drug use by age 50.
-Poor
relationships with parents did not predict the occurrence of
depression by age 50.
-Quality of sibling
relationships and family history did not predict alcohol abuse or
dependence.
1)
“…consider two men, one with and the other without a family
history of depression, and both with average-quality relationships
with parents and siblings, the one without a family history of
depression would have a 3.9% chance of developing depression,
whereas the one with a family history of depression would have a
13.5% chance of developing depression.”
2)
“consider two men, one with relatively good relationships
with his siblings and the other with relatinvely poor relationships
with his siblings, and both with average-quality relationships with
parents and no family history of depression, the one with better
sibling relationships would have a 2.3% chance of developing
depression , whereas the one with poor sibling relationships would
have a 9.9% chance of developing depression.”
3)
“Finally, if we consider a man who had parenting of average
quality but relatively poor relationships with his siblings and a
family history of depression, his risk of depression is 30%”
Comments:
Strengths – 30
years of longitudinal follow-up. Homogeneous groups. Impressive
follow-up.
Weaknesses-
Did not use categorical definition of depression, 19 people not
included in analysis, difficult to accurately stratify exposure in
meaningful way, Very homogeneous study population helps validity but
compromises generalizability (i.e., maybe parenting was generally
good in the entire population which prevented there from being a
difference in depression due to this factor?), why did the data only
come out now when it should have been “complete” in the mid-70s?
CLINICAL BOTTOM
LINE:
In this
prospective, longitudinal study of childhood influences on
depression in later life in healthy, upper-class white men, family
history and relationships with siblings in childhood were
independently predictive of depression while quality of
relationships with parents during childhood was not. While such a
study can not show causality, it certainly suggests that sibling
relationships are more important in developing depression later in
life than previously thought.
Return
to top of the page
AUTHOR
Gregory Weiss
Relationship between CAD, cardiac symptoms and depression in women
Chairman’s Rounds, June 25, 2007
Clinical Question:
What is the
relationship between CAD, cardiac symptoms and depression in women
who are referred for a coronary angiogram?
Reference:
Depression is Associated with Cardiac Symptoms, Mortality Risk, and
Hospitalization Among Women with Suspected Coronary Disease: The
NHLBI-Sponsored WISE study. Rutledge, T et al. Psychosomatic
Medicine 68:217-223, 2006.
Background:
The
connection between depression and cardiovascular events is well
established. While most of the early research was focused on men,
there is a growing body of literature that suggests this same
relationship holds for women. The majority of depression research on
cardiac patients was done on patients with severe atherosclerosis,
leaving patients with milder disease relatively unstudied. With
women about twice as likely to have depression than men,
understanding the relationship between CAD and depression in women
would seem to be of vital importance.
Methods:
Design/Setting -
Prospective, cohort study of 988 women referred for coronary
angiography.
Inclusion criteria –
Women older than
18 years of age referred for coronary angiography.
Exclusion criteria –pregnancy,
cardiomyopathy, recent MI, hx of CABG or PTCA, unable to complete
questionnaire, history of congenital heart disease.
Population –
988 agreed to
participate, 750 complete data on depression, 505 completed the BDI.
34 patients lost to follow-up, none of whom submitted depression
data.
Protocol
– Estimate CAD by maximal stenosis on angiography, some got
non-invasive cardiologic tests, questionnaires assessing symptoms at
baseline, telephone interview at 6 weeks and then yearly, Beck
Depression Inventory and question about hx of treatment for
depression at baseline.
Outcomes -
mortality,
hospitalization, cardiac symptoms.
Analysis -
T-tests for cardiac symptoms in different cohorts,
Cox regression models to evaluate age and CAD severity-adjusted
depression relationships and effects after controlling for CAD risk
factors.
Validity:
1) Comparison groups similar other than exposure of interest? –
no. Multiple confounders including SES, smoking, education, age,
social connections. Tried to control for these with regression
analysis.
2) Exposures and outcomes measured similarly in both groups? –
yes.
3) Follow-up sufficiently long and complete? yes. Only 34
drop-outs which occurred in patients whose data was use. 238 women
did not give depression data, not fully explained.
4)
Temporal relationship correct? – no. Not established that
depression preceded CAD.
5) Groups treated equally ? – yes.
6)
Dose-response gradient? – no. Not stratified according to
severity of depression.
Results:
(Table 1 & Table 2)
-18% reported current moderate depression or worse; 39% reported
history of treatment for depression.
-BDI and depression history were both reliably associated with CAD
risk factors and worse cardiac symptoms.
-BDI was associated with an increased mortality.
-depression hx associated with hospitalization, less severe CAD on
angiogram, and decreased likelihood of positive ischemia test.
Comments:
Strengths –
Clinically relevant group, BDI and angiography are well accepted to
predict depression and CAD respectively.
Weaknesses-
No assessment of mild vs. severe depression, many possible
confounders with groups being so different, low mortality rates did
not allow for assessment of causes of mortality.
CLINICAL BOTTOM
LINE:
In this
prospective cohort study of depression and CAD in women referred for
angiography, those women with histories of depression or a BDI>17
had increased cardiac symptoms, increased mortality and greater
chances at hospitalization. They had lower rates of stenosis on
angiography, indicating that their symptoms and mortality might not
have the same pathophysiology as non-depressed women and may need to
be treated differently.
Return
to top of the page
AUTHOR
Carolina Aponte Urdaneta
Ropinirole in
Treatment-Resistant Depression
Chariman’s Rounds, February 12, 2007
Citation:
Cassano P. et al.
Ropinirole in
Treatment-Resistant Depression: A 16-week Pilot Study. Can J
Psychiatry, Vol 50, No. 6, May 2005. p 357-361.
Clinical Question: Efficacy and
tolerability of Ropinorole augmentation in patients with treatment
resistant depression.
Background Info: Ropinirole is a nonergoline dopamine D2-D3 receptor
agonist which is used in the treatment of Parkinson’s disease.
Laboratory and clinical studies have shown that dopamine D2-D3
receptor agonists such as Pramipexole have antidepressant
properties. This study was conducted to assess efficacy and
tolerability of Ropinirole used for augmentation of antidepressant
pharmacotherapy.
Question Type: Therapy
Validity Criteria for Appropriate Question Type:
Follow-up:
follow-up was complete. The 10 patients who entered the trial where
properly accounted for and attributed at its conclusion.
Randomization:
No randomization took place. Open label study. All patients
underwent same intervention. No control group. Intention to treat:
All data was analyzed, with the last observation carried forward.
Similar groups: Only one group which was heterogeneous.
Blinding:
Open label study, no blinding. Equal treatment: Aside of
the intervention all the patients where on different medication
regimens and where therefore not treated equally.
Study Design Type: Clinical trial
Relevant Criteria
for Appropriate Design:
Allocation: One group. No control group. No randomization.
Blinding: Open trial.
Follow Up Period: 16 weeks outpatient.
Setting: Department of psychiatry of the University of Pisa, which
is a national referring center for mood and anxiety disorders.
Patients/Population: n=10, mean age 51 y, 7/10 women, patients who
did not respond to at least one antidepressant treatment,
(characterized by adequate doses and extended period of
administration), on treatment at least for 6 weeks prior to
augmentation. Exclusion criteria: physical illness, substance abuse,
serious suicide risk and psychotic symptoms.
Intervention/exposure: Ropinirole was added to current
antidepressant regimen at flexible doses: initial dose 0.25 mg for 3
days and increased to 0.75 mg at the end of first week and to a
maximum of 1.5 mg by the end of the second week. (Doses where then
adjusted in individual cases and final doses vary from 0.75 -2 mg
between patients).
Outcomes: 50% reduction in MADRS total score plus a 1 or 2 (very
much or much improved respectively) in the CGI-I scale. Secondary
outcome: Dosage Record Treatment emergent Symptom Scale. Remission (MADRS
score less than 12).
Patient Follow Up: At the end of the study only 5/10 of the
patients continued taking Ropinirole. All patients seem to be
accounted for at the end, presumably none where lost to follow up.
Main Results:
Mean scores on the MADRS decreased from 29.6 (SD 7.6) at baseline to
16.9 (SD 12.1) at endpoint (T test, / = 3.10, df 9, P < 0.02); the
CGl-S score decreased from 4.6 (SD 0.84) to 2.5 (SD 1.72) (Wilcoxon
Signed Rank test, z = -2.45; P < 0.02). Using the outcome
criteria, 4 of 10 patients (40%) to be were considered to be
responders. Rate of remission was 40% at the endpoint.
Conclusions: This is a study with several
flaws that make the results interesting but questionable. The
generalizability of results is limited due to sample size n=10 and
exclusion criteria (physically ill, substance abusers, high risk of
suicide). The group seems to be heterogeneous though specific
characteristics are not explicit in the article. The definition of
Treatment resistant depression is not clearly identified. The lack
of equal treatment aside of the intervention limits both the
assessment of efficacy and tolerability. The absence of a placebo
group makes results even more questionable in a disease that has had
placebo responses higher than 30% (though in the article authors say
it’s between 12 -20%). Also, the fact of being an open label trial
in which one of the primary outcomes is a subjective recording by
CGI- I scale, makes the possibility of expectation bias to be
higher. Since this is not randomized-control study their intention
to treat is limited.
Bottom line:
In this open-label non-placebo controlled pilot study of 10
patients with mood disorders, Ropinirole at a mean dose of 1.33 mg
improved MADRS scores (decrease in 50% of initial score) in 40 % of
the patients (4/10). However, the poor study design severely limits
the conclusions that can be drawnfrom this study.
Two
Teaching Points:
Initial studies with no placebo control groups such as this one give
are the initial steps to new Research approaches. The novelty of the
intervention might be a reason for publication. However, depending
on the quality of the design of the study the results can become
more or less confusing.
Two Questions for Discussion:
What would you do next to approach the research of
Ropinirole as an augmentation strategy? With these results would you
be willing to use it in one of your patients?
Return
to top of the page
AUTHOR
Christy Kubit
Effectiveness of Adjunctive Antidepressant Treatment for Bipolar
Depression
Chairman’s Rounds, April 9, 2007
Citation:
Sachs et al. Effectiveness of Adjunctive Antidepressant
Treatment for Bipolar Depression. NEJM 2007; 356. (Study was part of
the STEP-BD trials)
Clinical Question:
Do adjunctive
antidepressants reduce sx of bipolar depression without increasing
the risk of mania?
Question Type:
Therapy
Follow-Up:
All pts who entered the study were accounted for (Table 5). Pts were
treated for up to 26 weeks. Randomization: Using
equipoise-randomization strata, pts were randomized to 3 treatment
groups: 1) mood stabilizer (MS) + placebo, 2) MS + paroxetine, or 3)
MS + buproprion. (Allowed for entry of subjects who wanted to avoid
a specific antidepressant, AD.) Intention to Treat: Pts were
analyzed in the groups into which they were randomized. All
randomized pt data were analyzed. Similar Groups: There were
no significant differences between the 3 groups (see Table 2) in
such characteristics as site, gender, age, race, education, bipolar
type, +/- anxiety d/o, +/- subs abuse. Blinding: Pts,
physicians, and assessors were blinded to group assignment. Equal
Treatment: Beside the intervention, groups were treated equally.
All pts were given the option of remaining with a non-study
psychotherapist, assignment to STEP-BD intervention (long- or
short-term psychosocial intervention), or no psychosocial
intervention.
Study Design Type:
Randomized Control Trial
Duration:
26 weeks of treatment between 11/1999 and 7/2005. Setting: 22
centers, including Mass General (22.3% of pts), Baylor College
(10.1%), and CWRU (17.3%). Patients/Population: n=366/4360
STEP-BD pts: 179 assigned to MS+ antidepressant and 187 assigned to
MS + placebo. Pts were >18yo, met criteria for MDE in the context of
Bipolar I or II disorder per DSM-IV. Dx was established at entry
into study using Structured Clinical Interview for DSM-IV and
Mini-International Neuropsychiatric Interview. Exclusion Criteria:
Pts with a h/o intolerance or non-response to both buproprion and
paroxetine, pts requiring addition or change in dose of an
antipsychotic, or requiring short-term tx for coexistent substance
abuse. Intervention: Paroxetine (initial 10mg
à
max 40mg) or buproprion sustained-release (initial 150mg
à
max 375mg) were chosen for their low rates of switch to mania/
hypomania. All pts were also given FDA-approved mood stabilizer.
Follow-ups occurred at 6wks, at which time responding pts continued
meds, with q4wk f/u for 20 additional weeks. Non-responders were
offered dose increases or open-label dose increases with q2wk f/u
for the next 10wks. Outcomes: Primary outcome of “durable
recovery,” defined as at least 8 consecutive wks of euthymia.
Secondary outcome of “treatment-emergent affective switch.” Pts were
evaluated using Clinical Monitoring Form for mood disorders
(including SCID, SUM-D (continuous sx subscales for depression), and
SUM-ME).
Main Results:
-
23.5% of pts
receiving MS + AD achieved durable recovery, compared to 27.3%
receiving MS + placebo (p=0.4).
-
41.3% of pts
receiving MS + AD achieved durable recovery or transient
remission, compared to 48.7% receiving MS + placebo (p=0.23).
-
12.3% of pts
receiving MS + AD discontinued study meds d/t adverse events,
compared to 9.1% receiving MS + placebo (p=0.32).
Limitations:
1) No “pure” placebo group, 2) Only studied 2
antidepressants, 3) short study length: “durable recovery” lasted
only 8wks. 3) Many pts received psychosocial intervention. 4) Pts
with recent manic episodes were not likely to be referred for study.
Conclusions:
While adding an
antidepressant (paroxetine or buproprion) to a mood stabilizer in
treatment of bipolar depression was not demonstrated to induce
affective switching compared to pts receiving MS + placebo, this
combination was no better in treating depressive symptoms than was a
MS + placebo.
Synopsis:
The use of
adjunctive antidepressant medication, as compared with the use of
mood stabilizers alone, was not associated with increased efficacy
in treating bipolar depression or increased risk of
treatment-emergent affective switch.
Return
to top of the page
AUTHOR:
J.D. Harrison
Depression, Hopelessness, And Desire For Hastened Death In
Terminally Ill Patients With Cancer
Chairman’s Rounds, October 16,
2006
Reference:
Breitbart W,
Rosenfeld B, Pessin H, et al. Depression, hopelessness, and desire
for hastened death in terminally ill patients with cancer. JAMA.
2000;284(22): 2907-11.
Clinical
Question: What are the important determinants of how patients
cope with impending death and what is the role of depression?
Background:
In addition to the controversial debate over physician assisted
suicide, the elucidation of the possible relationships between the
desire for hastened death and depression, hopelessness, social
support, and physical symptoms are important in improving
end-of-life care. Hopelessness was characterized by a pessimistic
cognitive style rather than an assessment of one’s poor prognosis.
Other factors that may play a role in patient’s desire for hastened
death include spiritual well-being, quality of life, symptom
distress, physical functioning, and perception of oneself as a
burden to others.
Methods:
Study design:
Prospective cohort
Inclusion
criteria: Patients recruited at a palliative care hospital in
New York City between 6/1/98-1/31/99, life expectancy of less than 6
months, English speaking, cognitively able to provide informed
consent and valid data (MMSE>20), not considered likely by their
physician to suffer psychological harm from participation.
Final
population: 92 terminally ill cancer patients.
Data collection:
Measures
administered: Schedule of Attitudes Toward Hastened Death (SAHD),
Structured Interview for DSM IV (SCID), Hamilton Depression Rating
Scale, Beck Hopelessness Scale, Duke-UNC Functional Social Support
Questionnaire, Functional Assessment of Chronic Illness
Therapy-Spiritual Well-Being Scale, Brief Pain Inventory, Memorial
Symptom Assessment Scale, Karnofsky Performance Rating Scale, and
abbreviated version of McGill Quality of Life Questionnaire
Assessment
conducted jointly by 2 investigators to establish reliability.
Attempted to complete evaluations in a single interview.
Main outcome:
Scores on the SADH (Desire to hasten death)
Exposures:
SCID diagnosis of Depression, Hopelessness as classified by
endorsing more than 8 items on the BHS
Main Results:
Depressed and
nondepressed patients did not differ on the BHS (P=0.12)
Depression and
hopelessness were only moderately, but significantly, correlated
(r=0.29, p<0.008)
A SCID diagnosis of
depression was significantly associated with desire for hastened
death (χ2=11.44, p=0.001). 15 of 89 (17%) pts met criteria for
depressive episode. Of these 15, 7 (47%) had a desire for hastened
death indicated by a score of >10 on the SAHD. Of 74 pt not
depressed, 9 (12%) had a desire for hastened death. Thus pts with
depression were 4 times more likely to have a desire for hastened
death
2-way analysis of
variance demonstrated that depression and hopelessness had
significant, independent effects in SAHD scores. The presence of
either of these factors individually increased desire to hasten
death somewhat, while the presence of both increased the desire to
hasten death considerably.
Stepwise multiple
regression analysis was conducted. Hopelessness, depression, and
social support were the remaining variables in a significant model
that accounted for 51% of the variance in SAHD scores. Of note,
there was no significant association between desire for hastened
death and pain.
Conclusions:
Depression and Hopelessness are strong, independent predictors of
desire for hastened death in the terminally ill
Validity:
Focused clinical
question: Yes, but they did try to extrapolate to physician
assisted suicide
Inclusion Criteria
appropriate: Yes, explicit and comprehensive patient population, but
unavoidable selection bias to patients who are less ill,
generalizability is an issue since done in institution providing
aggressive state of the art palliative care
Data collection:
Assessment conducted jointly by 2 investigators to establish
reliability. Attempted to complete evaluations in a single
interview. Use of prior validation studies of tests administered.
Interrater reliability was assessed.
Data analyzed
appropriately: yes
Discussion Point:
Limitations inherent to palliative care studies-selection bias,
death or drop out rates, ethical issues
Separating accepted death from desire to hasten death
Hopelessness in the absence of depression in the terminally ill
Discussion
Questions:
Effectiveness of treatment of depression in terminally ill?
How to address hopelessness in a terminally ill patient?
Return
to top of the page
AUTHOR:
Christy Kubit
Chairman’s Rounds, September 18,
2006
Citation:
Daly JJ, Prudic et al. ECT in bipolar and unipolar
depression: differences in speed of response. Bipolar Disorders
2001: 3: 95-104.
Clinical Case:
39-yo WF with a h/o treatment-resistant Bipolar D/o admitted for a
depressive episode and ECT. Pt noted improvement in mood after first
treatment. (Pre-ECT BDI=37, mid-ECT BDI=9.)
Clinical
Question: How do patients with unipolar and bipolar depression
differ in response to ECT?
Validity
Criteria: Therapy Study.
Follow-up:
One week after final ECT session. Randomization: UP & BP pts
were randomized to 3 ECT protocols: 1) RUL ECT vs BL ECT,
Voltage=1xST, 2) RUL ECT V=1xST vs RUL ECT V=2.5xST vs BL ECT V=1xST
vs BL ECT V=2.5xST, 3) RUL ECT V=1.5xST vs RUL ECT V=2.5xST vs RUL
ECT V=6xST vs BL ECT V=2.5xST. Intention to Treat: Pts were
analyzed in the groups to which they were allocated (UPD vs
