Snips from the journals

Update on therapeutics
Shehan Williams and Thilini Rajapakse

Antipsychotic  drugs versus placebo 0·41) more than did oral drugs (RR 0·46, 95% CI
for  relapse  prevention in  schizophrenia: a 0·37-0·53 ); depot haloperidol (RR 0·14, 95% CI 0·04-
systematic review and meta-analysis.(1). 0·55) and fluphenazine (RR 0·23, 95% CI 0·14-0·39)
had had larger effects than other drugs. The effects
of antipsychotic drugs were greater in two unblinded
This is a systematic review and meta-analysis which aims trials (0·26, 0·17-0·39) than in blinded trials (RR 0·42,
to answer the question, are antipsychotics effective in 95% CI 0·35-0·51; p= 0·03). In a meta-regression, the
preventing relapse for patients with schizophrenia? The difference between drug and placebo decreased with
primary outcome was relapse occuring between 7 and study length.
12 months. Secondary outcomes include readmission,
dropout, improvement of disease, death, violence or This study strengthens current practice of maintenance
aggressive behaviour, adverse events, quality of life, treatment with antipsychotic drugs for patients with
satisfaction with care and employment. schizophrenia. In the Sri Lankan context, the use of
depot preparation especially the low cost fluphenazine
The Cochrane Schizophrenia Group’s specialised decanoate for maintenance treatment is also justified.
register for reports published before Nov 11,
2008; and PubMed, Embase, and ClinicalTrials.
gov for those before June 8, 2011 were searched. Adverse effects of second-generation
They also contacted pharmaceutical companies and antipsychotics in  children  and  adolescents: a
searched the reference lists of included studies and Bayesian meta-analysis (2).
previous reviews. Randomised trials of patients
with schizophrenia continued on or withdrawn from
any antipsychotic drug regimen after stabilisation were
eligible. They used the random effects model for analysis This is a systematic review and meta-analysis which
and verified results for the primary outcome with a aims to answer the question, what are the most
fixed effects model. Heterogeneity was investigated common short-term side effects of second-generation
with subgroup and meta-regression analyses. The study antipsychotics (SGAs) in children and adolescents?
was funded by the German ministry of education and The main outcomes considered were: Clinically
research. significant weight gain as declared by the investi-
gator or as defined in the trial (weight gain 97% or
They identified 116 suitable reports from 65 trials, weight gain 95%),; change in glucose, cholesterol and
with data for 6493 patients.  Antipsychotic  drugs triglycerides; change in prolactin; clinically significant
significantly reduced relapse rates at 1 year (drugs 27% hyperprolactinaemia; somnolence/sedation; clinical
vs placebo 64%; risk ratio [RR] 0·40, 95% CI 0·33- extrapyramidal symptoms (EPS) and or akathisia.
0·49; number needed to treat to benefit [NNTB] 3, 95% Studies were searched for in MEDLINE and EMBASE
CI 2-3). Fewer patients given antipsychotic drugs than (1996-2010), Food and Drug Administration and
placebo were readmitted (10% vs 26%; RR 0·38, 95% European Medicines Agency clinical trial registries,
CI 0·27-0·55; NNTB 5, 4-9), but less than a third of and reference lists of review articles. They found
relapsed patients had to be admitted. Limited evidence 41 short-term (3-12 weeks) controlled studies that
suggested better quality of life (standardised mean evaluated  SGA  adverse effects in youths. Using
difference -0·62, 95% CI -1·15 to -0·09) and fewer Bayesian meta-analysis, they analyzed odds ratios
aggressive acts (2% vs 12%; RR 0·27, 95% CI 0·15- (OR) or mean average effects. Numbers of arms and
0·52; NNTB 11, 6-100) with antipsychotic drugs than subjects in the 41 trials were aripiprazole, 10 (n = 671);
with placebo. Employment data were scarce and too few olanzapine, 14 (n = 413); quetiapine, 10 (n = 446);
deaths were reported to allow significant differences to risperidone, 25 (n = 1040); ziprasidone, 4 (n = 228);
be identified. More patients given antipsychotic drugs clozapine, 5 (n = 79); and placebo/untreated, 23 (n
than placebo gained weight (10%vs 6%; RR 2·07, 95% = 1138), totaling 93 arms (4015 patients). Clozapine
CI 2·31-3·25), had movement disorders (16%vs 9%; was assessed only for weight gain and somnolence.
RR 1·55, 95% CI 1·25-1·93), and experienced sedation Compared with placebo, significant treatment-related
(13%vs 9%; RR 1·50, 95% CI 1·22-1·84). Substantial increases were observed for weight gain with olanzapine
heterogeneity in size of effect was recorded. In subgroup (mean= 3.99 ± 0.42 kg; 95% credible interval, 3.17-
analyses, number of episodes, whether patients were in 4.84 kg), clozapine (2.38 ± 1.13 kg; 95% credible
remission, abrupt or gradual withdrawal of treatment, interval, 0.19-4.62 kg), risperidone (2.02 ± 0.32 kg;
length of stability before trial entry, first-generation or 95% credible interval, 1.39-2.66 kg), quetiapine (1.74
second-generation drugs, and allocation concealment ± 0.38 kg; 95% credible interval, 0.99-2.5 kg), and
method did not significantly affect relapse risk. Depot aripiprazole (0.89 ± 0.32 kg; 95% credible interval,
preparations reduced relapse (RR 0·31, 95% CI 0·21- 0.26-1.51 kg); glucose levels with risperidone (3.7 ±

20
Williams and Rajapakse

1.36 mg/dL; 95% credible interval, 1.08-6.42 mg/dL) However, weight increase with olanzapine, risperidone
and olanzapine (2.09 ± 1.08 mg/dL; 95% credible and clozapine and metabolic changes with olanzapine
interval, 0.13-4.32 mg/dL); cholesterol levels with were greater. Additional FES studies including broader
quetiapine (10.77 ± 2.14 mg/dL; 95% credible interval, based SGAs and FGAs are needed.
6.6-14.95 mg/dL) and olanzapine (4.46 ± 1.65 mg/dL;
95% credible interval, 1.24-7.73 mg/dL); triglyceride
levels with olanzapine (20.18 ± 5.26 mg/dL; 95% One-year, randomized, open trial comparing
credible interval, 9.85-30.53 mg/dL) and quetiapine olanzapine, quetiapine, risperidone and
(19.5 ± 3.92 mg/dL; 95% credible interval, 11.84-27.17 ziprasidone effectiveness in antipsychotic-naive
mg/dL); hyperprolactinemia with risperidone (OR,
38.63; 95% credible interval, 8.62-125.6), olanzapine patients with a first-episode psychosis (4).
(OR, 15.6; 95% credible interval, 4.39-41.1), and
ziprasidone (OR, 9.35; 95% credible interval, 1.24- The aim of this study was to compare the 12-month
37.03); and EPS with ziprasidone (OR, 20.56; 95% effectiveness of several second-generation
credible interval, 3.53-68.94), olanzapine (OR, 6.36; antipsychotic drugs, with that of haloperido/l in
95% credible interval, 2.43-13.84), aripiprazole (OR, never-treated patients with first-episode psychosis.
3.79; 95% credible interval, 2.17-6.17), and risperidone In total, 114 patients without life time exposure to
(OR, 3.71; 95% credible interval, 2.18-6.02). any psychotropic medication were randomized to
haloperidol, olanzapine, risperidone, quetiapine
All other SGAs except quetiapine significantly or ziprasidone. Primary outcome was time to all-
increased the risk of EPS compared with the placebo. cause discontinuation. Secondary outcomes included
All SGAs increased the risk of somnolence/sedation discontinuation rates and symptom change as measured
compared to placebo. by the Positive and Negative Syndrome Scale (PANSS).
The overall discontinuation rate was 64%. At 12 months,
the proportion of patients discontinuing treatment
Efficacy and safety of individual second- was 40.0% for olanzapine, 56.5% for quetiapine,
generation vs. first-generation antipsychotics in 64.0% for risperidone, 80.0% for ziprasidone and
first-episode psychosis: a systematic review and 85.7% for haloperidol. Mean time to antipsychotic
discontinuation was higher in patients randomized to
meta-analysis (3). second-generation antipsychotics than in those taking
haloperidol. Significantly lower discontinuation was
Because early treatment choice is critical in first- noted in patients on olanzapine than on haloperidol, or
episode schizophrenia-spectrum disorders (FES), this ziprasidone. The results suggest that olanzapine might
meta-analysis compared efficacy and tolerability of lead to longer treatment continuation in treatment naïve
individual second-generation antipsychotics (SGAs) FEP patients than haloperidol and, possibly ziprasidone.
with first-generation antipsychotics (FGAs) in FES. Global psychopathology was significantly less reduced
They conducted a systematic literature search (until by haloperidol than with each individual  SGA  in this
12 December 2010) and meta-analysis of acute, earliest phase of treatment.
randomized trials with FGA vs.  SGA  comparison;
patients in their first episode of psychosis and The findings of this randomised control study lend
diagnosed with schizophrenia-spectrum disorders; support to the use of SGA in antipsychotic-naïve
available data for psychopathology change, treatment patients with first episode psychosis.
response, treatment discontinuation, adverse effects,
or cognition. Across 13 trials (n  =  2509), olanzapine
(seven trials) and amisulpride (one trial) outperformed
FGAs (haloperidol: 9/13 trials) in 9/13 and 8/13 Polypharmacy withantipsychotics,antidepressants,
efficacy outcomes, respectively, risperidone (eight or benzodiazepines and mortality in schizophrenia
trials) in 4/13, quetiapine (one trial) in 3/13 and (5).
clozapine (two trials) and ziprasidone (one trial)
in 1/13, each. Compared to FGAs, extrapyramidal
symptom (EPS)-related outcomes were less frequent Polypharmacy is widely used in the treatment
with olanzapine, risperidone and clozapine, but of schizophrenia, although guidelines discourage
weight gain was greater with clozapine, olanzapine this practise. This was a registry based case linkage
and risperidone. Pooled SGAs were similar to FGAs study that investigated the use of benzodiazepines,
regarding total psychopathology change, depression, antidepressants, or multiple concomitant antipsychotics
treatment response and metabolic changes. SGAs and its association with increased mortality among
significantly outperformed FGAs regarding lower patients with schizophrenia. This study linked national
treatment discontinuation, irrespective of cause, databases of mortality and medication prescriptions
negative symptoms, global cognition and less EPS and among a complete nationwide cohort of 2588 patients
akathisia, while SGAs increased weight more hospitalized in Finland for the first time with a
(p < 0.05-0.01). Results were not affected by FGA dose diagnosis of schizophrenia between January 1, 2000,
or publication bias, but industry-sponsored studies and December 31, 2007.
favoured SGAs more than federally funded studies. To
summarize, in FES, olanzapine, amisulpride and, less Hazard ratios (HRs) were computed for all-
so, risperidone and quetiapine showed superior efficacy, cause mortality during the use of antipsychotics,
greater treatment persistence and less EPS than FGAs. antidepressants, or benzodiazepines in outpatient care,

Sri Lanka Journal of Psychiatry Vol 3(2) December 2012

21
 Snips from the journals

adjusting for the effects of socio-demographic and citalopram (165), sertraline (110), fluoxetine (102) and
clinical variables, geographic location, and current and bupropion (88).
past pharmacological treatments.
Most antidepressants did not differ in their risk for suicide
Compared with antipsychotic monotherapy, death. However, across several analytic approaches,
concomitant use of 2 or more antipsychotics was not although not instrumental variable analyses, fluoxetine
associated with increased mortality (HR, 0.86; 95% and sertraline had lower risks of suicide death than
CI, 0.51-1.44). Similarly, antidepressant use was not paroxetine. These findings are congruent with the Food
associated with a higher risk for mortality (HR, 0.57; and Drug Administration meta-analysis of randomized
95% CI, 0.28-1.16) and was associated with markedly controlled trials reporting lower risks for “suicidality”
decreased suicide deaths (HR, 0.15; 95% CI, 0.03-0.77). for sertraline and a trend toward lower risks with
However, benzodiazepine use was associated with a fluoxetine than for other antidepressants. Nevertheless,
substantial increase in mortality (HR, 1.91; 95% CI, divergence in findings by analytic approach suggests
1.13-3.22), and this was attributable to suicidal deaths caution when interpreting results.
(HR, 3.83; 95% CI, 1.45-10.12) and to non-suicidal
deaths (HR, 1.60; 95% CI, 0.86-2.97). In total, 826 of
904 patients (91.4%) who used benzodiazepines  had
purchased prescriptions that contained more than 28 Shehan Williams Senior Lecturer,
defined daily doses, violating treatment guidelines. Department of Psychiatry, Faculty of Medicine,
University of Kelaniya, Sri Lanka
Benzodiazepine use was associated with a marked Thilini Rajapakse Senior Lecturer,
increase in mortality among patients with schizophrenia, Department of Psychiatry, Faculty of Medicine,
whereas the use of an antidepressant or several University of Peradeniya, Sri Lanka
concomitant antipsychotics was not. Antidepressant
use was associated with decreased suicide deaths.

Although polypharmacy is best avoided, it is useful to References

keep these findings in mind when rational polypharmacy
is deemed necessary in certain clinical situations. 1. Leucht S,  Tardy M,  Komossa K,  Heres S,  Kissling
W,  Salanti G,  Davis JM. Antipsychotic  drugs versus
placebo for relapse prevention in schizophrenia:
Antidepressant agents and suicide death among a systematic review and meta-analysis. Lancet
US Department of Veterans Affairs patients in 2012; 379(9831):2063-71. CrossRef
depression treatment (6). 2. Cohen D, Bonnot O, Bodeau N, Consoli A, Laurent C.
Adverse effects of second-generation antipsychotics
in  children  and  adolescents: a Bayesian meta-analysis.
J Clin Psychopharmacol 2012; 32(3):309-16. CrossRef
This is a retrospective cohort study with a period 3. Zhang JP, Gallego JA, Robinson DG, Malhotra AK,
of follow up of five years that looked at the risk of Kane JM, Correll CU. Efficacy and safety of individual
suicide of people with depression on anti-depressants. second-generation vs. first-generation antipsychotics in
They assessed associations between suicide death first-episode psychosis: a systematic review and meta-
and treatment with the 7 most commonly used analysis. Int J Neuropsychopharmacol 2012; Dec 3:
antidepressants in a US national sample of Department 1-14. [Epub ahead of print] CrossRef
of Veterans Affairs patients in depression treatment. 4. San L, Arranz B, Perez V, Safont G, Corripio I, Ramirez
Multiple analytic strategies were used to address N,  Dueñas R, Alvarez E. One-year, randomized, open
potential selection biases. Department of Veterans trial comparing olanzapine, quetiapine, risperidone
Affairs patients with depression diagnoses and new and ziprasidone effectiveness in antipsychotic-naive
antidepressant starts between April 1, 1999, and patients with a first-episode psychosis. Psychiatry Res
September 30, 2004 (N = 502,179) were included. 2012; 200(2-3):693-701. CrossRef
Conventional Cox regression models, Cox models with 5. Tiihonen J, Suokas JT, Suvisaari JM, Haukka J, Korhonen
inverse probability of treatment weighting, propensity- P. Polypharmacy with antipsychotics, antidepressants,
stratified Cox models, marginal structural models or benzodiazepines and mortality in schizophrenia. Arch
(MSM), and instrumental variable analyses were Gen Psychiatry 2012;  69(5):476-83. CrossRef
used to examine relationships between  suicide  and 6. Valenstein M,  Kim HM,  Ganoczy D,  Eisenberg
exposure to bupropion, citalopram,  fluoxetine, D, Pfeiffer PN, Downing K, Hoggatt K, Ilgen M, Austin
mirtazapine, paroxetine,  sertraline, and venlafaxine. KL,  Zivin K,  Blow FC,  McCarthy JF. Antidepressant
Crude suicide rates varied from 88 to 247 per 100,000 agents and  suicide  death among US Department of
person-years across antidepressant agents. The highest Veterans Affairs patients in depression treatment. J Clin
rate per 100 000 person-years was among those Psychopharmacol 2012;32(3):346-53. CrossRef
prescribed mirtazapine (247, 95% CI 138 to 407),
followed by venlafaxine (195), paroxetine (180),

22