Download Combination Antipsychotics: Pros, Cons, and Questions

Combination Antipsychotics: Pros, Cons,
and Questions
by Alexander L. Miller and Catherine S. Craig
macy (Covell et al. 2002). Indeed, the evidence is that the
use of combination antipsychotic medications is increasing since the advent of second generation antipsychotics,
despite the extreme paucity of evidence for (or against)
this practice. It is worth noting that in the initial decades
after the introduction of first generation antipsychotics in
North America it was quite common to combine highpotency antipsychotics with sedating low-potency
antipsychotics. Over time it has become clear that there is
little, if anything, to be gained from this practice
(Hollister 1982). In this article, we attempt to explicate
the issues, explore possible rationales and explanations,
and suggest research strategies to initiate studies of combinations of antipsychotics as they are currently used.
Several authors have reviewed and discussed the use
of combination antipsychotics in the treatment of schizophrenia (Canales et al. 1999; Stahl 1999; Weiden and
Casey 1999; Kingsbury et al. 2001). The consensus is
that, except in cases where a patient has failed to respond
to adequate monotherapy trials of several antipsychotics
including clozapine, antipsychotic polypharmacy has little
support in the medical literature, but that there are circumstances in which clinician and patient may serendipitously
hit upon effective combinations. The prescribing patterns
of practitioners in this and other countries show frequent
and variable use of combination antipsychotics. A closer
look at the aforementioned VA Healthcare study reveals
significant variability in antipsychotic polypharmacy rates
among the five sites in the New York metropolitan area
region. The "high" site had a rate of 29 percent while the
"low" site had a rate of 11 percent (Weissman 2002). A
drug utilization survey conducted in Japan found that
more than 90 percent of patients on antipsychotics were
being treated with a combination of high- and lowpotency first generation antipsychotics in low doses (Ito et
al. 1999). A similar study in Italy revealed that approxi-
Abstract
In prescription surveys, use of combination antipsychotics is common and is increasing, despite little supporting evidence. This article highlights potential
problems with using combination antipsychotics, discusses paths to their maintenance use, and reviews the
efficacy evidence. Paths to maintenance antipsychotic
combinations include (1) failure or patient refusal of
all reasonable monotherapies, (2) indefinite continuation of combinations initially intended to be brief, (3)
trials of combinations in preference to reasonable
monotherapy trials, and (4) addition of a second
antipsychotic to counteract a problem (safety, tolerability, or adherence) arising during successful
monotherapy. Virtually all of the evidence on combination antipsychotics is on augmentation of clozapine,
with only one randomized controlled trial.
Research on combination antipsychotics is sorely
needed. Designing clinical trials is made difficult by
the very large numbers of possible combinations and
doses. It may be feasible to analyze existing data bases
to identify combinations that appear particularly
promising to investigate.
Keywords: Schizophrenia; drug therapy, combination; polypharmacy; antipsychotic agents.
Schizophrenia Bulletin, 28(1): 105-109,2002.
Other articles in this issue discuss the extent of use of
combination antipsychotics. A study of outpatient prescribing practices in the New York region of the VA
Healthcare system found an overall antipsychotic
polypharmacy rate of between 15 percent and 17 percent
(Weissman 2002). It should be noted that this study did
not discriminate between temporary (cross-tapering two
antipsychotics) and long-term antipsychotic polypharmacy. Researchers in Connecticut's public mental health
system showed that approximately 10 percent of patients
were maintained on long-term antipsychotic polyphar-
Send reprint requests to Dr. A.L. Miller, Department of Psychiatry, The
University of Texas Health Science Center at San Antonio, 7703 Floyd
Curl Drive, San Antonio, TX 78229-3900; e-mail: [email protected].
105
Schizophrenia Bulletin, Vol. 28, No. 1, 2002
A.L. Miller and C.S. Craig
mately 20 percent of surveyed patients were receiving two
antipsychotics, each prescribed at therapeutic doses
(Tognoni 1999). The most common combination was a
first generation antipsychotic with risperidone.
As noted above in regard to the Weissman study, a
cross-sectional analysis of frequency of use of combination antipsychotics does not distinguish between shortand long-term use. For example, in switching between
antipsychotics, most clinicians choose to overlap or
cross-titrate the two, resulting in a purposely brief period
of combination treatment. Other clinicians, particularly
in inpatient settings, view first generation antipsychotics
as temporarily useful adjuncts to second generation
antipsychotics in treating acute illness exacerbations
(figure 3, Ereshefsky 1999). In each of these instances,
the long-term goal is monotherapy but combinations are
used to achieve short-term goals (reduce risks of switching medications or promote more rapid resolution of
acute symptom exacerbation). While the empirical bases
for these short-term uses of combination antipsychotics
are not particularly strong, they have rationales that are
grounded in clinical experience and involve relatively
brief exposure to the risks of combinations that are
detailed below. On the other hand, long-term maintenance treatment with antipsychotic combinations raises
far more significant questions about efficacy, safety, tolerability, and cost. This article will focus mainly on
these long-term issues.
Why not combinations? The arguments against using
combination antipsychotics, except when monotherapies,
including clozapine, have failed, seem compelling.
Given all the reasons for not using combination
antipsychotics, why are they used so often? Conceptually,
there are four routes to long-term treatment with combination antipsychotics: (1) all reasonable monotherapies,
including clozapine, have been failures or have been
refused by the patient; (2) a combination intended to be
short term is not discontinued; (3) a combination is instituted because of a monotherapy's lack of efficacy, even
though further monotherapy trials would be reasonable; and
(4) a combination is used to partially deal with a particular
problem of monotherapy. We will address these scenarios
separately, examining the rationale and evidence for each.
The largest body of evidence on combination antipsychotics is on combinations with clozapine, to enhance
efficacy. This generally positive literature includes one
controlled trial (Shiloh et al. 1997) and a number of open
label trials. Shiloh and colleagues conducted a 10-week
randomized double-blind trial of the combination of
clozapine-sulpiride versus clozapine-placebo in 28 inpatients partially responsive to clozapine monotherapy. They
found that the average reduction in Brief Psychiatric
Rating Scale (BPRS) score was 20.7 percent in the clozapine-sulpiride group compared to 5.4 percent in the clozapine-placebo group (p < 0.05). Interestingly, patients in
the clozapine-sulpiride cohort fell into two major subgroups, with half demonstrating a mean reduction in
BPRS score of 42.4 percent ("responders") and a little
over a third showing a reduction of less than 5 percent
("nonresponders"). A limitation of the study was that, in
spite of the randomization, patients in the clozapineplacebo group had a significantly longer total duration of
previous hospitalization at baseline (p < 0.05).
Buckley et al. (2001) and Chong and Remington
(2000) have recently reviewed the clozapine augmentation literature. Buckley and colleagues examined the combination of clozapine and numerous psychotropics to treat
the target symptoms of schizophrenia in clozapine nonresponders or partial responders. The adjunctive agents
reviewed included first and second generation antipsychotics, mood stabilizers, selective serotonin reuptake
inhibitors, glycinergic agents, and electroconvulsive therapy. The authors concluded that while none of the above
agents stands out as an obvious first line choice for augmentation, these adjuncts are probably the clinician's best
option considering that very little evidence supports discontinuing clozapine in the hope of achieving an
improved response with a different atypical. Chong and
Remington evaluated the safety and efficacy of clozapine
augmentation in the treatment of schizophrenia, schizoaffective disorder, bipolar disorder, and major depression.
The authors found that, while some augmentation strategies were riskier than others, none of the reviewed combinations was absolutely unsafe. Like Buckley and col-
1. Oth^r than combinations with clozapine, there is an
absence of evidence to support the practice.
2. The likelihood of problematic side effects is increased.
3. The likelihood of problematic pharmacokinetic interactions is increased.
4. The likelihood of harmful pharmacodynamic interactions is increased.
5. Patients are less adherent to complex medical regimes
than simpler regimes (Chen 1991).
6. The costs are greater when second generation antipsychotics are combined in usual doses.
7. The risks of tardive dyskinesia may be as great with the
combination of a first and second generation antipsychotic as with a first generation antipsychotic alone.
8. The clinician often has no basis for deciding what dose
adjustments in which ingredient of the combination to
make in response to increased symptoms or side
effects.
106
Combination Antipsychotics
Schizophrenia Bulletin, Vol. 28, No. 1, 2002
leagues, they concluded that the literature does not support any specific augmentation strategy, and both reviews
emphasized the need for more controlled studies to elucidate the benefits and risks of clozapine augmentation.
The rationale for efforts to augment clozapine is
based on two observations: (1) clozapine is the best medication available for treatment-refractory schizophrenia,
and (2) about half of patients treated with clozapine do
not respond adequately (Lieberman et al. 1994).
Therefore, no other monotherapy is likely to benefit
patients who respond inadequately to clozapine, putting
the clinician in the position of having to resort to combination treatments to try to achieve at least some response.
There is very little information in the literature about
antipsychotic combinations with drugs other than clozapine. There is an open label report of further symptom
improvement in partial responders to olanzapine with
addition of sulphide (Raskin et al. 2000).
A second major route to long-term combination
antipsychotics is the result of a clinical "decision" to
extend a temporary combination indefinitely. Clinically,
the circumstances that most often produce this result are
(1) the continuation of a combination produced by crosstitration of two antipsychotics when the combination
appears to be beneficial, and (2) the continuation of the
component of a combination that was originally begun for
short-term reasons. As noted above, it is a fairly common
practice in inpatient settings to supplement a second generation antipsychotic with a first generation antipsychotic
(sometimes parenterally) in an effort to treat symptoms
such as aggression and agitation and to achieve a more
rapid response. If the patient is discharged on this combination, it may be unclear to the outpatient provider when,
if ever, the first generation antipsychotic should be discontinued. In both these instances, the central issue is
whether the need for the combination has been demonstrated in the individual patient. Clinicians should always
be on the lookout for serendipitously good treatment
results, and combination treatments can fall into this category. Given the problems with combinations, however, it
is incumbent on the prescriber to demonstrate that the
apparent benefits of the combination were not due to the
new drug alone, were not limited to the period of an acute
exacerbation, and were not merely a fortunate coincidence. This can be done only by progressing to a reasonable trial of monotherapy before reinstituting the combination, if the evidence still suggests that the patient was
better off on the combination.
The third route to combinations of antipsychotics is
the decision to use them in preference to further
monotherapy trials, even though reasonable monotherapies have not yet been tried. Often this is done when
clozapine has not yet been tried but there is reluctance on
the part of the patient, the physician, or both to undertake
a trial of clozapine. At other times the decision to use a
combination is rooted in the desire to further improve outcomes in a partial responder to monotherapy. This latter
approach is often accompanied by a pharmacological
rationale, such as adding a stronger dopamine receptor
antagonist if the patient has residual positive symptoms or
adding a stronger serotonin receptor antagonist if the
patient is still troubled by negative symptoms. Such rationales, while appealing, have no clinical trials to support
them and rely on theories about the properties and mechanisms of action of antipsychotic medications that are tentative and evolving. Particularly for patients in the early
years of their illness, one must question the potential
costs, in terms of illness progression, of the decision to
use unproven combination treatments rather than the single treatment that has been shown to be most effective for
treatment-refractory schizophrenia.
A fourth route to antipsychotic combinations is based
on safety and tolerability considerations. One published
instance of this approach addressed the problem of clozapine-induced weight gain and hyperglycemia by partially
substituting quetiapine for clozapine, finding that there
were improvements in both weight and glucose parameters without loss of efficacy (Reinstein et al. 1999).
Replication of this study would be useful. It is not uncommon for clinicians to combine a second generation
antipsychotic with a depot first generation antipsychotic.
There are two rationales for this, both based on the
premise that the patient will not be adequately adherent to
oral monotherapy. First, some clinicians partially replace
the depot medication with one less likely to produce
extrapyramidal symptoms or tardive dyskinesia, an oral
second generation antipsychotic, to improve the safety
and tolerability of the depot preparation. Second, when
patients are good responders to oral second generation
antipsychotics but repeatedly fail to take them regularly
when not closely supervised, some clinicians add depot
antipsychotics as a kind of safety net to prevent a precipitous return of psychosis if the patients miss doses of their
oral medications.
A problem with the first rationale is that, if the patient
is not going to adhere to oral therapy, he or she may be
left on a subtherapeutic dose of depot medication and
therefore be vulnerable to psychotic decompensation. A
problem with the second rationale is that adding a first
generation to a second generation antipsychotic may
reverse the atypical profile of the second generation medication (Kapur 1998). In this instance, adding back the
first generation medication may negate the newer medication's lower incidence of extrapyramidal symptoms and
tardive dyskinesia. As for the future, several depot second
generation antipsychotics are in development, and it will
107
Schizophrenia Bulletin, Vol. 28, No. 1, 2002
A.L. Miller and C.S. Craig
be interesting to see to what extent their advent reduces
the use or the study of the combination of a first generation depot antipsychotic and a second generation oral
antipsychotic.
Research on combinations of antipsychotics is considerably more difficult than research on monotherapies, for
many reasons. Given the very large number of possible
combinations, which ones should be studied? Should they
be dosed at full therapeutic doses of each component or at
a combined therapeutic dose? Should each combination be
studied in different ratios (1:1, 2:1, etc.)? In flexible dosing
studies, should both doses or just one dose be adjusted?
Should the focus be on maintenance and relapse prevention, which is when the long-term issues in using combinations are most problematic, but which are much more difficult studies to do than studies of the treatment of acute
episodes? Given the daunting nature of these research
questions, it is not surprising that almost all of the literature on combination antipsychotics involves clozapine
augmentation. Most patients on clozapine have no viable
monotherapy alternatives, leaving combination treatments
as the only options for improving response. In addition,
funding for trials of combination antipsychotics has not
been as readily available as funding for monotherapy trials. The manufacturers of the newer antipsychotics have
generally been intent on proving the efficacy of their products to the U.S. Food and Drug Administration and
demonstrating their singular value to prescribers. They
have been loath to support studies that could be interpreted
as showing the inadequacy of their products, in at least
some patients. This situation could change as the second
generation antipsychotic market matures and companies
begin to see combinations as a way of expanding their
market penetration, but payers, who are already facing
greatly increased drug costs, are likely to be increasingly
reluctant to fund expensive antipsychotic combinations in
the absence of a body of evidence supporting this practice.
Given the impossibly large number of trials that
would be required to examine the efficacy of all possible
antipsychotic combinations, it would seem worthwhile to
try to better define a more manageable number of combinations that merit investigation with randomized controlled trials. Again, the focus here is on long-term treatment. One approach is pharmacological, examining
combinations that "make sense" in terms of what we
understand about mechanisms of action and individual
drug properties. For example, it might "make sense" to
add a stronger dopamine receptor antagonist to the regime
of patients with persistent positive symptoms who are on
weaker dopamine receptor antagonists. As noted above,
however, the assumption that we know enough about the
pharmacology of antipsychotics to predict a priori which
are more likely to work in combination rests on shaky scientific grounds and an absence of clinical data. A second
approach is to identify patients who are already on combination antipsychotics and are doing well. After excluding
patients whose need for the combination had not been
well demonstrated (e.g., incomplete crossovers between
antipsychotics), one could study the remaining patients to
identify those combinations that appear most frequently in
this naturalistically defined group and, perhaps, their clinical and demographic characteristics. Such an approach
has become practical only recently, with the development
in some large mental health care systems of data bases
that include both pharmacy and clinical outcome information. It would still be necessary to do randomized controlled trials of specific combinations, but the choices
could be guided by the results of analyses that take advantage of the very large number of individual combination
antipsychotic trials that have already occurred but have
not been systematically examined.
References
Buckley, P.; Miller, A.L.; Olsen, J.; Garver, D.; Miller, D.;
and Csernansky, J. When symptoms persist: Clozapine
augmentation strategies. Schizophrenia
Bulletin,
27(4):615-628, 2001.
Canales, P.; Olsen, J.; Miller, A.L.; and Crismon, M.L.
Role of antipsychotic polypharmacotherapy in the treatment of schizophrenia. CNS Drugs, 12(3): 179-188, 1999.
Chen, A. Noncompliance in community psychiatry: A
review of clinical interventions. Hospital and Community
Psychiatry, 42(3):282-287, 1991.
Chong, S.A., and Remington, G. Clozapine augmentation:
Safety and efficacy. Schizophrenia Bulletin, 26(2):
421-^40, 2000.
Covell, N.H.; Jackson, C.T.; Evans, A.C.; and Essock,
S.M. Antipsychotic prescribing practices in Connecticut's
public mental health system: Rates of changing medications and prescribing styles. Schizophrenia Bulletin,
28(1): 17-29, 2002.
Ereshefsky, L. Pharmacological and pharmacokinetic considerations in choosing an antipsychotic. Journal of
Clinical Psychiatry, 60(Suppl 10):20-30, 1999.
Hollister, L.E. Polypharmacy: Which drug combinations
make sense? Hospital and Community Psychiatry,
33(6):433^35, 1982.
Ito, C ; Kubota, Y.; and Sato, M. A prospective survey on
drug choice for prescriptions for admitted patients with
schizophrenia. Psychiatry and Clinical Neurosciences,
53(Suppl):S35-S40, 1999.
Kapur, S. A new framework for investigating antipsychotic
action in humans: Lessons from PET imaging. Molecular
Psychiatry, 3:135-140, 1998.
108
Combination Antipsychotics
Schizophrenia Bulletin, Vol. 28, No. 1, 2002
Kingsbury, S.J.; Yi, D.; and Simpson, G.M. Rational and
irrational polypharmacy. Psychiatric
Services,
52(8): 1033-1036, 2001.
Stahl, S.M. Antipsychotic polypharmacy, part 1:
Therapeutic option or dirty little secret? Journal of
Clinical Psychiatry, 60(7):425-426, 1999.
Lieberman, J.A.; Safferman, A.Z.; Pollack, S.; Szymanski,
S.; Johns, C ; Howard, A.; Kronig, M.; Bookstein, P.; and
Kane, J.M. Clinical effects of clozapine in chronic schizophrenia: Response to treatment and predictors of outcome.
American Journal of Psychiatry, 151(12): 1744—1752,
1994.
Tognoni, G. Pharmacoepidemiology of psychotropic
drugs in patients with severe mental disorders in Italy.
Italian Collaborative Study Group on the Outcome of
Severe Mental Disorders. European Journal of Clinical
Pharmacology, 55(9):685-690, 1999.
Weiden, P.J., and Casey, D.E. "Polypharmacy":
Combining antipsychotic medications in the treatment of
Raskin, S.; Durst, R.; Katz, G.; and Zislin, J. Olanzapine
and sulphide: A preliminary study of combination/augmentation in patients with treatment-resistant schizophre-
nia. Journal of Clinical
schizophrenia. Journal of Practical Psychiatry and
Behavioral Health, July:229-233, 1999.
Psychopharmacology,
Weissman, E. Antipsychotic prescribing practices in the
VA-New York Metropolitan Region. Schizophrenia
Bulletin, 28(l):31-42, 2002.
2O(5):5OO-5O3, 2000.
Reinstein, M.J.; Sirotovskaya, L.A.; Jones, L.E.; Mohan,
S.; and Chasanov, M.A. Effect of clozapine-quetiapine
combination therapy on weight and glycaemic control:
The Authors
Preliminary findings. Clinical Drug Investigations,
18(2):99-104, 1999.
Alexander L. Miller, M.D., is Professor of Psychiatry and
Pharmacology at the University of Texas Health Science
Center at San Antonio, and Director of Clinical Research
at San Antonio State Hospital, San Antonio, TX.
Catherine S. Craig, Pharm.D., is Instructor, Department of
Psychiatry, The University of Texas Health Science
Center at San Antonio, San Antonio, TX.
Shiloh, R.; Zemishlany, Z.; Aizenberg, D.; Radwan, M.;
Schwartz, B.; Dorfman-Etrog, P.; Modai, I.; Khaikin, M.;
and Weizman, A. Sulpiride augmentation in people with
schizophrenia partially responsive to clozapine: A doubleblind, placebo-controlled study. British Journal of
Psychiatry, 171:569-573, 1997.
109