Letters to the Editor

Letters to the Editor

©

Association Between REFERENCES

Premenstrual Syndrome and Depression
Co

1. Roca CA, Schmidt PJ, Rubinow DR. A follow-up study of

premenstrual syndrome. J Clin Psychiatry 1999;60:763–766
py

Sir: We read with great interest the article by Roca and col- 2. Severino K. Premenstrual dysphoric disorder: controversies
leagues1 focusing on the follow-up of women with a previous surrounding the diagnosis. Harv Rev Psychiatry 1996;3:293–295
r

diagnosis of premenstrual syndrome (PMS). Roca and col- 3. Yonkers KA. The association between premenstrual dysphoric
ig

leagues demonstrated that PMS seems to be a stable diagnosis disorder and other mood disorders. J Clin Psychiatry 1997;58
ht

over time (5- to 12-year follow-up) by confirming previous (suppl 15):19–25

diagnosis using 2 prospective cycles of daily ratings (visual 4. Merikangas KR, Foeldenyi M, Angst J. The Zurich Study, 19:
patterns of menstrual disturbances in the community: results of
20

analogue scale) and a retrospective evaluation (DSM-IV criteria

for premenstrual dysphoric disorder [PMDD]). Despite the fact the Zurich Cohort Study. Eur Arch Psychiatry Clin Neurosci 1993;
243:23–32
01 ne p

that only 7 (26%) of 27 former PMS patients and 11 (52%) of 21

5. Harlow BL, Cohen LS, Otto MW, et al. Prevalence and predictors
O

controls completed 2 cycles of daily ratings, the revalidation of of depressive symptoms in older premenopausal women. Arch Gen
the PMS diagnosis was remarkable, and all PMS patients met
Pherso

Psychiatry 1999;56:418–424
DSM-IV criteria for PMDD. Thus, the stability of the diagnosis 6. Arpels JC. The female brain hypoestrogenic continuum from the
seen over time appears to strengthen the validity of PMDD as a premenstrual syndrome to menopause: a hypothesis and review
ys nal c

distinct syndrome. of supporting data. J Reprod Med 1996;41:633–639

ic op

The authors also reported a nonsignificant higher incidence

of depressive disorders among PMS patients (N = 27) com- Cláudio N. Soares, M.D., Ph.D.
ia y m

pared with controls (N = 21) in this small clinic-based study. Lee S. Cohen, M.D.
ns ay

Several authors2,3 have also addressed the question of the rela- Massachusetts General Hospital–Harvard Medical School
tionship between depression and severe PMS (lifetime and/or Boston, Massachusetts
Po be

current comorbidity), given the existing overlap of symptoms

(e.g., mood swings, lack of energy, and irritability). Previous
Dr. Roca and Colleagues Reply
stgprin

large community-based studies, on the other hand, have system-

atically pointed out the significant association between premen-
strual symptoms and a lifetime occurrence of psychiatric
ra ted

Sir: We appreciate the thoughtful and astute questions posed

disorders, as shown in a 10-year prospective epidemiologic co- by Drs. Soares and Cohen. They correctly raise the possibility
du

hort study of 218 young adults in Zurich, Switzerland,4 and, that women who display susceptibility to mood disturbance dur-
more recently, in a community-based cohort study of 4000 ing the menstrual cycle might be vulnerable to depression dur-
a

women from the Harvard Study of Moods and Cycles.5

te

ing other periods of reproductive endocrine change. Although

Moreover, it has been postulated that women with premen- one of the patients (and none of the controls) had a history of
strual depressive symptoms may constitute a subgroup with a
Pr

postpartum depression, the Structured Clinical Interview for

particular vulnerability to depression during periods of intense DSM-IV did not enable us to determine retrospectively the oc-
es

hormonal fluctuations, such as the postpartum period or the currence of postpartum blues. While the role of gonadal steroids
transition to the menopause.6 Thus, to better understand the pos- in postpartum blues, a fairly ubiquitous phenomenon, has not
s,

sible association between depressive episodes and PMS, it been determined, we have recently shown that euthymic women
would be helpful to know the percentage of PMS patients and
In

with a history of postpartum depression will experience depres-

controls in the study by Roca and colleagues who developed sion in association with gonadal steroid manipulations that are
c.

postpartum blues and/or postpartum depression. without effect in women lacking a history of postpartum depres-
In addition, taking into account the study design, Roca and sion,1 thus reinforcing the relevance of the first question raised
colleagues had to exclude PMS subjects and controls who were by Drs. Soares and Cohen.
menopausal at follow-up. Those women may have experienced Similarly, although we were unable, for obvious reasons, to
depressive symptoms during the menopausal transition. confirm the diagnosis of premenstrual syndrome (PMS) in
Similarly, PMS patients and controls with menstrual irregularity women who no longer were menstruating, it is possible that the
also had to be excluded. Considering that a significant percent- perimenopause might represent a period of increased vulner-
age of patients suffering from PMS and reporting menstrual ability to depression in those with a history of PMS. At
irregularity were probably becoming perimenopausal (mean follow-up, 9 women with histories of PMS were postmeno-
age = 47.1 ± 5.7 years), this could constitute a new opportunity pausal and 4 were probably perimenopausal (based, albeit im-
to investigate the extent to which the transition to the meno- perfectly, on age and recent-onset menstrual cycle irregularity).
pause would strengthen the association between depression and Of the 9, 3 were currently symptomatic and were being treated
reproductive-associated psychiatric disturbance. with antidepressants, and 2 were asymptomatic but on treatment

676
677 J Clin Psychiatry 61:9, September 2000
 Letters to the Editor

with an antidepressant. The 4 remaining subjects were asympto- 149:689–690

matic and on no psychotropic medication treatment. Regretta- 4. Basson B, Kinon BJ, Taylor CC, et al. Factors influencing weight
bly, the inaccuracy of retrospective estimates of the onset of change in patients with schizophrenia treated with olanzapine
perimenopause and depression precludes conclusion about the versus haloperidol or risperidone. Presented at the 51st Institute on
Psychiatric Services; Oct 29–Nov 2, 1999; New Orleans, La
proximity of depression to the perimenopause. Additionally, 5. Umbricht DSG, Pollack S, Kane JM. Clozapine and weight gain.
none of the 4 currently perimenopausal subjects was symptom- J Clin Psychiatry 1994;55(9, suppl B):157–160
atic. We conclude, therefore, that this data set is inadequate to 6. Bustillo JR, Buchanan RW, Irish D, et al. Differential effect of
answer the important question of the role of the perimenopause clozapine on weight: a controlled study. Am J Psychiatry 1996;153:
in the onset of depression in those with demonstrated reproduc- 817–819
tive endocrine-related affective disorders.
J. Steven Lamberti, M.D.
REFERENCE University of Rochester Medical Center
Rochester, New York
©

1. Bloch M, Schmidt PJ, Danaceau M, et al. Effects of gonadal steroids

Co

in women with a history of postpartum depression. Am J Psychiatry

2000;157:924–930 Dr. Ganguli Replies
py

Catherine A. Roca, M.D. Sir: I appreciate Dr. Lamberti’s comments on my article.1

Peter J. Schmidt, M.D.
r

He is correct in pointing out that, in his 1992 report, there was

ig

David R. Rubinow, M.D. an inverse correlation between weight gain and decrease in the
National Institute of Mental Health
ht

Brief Psychiatric Rating Scale score. However, the correlation

Bethesda, Maryland was not statistically significant (Spearman r = 0.31, df = 28,
20

p < .1). We were not aware of the second study cited by him,
which was apparently an unpublished oral presentation. With-
01 ne p

Is Antipsychotic Drug–Induced Weight Gain out further peer-reviewed evidence, we would still contend that
O

Associated With a Favorable Clinical Response? the relationship between weight gain and improvement in clini-
cal state remains tentative.
Pherso

Sir: I am writing to comment on Dr. Ganguli’s recent article1

REFERENCE
ys nal c

on weight gain associated with antipsychotic drugs. The article

stated that reports by Leadbetter et al.2 and Lamberti et al.3 dif-
ic op

fered as to whether clozapine-induced weight gain correlated 1. Ganguli R. Weight gain associated with antipsychotic drugs. J Clin
ia y m

with clinical improvement among patients with schizophrenia. Psychiatry 1999;60(suppl 21):20–24
ns ay

As a point of clarification, our findings were consistent with

those of Leadbetter et al. We found a trend for Brief Psychiatric Rohan Ganguli, M.D.
Rating Scale score changes to correlate inversely with amount School of Medicine, University of Pittsburgh
Po be

of weight gained such that lower (improved) scores were asso- Pittsburgh, Pennsylvania
ciated with larger gains. Most recently, this issue was examined
stgprin

in a large group of patients with schizophrenia or related disor-

ders receiving treatment with olanzapine, risperidone, or halo- Polypharmacy of
ra ted

peridol.4 Analyzing data from a total of 2266 patients, weight 2 Atypical Antipsychotics
du

gain was found to be associated with a favorable clinical re-

sponse for each of the 3 drugs.
Sir: I am writing to respond to the BRAINSTORMS article by
a

As noted by Dr. Ganguli, other investigators have not found

Dr. Stephen Stahl in the July 1999 issue of the Journal.1 I agree
te

an association between antipsychotic drug–induced weight gain

and clinical improvement.5,6 While it remains unclear whether that polypharmacy should be avoided. I was a resident in the
Pr

the association exists, the strength of the available evidence early 1970s, when even having a person on treatment with an
suggests that further study is needed. If an association between antidepressant and diazepam was considered a serious thera-
es

drug-induced weight gain and clinical improvement is estab- peutic misadventure. In his article, Dr. Stahl stated that no justi-
fication exists for using 2 expensive atypical antipsychotics and
s,

lished, it may provide a clue to the mechanisms underlying anti-

psychotic drug efficacy. Beyond the neuropharmacologic that this mixture usually results from being trapped in a
In

relevance of this issue, the clinical relevance of drug-induced cross-titration between 2 atypical agents without an adequate
weight gain is clear. Weight gain is a common side effect of trial of monotherapy. I describe a case of a patient who did re-
c.

antipsychotic medications. While not all patients will experi- ceive adequate trials of clozapine and olanzapine but neverthe-
ence the side effect to a concerning degree, effective clinical less did best on a mixture of the 2 drugs.
strategies are needed for management of those who do.
Case report. Ms. A, a 55-year-old woman with a 30-year his-
tory of schizoaffective disorder, had numerous admissions over
REFERENCES
the years for paranoia, delusions, agitation, and some mixed af-
fective components. Finally in 1993, Ms. A’s condition was sta-
1. Ganguli R. Weight gain associated with antipsychotic drugs. J Clin bilized with clozapine, 300 mg/day. She had tried most
Psychiatry 1999;60(suppl 21):20–24
2. Leadbetter R, Shutty M, Pavalonis D, et al. Clozapine-induced weight
neuroleptics as well as lithium, carbamazepine, and divalproex
gain: prevalence and clinical relevance. Am J Psychiatry 1992;149: sodium without success until she took clozapine. She was able
68–72 to live independently in an apartment and was involved in the
3. Lamberti JS, Bellnier T, Schwarzkopf SB. Weight gain among local mental health center’s socialization program. Unfortu-
schizophrenic patients treated with clozapine. Am J Psychiatry 1992; nately, despite attempts to manipulate her clozapine dosage or

J Clin Psychiatry 61:9, September 2000 677

678
Letters to the Editor

supplement clozapine with antidepressants or typical antipsy- tive patient who apparently did much better on treatment with a
chotics, Ms. A remained very sedated. Because she slept all combination of clozapine and olanzapine than she did with either
night and much of the day, she did not bother anyone, but she agent alone. A number of anecdotal observations like his are be-
would not attend the socialization program very often because ginning to make their way into the medical literature, usually as
she was sleeping. case reports or uncontrolled case series. For example, both ris-
The mental health center then tried to switch her to olanza- peridone2,3 and loxapine4 have also been reported to have favor-
pine. Over 3 months, clozapine was gradually decreased and able effects in augmenting clozapine’s efficacy in a small number
olanzapine, 10 mg/day, was added. When clozapine was of schizophrenic or schizoaffective patients. A recent report by
stopped, Ms. A’s condition greatly deteriorated. She became de- Taylor et al.3 is the largest series to date (13 patients with only
lusional and agitated and was admitted to the hospital for threat- partial responses to clozapine monotherapy) and, although an
ening neighbors and family members. The olanzapine dose was open study, did use standardized rating scales to rate symptoms
raised to 25 mg/day without any diminution in symptoms. Di- on nonclozapine monotherapy, clozapine monotherapy, and ris-
valproex, gabapentin, carbamazepine, haloperidol, and clonaze- peridone augmentation of clozapine. Their results suggest that 4
©

pam were tried without any real success. Eventually, Ms. A patients were much improved, 6 patients minimally improved,
Co

calmed down enough for discharge back to her apartment on and 3 patients were unchanged when risperidone was added to
treatment with olanzapine and divalproex. This status lasted clozapine.
only 2 weeks before the paranoia became too great, leading Although it is gratifying to find the occasional patient who
py

again to agitation and threats. She called her daughter all night finally responds to 2 antipsychotics when monotherapies and
complaining about things that were happening or about to hap- other augmentation strategies, such as lamotrigine,5 have failed,
r ig

pen to her. it should be emphasized that there is no way to predict who

Ms. A was readmitted, and her behavior in the hospital was might benefit from such an approach. Unfortunately, no con-
ht

similar to that during the prior admission: paranoia, delusions, trolled trials of antipsychotic augmentation by other antipsy-
agitation, intrusiveness, and rage lasting 30 minutes were fol- chotics have been conducted, even though the rate of use of 2
20

lowed by her being cooperative and pleasant, only to re-cycle concomitant antipsychotics is between 10% and 60%.6–10 This
30 minutes later. Her daughter mentioned that Ms. A was the lack of meaningful clinical guidelines for the use of 2 antipsy-
01 ne p

best she had been in 30 years in late January, when she was tak- chotics, plus the potentially huge costs of unlimited experimen-
O

ing clozapine and olanzapine. After a call to the mental health tation with 2 antipsychotic drugs in the hope of finding a
Pherso

center, it was determined that she had been on clozapine, 100 clinically meaningful responder, must be borne in mind before
mg/day, and olanzapine, 10 mg/day, at that time. This medica- this approach can be advocated.
ys nal c

tion regimen was then instituted with the approval of Ms. A and As the costs of antipsychotic drug treatment skyrockets, there
her daughter, and Ms. A promptly improved. Within 4 days, she is increasing pressure from payers to restrict their use. For ex-
ic op

was calmer, pleasant, helping elderly patients in the unit, help- ample, in California, the governor has recently convened a task
ia y m

ing the nurses with meals, and not only slept well at night but force to control dramatically increasing prescription medication
ns ay

was also awake and alert during the day. At follow-up 18 costs in the Medicaid (Medi-Cal) program, due predominantly to
months later, Ms. A continued to do well. She went to day treat- increasing costs of atypical antipsychotic drugs. The 2 most ex-
ment regularly, drove neighbors to their doctors’ appointments, pensive drugs in the fee-for-service Medi-Cal program for 1999
Po be

and even had been allowed to baby-sit her granddaughter alone. were olanzapine at $115,900,000 and risperidone at $61,827,000
before any rebates.11 The atypical antipsychotics were also the
stgprin

It is unclear why the combination of clozapine and olanza- most expensive medications per claim, per pill, and per day, with
pine worked when numerous other combinations did not. Nev- olanzapine costing on average $10.22 per day; risperidone,
ra ted

ertheless, this patient definitely has had a return to an almost $6.29 per day; clozapine, $11.88 per day; and quetiapine, $6.54
du

normal life. Fortunately, her daughter is very supportive and per day. Although the occasional patient may benefit from 2 such
caring toward her mother. I spoke to the daughter recently, and agents, approximately 25% of these patients are receiving 2 anti-
a

she commented that she has not known her mother to be this psychotics. Pressures to reduce escalating drug costs may soon
te

functional in 30 years. One thing that this case shows is the need lead to formulary restrictions unless prescribers are cautious
to listen to family members. Some of the old notions about about potentially very expensive and as yet unproven uses of
Pr

monotherapy are not as important anymore; however, the best these agents. Studies of the cost-effective use of 2 antipsychotics
practice is still to maintain the patient on treatment with the are long overdue. In the meantime, it might be best to administer
es

lowest dosage of medication possible as well as the least num- these powerful and important therapeutic agents as treasured re-
s,

ber of medications. sources whose use will be soon curtailed unless they are pre-
scribed prudently.
In

REFERENCE
REFERENCES
c.

1. Stahl SM. Antipsychotic polypharmacy, part 1: therapeutic option or

dirty little secret? [BRAINSTORMS] J Clin Psychiatry 1999;60:425–426 1. Stahl SM. Antipsychotic polypharmacy, part 1: therapeutic option or
dirty little secret? [BRAINSTORMS] J Clin Psychiatry 1999;60:425–426
2. Henderson DC, Goff DC. Risperidone as an adjunct to clozapine
Edward Rhoads, M.D., F.A.P.A. therapy in chronic schizophrenics. J Clin Psychiatry 1996;57:
Private Practice 395–397
Greensboro, North Carolina 3. Taylor CG, Flynn SW, Altman S, et al. An open trial of risperidone
augmentation of partial response to clozapine. Schizophr Bull. In press
4. Mowerman S, Siris SG. Adjunctive loxapine in a clozapine-resistant
Dr. Stahl Replies cohort of schizophrenic patients. Ann Clin Psychiatry 1996;8:193–197
5. Dursun SM, McIntosh D. Clozapine plus lamotrigine in
treatment-resistant schizophrenia [letter]. Arch Gen Psychiatry
Sir: In response to my article on antipsychotic polypharmacy,1 1999;56:950
Dr. Rhoads reports serendipitous observations of a schizoaffec- 6. Naber D, Holzbach R, Perro C, et al. Clinical management of cloza-

678
679 J Clin Psychiatry 61:9, September 2000
 Letters to the Editor

pine patients in relation to efficacy and side-effects. Br J Psychiatry REFERENCES

1992;160(suppl 17):54–59
7. Peacock L, Gerlach J. Clozapine treatment in Denmark: concomitant 1. Kumar V, Brecher M. Psychopharmacology of atypical antipsychotics
psychotropic medication and hematologic monitoring in a system and clinical outcomes in elderly patients. J Clin Psychiatry 1999;
with liberal usage practices. J Clin Psychiatry 1994;55:44–49 60(suppl 13):5–9
8. Stahl SM. Selecting an atypical antipsychotic by combining clinical 2. Pickar D. Prospects for pharmacotherapy of schizophrenia. Lancet
experience with guidelines from clinical trials. J Clin Psychiatry 1995;345:557–561
1999;60(suppl 10):31–41 3. Goldstein JM. Pharmacotherapy of schizophrenia [letter, comment].
9. Ereshefsky L. Pharmacologic and pharmacokinetic considerations Lancet 1995;346:450
in choosing an antipsychotic. J Clin Psychiatry 1999;60(suppl 10): 4. Seroquel [package insert]. Wilmington, Del: Zeneca Pharmaceuticals;
20–30 1997
10. Taylor D, Mace S, Mir S, et al. A prescription survey of the use of 5. Saller CF, Salama AI. Seroquel: biochemical profile of a potential
atypical antipsychotics for hospital inpatients in the United Kingdom. atypical antipsychotic. Psychopharmacology (Berl) 1993;112:
Int J Psychiatry Clin Pract 2000;4:41–46 285–292
©

11. California Department of Health. Drug Utilization Review Commit- 6. Goldstein JM. Preclinical profile of Seroquel (quetiapine): an atypical
tee Report. Sacramento, Calif: California Department of Health; 2000 antipsychotic with clozapine-like pharmacology. In: Holiday SG,
Co

Ancill RJ, MacEwan GW, eds. Schizophrenia: Breaking Down the

Stephen M. Stahl, M.D., Ph.D. Barriers. New York, NY: John Wiley & Sons; 1996:177–208
py

University of California, San Diego 7. Goldstein JM. Atypical antipsychotics: beyond acute psychosis,
San Diego, California new directions. Emerging Drugs 1999;4:127–152
r

8. Goldstein JM. Seroquel (quetiapine fumarate): a new atypical anti-

ig

psychotic. Drugs Today 1999;35:193–210

9. Goodman LS, Gilman AG, eds. Goodman and Gilman’s The
ht

Clarification of Pharmacological Basis of Therapeutics. 9th ed. New York, NY:

Anticholinergic Effects of Quetiapine McGraw-Hill; 1996:588
20

Jeffrey M. Goldstein, Ph.D., M.Sc.

01 ne p

Sir: In Table 1 of the article by Kumar and Brecher,1 the Martin Brecher, M.D.
O

muscarinic receptor binding affinity shown for quetiapine is AstraZeneca Pharmaceuticals

Pherso

not supported by the pharmacologic and clinical properties of Wilmington, Delaware

quetiapine and may mislead the readers regarding the anticho-
ys nal c

linergic effects of quetiapine. The reference used to support the

affinity for the muscarinic receptor (rated as “+++,” or “moder-
ic op

ate,” in Table 1) was taken from Pickar,2 but this value has Side Effect Profile of Enteric-Coated
previously been refuted by Goldstein.3 It is again important to Divalproex Sodium Versus Valproic Acid
ia y m

set the record straight, in particular because their article deals

ns ay

with elderly patients who are especially sensitive to the distress- Sir: Valproic acid and its derivatives are important agents in
ing adverse effects resulting from blockade of the muscarinic the treatment of patients with bipolar illness. This drug has been
Po be

receptor. extensively researched in the treatment of epilepsy, where 2 of

With regard to anticholinergic properties, the following are its common side effects have included gastrointestinal side ef-
stgprin

true about quetiapine: fects and thrombocytopenia.1–4 Two representative derivatives

are valproic acid and divalproex sodium (the enteric-coated de-
ra ted

• Quetiapine has no appreciable affinity for muscarinic rivative), the latter being more expensive but equally effective.
cholinergic receptors (IC50 > 5000 nM).4–8 However, in individual clinical cases, the side effect profile and
du

• Quetiapine demonstrates no muscarinic acetylcholine tolerability of each of these derivatives also need to be taken
antagonist activity in a standard isolated tissue assay into consideration. We present here a case report of a naturalis-
a

(guinea pig ileum) for anticholinergic activity (data tic A-B-A design, which included the replacement of divalproex
te

on file, AstraZeneca Pharmaceuticals). by valproic acid that resulted in the occurrence of gastrointes-
Pr

• Quetiapine demonstrates no muscarinic acetylcholine tinal symptoms and thrombocytopenia, while a switch back to
antagonist activity in a standard behavioral model divalproex resulted in the disappearance of these adverse ef-
es

(physostigmine-induced lethality in mice) for anti- fects.

cholinergic activity (data on file, AstraZeneca Phar-
s,

maceuticals). Case report. Mr. A, a 45-year-old white, divorced, unem-

In

ployed man diagnosed by DSM-IV with bipolar I illness, was

It is clear from preclinical pharmacology findings that que- admitted to a psychiatric hospital with depressed and irritable
c.

tiapine would be predicted to have minimal anticholinergic ef- (mixed) mood state, suicidal intent, and increasing anxiety and
fects in humans. agitation. He was started on buspirone, 10 mg t.i.d., and dival-
In clinical use, the incidence of anticholinergic effects proex, 750 mg t.i.d. Thyroxine was continued at a daily dose of
is low. Any observed anticholinergic effects are most likely 50 µg for hypothyroidism that had been diagnosed in the past.
due to quetiapine’s high affinity for the histamine H1 receptor, Four weeks later, Mr. A’s divalproex treatment was switched to
since it is known that histamine blockade can cause some valproic acid, given twice daily at the same dose. This occurred
“anticholinergic-like” actions that can be manifest in clinical when he was transferred to another facility, where unless
use.9 However, quetiapine is very well tolerated in special popu- divalproex is specifically requested by the physician, valproic
lations that are especially sensitive to anticholinergic effects, acid is dispensed instead of divalproex. His platelet count the
such as elderly patients with psychoses due to Parkinson’s dis- day before the change from divalproex to valproic acid was
ease and Alzheimer’s disease. Quetiapine’s use as a first-line 151,000/µL (normal range, 130,000–400,000/µL). Four and 5
agent for these populations and for patients with schizophrenia weeks later during routine checking of the complete blood
is a testament to this characteristic. count and differential for the appearance of leukopenia, his

J Clin Psychiatry 61:9, September 2000 679

680
Letters to the Editor

platelet count decreased to 122,000/µL and 110,000/µL, respec- Since other causes, including spontaneous changes and auto-
tively, while no thrombocytopenia-associated symptoms were immune processes, may be associated with the thrombocytope-
noted. Such a decrease to below the normal level is not expected nia as observed in our case, there is a need to compare the
to occur spontaneously. These changes were also accompanied incidence of thrombocytopenia associated with valproic acid
by nausea and emesis. These side effects led to a switch versus divalproex in larger databases. If there are differences in
back to divalproex and discontinuation of the valproic acid this regard, they may be important from several perspectives:
treatment. A gradual increase of the platelet count was noted, up patients’ acceptance (side effects, tolerability, adherence to
to 151,000/µL 2 weeks later. Symptoms of nausea and emesis long-term treatment), costs (use of ancillary medications—his-
also improved within 3 days of the switch. Valproate levels tamine-2 blockers for gastrointestinal symptoms, laboratory
were examined along with the platelet count and were found to costs of monitoring thrombocytopenia), and the risk-to-benefit
be within the therapeutic range. ratios in prescribing these different formulations for certain pa-
Mr. A’s medical conditions included mild obesity, hypothy- tients.
roidism, and type 2 diabetes mellitus. The history of adverse
©

effects of drug treatment included a history of rash from carba- REFERENCES

Co

mazepine treatment. Mr. A is a nonsmoker and has a remote his-

tory of alcohol abuse, but has been abstinent for 5 years. The 1. Zajecka J. Pharmacology, pharmacokinetics, and safety issues of
type 2 diabetes mellitus and hypothyroidism remained under
py

mood-stabilizing agents. Psychiatr Ann 1993;23:79

control with diet and thyroxine treatment, respectively. 2. Gidal B, Spencer N, Maly M, et al. Valproate-mediated disturbances
of hemostasis: relationship to dose and plasma concentration. Neurol-
r ig

Our data concur with the findings of Zarate and colleagues,3 ogy 1994;44:1418–1422
who reported fewer gastrointestinal side effects with divalproex 3. Zarate CA Jr, Tohen M, Narendran R, et al. The adverse effect profile
ht

compared with valproic acid. The literature for valproate sug- and efficacy of divalproex sodium compared with valproic acid: a
pharmacoepidemiology study. J Clin Psychiatry 1999;60:232–236
gests that gastrointestinal side effects occur in more than 20%
20

4. Physicians’ Desk Reference. 53rd ed. Montvale, NJ: Medical

of subjects taking this drug, whereas about 1% develop throm- Economics Co; 1999:417–443
bocytopenia with low doses and more than 20% with higher
01 ne p

5. Tohen M, Castillo J, Baldessarini R, et al. Blood dyscrasias with

doses.4 However, Tohen et al.5 reported no case of thrombocyto-
O

carbamazepine and valproate: a pharmacoepidemiological study

penia (platelet count < 100,000/µL) in a large sample of pa- of 2,228 cases at risk. Am J Psychiatry 1995;152:413–418
Pherso

tients treated with valproate. 6. Klotz U, Antonin KH. Pharmacokinetics and bioavailability of
We are not aware of previous reports that the frequency of sodium valproate. Clin Pharmacol Ther 1977;21:736–743
ys nal c

thrombocytopenia is decreased with divalproex versus valproic 7. Wilder BJ, Karas BJ, Hammond EJ, et al. Twice daily dosing of
acid administration. Divalproex, compared with valproic acid, valproate with divalproex. Clin Pharmacol Ther 1983;34:501–504
ic op

8. Beydoun A, Sackellares JC, Shu V, and the Depakote Monotherapy

seems to be associated with a less rapid increase of peak plasma for Partial Seizures Study Group. Safety and efficacy of divalproex
ia y m

levels, reaching peak levels after 3 hours.1 In contrast, peak sodium monotherapy in partial epilepsy: a double-blind, concentra-
ns ay

plasma levels for valproic acid were observed 0.6 to 1.2 hours tion-response design clinical trial. Neurology 1997;48:182–188
after its administration,4,6 and its rate of absorption was sug- 9. Delgado MR, Riela AR, Mills J, et al. Thrombocytopenia secondary
gested to be more rapid as compared with divalproex.7
Po be

to high valproate levels in children with epilepsy. J Child Neurol

No consensus exists as to the mechanism underlying valpro- 1994;9:311–314
ate-induced thrombocytopenia.2 Interestingly, Gidal et al.2 re-
stgprin

ported that valproate administration also alters some key Joseph Levine, M.D., M.A.Sci.
physiologic functions of platelets. Thrombocytopenia was sug- K. N. Roy Chengappa, M.D., F.R.C.P.C.
ra ted

gested to occur more frequently in those with high versus low Haranath Parepally, M.D.
du

plasma valproate levels.2,8,9 On the basis of data associated with Western Psychiatric Institute and Clinic
a variety of side effects such as tremor and lethargy,4 thrombo- Pittsburgh, Pennsylvania
a

cytopenia may also be associated with a rapid increase in

te

plasma levels. It might be that the less rapid increase of plasma

levels with a more flattened curve over time may be relevant to
Pr

the disappearance of thrombocytopenia in our subject after he

was switched from valproic acid back to divalproex. The throm- Correction
es

bocytopenia in the present report developed 4 weeks after the

s,

initiation of valproic acid. In general, blood dyscrasias are re- In the article “Mirtazapine Substitution in SSRI-Induced
ported with this drug after 2 to 4 weeks of treatment.4 The Sexual Dysfunction” by Alan J. Gelenberg, M.D., et al. (May
In

thrombocytopenia with this drug is reported to be transient, that 2000 issue, pp. 356–360), the order of authorship should have
is, it disappears on reduction or discontinuation of the drug. In shown Cindy McGahuey, B.A., as the second author. The cor-
c.

general, it is recommended to monitor for the appearance of rected byline is as follows: Alan J. Gelenberg, M.D.; Cindy
thrombocytopenia if other drugs (e.g., carbamazepine) with po- McGahuey, B.A.; Cindi Laukes, M.A.; Ghadeer Okayli, M.D.;
tential for inducing a reduction in platelet count are added or as Francisco Moreno, M.D.; Lynda Zentner, R.N.; and Pedro
a part of preoperative procedures.2,4 Delgado, M.D.

680
681 J Clin Psychiatry 61:9, September 2000