MANIC-DEPRESSIVE
`ILLNESS
`
`FREDERICK K. GOODWIN, M.D.
`Administrator
`
`Alcohol, Drug Abuse and Mental Health Administration
`and
`Senior Investigator
`National Institute of Mental Health
`
`KAY REDFIELD JAMISON, Ph.D.
`Associate Professor of Psychiatry
`The Johns Hopkins University School of Medicine
`
`New York Oxford
`OXFORD UNIVERSITY PRESS
`
`1990
`
`1 of 89
`
`Alkermes, Ex. 1065
`
`1 of 89
`
`Alkermes, Ex. 1065
`
`

`

`Oxford University Press
`
`Oxford New York Toronto
`Delhi Bombay Calcutta Madras Karachi
`Petaling Jaya
`Singapore HongKong Tokyo
`Nairobi Dares Salaam Cape Town
`Melbourme Auckland
`
`and associated companies in
`Berlin
`Ibadan
`
`Copyright © 1990 by Oxford University Press, Inc.
`
`Published by Oxford University Press, Inc.,
`200 Madison Avenue, New York, New York 10016
`
`Oxford is a registered trademark of Oxford University Press
`All rights reserved, No part of this publication may be reproduced,
`stored in a retrieval system, or transmitted, tn any form or by any means,
`electronic, mechanical, photocopying, recording, or otherwise,
`without the prior permission of Oxford University Press-
`
`Library of Congress Cataloging-in-Publication Data
`Goodwin, Frederick K,, 1936-
`Manic-depressive illness
`by Frederick K. Goodwin, Kay Redfield Jamison.
`
`p. em—Includes bibliographical references.
`ISBN 0-19-503934-3
`[. Jamison, Kay Redfield
`|. Manic-depressive psychoses.
`[DNLM: |. Bipolar Disorder, WM 207 G656m)
`RC316.G66
`1990 616.89'5—de20 DNLM/DLC for Library of Congress
`89-16396 CIP
`
`If, Title.
`
`987654321
`Printed in the United States of America
`on acid-free paper
`
`2 of 89
`
`Alkermes, Ex. 1065
`
`2 of 89
`
`Alkermes, Ex. 1065
`
`

`

`21
`
`Medical Treatment
`of Manic Episodes
`
`No one predicts how long it will be before the drugs take hold & [Robert Lowell]
`begins to be himself again. Meanwhile he writes and revises translations furiously
`and with a kind [of] crooked brilliance, and talks about himself in connection with
`Achilles, Alexander, Hart Crane, Hitler and Christ, and breaks your heart.
`—William Meredith!
`
`A patient in the throes of a manic episode can be
`intensely agitated, uncooperative, psychotic, ag-
`gressive, or dangerous. By the time the clinician
`is brought in, both patient and family are under-
`standably confused and distraught. The bizarre,
`frightening behavior obviously must be con-
`trolled humanely, but the clinician haslittle time
`to ponder available choices. Which drugsare best
`for this patient in this situation? Should the pa-
`tient be hospitalized? Should electroconvulsive
`therapy be used? Each decision calls for balanc-
`ing the ravages ofthe illness against the conse-
`quencesof intervention—a medication’s potency
`against
`its side effects,
`for example, or the
`patient’s safety against
`the stigma of hospi-
`talization.
`This chapter focuses on such issues in the med-
`ical management of acute manic episodes. Like
`others in this section, the chapter begins with a
`discussion of practical issues ofclinical manage-
`ment, an approach to treatment drawn from the
`research evidence and our own clinical experi-
`ence. The research literature is reviewed in the
`second part of the chapter, which some readers
`may choose to read first.
`Weare convinced that medical managementis
`necessary for all patients who are truly manic or
`are hypomanic and likely to become manic.
`Based on that assumption, we devote the follow-
`
`ing discussion largely to criteria for appropriate
`pharmacological treatment for acute mania. One
`important caveat is in order, however. Not all
`activated patients are necessarily manic, or even
`hypomanic, and not all mildly hypomanic pa-
`tients inevitably progress to mania. The line be-
`tween normal exuberance and clinical hypomania
`is sometimes difficult to discern, and clinicians
`must approach the task of differential diagnosis
`with care (see Chapters 4 and 5). Once the diag-
`nosis has been made, skillful psychological man-
`agement must accompany the drug treatment of
`emerging or acute mania, especially if the patient
`or family resists the idea of medications (see
`Chapter 25).
`the first of the modern antimanic
`Lithium,
`agents, remains the most important.Its therapeu-
`tic value was discovered by the Australian physi-
`cian John Cade (1949), whose post-World War I
`experiments with guinea pigs signaled a revolu-
`tion in the treatment of manic-depressiveillness.
`Several years were to pass before the importance
`of Cade’s pioneering work was recognized. Euro-
`pean psychiatrists began to take notice in 1954,
`when his observations were confirmed and ex-
`tended by Mogens Schou in Denmark. Although
`a handful of American psychiatrists were among
`the pioneers, lithium was not widely used in the
`United States until the late 1960s. This slow ac-
`
`603
`
`3 of 89
`
`Alkermes, Ex. 1065
`
`3 of 89
`
`Alkermes, Ex. 1065
`
`

`

`604
`
`ceptance was partly traceable to earlier adverse
`experiences with lithium as a salt substitute.
`Chlorpromazine, the prototypical antipsycho-
`tic medication for controlling the symptoms of
`schizophrenia, wasfirst used clinically for a psy-
`chiatric disorder in a manic patient (Schneider,
`1951, cited in Swazey, 1974), More extensive
`clinical observations in acutely manic patients
`followed (Lehmann and Hanrahan, 1954). Since
`lithium was still essentially unknown at
`that
`time, particularly to American psychiatrists,
`chlorpromazine quickly became the treatment
`of choice for acute mania. Haloperidol, a butyro-
`phenonethat also controls psychotic symptoms,
`wasintroducedin the late 1960s and was found to
`control psychotic behavior as effectively as
`chlorpromazine while producing less sedation
`and hypotension. As a result, many clinicians
`now prefer haloperidol and other high-potency
`neuroleptics, such as thiothixene.
`The use of anticonvulsant drugs to treat manic
`episodes dates back to the 1970s (Okumaet al.,
`1973). Some anticonvulsant drugs that have
`shown considerable therapeutic promise, particu-
`larly carbamazepine, clonazepam, and valproate,
`are already widely used with manic patients. Al-
`though not yet approved by the U.S. Food and
`Drug Administration for marketing as antimanic
`agents, they can, of course, be used by physicians
`at their own discretion.?
`
`CLINICAL MANAGEMENT
`
`Clinical Factors Influencing Drug Choices
`
`Clinical decisions in managing mania are influ-
`enced by the treatment setting,
`the nature and
`overall severity of the symptoms, and the pres-
`ence of medical complications. The following
`recommendations are based on findings of the
`studies reviewed later in this chapter, modified
`and amplified by our own clinical experience and
`that of colleagues we surveyed.
`
`Symptoms
`The most important consideration in choosing a
`treatment for manic symptomsis their nature and
`severity. Mild manic symptoms (hypomania or
`stage-[ mania) usually respond well to lithium
`alone. Restoring a normalsleep pattern (Hudson
`
`TREATMENT
`
`et al., 1989) can often avert escalation to more
`severe stages of mania. This might be accom-
`plished by using an adjunctive sedative hypnotic,
`such as
`the benzodiazepines clonazepam or
`lorazepam, during the evening.
`A neuroleptic may be needed to control severe
`symptoms, particularly gross hyperactivity and
`psychotic features. Whether to chose a neurolep-
`tic of high potency (e.g., haloperidol, thiothix-
`ene) or low potency (e.g., chlorpromazine, thio-
`ridazine)is still an unsettled issue. High-potency
`drugs have a relatively low level of hypotensive
`and sedative side effects, a feature that allows
`more rapid initial dose escalation and, therefore,
`presumably more rapid control of the psychosis.
`Low-potency neuroleptics, on the other hand, are
`more sedating—actually an advantage in achiev-
`ing early control of the acute mania. In addition,
`low-potency drugs carry less of a risk of extra-
`pyramidal effects,* including tardive dyskinesia,
`and neurotoxic reactions, and also the rare neu-
`roleptic malignant
`syndrome* (Casey, 1984;
`Pope et al., 1986).
`Both the research literature and our own clini-
`cal experience suggest that the anticonvulsants
`and neuroleptics are superior to lithium in the
`early phase of treating severe mania, that is, dur-
`ing the first week or two. After the first 2 weeks,
`lithium and, perhaps, carbamazepine are more
`effective than neuroleptics. Because of their
`greater specificity,
`lithium and carbamazepine
`calm the patient with a minimum ofsedation and
`nonspecific tranquilization. These drugs are also
`superior because theyareless likely to be associ-
`ated with postmania depressions and, even more
`important, carry no appreciable risk of tardive
`dyskinesia,
`The properrole of the anticonvulsantsin treat-
`ing acute mania has not yet been fully estab-
`lished. As reviewed later, carbamazepine is
`clearly effective, even when usedalone (although
`in most trials it was given in combination with
`lithium or neuroleptics). Existing data suggest
`that carbamazepine may be as effective in acute
`mania as lithium or neuroleptics, but its overall
`efficacy requires more study. Compared with
`lithium, carbamazepine is similar in its relative
`specificity against the affective core of mania and
`often faster in achieving its antimanic effects.
`Less clear is whether it can match the effective-
`ness of neuroleptics in the short-term control of
`
`4 of 89
`
`Alkermes, Ex. 1065
`
`4 of 89
`
`Alkermes, Ex. 1065
`
`

`

`MEDICAL TREATMENT OF MANIC EPISODES
`
`the extreme hyperactivity seen in psychotic ma-
`nia, although some evidence is encouraging.
`As a treatment for manic-depressive illness,
`carbamazepine is best established as an alterna-
`tive for patients who do not respondto lithium or
`cannot tolerate it. Thus, carbamazepine is the
`treatment of choice for managing acute mania in
`patients with a history oflithium-resistant rapid
`cycles, lithium failure or intolerance, or kidney
`dysfunction, Because of its antidepressant prop-
`erties, carbamazepine, alone or combined with
`lithium, may be particularly useful in the acute
`treatment of mixed states, which may not respond
`well to lithium alone (Secundaet al., 1985), Be-
`cause it lessens aggression, carbamazepine may
`also be a good choice for suicidal patients. Until
`further information is available, the other anti-
`convulsants should generally be reserved for pa-
`tients who do not respond satisfactorily to car-
`bamazepine. A possible exception to this rule
`may be clonazepam, which, because ofits seda-
`tive profile and safety, can be an important ad-
`junctin the initial treatment of mania.®
`
`Setting
`The treatmentsetting also influences the choice
`ofdrugs or electroconvulsive therapy (ECT), Ma-
`nia subsides more gradually with lithium than
`with neuroleptics, the anticonvulsants, or ECT.
`This lithium lag, 7 to 12 days when the maniais
`moderate to severe, might be tolerable in a well-
`staffed inpatient research unit, but very rapid con-
`trol of symptomshaspriority in mostsettings and
`is clearly a necessity in some, such as an emer-
`gency room without a closed psychiatric unit for
`backup.’ In these settings, neuroleptics and/or
`anticonvulsants (or, selectively, ECT) are prefer-
`able for highly agitated patients, A decision tree
`outlining the choice of treatments for mania is
`illustrated in Figure 21-1.
`
`Contraindications
`
`Medical conditions or medication needs some-
`times limit the choice of drugs.8 Although weare
`concerned here with the short-term use of drugsin
`treating acute mania,
`the medical factors dis-
`cussed subsequently are also relevant to discus-
`sions of long-term prophylactic treatment (see
`Chapter 23). Medical contraindications for anti-
`manic drugs, although rare, must alwaysbe bal-
`
`605
`
`anced against the risks of untreated mania. Table
`21-1 lists, in approximate rank order, contrain-
`dications to antimanic drugs. (The subjective and
`behavioral side effects oflithium andits effect on
`organ systemsare fully reviewed in Chapter 23.)
`Impaired kidney functionis a relative contrain-
`dication for lithium treatment becauselithium is
`eliminated principally through the kidney and can
`influence renal tubular activity. Lithium can be
`used for patients with moderate or stable impair-
`ment, but
`the blood level should be carefully
`monitored, since a therapeutic level usually can
`be reached with lower doses than those needed for
`patients with normally functioning kidneys. Car-
`bamazepine can be substituted for lithium when
`severe renal impairment precludesits use.
`Cardiac disease is another important con-
`siderationin treating mania. Byvirtue ofits ionic
`properties and especially its ability to substitute
`for potassium, lithium produces changes in the
`electrocardiogram (particularly T-wave flatten-
`ing) that are generally benign and reversible.
`There are, however, rare and scattered case re-
`ports of patients with certain kinds of cardiac pa-
`thology who experience lithium-induced com-
`plications (Jefferson et al., 1987).
`Myocardial infarction requires a balancing of
`risks, Lithium can conceivably produce com-
`plications in an already compromised myocar-
`dium (primarily because it can increase irri-
`tability). This risk must be weighed against
`possibly even greater risks, such as the effect of
`the untreated manic patient’s uncontrolled ac-
`tivity, psychophysiological stress, and uncertain
`compliance with cardiac medication, as well as
`the hypotension that may result from taking neu-
`roleptics. Lithium should,
`therefore, be con-
`sidered for managing a manic or hypomanic epi-
`sode, even during or shortly after myocardial
`infarction. Carbamazepine or perhaps clonaze-
`pam may provide useful alternatives to lithium or
`neurolepticsin this situation. (For comprehensive
`reviews of the cardiac effects of lithium, see Al-
`brecht and Miiller-Oerlinghausen, 1980; Jeffer-
`son et al., 1987.)
`Neurological conditions that influence treat-
`ment decisions in mania include epilepsy, parkin-
`sonism, dementia, cerebellar disease, and my-
`asthenia gravis. The risk of neuroleptic-induced
`tardive dyskinesia increases with age, particular-
`ly in women. In addition, the risk appears to be
`
`5 of 89
`
`Alkermes, Ex. 1065
`
`5 of 89
`
`Alkermes, Ex. 1065
`
`

`

`606
`
`TREATMENT
`
`Hypomania
`
`(Stage - | or - 1] Mania)
`
`Lithium
`
`Lithium
`(+ Clonazepam
`for sleep)
`
`beng
`Lithium +
`
`Carbamazepine 5
`se
`Clonazepam
`or
`Valproate =
`
`Mania
`
`(Stage - Il or - lil)
`Lithium
`(+ Clonazepam
`for sleep)
`
`Lithium + Neuroleptic
`
`Lithium+ { Valproate?
`See * ae
`
`Carbamazepine ®
`or
`or
`Clonazepam 4
`
`Figure 21-1. A treatment decision tree for mania. *The anticonvulsants are morelikely to be indicated when there
`is a history of rapid cycles, lithium resistance, or temporal lobe—like symptoms.
`
`substantially greater for patients with affective
`illness than for those with schizophrenia (Casey,
`1984). Intermittent use of a neuroleptic, more
`typical in manic-depressive illness than in schizo-
`phrenia, may also be associated with a greater
`risk of tardive dyskinesia, but this association is
`controversial.
`Neither lithium nor the neuroleptics are con-
`traindicated for acute mania in patients with
`classic epilepsy, although both drugs can produce
`activation of the electroencephalogram (EEG).
`The obvious choice for treating manic-depressive
`illness in patients with seizure disorders is car-
`bamazepine, which has anticonvulsant activity,
`Lithium can aggravate preexisting Parkinson’s
`disease, an effect that is not surprising, since
`lithium decreases dopamine synthesis in the brain
`(see Chapter 13) (Makeeva et al., 1974).? Car-
`bamazepine, which does not markedly affect the
`
`dopamine system,is preferable to neuroleptics in
`managing the mania that can emerge when par-
`kinsonian patients are treated with L-dopa. It is
`also best for manic patients with preexisting tar-
`dive dyskinesia.
`Neuroleptics or anticonvulsants may be better
`than lithium for manic patients with dementia,
`cerebellar disease, or other pathology of the CNS
`because lithium is more likely to intensify the
`underlying dysfunctions. However, some pa-
`tients with dementia are particularly sensitive to
`the organic confusional effects of neuroleptics or
`anticonvulsants. In the neuroleptics, this effectis
`probably due to their potent hypotensive action.
`The tendency of lithium to produce muscle
`weakness makes it unsuitable for treating manic
`patients with myasthenia gravis. It has been used
`successfully to treat the pathological mood la-
`bility associated with multiple sclerosis without
`
`6 of 89
`
`Alkermes, Ex. 1065
`
`6 of 89
`
`Alkermes, Ex. 1065
`
`

`

`MEDICAL TREATMENT OF MANIC EPISODES
`
`Table 21-1. Relative Contraindications
`for Antimanic Drugs
`
`Lithium
`
`Usually coniraindicated:
`Renal function impairment
`Acute myocardial infarction
`Myasthenia gravis
`Pregnancy - 1°" trimester
`Breast feeding
`Compromisedfluid or sall balance
`
`Use with close medical supervision, including limited
`dosage:
`Other cardiac pathology
`Parkinson's disease
`Pregnancy - 2% or 3" trimester
`Delivery
`Epilepsy
`Thyroid disorders
`
`Use with caution, including limited dosage:
`Cerebellar disorders
`Dementia
`Other CNS disorders
`Diabetes mellitus
`Ulcerative colitis
`Psoriasis
`Senile cataracts
`Osteoporosis
`Certain drugs (see text)
`
`Neuroleptics
`
`Myocardial infarction
`Parkinson's disease
`Compromisedliver function
`Porphyria
`Hypotension
`Tardive dyskinesia
`
`Carbamazepine
`
`Compromised liver function
`Porphyria
`Hematopoietic system abnormality
`A-V block
`
`Clonazepam
`
`Neurological disorders affecting balance
`CNS depression
`
`4 previous history of allergic reaction to any antimanic drug
`would be a contraindication for that particular drug.
`
`aggravating the neurological disorder(see, e.g.,
`Kemp et al., 1977), although such patients may
`have a lower threshold for someof lithium’s side
`effects in the CNS.
`Other medical conditions also may be affected
`by drug treatments for mania. Neuroleptics and
`perhaps carbamazepine should be ruled out for
`
`607
`
`patients with compromised liver function and
`porphyria, for example. Carbamazepine and the
`new atypical neuroleptic, clozapine, both have
`been associated with bone marrow suppression
`and should be avoided in patients with disturbed
`hematopoietic function. Although lithium is not
`contraindicated for patients with diabetes,
`the
`disease process should be monitored closely once
`the drug is started since it has been reported to
`exacerbate diabetes, especially in patients taking
`it for several years (see, e.g., Mellerup et al.,
`1983).
`Thyroid disease can be aggravated by the
`chronic use oflithium, butin the relatively brief
`acute treatment phase, the administration of thy-
`roid hormone can offset any effects of lithium.
`Hypothyroidism may also contribute to inade-
`quate lithium response. One severely manic pa-
`tient,
`for example, was unresponsive to a
`lithium—neuroleptic combination until after her
`hypothyroidism was corrected (Balldin et al.,
`1987). Similarly, postpartum mania may be asso-
`ciated with poor lithium response (Targum et al.,
`1979), which may be caused by a correctable low
`estrogen state (Wehr and Goodwin, 1981). Con-
`ditions in which electrolyte imbalance exists,
`such as severe diarrhea, complicate the use of
`lithium and perhaps also of carbamazepine, and
`neuroleptics might be favored. Any abnormality
`in the hematopoietic system may complicate the
`use of carbamazepine. To our knowledge, there
`are no medical contraindications to the use of
`clonazepam or other benzodiazepines.
`
`Pregnancy
`Birth defects, principally involving the cardiac
`system, occurat rates that are significantly higher
`than normal rates in babies whose mothers re-
`ceived lithium in the first 3 months of pregnancy.
`Thus, lithium should be avoided during thefirst
`trimester whenever possible. Mild manic epi-
`sodes during pregnancy should probably be man-
`aged without drugs, butit is prudent to treat more
`severe episodes, since the possible consequences
`of an untreated episode (such as injury, psycho-
`physiological stress, dehydration and malnutri-
`tion, profound sleep deprivation, and suicide)
`could pose a greaterrisk to the fetus than the side
`effects of lithium. The risk—benefit considera-
`tions for the use of lithium during pregnancy are
`thoroughly reviewed in Chapter 23. Clonazepam
`
`7 of 89
`
`Alkermes, Ex. 1065
`
`7 of 89
`
`Alkermes, Ex. 1065
`
`

`

`608
`
`is not known to be associated with fetal abnor-
`malities and, therefore, might be used in these
`circumstances. Another option is ECT, which
`can be used without specialrisk to the fetus. The
`clinical management of mania during pregnancy
`has been reviewed by Nurnberg (1980) and by
`Sitland-Marken and colleagues (1989).
`
`Concurrent Medications
`Although several drugs interact with lithium,
`neuroleptics, and the anticonvulsants, only a few
`combinations are contraindicated (see Table 23-5
`and discussion in this chapter). Knowledge of
`these interactions will influence the choice of one
`drug over another, but potential drug interactions
`generally should not take precedence over the
`clinical indications outlined previously.
`The concurrentuse oflithium and diuretics de-
`serves special attention. Loop diuretics, such as
`furosemide, do not substantially alter lithium ex-
`cretion and can be administered together safely
`(Saffer and Coppen, 1983; Jefferson et al.,
`1987). The thiazide drugs are more problematic,
`since they decrease tubular reabsorption of so-
`dium andindirectly increaselithium reabsorption
`and decrease its excretion. When these drugs are
`used, lithium should be started at a low dose and
`increased very gradually, with frequent monitor-
`ing of the bloodlevel.
`Other medical drugs with potential lithium in-
`teractions includeanti-inflammatory agents, such
`as indomethacin and phenylbutazone, which in-
`creaselithium levels (Reimannet al., 1983); car-
`diovascular medications, especially the anti-
`hypertensive methyldopa, which decreases renal
`clearance; and digoxin, which has been shown
`to reduce the acute manic efficacy of lithium
`(Chamberset al., 1982). Finally, some antibio-
`tics prescribed for lithium-associated acne have
`nephrotoxic potential.
`Because they compete for hepatic metabolism,
`certain drugs may significantly increase car-
`bamazepine blood levels and produce toxicity.
`Consequently, the combination of valproic acid
`and carbamazepine is contraindicated (Meyer et
`al., 1984; Lambert and Venaud, 1987; Meijer et
`al,, 1984). Among the other drugs that should be
`used cautiously with carbamazepinefor this same
`reason are verapamil, isoniazid, diltiazem, and
`erythromycin and related antibiotics (Berrettini,
`1986; Sovner, 1988). By contrast, other drugs—
`
`TREATMENT
`
`phenobarbital, primidone, and phenytoin—can
`decrease carbamazepine blood levels, presum-
`ably by inducing hepatic metabolism (Post etal.,
`1985).
`Concurrent administration of carbamazepine
`and neuroleptics has been reported in more than
`100 patients. The two drugs produce some addi-
`tive effects in the CNS, and there is some evi-
`dence that they enhance each other's effects. Car-
`bamazepine does not appear to alter lithium
`levels. Additive CNSeffects, especially sedation
`and cognitive and memory functions, should be
`kept in mind when deciding how fast to increase
`dosages and the ultimate dose level, Patients with
`preexisting CNS disease may be especially vul-
`nerable to neurotoxicity with this combination
`(Shukla et al., 1984).
`
`Determining Medication Dosage
`
`Neuroleptics
`Clinicians traditionally have used larger doses of
`neuroleptics for acute mania than for schizo-
`phrenia, but recent experience suggests that more
`modest doses can be effective. The lower doseis
`feasible if the patient is carefully monitored for
`early signs of improvement and takes lithium
`along with the neuroleptic. Chlorpromazine
`doses averaged more than | g per day in con-
`trolled studies, and comparably high doses have
`been reported for the high-potency neuroleptics,
`such as haloperidol and thiothixene. Blood level
`determinations for neuroleptics are not yet rou-
`tinely available as they are for lithium. Clinical
`state, age, sex, and weight must be considered in
`setting dose levels; higher doses are required for
`more disturbed and highly active patients and
`for patients who are male, young, or heavy.
`Haloperidolis usually started at 5 to 15 mg intra-
`muscularly (or 10 to 25 mg orally) every 4 to 6
`hours. For chlorpromazine, the preferred dosage
`is 50 to 100 mg, which can be administered intra-
`muscularly every 6 hours and then gradually re-
`placed by oral doses.!° The need for such high
`doses of neuroleptics should be reevaluated con-
`tinually throughout treatment of the acute manic
`episode. To minimize the possibility of neurotox-
`icity, extrapyramidal side effects, or postmania
`depression, dosage should be reduced as soon as
`manic symptoms begin to subside.
`
`8 of 89
`
`Alkermes, Ex. 1065
`
`8 of 89
`
`Alkermes, Ex. 1065
`
`

`

`MEDICAL TREATMENT OF MANIC EPISODES
`
`609
`
`Lithium
`The gap between therapeutic and toxic levels of
`lithium is the narrowest of any drug routinely
`used in psychiatry. Fortunately,
`the level of
`lithium in plasmais readily determined, and dos-
`age requirements have been studied extensively.
`In managing acute mania with lithium alone,it 1s
`best to use a dosage schedule that produces the
`highest plasma level consistent with acceptable
`side effects. These bloodlevels usually are higher
`than those considered necessary or safe for main-
`tenance therapy. The dose/blood level relation-
`ship is influenced by the individual's sex, age,
`weight (especially muscle mass), salt
`intake,
`amountof sweat, intrinsic renal clearance capaci-
`ty for lithium, and, as noted, other medications.
`A relatively higher dose/bloodlevel ratio is asso-
`ciated with being younger, male, and heavier and
`having a highersalt intake.
`the
`In the lithium treatment of acute mania,
`patient's clinical state is one of the most impor-
`tant factors affecting the dose/blood level rela-
`tionship. Some patients, when manic,
`retain
`lithium in body pools outside the plasma, proba-
`bly largely in bone (Greenspan et al., 1968; Almy
`and Taylor, 1973). In practice, more lithium is
`needed to achieve a given blood level during ma-
`nia than during euthymiaor depression (Goodwin
`et al., 1969; Serry, 1969; Kukopulos et al.,
`1985). When mania begins to subside, a dosage
`reduction usually is necessary to avoid lithium
`toxicity. Obviously, blood levels should be moni-
`tored more frequently when theclinical state is
`changing, especially from mania to euthymia or
`depression.
`To predict dosage requirements, some inves-
`tigators recommend a test dose of lithium fol-
`lowed 24 hourslater by a plasma level determina-
`tion (Cooper and Simpson, 1976; Perry et al.,
`1984). Fava and colleagues (1984) showed that,
`by using this technique, therapeutic levels were
`obtained faster, and fewer blood level determina-
`tions were required. Although this technique
`probably can be applied reliably when the mood
`State is stable, its practical value in treating acute
`mania is limited by the state-dependent kinetics
`of lithium. Errors in the predicted dose may, for
`example, be due to changesin patients’ sleep and
`activity, which presumably cause changes in re-
`nal clearance (Perry et al., 1984). In addition, use
`
`of this method necessitates a 24-hour delay in
`treatment. Norman and colleagues (1982) pro-
`posed a faster technique that can also account for
`changes in renal clearance. This technique may
`be impractical, however, since it requires a
`4-hour urine collection along with a blood
`sample.
`The plasmaleveloflithium needed to produce
`a clinical response differs substantially from one
`manic patient to another. The same is true of
`toxicity. These differences are partly caused by
`variability in tissue sensitivity, a variability en-
`countered with any drug. More important, how-
`ever, are individual differences in the ratio of
`plasmalithium to intracellular lithium, as re-
`flected in red blood cell (RBC) determinations.
`Toxic reactions
`reflect
`intracellular
`lithium,
`whereas serum levels reflect only the extracellu-
`lar compartment.
`These issues are important to treatment be-
`cause increasing plasma levels of lithium (up to
`1.4 mEq/liter)
`are
`associated with propor-
`tionately higher rates of therapeutic response
`(Stokes et al., 1976), Although there is reason to
`push the dose in patients who fail to respond,
`blood levels above 1.5 mEq/liter are not gener-
`ally recommended, and even levels between 1.2
`and 1.5 mEq/liter require considerable care
`to avoid toxicity.
`Indeed, an increase in the
`RBC/plasmalithium ratio often precedes the de-
`velopment of neurotoxicity (see, e.g., Dunneret
`al., 1978; Carroll and Feinberg, 1977). In most
`cases, blood levels in the therapeutic range can be
`achieved at doses between 900 and 1,800 mg
`daily of lithium carbonate.
`In deciding the maximum lithium level to use
`with a manic patient, the clinician should keep in
`mind that
`the most
`important potential
`toxic
`effects are those involving the CNS. This task is
`made moredifficult by the fact that the delirium-
`like symptoms that can occur in severe mania
`may be nearly indistinguishable from neurotoxic
`effects. (Specific neurotoxic effects oflithium are
`discussed in Chapter 23.)
`Some authors have suggested using a loading
`dose strategy for treating mania with lithium,'!
`both to achieve the maximum bloodlevel quickly
`and to speed therapeutic onset. The value ofthis
`strategy is questionable, however, especially in
`light of animal and human data indicating that
`lithium is slow to enter the brain from the blood,
`
`9 of 89
`
`Alkermes, Ex. 1065
`
`9 of 89
`
`Alkermes, Ex. 1065
`
`

`

`610
`
`TREATMENT
`
`even when plasmalevels are high. In one study,
`CSFlithium levels increased 50 percent, on aver-
`age, from thefirst to the third week on a constant
`lithium dose (Rey et al., 1979).
`
`Lithium Plus Neuroleptics
`The additive and possibly synergistic effects of
`lithium and neuroleptics must be considered
`when combining the two drugs, A severe enceph-
`alopathy syndrome was first reported in four
`manic patients treated with high doses of both
`lithium and haloperidol by Cohen and Cohen in
`1974, Since then, some 50 additional cases of
`neurotoxic syndromes resulting from the com-
`bination of lithium and a neuroleptic have been
`reported. Most of these conditions are reversible.
`On the other hand, eight prospective and retro-
`spective studies with a total of more than 600
`patients have generally failed to find any special
`neurotoxicity with this combination.!? This liter-
`ature suggests that the risk of neurotoxicity is
`associated with pre-existing encephalopathy and
`high dose levels, especially of the neuroleptics. '9
`Thus neuroleptics should be used in substantially
`lower dosages when combined withlithium than
`when used alone. We also recommendthat the
`lithium level be kept below 1.0 mEq/liter, in part
`because neuroleptics increase the RBC/plasma
`lithium ratio (Von Knorring et al., 1982), Al-
`thoughlithium and neuroleptics generally can be
`combined safely and effectively when done in
`this way, it is important to monitor CNS function
`and, in hospital settings,to alert the staff to watch
`for symptomsof neurotoxicity. Patients in seclu-
`sion rooms, who can rapidly become dehydrated,
`require special caution,
`including temperature
`monitoring (see later discussion of seclusion and
`restraints). One report of a high frequency of neu-
`rotoxicity with lithium—neuroleptic combina-
`tions in people over 65 suggests caution in this
`age group as well (Miller et al., 1986).
`
`Carbamazepine and Valproate
`When carbamazepine is used alone, the starting
`doseis usually 200 to 400 mg, which is increased
`to the 800 to 1,000 mg range during thefirst
`week. Further increases (up to about 1,600 mg)
`are appropriate if no responseis evidentafter the
`first 2 weeks and if not limited by unacceptable
`side effects, The blood level generally should be
`between 6 and 12 ng/ml. When carbamazepineis
`
`combined with lithium or neuroleptics, the dose
`and target blood level are typically somewhat
`lower, Over time, carbamazepine can Induceits
`own hepatic metabolism, and blood levels can
`fall, This problem is more troublesome in pro-
`phylactic treatment (see Chapter 23).
`Side effects are more likely to occur when dos-
`ages are increased rapidly in treating acute mania
`than when the dosageis built up slowly in thefirst
`phase of prophylactic treatment. These early side
`effects—drowsiness, dizziness, ataxia, confu-
`
`sion, double vision, and nausea—usually do not
`persist beyond the first week or two and often
`respond to temporary dosage reduction.
`Carbamazepine producesside effects aboutas
`frequently as lithium and less often than neu-
`roleptics. Skin rashes of varying degreesof sever-
`ity are a frequent problem (10 to 15 percent of
`patients). Those that are unaccompanied by evi-
`dence of a systemic allergic response can be treat-
`ed with 20 to 30 mg of prednisone administered
`daily for a few weeks, then gradually discon-
`tinued. Liver enzyme levels, complete blood
`count, and platelet coun