<<

. 53
( 68 .)



>>

advantages (Perugi et al., 1997).
The same problem of heterogeneous definitions of mixed state is true for studies
trying to characterize biological differences between mixed patients, manic
patients, and healthy controls. This is also true for rapid-cycling patients. Rapid
cycling has been defined by having four or more mood episodes within 1 year. This
definition has been made according to a pivotal trial conducted by Dunner and
Cambridge University Press, 2005.
#
312 H. Grunze and J. Walden


`
Fieve (1974) showing a rapid decline of lithium responsiveness with a cut-off at
four episodes. Recent observations of large patient samples however question this
˜˜magic line™™ defined by four episodes. Data from the Stanley Foundation Bipolar
Network (Kupka et al., unpublished data) demonstrate that there is a continuous
decline of treatment response with increasing number of episodes per year. Rapid
cycling also includes some rare manifestations which appear to have a highly
biological background, probably coupled to the circadian rhythm and Zeitgeber.
However, these patients with 48-rapid cycling only represent a very narrow
selection of rapid-cycling patients at large and conclusions from their biology
may be not transferable to rapid-cycling patients with a chaotic and irregular
course of illness.
The reason why it still makes sense to review the biology of rapid cycling and
mixed states in one chapter is that some aspects of these manifestations may be in
common. On a theoretical level, mixed states not following this strict definition of
DSM-IV but the wider definition of the Tenth Revision of International Classification
of Diseases (ICD-10: World Health Organization, 1992), which also allows timely
separation of depressive and manic syndromes, may also be called ultradian rapid
cycling. On the level of treatment, evidence from several open and some controlled
trials points towards common principles in drug treatment. It appears to be general
consensus that lithium is not very effective in both conditions, whereas antiepileptic
drugs, especially valproate and lamotrigine, and also atypical antipsychotics such as
olanzapine and risperidone may be more efficacious. To understand aspects of the
biology of rapid cycling and mixed states, another possible approach would there-
fore be to look for mechanisms of action of these drugs which are not or only
partially shared by lithium. We will use this approach in the second part of this
chapter.


Implications of catecholamines on mixed states and rapid cycling
The first level to be considered is the cell membrane and the action of different
neurotransmitters that may play a role in affective states. For bipolar disorder in
general, results are still conflicting about changes in neurotransmission. Increased
cortical norepinephrine (noradrenaline) and decreased 5-hydroxytryptamine and
dopamine turnover has been described in bipolar patients (Manji and Potter, 1997).
Within manic patients, the central nervous system levels of norepinephrine appear
to increase with the degree of dysphoria, anger, and anxiety (Post et al., 1989). Also a
central serotonergic deficit, in both manic and depressed patients, has been sug-
gested (Meltzer and Lowy, 1987) which may improve after treatment with valproate
(Maes et al., 1997). Tandon et al. (1988) compared selected cerebrospinal fluid
(CSF) parameters from patients with bipolar disorder, mixed, to those with mania
313 Rapid cycling and mixed states


and major depression. Fourteen patients in each category (DSM-III) were studied
with regard to the dopamine metabolite homovanillic acid (HVA) and the seroto-
nergic metabolite 5-hydroxyindoleacetic acid (5HIAA) in CSF under carefully
controlled conditions. Both CSF HVA and 5HIAA were found to be significantly
higher in manic than in major depressive patients. Discriminant analysis of the
biochemical variables of the mixed affective group identified two biochemically
distinct and clinically different subgroups of seven patients each, one resembling
the manic group and the other the major depressive group. Tandon et al. conclude
from these findings that mixed affective states do not exist as a separate entity,
but are composed of two subgroups obtained from the manic and major depressive
categories “ a view that remains controversial (McElroy et al., 2000).
However, other studies make a strong monoaminergic link to mixed states less
likely, as they reported no change in either 5-HT2A or 1A receptor density (Dean
et al., 2001) and, as far as the noradrenergic action is concerned, no change in beta-
receptor density (Werstiuk et al., 1990).
As bipolar illness is quite heterogeneous, more precise results may be expected
in a homologous group of patients. Therefore, several studies looked at a mono-
aminergic imbalance in 48-h rapid-cycling patients.
Characteristic and reproducible changes in norepinephrine and metanephrine
excretion, closely related to mood swings, were described by Juckel et al. (2000). In
general, the urinary excretion of norepinephrine and metanephrine was increased on
both manic and depressed days, with higher values during mania, and was generally
ameliorated after successful valproate treatment. As far as mixed states are concerned,
Swann et al. (1994) reported on increased 3-methoxy-4-hydroxyphenylethyleneglycol
excretion in mixed states. As especially tricyclic noradrenergic acting antidepressants
are capable of inducing a switch from pure depression into mixed states or mania, and
may also induce a rapid-cycling course, norepinephrine and its metabolites may play
at least a strongly modulating role in these conditions.
Concerning an impact of the serotonergic system, a very high and, compared to
bipolar disorder in general, increased comorbidity has been described between
mixed states, obsessive-compulsive disorder, and anxiety disorders, which are
generally considered as serotonergic disorders. Thus, for mixed states the seroto-
nergic system may also play a role. From clinical observation, this appears less
likely for rapid cycling, as selective serotonin reuptake inhibitors are considered a
generally safe treatment for possible induction of switch or rapid cycling.
To our knowledge, mixed states and rapid cycling have not yet been explored with
a special focus on the dopaminergic system, with the exception of the cited study of
Tandon et al. (1988). In general, a dopamine hypothesis of mania has been proposed
by several authors (Diehl and Gershon, 1992; Buki and Goodnick, 1998) and genetic
aberrations of dopamine receptors and transporters have been demonstrated in
314 H. Grunze and J. Walden


bipolar patients (Kelsoe et al., 1996; Manki et al., 1996; Waldman et al., 1997).
Recently, the D2-receptor region came out as a candidate locus associated specifically
with bipolar disorder (Massat et al., 2002). Decreased presynaptic dopamine
function in the basal ganglia after successful treatment of mania with valproate
has been demonstrated in a recent positron emission tomography study (Yatham
et al., 2001). However, so far no study on the dopaminergic system seems to have a
special focus on rapid cycling.
Taken together, data on aberrations of biogenic amines in rapid cycling and
mixed states are sparse, with the best evidence so far existing for distinct abnormal-
ities of the noradrenergic metabolism, at least in 48-h ultrarapid-cycling patients.
Generalizing these results to rapid cycling at large and mixed states, however,
would be premature.


Implications of hormonal aberrations on mixed states and rapid cycling
The case report of Juckel et al. (2000) not only looked into biogenic amines, but also
into changes of the limbic“hypothalamic“pituitary“adrenocortical axis in this 48-h
rapid-cycling patient. Changes of both human growth hormone and cortisol were
quite dramatic, with peaks during mania and troughs during depression. Again,
successful valproate treatment ameliorated this rollercoaster of the LHPA axis.
As far as mixed states are concerned, a small study of Cassidy et al. (1998)
compared seven mixed patients with purely manic patients concerning their
plasma dexamethasone concentration and cortisol response in the dexamethasone
suppression test (DST) during manic episodes. Measuring these parameters at 3
and 10 p.m., there was a tendency to decreased dexamethasone and increased
cortisol levels in the mixed group. Other studies, unfortunately not exceeding 10
patients, reconfirm a tendency for increased DST non-suppression (Evans and
Nemeroff, 1983; Krishnan et al., 1983).
Swann et al. (1992) investigated HPA function and its relationship to clinical
state in 19 hospitalized manic patients meeting Schedule for Affective Disorders
and Schizophrenia “ Research Diagnostic Criteria for acute manic episodes,
compared patients with and without a mixed presentation, and examined the
correlation between HPA activity and behavior. In this study, data were available
from 13“16 patients. Patients with mania had elevated CSF and urinary free
cortisol excretion compared with healthy subjects, and did not differ from
depressed patients in any cortisol measures. Mixed manic patients had signifi-
cantly higher morning plasma cortisol, postdexamethasone plasma cortisol and
CSF cortisol than pure manics. Five of seven mixed manics and three of nine pure
manics were DST non-suppressors. Afternoon plasma cortisol and CSF cortisol
correlated significantly with depressed mood; urinary free cortisol correlated with
315 Rapid cycling and mixed states


anxiety. None of the cortisol measures correlated with mania or agitation scores.
From these data, the authors suggest that increased cortisol secretion is a char-
acteristic of the depressed state in mixed manics, although pure manics may also
have increased DST non-suppression.
Thyroid dysfunction has been implicated by several authors in rapid-cycling
patients. Sack et al. (1988) reported on decreased nocturnal thyroid-stimulating
hormone (TSH) secretion in rapid-cycling patients. Hypothyroid metabolism has
also been reported by several authors in rapid cyclers (for a review, see Joffe and
Sokolov, 1997), and low basal TSH levels also appear to be associated with a higher
switch risk when bipolar depressed patients are exposed to antidepressants
(Bottlender et al., 2000). Preliminary clinical data (Bauer and Whybrow, 1990)
suggest that thyroxine addition may be a helpful augmentation strategy in refrac-
tory rapid-cycling patients. For mixed patients, however, a correlation to thyroid
dysfunction has not been convincingly observed so far. Joffe et al. (1994) found no
difference in the frequency of grade II subclinical hypothyroidsm or thyroid
hormone level between mixed (n = 10) and non-mixed (n = 57) manic patients.
However, a study of Zarate et al. (1997) in first-episode manic patients, 15 mixed
and 57 pure, showed a greater likelihood for elevated TSH in mixed patients, and a
small study of Chang points in the same direction of hypothyroidism in mixed
patients (Chang et al., 1998). In the largest trial looking for thyroid antibodies in
different mood states in 226 bipolar patients (including 28 patients with mixed
states), Kupka et al. (2002) found no difference in the number of thyroid antibody-
positive patients between euthymic, depressed, and mixed patients.
Besides these cited studies, there also several case reports on changes of neuro-
endocrinological abnormalities in rapid-cycling bipolar patients and their ameli-
oration with remission (e.g., Shimizu et al., 1997). But all these reports and studies
do not solve a general problem, and that is causality. Are aberrations of the
metabolism of biogenic amines and hormonal changes the cause of mixed states
and rapid cycling or are they simply effects of a different mood state? Or are they
even completely independent from mood and represent a variable triggered by
another unknown, underlying mechanism which also affects the mood state? Thus,
not only the evidence but also the causality of the impact of neurotransmitters
and hormones remains weak on the course of mixed state and rapid cycling.


The impact of transmembranous ion fluxes on rapid cycling and mixed states
Besides acting on different neurotransmitters (5-hydroxytryptamine, dopamine,
gamma-aminobutyric acid, and glutamate: for the latter two no studies have
so far been done for rapid cycling and mixed states), antiepileptic drugs mainly
target transmembranous ion fluxes. For carbamazepine, valproate, and lamotrigine,
316 H. Grunze and J. Walden


blockade of fast sodium inward currents may be a decisive mechanism of anti-
epileptic action. For bipolar disorder in general, calcium (and probably potassium)
fluxes may be the more important target.


A potential role of calcium in bipolar disorder?
Mobilization of calcium is a key event in presynaptic and postsynaptic signalling
and also in lasting neuronal changes, as long-term potentiation.
Increased intracellular calcium concentrations, under baseline conditions or after
mobilization following specific stimulation paradigms, are a solid finding in platelets
and lymphocytes of bipolar patients, in both manic and depressive episodes (cf.
Grunze et al., 1997; Hough et al., 1999). If we allow the speculative assumption that
these findings in peripheral cells also reflect the neuronal environment, we suggest
the following simplified model: mild elevations of intracellular calcium activate
metabolic processes, by activating adenylate cyclase. This increases, beside other
cyclic adenosine monophosphate-activated protein kinase-dependent steps, the
synthesis of catecholamines by phosphorylation of tyrosine hydroxylase, leading to
increased neuronal excitability. Excitability may also be increased by a partial
inhibition of the Na, K-ATPase activity (el Mallakh and Wyatt, 1995). The clinical
counterpart may be a manic syndrome. Further increase of intracellular calcium,
however, dampens the adenylate cyclase activity even below its normal level,
decreasing catecholamine synthesis, and, hypothetically, may lead to a state of lasting
neuronal depolarization by maximum inhibition of Na, K-ATPase. This state may
manifest itself clinically as depression. If patients are now recovering from depres-
sion, the intracellular calcium declines on its way to normalization, passing a level
that may again activate adenylate cyclase, causing the often-observed hypomanic
state after depression. Calcium concentrations fluctuating around the threshold
between the manic and depressed stage may be a hypothetical origin of mixed states
or ultradian cycling.
In conclusion, this hypothesis combines special potential factors of vulnerability
in bipolar patients, such as altered Na, K-ATPase (Antia et al., 1995) and adenylate
cyclase activity (Meltzer, 1986), with the potentiating effects of increased intra-
cellular calcium mobilization or calcium influx into the cell.


How do antiepileptic drugs used in treating mixed states and rapid cycling
potentially interfere with intracellular calcium signalling?
A well-known target of lithium action is the inositol phospholipid pathway
which shows increased sensitivity in bipolar patients (Moscovich et al., 1990).
Inositol 1, 4, 5-triphosphate (IP3) synthesis mobilizes intracellular calcium stores
317 Rapid cycling and mixed states


in close proximity to their calcium-dependent effector proteins. Lithium inhibits
IP3 (Berridge and Irvine, 1989). Recently, it has been shown that valproate is also
capable of inhibiting the synthesis of inositol monophosphate (Vaden et al., 2001).
This also means a reduction of the activation rate of Ca2þ calmodulin kinase II,
a enzyme that is critically involved in long-term synaptic changes (Lisman, 1994)
and of myristoylated alanine-rich C kinase substrate, implicated in synaptic
neurotransmission and cytoskeletal restructuring (Lenox and Watson, 1994),
thus inhibiting structural changes of neurons in the course of the disease.
Different mechanisms have been implied in the therapeutic action of carbama-
zepine (Grunze et al., 1999). With respect to calcium, carbamazepine exerts strong
calcium antagonistic properties in vitro by blocking L-type calcium channels
(Walden et al., 1992), whereas valproate inhibits another voltage-dependent calcium
channel, the T channel (Altrup et al., 1992). Lamotrigine also exerts part of its action
by calcium antagonistic effects on L, N and P-type channels (Wegerer et al., 1997; Xie
and Hagan, 1998).
Another recently discovered effect of valproate is its action on the chaperones
GRP 78, GRP 94, and calreticulin. These endoplasmatic reticulum stress proteins
bind calcium and are protective against cell death. Chronic treatment with valproate
increases the synthesis of all three chaperones in rat C6 glioma cells, thus contribut-
ing to calcium homeostasis in the endoplasmatic reticulum (Wang et al., 2001).
In summary, all currently used effective mood stabilizers show somehow direct
or indirect calcium antagonistic properties. Interestingly, a review by Hollister and
Trevino (1999) listed 61 partially successful reports on trials of calcium antagonists
in bipolar disorder.
Of special interest is the fact that a randomized study of the L-type calcium
channel blocker nimodipine showed efficacy in thus far refractory rapid-cycling
patients (Pazzaglia et al., 1993, 1998).
As already indicated in the introduction, antiepileptic drugs, especially valproate
and lamotrigine, are considered to be more efficacious than lithium in treating
mixed states and rapid-cycling patients. Retrospective analysis of the pivotal trial of
Bowden et al. (1994) by Swann et al. (1997) showed equal efficacy for valproate in
pure manic and mixed patients, whereas lithium showed no benefit when mania was
accompanied by depressive symptoms. The same appears true when analyzing other
controlled trials for the efficacy of lithium in pure mania versus mixed states. In
rapid-cycling patients, open data of 101 patients (Calabrese et al., 1993) supported
the clinical usefulness of valproate in patients with rapid cycling, both with pure
mania and mixed states. Lamotrigine, on the other hand, is so far the only mood
stabilizer whose efficacy for the prophylactic treatment of rapid cycling has been
backed up by a large randomized, placebo-controlled study (Calabrese et al., 2000).
Lamotrigine appeared to be especially helpful in bipolar II rapid-cycling patients;
318 H. Grunze and J. Walden


<<

. 53
( 68 .)



>>