- Email: [email protected]
The effect of neuroleptic discontinuation on psychopathology, involuntary movements, and biochemical measures in patients with persistent tardive dyskinesia
224
BIOL PSYCHIATRY 1989;26:224-233
The Effect of Neuroleptic Discontinuation on Psychopathology , Invo~unta~ Movements, and Biochemical Measures in Patients with Persistent Tardive Dyskinesia William M. Glazer, Malcolm B. Bowers, Jr., Dennis S. Chamey, and George R. Heninger
As some of the pharmacological activities of neuroleptic medication may involve pathophysiological mechanisms underlying schizophrenia and tardive dyskinesia (TD), it is useful to study patients undergoing medication discontinuation. In this study, 19 stable, neuroleptic-maintained patients with persistent TD underwent taper and discontinuation of their neurolepti~ meditation over a 3-week period, and rnuit~p~ebe~vioral and biochemical (plasma HVA, MHPG, and prolactin) measures were obtained. The major jinding was that early relapsing patients had lower baseline and a signiJcantly greater increase in plasma HVA levels after discontinuation than nonrelapsing patients. In addition, patients exhibiting withdrawal-exacerbated TD had significantly lower plasma MHPG levels than patients not exhibiting this p~nomenon. The clinical and p~rmaco~ogicaI implications of these endings are discussed.
Introduction As some of the pharmacological activities of neuroleptic medication are thought to involve the ~athophysioiogical mechanisms underlying schizophrenia and tardive dyskinesia (TD), it is useful to measure changes in psychological symptoms, involuntary movements, and indicators of neurotransmitter function in patients undergoing medication discontinuation. Numerous studies have concluded that both schizophrenia and TD may involve abnormal neurotransmitter function in the dopaminergic (Carlsson 1959; Klawans 1973) and noradrenergic (Hartman 1976; Jeste and Wyatt 1981) systems. Recent attention to peripheral indicators of central dopamine and norepinep~ne function have added to our understanding of neuroleptic activity as well as to our knowledge of the pathophysiology of schizophrenia and TD. For example, plasma homovanillic acid (pHVA), which is affected by diet (Davidsan et al. 19871, age (Moore et al. 1983), time
From the Depattment of Psychiatry, Yale University Schocl of Medicine, New Haven, CT. Supported in part by NIMH Grant MH 30929. Address reprint requests to Dr. W. M. Glazer. Department of Psychiatry, Yale University School of Medicine, 34 Park Street, New Haven, CT 06519. Received April 9, 1988; revised December 3, 1988.
e? 1989 Society of Biological
Psychiatry
NeurolepticWithdrawa and Biological Change
BIOL PSYCHIATRY 1989:26:224-233
225
of day (Davilla et al. 1987), and gender (Bowers et al. 1983), has been shown to change after exposure to neuroleptic medication (Pickar et al. 1986; Davidson et al. 1987) may predict response to neuroleptic medication (Bowers et al. 1983), and may correlate positively with severity of psychiatric symptomatology (Davis et al. 1985; Pickar et al. 1986). Plasma HVA has also been shown to correlate with the severity of involuntary movements in medicated patients with TD (Kirch et al. 1983). Serum prolactin (PRL), an indicator of tuberoinfundibular dopamine (DA) function, falls abruptly following neuroleptic discontinuation (Zander et al. 1981). Baseline serum PRL levels may predict early relapse (Brown and Laughren, 1981; Zander et al. 198 I) following neuroleptic discontinuation and are associated with TD severity in ~stmenopausal women (Glazer et al. 1981). Plasma MHPG (pMH~), which has been shown to reflect central noradrenergic function (Chamey et al. 1984), has been shown to be blunted in response to clonidine challenge in schizophrenia (Sternberg et al. 1982) and to be nonsignificantly elevated in TD as compared to non-TD patients (Glazer et al. 1987). Although there are no studies of changes in pMHPG in relation to discontinuation of neuroleptic medication, there are data indicating that plasma norepinephrine is lowered under these conditions (Zander et al. 1981). The purpose of this study is to investigate the effects of tapering and discontinuing neuroleptic medications on psychopathology, involuntary movements, and pHVA, pMHPG, and serum PRL in a group of neuroleptic-maintained patients with persistent TD.
Methods This is a I)-week, prospective, longitudinal study of the effect of neuroleptic medication discontinuation on behavioral and biochemical measures in 19 patients with TD. Specifically, the outcomes of interest include changes in pHVA, serum PRL, and pMHPG in the entire sample, as well as in subsamples stratified by relapse and withdrawalexacerbated TD status. In addition, we were interested to see if sociodemographic or clinical variables were associated with these subgroups.
The subjects in this study were psychiatric outpatients with TD who were selected from a larger cohort that had participated in a discontinuation study described elsewhere (Glazer et al. 1985). This larger cohort was drawn from the Tardive Dyskinesia Clinic population at the Connecticut Mental Health Center, in New Haven, CT (Glazer and Moore 1980; Glazer 1986). The inclusion criteria for the present study are as follows. (1) A diagnosis of TD, based on the presence of an Abnormal Involuntary Movement Scale (AIMS) (Smith et al. 1979) total score of 3 (out of a possible total of 28) or more on two examinations at least 3 months apart. The total AIMS score of 3 had to include 1 of the 7 anatomical areas scored 2 or more (Glazer and Moore 1980). In addition, patients had to have had at least 6 months exposure to neuroleptic medication and lack of evidence of an alternative medical diagnosis. (2) Patients had to be healthy men or nonpregnant women between the ages of 18 and 72; and (3) subjects had a Schedule for Affective Disorders and Schizophrenia (Lifetime Version)-Research Diagnostic criteria (Spitzer and Endicott, 1978a,b) diagnosis of chronic or subchronic schizophrenia or schizoaffective disorder. Exclusion criteria included (1) a history of substance abuse within the 5 years
226
BIOL PSYCHIATRY 1989:26:224-233
W.M.
Glazer et
al.
previous to the study and (2) use of other psychoactive medication, including antiparkinsonian medication, within the previous 6 months.
Behavioral measures included sociodemographic and clinical history, psychological symptoms, and involuntary movements. Sociodemographic and clinical history were obtained from patient interview and review of clinical records. Continuous variables included age, age at first psychiatric contact, age at first psychiatric hospitalization, cumulative hospitalization duration, cumulative duration of outpatient treatment, number of psychiatric hospitalizations, duration of neuroIeptic exposure, duration of TD, and neuroleptic dose in chlo~romazine equivalents (CPZE) (Hollister 1977). Categorical variabies included gender, race (white, black, other), and marital status (married, single, divorced, or widowed). Psychological symptoms were rated with the Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham 1962). Abnormal movements were rated with the AIMS (Smith et al. 1979). Ratings were performed by trained psychiatric nurses supervised by a study psychiatrist. To maximize interrater agreement, 28 joint interrater sessions were held during the study period. Inte~ater agreement between each nurse and the study psychia~st, as expressed by the intraclass correlation coefficient, ranged between 0.75 and 0.92 for the total AIMS score and 0.8 and 0.96 for the total BPRS score. To minimize interrater discrepancies with individual patients, each patient was evaluated by a single rater. Biochemical measures included serum PRL, pMHPG, and pHVA. Venipuncture was performed between 9:00 and 1I:00 AM, with patients having fasted since midnight. To avoid strenuous exercise, patients were instructed to use the elevator to get to their examination, and venipuncture was performed after the patient had been sitting at rest for at least 1.5 min. Plasma HVA and MHPG were measured by gas chromatographymass spectroscopy using deuterated internai standards (Dekirmenjian and Maas 1974; Maas et al. 1979; Glazer et al. 1987). Serum PRL was measured with a radioimmunoassay technique.
Procedure After signing informed consent, patients were given a complete history and physical examination by the study psychiatrist. Laboratory studies were ordered when clinically indicated. The ~~selinep~use occurred over a I-Z-week period prior to the tapering of neuroleptic medication. Demographic and clinical history was elicited, and the behavioral and biochemical measures were obtained on two occasions during this phase. The taper phase involved a 7-lo-day period (mean 8.3, SD 1.1) during which neuroleptic meditation doses were reduced by singIe-blind substitution of placebo capsules at a rate determined by the study team. During this phase, behavioral and biochemical measures were obtained on at least 1 and up to 3 separate days (mean 2.4, SD 0.6). Once neuroleptic medications were discontinued, patients were observed in the drugfree phase for up to 14 days, until there was demonstrable withdrawal-exacerbated TD or relapse, which will be defined below. During this phase, behavioral and biochemical measures were obtained on three separate days.
Neuroleptic Withdrawal and Biological Change
BIOL PSYCHIATRY 1989;26:224-233
221
Statistical Methods As mentioned above, the outcomes of interest included changes in the biochemical measures for the entire sample as well as for subgroups stratified by relapse and withdrawalexacerbated TD status. As the number and timing of the biochemical measures varied among patients, it was necessary to select observations carefully to ensure that they represented comparable conditions. Three observation periods were selected to yield the largest sample size. All patients in this study had at least one of two samples drawn at the baseline phase, and one of three samples drawn at both the taper and drug-free phases. For purposes of analysis, these values were averaged and expressed as a single score at each of the three time points. Relapse occurred if, during the drug-free phase, the study psychiatrist judged that neuroleptic medications were necessary. As this was an outpatient study, it was important to avoid an acute psychotic exacerbation that would result in hospitalization. Thus, this judgment was based on the presence of what was thought to be prodromal signs of impending psychosis (Herz and Melville 1980). These signs could include behavioral changes such as sleep disturbance, withdrawal, negativistic behavior, mood change, etc., and are not necessarily limited to delusions, hallucinations, or thought disorder. Withdrawal-exacerbated TD (WE-TD) was defined as an increase in involuntary mouth movements, as evidenced by three consecutively elevated AIMS ratings (on 3 separate days during the 14-day drug-free phase) of at least 3 points in items 2-4 (lips, tongue, and jaw). If patients relapsed early in the drug-free phase and had not yet met criteria for WE-TD, they were not considered to be WE-TD cases for the purpose of this analysis. To see if pHVA, serum PRL, or pMHPG changed significantly over time in this study, we performed repeated measures Analyses of Variance (ANOVAs), with the biochemical measure as the dependent variable and time, i.e., baseline, taper, and drug-free phases, as the independent variable. In addition, we looked for main effects and interactions among the following variables on changes in biochemical measures following neuroleptic discontinuation: age, gender, relapse status, WE-TD status, and time. When the ANOVA was significant (p < 0.05), t-test (two-tailed) analyses were used to determine if there were significant differences within and between groups.
Results Description
of Patient Sample
Nineteen patients had pMHPG and serum PRL measures taken at least once during each of the three phases of the study, i.e., baseline, taper, and drug-free. Of these 19, the mean age was 51 years, 37% were men, 84% were white, 90% were never married. The mean age at first psychiatric contact was 24, mean cumulative hospitalization time was 57 months, mean neuroleptic exposure duration was 191 months, mean duration of TD was 37 months, and the mean neuroleptic dose (CPZE) at baseline was 326 mg. Thirteen of these 19 patients had pHVA changes measured. The subgroup of 13 patients from this sample had very similar characteristics to the larger group from which it was drawn. Behavioral
Measures
Psychological relapse. Five (26%) of the 19 patients exhibited psychological relapse, as defined above in the Methods section. There were no statistically significant differences between relapsers and nonrelapsers among the demographic and clinical variables ex-
228
W.M.
BIOL PSYCHIATRY 1989;26:224-233
Table
1,
Glazer et al.
Mean HVA Levels @g/ml)and Change Over Time Stratified by Relapse Status Baseline
Mean HVA level @g/ml) Mean difference from baseine @g/ml)
Relapsers (II = 4)b 10.8 (1.3) -
Taper
Drugfree
12.2 (2.4) 1.4 (1.3)
15.0 (3.1) 4.2 (2.2)
12.7 (2.8) - 1.3 (2.7)
13.6 (2.3) - 0.4 (3.2)
12.5 (2.6) - 0.5 (2.6)
14.0 (2.5) I .O (3.6)
Nonrelapsers (n = 9) Mean HVA level (@ml) Mean difference from baseline (r&ml)
Mean HVA level (ng/ml) Mean difference from baseline (ng/ml)
14.0 (4.5)
-
Total (n = 13) 13.0 (4.0) -
“values are expressed as means (@ml) with SD in parentheses. %I two of the four relapsing patients, pHVA levels went up prior to the observation of prodromal signs, whereas in the other two patients, elevations in pHVA appeared to occur concurrently with psychopathology.
amined. However, relapsing patients, as compared to nonrelapsing patients, had fewer cumulative months in hospital (20 versus 71 months), longer neuroleptic exposure duration (222 versus 180 months), and a higher mean dose of neuroleptic medication at baseline (424 versus 291 mg CPZE). Upon entry into the study, all patients were symptomatically stable as evidenced by BPRS scores in the low 20s (lowest possible score 18). The mean total baseline BPRS score in the relapsing patients was 23.5 as compared to 23.6 in the nonrelapsing patients. Nonrelapsing patients exhibited these BPRS scores throughout the study period. Relapsing patients exhibited BPRS scores in the 30s. As mentioned above, symptoms exhibited by the relapsing patients were usually not psychotic in nature, but rather were consistent with the prodromal signs that have been described by Herz and Melville (1980). During the relapse, typical BPRS items that were rated included emotional withdrawal, uncooperativeness, tension, and guilt. Not included on BPRS items, but common to relapsing patients, were changes in sleep patterns. Withdrawal-exacerbated TD. Ten (53%) of the 19 patients exhibited WE-TD movements as defined above in the Methods section. When we stratified the total sample by WE-TD status, we were unable to find statistical differences among the clinical and demographic variables measured. However, the following differences were worthy of note: patients with WE-TD as compared to patients without WE-TD were younger (mean age 47 versus 56 years) were older at first psychiatric contact (mean age 27 versus 20 years) and had a lower duration of exposure to neuroleptic medication (mean exposure duration 173 versus 211 months).
Biochemical
Measures
HVA levels. HVA levels for the entire group of 13 patients over the three timepoints of the study are displayed in Table 1. Repeated measures ANOVA of mean pHVA levels, controlling for age, gender, and time, revealed a trend for women to have higher pHVA
229
BIOL PSYCHIATRY
Neuroleptic Withdrawal and Biological Change
1989:26:224-233
Table 2. Mean and Mean Change in PIasma, PRL, and MHPG Levels in Total Sample”~*
Mean PRL level Mean change in PRL level from baseline Mean MHPG level Mean change in MHPG level from baseline
Baseline
Taper
Drug free
15.2 (15.1) -
8.1 (7.9) -7.0 (10.2)
2.1 (1.1) - 13.1 (14.6)
4.0 (1.2)
-
3.9 (1.1)
-0.1 (0.7)
3.9(I.l)
-0.1 f0.9)
“n = 19. bAll values in this table are expressed in nglml, numbers in parentheses are standard deviations
values (F = 4.1, p < 0. lo), but no other direct effects or interactions were found. The significant negative correlation between baseline pHVA levels and baseline to drug-free HVA differences (Pearson R = 0.79, df = 12, p < 0.01) indicated that patients with the lowest baseline pHVA values had the largest increase during the drug-free period. Table 1 also displays mean pHVA values stratified by relapse status. The repeated measures ANOVA was not significant for the effect of relapse status on pHVA levels (F = 0.22, df = l,ll, p = 0.65) or change over the three sampling times (F = 1.4, df = 2,22, p = 0.27), but was significant for the interaction of relapse status and sampling times (F = 5.3, df = 2,22, p = 0.014). t-Test analyses revealed that relapsing patients had nonsignificantly lower mean baseline pHVA levels (10.8 ng/ml) as compared to nonrelapsing patients (14.0 ng/ml). The change in HVA from baseline to the drug-free time~int was positive and was signi~~~tly (t = 2.6, p < 0.025) greater in the relapsed patients (4.2 ngiml) as compared to the nonrelapsed patients, who experienced a nonsignificant change in a negative direction (-0.4 @ml). In the 7 WE-TD-positive patients, the mean pHVA levels (in ng/ml +- SD) at baseline, taper, and drug free timepoints were: 12.3 k 3.3, 12.0 2 2.4, and 13.3 -+ 2.6, respectiveIy. In comparison, the 6 WE-TD-negative patients had pHVA values of 13.5 4 4.7, 13.0 k 2.8, and 14.6 2 2.4 nglml at these timepoints, respectively. The repeated measures ANOVA for the effect of WE-TD status on mean pHVA levels following neuroleptic medication discontinuation was nonsignificant. PRL levels. Mean and mean differences in serum PRL levels for the sample of 19 patients over the three time~ints of the study are displayed in Table 2. A repeated measures ANOVA for the effect of time, age, and gender on serum PRL levels following neuroleptic medication taper and discontinuation was significant (F = I 1.9, df = 2,22, p < 0.01) for a main effect of time, but no other main effects or interactions. Specifically, serum PRL levels fall linearly from a baseline mean of 15.2 ng/ml to a taper level of 8.1 ng/ml to the drug-free mean level of 2.1 ng/ml following neuroleptic medication discontinuation. Repeated measures ANOVAs were performed to see if relapse and WE-TD status influenced serum PRL level changes following neuroleptic medication discontinuation. These analyses were not significant (data not shown).
Plasma MHPC. Table 2 displays the mean and mean differences in pMHPG levels over the three time~ints following neuroleptic medication discontinuation among the sample of I9 patients. Repeated measures ANOVA for the effect of neuroleptic discon-
230
BIOL PSYCHIATRY 1989:26:224-233
W.M. Glazer et al.
tinuation on pMHPG levels was nonsigni~cant for main effects or interactions of age, gender, and time. A repeated measures ANOVA for the effect of WE-TD status on mean pMHPG levels revealed a significant (F = 6.0, df = 1,l I, p = 0.025) main effects of WE-TD status, but no main effect of time or interaction of time and WE-TD status. Specifically, pMHPG levels, regardless of time, were lower in WE-TD patients (baseline 3.4 nglml, taper 3.4 ngiml, drug-free 3.6 @ml) as compared to non-WE-TD patients (baseline 4.6 nglml, taper 4.4 nglml, drug-free 4.3 ngimi). t-Test comparisons of mean pMHPG levels at each timepoint between the WE-TD and non-WE-TD group were significant at the p < 0.05 level.
Discussion Neuroleptic Discantinuation Biochemical Changes and Psychopathology The major finding of this study is that the change in pHVA level following neuroleptic medication discontinuation in symptomatically stable patients with TD is different in early relapsing patients (exhibiting prodromal signs) as compared to patients who do not manifest such psychopathology. As described in the Results section (see Table I), relapsing patients tend to have lower baseline pHVA levels and experience a significant increase in pHVA, whereas nonrelapsing patients have higher baseline levels, which seem to show little change following neuroleptic medication discontinuation. Two previous studies of pHVA in chronically psychotic schizophrenic patients have found positive associations between this metabolite and psychopathology (Davis et al. 1985; Pickar et al. 1986). However, this association was statistical in nature, as compared to our results, which suggest qualitative differences in the pattern of pHVA levels following discontinuation of neuroleptic medication in relapsing versus nonrelapsing patients. One study (Bowers et al. 1983) reported that psychotic patients who respond well to neuroleptic medication exhibit elevated drug-free pHVA levels that drop following treatment. Consistent with this observation, our relapsing patients exhibited the opposite pattern in pHVA levels following discontinuation. If these data are replicated in a prospective study containing a larger patient sample, pHVA levels may become a useful tool in low-dose treatment strategies, such as the intermittent or continuous low-dose approaches currently under investigation, One can only speculate about the pha~acological mech~isms underlying the lower baseline pHVA levels and robust increases after withdrawal in the early relapsing patients. Preclinical and clinical studies thus far suggest that chronic neuroleptic treatment results in an initial elevation of pHVA (Chang et al. 1986; Davidson et al. 1987; Davilla et al, 1987) that eventually returns to the pretreatment levels, or, in the case of some patients, may fall below pretreatment IeveIs (Pickar et al. 1986). It is possible that the early relapsing patients in this study represent a pha~acological subgroup of neuroleptic responders, characterized by depressed baseline pHVA levels that increase substantially with neuroleptic taper and withdrawal. Given the small sample size of this study, it is premature to speculate in greater detail around the pathophysiological mechanism(s) that may underly this phenomenon. Suftice it to say, the concomitant rise in pHVA levels before or during the early relapse phenomenon is consistent with the hypothesis that some forms of psychosis are associated with a hy~rdopaminergic condition.
Neuroleptic Withdrawal and Biological Change
BIOLPSYCHIATRY 1989:26:224-233
231
Our data did not replicate earlier reports of low serum PRL levels in early relapsing patients after neuroleptic taper and discontinuation (Brown and Laughren 1981; Zander et al. 1981). However, given the brief observational period in our study, it is difficult to know whether or not patients in our sample would have relapsed if maintained drug-free for a longer time period. To our knowledge, this is the first study of changes in pMHPG levels following neuroleptic discontinuation. A previous study (Zander et al. 1988) reported a significant decrease in plasma noradrenaline levels in 13 patients who had had their neuroleptic medication discontinued, but plasma noradrenaline levels did not correlate with CSF MHPG levels, which did not change with neuroleptic withdrawal. The data in the present study do not support a major role of noradrenergic dysfunction in schizophrenia.
Withdrawal-Emergent
TD
The phenomenology of WE-TD has been described in previous work (Crane 1968; Gardos and Cole 1980; Baldessarini 1985). However, little effort has been directed at determining the place of this phenomenon in the overall course of TD. This lack of information may explain why WE-TD is not included in the research diagnostic criteria for TD (Schooler and Kane 1982). To date, the incidence of WE-TD is unclear. To our knowledge, this is the first study that attempts to utilize biochemical measures to describe WE-TD. If circulating HVA and PRL levels are indicators of central DA function, we were unable to show that DA function in patients with WE-TD is different from those without it. Furthermore, WE-TD status did not appear to be associated with early relapse. Our previous studies of noradrenergic function in TD (Glazer et al. 1987a,b) have yielded little in comparison to other studies (Jeste and Wyatt 1981) that suggest that norepinephrine may play a role in the pathophysiology of TD. In this study, we found that overall pMHPG levels (regardless of time) were significantly lower in WE-TD patients. Perhaps noradrenergic function increases as patients evolve from WE-TD to non-WE-TD status. Alternatively, noradrenergic mechanisms may distinguish a WE-TD from a non-WE-TD subtype. Interestingly, in our study of changes in pMHPG following yohimbine challenge (Glazer et al. 1987), we found that TD patients exhibited nonsignificantly higher baseline levels of pMHPG than did non-TD patients. Another group (Albus et al. 1985) reported nonsignificantly elevated baseline serum norepinephrine levels in “severe TD” as compared to “TD” or “no TD” patients who received chronic neuroleptic maintenance. These data suggest that noradrenergic mechanisms may play a role in the pathophysiology of TD. Future studies are needed to help define this role more clearly.
References Albus M, Naber D, Muller-Spahn F (1985): Tardive dyskinesia: Relation to computer-tomographic endocrine, and psychopathological variables. Biol Psychiatry 20:1082-1089. Baldessarini RJ (1985): Clinical and epidemiologic aspects of tardive dyskinesia. J Clin Psychiatry 46:8-13. Bowers MB, Swigar ME, Jatlow PI (1983) Sex differences in plasma homovanillic acid in acute psychosis. N Engl J Med (letter to the editor). 308:845-846. Bowers MB, Swigar ME, Jatlow PI, Goicoechea N (1984): Plasma catecholamine metabolites and early response to haloperidol. J Clin Psychiatry 45:248-25 1.
232
BIOL PSYCHIATRY 1989;26:224-233
W.M. Glazer et al.
drown w, Laughren T (1981): Low serum prolactin and early relapse following neuroleptic withdrawal. Am J Psychiatry 138:237-239. t&lsson A (1959): The occurrence, distribution, and physiological role of catecholamines in the nervous system. Phurmacol Rev 11:490-493. Chang WH, Yeh EK, Hu WH, Tseng YT, Chung MC, Chang HF (1986): Acute and chronic effects of haloperidol on plasma and brain homovanillic acid in the rat. Biol Psychiatry 21:374-381. Chmey DS, Heninger GR, Breier A (1984): Noradrenergic function in panic anxiety. Arch Gen Psychiatry 41:751-763.
Crane GE (1968): Tardive dyskinesia in patients treated with major neuroleptics: A review of the literature. Am J Psychiatry 124:40-48. Davidson M, Giordani AB, Mohs RC, Mykytyn VV, Platt S, Aryan ZS, Davis KL (1987a): Control of exogenous factors affecting plasma homovanillic acid concentration. Psychiatry Res 20:307312.
Davidson M, Giordani AB, Mohs RC, Horvath RB, Davis BM, Powchik R, Davis KL (1987b) Short-term haloperidol administration acutely elevates human plasma homovanillic acid concentrations. Arch Gen Psychiatry 44: 189-190. Davidson M, Losonczy MF, Mohs RC, Lesser JC, Powchik P, Freed LB, Davis BM, Mykytyn VV. Davis KL (1987~): Effects of debrisoquin and haloperidol on plasma homovanillic acid concentration in schizophrenic patients. Neuropsychophurmacoiogy 1: 17-23. Davilla R, Zumarraga M, Perea K, Andia MA, Friedhoff AJ (1987): Elevation of plasma homovanillic acid level can be detected within four hours after initiation of haloperidol treatment (letter to the editor). Arch Gen Psychiatry 44:837-838. Davis KL, Davidson M, Mohs RC, Kendler K, Davis B, Johns C, DeNigris Y, Horvath TB (1985): Plasma homovanillic concentration and the severity of schizophrenic illness. Science 227: 16011602. Dekirmenjian H, Maas JW (1974): MHPG in plasma. Clin Chem 52:203-208. Gardos G, Cole JO (1980): Overview: Public health issues in tardive dyskinesia. Am J Psychiatry 137:776-781.
Glazer WM (1986): Notes from a tardive dyskinesia clinic. Yale Psychiatric Q 84-15. Glazer W, Moore DC (1980): The diagnosis of rapid abnormal involuntary movements associated with fluphenazine decanoate. J Nerv Ment Dis 168:439-441. Glazer WM. Moore DC, Bowers MB, Brown WA (1981) Serum prolactin and tardive dyskinesia. Am J Psychiatry 138:1493-1496.
Glazer W, Hafez H, Benarroche C (1985): Molindone and haloperidol in tardive dyskinesia. J Cfin Psychiatry 464-7.
Glazer W, Charney D, Heninger G (1987a) Noradrenergic function in schizophrenia. Arch Gen Psychiatry 44:898-904.
Glazer WM. Morgenstem H, Jeste D, Zahner G, Hafez H, Benarroche C (1987b): Serum dopamine beta hydroxylase activity and tardive dyskinesia. Psychoneuroendocrinology 12:289-294. Hartmann E (1976): Schizophrenia: A theory. Psychopharmacology 49: l-15. Herz MI, Melville C (1980): Relapse in schizophrenia. Am J Psychiatry 137:801-805. Hollister LE (1977): Anti-psychotic medications and the treatment of schizophrenia. In Barchas JD, Berger PA, Ciaranello RD, et al (eds), Psychopharmacology: From Theory to Practice. New York: Oxford University Press, pp 121-150. Jeste DV, Wyatt RJ (1981): Dogma disputed: Is tardive dyskinesia due to postsynaptic dopamine receptor supersensitivity?. J Clin Psychiatry 42:455-457. Kirch D, Hattox S, Bell J, Murphy R, Freedman R (1983): Plasma homovanillic acid and tardive dyskinesia during neuroleptic maintenance and withdrawal. Psychiatry Res 9:2 17-223. Klawans H (1973): The pharmacology of tardive dyskinesia. Am J Psychiatry 130:82-86. Maas JW, Hattox SE, Martin DM, Landis DH (1979): A direct method for determining dopamine synthesis and output of dopamine metabolites from brain in awake animals. J Neurochem 32:839843.
Neuroleptic
Withdrawal
BIOL PSYCHIATRY 1989;26:224-233
and Biological Change
233
Moore D, Glazer WM, Bowers M, Heninger G (1983): Tardive dyskinesia and plasma homovanillic acid. Biol Psychiatry 18: 1393-1402. Overall JE, Gorham DR (1962): The Brief Psychiatric
Rating Scale. Psychol Rep 10:799-812.
Pickar D, Labarca R, Doran AR, Wolkowitz 0, Roy A, Breier A, Linnoila M, Paul S (1986) Longitudinal measurements of plasma homovanillic acid levels in schizophrenic patients. Arch Gen Psychiatry 43:669-676. School& N, Kane J (1982): Research diagnoses for tardive dyskinesia. Arch Gen Psychiatry 39:486487. Smith JM, Kucharski LT, Oswald M (1979): A systematic inpatients. Am J Psych&y 136:918-922.
investigation
of tardive dyskinesia
Spitzer R, Endicott J (1978a): Schedulefor AfSectiveDisorders and Schizophrenia-Ltfetime (SADS-L) (ed 3). New York: New York State Psychiatric Institute.
in
Version
Spitzer R, Endicott J (1978b): Research Diagnostic Criteria (ed 3). New York: New York State Psychiatric Institute. Stemberg DE, Chamey DS, Heninger GR, Leckman JF, Hafstad KM, Landis DH (1982) Impaired presynaptic regulation of norepinephrine in schizophrenia. Arch Gen Psychiatry 39:285-289. Zander KJ, Fischer B, Zimmer R, Ackenheil M (1981): Long-term neuroleptic treatment of chronic schizophrenic patients: Clinical and biochemical effects of withdrawal. Psychopharmacology 73:43-47.
BIOL PSYCHIATRY 1989;26:224-233
The Effect of Neuroleptic Discontinuation on Psychopathology , Invo~unta~ Movements, and Biochemical Measures in Patients with Persistent Tardive Dyskinesia William M. Glazer, Malcolm B. Bowers, Jr., Dennis S. Chamey, and George R. Heninger
As some of the pharmacological activities of neuroleptic medication may involve pathophysiological mechanisms underlying schizophrenia and tardive dyskinesia (TD), it is useful to study patients undergoing medication discontinuation. In this study, 19 stable, neuroleptic-maintained patients with persistent TD underwent taper and discontinuation of their neurolepti~ meditation over a 3-week period, and rnuit~p~ebe~vioral and biochemical (plasma HVA, MHPG, and prolactin) measures were obtained. The major jinding was that early relapsing patients had lower baseline and a signiJcantly greater increase in plasma HVA levels after discontinuation than nonrelapsing patients. In addition, patients exhibiting withdrawal-exacerbated TD had significantly lower plasma MHPG levels than patients not exhibiting this p~nomenon. The clinical and p~rmaco~ogicaI implications of these endings are discussed.
Introduction As some of the pharmacological activities of neuroleptic medication are thought to involve the ~athophysioiogical mechanisms underlying schizophrenia and tardive dyskinesia (TD), it is useful to measure changes in psychological symptoms, involuntary movements, and indicators of neurotransmitter function in patients undergoing medication discontinuation. Numerous studies have concluded that both schizophrenia and TD may involve abnormal neurotransmitter function in the dopaminergic (Carlsson 1959; Klawans 1973) and noradrenergic (Hartman 1976; Jeste and Wyatt 1981) systems. Recent attention to peripheral indicators of central dopamine and norepinep~ne function have added to our understanding of neuroleptic activity as well as to our knowledge of the pathophysiology of schizophrenia and TD. For example, plasma homovanillic acid (pHVA), which is affected by diet (Davidsan et al. 19871, age (Moore et al. 1983), time
From the Depattment of Psychiatry, Yale University Schocl of Medicine, New Haven, CT. Supported in part by NIMH Grant MH 30929. Address reprint requests to Dr. W. M. Glazer. Department of Psychiatry, Yale University School of Medicine, 34 Park Street, New Haven, CT 06519. Received April 9, 1988; revised December 3, 1988.
e? 1989 Society of Biological
Psychiatry
NeurolepticWithdrawa and Biological Change
BIOL PSYCHIATRY 1989:26:224-233
225
of day (Davilla et al. 1987), and gender (Bowers et al. 1983), has been shown to change after exposure to neuroleptic medication (Pickar et al. 1986; Davidson et al. 1987) may predict response to neuroleptic medication (Bowers et al. 1983), and may correlate positively with severity of psychiatric symptomatology (Davis et al. 1985; Pickar et al. 1986). Plasma HVA has also been shown to correlate with the severity of involuntary movements in medicated patients with TD (Kirch et al. 1983). Serum prolactin (PRL), an indicator of tuberoinfundibular dopamine (DA) function, falls abruptly following neuroleptic discontinuation (Zander et al. 1981). Baseline serum PRL levels may predict early relapse (Brown and Laughren, 1981; Zander et al. 198 I) following neuroleptic discontinuation and are associated with TD severity in ~stmenopausal women (Glazer et al. 1981). Plasma MHPG (pMH~), which has been shown to reflect central noradrenergic function (Chamey et al. 1984), has been shown to be blunted in response to clonidine challenge in schizophrenia (Sternberg et al. 1982) and to be nonsignificantly elevated in TD as compared to non-TD patients (Glazer et al. 1987). Although there are no studies of changes in pMHPG in relation to discontinuation of neuroleptic medication, there are data indicating that plasma norepinephrine is lowered under these conditions (Zander et al. 1981). The purpose of this study is to investigate the effects of tapering and discontinuing neuroleptic medications on psychopathology, involuntary movements, and pHVA, pMHPG, and serum PRL in a group of neuroleptic-maintained patients with persistent TD.
Methods This is a I)-week, prospective, longitudinal study of the effect of neuroleptic medication discontinuation on behavioral and biochemical measures in 19 patients with TD. Specifically, the outcomes of interest include changes in pHVA, serum PRL, and pMHPG in the entire sample, as well as in subsamples stratified by relapse and withdrawalexacerbated TD status. In addition, we were interested to see if sociodemographic or clinical variables were associated with these subgroups.
The subjects in this study were psychiatric outpatients with TD who were selected from a larger cohort that had participated in a discontinuation study described elsewhere (Glazer et al. 1985). This larger cohort was drawn from the Tardive Dyskinesia Clinic population at the Connecticut Mental Health Center, in New Haven, CT (Glazer and Moore 1980; Glazer 1986). The inclusion criteria for the present study are as follows. (1) A diagnosis of TD, based on the presence of an Abnormal Involuntary Movement Scale (AIMS) (Smith et al. 1979) total score of 3 (out of a possible total of 28) or more on two examinations at least 3 months apart. The total AIMS score of 3 had to include 1 of the 7 anatomical areas scored 2 or more (Glazer and Moore 1980). In addition, patients had to have had at least 6 months exposure to neuroleptic medication and lack of evidence of an alternative medical diagnosis. (2) Patients had to be healthy men or nonpregnant women between the ages of 18 and 72; and (3) subjects had a Schedule for Affective Disorders and Schizophrenia (Lifetime Version)-Research Diagnostic criteria (Spitzer and Endicott, 1978a,b) diagnosis of chronic or subchronic schizophrenia or schizoaffective disorder. Exclusion criteria included (1) a history of substance abuse within the 5 years
226
BIOL PSYCHIATRY 1989:26:224-233
W.M.
Glazer et
al.
previous to the study and (2) use of other psychoactive medication, including antiparkinsonian medication, within the previous 6 months.
Behavioral measures included sociodemographic and clinical history, psychological symptoms, and involuntary movements. Sociodemographic and clinical history were obtained from patient interview and review of clinical records. Continuous variables included age, age at first psychiatric contact, age at first psychiatric hospitalization, cumulative hospitalization duration, cumulative duration of outpatient treatment, number of psychiatric hospitalizations, duration of neuroIeptic exposure, duration of TD, and neuroleptic dose in chlo~romazine equivalents (CPZE) (Hollister 1977). Categorical variabies included gender, race (white, black, other), and marital status (married, single, divorced, or widowed). Psychological symptoms were rated with the Brief Psychiatric Rating Scale (BPRS) (Overall and Gorham 1962). Abnormal movements were rated with the AIMS (Smith et al. 1979). Ratings were performed by trained psychiatric nurses supervised by a study psychiatrist. To maximize interrater agreement, 28 joint interrater sessions were held during the study period. Inte~ater agreement between each nurse and the study psychia~st, as expressed by the intraclass correlation coefficient, ranged between 0.75 and 0.92 for the total AIMS score and 0.8 and 0.96 for the total BPRS score. To minimize interrater discrepancies with individual patients, each patient was evaluated by a single rater. Biochemical measures included serum PRL, pMHPG, and pHVA. Venipuncture was performed between 9:00 and 1I:00 AM, with patients having fasted since midnight. To avoid strenuous exercise, patients were instructed to use the elevator to get to their examination, and venipuncture was performed after the patient had been sitting at rest for at least 1.5 min. Plasma HVA and MHPG were measured by gas chromatographymass spectroscopy using deuterated internai standards (Dekirmenjian and Maas 1974; Maas et al. 1979; Glazer et al. 1987). Serum PRL was measured with a radioimmunoassay technique.
Procedure After signing informed consent, patients were given a complete history and physical examination by the study psychiatrist. Laboratory studies were ordered when clinically indicated. The ~~selinep~use occurred over a I-Z-week period prior to the tapering of neuroleptic medication. Demographic and clinical history was elicited, and the behavioral and biochemical measures were obtained on two occasions during this phase. The taper phase involved a 7-lo-day period (mean 8.3, SD 1.1) during which neuroleptic meditation doses were reduced by singIe-blind substitution of placebo capsules at a rate determined by the study team. During this phase, behavioral and biochemical measures were obtained on at least 1 and up to 3 separate days (mean 2.4, SD 0.6). Once neuroleptic medications were discontinued, patients were observed in the drugfree phase for up to 14 days, until there was demonstrable withdrawal-exacerbated TD or relapse, which will be defined below. During this phase, behavioral and biochemical measures were obtained on three separate days.
Neuroleptic Withdrawal and Biological Change
BIOL PSYCHIATRY 1989;26:224-233
221
Statistical Methods As mentioned above, the outcomes of interest included changes in the biochemical measures for the entire sample as well as for subgroups stratified by relapse and withdrawalexacerbated TD status. As the number and timing of the biochemical measures varied among patients, it was necessary to select observations carefully to ensure that they represented comparable conditions. Three observation periods were selected to yield the largest sample size. All patients in this study had at least one of two samples drawn at the baseline phase, and one of three samples drawn at both the taper and drug-free phases. For purposes of analysis, these values were averaged and expressed as a single score at each of the three time points. Relapse occurred if, during the drug-free phase, the study psychiatrist judged that neuroleptic medications were necessary. As this was an outpatient study, it was important to avoid an acute psychotic exacerbation that would result in hospitalization. Thus, this judgment was based on the presence of what was thought to be prodromal signs of impending psychosis (Herz and Melville 1980). These signs could include behavioral changes such as sleep disturbance, withdrawal, negativistic behavior, mood change, etc., and are not necessarily limited to delusions, hallucinations, or thought disorder. Withdrawal-exacerbated TD (WE-TD) was defined as an increase in involuntary mouth movements, as evidenced by three consecutively elevated AIMS ratings (on 3 separate days during the 14-day drug-free phase) of at least 3 points in items 2-4 (lips, tongue, and jaw). If patients relapsed early in the drug-free phase and had not yet met criteria for WE-TD, they were not considered to be WE-TD cases for the purpose of this analysis. To see if pHVA, serum PRL, or pMHPG changed significantly over time in this study, we performed repeated measures Analyses of Variance (ANOVAs), with the biochemical measure as the dependent variable and time, i.e., baseline, taper, and drug-free phases, as the independent variable. In addition, we looked for main effects and interactions among the following variables on changes in biochemical measures following neuroleptic discontinuation: age, gender, relapse status, WE-TD status, and time. When the ANOVA was significant (p < 0.05), t-test (two-tailed) analyses were used to determine if there were significant differences within and between groups.
Results Description
of Patient Sample
Nineteen patients had pMHPG and serum PRL measures taken at least once during each of the three phases of the study, i.e., baseline, taper, and drug-free. Of these 19, the mean age was 51 years, 37% were men, 84% were white, 90% were never married. The mean age at first psychiatric contact was 24, mean cumulative hospitalization time was 57 months, mean neuroleptic exposure duration was 191 months, mean duration of TD was 37 months, and the mean neuroleptic dose (CPZE) at baseline was 326 mg. Thirteen of these 19 patients had pHVA changes measured. The subgroup of 13 patients from this sample had very similar characteristics to the larger group from which it was drawn. Behavioral
Measures
Psychological relapse. Five (26%) of the 19 patients exhibited psychological relapse, as defined above in the Methods section. There were no statistically significant differences between relapsers and nonrelapsers among the demographic and clinical variables ex-
228
W.M.
BIOL PSYCHIATRY 1989;26:224-233
Table
1,
Glazer et al.
Mean HVA Levels @g/ml)and Change Over Time Stratified by Relapse Status Baseline
Mean HVA level @g/ml) Mean difference from baseine @g/ml)
Relapsers (II = 4)b 10.8 (1.3) -
Taper
Drugfree
12.2 (2.4) 1.4 (1.3)
15.0 (3.1) 4.2 (2.2)
12.7 (2.8) - 1.3 (2.7)
13.6 (2.3) - 0.4 (3.2)
12.5 (2.6) - 0.5 (2.6)
14.0 (2.5) I .O (3.6)
Nonrelapsers (n = 9) Mean HVA level (@ml) Mean difference from baseline (r&ml)
Mean HVA level (ng/ml) Mean difference from baseline (ng/ml)
14.0 (4.5)
-
Total (n = 13) 13.0 (4.0) -
“values are expressed as means (@ml) with SD in parentheses. %I two of the four relapsing patients, pHVA levels went up prior to the observation of prodromal signs, whereas in the other two patients, elevations in pHVA appeared to occur concurrently with psychopathology.
amined. However, relapsing patients, as compared to nonrelapsing patients, had fewer cumulative months in hospital (20 versus 71 months), longer neuroleptic exposure duration (222 versus 180 months), and a higher mean dose of neuroleptic medication at baseline (424 versus 291 mg CPZE). Upon entry into the study, all patients were symptomatically stable as evidenced by BPRS scores in the low 20s (lowest possible score 18). The mean total baseline BPRS score in the relapsing patients was 23.5 as compared to 23.6 in the nonrelapsing patients. Nonrelapsing patients exhibited these BPRS scores throughout the study period. Relapsing patients exhibited BPRS scores in the 30s. As mentioned above, symptoms exhibited by the relapsing patients were usually not psychotic in nature, but rather were consistent with the prodromal signs that have been described by Herz and Melville (1980). During the relapse, typical BPRS items that were rated included emotional withdrawal, uncooperativeness, tension, and guilt. Not included on BPRS items, but common to relapsing patients, were changes in sleep patterns. Withdrawal-exacerbated TD. Ten (53%) of the 19 patients exhibited WE-TD movements as defined above in the Methods section. When we stratified the total sample by WE-TD status, we were unable to find statistical differences among the clinical and demographic variables measured. However, the following differences were worthy of note: patients with WE-TD as compared to patients without WE-TD were younger (mean age 47 versus 56 years) were older at first psychiatric contact (mean age 27 versus 20 years) and had a lower duration of exposure to neuroleptic medication (mean exposure duration 173 versus 211 months).
Biochemical
Measures
HVA levels. HVA levels for the entire group of 13 patients over the three timepoints of the study are displayed in Table 1. Repeated measures ANOVA of mean pHVA levels, controlling for age, gender, and time, revealed a trend for women to have higher pHVA
229
BIOL PSYCHIATRY
Neuroleptic Withdrawal and Biological Change
1989:26:224-233
Table 2. Mean and Mean Change in PIasma, PRL, and MHPG Levels in Total Sample”~*
Mean PRL level Mean change in PRL level from baseline Mean MHPG level Mean change in MHPG level from baseline
Baseline
Taper
Drug free
15.2 (15.1) -
8.1 (7.9) -7.0 (10.2)
2.1 (1.1) - 13.1 (14.6)
4.0 (1.2)
-
3.9 (1.1)
-0.1 (0.7)
3.9(I.l)
-0.1 f0.9)
“n = 19. bAll values in this table are expressed in nglml, numbers in parentheses are standard deviations
values (F = 4.1, p < 0. lo), but no other direct effects or interactions were found. The significant negative correlation between baseline pHVA levels and baseline to drug-free HVA differences (Pearson R = 0.79, df = 12, p < 0.01) indicated that patients with the lowest baseline pHVA values had the largest increase during the drug-free period. Table 1 also displays mean pHVA values stratified by relapse status. The repeated measures ANOVA was not significant for the effect of relapse status on pHVA levels (F = 0.22, df = l,ll, p = 0.65) or change over the three sampling times (F = 1.4, df = 2,22, p = 0.27), but was significant for the interaction of relapse status and sampling times (F = 5.3, df = 2,22, p = 0.014). t-Test analyses revealed that relapsing patients had nonsignificantly lower mean baseline pHVA levels (10.8 ng/ml) as compared to nonrelapsing patients (14.0 ng/ml). The change in HVA from baseline to the drug-free time~int was positive and was signi~~~tly (t = 2.6, p < 0.025) greater in the relapsed patients (4.2 ngiml) as compared to the nonrelapsed patients, who experienced a nonsignificant change in a negative direction (-0.4 @ml). In the 7 WE-TD-positive patients, the mean pHVA levels (in ng/ml +- SD) at baseline, taper, and drug free timepoints were: 12.3 k 3.3, 12.0 2 2.4, and 13.3 -+ 2.6, respectiveIy. In comparison, the 6 WE-TD-negative patients had pHVA values of 13.5 4 4.7, 13.0 k 2.8, and 14.6 2 2.4 nglml at these timepoints, respectively. The repeated measures ANOVA for the effect of WE-TD status on mean pHVA levels following neuroleptic medication discontinuation was nonsignificant. PRL levels. Mean and mean differences in serum PRL levels for the sample of 19 patients over the three time~ints of the study are displayed in Table 2. A repeated measures ANOVA for the effect of time, age, and gender on serum PRL levels following neuroleptic medication taper and discontinuation was significant (F = I 1.9, df = 2,22, p < 0.01) for a main effect of time, but no other main effects or interactions. Specifically, serum PRL levels fall linearly from a baseline mean of 15.2 ng/ml to a taper level of 8.1 ng/ml to the drug-free mean level of 2.1 ng/ml following neuroleptic medication discontinuation. Repeated measures ANOVAs were performed to see if relapse and WE-TD status influenced serum PRL level changes following neuroleptic medication discontinuation. These analyses were not significant (data not shown).
Plasma MHPC. Table 2 displays the mean and mean differences in pMHPG levels over the three time~ints following neuroleptic medication discontinuation among the sample of I9 patients. Repeated measures ANOVA for the effect of neuroleptic discon-
230
BIOL PSYCHIATRY 1989:26:224-233
W.M. Glazer et al.
tinuation on pMHPG levels was nonsigni~cant for main effects or interactions of age, gender, and time. A repeated measures ANOVA for the effect of WE-TD status on mean pMHPG levels revealed a significant (F = 6.0, df = 1,l I, p = 0.025) main effects of WE-TD status, but no main effect of time or interaction of time and WE-TD status. Specifically, pMHPG levels, regardless of time, were lower in WE-TD patients (baseline 3.4 nglml, taper 3.4 ngiml, drug-free 3.6 @ml) as compared to non-WE-TD patients (baseline 4.6 nglml, taper 4.4 nglml, drug-free 4.3 ngimi). t-Test comparisons of mean pMHPG levels at each timepoint between the WE-TD and non-WE-TD group were significant at the p < 0.05 level.
Discussion Neuroleptic Discantinuation Biochemical Changes and Psychopathology The major finding of this study is that the change in pHVA level following neuroleptic medication discontinuation in symptomatically stable patients with TD is different in early relapsing patients (exhibiting prodromal signs) as compared to patients who do not manifest such psychopathology. As described in the Results section (see Table I), relapsing patients tend to have lower baseline pHVA levels and experience a significant increase in pHVA, whereas nonrelapsing patients have higher baseline levels, which seem to show little change following neuroleptic medication discontinuation. Two previous studies of pHVA in chronically psychotic schizophrenic patients have found positive associations between this metabolite and psychopathology (Davis et al. 1985; Pickar et al. 1986). However, this association was statistical in nature, as compared to our results, which suggest qualitative differences in the pattern of pHVA levels following discontinuation of neuroleptic medication in relapsing versus nonrelapsing patients. One study (Bowers et al. 1983) reported that psychotic patients who respond well to neuroleptic medication exhibit elevated drug-free pHVA levels that drop following treatment. Consistent with this observation, our relapsing patients exhibited the opposite pattern in pHVA levels following discontinuation. If these data are replicated in a prospective study containing a larger patient sample, pHVA levels may become a useful tool in low-dose treatment strategies, such as the intermittent or continuous low-dose approaches currently under investigation, One can only speculate about the pha~acological mech~isms underlying the lower baseline pHVA levels and robust increases after withdrawal in the early relapsing patients. Preclinical and clinical studies thus far suggest that chronic neuroleptic treatment results in an initial elevation of pHVA (Chang et al. 1986; Davidson et al. 1987; Davilla et al, 1987) that eventually returns to the pretreatment levels, or, in the case of some patients, may fall below pretreatment IeveIs (Pickar et al. 1986). It is possible that the early relapsing patients in this study represent a pha~acological subgroup of neuroleptic responders, characterized by depressed baseline pHVA levels that increase substantially with neuroleptic taper and withdrawal. Given the small sample size of this study, it is premature to speculate in greater detail around the pathophysiological mechanism(s) that may underly this phenomenon. Suftice it to say, the concomitant rise in pHVA levels before or during the early relapse phenomenon is consistent with the hypothesis that some forms of psychosis are associated with a hy~rdopaminergic condition.
Neuroleptic Withdrawal and Biological Change
BIOLPSYCHIATRY 1989:26:224-233
231
Our data did not replicate earlier reports of low serum PRL levels in early relapsing patients after neuroleptic taper and discontinuation (Brown and Laughren 1981; Zander et al. 1981). However, given the brief observational period in our study, it is difficult to know whether or not patients in our sample would have relapsed if maintained drug-free for a longer time period. To our knowledge, this is the first study of changes in pMHPG levels following neuroleptic discontinuation. A previous study (Zander et al. 1988) reported a significant decrease in plasma noradrenaline levels in 13 patients who had had their neuroleptic medication discontinued, but plasma noradrenaline levels did not correlate with CSF MHPG levels, which did not change with neuroleptic withdrawal. The data in the present study do not support a major role of noradrenergic dysfunction in schizophrenia.
Withdrawal-Emergent
TD
The phenomenology of WE-TD has been described in previous work (Crane 1968; Gardos and Cole 1980; Baldessarini 1985). However, little effort has been directed at determining the place of this phenomenon in the overall course of TD. This lack of information may explain why WE-TD is not included in the research diagnostic criteria for TD (Schooler and Kane 1982). To date, the incidence of WE-TD is unclear. To our knowledge, this is the first study that attempts to utilize biochemical measures to describe WE-TD. If circulating HVA and PRL levels are indicators of central DA function, we were unable to show that DA function in patients with WE-TD is different from those without it. Furthermore, WE-TD status did not appear to be associated with early relapse. Our previous studies of noradrenergic function in TD (Glazer et al. 1987a,b) have yielded little in comparison to other studies (Jeste and Wyatt 1981) that suggest that norepinephrine may play a role in the pathophysiology of TD. In this study, we found that overall pMHPG levels (regardless of time) were significantly lower in WE-TD patients. Perhaps noradrenergic function increases as patients evolve from WE-TD to non-WE-TD status. Alternatively, noradrenergic mechanisms may distinguish a WE-TD from a non-WE-TD subtype. Interestingly, in our study of changes in pMHPG following yohimbine challenge (Glazer et al. 1987), we found that TD patients exhibited nonsignificantly higher baseline levels of pMHPG than did non-TD patients. Another group (Albus et al. 1985) reported nonsignificantly elevated baseline serum norepinephrine levels in “severe TD” as compared to “TD” or “no TD” patients who received chronic neuroleptic maintenance. These data suggest that noradrenergic mechanisms may play a role in the pathophysiology of TD. Future studies are needed to help define this role more clearly.
References Albus M, Naber D, Muller-Spahn F (1985): Tardive dyskinesia: Relation to computer-tomographic endocrine, and psychopathological variables. Biol Psychiatry 20:1082-1089. Baldessarini RJ (1985): Clinical and epidemiologic aspects of tardive dyskinesia. J Clin Psychiatry 46:8-13. Bowers MB, Swigar ME, Jatlow PI (1983) Sex differences in plasma homovanillic acid in acute psychosis. N Engl J Med (letter to the editor). 308:845-846. Bowers MB, Swigar ME, Jatlow PI, Goicoechea N (1984): Plasma catecholamine metabolites and early response to haloperidol. J Clin Psychiatry 45:248-25 1.
232
BIOL PSYCHIATRY 1989;26:224-233
W.M. Glazer et al.
drown w, Laughren T (1981): Low serum prolactin and early relapse following neuroleptic withdrawal. Am J Psychiatry 138:237-239. t&lsson A (1959): The occurrence, distribution, and physiological role of catecholamines in the nervous system. Phurmacol Rev 11:490-493. Chang WH, Yeh EK, Hu WH, Tseng YT, Chung MC, Chang HF (1986): Acute and chronic effects of haloperidol on plasma and brain homovanillic acid in the rat. Biol Psychiatry 21:374-381. Chmey DS, Heninger GR, Breier A (1984): Noradrenergic function in panic anxiety. Arch Gen Psychiatry 41:751-763.
Crane GE (1968): Tardive dyskinesia in patients treated with major neuroleptics: A review of the literature. Am J Psychiatry 124:40-48. Davidson M, Giordani AB, Mohs RC, Mykytyn VV, Platt S, Aryan ZS, Davis KL (1987a): Control of exogenous factors affecting plasma homovanillic acid concentration. Psychiatry Res 20:307312.
Davidson M, Giordani AB, Mohs RC, Horvath RB, Davis BM, Powchik R, Davis KL (1987b) Short-term haloperidol administration acutely elevates human plasma homovanillic acid concentrations. Arch Gen Psychiatry 44: 189-190. Davidson M, Losonczy MF, Mohs RC, Lesser JC, Powchik P, Freed LB, Davis BM, Mykytyn VV. Davis KL (1987~): Effects of debrisoquin and haloperidol on plasma homovanillic acid concentration in schizophrenic patients. Neuropsychophurmacoiogy 1: 17-23. Davilla R, Zumarraga M, Perea K, Andia MA, Friedhoff AJ (1987): Elevation of plasma homovanillic acid level can be detected within four hours after initiation of haloperidol treatment (letter to the editor). Arch Gen Psychiatry 44:837-838. Davis KL, Davidson M, Mohs RC, Kendler K, Davis B, Johns C, DeNigris Y, Horvath TB (1985): Plasma homovanillic concentration and the severity of schizophrenic illness. Science 227: 16011602. Dekirmenjian H, Maas JW (1974): MHPG in plasma. Clin Chem 52:203-208. Gardos G, Cole JO (1980): Overview: Public health issues in tardive dyskinesia. Am J Psychiatry 137:776-781.
Glazer WM (1986): Notes from a tardive dyskinesia clinic. Yale Psychiatric Q 84-15. Glazer W, Moore DC (1980): The diagnosis of rapid abnormal involuntary movements associated with fluphenazine decanoate. J Nerv Ment Dis 168:439-441. Glazer WM. Moore DC, Bowers MB, Brown WA (1981) Serum prolactin and tardive dyskinesia. Am J Psychiatry 138:1493-1496.
Glazer W, Hafez H, Benarroche C (1985): Molindone and haloperidol in tardive dyskinesia. J Cfin Psychiatry 464-7.
Glazer W, Charney D, Heninger G (1987a) Noradrenergic function in schizophrenia. Arch Gen Psychiatry 44:898-904.
Glazer WM. Morgenstem H, Jeste D, Zahner G, Hafez H, Benarroche C (1987b): Serum dopamine beta hydroxylase activity and tardive dyskinesia. Psychoneuroendocrinology 12:289-294. Hartmann E (1976): Schizophrenia: A theory. Psychopharmacology 49: l-15. Herz MI, Melville C (1980): Relapse in schizophrenia. Am J Psychiatry 137:801-805. Hollister LE (1977): Anti-psychotic medications and the treatment of schizophrenia. In Barchas JD, Berger PA, Ciaranello RD, et al (eds), Psychopharmacology: From Theory to Practice. New York: Oxford University Press, pp 121-150. Jeste DV, Wyatt RJ (1981): Dogma disputed: Is tardive dyskinesia due to postsynaptic dopamine receptor supersensitivity?. J Clin Psychiatry 42:455-457. Kirch D, Hattox S, Bell J, Murphy R, Freedman R (1983): Plasma homovanillic acid and tardive dyskinesia during neuroleptic maintenance and withdrawal. Psychiatry Res 9:2 17-223. Klawans H (1973): The pharmacology of tardive dyskinesia. Am J Psychiatry 130:82-86. Maas JW, Hattox SE, Martin DM, Landis DH (1979): A direct method for determining dopamine synthesis and output of dopamine metabolites from brain in awake animals. J Neurochem 32:839843.
Neuroleptic
Withdrawal
BIOL PSYCHIATRY 1989;26:224-233
and Biological Change
233
Moore D, Glazer WM, Bowers M, Heninger G (1983): Tardive dyskinesia and plasma homovanillic acid. Biol Psychiatry 18: 1393-1402. Overall JE, Gorham DR (1962): The Brief Psychiatric
Rating Scale. Psychol Rep 10:799-812.
Pickar D, Labarca R, Doran AR, Wolkowitz 0, Roy A, Breier A, Linnoila M, Paul S (1986) Longitudinal measurements of plasma homovanillic acid levels in schizophrenic patients. Arch Gen Psychiatry 43:669-676. School& N, Kane J (1982): Research diagnoses for tardive dyskinesia. Arch Gen Psychiatry 39:486487. Smith JM, Kucharski LT, Oswald M (1979): A systematic inpatients. Am J Psych&y 136:918-922.
investigation
of tardive dyskinesia
Spitzer R, Endicott J (1978a): Schedulefor AfSectiveDisorders and Schizophrenia-Ltfetime (SADS-L) (ed 3). New York: New York State Psychiatric Institute.
in
Version
Spitzer R, Endicott J (1978b): Research Diagnostic Criteria (ed 3). New York: New York State Psychiatric Institute. Stemberg DE, Chamey DS, Heninger GR, Leckman JF, Hafstad KM, Landis DH (1982) Impaired presynaptic regulation of norepinephrine in schizophrenia. Arch Gen Psychiatry 39:285-289. Zander KJ, Fischer B, Zimmer R, Ackenheil M (1981): Long-term neuroleptic treatment of chronic schizophrenic patients: Clinical and biochemical effects of withdrawal. Psychopharmacology 73:43-47.