Schizophrenia Research, 3 (1990) 3 15-320 Elsevier

315

SCHIZO 00138

A comparison of masking effects of haloperidol versus molindone in tardive dyskinesia

William M. Glazer and Hisham Hafez

Yale University School of Medicine, TD Clinic, New Haven, CT 06519, U.S.A., and Nashua Brookside Hospital. Nashua. NH 03063, U.S.A.

(Received 30 March 1990, revised received 21 May 1990, accepted 24 May 1990)

An experimental method was utilized to compare the masking effects of two neuroleptic agents-molindone and haloperidol-on 18 neuroleptic-treated schizophrenic patients exhibiting operationally defined with- drawal-exacerbated tardive dyskinesia. After a week on one of these two medications at preestablished doses equivalent to that of the pre-study neuroleptic, molindone-masked total AIMS scores by significantly less (12%) than haloperidol(27%). Similarly, during a second week when the dose of these neuroleptics was equivalent to 200% that of the pre-study dose, molindone masked the total AIMS score significantly less (23%) as compared to haloperidol (53%). Several interpretations of this finding are considered. This study demonstrates the feasibility of a method that may offer a model for understanding pharmacological differences among neuroleptic medications.

Key words: Haloperidol; Molindone; Neuroleptic; Tardive dyskinesia

INTRODUCTON

Relatively little is known about the ability of different neuroleptic medications to alter the move- ments of tardive dyskinesia (TD) (Gerlach and Simmelsgaard, 1978; Small et al., 1987). Such knowledge is important if we are to understand the mechanisms of action of these medications as well as the nature of the disorders they treat and the side effects they cause. In a previous preliminary publication (Glazer et al., 1985), we reported re- sults of an attempt to compare the ability of two neuroleptic medications-molindone and haloperi- dol-to mask the movements of TD. We found a marginally significant tendency for haloperidol to mask TD movements more than molindone in 11 patients. The purpose of this paper is to report the final results of this study, having added more

Correspondence to: William Glazer, Tardive Dyskinesia

Clinic, 34 Park Street, New Haven, CT 06519, U.S.A., Tel. 203

789 6985.

patients to increase the statistical power to detect differences.

The hypothesis of this study is based on two observations regarding the course of TD. First, it is known that the withdrawal of neuroleptic medica- tion from some patients with TD exacerbates the involuntary movements (‘withdrawal-exacerbated TD’). Second, neuroleptic medication can suppress or ‘mask’ TD movements in many patients. If patients with withdrawal-exacerbated TD are ran- domly assigned to equivalent doses of molindone or haloperidol, it is possible to measure differences in the ability of these two medications to mask TD movements. Molindone shares a pharmacologic profile with haloperidol and other neuroleptic medications (Owen and Cole, 1989). However, in the rat (Bunney et al., 1975; Greengard, 1975; Alander et al., 1980; Meller, 1982; Fuller and Snoddy, 1983; McMillen and McDonald, 1983; White and Wang, 1983) but not the guinea pig (Koller et al., 1984) molindone may be different from the typical neuroleptic agents. In the rat, molindone has been reported not to inhibit striatal

0920-9964/90/$03.50 0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

316

and mesolimbic dopamine adenylate cyclase (Greengard, 1975), not to cause a sustained in- crease in firing rates of dopamine cells in the substantia nigra (Bunney et al., 1975), and not to produce a significant increase in striatal binding and apomorphine-induced stereotypy (Meller, 1982) or a decrease in zona compacta dopamine cell firing activity (White and Wang, 1983) after chronic treatment. Furthermore, in low to moder- ate doses, molindone is comparatively a more potent inhibitor of presynaptic than postsynaptic dopamine activity (Alander et al., 1980; Fuller and Snoddy, 1983; McMillen and McDonald, 1983). Consistent with these animal studies, a study in normal human brain caudate nucleus found that molindone was one of the least potent drugs studied at the D, receptor with a KD equal to 120 f 40, compared to haloperidol with a KD equal to 4+ 1 nM (Richelson, 1985). Assuming that TD is due to striatal dopamine dysfunction following neuroleptic treatment (Klawans, 1973), and assum- ing that the mechanism for masking involves this system, we hypothesized that molindone would be a less potent masker of withdrawal-exacerbated TD than haloperidol.

MATERIALS AND METHOD

Patients

A total of 3 1 outpatients who had been receiving at least a year of continuous treatment with neurolep- tics other than molindone or haloperidol agreed to participate in this study. As will be explained below, 18 patients completed the study. The patients, all diagnosed as having TD, were fol- lowed for at least 3 months in the Tardive Dyskine- sia Clinic at the Connecticut Mental Health Center (Glazer and Moore, 1981). To be included for study, patients had to (1) be healthy males and nonpregnant females between the ages of 18 and 70 who met research diagnostic criteria (Endicott and Spitzer, 1978; Spitzer and Endicott, March 1, 1978) for chronic or subchronic schizophrenia or schizo- affective disorder; (2) meet diagnostic criteria for TD (Glazer and Moore, 1981); (3) have had at least 12 months exposure to neuroleptic medication and lack of evidence of an alternative medical diagno- sis. Excluded from the study were (1) patients who reported or were documented to have had a history

of substance abuse in the previous 5 years; and (2) patients who used other psychoactive medications including antiparkinsonian agents, within the pre- vious 6 months.

Procedure After the nature of the study was explained to the patients and informed consent was obtained, they were given a complete history and physical exam- ination. Laboratory studies were ordered where clinically indicated. Neuroleptic medications were tapered over a 7-10 day period and then with- drawn, with single-blind substitution of placebo for 7-14 days. Each patient was followed in the drug-free state until there was a demonstrable increase in involuntary movements, as evidenced by an increase in scores on the AIMS that averaged at least three units in Items 2-4 (lips, tongue, and jaw), or until they developed psychological symp- tomatology that required the use of neuroleptic medication. To proceed to the medication assign- ment, patients had to demonstrate this increase on three consecutive ratings during the 14-day drug- free period. This criterion assured that only those patients whose TD could be at least partially masked were identified. The definition of with- drawal-exacerbated TD was limited to the lips, tongue and jaw because it was felt that movements in these areas were most specific to TD.

Of the 31 patients entered into the study, 18 met this criterion for withdrawal-exacerbated TD and were then admitted to a masking phase in which they were randomly assigned to receive either molindone or haloperidol in a double-blind fash- ion. Medication dosing was determined by the occurrence of side effects or psychiatric symptoms. If the patient experienced no side effects or psychi- atric symptoms, the dose was raised during the first week from 100% to the second week when about 200% dose equivalency of the patient’s prestudy neuroleptic medication was given.

Medication was supplied in identical-appearing red capsules containing 25 mg and 5 mg, respec- tively, of molindone and haloperidol. Relative dose relationships among the various previously used neuroleptic medications were determined from a review of the general literature and a poll of experts in the field. Based on these data, doses were calculated in the following mg equivalents: chlor- promazine, 100; thioridazine, 100; perphenazine,

317

10; trifluoperazine, 5; fluphenazine, 2; thiothixene, 5; haloperidol, 3; molindone, 1.5; and loxapine, 10. At each visit, the remaining medication was col- lected and counted to estimate patient compliance. None of the patients completing the study required antiparkinsonian medication during the study.

Ratings Psychiatric symptoms, dyskinetic movements and parkinsonism were measured on the Brief Psychi- atric Rating Scale (BPRS) (Overall and Gorham, 1962), the AIMS (Smith et al., 1979) and the Webster Parkinsonism Rating Scale (Webster 1968). AIMS ratings were performed by the same psychiatric nurse rater at least twice before the

taper phase, at least three times during the taper phase and withdrawal (drug-free) phase, and twice a week during the medication assignment period. Interrater agreement among raters on the AIMS total score ranged from intraclass correlations of 0.7550.92.

Statistical methods Since the number and timing of the ratings and the dosages of study medications varied among patients, it was necessary to select observations carefully to ensure that they represented compar- able conditions. Three observation periods were selected to yield the largest sample size. Eighteen of the 31 patients entered into the study had 2 or more observations each at the drug-free (with- drawal-exacerbation) phase, the 100% and 200% masking phases. Thirteen patients never received the two masking doses of study medication be- cause they developed psychotic symptoms on re- duced dosages and had to be given neuroleptic treatment or failed to exhibit withdrawal-exacer- bated TD.

Changes in total AIMS scores were analyzed by an ANOVA containing one between-groups factor (two drug groups) and one within-subjects factor (the drug-free AIMS score, the AIMS score at the 100% masking dose, and the AIMS score at the 200% masking dose). Comparisons were made in conjunction with the main effect and interaction. To assess the effectiveness of the masking properties of each drug, two specific comparisons were also made in conjunction with the interaction of time and group. First, the differences in change from the drug-free phase to the 100% and 200%

masking doses are compared between molindone and haloperidol. Second, the differences in the 100% and 200% masking doses are compared between the two drugs. A percentage change in the total AIMS score was calculated for each patient by dividing the difference between the AIMS score at the drug-free phase and the masking phase by the drug-free AIMS score, and then multiplying by 100. Analysis of variance (ANOVA) was then performed for the two groups to compare percent change scores.

RESULTS

Of the 18 patients who received study medication at the 100% and 200% masking doses, 9 (50%) were assigned to molindone and received a mean 100% dose of 75 mg and a mean 200% dose of 145 mg. Nine (50%) patients were assigned to haloperidol at a 100% dose of 19.3 mg and a mean 200% dose of 34.3 mg. The mean age of the sample was 47 years, 55% were white and 55% were women. The population suffered from prolonged illness as evidenced by a mean neuroleptic expo- sure duration of 152 months, and a mean number of 4.6 hospitalizations. The two groups were com- parable except for a greater past hospitalization duration in the molindone (62h 123 months vs. 16f 18 months) as compared to the haloperidol- treated group.

Group differences in total AIMS scores at drug- free, 100% and 200% masking doses are displayed in Table 1. A repeated measures ANOVA of the AIMS scores revealed a marginally significant overall difference between the two groups summed over the 3 observation periods (F= 3.43, df= 1, P=O.O86). However, both the time period effect and, most importantly the interaction of time and group, i.e. the effect of drug exposure on the pattern of AIMS scores over time, were highly significant (time effect F= 75.9, df=2, P<O.OOOl; time x group effect F= 9.9, df=2, P<O.OOOl). To see the source and magnitude of these effects we calculated contrast scores comparing the two drugs for changes from drug-free to the two masking doses and from the 100% to the 200% masking doses. In terms of the hypothesis of this study, the time-drug contrasts were the most relevant. The

318

TABLE 1

Means and standard deviations for total” AIMS SCOWS from drug-free phase to the two masking doses

Variable Molindone (n = 9) Haloperidol (n = 9)

Mean SD Mean SD

Drug-free phase 11.35 2.68 10.97 1.78

100% masking dose 9.94 2.20 8.07 2.28 200% masking dose 8.68 2.86 5.24 2.10

“Total AIMS score refers to the sum of scores on the 7

anatomical areas. Total score cannot exceed 28.

average AIMS score decrease from drug-free to the masking doses was significantly greater for halo- peridol (t=3.5, P=O.O03). The further decrease from 100% to 200% masking doses was also significantly greater for haloperidol (t = 2.5, P= 0.022).

At the 100% masking dose, haloperidol reduced the total AIMS score by 27.2% as compared to an 11.7% reduction in the molindone-treated group (ANOVA F= 50.0, df= 1, P<O.OOl). At the 200% masking dose, the haloperidol-treated group’s mean total AIMS score decreased by 53.2% of the baseline as compared to 23.4% in the molin- done-treated group (ANOVA F= 81.4, df = 1, P<0.0001).

DISCUSSION

The main finding of this study strengthens our preliminary observation that haloperidol masks withdrawal-exacerbated TD more than molindone. This difference in masking effect, as measured by absolute change and percent change in AIMS scores, is significant when the dose is 100% and 200% of the pre-study neuroleptic dose. These results contrast with those reported by another group who utilized the same study design in 26 chronically hospitalized male veterans (Borison et al., 1987). That group reported a ‘marginally sig- nificant trend’ for a greater masking of TD with molindone. This difference might be explained in part by the fact that their population was entirely male and more severely ill than ours. Additionally, all patients withdrawn from neuroleptics were studied whether they did or did not exhibit with-

drawal-exacerbated TD. As we have reported else- where (Glazer, June, 1989), patients with with- drawal-exacerbated TD may be pharmacologically distinct from patients who do not exhibit this phenomenon.

There are several interpretations for the finding in this study, including some that involve limita- tions to the study design. First, there is the possi- bility that we made incorrect dose equivalency calculations among the various neuroleptic agents and, specifically, between molindone and haloperi- dol. We attempted to control for potency differ- ences by dosing patients in terms of percentage of the neuroleptic dose received prior to the study. By treating patients at a range of doses up to 200% of the pre-study dose, we attemped to address the possibility that we underestimated the potency of molindone. It must be kept in mind that the literature pertaining to dose equivalency among neuroleptic medications is unclear. There are no data to compare therapeutic dose-response curves of the various types of neuroleptics (Baldessarini, 1985) available. Many investigators use different ratios of equivalence, but the estimated therapeutic dose range for molindone is 10-l 5 mg for every 2-3 mg of haloperidol. Thus if the real equivalence was 2 instead of 3 mg of haloperidol to 15 mg of molindone, then the haloperidol-treated group would have received one third more neuroleptic than the molindone group, which, given the small sample size, could explain the findings. Therefore, it is still possible that molindone appears to be a less potent masker of TD because we underesti- mated its antipsychotic potency or overestimated haloperidol’s potency.

Second, the findings could be explained by the secondary effects of parkinsonism rather than the direct drug effects on TD (Gerlach and Simmels- gaard, 1978; Nordic Dyskinesia Study Group, 1986). The results of the present study do not support this notion. We saw no inverse relation- ship between TD movements and parkinsonism in either treatment group (data not presented).

Third, the course of TD may impact on the interpretation of the results of this study. The acute effects of these two medications on involuntary movements may not be evident after prolonged treatment. Indeed, another group reported no dif- ferences in masking or unmasking effects of several medications in a geriatric population after a

319

6-month observation period (Gerlach and Sim-

melsgaard, 1978). Since little is known about the meaning of withdrawal-exacerbated TD, it is possi- ble that patients exhibiting this phenomenon are at a different phase in the course of TD or represent a distinct pharmacological subgroup of the disorder. Thus, the findings of this study may not be general- izable to all cases of TD.

Fourth, assuming that the weaker masking effect of molindone as compared to haloperidol is due to the pharmacological properties of the neuroleptics studied, the mechanism(s) could involve molin- done’s preferential activity at presynaptic dopam- ine receptor sites (Bunney et al., 1975) or its diminished activity on A9 neurons as compared to

haloperidol (White and Wang, 1983). An important remaining question for future

study is if a neuroleptic’s ability to mask with- drawal-exacerbated TD is related to its ability to cause TD. Given that clozapine, which is felt to have no dyskinetogenic potential, in some studies tends to mask TD movements (e.g., Small et al., 1987) it is not obvious that weaker masking properties of a medication are associated with diminished potential to cause TD. However, as the exploration for novel neuroleptics proceeds, knowledge about effects of different neuroleptics on various human behaviors and the correlation of these effects in animal models will help to under- stand the mechanisms of action of these agents and

the diseases they are treating. An interesting question for future study is if the

ability of neuroleptics to mask withdrawal emer- gent TD is directly related to antipsychotic po- tency. If such a relationship exists, then the method employed in this study could be used to determine dose equivalencies of different neuroleptics in hu- man beings without resorting to measurements of psychosis.

In conclusion, this study demonstrates a feasible method to compare the masking activity of neuro- leptic medications. The results of this study suggest that molindone has a weaker TD masking poten- tial as compared to haloperidol, but this finding should not lead clinicians to conclude that molin- done is less likely to contribute to TD. It is hoped that the method employed in this study will con- tribute to the clarification of mechanisms underly- ing TD and the development of novel neuroleptic agents.

ACKNOWLEDGEMENTS

This study was presented at the American Psychi- atric Association Annual Meeting, Chicago, IL, May 14, 1987. It was supported in part by a grant from E.I. DuPont Pharmaceutical Company and National Institute of Mental Health Grant MH- 30929. The authors wish to thank Ann Armas for her help in preparing this manuscript.

REFERENCES

Alander, T., Grabowska-Andan, M. and Anden, N.E. (1980)

Physiological significance of dopamine autoreceptors as studied following their selective blockade by molindone. J.

Pharm. Pharmacol. 32, 780-782.

Baldessarini, R.J. (1985) Antipsychotic agents. Chemotherapy in Psychiatry: Principles and Practice. Harvard University Press, Cambridge, MA. pp. 14-92.

Borison, R.L.. Shah, T.H., White, T.H. and Diamond, B.I. (1987) Atypical and typical neuroleptics and tardive dyski- nesia. Psychopharmacol. Bull. 23. 218-220.

Bunney, S.B., Roth, R.M. and Aghajanian, G.K. (1975) Effects

of molindone on central dopaminergic neuronal activity and

metabolism: Similarity to other neuroleptics. Psychopharma-

col. Commun. 1, 3499358. Endicott, J. and Spitzer. R.L. (1978) A diagnostic interview:

The schedule for affective disorders and schizophrenia. Arch.

Gen. Psychiatry 35, 837-844. Fuller, R.W. and Snoddy, H.D. (1983) Molindone: higher doses

needed to block pergolide-induced elevation of serum corti-

costerone than to elevate dopamine metabohtes in brain. Life Sci. 33, 235772361.

Gerlach, J. and Simmelsgaard, H. (1978) Tardive dyskinesia

during and following treatment with haloperidol, haloperi- dol+ biperiden, thorazine and clozapine. Psycharmacology

59, 105-l 12. Glazer, W.M. (1989) Noradrenergic function and tardive dyski-

nesia. Psychiatr. Ann. 19, 297-301. Glazer, W.M. and Moore. D.C. (1981) A tardive dyskinesia

clinic in a mental health center. Hosp. Commun. Psychiatry 32, 572-574.

Glazer, W.. Hafez, H. and Benarroche, C. (1985) Molindone

and haloperidol in tardive dyskinesia. J. Clin. Psychiatry 46, 4-7.

Greengard, P. (1975) Presynaptic nd postsynaptic roles of cyclic AMP and protein phosphorylation at catecholaminergic syn- apses. In: 0. Almgren, A. Carlsson and J. Engel (Eds.), Chemical Tools in Catecholamine Research. Elsevier, Am- sterdam.

Klawans, H. (1973) the pharmacology of tardive dyskinesia. Am. J. Psychiatry 130, 82-86.

Keller, W., Curtin, J. and Fields, J. (1984) Molindone com-

pared to haloperidol in a guinea-pig model of tardive dyski- nesia. Neuropharmacology 23. 1191~1194.

320

McMillen, B.A. and McDonald, C.C. (1983) Selective effects of

buspirone and molindone on dopamine metabolism and

function in the striatum and frontal cortex of the rat.

Neuropharmacology 22, 273-278.

Meller, E. (1982) Chronic molindone treatment: Relative inabil-

ity to elicit dopamine receptor supersensitivity in rats. Psy-

chopharmacology 76, 2222227.

Nordic Dyskinesia Study Group (1986) Effect of different

neuroleptics in tardive dyskinesia and parkinsonism. Psycho-

pharmacology 90, 4233429.

Overall. J.E. and Gorham, D.R. (1962) The brief psychiatric

rating scale. Psychol. Rep. i0, 7999812.

Owen, R.R. and Cole, J.O. (1989) Molindone hydrochloride: A

review of laboratory and clinical findings. J. Clin. Psychophar-

macol. 9, 2688276.

Richelson, E. (1985) Pharmacology of neuroleptics in the

United States. J. Clin. Psychiatry 46, 8814.

Small, J.G., Milstein, V., Marhenke, J.D., Hall, D.D., and

Kellams, J.J. (1987) Treatment outcome with clozapine in

tardive dyskinesia, neuroleptic sensitivity, and treatment-

resistant psychosis. J. Clin. Psychiatry 48. 263-267.

Smith, J.M., Kucharski, L.T. and Oswald, M. (1979) A system-

atic investigation of tardive dyskinesia in inpatients. Am. J.

Psychiatry 136, 918-922.

Spitzer, R. and Endicott, J. (1978) Schedule for affective

disorders and schizophrenia lifetime version (SADS-L), 3rd

edn. New York State Psychiatric Institute. New York.

Webster. D. (1968) Critical analysis of the disability in Parkin-

son’s disease. Mod. Treatm. 5, 2577282.

White, R.J. and Wang. R.Y. (1983) Differential effects of

classical and atypical antipsychotic drugs on A9 and AI0

dopamine neurons. Science 22 I, lO54- 1057.