Clinical characteristics of neuroleptic-induced parkinsonism

Transcription

1 J Neural Transm (2001) 108: Clinical characteristics of neuroleptic-induced parkinsonism S. Hassin-Baer 1,*, P. Sirota 2, A. D. Korczyn 1,3, T. A. Treves 1,**, B. Epstein 2, H. Shabtai 1, T. Martin 2, Y. Litvinjuk 2, and N. Giladi 1 1 Movement Disorders Unit, Department of Neurology, Tel-Aviv Medical Center, Tel-Aviv, 2 Department 6A, Abarbanel Mental Health Center, Bat-Yam, 3 Sieratzki Chair of Neurology, Tel-Aviv University, Ramat Aviv, and Sackler School of Medicine, Tel-Aviv University, Tel-Aviv, Israel Received May 8, 2001; accepted June 10, 2001 Summary. In order to characterize the clinical spectrum of neurolepticinduced parkinsonism (NIP), we studied a population of consecutive psychiatric in-patients treated with neuroleptics for at least two weeks, who were diagnosed by their psychiatrist as having parkinsonism. Parkinsonism was confirmed by a movement disorders specialist who performed neurological assessment including the motor examination and the activities of daily living (ADL) sections of the Unified Parkinson s Disease Rating Scale (UPDRS), and the Hoehn and Yahr (H&Y) staging. Seventy-five patients (54 males), aged years (range 21 to 73 years) were included in the analysis. The mean duration of neuroleptic therapy was years, while 61% were treated for more than 10 years. Most of the patients (n 66, 88%) were scored as H&Y stage 2.5 or less. Rest tremor was present in 44% of the patients, and usually persisted in action. Forty-one patients (61%) had symmetrical involvement. Parkinsonian signs were significantly more common and pronounced in the upper in comparison with the lower limbs (p ). Gait disturbances were mild and freezing of gait was very rare (n 2). Neither age nor duration of therapy or their interaction affected the total motor score or any of the motor sub-scores. In conclusion, NIP differs from PD for more bilateral involvement with relative symmetry, and by affecting upper limbs more often than the lower ones. NIP tends to be associated with the triad of bradykinesia, tremor and rigidity while PD tends Present addresses: * Department of Neurology, Sheba Medical Center, Tel-Hashomer, Israel ** Department of Neurology, Rabin Medical Center, Campus Beilinson, Petach Tikva, Israel

2 1300 S. Hassin-Baer et al. to involve gait and posture more often. NIP develops unrelated to duration of neuroleptic treatment or age of the patient, suggesting an individual predisposition to blockage of the dopaminergic receptors. Keywords: Drug induced parkinsonism, neuroleptic induced parkinsonism, antipsychotics, extrapyramidal side effects, symmetry. Introduction Patients treated with neuroleptic agents, that is dopamine receptor blocking agents (DRBA) frequently exhibit extrapyramidal adverse events. Among these are dystonia, akathisia, dyskinesia and parkinsonism (Korczyn and Goldberg, 1976; Diederich and Goetz, 1998). Ever since the introduction of DRBA for the treatment of psychosis forty years ago, neuroleptic-induced parkinsonism (NIP) became a common movement disorder. Epidemiological studies provide a wide range of NIP prevalence in different studies from 5% to 90% (Ayd, 1961; Kennedy et al., 1971; Korczyn and Goldberg, 1976; Stephen and Williamson, 1984). Distinct diagnostic criteria and patient populations, as well as exposure to different types and doses of neuroleptics may explain these differences. The clinical features of NIP develop within 1 to 3 months after introduction of DRBA or dose increment (Stephen and Williamson, 1984). After discontinuation, recovery occurs in 60 70% of the patients within 7 weeks, although complete recovery may take 6 months or more (Klawans et al., 1973). Some patients do not recover and continue to exhibit parkinsonian features and a few deteriorate following drug withdrawal (Melamed et al., 1991). Another group of patients recovers temporarily until a syndrome identical to Parkinson s disease (PD) reappears months to years later (Kennedy et al., 1971; Wilson and Primrose, 1986). These observations cannot be fully explained by the accepted pathogenetic hypothesis, which is based on dopamine receptor blockade. Positron emission tomography (PET) studies showed reduced F 18 -fluoro-dopa uptake in some patients with NIP, implying presynaptic dopaminergic nerve terminal dysfunction as a contributory factor (Burn and Brooks, 1993). Dopaminergic presynaptic dysfunction can be the result of primary degenerative process i.e. PD (Rajput et al., 1982) or the result of toxic effects of neuroleptics on these terminals. Several groups have described the clinical features of NIP. In most of the studies the presence of various signs or symptoms were reported, without rating their severity or mentioning their exact topographical distribution (Kennedy et al., 1971; Korczyn and Goldberg, 1976; Stephen and Williamson, 1984; Diederich and Goetz, 1998). The frequency of axial involvement (e.g. falls and gait disturbances) was rarely addressed and comparison to PD was not formally attempted. The goal of the present study was to examine the clinical features of NIP with the eyes and tools of movement disorders specialists, to find associated factors and to compare NIP to PD, in an attempt to improve our understanding of the relationship between these two disorders.

3 Clinical characteristics of neuroleptic-induced parkinsonism 1301 Methods The source population was composed of patients hospitalized in psychiatric departments at Abarbanel Mental Health Center in Bat-Yam, Israel, from December 1997 or July 1998 were receiving neuroleptic therapy for at least 2 weeks and were suspected of having parkinsonism by their treating psychiatrist. The study was approved by the hospital Helsinki committee. All patients agreed to take part in the study and signed an informed consent. The patients were examined by a movement disorders specialist, to confirm the presence of parkinsonism. Patients were included if they fulfilled the DSM-IV (1994) criteria for NIP, while parkinsonism was considered if at least 2 out of the 4 cardinal signs (tremor, bradykinesia, rigidity and postural instability) were present. All patients who fulfilled the inclusion criteria were taken off anticholinergic treatment for at least 24 hours and in most cases 48 hours before the motor assessment was performed according to the Unified Parkinson s Disease Rating Scale (UPDRS) (Lang and Fahn, 1989). The presence of other extrapyramidal features was not analysed. Demographic information, medical history, family history of psychiatric or neurological disorders, and careful history regarding pharmacotherapy were obtained from the patient, treating psychiatrist and hospital records. For the sake of statistical analysis UPDRS subscales were calculated for items from the ADL section and from the motor examination. The subscores definitions are listed below: TADLS Total ADL score the sum of all items of the ADL section (items 5 17). TMS Total motor score the sum of all motor examination items (items 18 31). GTS Global tremor score the sum of the tremor item by history and tremor at rest items by examination (items 16 and 20 for the face and four limbs). GBS Global bradykinesia score the sum of dexterity items (3 in each upper limb and 1 in each lower limb) and the general bradykinesia item of the motor examination (items 23, 24, 25, 26 and 31). GRS Global rigidity score the sum of 5 items of rigidity from the motor examination (item 22). PIGD score Postural instability gait difficulty score the sum of items from history regarding falling, freezing and walking with items of gait and postural instability from the motor examination (items 13, 14, 15, 29 and 30). GS Gait score the sum of the item from history concerning walking and freezing and the gait item from examination (items 14, 15, 29). UBS Upper body score the sum of all motor items from the upper limbs (the mean of right and left upper limbs rest tremor, the mean of right and left upper limbs rigidity and the mean of 3 items of dexterity in the right and left upper limbs) (items 20, 22, 23, 24 and 25). LBS Lower body score the sum of all motor items from the lower limbs (the mean of right and left lower limbs rest tremor, the mean of right and left lower limbs rigidity and the mean of right and left leg agility) (items 20, 22 and 26). RS Right score the sum of all motor items from the right upper and right lower limbs (tremor at rest, rigidity, 3 items of dexterity for right upper limb and one item for the lower) (items 20, 22, 23, 24, 25 and 26 of right side). LS Left score the sum of all motor items from the left upper and left lower limbs (rest tremor, rigidity, 3 items of dexterity for left upper limb and one item for the lower limb) (items 20, 22, 23, 24, 25 and 26 of left side). The associations between the different extrapyramidal features [e.g. tremor (GTS), rigidity (GRS), bradykinesia (GBS) or postural instability and gait disturbances (PIGD)] were assessed using Pearson correlation test. The comparisons between upper to lower and right to left body scores were done using pairedt test. However for the comparison of

4 1302 S. Hassin-Baer et al. asymmetry, we used only the absolute differences between sides in the first case. The scores obtained in the right and left sides were also reported on a scatter plot and linear regression lines were used to reflect asymmetry: in case of perfect symmetry the line is expected to have a slope of 45 degrees and regression coefficient of 1. The number of symmetrical cases, defined by RS-LS 1 was determined. The association between clinical features (expressed by subscores) and age, duration of therapy, or their interaction was evaluated using multiple linear regression analysis. Statistical significance was defined for p Results Out of 90 patients who were referred for evaluation by the treating psychiatrist, 75 patients (54 male and 21 females, male to female ratio (2.6:1) met inclusion criteria and were included in the analysis. The male predominance found in our study population was comparable to the patient population in the hospital at the time of the study. Ten patients did not meet inclusion criteria of parkinsonism, 2 did not stop therapy with anticholinergics and 3 refused to be examined. The mean age of the study population was 46.0 years (standard deviation (SD) 13.3, median age 43.0, range years). Sixty-one patients (81%) had been diagnosed with schizophrenia, 6 (8%) with a schizoaffective disorder and 2 (3%) with affective disorders (the remainder had other diagnoses). Thirty-four patients (45%) were treated with haloperidol and the others with other DRBA s, of whom 10 (13%) were treated with atypical neuroleptics (risperidone, clozapine or olanzapine) and 3 patients were on a combination of two neuroleptics each. The mean duration of exposure to neuroleptic therapy was 15.4 years (SD 11.8, range: 2 weeks 43 years) and 61% of the patients were treated for longer than 10 years (Fig. 1). Sixty-six patients (87%) were treated with an anticholinergic agent, which was withheld at least 24 hours prior to evaluation. Only one of the patients reported family history of parkinsonism unrelated to neuroleptic treatment, while 25 patients (33.3%) had a positive family history of psychiatric disorders. > <0.5 Duration of Treatment (years) Number of Patients Fig. 1. Duration of neuroleptic treatment among 75 patients with neuroleptic-induced parkinsonism. The mean duration of neuroleptic therapy was 15.4 years (SD 11.8, range: 2 weeks 43 years) and 61% of the patients were treated for more than 10 years

5 Clinical characteristics of neuroleptic-induced parkinsonism 1303 The ADL section of the UPDRS Most patients had only minimal disability or none at all, but drooling, tremor and speech difficulties were the most common symptoms (data not shown). Falling or prominent gait difficulties were rare (n 10 and n 9, respectively) and only two patients reported freezing of gait. The motor examination section of the UPDRS Most patients had mild parkinsonism according to the motor examination, as reflected by the mean motor subscores (Table 1). Hoehn and Yahr (H&Y) stage Sixty-six patients (88%) had a H&Y stage 2.5, of whom 45 were in stage 2. Unilateral cases, either with (stage 1.5, n 2) or without (stage 1, n 3) axial involvement were very rare (4% and 3%, respectively). Only eight patients (11%) were in stage 3, 1 patient in stage 4 and none in stage 5. The mean H&Y stage of the study population was Tremor Rest tremor was present in 44% of the patients (n 33), being more common in the upper limbs than in the lower limbs (41% versus 20%, respectively). Only 2 patients had tremor limited to the lower limbs. In most patients (n 53, 71%) tremor was strictly symmetrical and in others there was only a small difference between the right and left sides in severity. Tremor involving the facial structures was present in 15 patients (20%). Action tremor of the upper limbs was present in 44% of the patients (n 33), strongly correlating with Table 1. The motor performance as scored in subscales of the UPDRS and the ADL score of the UPDRS in 75 patients with NIP Subscales 8 Maximum Mean SD Range obtainable score Total motor score , 72 Global tremor score , 18 Global bradykinesia score , 28 Global rigidity score , 18 Upper body score , 9 Lower body score , 8 Gait score , 8 Postural impairment gait difficulty , 20 Right score , 23 Left score , 25 Total ADL score* , 40 8 for UPDRS items included in each subscales see text. UPDRS Unified Parkinson s Disease Rating Scale, ADL Activities of Daily Living, NIP Neuroleptic-induced parkinsonism, SD standard deviation. *For the total ADL score, median 4

6 1304 S. Hassin-Baer et al. rest tremor in these limbs (Spearman correlation coefficient 0.5, p ). Thirteen patients (17%) had action tremor of the upper limbs without rest tremor. Rigidity All patients exhibited rigidity, most common in the neck (93%), followed by the upper limbs (84%) and least in the lower limbs (57%). Gait impairment Gait impairment was observed in 31 patients (41%) mostly in the form of mild slowing (27 patients). Only 4 patients had clear parkinsonian gait with some shuffling and one of them could not walk unassisted. Only two patients reported freezing of gait, which was not seen during our exam. Postural stability Significant postural instability was observed in 9 patients (12%), while 17 patients (23%) exhibited minimal impairment in the pull test. Symmetry The parkinsonian signs were symmetrical in 61% of the patients (n 46). The right and left scores had a linear correlation with a slope of 0.95 (p ) (Fig. 2). This indicates an almost perfect symmetrical clinical picture. Upper versus lower body involvement UBS ( , mean SD) was significantly higher than LBS ( , mean SD; p ; Table 1). Fig. 2. Linear regression curve of right score (RS) versus left score (LS)

7 Clinical characteristics of neuroleptic-induced parkinsonism 1305 Table 2. Correlation between the motor subscales used for assessment of parkinsonism GTS GS PIGD GBS GRS (0.4) (0.0001) (0.0001) (0.0001) GBS (0.003) (0.0001) (0.0001) PIGD 0.21 (0.072) GS 0.23 (0.046) * The numbers in the cells are Pearson correlation coefficients and their corresponding p values. The correlation was performed only for subscales which do not include common items. GTS Global tremor score, GRS Global rigidity score, GBS Global bradykinesia score, PIGD score Postural instability gait difficulty score, GS Gait score Intercorrelation between parkinsonian features All parkinsonian features correlated well with each other except between tremor and rigidity, while correlation between tremor PIGD was borderline (Table 2). Association with age and treatment duration Using multiple linear regression analysis we found no association between the total motor score and the patients age or duration of therapy. Neither were any of the motor subscores (GTS, GRS, GBS, PIGD or GS) associated with either the patient s age or duration of neuroleptic therapy. Discussion We describe the clinical features of 75 psychiatric in-patients with NIP. Only 1% of our patients reported a first-degree or second-degree relative affected with parkinsonism unrelated to neuroleptic treatment, a finding that does not support an association between NIP and PD. The severity of parkinsonian features in NIP is variable (Ayd, 1961; Kennedy et al., 1971; Korczyn and Goldberg, 1976; Stephen and Williamson, 1984; Hardie and Lees, 1988). Most patients with NIP in the study by Korczyn and Goldberg were asymptomatic (Korczyn and Goldberg, 1976). In Kennedy s series (Kennedy et al., 1971), where all chronic schizophrenic patients from two hospitals who have been treated with trifluoperazine were assessed, extrapyramidal symptoms did have clinical significance: 14% had dressing difficulties, 27% postural abnormalities and 49% gait disturbances. These disabilities were not significantly associated with the parkinsonian signs, but rather to other extrapyramidal motor abnormalities, like akathisia

8 1306 S. Hassin-Baer et al. or dyskinesia (Kennedy et al., 1971). On the contrary, in the series by Stephen and his colleagues, of symptomatic elderly patients with drug-induced parkinsonism, many patients had severe parkinsonism, very similar to PD. These patients were referred to hospitalization due to severe motor impairment. The authors concluded that the similarity between NIP and PD in their series might be due to the fact that neuroleptics unmask latent PD, which is common in the elderly population (Stephen and Williamson, 1984). In contrast, most of our NIP patients had no functional disability although some were bothered by symptoms such as speech difficulty, drooling or tremor. The methodology of the study may account for the preponderance of mild cases, since our patients were selected by their psychiatrists due to signs of parkinsonism regardless of any complaints. Although most of the patients (87%) were treated by anticholinergics, these drugs were withdrawn hours before evaluation; therefore the symptomatic benefit of the anticholinergic drugs could play a minor role by decreasing the severity of the parkinsonian signs. Except for accentuation of parkinsonism, anticholinergic withdrawal precipitated motor and behavioral disorders such as restlessness and akathisia, discussion of which is beyond the scope of this article. The quantification of parkinsonian features in NIP patients using the UPDRS allowed us to discriminate between upper and lower limb involvement, and depict the relative segmental distribution of symptoms (e.g. right versus left or upper versus lower body involvement). Korczyn and Goldberg (1976) rated parkinsonian features of hypokinesia, rigidity, tremor, facial mask, gait disturbance, drooling and micrographia, using an arbitrary severity scale of 0 3, but the authors did not describe the relative distribution of signs in different body parts. Hardie and Lees assessed 26 patients with NIP aged 21 to 71 using the Webster rating scale: they found no specific features for NIP. The findings were commonly asymmetric and tremor was of greater or equal severity to other cardinal signs in 53% of the patients (Hardie and Lees, 1988). Kennedy and his colleagues (Kennedy et al., 1971) reported that tremor was very common, occurring in 88% of the patients (when doubtful and mild cases were excluded it was 48%), and rigidity was found in 68% of the patients (38% withsignificant manifestations). In Ayd s series tremor occurrence was 60%, usually beginning in the upper limb (Ayd, 1961). These figures are similar to ours, where tremor prevalence was 44%. Is NIP part of the spectrum of PD? In order to answer this question we attempted to compare our data with similar data concerning PD patients. In the PD database at Columbia Presbyterian Medical Center (CPMC) in 1992 the prevalence of tremor in PD was 77% (from a total of 1,347 PD patients), while in patients with drug-induced parkinsonism it was 52% (from a total of only 27 patients included in the analysis). Motor blocks (freezing phenomenon) occurred in 31% (out of 1,227) of PD patients and in 8% (out of 24) of drug-induced parkinsonism patients. Segmental distribution cannot be compared since it was not reported for PD patients. However, the CPMC database was composed of early as well as advanced cases with a preferential referral bias of more advanced cases (Fahn, 1997). For this reason we looked at the data reported from the Deprenyl and Tocopherol Antioxidative Therapy of

9 Clinical characteristics of neuroleptic-induced parkinsonism 1307 Parkinsonism (DATATOP) study of recently diagnosed, unmedicated patients. In our own present series the parkinsonism in NIP is different from that reported from the DATATOP study, where the baseline data from early dopa-naive PD patients is given. Most NIP patients were at H&Y stages of 2 3 and only 7% (n 5) were unilateral. In most of our cases, this implies a bilateral disease with or without some postural involvement and not an advanced disease. This is different from the DATATOP study where one of the inclusion criteria was H&Y stage under 2.5 and 50% of the patients had a unilateral disease a typical feature of early PD (Jankovic et al., 1990). The prevalence of rest tremor is lower in NIP than in PD as reflected in the DATATOP study, where more than 70% of the patients had rest tremor (Jankovic et al., 1990). In NIP the rest tremor is usually associated with an action component in the same limb. Many patients have an action tremor alone, an uncommon finding in PD. Upper limbs were involved significantly more often than the lower limbs. In the DATATOP study 7.1% of the patients had freezing of gait (Giladi et al., 2001), while we found 3% (n 2) in the NIP patients. Prominent gait and postural abnormalities among patients with NIP are uncommon as well. These differences could be partially attributed to the fact that most patients had mild parkinsonism, which did not progress to affect gait and postural reflexes in spite of long-term exposure to neuroleptics. We found no association between the severity of the parkinsonian syndrome or any of the motor subscores to the age of the patients or the duration of therapy. The same was shown by Korczyn and Goldberg (1976). Several other authors, however, found that increasing age rather than the duration or the dosage of neuroleptic therapy was associated with increased parkinsonian severity (Ayd, 1961; Kennedy et al., 1971; Hardie and Lees, 1988). The present study contributes little to the understanding of the basic question of the pathogenesis of NIP. Our results do not provide further evidence in favor of the pathogenetic theory of neuroleptic toxicity to nigrostriatal dopaminergic system, or the unmasking of age-related dopaminergic system degeneration in most cases of NIP. Dopamine receptor blockade is probably the main mechanism of NIP. Individual susceptibility to blockade of the D 2 dopamine receptors as well as to interactions with other neurochemical circuits may be responsible for the variable severity of NIP. This susceptibility may be related to receptor affinity to DRBA s, or perhaps efficiency of compensatory mechanisms that can bypass the dopaminergic blockage. Dose or potency effect of the neuroleptic agent was not assessed in our study due to the changes in medications in these hospitalized patients making it one of the study s limitations (Ayd, 1961). In conclusion, our study showed that NIP differs from PD in terms of some clinical symptoms. NIP is more symmetrical and tends to be associated with the triad of bradykinesia, rigidity and tremor, while PD tends to involve gait and posture more commonly. In contrast to PD, which worsens with time, NIP seems to be unassociated with duration of neuroleptic treatment. In spite of the presence of parkinsonian symptoms the functional disability and awareness in NIP is mild and partial. Future studies should look more carefully at

10 1308 S. Hassin-Baer et al.: Clinical characteristics of NIP the metabolic changes of brain function secondary to dopaminergic blockage by DRBA s and specifically at possible compensatory mechanisms. References American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders. American Psychiatric Press, Washington DC Ayd F (1961) A survey of drug-induced extrapyramidal reactions. JAMA (175): Burn DJ, Brooks DJ (1963) Nigral dysfunction in drug induced parkinsonism: an F-dopa PET study. Neurology 43: Diederich NJ, Goetz CG (1998) Drug-induced movement disorders. Neurol Clin 16(1): Fahn S (1997) From the PD database at Columbia Presbyterian Medical Center, A comprehensive review in movement disorders for the clinical practitioner. Columbia University, pp Giladi N, McDermott M, Fahn S, Prodborski S and the Parkinson study group (2001) Freezing of gait in Parkinson s disease. Neurology 56(12): Hardie RJ, Lees AJ (1988) Neuroleptic-induced Parkinson s syndrome: clinical features and results of treatment with levodopa. J Neurol Neurosurg Psychiatry 51(6): Jankovic J, McDermott M, Carter J, et al (1990) Variable expression of Parkinson s disease: a base-line analysis of the DATATOP cohort. The Parkinson Study Group. Neurology 40(10): Kennedy PF, Hershon HI, Mcguire RJ (1971) Extrapyramidal disorders after prolonged phenothiazine therapy. Br J Psychiatry 118(546): Klawans HL, Bergen D, Bruyn GW (1973) Prolonged drug-induced Parkinsonism. Confin Neurol 35(6): Korczyn AD, GJ Goldberg (1976) Extrapyramidal effects of neuroleptics. J Neurol Neurosurg Psychiatry 39(9): Lang AE, Fahn S (1989) Assessment of Parkinson s disease. Quantification of neurological deficit. T Munsat Stoneham, Mass, Butterworths, pp Melamed E, Achiron A, Schapira A, Davidovicz S (1991) Persistent and progressive parkinsonism after discontinuation of chronic neuroleptic therapy: an additional tardive syndrome? Clin Neuropharmacol 14(3): Rajput AH, Rozdilsky B, Hornykiewicz O, Shannak K, Lee T, Seeman P (1982) Reversible drug-induced parkinsonism. Clinicopathologic study of two cases. Arch Neurol 39(10): Stephen PJ, Williamson J (1984) Drug-induced parkinsonism in the elderly. Lancet 2(8411): Wilson JA, Primrose WR (1986) Drug induced parkinsonism. Br Med J (Clin Res Ed) 293(6552): 957 Authors address: Dr. N. Giladi, Movement Disorders Unit, Tel-Aviv Medical Center, 6 Weizmann St., Tel-Aviv 64239, Israel, ngiladi@tasmc.health.gov.il