Updated Perspectives on the Management of Drug-Induced Parkinsonism (DIP): Insights from the Clinic

Introduction

Parkinsonism refers to the clinical combination of bradykinesia, rigidity, tremor, and postural instability. Parkinsonism can be due to a primary neurodegenerative disorder, as in Parkinson’s disease; or it can be secondary, as can be seen in metal deposition diseases (like manganese toxicity or Wilson’s disease). Parkinsonism can also be iatrogenic, which is the topic of this review.

The use of neuroleptics, such as typical and atypical antipsychotics (among other drug classes), can cause an acute to subacute drug-induced parkinsonism (DIP) or may unmask underlying idiopathic Parkinson’s disease (PD). DIP was first described in 1954 in patients treated with chlorpromazine and reserpine.¹ Symptoms can manifest within days of medication initiation, with the vast majority emerging by 3 months.² A second peak of symptom onset at 12 months of exposure has been reported, mainly with the use of calcium channel blockers.³ The parkinsonism tends to be bilateral and symmetric, but this is not a rule.

Typical antipsychotics are found to bind more tightly than dopamine to the D2 receptor and have lower dissociation constants. In contrast, atypical antipsychotics are found to bind more loosely than dopamine, with higher dissociation constants, adding to the “fast-off” hypothesis.⁴ One pharmacokinetic study showed that the typical antipsychotic haloperidol had 6.4- and 3.9-times slower rates of dissociation compared to the atypical antipsychotics clozapine and quetiapine, respectively.⁵ Additionally, clozapine exerts its therapeutic effect primarily by D1 receptor antagonism, whereas typical antipsychotics such as haloperidol exert their effects via D2 receptor blockade.⁶ This difference in dopamine receptor affinity may contribute to the increased risk of extrapyramidal side effects.

With discontinuation of the offending medication, most cases of parkinsonism will reverse within days to months – strongly suggesting DIP.² About two-thirds of patients recover within 7 weeks, some persist for up to 18 months.⁷ Alternatively, symptoms may persist and slowly progress or cease only to recur with time. When symptoms persist and worsen with time, idiopathic PD (which was preclinical at the time of neuroleptic exposure) should be suspected. The term tardive parkinsonism has been used for patients with persistent symptoms after discontinuation, but pathologic evidence suggests that most cases of tardive parkinsonism are really unmasked PD.⁸ Pre mortem, these entities can be differentiated with a dopamine transporter (DAT) scan – which will be normal in drug induced or tardive parkinsonism.⁹

In the clinic, a neurologist may be asked to help differentiate drug-induced parkinsonism from idiopathic PD. While ancillary tests, such as metaiodobenzylguanidine (MIBG) cardiac scans¹⁰ and DAT scans can be confirmatory, a proper history, with a focus on medication review, nonmotor features, and physical exam can help guide one towards the correct diagnosis. Early recognition and management are important and have long-term implications as DIP is a treatable condition, and without appropriate management can lead to falls, nursing home placement and increased mortality.¹¹ Recognition will lead to early withdrawal of culprit agents and negate the need for dopaminergic therapies.

In this manuscript, we differentiate DIP and idiopathic PD, list culprit medications in the development of DIP, discuss motor and nonmotor signs and symptoms that can help with differentiation, and detail the management of DIP.

Diagnostic Criteria

In the DSM-V, DIP is defined as:

Parkinsonian tremor, muscular rigidity, akinesia (i.e., loss of movement or difficulty initiating movement), or bradykinesia (i.e., slowing movement with decremental response) developing within a few weeks of starting or raising the dosage of a medication (e.g., a neuroleptic) or after reducing the dosage of a medication used to treat extrapyramidal symptoms.¹²

It is important to note that unlike the diagnostic criteria for Parkinson’s disease, bradykinesia is not required for the diagnosis of DIP.

Pathophysiology

The classical schematic of the basal ganglia circuit is useful in understanding the pathophysiology of drug-induced parkinsonism. Neuroleptics, including antipsychotics, antiemetics, and other dopamine antagonists block D2 dopamine receptors in the striatum. Reduction of dopamine transmission in the striatum leads to disinhibition of the indirect pathway via GABA- and Enkephalin-containing striatal neurons. The subthalamic nucleus is disinhibited, leading to excitation of the globus pallidus and substantia nigra pars reticulata. In turn, the thalamocortical projections are further inhibited. Bradykinesia and other signs of parkinsonism are the manifestations of this biochemical cascade and subsequent imbalance in the direct and indirect pathways.¹³ The delay in improvement after withdrawal of the medication may be due to long half-lives and the persistence of neuroleptic agents in body tissue. Older neuroleptics are highly surface-active and fat-soluble and readily concentrated in cell membranes.

The disruption in the dopaminergic-cholinergic balance caused by neuroleptics may explain why parkinsonism occurs, and why anticholinergics are beneficial in addressing these symptoms. Neuroleptics cause an increase in dopamine turnover which persists following destruction of the strio-nigral pathway. Anticholinergics antagonize the increase in dopamine turnover caused by neuroleptics by decreasing activity of tyrosine hydroxylase. Anticholinergics such as benztropine can also inhibit reuptake of dopamine into nerve terminals.¹⁴

Causative Drugs

DIP has been described as the sequela of neuroleptics including typical and atypical antipsychotic medications, though the risk is low with quetiapine and clozapine. A summary of culprits is listed in Table 1.

Table 1 Potentially Causative Agents in the Development of DIP

Antipsychotics

Antipsychotic agents cause DIP by antagonizing central dopamine receptors as described above. A recent meta-analysis suggests that the pooled prevalence of extrapyramidal symptoms in patients taking antipsychotics is as high as 20%.¹⁹ Another study involving every resident of Piedmont, Italy, older than 39 years suggests that the risk of Parkinsonism is threefold higher in residents who were taking antipsychotics.²⁰ The data also indicated that typical antipsychotics were more likely than atypical antipsychotics to put the patient at risk for parkinsonism. A significant caveat is that it is not clear how many of the patients studied had underlying idiopathic PD, which was simply unmasked by exposure to the medication.

Nonetheless, literature such as this underscores the significant burden of parkinsonism in patients taking both first- and second-generation antipsychotics. It is also crucial to understand the differential risk for DIP the two generations carry, which is rather nuanced.

Typical (also known as first-generation) antipsychotics have a higher affinity for central D2 receptors than atypical (second-generation) antipsychotics.²⁰ The preference of atypical antipsychotics to antagonize serotonin receptors rather than D2 receptors is thought to contribute to the lower overall incidence of extrapyramidal symptoms compared to typical antipsychotics.²¹ Additionally, it is thought that the more rapid dissociation from D2 receptors contributes to the overall lower risk.²²

However, the difference likely depends on the dose and also the dopamine receptor affinity. For example, a meta-analysis of all randomized controlled trials comparing second-generation antipsychotics with first-generation antipsychotics at lower doses indicated that the difference was not significant.²³ In corollary to this, a retrospective cohort study shows that high-dose second-generation antipsychotics have a similar risk of developing DIP to first-generation antipsychotics.²⁴ In particular, only the typical agents with the highest potency (haloperidol, perphenazine, and thiothixene) were found overall to pose a higher risk than the atypical antipsychotics.

There are additional differences in risk between the different second-generation antipsychotics. One meta-analysis of 54 studies suggests that patients on risperidone were more likely to require treatment for parkinsonism than patients taking ziprasidone, who were more likely to require treatment than patients taking olanzapine, quetiapine, and zotepine. All of these populations were more likely to need treatment than patients taking clozapine.³ Usage in the elderly further brings out the differences between second-generation antipsychotics. To date, only clozapine and quetiapine have evidence of lower DIP rates in the elderly.²⁵

Non-Antipsychotic Neuroleptics

Certain antihistamines are known to increase the risk of DIP.³ In fact, the first-generation antihistamines portend a much higher risk, which is thought to be due to the fact that these drugs are phenothiazine derivatives.³ Phenothiazine is a peripheral and central D2 antagonist, which is why it is classified in this review as a neuroleptic.

Certain antiemetics are also known to increase the risk of DIP, particularly those with dopaminergic antagonism. These anti-emetics are listed in Table 1 and are metoclopramide, clebopride, promethazine, and domperidone. Although domperidone, for example, is known to cross the blood–brain barrier less easily,²⁶ this relative restriction is less of a factor in children and the elderly. This is one of the reasons why it is suspected that these antiemetics have a higher risk of creating DIP in children and the elderly compared to the remaining population.³

The benzamide derivatives are neuroleptic medications used for various indications including Huntington’s chorea (tiapride) and as a gastroprokinetic (cisapride). They have central anti-D2 receptor action and have been observed less frequently, to put patients at risk for DIP; therefore, the risk exists, but is much lower.³

Non-Neuroleptics

Numerous other medications without clearly denoted dopamine-receptor blockade in the literature are reported to cause DIP. Of particularly high risk for inducing the development of DIP are the agents designed to deplete central dopamine storage or synthesis.¹ These medications include α-methyldopa, tetrabenazine, deutetrabenazine, valbenazine, and reserpine.^32,33 This can lead to many challenges when treating, amongst other disorders, patients with Huntington’s Disease.

One area of some controversy is the relation of antidepressants to DIP. Although in general the mechanism of the medications should have no relation to parkinsonism, the reports in the literature are not infrequent. In the previously mentioned French pharmacovigilance study spanning 17 years, there were 21 reports of parkinsonism after using antidepressants, which constituted 8% of all reports. 12/21 reports were from patients taking SSRIs, 3 were from imipraminics (1 amitriptyline, 1 clomipramine, and 1 dosulepin), 5 were from patients taking venlafaxine, and 1 was from a patient taking mirtazapine. A different pharmacologic review found that starting serotonergic antidepressants in patients already using L-dopa was followed by a more rapid increase in antiparkinsonian drugs.³⁴ One case series in 2001 compiled about 100 cases of extrapyramidal symptoms associated with SSRI use.³⁵ That same report estimates that the annual incidence is about 1–2 cases per 1000 patients. One study of a 5-HT1A agonist, Sarizotan, was found to worsen symptoms of parkinsonism.³⁶ Serotonergic antidepressants have been shown in animals to decrease the concentration of dopamine in the striatum.³⁷ Therefore, some low-level evidence exists for a pathophysiological link, and a low but nontrivial rate of case reports suggests that antidepressants should be considered low but non-zero risk for causing DIP.

Of the antiepileptics, valproic acid and (less frequently) phenytoin³⁸ and per one case report oxcarbazepine³⁹ in particular have been reported to cause DIP. A collection of studies suggests that as high as 5% of patients taking valproic acid suffer from drug-induced parkinsonism.^31,40,41 The mechanism is yet unknown, but some authors suggest that oxidative stress and mitochondrial dysfunction are known to occur with valproate-induced toxicity, and note that this is also implicated in Parkinson's Disease.²³

Two calcium channel blockers, flunarizine and cinnarizine, have a long history with ties to DIP²⁷ and are widely reported to the point where in most sources it has its own name: FCIP, flunarizine-cinnarizine induced parkinsonism. The presence of pre-existing essential tremor and older age appear to be independent risk factors.²⁸ The majority of patients tend to be female, as high as 89% in one study with 74 patients.²⁹ The mechanism is not fully understood, but the hypotheses are important to understand in order to contrast these medications with other calcium-channel blockers, which appear to have a lower risk of DIP. Some authors believe that FCIP is caused by loss of dopamine precursor depletion as well as postsynaptic dopamine blockade.²⁷ Flunarizine and cinnarizine differ from other calcium-channel blockers in that they carry a piperazine nucleus, which is known to interfere with dopamine storage, release, and dopamine receptor functionality.³⁰ The presentation of FCIP tends to be an asymmetric akinetic-rigid syndrome, whereas traditional DIP caused by other agents is more frequently symmetric.³¹

Outside of cinnarizine and flunarizine, discussed above, calcium-channel blocker antihypertensives are considered low-risk for DIP. Verapamil and diltiazem are implicated in just a few case reports to date.^42–44 There is a case report of captopril-induced DIP,⁴⁵ despite a 2013 study suggesting that captopril protects nigrostriatal dopamine neurons in animal models of parkinsonism.⁴⁶ No other blood pressure agents appear to have been associated with DIP to date.

One non-neuroleptic mood stabilizer that has appeared in the literature related to DIP is lithium. Lithium is notorious, like many of the drugs mentioned above, for drug–drug interactions. Interestingly, a few of the reports of extrapyramidal symptoms related to lithium use were linked to pharmacokinetic interactions with bromocriptine in patients who already had pre-existing PD.³ Other cases have been published in patients who developed apparent DIP despite no prior symptoms.^46–48 The incidence appears infrequent, but underscores the importance of considering drug–drug interactions in the genesis of DIP.

Another rare but interesting class of medications associated with DIP is the cholinesterase inhibitors. This compound is already known to exacerbate symptoms of Parkinson's disease.⁴⁹ One author notes that one of the hypotheses behind why lithium may cause DIP is that it has some anticholinesterase activity.⁵⁰ Of note, this medication class (including donepezil, rivastigmine, and galantamine) shares a common piperazine nucleus, which is also true of flunarizine and cinnarizine as mentioned above.¹³ Despite this, anticholinergic medication has not yet shown clear evidence of benefit in DIP patients.⁵¹

Other miscellaneous medications that have been rarely reported to cause DIP are listed in Table 1. These medications include antiarrhythmics, some antibiotics, some immunosuppressants, and endocrine agents.

Epidemiology and Risk Factors

The accuracy of DIP prevalence is limited by misdiagnosis and myriad of criteria, leading to a lack of a single universally accepted criteria.¹ There are likely several reasons for underdiagnosis including focus on the illness for which the culpable agent has been started, and a propensity to underestimate the likelihood of newer neuroleptics to cause DIP, albeit in lower frequency. Additionally, in other missed cases, motor features may have been mild, asymptomatic, or delayed.¹ Conversely, there is evidence of overdiagnosis, namely in one academic center where only 1 of 24 referral for DIP ultimately retained the diagnosis.¹¹ Therefore, prevalence data are highly variable between studies. Population studies attempting to identify what percentage of patients with parkinsonian symptoms had a diagnosis of DIP yield numbers such as 22% in Central Spain,⁵² 20% in Olmsted, MN,⁵³ 37% in Bambui, Brazil,⁵⁴ and 5% of all cases in a collaborative study across Europe.⁵⁵ The disease skews towards women and the elderly, potentially due to the frequency of comorbidities, conditions requiring neuroleptics, hormonal influences, and medication interactions.³ Other independent risk factors include baseline extra-pyramidal signs, type of neuroleptic, severity of dementia,⁵⁶ family history of parkinsonism,⁵⁷ drug dosing,³ prior brain injury,⁵⁸ severity of psychiatric disease,⁵⁹ unexplained hyposmia⁶⁰ and HIV infection.⁶¹ Careful history taking should elicit potential drug exposure and other risk factors.

Clinical Manifestations/Diagnosis

Motor manifestations in DIP can be indistinguishable from idiopathic PD. The prevalence of certain motor features in DIP can be gleaned by published series of patients. Hassin-Baer et al reported the motor characteristics of 75 patients with PD.⁶² In this cohort, most, 61%, of patients had symmetrical symptoms. Tremor, including rest and action tremor, was present in 44%. Gait disorders were mild in this cohort, and freezing of gait was rare (seen in only 2 of 75 patients). A similar prevalence of asymmetric symptoms was reported by Sethi and Zamrini, in 60% of 20 patients.⁶³

The presence of associated signs and symptoms can help make a clinical distinction between DIP and PD. A diagnosis of DIP is supported by the co-occurrence of tardive syndromes.² These syndromes include the classic oral-buccal lingual and other facial dyskinesia, respiratory dyskinesia, akathisia, humming, dystonia, myoclonus, tics, and chorea.¹¹ Considering the non-motor symptom sphere can be helpful. Kim et al found that when compared to DIP patients, PD patients had significantly higher scores on the Non-Motor Symptoms Scale (NMSS). Urinary symptoms, sleep disturbances, concentration issues, and anosmia were significantly associated with PD compared to DIP.⁶⁴ Morley et al compared 97 patients with DIP and 97 age-matched control with PD and determined that abnormal olfaction, sexual dysfunctional, and constipation were significantly more common in PD, present in 88%, 49%, and 47%, respectively, compared to 57%, 30% and 30% in DIP patients.⁶⁵ Up to 60% of patient with PD have REM sleep behavior disorder (RBD),⁶ compared to less than 1% in the general population;¹⁰ and RBD is more prevalent in PD compared with DIP.⁶⁵ Sialorrhea has long been recognized as an extrapyramidal side effect of neuroleptic medications,⁶⁶ but can be seen in both PD and DIP, and the comparative prevalence has not been studied to our knowledge.

The diagnosis of DIP is grounded on the patient’s history of exposure to neuroleptics, or other culprit medications. The motor features of DIP and PD can be identical. Nonmotor symptoms should be elicited, and the absence of RBD, anosmia, or constipation are very pertinent negatives in the review of symptoms. Among the nonmotor symptoms, hyposmia may serve as a potential biomarker, assessable at the bedside, to distinguish DIP from an underlying degenerative synucleinopathy. In one study, among 16 DIP patients, those who had reduced tracer binding on DAT scans were also found to have abnormal olfactory function, suggesting a neurodegenerative etiology of their parkinsonism.⁶⁰ In another study, among patients with DIP, lower age- and sex-adjusted olfactory scores and higher rates of anosmia were found in patients with abnormal DAT scans.⁶⁷ Additionally, olfactory impairment may be able to predict response to drug withdrawal. In one study among 13 DIP patients, not only was hyposmia significantly more common in the underlying PD group but in 11 of 13 patients this was able to correctly predict symptomatic recovery after drug removal.⁶⁸ While nonmotor features provide suggestive clues, they do not provide diagnostic certainty.

Resolution of parkinsonism with discontinuation of medication is diagnostic. If you cannot wait, then Dopamine Transporter (DAT) imaging can provide rapid clarity, with abnormal imaging suggesting underlying idiopathic PD.⁶⁹ It is important to point out that DAT scans evaluate pre-synaptic terminal density, but it is the post-synaptic receptors that are typically stimulated by dopamine-receptor blocking agents. This is the core mechanism behind the scan’s utility in distinguishing DIP from PD. The use of DAT imaging has provided us with interesting retrospective data. Patients diagnosed with DIP who then go on to have positive DAT scans may differ clinically to those who go on to have negative scans. One article studied a population of 34 US veterans diagnosed with DIP, 12 of whom were later found to have nigrostriatal degeneration on DAT scan.⁷⁰ In the population with degeneration, there was a significantly higher burden of gait abnormalities and non-motor symptoms. Specifically, lower performance in olfactory function (also reported elsewhere), the Timed Up and Go test, and total nonmotor symptom burden was a significant predictor of an abnormal DAT scan in this population.⁶³ A similar retrospective study analyzed 51 patients at a movement disorders specialty center for whom a DAT scan was ordered to differentiate DIP and PD.⁷¹ The analysis noted that patients with greater than 2 cardinal manifestations of parkinsonism (tremor, rigidity, akinesia, postural instability) were significantly more likely (63.89% vs 93.33%, p = 0.04) to have a positive DAT scan, but that otherwise the syndromes were clinically similar. An interesting investigation of early-onset vs late-onset DIP, in relationship to neuroleptic initiation, notes that early-onset patients were more likely to have DAT abnormalities.⁷² This lends credence to the discussion that many patients diagnosed with DIP already have an underlying susceptibility to develop parkinsonism. Overall, these data suggest that some DAT-positive and DAT-negative patients undergoing workup for DIP vs PD may have distinct clinical features. These features are likely not sufficiently specific to make a diagnosis on their own, but still merit consideration for use alongside the other available clinical data when making practical management decisions.

Treatment

Prevention of DIP may be achieved by clinical monitoring of parkinsonism during the initial titration of culprit medications in order to determine the “neuroleptic threshold.”⁷³ Signs of DIP may be detected as early as within 24 to 48 hours of neuroleptics in young patients, and within 96 hours in the elderly.⁷⁴ Frequency of clinical monitoring can be decreased after the first three months post initiation or increase in dosage of a neuroleptic, as most cases of DIP present themselves within this time period.^75,76

Treatment of DIP can be achieved by dose reduction, discontinuation of the causative agent, or switching to an alternative agent. This process should be done in conjunction with the patient’s treating psychiatrist. Discontinuation of the causative drug is advised when feasible while maintaining the risk of relapse in mind.¹ For patients in whom their symptoms are stabilized, neuroleptic dose reduction would be a viable option. As noted above, due to the long half-lives of neuroleptic medications, the improvement of parkinsonism will not be immediate. With both dose reduction and discontinuation, there is a risk of withdrawal-emergent dyskinesia.

Transitioning to a drug with a lower affinity to D2 receptors is suggested if psychiatric symptoms worsen on a lower dose. Only quetiapine and clozapine have been shown not to worsen parkinsonism, with clozapine receiving Level A evidence in one review.⁷⁷ Of note, patients transitioning to clozapine must be monitored for the development of agranulocytosis and be warned about the potential development of hypersalivation, which can occur in up to 54% of those receiving this medication.⁷⁸

Anticholinergics such as trihexyphenidyl and benztropine are commonly used to treat DIP. Benztropine was found to effectively treat extrapyramidal symptoms in DIP at a dose of 4mg per day.⁷⁹ These medications should be avoided in the elderly due to their central effects of cognitive impairment, hallucinations, delusions, and delirium in addition to their peripheral side effects of tachycardia, dry mouth, urinary retention, and constipation. Use of anticholinergics may also lead to a decrease in the plasma concentration of neuroleptics.⁸⁰ Additionally, one review found that the use of benztropine hindered the efficacy of neuroleptic drugs among patients with schizophrenia, leading to worsening of positive symptoms.⁸¹ For these reasons, use of preventive anticholinergics in patients taking neuroleptics is discouraged.⁸²

Amantadine, including the extended-release formulation Osmolex, has FDA approval for the treatment of DIP. It may be a better option in elderly patients with DIP or in patients who cannot tolerate anticholinergic side effects. In a double-blind placebo-controlled crossover study, amantadine at a dose of 100mg twice daily was as effective as trihexyphenidyl 4mg twice daily in treating DIP and produced fewer side effects.⁵¹ One double-blind crossover study found no difference in efficacy between amantadine and benztropine.⁸³ In another double-blind study, amantadine was found to have fewer side effects than and comparable efficacy to benztropine in treating DIP, with the exception of benztropine having a greater effect on rigidity.⁸⁴ Additionally, a systematic review provided Level C evidence for the use of amantadine in tardive syndromes.⁸⁵

Regarding dopaminergics, levodopa is not commonly used for the treatment of DIP, and caution is encouraged in patients with psychosis.⁸⁶ One study found that among patients whose DIP symptoms were controlled by anticholinergics, when switched to levodopa, they experienced worsening of tremor, rigidity, ataxia, diaphoresis, anxiety, lethargy, and depression.⁸⁷ In another study, patients with parkinsonism and abnormal SPECT treated with levodopa experienced improvement in motor UPDRS scores over a three-month period, while patients with normal SPECT did not. However, this group with abnormal SPECT scan and levodopa responsiveness likely had idiopathic PD unmasked by neuroleptics, as this combination of findings would suggest an underlying degenerative cause for the observed parkinsonism rather than DIP.⁸⁸ One study showed positive results with levodopa: 15 patients with DIP due to antipsychotics were treated with levodopa (doses of 300–1000mg) for up to four years; most did not experience adverse effects: only one patient had worsening of their psychiatric state and two patients developed dyskinesia. However, of these patients, eight experienced either no or slight improvement in motor symptoms as measured by the Webster score. Only two of the patients who had the culprit drug withdrawn in this study experienced either complete or transiently complete (12 months) resolution of extrapyramidal symptoms.⁸⁹ Another study evaluated the use of the dopamine agonist rotigotine in patients with DIP due to antipsychotics. Twenty patients were treated with 3–8mg of rotigotine and were found to have a significant improvement in total UPDRS, UPDRS Part III, Simpson-Angus Scale, and Barnes Akathisia Rating Scale scores; this improvement was sustained for at least one month. All patients except one tolerated rotigotine well without worsening of psychotic symptoms.⁹⁰

Electroconvulsive therapy (ECT) has also been found to improve the motor symptoms of DIP. In one randomized control trial, fourteen of twenty-five patients experienced transient improvement in their parkinsonism and psychosis after ECT.⁹¹ ECT has also been found useful in elderly patients with DIP, with improvement in motor symptoms noted up to six months after treatment.⁹²

Though exercise has been found to be beneficial in PD, with improvements in disease symptoms, mobility, balance, gait and quality of life, studies have not proven its efficacy in DIP.⁹³ Despite lack of evidence, exercise and physiotherapy should be part of the treatment regimen for patients with DIP.

Prognosis

Symptoms of DIP should resolve by six months after discontinuation of the causative medication.⁵³ However, one case series found persistent parkinsonism nine months after discontinuation of medication in patients with a normal DAT scan, suggesting a longer period of surveillance of up to one year is required.⁹⁴ One study found that the majority of patients who had been receiving treatment for DIP did not experience relapse of symptoms once the anti-parkinsonism treatment was discontinued after 3 months. The minority of patients who required resumption of treatment experienced return of symptoms within two to four weeks.⁹⁵ Another study found a low risk (8%) of relapse among patients with DIP treated with anticholinergics.⁹⁶

Discontinuation of the medication may lead to no change in symptoms. In one study, 16 weeks after the discontinuation of phenothiazine antipsychotics, no change in parkinsonism was observed; of note, 28% of patients in this study experienced relapse in psychiatric symptoms.⁹⁷

In a large retrospective study, patients receiving either typical or atypical antipsychotics were three times as likely as unexposed subjects to develop either PD or atypical parkinsonism in the period of 18 months after initiation of drug use to two years after cessation of treatment. This risk decreased by about one-third after two years post-cessation of treatment.¹⁹ Another prospective cohort study found an increased risk of developing incident PD among patients exposed to neuroleptics of 3.16 for benzamides and 2.59 for phenothiazines.⁹⁸ In one case series, two young patients (both age 35) treated with neuroleptics developed parkinsonism, which persisted and progressed after discontinuation of the medication; additionally, their symptoms were responsive to levodopa. The authors propose a potential mechanism of unmasking of idiopathic PD by neuroleptic use: neuroleptics are cytotoxic,^99,100 leading to creation of free radicals, which leads to membrane and cell damage;¹⁰¹ concurrently, neuroleptics lead to increased firing, and increased dopamine synthesis and release in nigrostriatal dopaminergic neurons.^102,103 This effect would be enhanced in a vulnerable population of neurons, such as those in patients with underlying PD, and may explain increased susceptibility to DIP.¹⁰⁴

Transcranial sonography has been found helpful in distinguishing between DIP and suspected neurodegenerative parkinsonism; patients with suspected degenerative parkinsonism were found to have significantly higher substantia nigra echogenicity. Additionally, normal echogenicity was associated with complete recovery after drug withdrawal.¹⁰⁵

One prospective study found shortest onset of DIP among patients treated with typical neuroleptics (1.2 ± 0.7 months), while those treated with calcium channel blockers (CCB), flunarizine or cinnarizine, had the longest time to onset (6 ± 1.4 months). The duration of parkinsonism was also longest in the CCB group (3.2 years).¹⁰⁶

A history of prior extrapyramidal syndromes, in addition to age, sex, and medication dose may be predictive of future extrapyramidal syndromes.¹⁰⁷ One study found that pre-treatment levels of dopamine and its metabolites correlated with development of akinesia, rigidity, and extrapyramidal symptoms; patients who excreted less free dopamine were found to develop more extrapyramidal symptoms.¹⁰⁸

Our Approach

Our clinical approach to a patient with parkinsonism and potential DIP begins with a focus on the history. This includes a detailed medication exposure history and the duration of exposure (do not forget to ask about metoclopramide), eliciting risk factors for DIP or PD (eg, family history of PD), and considering the presence or absence of non-motor symptoms suggestive for PD (eg, RBD, constipation, and anosmia). Remember, that the physical exam, including the MDS-UPDRS part 3, may not help distinguish DIP vs PD, DIP can manifest with asymmetric or tremulous presentations, and comorbid tardive syndrome is a clue supporting DIP.

We are wary of parkinsonism or tremor (and other movement disorders) that develop after the use of non-classic culprit medications—including those implicated from the realm of case reports. Neuroleptics and VMAT inhibitors inducing parkinsonism, however, are well established and make pathophysiologic sense. Cases where symptoms develop acutely and persistent after short-term exposure of a medication are also suspect—functional movement disorders should be considered in these cases.

Stopping the potentially offending agent can be diagnostic and therapeutic. This should be done in conjunction and in conversation with psychiatry. Quetiapine or clozapine are the ideal alternatives to other neuroleptics in these scenarios. Both DIP and PD patients can benefit motorically from discontinuation of dopamine blocking agents. Then, watch and wait. Short- and long-term follow-up of the offending agent will help clarify the diagnosis. Without patience and time, DIP and PD unmasked by neuroleptics can be indistinguishable, and a DAT scan (or perhaps sympathetic cardiac scintigraphy) can quickly differentiate. DAT scan requires holding certain mood regulating medications for variable lengths of time (eg 45 days for fluoxetine), which is not always ideal or possible. A levodopa trial can be considered to aid in diagnosis, but one must use caution in patients with psychosis.

Amantadine (including its extended-release formulations) or anticholinergic agents (trihexyphenidyl, benztropine) can be considered for bothersome motor symptoms in DIP. If all goes well with the discontinuation of culprit agents, then motor treatment may not be necessary long term. Physical, occupational, and speech therapy can be utilized. Botulinum toxin injection can be used for bothersome sialorrhea. Anticholinergics being prescribed for motor symptoms can serve a dual purpose in also mitigating sialorrhea, employing their antimuscarinic and drying side effect. Similarly, amantadine can help with concurrent oro-buccal-lingual dyskinesia.

Drug-induced parkinsonism can be comorbid with other neurologic disorders. Parkinsonism can develop in a patient with Huntington’s disease using a VMAT inhibitor for chorea. Or, in an adolescent with Tourette syndrome using risperidone (or any one of many other neuroleptics) for tics. Remember to monitor for the development of parkinsonism for such patients after initiating neuroleptics for any indication.

Future Perspective

Clinicians and patients would benefit from the identification of cost-effective biomarkers to aid in the diagnosis of DIP and its differentiation from PD. Early and accurate diagnosis will facilitate discussions about prognosis and guide treatment. Skin biopsies for phosphorylated synuclein have yielded positive results in diagnosing synucleinopathies, with sensitivity and specificity >90% in differentiating idiopathic PD from controls.¹⁰⁹ Abnormal biopsies results could help rule out DIP and would portend a worse prognosis and less expected improved with culprit drug discontinuation. DAT scan offers high sensitivity (98%), but lower specificity (67%) in early PD.¹¹⁰ Skin biopsy may be more cost-effective than DAT scan, and the biopsy specimen can be shipped to central processing labs, whereas patients need to travel to the DAT scan (which may not be locally available).

The Movement Disorders Society has published diagnostic criteria, intended to be utilized for research purposes, for prodromal PD.¹¹¹ As a patient’s risk factors accumulate, including non-motor symptoms (particularly RBD), environmental exposures, family history, personal genetic information, and imaging biomarkers (DAT scan), the likelihood of having prodromal PD increases. Perhaps, this scoring system (or a similar one) could be applied to patients with suspected DIP to help differentiate neuroleptic induced unmasked PD, in a previously prodromal patient, from true DIP. Future studies could focus on identifying additional clinical diagnostic biomarkers discriminating PD from DIP.

There are no recent studies regarding therapeutics for DIP. Can we do better than anticholinergics or amantadine? And what should we do for patients with normal DAT scans and who do not improve after suspected culprit drug discontinuation? Is there a “tardive” parkinsonism with expected permanent deficits, similar to the oro-buccal-lingual tardive dyskinesia that persists after medication discontinuation? Better understanding of the pathophysiology that differentiates these syndromes can lead to more targeted therapy and perhaps prevention of DIP.

Conclusions

Drug-induced parkinsonism is common with prolonged neuroleptic exposure and can mimic idiopathic PD. Expeditious recognition of this syndrome can lead to timely treatment (discontinuation of the culprit medication) and thus minimize morbidity. If the patient is not improving in the months following drug discontinuation, then one must consider a case of unmasked idiopathic PD—DAT scan, levodopa trial, and perhaps skin biopsy for synuclein can arbitrate.