In this chapter we will review the different types of drugs that may cause cerebellar ataxia as part of their side effect profile. Ataxia as a drug-induced movement disorder occurs most often in the setting of acute intoxication with phenytoin (PHT) and carbamazepine (CBZ) as the two drugs most commonly implicated . The frequent occurrence of ataxia due to drug exposure stems from the fact that the cerebellum is particularly susceptible to intoxication. Among all cerebellar cells, Purkinje neurons are especially susceptible to this form of injury. Cerebellar circuits are also a main target of drug exposure. However, figures on the prevalence and incidence of cerebellar involvement secondary to drug intoxication are still lacking for most drugs .
As is often the case with exposure to neuroleptics and the development of parkinsonism or tardive dyskinesia, the occurrence of ataxia in the setting of an epileptic patient treated either with PHT or CBZ is an expected event. However, it may not be the case with many other drugs that have been only occasionally reported as causing ataxia, and for that reason the treating physician may be unaware. Therefore, when faced with an unexpected event, such as the development of ataxia in a patient undergoing any pharmacological treatment, it may be useful to bear in mind the existence of certain instruments that may help in the ascertainment of the drug-induced nature of the phenomenon, such as the Adverse Drug Reaction (ADR) Probability Scale (Naranjo’s Scale).
Definition of ataxia
Ataxia is both a neurologic symptom and a sign seen only in association with movement or with an attempt to maintain position against a deflecting force such as gravity. Ataxic movements are poorly organized and usually are due to dysfunction of the cerebellum or its numerous connections with other brain regions . Limb and gait ataxia are quite common, but incoordination can also affect speech articulation, swallowing, vision, and truncal movements. The most common phenomenology in ataxia involves limb incoordination, unsteady and broad based gait, tremor, nystagmus, and scanning speech. Like all movement disorders, ataxia can be described in terms of its location, amplitude, frequency, modifying factors, course over time, and associated neurologic and nonneurologic symptoms. The differential diagnosis of ataxia includes a large number of neurologic disorders, so efficient evaluation requires a careful history and a thorough physical and neurologic examination to place ataxia in its appropriate context.
An appreciation of the broad differential diagnosis of movement disorders is also necessary when attempting to implicate a specific drug as causative. Classification and differential diagnosis of ataxic syndromes have an intrinsic complexity owing to the variability in phenotypic presentations and in etiologies, which include trauma, toxic and metabolic causes, neoplasms, immune mechanisms, drugs, and genetic diseases. Pure cerebellar symptoms are rarely observed, while the clinical picture of both genetic and sporadic ataxia syndromes is sometimes complicated by the presence of extracerebellar neurological or multisystem extraneural pathology . Even when the suspected cause is attributable to drug exposure there should be a comprehensive evaluation both clinical and through the use of ancillary diagnostic methods, to rule out other etiologies, as well as a causality assessment. Drug-induced ataxia should always be considered in the differential diagnosis in sporadic cases of this disorder.
The drugs that have been most frequently held responsible for the development of drug-induced ataxia are antiepileptic drugs, and there is ample experience and a comprehensive body of literature, particularly related to PHT and CBZ. However, there are a growing number of published reports describing similar side effects with other AEDs besides the two most often cited in the literature. Similarly, other classes of drugs have been held responsible for the occurrence of drug-induced ataxia through anecdotal case reports or small series of patients. In Table 14.1 the case reports or reviews of drugs that produce ataxia published from 1990 through 2014 are cited. For PHT, due to the large number of published cases, a single review is cited.
|Drug class||Drug (specific)||Evidence||Reference|
|Antiepileptic||Phenytoin||Review||(Awada et al., 2001) |
|Retrospective review (n: 519)||(Zaccara et al., 2013) |
|Other antiepileptics||Phenobarbital||Clinical review||(Factor et al., 2008) |
|Gabapentin||Case report||(Steinhoff et al., 1997) |
|Vigabatrin||Case report||(Sander et al., 1991) |
|Lamotrigine||Case report||(Fife et al., 2006) |
|Levetiracetam||Case report||(Chayasirisobhon et al., 2010) |
|Zonisamide||Systematic review||(Carmichael et al., 2013) |
|Rufinamide||Randomized controlled trial||(Brodie et al., 2009) |
|Lacosamide||Randomized controlled trial||(Ben-Menachem et al., 2007) |
(Pirzada et al., 2000) 
(Noguerón et al., 1997) 
(Barbieux et al., 1996) 
|Cytarabine (Ara-C)||Case report|
(Friedman et al., 2001) 
(Yeshurun et al., 2001) 
(Hasle et al., 1990) 
(Gounaris et al., 2010) 
(Lam et al., 2008) 
(Renouf et al., 2006) 
Case report (n:2)
(Masterson et al., 2008) 
(Ferhanoglu et al., 2003) 
|Platinum (Cisplatin, Oxaliplatin, Paclitaxel)||Review (phase I)||(Kobrinsky et al., 2013) |
|Epothilone D||Review (phase II)||(Beer et al., 2007) |
Case report (n:5)
(Chaubey et al., 2013) 
(Willis et al., 2009) 
(Hindle et al., 2008) 
(Krauser et al., 2005) 
(Garreto et al., 1994) 
(Arnaud et al., 1992) 
|Propafenone||Case report (n:3)||(Odeh et al., 2000) |
|Antibiotic||Metronidazole||Systematic review||(Kuriyama et al., 2011) |
|Isoniazid||Case report (pediatric)||(Lewin et al., 1993) |
Case report (n:3)
Case report (n:6)
(Niethammer et al., 2007) 
(Manto et al., 1996) 
|Calcineurin (Tacrolimus, cyclosporine)|
Case report (pediatric)
(Yamaguchi et al., 2007) 
(Kaczmarek et al., 2007) 
(Kaleyias et al., 2006) 
(Masannat et al., 2013) 
(Gordon et al.,1995) 
(Playford et al., 1990) 
|Statins||Case report||(Teive et al., 2012) |
Among the common movement disorders associated with PHT, cerebellar ataxia is the most frequently reported, followed by asterixis and myoclonus, although a variety of other movement disorders, including chorea, orofacial dyskinesias, tremor, tics, and dystonia, can occur. The classical signs of PHT intoxication are coarse nystagmus, limb and gait ataxia, and dysarthria, which are all signs of cerebellar dysfunction. In a review of 94 cases of PHT intoxication in a general hospital, ataxia was observed in 59/94 patients (63%), and of these, 18 patients had fallen, and nine of these had suffered injuries from their falls severe enough to require medical care. In the reported cases serum PHT levels ranged from 21.4–90 micrograms/ml, with a mean level 44.4 +/–12.5 micrograms/ml . While the outcome in these patients is usually good, with the resolution of ataxia after the drug is metabolized to nontoxic levels, some patients develop chronic complications, such as a slight cognitive deficit or a peripheral neuropathy.
Although acute intoxication is quite frequent, cerebellar signs may develop after several years of chronic treatment. Complete recovery from cerebellar deficits is the usual outcome in these patients once the offending drug has been discontinued or tapered. However, irreversible lesions may develop. Usually, patients who develop irreversible cerebellar deficits have been exposed to higher doses for longer periods of time [7, 8].
Computed tomography (CT) and brain magnetic resonance imaging (MRI) show cerebellar atrophy of variable degree in patients developing irreversible cerebellar signs following chronic treatment with PHT , with the vermian region predominantly affected . In a case-controlled study, cerebellar atrophy was seen in patients with epilepsy exposed to PHT in the absence of seizures and preexisting cerebellar damage . Cerebellar atrophy due to acute PHT intoxication is unusual, but a few cases have been reported , although it is controversial whether this effect is due to PHT itself, the underlying seizure disorder, or other causes.
Hospital admission is indicated in symptomatic cases until a declining serum PHT level is observed and ataxia resolves .
Carbamazepine (CBZ) is also commonly implicated as a cause of ataxia. Typically, patients complain of dizziness and exhibit gaze-evoked nystagmus, action tremor, and ataxia of stance/gait. CBZ induces dose-dependent ataxic effects. In a study by Seymour , of 33 cases of intentional or accidental CBZ overdose, ataxia, nystagmus, or ophthalmoplegia were seen in 48% of adults with a mean overdose of 12 grams (range 1.6–45). It appears that the presence of cerebellar atrophy as seen on MRI predisposes CBZ-treated patients to ataxia at significantly lower serum levels compared to patients without cerebellar atrophy .
Oxcarbazepine (OXC) is the 10-keto analog of carbamazepine which blocks high-frequency voltage–dependent repetitive firing of sodium channels. Indirect comparisons between these AEDs, taking into account dose effect, showed that OXC may be associated with more frequent neurological adverse events than lacosamide (LCM) and eslicarbazepine acetate (ESL). Abnormal coordination/ataxia and diplopia were significantly more frequently observed in patients treated with OXC compared to patients treated with LCM and ESL. The study reveals that the number needed to harm (NNH) associated with OXC 1200 mg (95% IC) was 5 for ataxia/abnormal coordination .
Phenobarbital (PB) may induce transient ataxic signs at toxic levels, both acutely and after chronic exposure. The most common deficits are gaze-evoked nystagmus, kinetic tremor, and ataxia of stance and gait. Most patients will exhibit concomitant drowsiness. It is estimated that about 5% of epileptic patients treated with barbiturates show cerebellar deficits. Nevertheless, evidence of cellular toxicity and neuronal loss in adults is lacking. Experimental studies underline the vulnerability of the cerebellum during growth, but clinical implications are unclear .
The mechanism of action of gabapentin (GBT) is through the inhibition of calcium influx and the subsequent release of excitatory neurotransmitters. It is a well-tolerated drug when used for monotherapy or associated with other AEDs. However, movement disorders have been reported previously as rare side effects in individual patients. Steinhoff  reported two patients, who developed isolated severe ataxia on low-doses of GBT which resolved abruptly after discontinuation of the drug. One of the patients was receiving GBP as monotherapy and in the other patient, GBP was taken with CBZ. When GBP was stopped for one week, the ataxia disappeared.
Vigabatrin (VGB) is a structural analog of gamma aminobutyric acid (GABA) which irreversibly inhibits the enzyme GABA transaminase. It has been used in refractory epilepsy and infantile spasms (IS) for more than 10 years. Use was restricted due to concern about its safety profile. VGB can induce mild ataxic posture in adults with poorly controlled epilepsy. Gait ataxia appears to be dose-related . There is no data on ataxia as an ADR of VGB in children with epilepsy.
Lamotrigine (LTG) acts by blocking the voltage-dependent sodium channels and thus blocks the release of glutamate through stabilization of the presynaptic membrane. The indications are partial and generalized seizures and worsening severe myoclonic epilepsy. Lamotrigine causes less dysequilibrium than does CBZ in older people on monotherapy for epilepsy .
Levetiracetam (LVT) selectively inhibits high-voltage–activated calcium channels and reduces calcium release from intraneuronal stores. It also binds to a specific target in the brain, the synaptic vesicle protein 2 A (SV2A), an integral membrane glycoprotein, which is involved in the control of vesicle fusion and exocytosis. There is no substantial evidence that LVT can induce ataxia. However, in a single case report Chayasirisobhon et al.  comment on a patient with epilepsy who took an overdose of 63 grams of LVT, resulting in mild ADRs, including mild blurred vision and mild ataxia that rapidly subsided one day after drug discontinuation. However, a metaanalysis evaluating the risk of balance disorders in patients under treatment with second generation AEDs (LVT, GBP, LMT, OXC, pregabalin, tiagabine, topiramate, or zonisamide) at standard therapeutic doses revealed that LVT did not increase imbalance risk at any dose, while the other drugs included in this metaanalysis were found to increase the risk of imbalance in a dose-response fashion .
Zonisamide (ZNS) is a sulphonamide derivative, a broad spectrum AED that acts through multiple actions: facilitation of dopaminergic and serotoninergic neurotransmission through the blockade of T-type calcium channels, prolongation of sodium channel inactivation, and as a weak inhibitor of carbonic anhydrase. It is approved for use as an adjunctive therapy in adults with partial-onset seizures, IS, mixed seizure types of Lennox-Gastaut syndrome (LGS), myoclonic, and generalized tonic clonic seizure. Neurological side effects include ataxia, in addition to somnolence, agitation, and anorexia .
Rufinamide (RUF) is an AED that was approved in the European Union in January 2007 as an orphan drug for the adjunctive treatment of seizures associated with LGS. Its mechanism of action is not completely understood but it is believed to work by prolonging the inactive state of sodium channels and therefore limiting excessive firing of sodium-dependent action potentials. The common ADRs with RUF are consistent with the typical profile of AEDs, mainly involving the CNS and gastrointestinal systems. In a placebo-controlled trial  of RUF (n: 156) at therapeutic levels (400 mg twice daily, titrated up to 1600 mg twice daily) versus placebo (n: 157) for the adjunctive treatment of partial seizures in adults and adolescents, the incidence of ataxia versus placebo was 13.5 versus 0.6; they occurred primarily during the titration phase.
Lacosamide (LCS) is a functionalized amino acid that selectively enhances slow inactivation of voltage-gated sodium channels, increasing the proportion of sodium channels unavailable for depolarization. In a double blind, placebo-controlled, randomized trial, the most common ADRs which occurred in at least 10% of patients in any randomized treatment group were in the CNS and gastrointestinal system (dizziness, headache, nausea, fatigue, ataxia, abnormal vision, vomiting, diplopia, somnolence, and nystagmus). In their 2007 review Ben-Menachem et al. , reported that of 421 patients exposed to LCS, ataxia as an ADR was observed in 42 patients, corresponding to13% of the total population under analysis. Ataxia appeared to be related to LCS dose (600 mg/day), although data on serum levels of the drug was unavailable.