Secondary and Primary Dystonia

Pathophysiological Differences

Maja Kojovic; Isabel Pareés; Panagiotis Kassavetis; Francisco J. Palomar; Pablo Mir; James T. Teo; Carla Cordivari; John C. Rothwell; Kailash P. Bhatia; Mark J. Edwards


Brain. 2013;136(7):2038-2049. 

In This Article


The main findings of the present study are (i) the response to PAS in patients with secondary dystonia is no different to that in healthy participants, in contrast to the enhanced response in patients with primary dystonia; (ii) patients with secondary dystonia have reduced SICI, on the side of the lesion only; and (iii) eye blink classical conditioning is worse in patients with primary than in those with secondary dystonia.

Differences in Paired Associative Stimulation Induced Plasticity Between Secondary and Primary Dystonia

The enhanced response to PAS that we found in our patients with primary dystonia is in line with several previous studies using a variety of plasticity-testing protocols (Quartarone et al., 2003; Edwards et al., 2006; Weise et al., 2006). It was, however, surprising to find that the response to PAS was normal in secondary dystonia. This is unlikely to be due to differences in baseline corticospinal excitability, as the input-output curves and motor thresholds were the same in all three groups that we studied. Nor is it likely to be a result of the longer duration and more severe dystonic symptoms in the patients with secondary dystonia. Although the present study only examined cases of primary segmental dystonia, previous investigations from this laboratory have found enhanced responses to experimental plasticity protocols even in patients with primary generalized dystonia, whose symptoms began in childhood (Edwards et al., 2006) and were so severe as to require bilateral pallidal deep brain stimulation (Ruge et al., 2011). In addition, there was no correlation between disease duration and the response to PAS in our patients with secondary dystonia.

We have reported previously (Kojovic et al., 2011) that botulinum toxin treatment can transiently reduce the response to PAS in patients with primary dystonia, which then returns to the level present before botulinum toxin injection after a few months. Since all the patients in the present study were investigated at least 15 weeks after their last injection, this acute effect of botulinum toxin is unlikely to have influenced the present results. Nevertheless, it is difficult to speculate on whether there might have been possible chronic effects of botulinum toxin on motor cortex plasticity, as this has not been previously investigated. Several of the patients with secondary dystonia had been treated for many years and it is possible that this could have permanently reduced their PAS response and skewed the group data even though there was no difference in mean duration of treatment in the primary and secondary cases. This seems unlikely to have been the case as there was no correlation between duration of botulinum toxin treatment and the response to PAS protocol.

In the absence of other explanations, we suggest that enhanced motor cortex plasticity is an inherent, genetically determined trait (endophenotype) specific for primary dystonia that predisposes some individuals to develop dystonia. As suggested by Quartarone et al. (2006b) this may result in an excessive tendency to form associations between sensory input and motor output, leading to dystonia, particularly under circumstances involving frequent repetition of specific movements. The fact that sensorimotor plasticity is normal not only in secondary but also in psychogenic dystonia (Quartarone et al., 2009) further confirms that abnormally enhanced plasticity is an endophenotypic trait specific to primary dystonia.

Secondary dystonia is believed to be related to functional changes in sensorimotor circuits after brain injury (Burke et al., 1980), but the exact mechanism underlying the changes and the anatomical regions in which they occur are not well understood. Diffusion tensor imaging and functional MRI studies in patients with subcortical strokes suggest that no significant reorganization occurs in the ipsilesional primary motor cortex per se, but rather within the white matter tracts (Fries et al., 1993; Jang, 2007) with a contribution of the corticospinal tract from the unaffected hemisphere (Jankowska and Edgley, 2006). Thus, it may be that the principal pathological processes spare the function of ipsilesional primary motor cortex. In a PET activation study Ceballos-Baumann et al. (1995) showed that the pattern of primary motor cortex activity differs between patients with acquired hemidystonia and idiopathic torsion dystonia. Similarly, a combined functional MRI and diffusion tensor imaging study on a patient with hemidystonia caused by a penetrating injury of caudate and lentiform nucleus showed that there was no significant functional reorganization in the primary motor cortex after injury (Werring et al., 1998). This would be consistent with the normal response to PAS in our patients.

Eye Blink Classical Conditioning and its Possible Relation to Paired Associative Stimulation Response in Dystonia

Eye blink classical conditioning, as used in human studies, is a form of predictive learning that lesion studies have shown to depend on the integrity of the olivo-cerebellar circuit (Gerwig et al., 2007). Indeed, in healthy individuals, continuous theta burst stimulation over cerebellum, which is thought to interfere with function in cerebellar circuits, abolishes eye blink classical conditioning (Hoffland et al., 2012). Previously we had found that eye blink classical conditioning was markedly reduced—compared with healthy volunteers—in patients with primary focal hand and/or cervical dystonia and had speculated that this was further evidence in favour of a cerebellar involvement in primary dystonias (Teo et al., 2009). In the present study, patients with secondary dystonia showed preserved eye blink classical conditioning that did not differ from healthy control subjects. Eye blink classical conditioning decreases with age (Finkbiner and Woodruff-Pak, 1991; Bellebaum and Daum, 2004) and therefore the age difference between compared groups could have been a confounding factor. However, even though our patients with secondary dystonia were younger than both the healthy control subjects and patients with primary dystonia, their eye blink classical conditioning was similar to healthy control subjects and superior to eye blink classical conditioning in primary dystonia. Therefore, younger age is unlikely to be a reason for apparently normal eye blink classical conditioning in our patients with secondary dystonia. The implication of our findings is that the pathophysiology of secondary dystonia is more localized than that of primary dystonia.

Although eye blink classical conditioning and PAS are usually thought to test quite different circuits in different parts of the brain, there may be some connection between the two that could potentially link the present results in primary and secondary dystonias. Recent work has shown that the response to some PAS protocols is modulated by inputs from the cerebellum; thus a disordered cerebellum could potentially lead to abnormal PAS (Hamada et al., 2012, Popa et al., 2013). In healthy volunteers, the effect of a PAS25 protocol (that is, with an interval of 25 ms between median nerve and TMS pulse) is reduced or abolished by concurrent andoal direct current stimulation over the cerebellum or by preconditioning with excitatory intermittent theta burst stimulation (Hamada et al., 2012, Poppa et al., 2013); in contrast, preconditioning the cerebellum with continuous theta burst stimulation enhanced PAS (Popa et al., 2013). Thus, the effect of motor cortex PAS25 depends on the functional state of cerebellar output.

From the data outlined above, the combination of enhanced response to PAS25 and decreased eye blink classical conditioning in primary dystonia is similar to what occurs with cerebellar continuous theta burst stimulation in healthy volunteers: eye blink classical conditioning is reduced and PAS25 plasticity increased. The conclusion is that a cerebellar disorder in patients with primary dystonia may be related to their abnormal response to PAS. However, this is unlikely to be the whole story. The response to PAS21.5 (that is, PAS with a 21.5 ms interval between stimuli) is unaffected by cerebellar direct current stimulation (Hamada et al., 2012) in healthy participants yet it is still enhanced in primary dystonias (Weise et al., 2006), suggesting that there is an intrinsic disorder of cortical plasticity in addition to any secondary influence from a disordered cerebellum.

The Role of Reduced Intracortical Inhibition in Dystonia

The final finding of our study is that patients with secondary dystonia had reduced SICI on the affected side. This is in line with a recent finding of reduced SICI in patients with dystonia caused by lentiform nucleus lesions (Trompetto et al., 2012). The present data showed only a non-significant trend toward reduced SICI in patients with primary dystonia, in contrast with reduced SICI reported in some previous studies (Ridding et al., 1995a; Edwards et al., 2003; Quartarone et al., 2003). However, others have found SICI to be normal in primary dystonia (Rona et al., 1998; Stinear and Byblow, 2004; Brighina et al., 2009). This inconsistency probably reflects the large between-subject variability of intracortical inhibition that is present even in healthy subjects (Wassermann, 2002), as well as methodological differences between studies (conditioning stimulus intensity, unconditioned MEP amplitude, interstimulus intervals, rest versus active condition of the target muscle). The pathophysiological significance of reduced intracortical inhibition in dystonia remains obscure (Berardelli et al., 2008) and there is still no uniform hypothesis to account for reduced SICI in all forms of dystonia. Reduced SICI is not specific for dystonia and is found in other basal ganglia diseases, including Parkinson's disease and Tourettes syndrome (Ridding et al., 1995b; Ziemann et al., 1997). Therefore, a loss of intracortical inhibition may be regarded as a non-specific maladaptive change within the motor cortex, caused by chronic disorganized basal ganglia output. Our finding would fit with this hypothesis, as SICI was only abnormal on the clinically affected side of our patients with secondary dystonia. Alternatively, reduced SICI may arise as a consequence of maintaining an abnormal dystonic posture that could have triggered cortical reorganization through aberrant afferent input (Espay et al., 2006). This hypothesis is nevertheless insufficient to explain the reduced SICI in non-affected body parts in primary focal dystonia (Sommer et al., 2002), or in non-manifesting DYT1 mutation carriers (Edwards et al., 2003). Considering the pathophysiological importance of reduced SICI in different types of dystonia, the conclusion is that reduced intracortical inhibition must co-exist with other abnormalities, to cause clinical expression of dystonia: for example, increased plasticity and/or abnormal cerebellar function in primary dystonia (Quartarone et al., 2003), psychogenic factors in non-organic dystonia (Espay et al., 2006; Quartarone et al., 2009) or injury to basal ganglia and its connections in secondary dystonias (Trompetto et al., 2012).

There was no significant difference in cortical silent period duration between groups, although there was a tendency toward a shortening of the cortical silent period on the affected side in both secondary and primary dystonia, compared with control subjects. The literature on cortical silent period in dystonia has been less consistent than for SICI, with studies reporting normal cortical silent period (Stinear and Byblow, 2005) or reduced cortical silent period (Chen et al., 1997) or an abnormality was restricted to a specific task (Tinazzi et al., 2005). SICI and the cortical silent period are thought to depend on GABA-A and GABA-B cortical interneurons, respectively and therefore could be differentially affected by disease (Werhahn et al., 1999; Di Lazzaro et al., 2006; Hallett, 2011). This might explain the abnormal SICI and normal cortical silent period in our patients with secondary dystonia. Trompetto et al. (2012) suggested that the cortical silent period is reduced in secondary dystonia when the lesion is restricted to striatum, while it might be normal if the lesion involves pallidum or thalamus. We did not find the duration of cortical silent period to be related to the anatomical site of the lesion.

Limitations of the Study

We acknowledge the limitations of our study. Our sample of patients with secondary dystonia is heterogeneous regarding aetiology and anatomical site of the lesion. Although it is possible that different lesions could have different functional effects on motor cortex plasticity, we believe this is unlikely given the similar response to PAS among all patients with secondary dystonia, including the lack of spread into the non-target adductor digiti minimi muscle (Fig. 4). Another limitation of our data is the long interval between the brain injury and TMS study. With the present design, we cannot exclude the possibility that motor cortex plasticity was affected at the time of emergence of dystonia, and then over time has reverted to normal. This issue could be addressed in a prospective study that would need to include a large number of patients, given that only a small proportion of patients with subcortical lesions will go on to develop dystonia.