Demystifying Spasticity: Reply to Dietz.

REPLY: I was glad to read Dr. Volker Dietz’s recent letter to the editor (Dietz 2007) comparing his work on spasticity in humans to my studies of spasticity in the sacral spinal rat (Bennett et al. 2004). His comments highlight many of the confusions that have shrouded the study of spasticity and give me a chance to clear the air on this topic. The first confusion relates to the complexity of spasticity and the lack of consensus on which aspects of spasticity are most relevant. The spastic syndrome in humans (or what Dietz calls the “spastic movement disorder”) is a complex collection of clinical conditions, including excess muscle tone (hypertonus), changes in muscle properties (like contractures), excess reflex activity (hyperreflexia), muscle oscillations (clonus), and massive uncontrolled motoneuron firing and contractions (spasms; Bennett et al. 2004; Dietz and Sinkjaer 2007; Kuhn and Macht 1948). In my laboratory, we mostly study spasms triggered by brief cutaneous stimulation and we have shown these to result from excess motoneuron excitability (Bennett et al. 2004; Li et al. 2004a). Dietz has spent a lot of time studying muscle properties (Dietz et al. 1986) and emphasizes their importance in his letter to the editor (Dietz 2007). From this perspective, he criticizes research that does not focus on muscle properties, suggesting that it is not relevant to the study of spasticity. This seems a bit odd because upward of 80% of spinal cord injured humans suffer from spasms caused by excess motoneuron activity (Maynard et al. 1990), and thus understanding spasms is important. The second confusion has arisen because spasticity research has for decades focused exclusively on correlating spastic behavior with short-lasting reflexes (monoand polysynaptic) that can be conveniently measured in the laboratory in humans, as though these reflexes were the only forms of neuronal activity possible from the spinal cord. I think that research should instead focus on measuring neuronal activity during actual spastic episodes, such as during spasms, even though this means studying complex long-lasting reflexes, rather than simple short-lasting reflexes (Bennett et al. 2004). Dietz rightly points out that after years of studying short-latency monosynaptic reflexes and other short-lasting polysynaptic reflexes, it is clear that they do not “contribute greatly to the spastic movement disorder” (Dietz 2007; Dietz and Sinkjaer 2007). However, he then seems to fall into the classic trap of assuming that these reflexes are the only forms of neuronal activity possible from the spinal cord, and thus makes the overgeneralization that human spasticity is not “primarily caused by an increased neuronal activity” (Dietz 2007). Because muscle properties change somewhat in spastic patients (Dietz and Sinkjaer 2007), Dietz then concludes, without any further evidence, that properties of muscles must instead be primarily responsible for spasticity (Dietz 2007; Dietz and Sinkjaer 2007). Importantly, Dietz has not considered longer-lasting reflexes more closely related to the pathological motor behavior. For example, in spinal cord injured patients cutaneous stimulation evokes a long-lasting reflex that underlies muscle spasms. We have shown in both humans and rats with spinal cord injury (Gorassini et al. 2004; Li et al. 2004a) that such long-lasting reflexes are produced by sustained motoneuron discharges maintained by currents intrinsic to the motoneuron (persistent sodium and calcium currents, PICs). After a brief synaptic input (cutaneous), these currents remain activated and produce uncontrolled firing that cannot be turned off because of the loss of descending inhibitory control following spinal cord injury. Thus, in patients that suffer from muscle spasms, neuronal activity in the form of aberrant motoneuron firing plays a major role in their clinical spasticity condition. The question of whether changes in properties of muscles or changes in neuronal activity are more important depends on the particular patient and the muscle studied. Following spinal cord injury some muscles suffer from atrophy, contractures, and tendon shortening, and I would agree with Dietz that in these muscles the intrinsic muscle property changes dominate the clinical picture (Dietz 2007). However, in many muscles (both leg flexor and extensor muscles) there can be significant motoneuronal activity following spinal cord injury, in the form of clonus and muscle spasms (Gorassini et al. 2004; Kuhn and Macht 1948; Thomas and Ross 1997). It is also very telling that BOTOX injections can rapidly reduce muscle tone (Richardson et al. 2000), which can happen only if the muscles are centrally activated, rather than under peripheral contractures. Thus, it is somewhat of an overgeneralization for Dietz to claim that changes in muscle properties, rather than neuronal activity, always determine spasticity. It is sometimes even quite the opposite. Indeed, we have recently shown that neuronal activity in the form of muscle spasms actually prevents muscle atrophy and changes in muscle properties (Harris et al. 2006, 2007). So the nonspasming muscles atrophy and the spasming muscles do not atrophy, which raises the question of how vigorously we should treat spasticity with drugs like baclofen (Li et al. 2004c). The third confusion in the study of spasticity has arisen over the years because researchers have mixed together studies of patients with stroke and spinal cord injury. The clinical spasticity syndrome following stroke is characterized in large part by excess muscle tone (especially in arm flexors and leg extensors) and a relative absence of spasms, whereas the clinical spasticity syndrome following spinal cord injury is characterized by less muscle tone and, instead, flexor and extensor muscle spasms triggered by cutaneous stimulation are the hallmark (Kuhn and Macht 1948). Dietz’s early research largely focused on stroke (Dietz et al. 1986), and from this stroke work he came to the somewhat extreme conclusion that changes in muscle properties (contractures) dominate spasticity following stroke (see Dietz and Sinkjaer 2007). He then makes Address for reprint requests and other correspondence: D. J. Bennett, 513 HMRC, University of Alberta, Edmonton, AB, Canada T6G 2S2 (E-mail: [email protected]). J Neurophysiol 99: 1041–1043, 2008. doi:10.1152/jn.01279.2007.