‘Don’t delay, start today’: delaying levodopa does not delay motor complications

This scientific commentary refers to ‘The modern pre-levodopa era of Parkinson’s disease: insights into motor complications from sub-Saharan Africa’, by Cilia et al. (doi:10.1093/brain/awu195).

Preventing the development of motor complications is one of the principal concerns when treating patients with Parkinson’s disease, and ‘levodopa-sparing’ approaches have been commonly touted as the best method of achieving this. Early use of dopamine agonists, rather than levodopa, was the preferred management strategy in the 1990–2000s, and several large randomized controlled trials reported delayed development of motor complications with this approach. However, follow-up studies revealed that once patients were started on levodopa, they developed motor complications of the same severity and at the same rate irrespective of whether levodopa had been initiated earlier or later. Indeed, after 10–14 years of treatment, patient profiles were essentially identical regardless of how they began dopaminergic therapy (Katzenschlager et al., 2008). These results, combined with increasing awareness of the major side-effects of dopamine agonists—particularly sleepiness and impulse-control disorders—have led many physicians to switch to earlier use of levodopa as monotherapy, particularly in patients over 60 years of age in whom the risk of dyskinesia is lower.

In this issue of Brain, Cilia and colleagues present data that help endorse this strategy (Cilia et al., 2014). This cross-sectional and 4-year longitudinal study recruited patients with Parkinson’s disease from Ghana (n = 91) and Italy (n = 2282). The authors took advantage of the opportunity to study patients from sub-Saharan Africa who typically experience longer delays in receiving levodopa therapy than patients in first world countries. Demographics and disease-related factors influencing the development of motor fluctuations and levodopa-induced dyskinesia (LID) were evaluated for both groups. Although accurate recording of LID can be difficult, the authors attempted to determine dates of onset as precisely as possible.

A subgroup of subjects from Italy (n = 50) who were levodopa-naive were matched to 59 subjects from Ghana who, for economic and medical reasons, had longer disease duration before starting levodopa (mean 4.2 years) than the therapy-, age- and sex-matched patients in Italy (mean 1.8 years). The subjects had comparable disease severity [as per Unified Parkinson’s Disease Rating Scale (UPDRS) part III motor scores OFF drug]. Disease duration at the time of developing motor fluctuations (mean 5–6 years) and LID (mean 6–7 years) was also similar in the two groups; thus the group from Ghana developed motor complications more quickly after starting levodopa. Logistic regression analysis for predictors of motor complications showed that higher levodopa dose, expressed as mg/kg, and longer disease duration at the time of starting levodopa, were the main risk factors. There was no association with duration of levodopa use.

The pathophysiology of levodopa-induced motor complications is likely due to a combination of disease-related factors, as well as the pharmacokinetics of levodopa itself. The relative contributions of these are hard to tease apart, as individuals with Parkinson’s disease are generally not left untreated for long periods, thus the effects of disease progression cannot be separated from those of chronic levodopa use. However, the fact that individuals with Parkinson’s disease develop motor complications over time suggests the involvement of progressive pathological changes, particularly degeneration of nigro-striatal dopamine terminals. Thus, early and more severe dyskinesia occurs in patients with more severe dopamine depletion, dyskinesia occurs earlier and more prominently on the side of the body most affected by parkinsonism, and subjects with severe secondary parkinsonism, e.g. MPTP-parkinsonism or multiple system atrophy type parkinsonism, have early onset of dyskinesia.

In the MPTP non-human primate model of Parkinson’s disease, at least 50% loss of dopamine is needed for normal doses of levodopa to induce dyskinesia. In the typical model where there is >90% dopamine depletion, dyskinesia can often be seen on first exposure to levodopa (Fox et al., 2010). However, Cilia and colleagues report that disease duration and clinical severity (as opposed to severity of striatal dopamine deficiency) do not necessarily equate as risk factors for the development of LID. This ‘disconnect’ is possibly due to the roles of various pre- and postsynaptic mechanisms in compensating for striatal dopamine deficiency. As the authors point out, historical data from the very early levodopa era (as described by Cotzias et al., 1969) are consistent with the conclusions of this ‘modern day’ study, i.e. that it is disease duration, not clinical severity or duration of levodopa, that seems to be most important. Further studies into the nature of these compensatory mechanisms may help identify potential therapeutic strategies.

However, the unusual pharmacokinetic properties of levodopa per se also contribute to the pathogenesis of motor complications, as the very short half-life of the drug leads to abnormal pulsatile stimulation of striatal dopamine receptors. Indeed, LID can occur without nigro-striatal dopamine loss. Normal primates can develop LID in response to chronic treatment with high-dose levodopa (Pearce et al., 2001), whereas patients with DOPA-responsive dystonia can develop chorea with low-dose levodopa typically upon treatment initiation.

Separating the effects of chronic levodopa from those of disease state is possible to some degree in the MPTP primate, in which disease (once stabilized) does not progress. Motor fluctuations accompanied by a reduced duration of anti-parkinsonian action (‘wearing-off’) also develops in MPTP primates, suggesting a mechanism that involves levodopa per se, rather than disease progression (Fox et al., 2010). We have also shown that when there is a stable severe disease state in MPTP primates, dyskinesia can develop on first exposure to levodopa, but the effect is dose-related: higher doses are more likely to induce LID than lower doses. Chronic treatment with a high dose of levodopa does not produce any further increase in LID; however, chronic treatment with lower doses causes a progressive increase in LID (Mestre et al., 2010). These studies emphasize that both disease severity and levodopa dose are likely more important than duration of levodopa for the development of motor complications.

Nutt et al. (2002, 2010) have also attempted to explain the development of motor complications by objectively evaluating the effects of long-term levodopa therapy in patients with early stage Parkinson’s disease over 4 years. Their results show that the anti-parkinsonian and dyskinesia-inducing actions of levodopa coincide, and that chronic levodopa is associated with a reduced latency to peak-response and a longer duration of dyskinesia, rather than with an increase in the peak response. The so-called ‘long-duration response’ to levodopa (the effect seen in early Parkinson’s disease where a delayed or missed dose of levodopa is not associated with re-emergence of symptoms) is what seems to be lost with progressive disease and chronic levodopa therapy. The chronic pulsatile stimulation of dopamine receptors in the striatum seems to switch the pharmacological response to levodopa from long- to short-duration. Augmenting the long-duration response, for example through the early use of longer-acting levodopa preparations, may provide an effective anti-parkinsonian action with reduced risk of motor complications; however, the pharmacology of levodopa remains a challenge for the development of such agents. Recent studies with novel delivery modalities have shown promise, although technical and logistic challenges remain (e.g. levodopa/carbidopa intestinal gel). The early use of deep brain stimulation of the subthalamic nucleus has also been proposed as a possible therapy (Schuepbach et al., 2013), which may indirectly induce a long-duration response. However, practical issues and risks associated with very early surgery limit this option at present.

The findings of Cilia and colleagues seem to suggest that the chronic use of levodopa, inducing ‘priming’, is not as important a factor in the development of motor complications as previously thought. The pathophysiology of LID is believed to involve chronic pulsatile dopamine receptor stimulation leading to altered dopamine D1 receptor activation and overactivity of the direct cortico-striato-pallido-thalamo-cortical pathway, with resultant reduced inhibition from pallidal-thalamic outputs. The pathogenesis of LID occurring on first or early exposure to levodopa, and its relation to disease factors, is less clear. Possible mechanisms may involve conversion of levodopa to dopamine in remaining serotonergic terminals with unregulated, abnormal release (Carta and Bezard, 2011). Where no levodopa has been administered previously, the striatal dopamine imbalance may result in a marked increase in dopamine D2 receptor sensitivity due to a loss of the normal endogenous dopamine-induced inhibition (i.e. a form of denervation supersensitivity). Exogenous dopamine from levodopa would thus bind to these overactive D2 receptors and reduce activity in the indirect striato-pallidal pathway, possibly to below normal functional levels, with subsequent reduced activation of the subthalamic nucleus and output regions of the basal ganglia, resulting in dyskinesia. The higher the dose of levodopa administered, the more likely this scenario.

There are other potential risk factors for LID that will require evaluation in future studies. These include genetic factors, such as haplotype variations in dopamine D1, D2 and D3 receptors and enzymes involved in levodopa metabolism. In particular, Cilia and colleagues highlight the single case of a patient from Ghana who developed LID on first dose of levodopa despite having mild disease (UPDRS motor score 8 OFF drug); suggesting that genetic factors may reduce the threshold for expression of LID in this population.

Overall, this interesting study—which took advantage of a unique clinical opportunity—indicates that physicians should not be afraid of using levodopa (in low doses) to treat patients early in the course of Parkinson’s disease. Withholding the most effective anti-parkinsonian drug for fear of motor complications seems inappropriate. This conclusion concurs with evidence from the recent report of the PD-MED trial, which demonstrated that patients started on levodopa versus levodopa-sparing therapies have very similar long-term outcomes (PD Med Collaborative Group, 2014).

References