The Human Experience with Intravenous Levodopa

Objective: To compile a comprehensive summary of published human experience with levodopa given intravenously, with a focus on information required by regulatory agencies. Background: While safe intravenous (IV) use of levodopa has been documented for over 50 years, regulatory supervision for pharmaceuticals given by a route other than that approved by the U.S. Food and Drug Administration (FDA) has become increasingly cautious. If delivering a drug by an alternate route raises the risk of adverse events, an investigational new drug (IND) application is required, including a comprehensive review of toxicity data. Methods: Over 200 articles referring to IV levodopa were examined for details of administration, pharmacokinetics, benefit, and side effects. Results: We identified 142 original reports describing IVLD use in humans, beginning with psychiatric research in 1959–1960 before the development of peripheral decarboxylase inhibitors. At least 2760 subjects have received IV levodopa, and reported outcomes include parkinsonian signs, sleep variables, hormone levels, hemodynamics, CSF amino acid composition, regional cerebral blood flow, cognition, perception and complex behavior. Mean pharmacokinetic variables were summarized for 49 healthy subjects and 190 with Parkinson's disease. Side effects were those expected from clinical experience with oral levodopa and dopamine agonists. No articles reported deaths or induction of psychosis. Conclusion: At least 2760 patients have received IV levodopa with a safety profile comparable to that seen with oral administration.


INTRODUCTION
Impairments in dopaminergic neurotransmission in the basal ganglia are a hallmark of Parkinson disease (PD), the second most common neurodegenerative disease. Replacement of dopamine has been the cornerstone of treatment for PD. Because dopamine itself does not cross the blood-brain barrier (BBB), its immediate precursor levodopa (L-3,4-dihydroxphenylalanine, L-DOPA) is administered since it crosses the BBB (Hornykiewicz, 1963;Cotzias et al., 1967;Birkmayer and Hornykiewicz, 2001). Although purified levodopa was first ingested by mouth in 1913 (Roe, 1997), it was first used for medical treatment by intravenous (IV) rather than oral administration (Pare and Sandler, 1959;Birkmayer and Hornykiewicz, 2001).
Oral levodopa has become the preferred method of treatment clinically, but IV levodopa administration still holds advantages over the oral form for some purposes. First, the rapid administration of IV levodopa is often necessary for certain study designs, including those focused on the pharmacokinetics and pharmacodynamics of the drug. Additionally, the IV route leads to more predictable plasma levodopa concentration because oral medications have highly variable absorption characteristics, especially in PD patients (Bushmann et al., 1989), with differences in absorption based on sex and age (Robertson et al., 1989;Kompoliti et al., 2002). IV administration also permits researchers to keep brain levodopa concentrations constant while assessing physiological responses over time. Recent years have seen increasing interest in potential benefits of continuous dopaminergic stimulation in the treatment of PD (Jenner et al., 2011). Continuous stimulation helps avoid wearing off of motor benefit during levodopa nadirs, and there is also some evidence that it may reduce the risk of, or mitigate, dyskinesias and other peak-dose side effects. Thus, IV levodopa may prove useful for human studies investigating the pathophysiology of continuous vs. pulsatile dopaminergic stimulation in humans. Finally, IV levodopa is sometimes used clinically in patients who cannot tolerate oral medications, such as PD patients during surgery or on total parenteral nutrition.
Current U.S. FDA regulations focus heightened scrutiny on research in which drugs are delivered by a route for which the drug has not been approved. Predictably, in addition to any safety benefits, the heightened scrutiny has created practical obstacles to research with IV levodopa, as described for instance by Rascol et al. (2001, p. 250). Specifically, an IND (Investigational New Drug) application must be submitted if the risks of IV administration significantly exceed those of oral levodopa [ §21 CFR 312.2(b)(iii)]. Therefore, the overall goal of this paper is to determine whether or not IV levodopa carries risks greater than oral administration by compiling a literature review that comprehensively summarizes the human experience with intravenously administered levodopa. We tabulate the extent of human exposure, side effects, benefits, and efficacy. We also summarize pharmacokinetic (PK) and pharmacodynamic (PD) parameters from these studies. These data should help inform decisions about whether IV administration of levodopa requires an IND.

METHODS
The authors searched MEDLINE and OVID, reviewed selected books, searched toxicity databases, and followed references cited in those sources. Articles written completely in languages other than English, French, German, Italian, Spanish, or Portuguese were excluded. Search terms included (levodopa/L-dopa/DOPA) AND (intravenous/intravascular/infusion/injection/i.v.); limit to humans; search date through May, 2015. Studies using oral or intraduodenal l-DOPA administration were excluded except for PK/PD studies cited in Table 2. Studies in which IV levodopa was always coadministered with monoamine oxidase inhibitors (MAOIs) or catechol-O-methyltransferase (COMT) inhibitors were excluded. Levodopa methyl ester  and d,l-DOPA (Pare and Sandler, 1959) were included, but PK/PD calculations were corrected for the difference in molecular weights. Co-administered drugs were reported if included by the authors.
We recorded total dose and maximum infusion rate. We also recorded pharmacokinetic (PK) and pharmacodynamic (PD) parameters where available, including steady state volume of distribution (VOD), clearance, distribution half life (t ½α ), elimination half life (t ½ or t ½β ), E max , and EC 50 . Reported data were used to calculate any missing PK parameters where possible. Additionally, any reports on efficacy were noted. Side effect frequency was recorded if reported. The number of subjects and subject conditions (Parkinson disease, other disease states or healthy volunteers) were recorded for each study. Average PK parameters were calculated across studies, weighted by the number of subjects.

RESULTS
One hundred forty-two articles reporting intravenous levodopa administration were identified. Most subjects with parkinsonism were diagnosed with idiopathic PD, but some studies reported a variety of etiologies including postencephalitic and vascular parkinsonism and PSP. PD patients differed in their history of prior drug treatment before the studies with conditions including de novo, fluctuating, on-off, and stable. Some subjects were treated with levodopa for conditions other than PD (see Table 1: Patient Populations and Response Parameters), including other movement disorders (dystonia, progressive supranuclear palsy [PSP], neuroleptic malignant syndrome [NMS], primary psychiatric disorders (schizophrenia, mood disorders, personality disorders), endocrine disorders (diabetes mellitus, essential obesity, hypopituitarism), hepatic disease (alcoholic cirrhosis, steatohepatitis, hepatic encephalopathy), cardiac valvular disease, and asthma. Healthy controls were also included in some studies.
Pharmacokinetic data were reported for a total of 251 human subjects (see Table 2: Pharmacokinetics of Levodopa). Co-administration of a peripheral decarboxylase inhibitor (PDI) lowered the clearance and increased the elimination half-life of intravenously administered levodopa, while there was no notable effect of PDIs on volume of distribution. Additional PK data are available from studies that gave levodopa by

Patient populations
Response parameters other routes (Sasahara et al., 1980a;Poewe, 1993;Muhlack et al., 2004;LeWitt et al., 2009), and several studies report the bioavailability of oral doses relative to IV administration (Sasahara et al., 1980b;Robertson et al., 1989;Kompoliti et al., 2002). The pharmacodynamic data (see Table 3: Reports of Human Experience with IV Levodopa) represent a total of 2760 human subjects, with a significant variety of patient groups and a multitude of response parameters (see Table 1). No side effects were reported for 1260 subjects. The highest total IV dose was 4320 mg in 1 day, given to a patient with idiopathic PD and carcinoma of the retina. The patient reported no adverse effects at this dose. The highest reported single bolus dose was 200 mg, and the highest infusion rate was 5.0 mg/kg/h. Concomitantly administered peripheral decarboxylase inhibitors included carbidopa and benserazide. PDIs affected clearance and volume of distribution (as mentioned above), minimized gastrointestinal symptoms, and allowed subjects to be given lower doses of levodopa. Other concomitant drugs are listed, to help explain any side effects that might be caused by concomitant drug administration or an interaction with levodopa rather than by levodopa alone. These include adenosine receptor antagonists (istradefylline,tozadenant [SYN115], aminophylline, caffeine), stimulants (amphetamines, methylphenidate), dopamine receptor agonists (apomorphine, terguride, SKF38393), monoamine oxidase (MAO) inhibitors, dextromethorphan, estradiol, paroxetine, and dantrolene.
A variety of neurological, psychiatric, cardiovascular, and other physiological effects of levodopa were monitored (see Table 1). There were no reported deaths. There were no instances of psychosis, even when attempting to elicit it in susceptible subjects (Goetz et al., 1998). There were also no life-threatening events (serious adverse effects) following IV levodopa administration at high doses, regardless of whether a PDI was co-administered. With co-administration of a PDI, the dosage range causing side effects (mainly nausea and asymptomatic hypotension) was 45-150 mg as a single bolus or infusions of 0.5-2.0 mg/kg/h. Without a co-administered PDI, side effects were reported with a bolus of 60-200 mg or an infusion of 1.5-3.0 mg/kg/hr. Side effects were more likely with higher doses, but other factors such as age, sex, disease severity, and prior treatment also played a role in side effects of levodopa.
Other than these side effects found at high doses, several milder or less frequent side effects were reported. These primarily included mild autonomic changes (orthostasis and tachycardia), psychiatric changes (sedation, anxiety, insomnia, and improvement in mood), and neurologic effects (improvements in tics, REM sleep changes, subjective weakness, headaches, and increased dyskinesias). Various other effects were noted in isolated reports (listed in Table 3). It is important to note that both side effects and efficacy depended strongly on subject factors including gender, age, past treatment, and disease state. Also, dsykinesia was mentioned as a side effect only in patients with PD, and most often in those with a long history of previous levodopa treatment.
Motor benefits of levodopa in PD have been demonstrated conclusively. Additional reported benefits of IV levodopa treatment in PD included improved sleep (Hardie et al., 1984) and attenuation of early morning akinesia or dystonia . In other patient groups, benefits of IV levodopa included improvement of the comatose state in hepatic encephalopathy (Abramsky and Goldschmidt, 1974) and improvement in depressive and somatoform symptoms (Ingvarsson, 1965a). One report found it more effective than dantrolene for treating neuroleptic malignant syndrome (Nisijima et al., 1997). More recently, IV levodopa treatment was found to alleviate the neuropsychiatric adverse effects associated with interferon-alpha, namely lethargy, hypersomnia, depression, agitation, akathisia, and confusion (Sunami et al., 2000).

DISCUSSION
The existing literature strongly supports the safety of IV levodopa, which has been used in humans for more than half a century (Pare and Sandler, 1959). IV levodopa has been administered to over 2700 human subjects. Despite infusion rates as high as 5.0 mg/kg/h and boluses as large as 200 mg, there are no recorded instances of death or of other serious adverse effects of IV levodopa, nor have there been documented cases of other serious side effects, such as psychosis, that might limit its use in humans. Milder side effects, the most significant of which are nausea and vomiting, were most prominent with rapid infusions in the range of 1-2 mg/kg or 100-200 mg over less than 15 min (Bruno and Bruno, 1966;Fehling, 1966;Rinne and Sonninen, 1968;Moorthy et al., 1972;Quinn et al., 1984;Black et al., 2003). These conclusions are supported by safety data from other species. The Registry of Toxic Effects of Chemical Substances reports the lowest published toxic dose of levodopa in any nonhuman species as 2.5 mg/kg, referring to a subtle behavioral effect on a learning measure in a mouse (NIOSH and Biovia, 2015) 1 . The lowest IV levodopa dose that was lethal to half of subjects (LD50) was ">100 mg/kg" in rats. In mice, the LD50 ranges from 450 mg/kg (administered intravenously) to 4449 mg/kg (administered subcutaneously). Typical human doses are in the range of only 1 mg/kg; thus, human studies with IV levodopa administer doses substantially lower than those dangerous to nonhuman mammals.
In summary, IV levodopa has similar efficacy and side effects as oral levodopa (Connolly and Lang, 2014) and dopamine agonists (Bonuccelli and Ceravolo, 2008). These include gastrointestinal (nausea, vomiting, and abdominal discomfort) and neuropsychiatric effects (sedation, dyskinesias). Nausea and orthostatic hypotension, side effects of both IV and oral levodopa, are largely blocked by PDIs and are less common in patients accustomed to dopamimetic treatment. The other side effects are infrequent and neither serious nor life-threatening (Connolly and Lang, 2014). When given with adequate PDI pretreatment, IV levodopa has minimal if any cardiovascular effects (Siddiqi et al., 2015).
The safety of IV levodopa is important for patients but also for regulatory review. Changing the route of administration of any drug in a study traditionally necessitates submitting an IND application if changing the route of administration "significantly increases the risks . . . associated with the use of the drug product" [ §21 CFR 312.2(b)(iii)]. The data from our review of the literature suggest that IV administration of levodopa does not significantly increase the associated risks of levodopa in comparison to oral administration. In summary, studies conducted throughout the past half century support the safety of IV levodopa administration in human patients.