Predicting Long-Term Outcome of Patients of Early Parkinsonism with Acute Levodopa Challenge Test
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.314539
Source of Support: None, Conflict of Interest: None
Keywords: Acute levodopa challenge test, idiopathic Parkinson's disease, levodopa-induced dyskinesia, levodopa responsiveness, multisystem atrophy, predictabilityKey Message: In this study acute levodopa challenge test (ALCT) was performed in patients with early parkinsonism. It was useful in predicting a future diagnosis of idiopathic Parkinson's disease. Appearance of dyskinesia and levodopa intolerance predicted future diagnosis of multisystem atrophy.
Definitive diagnosis of parkinsonian syndromes is important for proper management and prognostication. Differentiating idiopathic Parkinson's disease (IPD) from other parkinsonian syndromes is extremely important to plan for management strategies, but this is a daunting task even for experienced clinician because of lack of biomarker. Diagnosis often requires longitudinal observation making it difficult to make a correct diagnosis at early stage. For early and short-term diagnostic purposes, acute pharmacological challenge tests are often used. These include acute challenge tests using a single dose of antiparkinsonian drugs (levodopa or apomorphine) and 123I-metaiodobenzylguanidine (MIBG) myocardial scintigraphy. As sustained dopaminergic responsiveness is one of the diagnostically significant features of IPD, whether a single dose of levodopa can reproduce the effect of sustained and long-term effects of levodopa is a matter of controversy.
Several investigators have evaluated the significance of acute levodopa challenge in predicting a clinical diagnosis of IPD and long-term response to levodopa in subjects with parkinsonism. But these studies were performed with diverse methodology, particularly in relation to the methods of assessment and the cut-off used to define a positive response.,,,,,, Excepting a few, most of these studies used the criteria that were always established a priori to call it a positive acute levodopa challenge.,, Moreover, methods used to assess the motor responses in these studies were clinical scales like modified Webster score, tremor score, and Unified Parkinson's Disease Rating Scale subset III (UPDRS-III) score., Some have used clinical measurement (finger tapping speed, walking time,,), or instrumental measurement for assessing motor responses. But some of the main flaws of previous studies were retrospective performance,,,, lack of blindness of the test examiner regarding the clinical diagnosis,,,, and the inclusion of only IPD patients.,
All these studies have evaluated the significance of motor improvement after a single high dose of levodopa. But there are other effects of the test e.g., the development of dyskinesia or dopaminergic side effects like nausea, vomiting, hypotension, and profuse perspiration. None of the earlier studies had looked into the significance of these manifestations in the differentiating various parkinsonian syndromes or predicting the development of levodopa-induced dyskinesia (LID) during long-term levodopa therapy. As early development of autonomic dysfunction is a feature of multisystem atrophy (MSA), we hypothesized that symptoms of levodopa intolerance (SLI) during the ALCT could predict a follow-up diagnosis of MSA. Reports of development of peak-dose dyskinesia at relatively low dose of levodopa in MSA also lead us to hypothesize that appearance of dyskinesia during acute challenge could be another predictor of MSA diagnosis.
In this study, we planned to evaluate the sensitivity and specificity of various outcomes of ALCT namely improvement of motor function, development of dyskinesia and intolerance, and tried to look into the power of ALCT to predict diagnosis of IPD, or any Parkinson-plus syndrome, and also to predict long-term levodopa responsiveness and levodopa-induced dyskinesia (LID) during long-term therapy.
In this prospective study 89 consecutive patients presenting with parkinsonism of less than two years duration, admitted to the Institute for diagnostic and treatment purposes between November 2015 to October 2016 were enrolled. Parkinsonism was defined by the presence of any two of bradykinesia, rigidity, resting tremor, and postural instability. Written informed consent was taken from all patients. Acute levodopa challenge test was performed in all patients who gave written consent. Afterwards all patients were treated with levodopa and were followed up for an average of 18 months. Brain imaging (MRI in most cases, or CT scan) was performed in all patients. Cognitive assessment, autonomic function testing, cerebrospinal fluid (CSF) tap test, positron emission tomography (PET) scan of brain and were done in selected cases. The study was approved by the institutional ethics committee.
Acute levodopa challenge test
We carried out test as per the method described by Merello on all of our patients. Domperidone (30 mg/day) was given for three days before the challenge test to prevent any unwanted gastrointestinal side effects of levodopa. Antiparkinsonian drugs were withdrawn the night before the evaluation day, in patients receiving these drugs. On the 4th day after starting domperidone, levodopa (250 mg)/carbidopa (25 mg), was given under fasting conditions to avoid the effects of meals on levodopa absorption in the morning. Motor symptoms were examined using the UPDRSm rating scale before and after the challenge test. Patients were evaluated clinically at baseline and every thirty minutes up to four hours. Symptoms of levodopa intolerance (SLI) could either be reported by the patient or observed by the physician, followed by the clinical evaluation of the patient. Development of any dyskinesia was also noted during the assessment. Time of appearance of these symptoms or dyskinesia after levodopa intake was systematically recorded.
Follow-up and diagnosis
At the end of 18 months patients were assessed for the presence of exclusion criteria for the diagnosis of IPD, for long-term responsiveness to levodopa, the progression of symptoms, development of dyskinesias, and finally to make a provisional clinical diagnosis [Figure 1]. The diagnosis of IPD was made as per the criteria proposed by the task force of Movement Disorders Society. The diagnosis of MSA and progressive supranuclear palsy (PSP) was made according to the standard criteria for clinical diagnosis. Diagnosis of other parkinsonian syndromes were made based on their respective clinical diagnostic criteria., Patients who did not fulfil any of these diagnostic criteria were classified as having an unclassified parkinsonian syndrome (UcPS). All those other than IPD were classified as non-IPD.
Positive acute levodopa challenge test - 30% improvement in UPDRSm score from baseline after giving 250 mg levodopa-carbidopa combination.
Symptoms of Levodopa intolerance (SLI) during the acute levodopa challenge test – Presence of any one of the following - nausea, vomiting, profuse perspiration, and hypotension during the test.
Long-term levodopa responsiveness – Any one of the below at the end of 18 months
Data were tabulated and analyzed using Stata (version 13) software and analyzed using Chi-square tests. The sensitivity and specificity of ACLT in prediction of future diagnosis were also calculated.
Out of 89 patients, 85 patients completed the ALCT and owing to development of akinetic-rigid state during antiparkinsonian drug withdrawal four patients were not subjected to the test. Ten patients were lost in follow-up over a period of 18 months, and thus seventy-five patients were available for the final analysis [see [Figure 1]]. Among them 42 were men and 33 were women, with mean age of 65.3 (SD = 9.4) years. Demographic profile of the patients are listed in [Table 1]. The mean duration of the disease was 16.9 (SD = 4.1) months, and the mean age of onset of parkinsonism was 63.7 (SD = 9.3) years.
The test was positive in 37 (43.5%) and negative in 48 (56.5%) patients. At the end of 18 months follow-up, 34 (45.3%) were diagnosed as having IPD. Of the 41 (54.7%) non-IPD parkinsonism, 12 (16%) patients could not be classified into any specific parkinsonian syndrome.
Overall sensitivity and specificity of ALCT to predict the diagnosis of IPD was 79.4% and 70.7%, respectively [see [Table 2]]. The sensitivity and specificity of the test to predict long-term levodopa response were both 60%. However, when patients with final diagnosis of IPD are considered, ALCT could predict long-term levodopa responsiveness with a sensitivity and specificity of 81.2% and 50% respectively.
Dyskinesia developed in 9 (12%) out of 75 patients during the test. We found that 33% of the patients who developed dyskinesia during the ALCT, compared to 27.3% of those who did not, developed levodopa-induced dyskinesia (LID) on long-term levodopa therapy, a difference which did not attain statistical significance (P = 0.7).
Regarding the sensitivity and specificity of the appearance of dyskinesia during ALCT to predict the future development of LID, it was 14.3% and 88.9% respectively. However, in the IPD group only it was 7.7% and 85.7%, respectively [see [Table 2]].
At the end of 18 months' follow up, 9 developed dyskinesia during ALCT and received a final diagnosis of MSA or non-MSA, and underwent further analysis. Out of nine patients with dyskinesia during ALCT, 4 (44.5%) received a final diagnosis of MSA. Whereas, out of 66 patients (who were classified as non-MSA), 5 patients (7.5%) had dyskinesia during ALCT (P = 0.019). Thus, the sensitivity and specificity of development of dyskinesia during ALCT to predict the diagnosis of MSA were 37.5% and 91% respectively [see [Table 2]].
We found significant difference (P = 0.0) between patients who developed (57.4%) vs who did not develop (5.9%) SLI during the test to get a diagnosis of MSA after 18 months. The sensitivity and specificity of the appearance of symptoms of SLI during the test to make a diagnosis of MSA was 50% and 95.5%, respectively [see [Table 2]].
Parkinsonisms are clinicopathological entities, and with the lack of biomarkers, they need autopsy to make a definite diagnosis. Clinical criteria of various parkinsonian syndromes require long-term follow-up, observing the response to levodopa and appearance of LID.  To make an early diagnosis, ALCT has been tried in the past, although, with diverse methodologies and results, including discrepancies regarding levodopa dosage, and the optimal cut-off point for a positive result.,, Similar to Merello et al., we considered the test as positive when there was a minimum 30% of improvement on UPDRSm. Our patients had parkinsonism of <two years duration and were followed up for 18 months. The sensitivity and specificity of a positive ALCT to predict the clinical diagnosis of IPD was 79.4% and 70.7%, respectively in our study. This makes this test a good diagnostic tool for IPD in short-term observation over a period of 18 months. This observation is comparable to those achieved by other researchers. Clarke and Davies reported the sensitivity varying between 64% and 85% and specificity between 59% and 80% (95% confidence interval). Rossi et al. reported sensitivity and specificity of the acute challenge test as 77.1% and 71.1% respectively, and almost similar rate of sensitivity (70.9%) and specificity (81.4%) was reported by Merello et al. Due to heterogeneous methodologies, it is difficult to compare directly the predictive value of the test in our study with the others.
For prediction of long-term levodopa responsiveness, the sensitivity and specificity of ALCT in our study were both 60%, which was somewhat poor. Thus, ALCT could predict a diagnosis of IPD more reliably, than predicting chronic levodopa responsiveness, an observation which is similar to the study of Rossi et al. However, when our study population was classified as IPD and non-IPD groups, the test sensitivity was much higher in IPD to detect long-term levodopa responsiveness (81.2%). Previously, some histologically proven cases of IPD were found non-responsive to adequate doses of levodopa. There is also report of MSA patients showing significant improvement with ALCT or apomorphine who could also be treated chronically with a dopamine agonist.
LID usually occurs when levodopa dose reaches its peak concentration in the brain. It was hypothesized that the appearance of dyskinesia during ALCT may be predictive of LID during long-term therapy. However, our study showed a very low sensitivity (14.3%) in this regard (though with specificity of 88.9%). When only the IPD group was considered, still LID could not be predicted accurately by the appearance of dyskinesia during ALCT. LID usually appears after few years of levodopa therapy in IPD. As average follow up of our cohort was only 18 months, development of LID on long-term treatment might not be detected in this short time.
There are reports of MSA patients developing peak-dose dyskinesia at relatively low levodopa dosages, indicating the possibility that dyskinesia during ALCT might be a predictor of MSA diagnosis. In our study, we have found a sensitivity and specificity of 37.5% and 91% respectively for the dyskinesia during an acute challenge in predicting a diagnosis of MSA.
SLI such as nausea, vomiting, profuse perspiration, and hypotension during the ALCT developed in a significant number of patients who underwent the challenge test, even if they were pre-treated with domperidone. Based on the description of Wenning et al., that there is early and profound autonomic dysfunction in MSA, we tried to assess whether the appearance of SLI during ALCT could predict an MSA diagnosis. The sensitivity and specificity of SLI during an ALCT for predicting MSA was 50% and 95.5% respectively. This is in sharp contrast to a retrospective study where a high sensitivity (80%) and a low specificity (44%) was documented nullifying the hypothesis. However, from our study, we can conclude that development of SLI during an ALCT increases the likelihood of getting a diagnosis of MSA.
From this study we conclude that positive ALCT is useful in predicting a clinical diagnosis of IPD, however, it cannot effectively predict long-term responsiveness to levodopa. Although the development of dyskinesia during the test could not correctly predict the development of LID during long-term levodopa therapy, it can predict a diagnosis of MSA. Finally, the appearance of SLI during the test also increases the likelihood of a future diagnosis of MSA.
The strength of our study is its prospective design using standard criteria, and its ability to correlate future diagnosis with ALCT. However, the study has few limitations. Firstly, the follow-up was for 18 months, which was less than the average time required to make a definitive diagnosis of IPD or the development of LID. Secondly, being a open-label study, there could be bias in interpretation. Thirdly, the lack of pathological correlation. Finally, majority of our patients were on some anti-parkinsonian medication, and thus the response of drug-naïve patients could not be evaluated.
We dedicate this article to our beloved teacher Late Prof. Shyamal Kumar Das who was involved in conceptualizing this study and supervised in every stage of data collection and analysis.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Table 1], [Table 2]