Method of levodopa response calculation determines strength of association with clinical factors in Parkinson disease

Abstract

Background: The levodopa challenge test is routinely used in Parkinson disease (PD) to determine a patient's motor improvement following levodopa administration [levodopa response (LR)]. LR is most commonly reported as a percent OFF to ON change in the Unified Parkinson Disease Rating Scale (UPDRS) part III score, and occasionally as an absolute difference in score. This inconsistency in LR determination alters how clinical factors such as patient age and disease duration are understood in relation to LR in PD. Objective: The aim of this study was to compare the calculation of the LR as either a percent change or difference in UPDRS-III motor score between OFF and ON medication. These two scores were then used to correlate to disease duration, patient age, levodopa duration, levodopa equivalent dose (LED), OFF score, cognition, mood, gait, and quality of life (QOL). Methods: 70 PD patients underwent the levodopa challenge test. The UPDRS-III motor examination was performed in the defined OFF and ON medication states to determine LR. Each patient was assessed after 12-14 h without anti-parkinsonian medication and then given three 100/25 mg levodopa/carbidopa tablets. LR was reported as both a difference in score [OFF - ON; absolute LR (aLR)] and as a percent change in score [(OFF - ON)/OFF*100%; %LR]. Patients completed the following non-motor symptom assessment scales: Montreal Cognitive Assessment, Freezing of Gait Questionnaire, Activities-specific Balance Confidence Scale, Parkinson's Disease Questionnaire, and Geriatric Depression Scale. The effect of the LR calculation method was correlated to the clinical measures. Results: The aLR was significantly associated with disease duration (r = 0.40), levodopa duration (r = 0.47), OFF motor score (r = 0.58), and LED (r = 0.31), but not age. The aLR was also found to have a significant relationship with clinical scales assessing cognition (r = 0.41), freezing of gait (r = 0.35), QOL (r = 0.40), and depression (r = 0.30). By contrast, the more commonly used %LR demonstrated no significant relationships with any of the variables tested. Conclusion: Although the %LR is more commonly employed in clinical protocols and research studies, the aLR is the superior method for reporting motor response to levodopa in PD given its significant associations with the clinical factors evaluated.