Accuracy of acute levodopa challenge for clinical prediction of sustained long‐term levodopa response as a major criterion for idiopathic Parkinson's disease diagnosis

Clinical idiopathic Parkinson's disease (PD) diagnosis requires following strict clinical criteria. Final definitive diagnosis can only be made after pathological confirmation and, despite following clinical criteria, several cases are misdiagnosed. We assessed sensitivity and specificity of acute challenge with levodopa (L-dopa) to predict sustained long-term L-dopa responsiveness as a major criterion for clinical diagnosis of PD. A consecutive series of 82 patients first seen at a movement disorders clinic with a parkinsonian syndrome without specific diagnosis was included. A second examiner, blind to the presumptive diagnosis, performed in each patient an acute challenge with 250/50 mg of L-dopa-carbidopa and rated the test as positive or negative according to whether values reached a minimal 30% of improvement on UPDRS scores. Positive tests were considered supportive of presumptive clinical diagnosis of Parkinson's disease. Blind to test results and according to clinical presumption, the first examiner started patient treatment with the necessary L-dopa dose or, alternatively, until reaching 1 g for 1 month in those who failed to display a positive test response. At 24 month follow-up, they were re-tested with 1 g for 1 month when required. At this point, clinical criteria of the U.K. Parkinson's Disease Society Brain Bank were applied and definitive clinical diagnosis of PD was made. Sensitivity, specificity, and positive predictive ratio for acute challenge were calculated. Overall sensitivity and specificity of acute L-dopa challenge to predict clinical diagnosis of PD was 70.9% and 81.4%, respectively; positive predictive ratio was 88.6%. When patients were divided into three groups according to their UPDRS motor section score at initial examination, sensitivity and specificity varied: Group I (or=21), 36.4% and 87%, respectively. Positive predictive ratio increased to 100% in Group I and to 87.5% in Group III. The positive result of initial acute L-dopa challenge predicts chronic L-dopa responsiveness as major criterion of PD in all patients with UPDRS motor scores lower than 10.