ANALYSIS OF PREDICTIVE FACTORS OF NEURODEGENERATION IN PATIENTS WITH BIPOLAR DISORDER AND PARKINSONISM: A NESTED CASE-CONTROL STUDY

INTRODUCTION AND AIM OF STUDY: BIPOLAR DISORDER (BD) HAS BEEN SUGGESTED TO BE A RISK FACTOR FOR THE DEVELOPMENT OF PARKINSON’S DISEASE (PD). STANDARD TREATMENT OF BD INCLUDES DRUGS THAT ARE KNOWN TO INDUCE DRUG-INDUCED PARKINSONISM (DIP). CLINICAL DIFFERENTIATION BETWEEN PD AND DIP IS CRUCIAL AND MIGHT BE AIDED BY FUNCTIONAL NEUROIMAGING OF THE DOPAMINERGIC NIGROSTRIATAL PATHWAY. THUS, THE AIM OF THIS THESIS IS TO PERFORM A CLINICAL AND SCINTIGRAPHIC CHARACTERIZATION OF CONSECUTIVE BD PATIENTS WITH PARKINSONISM, THE LATTER BY EXPLORING THE FUNCTIONAL INTEGRITY OF THE DOPAMINERGIC NIGROSTRIATAL PATHWAY WITH THE USE OF DATSCAN, AND TO ANALYZE ANY POSSIBLE PREDICTIVE FACTORS OF NEURODEGENERATION IN THE BD POPULATION. METHODS: WE ENROLLED CONSECUTIVE BD PATIENTS WITH PARKINSONISM. SCREENING FOR PARKINSONISM WAS PERFORMED WITH THE SCANMOVE INSTRUMENT. CLINICAL EVALUATION WAS PERFORMED USING THE MDS-UPDRSIII SCALE. PATIENTS WERE ALSO SCREENED FOR REM SLEEP BEHAVIOR DISORDER, USING THE RBD SCREENING QUESTIONNAIRE, AND HYPOSMIA, USING THE UNIVERSITY OF PENNSYLVANIA SMELL IDENTIFICATION TEST (UPSIT). A FIRST LEVEL COGNITIVE ASSESSMENT WAS OBTAINED WITH THE MONTREAL COGNITIVE ASSESSMENT. PATIENTS UNDERWENT 123I-IOFLUPANE DOPAMINE TRANSPORTER SINGLE-PHOTON EMISSION COMPUTER TOMOGRAPHY (SPECT). A SUBSET OF BD PATIENTS PERFORMED GAIT ANALYSIS, AND THE OBTAINED PARAMETERS WERE COMPARED TO THOSE OF A POPULATION OF AGE-MATCHED PD PATIENTS. WE EMPLOYED MONOVARIATE AND MULTIVARIATE (MACHINE-LEARNING-BASED) STATISTICAL METHODS IN ORDER TO LOOK FOR POSSIBLE PREDICTIVE FACTORS OF NEURODEGENERATION. RESULTS: FORTY-ONE CONSECUTIVE BD PATIENTS WERE ENROLLED. EIGHT OF THEM (19,5%) HAD ABNORMAL SCANS (BD+). THE TWO POPULATIONS SIGNIFICANTLY DIFFERED IN TERMS OF AGE AT EVALUATION, WITH BD+ POPULATION SIGNIFICANTLY OLDER THAN THE POPULATION WITH NORMAL SCAN (BD-). MACHINE LEARNING-BASED CLASS DISCRIMINATION, PERFORMED WITH THE PLS-DA, RANDOM FOREST AND LOGISTIC REGRESSION ALGORITHMS, DID NOT REACH STATISTICAL SIGNIFICANCE. HOWEVER, WE DISCOVERED TRENDS IN WORST MDS-UPDRSIII “GAIT” SCORES, MORE FREQUENTLY REFERRED HYPOSMIA AND HIGHER RBD SCORES IN THE BD+ CLASS, COMPARED TO BD- CLASS. LEFT ACTION TREMOR TENDED TO BE SLIGHTLY WORSE IN THE BD- CLASS. MOREOVER, THE BD - CLASS TENDED TO REPORT A HIGHER NUMBER OF EPISODES AND HIGHER DOSES OF ANTIPSYCHOTICS. BD UPSIT SCORES WERE COMPARED TO THOSE OBTAINED BY A POPULATION OF AGE-MATCHED PD PATENTS AND CONTROLS. DIFFERENCE WAS STATISTICALLY SIGNIFICANT AMONG THE THREE CLASSES, WITH THE BIPOLAR POPULATION SCORING INTERMEDIATELY BETWEEN PD AND CONTROLS. GAIT ANALYSIS SHOWED AN OVERALL WORSE GAIT PERFORMANCE IN BD PATIENTS COMPARED TO PD, ASSOCIATED TO A HIGHER INTERFERENCE DURING COGNITIVE DUAL TASK. A FURTHER TREND IN WORSE PARAMETERS IN THE BD+ POPULATION, COMPARED TO BD-, WAS ALSO DETECTED. CONCLUSIONS: ABOUT 20% OF BD PATIENTS WITH PARKINSONISM MIGHT HAVE AN UNDERLYING DOPAMINERGIC DEFICIT, WHICH WOULD NOT BE DUE TO CUMULATIVE EXPOSURE TO OFFENDING DRUGS AND IS HIGHER THAN EXPECTED IN THE GENERAL POPULATION. THIS SUPPORTS THE EVIDENCE THAT BD MAY REPRESENT A RISK FACTOR FOR SUBSEQUENT DEVELOPMENT OF NEURODEGENERATIVE PARKINSONISM. TRENDS IN MORE FREQUENT REFERRED HYPOSMIA AND HIGHER RBD SCORES, AS WELL AS WORSE GAIT FEATURES, IN THE BD+ POPULATION, MAY REPRESENT POSSIBLE PREDICTIVE FACTORS OF NEURODEGENERATION, WHICH MUST BE VERIFIED ON A LARGER COHORT.