Observational, prospective, longitudinal study.
Consecutive patients will be recruited from the Neurology Outpatient Clinic of two hospitals: the University Hospital Maastricht in Maastricht and the Maasland Hospital in Sittard, The Netherlands. TCD will be done in the departments of Clinical Neurophysiology of the two above mentioned hospitals. SPECT scanning will be done in the departments of Nuclear Medicine of the two hospitals
The Institutional Review Board (IRB) of the University Hospital Maastricht has approved the study (MEC 05–228, 4 April 2006). (This IRB also functions as IRB for the Maasland Hospital in Sittard, The Netherlands). All patients will be asked for informed consent through a standardised information form that is also approved by the Institutional Review Board.
250 consecutive patients with new parkinsonian signs and symptoms (of unclear origin at the time of visit) referred to the Neurology Outpatient Clinics of the University Hospital Maastricht (n = 150) and the Maasland Hospital Sittard, (n = 100).
1. Patients with parkinsonian signs and symptoms of unclear origin at the time of visit at the Neurology Outpatient Clinic. In his/her differential diagnosis the treating neurologist should be considering one of the following conditions: PD, MSA, and PSP. ET, VP or DIP.
2. Age older than 18 years.
1. Patients presenting with a clear unequivocal diagnosis of their parkinsonism.
2. Patients whose life expectancy is less than the required follow-up of two years.
After informed consent all subjects will undergo a structured interview, neurological examination (See Additional file 1
), TCD and SPECT within 6 weeks of the initial visit at the Outpatient Clinic. After two years all patients are re-examined by two movement disorder specialist neurologists for a final clinical diagnosis. This diagnosis serves as a surrogate gold standard to calculate the accuracy of SPECT and TCD to differentiate between PD and other types of parkinsonism.
A. Interview and neurological examination
After informed consent the patient is seen on the outpatient clinic for the inclusion interview and neurological examination by a third party physician (i.e. a physician not treating the patient, and blinded for information in the routine clinical records). In the structured interview a standard form with the following items are discussed: medical history, used drugs and effect, intoxications, duration of complaints, and most affected body- side (See attachment 1). The following clinimetric scales are scored: UPDRS (parts I, III and IV) [32
], Hoehn and Yahr score [33
], Hamilton Rating Scale [34
] for depression and the SCOPA cognition scale [35
]. The Sniffin Sticks smell test is done according to a standardised protocol [36
]. Finally, the including physician will try to reach a probable diagnosis, strictly applying the UK Parkinson's Disease Society Brain Bank criteria [2
All subjects will undergo SPECT scanning within 6 weeks of inclusion in the study. In this study FP-CIT (123I-ioflupane, Nycomed, Amersham, U.K.) is used as presynaptic radiotracer. Medication (amphetamine, citalopram, fentanyl, cocaine, fluoxetine, fluvoxamine, paroxetine, sertraline, venlafaxine) which could interfere with the radiotracer is discontinued at least 5 half life times. After intravenous injection of the tracer, SPECT measures baseline dopamine transporter integrity in the brain. SPECT is performed with a triple head camera (MultiSPECT3, Siemens, Ohio, USA) equipped with high-resolution collimators. A semi-automatic template model programme is used to calculate the ratios between left striatal and right striatal and occipital regions respectively. Total time of acquisition is 30 minutes (45 seconds per frame for 40 views per detector). Zoom factor: 1.00 and the matrix size: 128 × 128. Filtered back-projection acquisition is performed. Images are filtered using a Butterworth filter with a cut-off value of: 0.4–0.5 and an order of 5. A division between the caudate nucleus and putamen is made. The ratios are corrected using Alderson's brain phantom, with known activities in the caudate nucleus and putamen. A binding of two standard deviations below healthy controls is considered as abnormal (FP-CIT 8.25 +-1.85 for putamen and 7.76 +-1.77 for caudate nucleus). Beside quantitative the scans will be also judged visually by the same nuclear specialist blinded for the final clinical diagnosis. If quantitative and visual judgments do not agree the conclusion of visual judgment is taken (unpublished data).
C. Transcranial Duplex Scanning (TCD)
TCD investigation is performed bilaterally through the pre-auricular bone window with a 2–4 MHz phased array transducer (SONOS 5500; Philips, Eindhoven, the Netherlands) by an experienced sonographer, blinded for the clinical data and SPECT results. The quality of the temporal bone window, the SN and Raphe nuclei (RN) of all subjects are scored directly by the sonographer blinded for the final clinical diagnosis and SPECT result. The quality of the bone window is scored as good, moderate or inconclusive.
Two different methods are applied for the evaluation of the echointensity of the SN. Firstly, the presence or absence of an obviously visible bilateral hyperechogenic SN is scored (qualitative method). The SN are scored as hyperechointens, not hyperechointens or inconclusive (= no typical configuration of hyperechointensity or low quality of the temporal bone window). Secondly, the area of an eventually hyperechogenic SN will be measured quantitatively (quantitative method). Both the right and left SN are measured from both sides, i.e. both temporal bone windows. After encircling, the area is automatically calculated. A hyperechogenic area of at least 0.2 cm2 is classified as characteristic for PD. The RN are scored as: invisible (= iso-intense), just visible, visible (= hyper-intense) or inconclusive (= doubtful echointensity or low quality temporal bone window).
To determine inter-observer variability and to increase the power of the study a second sonographer will also judge the acquired echo data. A loop of 64 images of each patient will be acquired scanning the brainstem cranio-caudally and will be stored in order to allow for off-line analysis. Off-line the quality of the temporal bone window, SN and RN will scored by the second sonographer.
D. Regular Outpatient follow-up
The initial treating neurologist will remain responsible for the regular outpatient management of the patient included in the study. He or she will discuss the test results with the patient, and base his/her treatment plan on these. All further clinical decisions will be made by the treating neurologist.
E. Re-examination at two-year follow-up
Two years after inclusion, all patients will be re-examined separately by two independent movement disorder specialist neurologists blinded for the tests results. They will also be blinded for the clinical records of the treating neurologist. The same standard form as in the first visit is filled in (see Additional file 1
) and they will be asked to reach a clinical diagnosis, independently from each other, according to generally accepted clinical criteria [38
]. If these two diagnoses are not identical, the final diagnosis of this patient will be coded as inconclusive.
Our main hypothesis is that TCD is as sensitive as SPECT to differentiate PD from other parkinsonian disorders. For the power analyses we assumed a sensitivity of SPECT of 90%, based on the analysis of our own data on 248 consecutive patients [19
]. Assuming this 90% for TCD sensitivity we can accept as lowest border of the 95% confidence interval, 86% or higher:
SD = (√ (p1 (1-p1))/n → p1 = 0.9, sd 0.02, implying n = 190 patients, needed who have the hyperintensity of SN with TCD scanning.
In our pilot study 15% of the patients had an insufficient temporal bone window, so 224 patients are needed to compensate for the amount of inconclusive TCDs [28
]. Based on this study we expect that 90% of all patients with inconclusive parkinsonism will ultimately have PD, so we set a target of initial 250 patients with unclear parkinsonism needed for this trial. We will calculate the sensitivity and specificity, positive predictive value, negative predictive value and diagnostic odd's ratio (OR) with its 95% confidence intervals (95% CI) of the first clinical judgment, TCD, FP-CIT SPECT scan and smell tests to predict the clinical diagnosis after 2 year follow-up. Accuracy is determined for all parkinsonian subgroups separately (PD versus APS, PD versus ET, and PD versus VP, PD versus DIP, and PD versus all other types of parkinsonism). For expected SPECT, TCD and smell tests scores for each parkinsonian disorder, see table .
Expected TCE, FP-CIT SPECT and odour recognition results for all parkinsonian disorders
Additionally we will determine the predictive value of TCD compatible with PD for an abnormal FP-CIT SPECT scan. Finally the inter-observer reliability for SN and RN judgement by TCD will be determined (Cohen's kappa test).
SPSS 11.0 for windows (SPSS, Chicago, IL, USA) and StataSE9 (Stata corporation, Texas, USA) will be used for statistical analysis.