Objectives To analyse the differences in the clinical features and characteristics of 123I-labelled 2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl) nortropane (123I-FP-CIT) single photon emission CT (SPECT) imaging among patients with vascular parkinsonism (VP) and Parkinson's disease (PD).
Methods We performed a case–control study to compare clinical features and qualitative and semi-quantitative analyses of 123I-FP-CIT SPECT images between 106 patients with VP and 280 patients with PD. A case series study was used to search for clinical features related to SPECT or neuroimaging findings among patients with VP.
Results Patients with VP had a higher age at symptom onset and lower disease duration than patients with PD. The most frequent symptom at onset was gait disorder in VP and tremor in PD. Gait disorder, postural instability and falls were more frequent in VP. Rest and mixed tremor were more prevalent in PD. Of the patients who received levodopa treatment in the VP group, only about half had a good response. Qualitatively 123I-FP-CIT SPECT images were normal in 32.5% of patients with VP and abnormal in all patients with PD. The use of different visual score patterns showed higher ability to differentiate VP from PD. Semi-quantitative analysis showed significantly higher uptake in the striatum, caudate and putamen in VP. The asymmetry index was higher in patients with PD. Among patients with VP, falls were the only clinical feature that demonstrated a correlation with the SPECT visual pattern.
Conclusion Our data contribute to the confirmation that VP and PD are two different clinical entities. Neurological signs, response to treatment and qualitative and semi-quantitative 123I-FP-CIT SPECT analyses may help to make the diagnosis.
- Cerebrovascular disease
- Parkinson's disease
- nuclear medicine
- 123I-FP-CIT SPECT imaging
- evoked potentials
- movement disorders
- functional imaging
- Alzheimer's disease
Statistics from Altmetric.com
- Cerebrovascular disease
- Parkinson's disease
- nuclear medicine
- 123I-FP-CIT SPECT imaging
- evoked potentials
- movement disorders
- functional imaging
- Alzheimer's disease
Critchley described a syndrome called ‘arteriosclerotic parkinsonism’ in 1929, which included short-stepped gait, bradykinesia and rigidity in older patients with a history of arteriosclerosis and several strokes. This author suggested that vascular lesions in basal ganglia should be involved in the aetiology of this kind of parkinsonism.1–3
Since Critchley's first description, the term ‘vascular parkinsonism’ (VP) or ‘lower-body parkinsonism’ has been a controversial concept in neurology, considering the high proportion of patients with vascular lesions who do not develop signs of parkinsonism. Fazekas reported frequencies of 10–30% of asymptomatic older patients with vascular risk factors and widely confluent lesions on MRI who did not develop parkinsonian signs.4 However, in recent decades several neuropathological studies have re-established VP as a distinct entity, although there have only been few studies on the histology of this disease and it is not clear how ischaemic lesions give rise to parkinsonism.5–7
The differentiation of VP from Parkinson's disease (PD) represents a diagnostic challenge for physicians and even for neurologists. Several clinical features have been associated with VP, including bilateral and symmetrical involvement of the lower limbs, gait disorder, short steps, postural instability, falls and absence of rest tremor. Furthermore, patients with VP usually have poor or no response to levodopa.8–10 However, these symptoms can vary greatly and the diagnosis of VP cannot be accurately confirmed on the basis of clinical features alone.
Single-photon emission computed tomography (SPECT) is one of the most useful tools to differentiate degenerative parkinsonisms from other parkinsonian syndromes in which symptoms or clinical course are unusual. The cocaine derivative 123I-labelled 2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl) nortropane (123I-FP-CIT) binds to dopamine transporter (DAT) in the striatum and acts as a good marker to differentiate presynaptic disorders from other parkinsonisms. In PD, striatal binding of 123I-FP-CIT is significantly reduced and the degree of reduction correlates with the severity of the disease.8 ,9 In VP, however, some patients have a normal presynaptic tracer binding, whereas others have a diffuse reduction of DAT with a pattern similar to that described in PD, in which a predominant reduction of tracer uptake is seen in the posterior putamen. The asymmetry index is a value which compares right and left striatal 123I-FP-CIT binding. In patients with VP, this value is normal or at least lower than in PD, showing the presence of a symmetric 123I-FP-CIT uptake in the basal ganglia.2 ,10–14 However, in patients with VP with a striatal vascular lesion, the asymmetry index can be much higher than in PD due to a strictly unilateral reduction of striatal tracer binding.14 All this supports the idea that VP is a heterogeneous entity and despite the advances in the last decades, there are still several gaps in our understanding of the disease.
In a typical patient, with classical lower limb involvement and a normal 123I-FP-CIT SPECT result, there can be a strong suspicion of VP. However, the distinction from PD is more difficult if patients show a more atypical clinical presentation or have abnormal 123I-FP-CIT SPECT imaging. In this study we describe the clinical features and the results of 123I-FP-CIT SPECT imaging in a cohort of patients with VP, comparing them to a group of patients with PD. Our aim was to find clinical or imaging differences between the two groups. We performed qualitative and semi-quantitative analyses of 123I-FP-CIT SPECT images to determine whether the degree and pattern of dopaminergic degeneration can help us to distinguish between the two groups. We performed cluster analyses based on quantitative data to determine different imaging patterns which contribute to predict a specific diagnosis. Furthermore, among patients with VP, we searched for the existence of clinical features related to the SPECT pattern or to structural neuroimaging findings.
Materials and methods
We retrospectively evaluated a cohort of 106 patients with VP from an initial sample of 137 patients who were seen at our centre from 2006 to 2011. The diagnosis of VP was made according to the diagnostic criteria proposed by Zijlmans et al 15 (see details about the selection of patients with VP in the online supplemental material). The control group was composed of 280 consecutive patients with PD seen in our clinics who fulfilled the UK Parkinson's Disease Society Brain Bank clinical diagnostic criteria.7 The diagnosis of patients in our study was based on clinical information collected from medical records regardless of the 123I-FP-CIT SPECT result.
We designed two types of studies: a case–control study comparing clinical features and the characteristics of 123I-FP-CIT SPECT imaging in both groups; a case series study to search for clinical features related to the SPECT pattern or to vascular lesions in neuroimaging studies among patients with VP.
This study was approved by the local ethics committee and complies with the Declaration of Helsinki.
The clinical features of patients in the two groups were collected from the medical records. This clinical information and the results of 123I-FP-CIT SPECT imaging in the VP group was compared with that observed in the PD group (see details about the clinical features analysed in the online supplemental material).
Neuroimaging studies (CT or MRI) was performed in all 106 patients with VP. These studies were performed less frequently in patients with PD, with a total of 177 out of the 280 patients (63.2%). The neuroimaging result was divided into: normal, periventricular and deep subcortical white matter lesions, lacunar strokes and territorial infarction.
Eighty patients in the VP group and 171 patients in the PD group underwent brain 123I-FP-CIT SPECT (see the SPECT protocol in the online supplemental material). In each group separately, we analysed the variables that could affect the 123I-FP-CIT uptake (age, disease duration time, Hoehn and Yahr scale and clinical variables) in patients who underwent SPECT and in patients who did not.
The 123I-FP-CIT SPECT images were analysed by two experienced nuclear-medicine specialists (DGS, VAMO), who were blinded to the patient's clinical status and who evaluated each DAT scan based on a qualitative analysis according to the degree of reduction in radioligand uptake. The image results were classified as normal or abnormal. Furthermore, for visual assessment of the images we used two types of scoring systems. The first system was a commonly used standardised visual system16 (see details about this scoring system in the online supplemental material). The second one was the following visual score system which was developed by our nuclear medicine physicians (figure 1):
0. Bilateral normal uptake.
1. Mild or moderate homogeneous and bilateral decreased striatal uptake.
2. Focal deficit of tracer accumulation on any striatal region or homogeneous unilateral striatal diminished uptake.
3. Symmetric or asymmetric striatal reduced uptake, more pronounced in putamen than in caudate, showing a rostrocaudal gradient, with or without other irregular associated defects.
Visual score results were categorised into two patterns (VP pattern/PD pattern), which separately grouped the majority of patients corresponding to each disease.
In addition, an automated semi-quantitative analysis was performed to evaluate specific/non-specific 123I-FP-CIT uptake ratios (see details about the semi-quantitative analysis protocol in the online supplemental material). We performed cluster analyses using the quantitative results provided by 123I-FP-CIT SPECT. For these analyses we used the values of the most affected putamen, the caudate ipsilateral to the most affected putamen and the asymmetry index. Similarly to visual score, the imaging cluster results were used to create a new dichotomous variable consisting of two patterns (VP pattern/PD pattern), which also grouped the majority of patients of each type. Finally, we analysed the relationship between our visual score system and the imaging clusters based on quantification.
To evaluate the ability to differentiate VP and PD, for qualitative analyses we used the dichotomous visual score patterns (VP pattern/PD pattern) and the normal/abnormal result, considering normal SPECT images as a positive test to diagnose VP. For quantitative analyses we used the established dichotomous imaging cluster patterns (VP pattern/PD pattern).
For the case series study, among patients with VP we analysed several clinical features (symptom at onset, arterial hypertension, tremor, impaired postural reflexes, gait disturbances, freezing, falls, cognitive impairment and response to treatment), searching for an association with the SPECT pattern. We compared all these clinical features with dichotomous qualitative analyses (normal/abnormal), visual scores patterns (VP pattern/PD pattern) and imaging cluster patterns (VP pattern/PD pattern). Furthermore, we looked for an association between these SPECT patterns and vascular lesions in neuroimaging studies, and between clinical features and the results of neuroimaging studies.
See details in the online supporting information about the statistical analyses.
As shown in table 1, the mean age at symptom onset and the male to female ratio were higher in the VP group than in the PD group. The mean disease duration time was lower in patients with VP. The Hoehn and Yahr scale was higher in patients with VP than in the PD group.
Among the vascular risk factors observed, hypertension was more frequent in patients with VP than in the PD group. We also observed significant differences in the symptoms at onset: gait disorder was more prevalent in the VP group, whereas tremor was more frequent in the PD group. Rest and mixed tremor were more common in patients with PD but we did not find significant differences in postural tremor. Most patients with VP developed gait disorder during the progression of the disease, which was significantly different from the PD group. Postural instability, falls, dysphagia and cognitive impairment were also more frequently observed in the VP group. After adjusting for age and sex, the differences in all these clinical features remained statistically significant. Gait freezing, urinary incontinence and emotional lability showed no significant difference between the two groups. In the multivariate analysis with imputed values some of the most important clinical findings remained significant, such as gait disorder as symptom at onset, hypertension, tremor, gait disorder during clinical course and dysphagia (see table S1 in the online supplemental material).
In the VP group, 73 of 106 patients (68.9%) received dopaminergic therapy, all of them receiving levodopa and 11 (10.4%) also received dopamine agonists. An improvement was seen in 35 patients (47.9%). In the PD group, 49 of 280 patients (17.6%) received treatment with levodopa, 33 (11.9%) with dopamine agonists and 196 (70.5%) with both, which was statistically different from the VP group. All patients with PD improved after treatment. The mean equivalent dose of levodopa was 409.9 mg in the VP group and 706.6 mg in the PD group, which was also significantly different (t=4.67, df=85.17, p<0.001).
For neuroimaging findings (CT or MRI), normal scans were described only in the PD group. Evidence of vascular lesions was seen on neuroimaging studies in both groups, but was significantly more common in the VP group (table 1).
Table 2 shows the distribution of normal and abnormal scans between the VP and PD groups. Qualitatively, 123I-FP-CIT SPECT images were normal in 32.5% of patients with VP and abnormal in all patients with PD.
Tables 3 and 4 show the distribution of the VP and PD groups in the two different scoring systems. In the standardised scoring analyses, visual scores 0 and 1 mainly grouped patients with VP while visual scores 2 and 3 mostly included patients with PD. However, in the newly developed scoring system, visual scores 0, 1 and 2 mainly grouped patients with VP while visual score 3 mostly included patients with PD. As shown in tables 5 and 6, considering this tendency in the grouping, visual score results were transformed into a new dichotomous variable with two possible patterns: VP pattern and PD pattern.
Our nuclear medicine specialists performed inter-observer and intra-observer agreement for qualitative analyses. For the normal/abnormal analysis the inter-observer agreement was very good (Cohen's κ =0.83), the intra-observer agreement was very good for observer 1 (Cohen's κ =0.96) and good for observer 2 (Cohen's κ =0.62). For the standardised visual score the inter-observer agreement was moderate (Cohen's κ =0.59) and the intra-observer agreement was very good for both observers (Cohen's κ =0.95 for observer 1 and 0.83 for observer 2). For the newly developed visual score, the inter-observer agreement was very good (Cohen's κ =0.81) and the intra-observer agreement was very good for observer 1 (Cohen's κ =0.91) and good for Observer 2 (Cohen's κ =0.74).
DAT density expressed as specific/non-specific 123I-FP-CIT SPECT uptake ratios was significantly higher in the most affected putamen and caudate and striatum ipsilateral to the most affected putamen in patients with VP compared with the PD group (figure 2). The putaminal area showed the greatest difference in DAT density between the two groups. The putamen/caudate ratio was also higher in patients with VP. The mean asymmetry index was higher in the PD group, showing the presence of a more symmetric 123I-FP-CIT uptake in basal ganglia in patients with VP.
In each group separately, we compared age, disease duration, Hoehn and Yahr scale and clinical features in patients with and without SPECT. No statistically significant differences were found in the VP group. In patients with PD, we found significant differences in age (t=4.31, df=278, p<0.001) and disease duration (t=4.09, df=178.77, p<0.001). Since patients with PD who underwent 123I-FP-CIT SPECT seemed to be younger and with shorter disease duration, we analysed the features of the scans in this cohort and we found that higher age and disease duration corresponded to lower putamen uptake (Spearman ρ=−0.311, p<0.001) and lower caudate uptake (Spearman ρ=−0.341, p<0.001). Therefore, the subjects who underwent SPECT are generally less affected than the general PD population.
Cluster analysis results
Three main groups were observed in the imaging clustering (see table S2 in the online supplemental material):
Cluster number 1 showed small asymmetry between both striata (asymmetry index 6.39). Putamen and caudate uptake ratios were rather high (1.48 and 1.94, respectively).
Cluster number 2 was characterised by a medium asymmetry index (19.39), low level of the most affected putamen uptake ratio (0.58) and slightly decreased ipsilateral caudate uptake ratio (1.10).
Cluster number 3 was characterised by a high asymmetry index (74.44) and very low levels of most affected putamen and ipsilateral caudate specific uptake ratios (0.27 and 0.61, respectively).
As shown in table 7, cluster 1 was mostly made up of patients with VP, whereas clusters 2 and 3 mainly included patients with PD.
In a similar way to visual analyses, imaging clusters were divided into two patterns: VP pattern (cluster 1) and PD pattern (clusters 2 and 3) (table 8).
Lastly we analysed the relationship between our visual score system and the imaging clusters based on quantification. We found a significant association between VP patterns and PD patterns (see tables S3 and S4 in the online supplemental material). Furthermore, in patients with VP we analysed the association between imaging clusters and normal/abnormal 123I-FP-CIT SPECT studies and we found that most of the normal studies were grouped in cluster 1, while the abnormal studies were spread over the three clusters (see table S5 in the online supplemental material).
As shown in table 9, analyses using patterns of the newly developed visual score showed the highest ability to differentiate between VP and PD compared with qualitative analyses expressed as normal/abnormal imaging, standardised visual score and imaging cluster patterns.
Clinical features related to neuroimaging findings
Among clinical features analysed in patients with VP, by searching for a correlation with the neuroimaging findings we found the following results:
Patients with subcortical and periventricular white matter disease more frequently had postural instability (χ2=7.48, df=1, p=0.02; OR=5.75, 95% CI 1.47 to 22.39).
Patients with lacunar strokes more frequently had falls (χ2=7.32, df=1, p=0.01; OR=3.61, 95% CI 1.38 to 9.42).
Patients with territorial infarction had a lower response to treatment (χ2=6.68, df=1, p=0.01; OR=0.09, 95% CI 0.01 to 0.79).
Clinical features related to the SPECT results
Among clinical features analysed in patients with VP, we only found a correlation between the visual score patterns and the presence of falls, such that, patients with VP and PD visual score pattern showed a lower probability of having falls (χ2=5.36, df=1, p=0.02; OR=0.25, 95% CI 0.72 to 0.85). Binary logistic regression was further assessed adjusting for age and sex with a significant result for the presence of falls (p=0.03).
Neuroimaging findings related to the SPECT pattern
No association was found between vascular lesions and SPECT patterns in the VP group.
In this study, we observed significant differences regarding sex distribution. As reported previously,5 ,12 ,17 ,18 the male to female ratio was higher in the VP group. These results suggest that men have a greater risk of developing VP than women. We also observed significant differences in the age at onset of symptoms and the duration of the disease. Winikates and Jankovic17 compared 69 patients with VP and 277 patients with PD. They concluded that patients in the vascular group were on average 8.6 years older than patients in the PD group and that the disease duration at the moment of evaluation was shorter among patients with VP (0.82 years). These differences have also been confirmed in some previous studies.1 ,12 ,19 ,20 We found greater differences in our study, with an age at onset 17.3 years older and a mean disease duration time 4.4 years shorter in the VP group. However, it must be taken into consideration that patients with PD seen at our centre are on average younger than those described in the general population. This may result in stronger differences between the two groups and could be a limitation of our study. To solve this problem, further studies should be carried out including patients from the general population instead of patients from clinical bases.
The higher prevalence of vascular risk factors in patients with VP has also been previously reported. Several authors12 ,17 ,18 ,21 ,22 have found that hypertension is significantly more common in the VP group than in the PD group, whereas there are no data on smoking or hypercholesterolaemia as vascular risk factors for VP.21 We also analysed the prevalence of hypertension and found similar results.
In our study, the clinical features of VP at onset and during clinical course were similar to those described by Critchley and other authors.1 ,5 ,19 ,23 We found that the most common symptom at onset in PD was tremor, whereas patients with VP presented mostly with gait disorder. Patients with VP in our study were more likely to have gait disorder, postural instability, falls and dementia and were less likely to experience tremor compared with patients with PD. Although these results could be partially explained by the difference in age between patients with PD and those with VP in our population, the differences remained statistically significant after adjusting for age and sex. These findings are in agreement with those obtained by other authors.12 ,17 ,24–28 However, unlike other reports, we have not found significant differences in gait freezing. Consistent with other studies,12 ,17 ,19 ,29–31 our patients with VP were less likely to respond to levodopa treatment. Approximately half of them did not improve after treatment. Considering the proportion of patients with a positive response, in doubtful cases a therapeutic trial with levodopa should be considered.19 One of the limitations of our study is that the data on neurological features were based on clinical records and not on prospective follow-up observations. Thus, we did not have chronologically complete information for all patients. For this reason multivariate analyses including all clinical variables could not be performed in our study, since a single missing value of any variable excluded the patient in question from the analyses. This results in a substantial reduction in cohort size and the consequent loss of statistical power. Future prospective studies should be carried out to solve these problems.
In line with previous literature,6 ,10 ,14 ,17 the predominant neuroimaging findings (CT or MRI) in our study in patients with VP were subcortical white matter lesions, followed by lacunar strokes and, less frequently, territorial infarction. Normal scans were described only in the PD group. Evidence of vascular lesions was seen on neuroimaging studies in both groups, but was significantly more common in the VP group.
In our study the qualitative 123I-FP-CIT SPECT imaging analysis was abnormal in about two-thirds of patients with VP. This proportion of abnormal imaging is higher than in previous studies.9 ,13 To try to further differentiate the SPECT results between patients with VP and those with PD, we used two different visual scores, the standardised one16 and one proposed by our group, both of which were made up of four different items. Based on SPECT semi-quantitative data, we performed a cluster analysis obtaining three different clusters. Our visual score patterns showed higher ability to differentiate VP and PD compared with imaging cluster, with the standardised scoring system and with qualitative analyses expressed as normal/abnormal imaging. However, when considering predictive values of diagnostic tests, it is important to recognise the influence of the prevalence of the disease, so it should be taken into account that it is possible that the reported data do not completely reflect the usual clinical conditions. Although one might expect that a quantitative approach would be more accurate in differentiating VP from PD, the explanation for a better precision of visual scoring is probably related to the distribution of alterations in striatal uptake in VP which are very irregular in distribution and size. The use of ‘regions of interest’ (ROIs) comprising the head of the caudate nucleus or entire putamen is not able to reflect the heterogeneous changes that occur in VP, whereas such changes can be detected by the visual analysis made by a nuclear medicine physician.
After comparing clinical features and the response to levodopa with SPECT results (dichotomous qualitative analyses (normal/abnormal), visual scores, imaging clusters), we only found an association between our visual score and the presence of falls. We did not find any association between the neuroimaging findings and the SPECT result.
Many reports9 ,13 ,32 claim that striatal 123I-FP-CIT binding is preserved or only mildly reduced in VP. However, Zijlmans et al 32 also described that dopaminergic deficit in patients with VP can sometimes be as marked as in patients with PD. Contrafatto et al 33 recently reported that the striatal asymmetry index is usually significantly higher in PD compared with VP, although occasionally, high asymmetry in bilateral uptake ratios can be found in these patients.32 In our study, we also found higher caudate and putamen uptake and higher putamen/caudate ratio in patients with VP compared with those with PD. The asymmetry index in our patients was also significantly higher in the PD group than in the VP group. This is consistent with the idea that the disease in VP is usually more diffusely distributed than in PD. All these data contribute to supporting the notion that this disorder is distinct from PD. Visual analysis and semi-quantitative analysis using ROIs provide a similar diagnostic performance in the identification of nigrostriatal degeneration. In general, semi-quantitative analysis does not provide additional information to the visual system;34 in some cases, experienced observers are able to identify patterns with extremely subtle changes, which cannot be observed via the ROI analysis.35 It would be necessary to implement new methods of quantitative analysis, specifically designed for this condition, to overcome these limitations.
A limitation of our study was that 123I-FP-CIT SPECT was not performed in all patients. After analysing the variables that could affect the 123I-FP-CIT uptake (age, disease duration, Hoehn and Yahr scale and clinical features) among patients with and without SPECT, no statistically significant differences were found in the VP group. However, we found lower age and shorter disease duration in patients with PD who underwent SPECT. We also found that higher age and longer disease duration corresponded to lower putamen and caudate uptake. Taking this into account, although in the PD group patients who underwent SPECT are not completely representative of the whole PD cohort, it seems that if these patients were older or had a higher evolution time our results would have been at least equal to or better in terms of sensibility and specificity.
To our knowledge, this is the study with the largest number of patients with VP included until now. Furthermore, it was conducted in a population from southern Spain, obtaining similar results to previous studies, which suggests that VP features are not influenced by the population characteristics. Thus, all the clinical features and the imaging findings which have been found to differ between both groups could be considered useful tools in differentiating VP from PD.
We would like to thank Juan Manuel Praena Fernández for his help with the statistical analyses.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement material
Funding IH-F, SJ, MTC, FC, MC and MP received support in the form of grants from the Ministerio de Ciencia e Innovación de España (SAF2007-60700); the Instituto de Salud Carlos III (PI10/01674); the Consejería de Innovación, Ciencia y Empresa de la Junta de Andalucía (CVI-02526, CTS-7685); the Consejería de Salud de la Junta de Andalucía (PI-0377/2007, PI-0741/2010), the Sociedad Andaluza de Neurología and the Jaques and Gloria Gossweiler Foundation.
Competing interests None.
Ethics approval Ethics approval was provided by the local ethics committee of Hospital Universitario Virgen del Rocio.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.