Pre- and postsynaptic dopamine SPECT in the early phase of idiopathic parkinsonism: a population-based study
- First Online:
- Cite this article as:
- Mo, S.J., Linder, J., Forsgren, L. et al. Eur J Nucl Med Mol Imaging (2010) 37: 2154. doi:10.1007/s00259-010-1520-3
- 218 Views
The aim of this study was to assess the diagnostic contribution of pre- and postsynaptic dopamine SPECT in drug-naïve patients with early idiopathic parkinsonism and to investigate possible differences between idiopathic Parkinson’s disease (PD) and atypical parkinsonian syndromes (APS) and possible differences in motor subtypes of parkinsonism.
A group of 128 newly diagnosed idiopathic parkinsonian patients and 48 healthy controls was studied. Presynaptic baseline SPECT with 123I-FP-CIT was performed in all patients and in 120 patients also a baseline postsynaptic SPECT with 123I-IBZM. Clinical diagnoses were reassessed after 12 months.
Presynaptic uptake in the putamen and caudate was significantly reduced in patients compared to controls. Presynaptic uptake ratios were not different between PD patients and patients with APS, and postsynaptic uptake in APS was not significantly reduced compared to PD or controls. In half of the APS patients both pre- and postsynaptic uptake ratios were reduced on the same side in the striatum. Impaired motor performance was associated with decreased presynaptic uptake in the putamen in PD. The postural instability and gait difficulty (PIGD) subtype of PD had lower presynaptic uptake ratios than patients with tremor-dominated (TD) symptoms.
Not only presynaptic putamen uptake ratios, but also caudate ratios were reduced in a majority of the patients in our study. At baseline scan, i.e. in an early stage of the disease, the accuracy of excluding APS in the whole study population was 85% using a combination of pre- and postsynaptic SPECT. Already at baseline, lower presynaptic SPECT ratios were seen in PD with PIGD at onset compared to those with TD subtype.
KeywordsParkinson’s diseaseParkinsonian syndromes123I-IBZM123I-FP-CITSPECT
The accuracy for diagnosis of atypical parkinsonian syndromes (APS) is typically low, especially in the early stages of the disease. Even in late stages a substantial proportion of patients do not receive the correct diagnosis [1, 2]. The current clinical diagnostic criteria for Parkinson’s disease (PD) , progressive supranuclear palsy (PSP)  and multiple system atrophy (MSA) [5, 6] are not designed for diagnostic purposes in early stages of these disorders and do not make use of laboratory tests or imaging. Many imaging methods, e.g. magnetic resonance imaging (MRI), dopamine studies with single photon emission computed tomography (SPECT) or positron emission tomography (PET) and other auxiliary tests such as anal sphincter electromyography (EMG), facial reflexes, eye movement analysis and cerebrospinal fluid markers, have been studied in this respect. Most of these tests have been studied in patients with moderate to advanced disease and the results are not readily applicable for use in early stages of the disorders.
Both pre- and postsynaptic imaging of the dopaminergic system have been extensively studied in these disorders and 123I-FP-CIT SPECT (123I-ioflupane, GE Health B.V., Eindhoven, The Netherlands), a presynaptic radioligand with affinity to the presynaptic dopamine transporter protein, has been shown to be useful to separate patients with PD from essential tremor (ET) , psychogenic parkinsonism and drug-induced parkinsonism (DIP) [8–10]. It has not been possible to separate idiopathic PD from APS with 123I-FP-CIT SPECT .
The postsynaptic receptor ligand 123I-IBZM (123I-iolopride, GE Health B.V., Eindhoven, The Netherlands) with affinity to the D2 and D3 receptors located on postsynaptic neurons in the brain, with high concentration in the striatum , had earlier been reported to be of use to differentiate PD from APS , and the combination of both presynaptic and postsynaptic investigations has been reported to be of use in the differentiation between PD and certain forms of APS [14–17]. In the literature, uncertainty still remains about the benefit of IBZM SPECT in the diagnostic arsenal for corticobasal degeneration (CBD) and whether a normal IBZM SPECT can exclude APS . The diagnostic power seems to increase using a combination of pre- and postsynaptic dopamine SPECT in the differential diagnosis of PD versus APS . A recent review by Felicio et al. states that postsynaptic dopamine receptor imaging in addition to presynaptic dopamine transporter imaging is necessary, together with clinical reassessment and follow-up imaging, to improve diagnostic accuracy in parkinsonian diseases .
The aim of this study was to assess the diagnostic contribution of the combined results of pre- and postsynaptic imaging of dopamine neurons in a community-based, dopa-naïve incidence study population with early idiopathic parkinsonism. Our focus was on the possible differences between PD and the APSs, and the motor subtypes of parkinsonism, i.e. postural instability and gait difficulty (PIGD) vs tremor-dominant (TD) parkinsonism, as well as on possible gender differences.
The study was approved by the Ethics and the Radiation Safety Committees of the University Hospital of Norrland in Umeå, Sweden. All patients and healthy control subjects gave spoken and written informed consent before inclusion.
Materials and methods
During 2004 through 2008, patients with idiopathic parkinsonism were identified for a prospective population-based study. The catchment area consists of the city of Umeå with surrounding municipalities with approximately 142,000 inhabitants. All physicians in the catchment area were, during the investigation period, repeatedly requested for referral of all suspected cases with parkinsonism to the Department of Neurology at Umeå University Hospital, which is the only neurological department in the catchment area. All referral letters to the Department of Neurology were screened for eligibility. All referred patients with suspected parkinsonism underwent a standardized clinical examination by an experienced neurologist, specialized in movement disorders. Patients were then included if they fulfilled the clinical criteria for parkinsonism according to the UK Parkinson’s Disease Society Brain Bank (UK PDSBB) criteria, i.e. bradykinesia plus one other feature (rigidity, tremor or postural instability). Patients were thoroughly examined for subclassification of different forms of parkinsonism by two movement disorder specialists independently applying solely the established clinical criteria for PD , possible or probable MSA (using the Gilman criteria for MSA from 1999 where imaging is not included) , possible or probable PSP , CBD  and possible or probable dementia with Lewy bodies (DLB) . A wider definition of PD was also used: parkinsonism according to UK PDSBB plus one or two supporting criteria and no exclusion criteria (i.e. probable PD). When none of the criteria for the specific forms of parkinsonism were fulfilled patients were categorized as having unclassifiable parkinsonism. Diagnosis at baseline and at follow-up after 1 year was made independently of results of dopamine SPECT investigations. The clinical assessment and diagnosis was performed prior to, and independently of the SPECT investigations, both at baseline and at follow-up, avoiding clinical diagnoses being biased by the results of the complementary SPECT results.
Patients were also classified according to the Unified Parkinson’s Disease Rating Scale motor score (UPDRS III) and the Hoehn and Yahr (H&Y) severity scale . Patients with signs of cognitive decline defined as Mini-Mental State Examination (MMSE) < 24  and patients with secondary parkinsonism, e.g. vascular parkinsonism (VaPD), DIP), were excluded.
During the study period, 136 patients (76 men and 60 women) were reassessed by clinical examination at follow-up after 1 year. At follow-up, 111 patients had PD (definitive PD or probable PD), 7 patients had MSA (probable MSA or possible MSA) and 3 patients had PSP (probable PSP or possible PSP). Seven cases were unclassifiable, and eight patients fulfilled the diagnostic criteria for more than one disease.
At the time of inclusion, prior to the start of any pharmacological treatment affecting dopaminergic neurons, 128 patients (72 men and 56 women), which is the study group evaluated in this paper, were examined at baseline with 123I-FP-CIT SPECT and 120 of these (68 men and 52 women) were also examined with 123I-IBZM SPECT.
Number of patients with pre- and postsynaptic 123I-FP-CIT SPECT, by diagnoses and cameras
12-month clinical diagnosis
Presynaptic 123I-FP-CIT SPECT
Postsynaptic 123I-FP-CIT SPECT
Total cases with baseline SPECT (all cases)
Total cases with baseline SPECT (all cases)
Definite PD/probable MSA
Definite PD/probable MSA
Probable PD/possible MSA
Probable PD/possible PSP
Demographics and clinical presentation of patients with baseline presynaptic 123I-FP-CIT SPECT by diagnosis at 12-month clinical follow-up reassessment
Patients with baseline presynaptic 123I-FP-CIT SPECT and 12-month clinical reassessment, by diagnosis
Age at onset (years)
Mean (± 1 SD)
Symptom duration (months)
Mean (± 1 SD)
Age at baseline presynaptic scan (years)
Mean (± 1 SD)
UPDRS III subtotal
Hoehn & Yahr stage
Classification of the PD subtypes was made according to Jankovic et al. , i.e. PIGD, TD or indeterminate (IND).
Of the 128 patients included, 58.6% (75/128) were of the PIGD motor subtype, 30.5% (39/128) were TD and 10.9% (14/128) were of the IND type at baseline.
In PD patients (PD and probable PD), 53% (55/104) were PIGD and in APS patients 78% (14/18) were PIGD.
In men with PD (PD and probable PD), 50% (29/58) were PIGD, 34.5% (20/58) TD and 15.5% (9/58) IND. Among women with PD (PD and probable PD) 56.5% (26/46) were PIGD, 34.8% (16/46) TD and 8.7% (4/46) IND.
Forty-eight healthy controls without signs of parkinsonism or any other neurological disorder participated. Thirty-one controls, age- and sex-matched to the first 50 patients included (17 men and 14 women), with a mean age of 68.2 ± 6.6 (years ± 1 SD) years, were recruited through announcement in a local newspaper. Seventeen male controls with a mean age of 58.7 ± 4.8 (years ± 1 SD) years were recruited from an ongoing prospective research project on cognitive functions; of these, 16 controls performed only IBZM.
Cameras and image reconstruction
Due to equipment replacement, the SPECT data in this study were acquired using two different gamma cameras at the Umeå University Hospital, Umeå, Sweden: (1) a triple-head gamma camera, Neurocam (General Electric, Milwaukee, WI, USA), which is dedicated for brain SPECT imaging and (2) a dual-head hybrid system, Infinia Hawkeye (Infinia Hwk) (General Electric, Milwaukee, WI, USA), which is a multipurpose camera equipped with a low-dose CT (“Hawkeye”). Both cameras were equipped with low energy general purpose (LEGP) collimators, which gave a resolution (i.e. full-width at half-maximum, FWHM) of 8.5 mm (Neurocam) and 9.0 mm (Infinia), at a distance of 10.0 cm from the collimator surface. For both devices, a 20% energy window was centred at the 123I photon energy of 159 keV.
The acquisition protocols and reconstruction methods were as follows:
The images were acquired in 128 × 128 image matrices with a pixel size of 2.00 × 2.00 mm. A 360° stepwise rotation, with 45 s acquisition time in 128 equally spaced angles, was performed with a fixed rotation radius of 12.2 cm. The images were reconstructed with the method that traditionally has been used for this equipment, i.e. the filtered backprojection technique (FBP) without attenuation correction. A Butterworth pre-filter (cut-off 0.45 cm−1 power 7) was used for reconstruction.
The images were acquired in 128 × 128 image matrices using a zoom factor of 1.5, which resulted in a pixel size of 2.95 × 2.95 mm. A 360° stepwise rotation, with 30 s acquisition time in 120 equally spaced angles, was performed with a rotation radius of about 15 cm. The low-dose CT images were used for calculation of attenuation maps for attenuation correction. The images were reconstructed with ordered subset expectation maximization (OSEM) reconstruction, using 2 iterations and 10 subgroups , with attenuation correction and scatter correction using the triple energy window method . A Butterworth post-filter (cut-off 0.45 cm−1, power 8) was used as a post-filter for the OSEM reconstructions. This method is typically used for this system.
Imaging and measurement
Subjects received 200 mg potassium perchlorate po before and after scanning, for thyroid protection. All medications that could interact with the ligand (both 123I-IBZM and 123I-FP-CIT) were discontinued 6 T1/2 before scanning.
Presynaptic dopamine transporter SPECT
For presynaptic dopamine transporter SPECT, 123I-FP-CIT was used. An average bolus dose of 187 MBq (range: 168–208 MBq) was administered intravenously, and the SPECT study was performed 3 h post-injection. Scanning time was approximately 35 min on both cameras. The effective dose was estimated to be 4.3 mSv for 185 MBq injected activity.
Postsynaptic D2 receptor SPECT
For postsynaptic D2 receptor SPECT, 123I-IBZM 3.4 × 10−9 g/ml was used. The substance was delivered ready to inject, with a radiochemical purity of >95% and a specific activity >74 GBq/μmol in a concentration of 74 MBq/ml. An average bolus dose of 185 MBq 123I-IBZM (range: 167–194 MBq) was administered intravenously and the SPECT was performed 1.5 h post-injection. Scanning time was approximately 35 min on both cameras. The effective dose was estimated to be 6.3 mSv for 185 MBq injected activity.
Measurement of pre- and postsynaptic uptake
Image analysis was made with the evaluating physicians blinded to all clinical information and status of the case (patient or control) and was performed by two experienced nuclear medicine specialists (SJM and KR).
Only striatum to background ratios were of interest and therefore it was not necessary to correct for physical decay. The same procedure was used both for pre- and postsynaptic dopamine scans and was as follows:
Imaging of patients and controls
At baseline, 77 patients were examined with 123I-FP-CIT SPECT on the Infinia Hwk and 51 patients were examined on the Neurocam, in total 128 patients. Sixty-nine patients were examined with 123I-IBZM SPECT at baseline on the Infinia Hwk and 51 patients were examined on the Neurocam, in total 120 patients. Two patients were examined with 123I-IBZM SPECT on a different camera than the presynaptic scan. The median interval between pre- and postsynaptic SPECT in patients was 7 days (range: 4–146 days). Four patients had an interval >100 days between pre- and postsynaptic SPECT.
Twenty-six healthy controls were examined with 123I-FP-CIT SPECT on the Infinia Hwk (14 men with a mean age of 63.5 ± 8.1 years and 12 women with a mean age of 71.5 ± 3.0; age difference, p = 0.004) and 12 controls were examined with 123I-FP-CIT SPECT on the Neurocam (6 men with a mean age of 70.8 ± 2.5 years and 6 women with a mean age of 70.0 ± 3.7 years). Of these, seven healthy controls were examined with 123I-FP-CIT SPECT on both cameras the same day. Forty-two healthy controls were examined with 123I-IBZM SPECT on the Infinia Hwk (30 men and 12 women) and 21 controls were examined on the Neurocam (9 men and 12 women). Of these, 15 controls were examined with 123I-IBZM SPECT on both cameras. Half of the group of controls that were scanned on both cameras were examined with the triple-head system first and subsequently with the dual-head hybrid system; for the other controls this scanning procedure was reversed.
The statistical analysis was made on a personal computer using SPSS 17.0 for Windows (SPSS Inc., Chicago, IL, USA). The significance level was set at p values < 0.05. The one-way analysis of variance (ANOVA) routine, an extension of the two-sample t test, was used for comparison of means; Bonferroni’s post hoc test was used for multiple comparisons.
The Mann-Whitney U test was used to compare medians of non-normal distributions and the chi-square test for proportions. The Kruskal-Wallis H test, an extension of the Mann-Whitney U test, was used as the non-parametric analogue to the ANOVA test. Correlation analysis was made using Pearson’s correlation coefficient or Spearman’s rho as a measure of linear association between two variables. In the ROC analyses, cut-off values for computation of likelihood ratios and measures of accuracy for 123I-FP-CIT were derived from the ROC analysis.
In patients, calculations were performed using the uptake ratios in the hemisphere with the lowest uptake, defined as the most affected side. Since the patients included were early in the course of parkinsonian disease, it was reasonable to assume that the hemisphere with the lowest uptake in the putamen can be defined as the most affected side, which would be true even if the patient experiences symptoms from both sides of the body. Especially if the patient has predominantly rigidity and gait disturbances and not as much tremor, it can be difficult to define which side is the most affected in the brain only by clinical findings or the patient’s own reporting. Thus, by using only the uptake ratios in the hemisphere with the lowest uptake in the putamen, we avoided including normal or subnormal uptake ratios in the not yet as affected or symptomatic striatum, since the patients included were early in the course of parkinsonian disease and the disease is predominantly unilateral.
For statistical analysis of postsynaptic 123I-IBZM uptake ratios in patients, calculations were performed using the striatal 123I-IBZM uptake ratio in the same hemisphere as the lowest putamen ratio of the 123I-FP-CIT uptake (i.e. the most affected side). By defining the most affected side as above, it was rational to study the corresponding postsynaptic uptake to investigate possible upregulation or depletion of D2 receptors at that side (the “mirror effect”), which is hypothesized to discriminate between PD and atypical parkinsonism.
In healthy controls, means of the right and left uptake ratios were used in calculations for both pre- and postsynaptic SPECT to compensate for normal side-to-side asymmetry.
Presynaptic dopamine transporter SPECT
Sensitivity and specificity of 123I-FP-CIT SPECT in patients with parkinsonism, by SPECT camera. Area under the curve data; cut-off values derived from ROC analysis
123I-FP-CIT SPECT on Neurocam
123I-FP-CIT SPECT on Infinia Hwk
Healthy controls, n = 12
Healthy controls, n = 26
Patients, n = 51
Patients, n = 77
Positive likelihood ratio
Negative likelihood ratio
Presynaptic 123I-FP-CIT uptake ratios by diagnosis and SPECT camera
± 1 SD
± 1 SD
± 1 SD
Definitive and probable PD
Definitive and probable PD
Age and disease severity
Presynaptic uptake in the putamen, caudate and the whole striatum did not significantly correlate to disease duration or age at baseline in PD patients. The UPDRS III score was negatively associated with presynaptic caudate and putamen uptake in PD patients (Spearman’s rho; Infinia: caudate −0.326, p = 0.016, putamen −0.289, p = 0.034; Neurocam: caudate −0.360, p = 0.028, putamen −0.359, p = 0.029).
The median MMSE score in the PD group was 29, ranging between 24 and 30 (3 cases with less than 26; no data in 7 cases) without any significant differences between PIGD-PD and TD-PD. There were no statistical correlations between MMSE and presynaptic uptake.
Disease severity in patients with definitive and probable PD by motor subtype
PDa patients with baseline scan (n = 104)
UPDRS III subtotal score
Hoehn & Yahr stage
There were no significant differences in presynaptic uptake ratios between men and women with PIGD-PD. Female patients with TD-PD were older at disease onset than male patients (p = 0.04). There was no statistical difference in presynaptic uptake between genders, either in the control group or in the patient group; however, mean uptake ratios in all presynaptic compartments tended to be slightly higher in female TD-PD patients and in female healthy controls. Disease duration and UPDRS III score were equal between genders. There was a weak positive association between age at disease onset and disease severity as measured by UPDRS III score that was more pronounced in women with PD than men (Spearman’s rho 0.320, p = 0.017; female linear r2 = 0.28, male linear r2 = 0.09).
Postsynaptic D2 receptor SPECT
There were no significant differences between patients and controls in postsynaptic uptake ratios on the same side as the lowest putamen uptake ratio. The postsynaptic ratios in the dopa-naïve PD patients in this study were not generally higher than in healthy controls. Nor did 123I-IBZM ratios differ significantly between PD and APS. Postsynaptic uptake ratios were not related to age at baseline postsynaptic SPECT or age at disease onset. It was not associated with disease severity, as measured by UPDRS III subscore or H&Y stage. Postsynaptic uptake ratios did not differ significantly between PD motor subtypes or gender.
Combined pre- and postsynaptic SPECT
In this study, presynaptic dopamine transporter SPECT showed high accuracy for excluding APS at baseline. It is worth noting that caudate uptake ratios were affected in a great majority of patients with idiopathic parkinsonian syndromes, already at baseline. Low presynaptic function was, as reported in other studies, associated with severity of symptoms [27, 28], but not related to symptom duration or patient age. Reduced caudate uptake ratios have previously been described for subgroups in early PD populations, but usually emphasis is put on the pathology in the putamen. Studies with a focus on PD motor subgroups suggest a significantly lower presynaptic uptake in both the caudate nucleus and the putamen in PIGD-PD compared to TD-PD . We found caudate ratios on the same side as the lowest putamen uptake ratio significantly reduced in the majority of the newly diagnosed patients with parkinsonian disease. A large proportion of the patients were of the PIGD type. However, also many TD-PD had significantly reduced caudate uptake.
The presynaptic uptake ratios were not significantly different between PD and APS in this study, but the number of patients with APS was limited, hence the statistical power is low.
The postsynaptic ratios in the dopa-naïve PD patients in this study were not generally significantly higher than in healthy controls, in concordance with the findings of Schwarz et al. .
Patients with PIGD-PD had lower presynaptic ratios and were more affected by the disease than TD-PD, as indicated by higher UPDRS III subtotal score and H&Y stage. In our study, PD patients with PIGD and TD motor subtype differ in disease severity but not in age or symptom duration. This could indicate that the PIGD subtype PD at early onset represents a more aggressive form with more pronounced affection of the dopamine system.
It is known that PIGD-PD patients are at higher risk of developing dementia . Cognitive status, measured by MMSE was, however, neither correlated with pre- and postsynaptic striatal dopaminergic function nor with PD motor subtype or gender in our study. No significant decrease in the MMSE scores was seen in PD cases compared to healthy controls. In a recent study from Norway, the proportion of mild cognitive impairment was nearly 20% of the newly diagnosed PD patients .
Even though the APS group is small, our results indicate that the combination of pre- and postsynaptic dopamine SPECT can serve as an indicator for excluding APS in a parkinsonian group of early diagnosed, dopa-naïve idiopathic parkinsonian patients with a reasonably high accuracy of 85%. Half of the patients with the 12-month follow-up clinical diagnosis of APS revealed a combination of severely reduced both pre- and postsynaptic uptake on the same side already at the baseline scans. With longer follow-up, also patients with mixed diagnoses due to overlapping clinical criteria will probably develop a clinical picture which allows more precise clinical diagnoses . The development of the dopamine SPECT findings in these cases will be further investigated in forthcoming studies.
There was a tendency, however not statistically significant, towards higher age at onset at baseline and towards higher presynaptic uptake ratios in women with TD-PD. This is in agreement with a study from The Netherlands . As opposed to that, PIGD-PD was more common than TD-PD at baseline in women in our study.
The use of two different cameras in the study is a limitation, since semi-quantitative uptake ratios differed significantly between these two cameras which rendered relatively small statistical groups on each camera. However, our results were consistent between the two cameras. It is clear that the standard deviations within groups are higher for the image data acquired on Infinia Hwk. There may be several reasons for this, which have to do with image acquisition and reconstruction but also the ROI method used for semi-quantification. The Infinia Hwk reconstructions have an overall higher quantitative accuracy, because of attenuation and scatter correction, and it is therefore likely that the higher standard deviations reflect true individual variations, at least to some degree. This assumption is supported by the findings of Berti et al. . Image reconstruction with OSEM is also described as giving more reproducible results in 123I-IBZM than image reconstruction with FBP . It should however be noted that the higher noise level and somewhat lower resolution in the Infinia Hwk reconstructions also may contribute to the increased standard deviations. The higher noise level, caused by scatter correction and the difference in sensitivity due to the lower number of camera detectors, may have an influence on the statistics, especially in the background ROI.
The study population was community based and therefore expected to be representative of the general population with idiopathic parkinsonism, which adds strength to the study. Furthermore, the clinical assessments were made without taking into account auxiliary tests or imaging results by two experienced specialists in movement disorders. Moreover, the semi-quantitative analysis of the patients and the healthy controls was performed blinded to the physician who performed the quantification of the SPECT studies to avoid bias. The ROI method for quantification is clearly a method, which demands carefulness and constancy, and the SPECT analysis was made by two specialists in nuclear medicine and diagnostic radiology with long experience in this method.
In conclusion, presynaptic SPECT uptake was significantly reduced in newly diagnosed dopa-naïve patients with idiopathic parkinsonism, and in the majority of cases, even the caudate nucleus was affected already in this early phase of the disease.
There were no statistically significant differences in dopamine SPECT uptake, disease severity or age at onset between female and male PD patients in this study.
There was no significant association between decreased caudate presynaptic uptake and MMSE results at baseline in this study. In further studies, the prognostic importance of low caudate SPECT uptake ratios in relation to the performance in other cognitive tests will be evaluated.
Differences were found in presynaptic uptake between PD motor types which might indicate differences in prognosis.
Finally, although it has to be confirmed by follow-up and other studies, our study shows that not the 123I-IBZM SPECT alone, but the pattern of combined pre- and postsynaptic dopamine SPECT on the most affected side in dopa-naïve patients with idiopathic parkinsonism can contribute in excluding APS already in an early phase of the disease.
This study was supported by grants from The Swedish Medical Research Council, The Parkinson Foundation in Sweden, The Swedish Association of Persons with Neurological Disabilities, The University of Umeå, the Foundation for Clinical Neuroscience at Umeå University Hospital, Västerbotten County Council (ALF) and King Gustaf V’s and Queen Victoria’s foundation.
Conflicts of interest