Dopamine Transporter Imaging

Fig. 11.1

Nigrostriatal dopaminergic synapse and targets of PET and SPECT ligands. Tyrosine is converted to L-DOPA by tyrosine hydroxylase. L-DOPA is then converted to dopamine, which is gathered into vesicles by vesicular monoamine transporter and from which it is released into the synaptic cleft. Dopamine binds to D1 and D2 receptors on the postsynaptic membrane. Some dopamine re-enters the presynaptic terminal via dopamine transporters (DaTs) (Reproduced with permission from Cambridge University Press [6])

The radionuclide attached to the ligand emits radiation which can be detected using PET or SPECT. SPECT uses gamma rays to provide 3D information from 2D images. For PET radionuclides emit positrons which travel a short distance before hitting an electron, producing a pair of photons moving in opposite directions. The detection of these photons enables production of 3D images. PET produces higher resolution images than SPECT. Manipulation of scanning parameters enables slices to be taken along any axis of the brain.

The active ligands are labelled by radionuclides with different half-lives. The most practical radionuclide is ¹²³Iodine which has a half-life of 13 h. Examples of ligands labelled with ¹²³Iodine include 2β-carbomethoxy-3β-(4-iodophenyl)tropane (β-CIT) and n-fluoropropyl-2β-carbomethoxy-3β-(4-iodophenyl)nortropane (FP-CIT).

The ligands used in PET scanning have shorter half-lives which pose more practical difficulties. SPECT scanning is more widely available and cheaper than PET. Ligands used with PET can be labelled with ¹⁸Fluorine (half-life 110 min) or ¹¹Carbon (half-life 20 min).

The most researched ligand in DLB is FP-CIT which is used with SPECT. FP-CIT has high affinity and relatively good selectivity for DaT with the optimal scanning window being 3–6 h after injection of the ligand.

11.4 Interpretation of Dopaminergic Imaging

Interpretation of FP-CIT scans is normally done with visual assessment (see Fig. 11.2). Benamer et al. [11] found visual inspection was sufficient in patients with normal scans or those typical of dopamine degeneration. However, visual interpretation of scans is subjective, and trials tend to use highly experienced nuclear physicians and are therefore likely to provide greater accuracy than might be achieved in ordinary clinical settings. In patients with marginal or difficult to interpret scans, a semiquantitative analysis can be helpful; this can yield a more objective interpretation of results. In clinical practice regions of interest with predefined templates of the striatum and nonspecific binding regions are commonly used. More recently semi-automated quantification tools have been developed, and there are several programmes able to quantify DaT studies. Söderland et al. [12] performed visual and semiquantitative analysis of FP-CIT scans. Although visual interpretation of scans provided good interobserver agreement (mean k ± SD, 0.80 ± 0.05), this was improved as striatal binding ratios (0.86 ± 0.07), and caudate-to-putamen ratios (0.95 ± 0.04) were provided to the raters. The most experienced rater in the study made only minor changes in diagnosis after being provided with semiquantitative data in a small number of patients with more complex scans. Less experienced raters more frequently changed their diagnosis with additional information, and the updated diagnoses were in better agreement with the experienced rater. In clinical practice nuclear medicine physicians are likely to be less experienced than those used in research studies. This highlights that less experienced readers can perform as effectively as more experienced readers if provided with semiquantitative information.

Fig. 11.2

Example of visual scan rating categories. Categories used here are normal (normal uptake in all regions); abnormal type 1 (asymmetric activity with one putamen showing reduced uptake); type 2 (absent activity in the putamen of both hemispheres); and type 3 (absent activity in the putamen of both hemispheres and greatly reduced in one or both caudate nuclei). Reproduced with permission from the Royal College of Psychiatrists [13]

11.5 Dopamine Transporter Imaging

Studies using DaT imaging have focused on a number of research questions. These include evaluating the accuracy of DLB diagnosis with and without the support of imaging and against autopsy diagnosis, imaging to monitor disease progression, comparison of dopaminergic imaging with other imaging techniques, evaluating the performance of different ligands and how imaging findings correlate with clinical symptoms and signs. We will discuss these in some detail below.

11.6 DLB Diagnosis Using Dopaminergic Imaging

11.6.1 DLB Compared to Normal and AD

The main studies comparing FP-CIT imaging in DLB and AD are summarised in Table 11.1.

Table 11.1

Studies comparing DaT imaging results between DLB and AD

Paper	Ligand	Patients	Aim	Sensitivity of FP-CIT scan	Specificity of FP-CIT scan	Other information
Walker et al. [14]	FP-CIT	27 DLB	Distinguish DLB from AD	88 % in subsample with autopsy	100 % in subsample with autopsy	Lower striatal ligand uptake in DLB and PD compared to AD and controls
		17 AD
		19 early PD
		16 controls
Ceravolo et al. [17]	FP-CIT	13 AD-P	To investigate DaT loss in AD patients with parkinsonism (AD-P) compared to DLB and controls	–	–	FP-CIT striatal uptake in patients with AD-P was similar to controls. DLB and PD had lower uptake in all striatal areas when compared to AD-P and controls
		15 DLB
		20 PD
		8 controls
O’Brien et al. [15]	FP-CIT	23 DLB	To investigate DaT loss in DLB compared to AD, PD, PDD and controls	78 %	94 %	Striatal binding reduced in DLB, PD and PDD compared with AD and controls. Relatively greater caudate loss in DLB and PDD compared with PD
		36 PDD
		34 AD
		38 PD
		33 controls
McKeith et al. [18]	FP-CIT	151 DLB	Investigate the sensitivity and specificity of DaT against consensus diagnosis	78 %	90 %	Overall diagnostic accuracy 86 %. Positive predictive value 82 %. Negative predictive value 88 %
						Inter-reader agreement k = 0.87
						Possible DLB 38 % had an abnormal scan
		147 non-DLB dementia (mainly AD)				Possible DLB 38 % had an abnormal scan
Walker et al. [20]	FP-CIT	8 DLB	Determine sensitivity and specificity of DaT against autopsy	88 % visual rating	83 % visual rating	FP-CIT good sensitivity and very good specificity
		12 non-DLB			83 % visual rating
					100 % semiquantitative
				88 % semiquantitative	100 % semiquantitative
O’Brien et al. [19]	FP-CIT	44 possible DLB	Accuracy of FP-CIT in diagnosis of possible DLB	63 %	100 %	12/19 cases of probable DLB at follow-up had abnormal baseline scan. 7/7 cases of AD at follow-up had normal scan
Walker et al. [13]	FP-CIT	114 possible DLB	To investigate whether performing a scan in patients with possible DLB would aid diagnosis	–	–	Use of the scans allowed the physicians to be more certain about their diagnosis. A positive scan was more likely to lead to a change in diagnosis

FP-CIT n-fluoropropyl-2β-carbomethoxy-3β-(4-iodophenyl)nortropane, DLB dementia with Lewy bodies, AD Alzheimer’s disease, PD Parkinson’s disease, DaT dopamine transporter, PDD Parkinson’s disease dementia

Early single-centre studies showed good sensitivity and specificity for FP-CIT compared to clinical diagnosis. Walker et al. [14] scanned patients with DLB, AD, PD and controls using FP-CIT. Patients with DLB had reduced binding in both the posterior and anterior putamen and in the caudate nucleus when compared with AD and controls.

O’Brien et al. [15] extended a similar methodology to a larger cohort of 164 participants including PDD cases. As in Walker et al. [14], DLB subjects had significantly reduced FP-CIT binding in the caudate and anterior and posterior putamen compared to AD and controls. DaT loss in DLB was similar to PD, but with a less prominent caudate-putamen gradient. Patients with PDD had the greatest DaT loss in all striatal areas. The less pronounced caudate-putamen gradient in DLB was also observed by Walker et al. [16]. In that study the main emphasis was on comparison of DLB and PD cases.

Ceravolo et al. [17] examined patients with AD and extrapyramidal features using FP-CIT and compared them to patients with DLB and PD and controls. The AD patients showed similar results to control subjects in the putamen and caudate nucleus despite their extrapyramidal signs. There was, as expected, significant differences between DLB and AD in DaT binding across all the striatal areas.

McKeith et al. [18] completed a multicentre, multinational phase-III study using FP-CIT to examine 326 patients with diagnoses of probable or possible DLB and compared them to cases of non-DLB dementia. Clinical diagnosis was determined using consensus panel methodology (consensus of three independent clinical experts). The scans were visually rated by three nuclear medicine physicians as normal or abnormal. The mean sensitivity of FP-CIT imaging for a diagnosis of probable DLB was 78 % and mean specificity was 90 %. The mean overall diagnostic accuracy was 86 %, with 82 % for positive predictive value and 88 % for negative predictive value. Inter-reader agreement for rating scans was high (k = 0.87).

In a 1-year follow-up of the McKeith et al. [18] cohort, of the 44 patients given an initial diagnosis of possible DLB, 18 had no change in diagnosis, 19 converted to probable DLB and seven were diagnosed with AD. Of the 19 who converted to probable DLB, 12 had an abnormal scan at baseline and all seven patients that were assigned a diagnosis of AD at follow-up had a normal baseline scan. This indicated that FP-CIT result was able to point towards diagnosis at follow-up at a time when clinicians could only assign a possible DLB diagnosis [19].

The performance of the FP-CIT scan to facilitate an earlier diagnosis was further confirmed in a study with autopsy diagnosis. Walker et al. [20] found that reduced striatal uptake of FP-CIT had 88 % sensitivity and 100 % specificity when making a diagnosis of DLB compared to autopsy diagnosis (the gold standard). In comparison, an early clinical diagnosis without imaging had a sensitivity of 75 % and specificity of 42 %, indicating that the FP-CIT imaging facilitated a more accurate diagnosis. In an extension of this cohort, a modified analysis was used, whereby an abnormal scan was defined as uptake below 2 standard deviations of the mean of controls in the worse affected posterior putamen. Using this technique in the larger cohort showed the sensitivity of FP-CIT SPECT scan for diagnosing DLB to be 100 % and the specificity 92 % [21].

A recent phase-IV multinational study [13] randomised patients with possible DLB diagnoses (difficult to diagnose cases) to FP-CIT imaging or no imaging. Of the 170 patients, 114 were randomised to imaging and 56 were controls. Of the 114 patients with possible DLB, 43 % had an abnormal scan. More patients in the imaging group had a change to a more certain diagnosis (probable DLB or non-DLB) compared with controls at 24 weeks (71 % vs. 16 %). Interestingly, clinicians were much more likely to change the diagnosis if the scan was abnormal (82 %) than normal (46 %).

11.6.2 DLB Compared to PD and PDD

DLB, PD and PDD all feature nigrostriatal degeneration. However, there are subtle differences between PD/PDD and DLB on FP-CIT imaging, giving us some insight into the distribution of underlying pathology. However, in clinical practice FP-CIT imaging is never used for differential diagnosis between DLB and PD/PDD as this is always done on the basis of clinical and cognitive assessment.

Ransmayr et al. [22] showed DaT binding was lower in DLB patients compared to those with PD. The authors also showed that PD patients had a more asymmetrical uptake of β-CIT than DLB cases. O’Brien et al. [15] showed greater caudate involvement in DLB and PDD when compared to PD.

Walker et al. [16] compared both DLB and PD patients to controls. They found a difference in both groups in the binding in the caudate and putamen bilaterally relative to controls. The PD group had a more marked gradient of uptake than the DLB group. DLB showed a uniform decrease in the uptake of dopamine in the caudate and anterior and posterior putamen. PD patients had less severe loss in the caudate but increased loss in the putamen, particularly the posterior putamen contralateral to the most affected side. This resulted in marked asymmetry in posterior putaminal uptake compared to DLB and controls and fits with the clinical picture of more asymmetrical motor symptoms in PD than DLB.

Rossi et al. [23] investigated striatal DaT uptake in 30 patients with DLB and 30 with PDD using FP-CIT. They found that the striatal uptake of the DaT tracer was not significantly different between PDD and DLB.

Only gold members can continue reading. Log In or Register to continue