Prospective F-18 FDOPA PET Imaging Study in Human PD

Abstract

Purpose

We present the findings of our final prospective study submitted to the U.S. Food and Drug Administration (FDA) for New Drug Application (NDA) approval for the use of 3,4-dihydroxy-6-[18F]fluoro-l-phenylalanine (F-18 FDOPA) positron emission tomography (PET) imaging for Parkinson’s disease (PD). The primary aim was to determine the sensitivity, specificity, and predictive values of F-18 FDOPA PET in parkinsonian patients with respect to clinical standard-of-truth (SOT). Secondary outcomes included the inter-rater reliability, and correlation of quantitative measures for PET with dopaminergic status.

Methods

In 68 parkinsonian subjects, F-18 FDOPA PET scan from 80 to 100 min was acquired following a CT scan. Scan images were presented to one expert in F-18 FDOPA image interpretation and two physicians with prior experience in I-123 FPCIT single-photon emission computed tomography image interpretation. Fifty-six subjects completed the study with a follow-up for SOT determination. Image readers were blind to the clinical/quantitative data; SOT clinician was blind to the image data.

Results

For 47 of the 56 patients, SOT was in agreement with the PET scan results. For nine patients, SOT suggested dopaminergic deficit, whereas the imaging showed normal uptake. The specificity and positive predictive values are 91% and 92%, respectively, suggesting high probability that those who test positive by the PET scan truly have dopaminergic degeneration. The sensitivity was 73%. Inter-rater agreement was 0.6–0.8 between the different readers.

Conclusion

Our prospective study demonstrates high specificity and moderate sensitivity of F-18 FDOPA PET for PD. We received NDA approval in October 2019.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13139-022-00748-4.

Introduction

The latest estimates of the number of people living with Parkinson’s disease (PD) in the USA is 1 million, and approximately 60,000 Americans are diagnosed with PD each year. Accurate and timely diagnosis of PD is essential for appropriate therapy initiation as well as preventing unnecessary and possibly harmful treatments when symptoms mimic PD but are caused by other underlying conditions. Recently, some prominent clinical trials in the Parkinson’s disease field have used clinical confirmation by a movement disorders specialist after 6 months or more of follow-up which includes Unified Parkinson’s Disease Rating Scale (UPDRS) assessment for motor and non-motor deficits [1, 2]. There are a number of radiotracers for single-photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging that have been developed and two of the widely used radiopharmaceuticals that are now approved in Europe and Asia for clinical imaging are N-3-fluoropropyl-2-(R)-carboxymethoxy-3-(R)-(4-[123I]iodophenyl) nortropane (I-123 FPCIT, DaTscan) and 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine (F-18 FDOPA). Numerous studies demonstrating the high sensitivity and specificity of these two radiotracers have been published [3, 4], including a review article with extensive bibliography about the historical use of F-18 FDOPA [5]. In addition to qualitative image interpretation, simplified quantitative approaches involving short imaging times also have been developed; these approaches increase patient compliance, reduce motion artefacts, and provide increased confidence levels in visual image interpretation [6, 7]. A high correlation has been demonstrated between F-18 FDOPA uptake rate constants and dopamine cell counts and levels [8]. Guidelines for the use of dopaminergic radiotracers in Parkinson’s disease have been periodically published, the last one being by Morbelli in 2020 [9].

While F-18 FDOPA PET has been used in the USA since the 1980s in clinical research studies without full approval and Medicare reimbursement, this radiotracer was approved for clinical applications in Europe and many other places in the world by 2010. Compared to DaTscan SPECT imaging, F-18 FDOPA PET offers significant advantages including improved resolution, lower radiation burden, shorter imaging sessions, no pre-scan administration of Lugol’s solution, and similar or lower cost. Although a large body of published work exists on the use of F-18 FDOPA in PD, the lack of prospective studies in parkinsonian disorders has prevented the approval of this tracer for reimbursement in the USA. We initiated the process of obtaining a U.S. Food and Drug Administration (FDA) New Drug Application (NDA) for F-18 FDOPA with our first application submission in 2012 (study #15) that included published data as well as retrospective data on 185 PD subjects. As per feedback from FDA, we submitted our second application in 2014 (study #17) that included a prospective trial of F-18 FDOPA in 53 PD subjects. The pertinent details of these two studies are available in the FDA document [10]. A third and final application (study #18) was submitted to the FDA in 2019, the results of which are presented in this manuscript. The following aims were addressed:

1. To determine the sensitivity, specificity, and predictive values of F-18 FDOPA PET in patients with at least one sign of parkinsonism, with respect to clinical diagnosis at 6 months to one year after the scan.

2. To determine the inter-rater reliability of the three PET readers.

3. To determine how quantitative measures for PET correlate with dopaminergic degeneration.

4. To monitor and evaluate safety in patients undergoing F-18 FDOPA PET scans.

Based on all three submissions, FDA issued an NDA to us in October 2019 (No. 200655). We are currently pursuing reimbursement for F-18 FDOPA PET imaging.

Materials and Methods

Subject Population

We enrolled 69 patients with suspected parkinsonism (over 2 years from 2016 and 2018) with a target of at least 50 subjects to complete PET imaging and follow-up clinical evaluation within the specified time window (6–12 months). One subject did not complete PET imaging after declining to reschedule imaging appointment following radiotracer synthesis failure. Subject age was 65.1 ± 13.2 (mean ± standard deviation) years with disease duration of 35.5 ± 38.3 months, baseline disease severity was Hoehn and Yahr (H&Y) stages 1–3 (1.7 ± 0.9), and UPDRS Part III score range of 20–35 (14.3 ± 11.7). The majority of subjects (80%) reported having PD symptoms for less than 5 years, though a few subjects had long duration parkinsonism with uncertain diagnosis (see Appendix Table 3 and Online Resource 1). For quantitative analysis, data from 15 normal control subjects from the first study (study #15) were utilized. (See Online Resource 2 for a summary of the NDA application process).

Table 3.

A summary of the baseline variables for clinical study #18

Subject ID	Age (years)	Sex	Disease duration (months)	Injection dose (MBq)	Mass dose (mg)	Scan to SOT duration (days)	UPDRS part III score
							Baseline	6 months
54	54	M	12	188.7	1.71	225	12	17
55	52	F	6	196.1	1.30	183	8	14
56	77	M	4	199.8	1.32	231	24	28
57	55	F	60	196.1	1.21	207	3	3
59	49	M	36	166.5	1.23	189	49	47
60	67	M	84	199.8	1.67	222	19	19
61	38	M	48	199.8	1.44	266	6	8
62	79	M	12	199.8	1.13	184	33	19
63	78	M	24	199.8	1.18	196	14	4
64	70	M	12	203.5	1.20	215	8	10
66	22	F	36	203.5	1.18	189	18	9
67	65	M	1	203.5	1.26	201	3	5
69	69	M	108	185.0	1.67	186	54	43
72	58	F	8	192.4	1.28	232	15	9
73	67	M	18	196.1	1.31	211	10	2
74	74	M	2	192.4	1.19	233	16	12
75	61	M	6	199.8	1.24	199	28	23
76	56	M	84	203.5	0.98	193	18	38
77	81	F	4	203.5	1.06	189	24	25
79	41	M	36	203.5	1.03	190	31	40
80	90+	F	1	192.4	1.26	207	5	0
81	74	M	36	192.4	1.26	193	7	6
82	68	M	36	199.8	1.21	185	15	18
83	61	M	60	199.8	1.21	226	18	18
84	79	M	24	196.1	1.31	193	23	30
85	61	M	12	192.4	1.30	194	38	31
86	68	M	12	199.8	1.08	189	25	24
87	77	M	24	185.0	1.24	181	8	8
88	49	F	180	196.1	1.31	181	5	2
90	49	M	60	196.1	1.33	182	6	3
91	63	F	12	192.4	1.20	189	28	12
92	65	M	5	196.1	1.31	187	14	24
93	72	M	8	188.7	1.26	210	7	5
94	72	M	60	199.8	1.29	196	14	4
96	89	M	120	177.6	1.05	188	19	13
97	51	F	24	173.9	1.25	221	15	16
98	42	F	15	181.3	1.21	199	3	4
99	74	M	2	181.3	1.10	189	4	0
101	54	M	48	185.0	1.28	189	2	0
102	63	M	24	173.9	1.19	203	5	11
103	53	M	18	181.3	1.01	184	5	2
104	65	M	24	181.3	1.41	191	21	13
106	74	F	96	181.3	1.10	195	8	12
107	86	M	36	181.3	1.31	188	6	10
108	69	M	3	177.6	1.14	203	6	9
109	84	M	12	181.3	1.16	203	7	4
110	65	F	72	177.6	1.15	183	10	14
111	71	F	48	181.3	1.17	201	12	31
112	77	F	48	181.3	1.06	200	7	6
113	73	M	8	173.9	1.02	209	19	11
115	70	F	1	181.3	1.30	184	7	13
116	69	F	6	177.6	1.24	200	12	7
117	70	M	96	185.0	1.29	215	6	4
119	79	F	132	181.3	1.36	184	5	6
120	81	F	9	181.3	1.24	189	15	21
121	52	M	12	177.6	1.33	182	25	34

Out of the 68 patients who completed the PET scan, the first 56 patients who reported after 6 months for follow-up examination were included for further analysis; this met the criteria of a minimum of 50 subjects stipulated by the FDA. The demographics of the subject population and UPDRS motor scores are provided in Appendix Table 3. Concomitant medications as well as drug interaction potential for F-18 FDOPA and effect of the concomitant drugs on F-18 FDOPA PET scan (such as antipsychotic drugs like Haloperidol, Monoamine oxidase inhibitors, Reserpine, Aripiprazole, Dopamine agonist and antagonist drugs) are presented in Online Resource 3 [11–16]. The eligibility criteria were as follows:

Inclusion criteria: (1) All patients will be 18 years of age or older (F-18 FDOPA PET is not indicated for pediatric patient under this protocol); (2) Patients must have a clinical diagnosis exhibiting at least one cardinal parkinsonian symptom (muscle rigidity, tremor, bradykinesia, gait disturbances, postural instability, etc.).

Exclusion criteria: (1) Pregnant or breastfeeding. Urine pregnancy testing was required for all women of childbearing potential prior to PET imaging; (2) Patients with a confirmed diagnosis of essential tremor or a previous F-18 FDOPA PET scan to confirm PD; (3) Patients with clinically confirmed atypical parkinsonism.

Patients were referred for F-18 FDOPA PET imaging by local physicians. This study was a single site study (The Feinstein Institutes for Medical Research) and was performed with Institutional Review Board of Northwell Health approval and in compliance with the HIPAA Privacy Rule. Informed consent was obtained from all individual participants included in the study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 2013 Helsinki declaration and its later amendments or comparable ethical standards. No specific populations have been studied in this clinical study.

Preparation for F-18 FDOPA PET

All subjects undergoing F-18 FDOPA PET were asked to preferably refrain from eating but if not possible, to consume only a light low-protein meal 4 hours prior to the scan. All PD medications were stopped 12 h before the scan. Then, 150 mg of Carbidopa was administered to the patient orally 60–120 min prior to the administration of F-18 FDOPA injection. Approximately 185 MBq (5 mCi) of F-18 FDOPA was administered to the patient (see Appendix Table 3). Adverse event monitoring took place during the PET scan and with a 24–72 h follow-up phone call.

Imaging Protocol

Approximately 70 min post-injection, the subject was asked to void to remove radioactivity collected in the bladder, thus reducing radiation exposure and increasing comfort during the scan. Subjects were then positioned for scanning using a cushioned head-holder. Images were acquired as follows: A 9-s ultra-low-dose CT scan (80 KeV, 10 mA) was performed just before a 3D PET scan from 80 to 100 min. All scans were performed on a GE Discovery IQ scanner (General Electric, Boston, Massachusetts, USA) with 5-mm resolution in all directions and images reconstructed using Q.CLEAR software (a Bayesian penalized likelihood reconstruction algorithm designed for full convergence; β = 200); matrix size 128 × 128. Performance of this scanner has been previously published [17].

Image Processing

Central 6 slices of striatum (usually visible on 8–9 slices) were selected and a single summed slice (19.8 mm thickness) was displayed in a single image display (SID). A threshold (window) was used to remove the bottom 20% of the image pixels to focus on gray matter. This SID image was used for visual reading. Over smoothness of the images results from a composite slice thickness of 19.8 mm which was deemed to visually provide accurate reflection of the dopaminergic functional state of the basal ganglia.

Image Interpretation

We selected three readers blinded to the clinical status of each subject to read the PET scan visually. None of the readers were involved in conducting the PET procedures to acquire the PET images. Two of the non-expert readers (radiologists/nuclear medicine) had no prior experience in reading F-18 FDOPA PET scans, though they had previous experience in reading I-123 FPCIT SPECT (DAT) scans for a few years. The visual reading of the F-18 FDOPA PET scan of these two readers was compared with that of the expert reader who has numerous years of experience in reading F-18 FDOPA PET scans. The justification of using one expert and two non-experts was based upon our previous data that demonstrated a very good agreement within experts in reading F-18 FDOPA PET images ([10], study #17) and, to evaluate its utility in real-world clinical practice. All three blinded readers were provided with a de-identified scan, without any supportive quantitative data, and directed not to consult with anyone for scan interpretation. There was no formal training for the new readers prior to the scan reading. However, they were provided with written guidelines on how to read the F-18 FDOPA PET scan (see below). The research coordinator was responsible for maintaining the integrity of the blinding process and the database.

Binominal Visual Assessment of the F-18 FDOPA PET Scan

Readers were instructed to note any visual defect in the size and shape of putamen and caudate in the PET scan (Fig. 1). A full crescent-shaped putamen and caudate image will indicate a normal scan (−) as shown in the image below. A reduction in the size and shape of putamen (unilaterally or bilaterally) or both putamen and caudate (unilaterally or bilaterally) will indicate an abnormal scan (+). The degeneration starts at the posterior putamen at the earliest stage of the disease, with just one or two clinically observable parkinsonian symptoms when a 30–40% dopaminergic neuronal loss has been demonstrated. F-18 FDOPA uptake in the putamen and caudate reflect the functional availability of dopaminergic neurons. Readers are requested to evaluate the scan by visual examination only to report positive (+) and negative (−) scan results. For this clinical study, image interpretation did not involve staging.

Fig. 1.

F-18 FDOPA staging. Normal: F-18 FDOPA uptake is obviously increased from the background in both caudate and putamen (cashew-like appearance). Stage 1, Mild: F-18 FDOPA uptake is reduced in posterior putamen on one contralateral side but is preserved on both sides in the caudate nuclei. Stage 2, Moderate: F-18 FDOPA uptake is preserved in caudate but reduced in putamen on both ipsilateral and contralateral sides. Stage 3, Severe: F-18 FDOPA uptake is significantly reduced in putamen on both sides and to a much lesser degree in caudate. The images look oversmoothed because of the slice thickness (19.8 mm) resulting from a sum of central 6 slices of 3.3 mm each; matrix size 128 × 128. For reconstruction parameters, see text in Imaging section

Standard of Truth

Clinical examination utilizing the UPDRS Part III was performed by a movement disorders specialist 6–12 months following F-18 FDOPA PET imaging. This physician was blinded to the PET scan results and other clinical information (provided by the treating physician at the time of referral). The final SOT report is based on the best possible clinical diagnosis made by the blinded physician following this examination. The final SOT and the blinded PET scan readings have been used for the computation of sensitivity, specificity, and predictive values.

Primary Endpoint of Imaging

The primary endpoint for this study is the visualization of either the loss or no loss of dopaminergic neurons in the striatum based on the uptake of F-18 FDOPA. Visualization is based on the assessment of the shape and size of the putamen/caudate region. Reduced uptake of F-18 FDOPA in the area of the putamen or both putamen and caudate unilaterally or bilaterally will indicate dopaminergic neuronal loss.

No attempt to differentiate the subdivisions of dopaminergic parkinsonism such as MSA, PSP, and CBD were made either qualitatively or quantitatively based upon the imaging data. The movement disorders expert did note the subdivisions of parkinsonism based upon clinical examination. These subdivisions are provided for reference purposes only (Online Resource 4).

To quantitate the PET scan results for patients with dopaminergic and non-dopaminergic parkinsonism, we used striatal/occipital ratio (SOR) calculations [6]. SOR quantification was not used for the determination of the primary outcome (sensitivity, specificity, PPV, and NPV) but may be used as a complementary tool to assess the PET scan visual reading results.

Statistical Analysis

Descriptive statistics (means and standard deviations or medians and interquartile ranges) for continuous variables were calculated. Frequencies, tabulations, and percentages were calculated for categorical variables.

The primary objective of this study was to estimate sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) using the F-18 FDOPA PET scan read as a diagnostic test and the final SOT as the clinical gold standard for diagnosis.

Exact binomial methods were used to calculate 95% confidence intervals for sensitivity, specificity, PPV, and NPV. (Asymptotic methods were not used due to small sample size.) Estimates of PPV and NPV were computed directly from the applicable 2 × 2 table, based on the assumption that the observed prevalence rate of dopaminergic degeneration in the sample is an unbiased estimate of the prevalence rate in the referral population.

Inter-rater reliability analysis to determine the level of agreement of three PET readers was performed using the kappa statistic to determine consistency between raters. The data analysis for this study was generated using SAS software (SAS Institute Inc., Cary, NC).

Results

Of the 56 patients who completed F-18 FDOPA PET imaging and SOT, 26 were assigned a positive scan reading and 30 were assigned a negative scan reading (see Fig. 2). Furthermore, the physician responsible for the final SOT agreed with the PET scan results of 47 patients; for the remaining nine patients, SOT pointed to dopaminergic degeneration whereas the PET scan showed no dopaminergic degeneration (Table 1). For these nine patients, PET scans were normal based on blinded reading. Even though the SOR values are not used for the calculation of the sensitivity, specificity, PPV, and NPV, it is worth mentioning the usefulness of SOR values as an additional tool. For our center, an SOR value above 2.0 is considered normal [6]. As shown in Appendix Table 4, the SOR values for these patients are in the normal range. Our manual method of ROI selection is also briefly mentioned in the Table 2 legend. For 8 of the patients, the SOR values are greater than 2.0. The distribution of the SOR values for the entire cohort is graphically displayed in Fig. 3 (Appendix).

Fig. 2.

Study flow diagram. Of the 69 subjects enrolled, 68 underwent F-18 FDOPA PET imaging and 56 of these subjects were further evaluated clinically 6–12 months later by a movement disorder specialist blinded to the PET scan (SOT, standard of truth). Imaging revealed loss of dopaminergic neurons in 26 subjects; loss of dopaminergic neurons was not detected in 30 subjects

Table 1.

Results of visual assessment and diagnostic values of F-18 FDOPA PET

PET scan (expert reader) vs. final SOT
PET scan	Final SOT		Total
	Dopaminergic degeneration	No dopaminergic degeneration
	(n)	(n)	(n)
Positive	24	2	26
Negative	9	21	30
Total	33	23	56
Statistics	Results	Exact binomial 95% confidence interval
Sensitivity	72.7%	(54.5%, 86.7%)
Specificity	91.3%	(72.0%, 98.9%)
PPV	92.3%	(74.9%, 99.1%)
NPV	70.0%	(50.6%, 85.3%)
PET scan (majority reader) vs. final SOT
PET scan	Final SOT		Total
	Dopaminergic degeneration	No dopaminergic degeneration
	(n)	(n)	(n)
Positive	20	3	23
Negative	13	20	33
Total	33	23	56
Statistics	Results	Exact binomial 95% confidence interval
Sensitivity	60.6%	(42.1%, 77.1%)
Specificity	87.0%	(66.4%, 97.2%)
PPV	87.0%	(66.4%, 97.2%)
NPV	60.6%	(42.1%, 77.1%)

Table 4.

SOR quantification (n = 56)

	L posterior putamen	R posterior putamen	L anterior putamen	R anterior putamen	L caudate	R caudate
SOR quantification for positive scans (n = 26)
Range	1.446–2.268	1.392–2.472	1.647–3.210	1.636–2.899	1.930–3.262	1.938–3.324
Average	1.790	1.736	2.293	2.263	2.581	2.573
Standard deviation	0.238	0.261	0.352	0.365	0.392	0.388
SOR quantification for negative scans (n = 30)
Range	1.899–3.486	1.717–3.288	1.964–4.125	2.067–4.107	2.034–4.117	1.966–4.087
Average	2.627	2.543	3.063	3.023	3.104	3.080
Standard deviation	0.388	0.379	0.463	0.483	0.473	0.432

Table 2.

Inter-rater reliability statistics

Readers compared	Kappa	Kappa 95% confidence interval
Reader 1 vs. Reader 2	0.634	(0.437, 0.830)
Reader 1 vs. Reader 3	0.747	(0.572, 0.921)
Reader 2 vs. Reader 3	0.733	(0.552, 0.914)

Fig. 3.

SOR values for positive and negative scans

Table 1 summarizes the main results for this study and includes positive and negative scan readings, final SOT as determined by the expert movement disorders specialist, sensitivity, specificity, PPV, and NPV. A comparison of the visual scan readings provided by the three blinded readers is provided in Table 2. These tables illustrate very high specificity and PPV as well as moderately high sensitivity and NPV. The inter-rater reliability is also in the significant range of 0.634–0.747.

Comparison of Baseline Diagnosis with the Diagnosis after 6 Months

All subjects were evaluated by the movement disorders specialist at baseline (at the time of imaging) and follow-up (6–12 months after imaging) and were assigned a diagnosis of dopaminergic degeneration or no dopaminergic degeneration. In six patients, the baseline and follow-up diagnosis were not in agreement. Out of the six patients, five of them had the initial diagnosis as dopaminergic degeneration, but the diagnosis was changed to no dopaminergic degeneration after 6 months. The final diagnosis of these patients was in agreement with the F-18 FDOPA PET scan (negative scan). In one patient (patient ID 093), the baseline diagnosis was dopaminergic degeneration and the diagnosis after six months was no dopaminergic degeneration, but the PET scan showed a positive scan with reduced uptake in the putamen bilaterally.

The subdivisions of the parkinsonian syndromes reported by the physician responsible for the SOT are purely based on physical examination (UPDRS) of patients (Online Resource 4). F-18 FDOPA PET cannot make this distinction.

Discussion

F-18 FDOPA PET scan has very high specificity and PPV, which provide evidence that the test performs well in a patient referral population where the prevalence of dopaminergic degeneration is about 59%. From data in Table 1, we draw the following conclusions: (1) When the PET scan is performed on a patient without dopaminergic degeneration, there is a 91.3% (95% CI: 72.0 to 98.9%) probability that this patient will be identified as negative (no dopaminergic deficit), and there is a 92.3% (95% CI: 74.9 to 99.1%) probability that those who test positive by the PET scan truly have dopaminergic degeneration; (2) The sensitivity and NPV were slightly lower, at 72.7% and 70.0%, respectively. When the PET scan is performed on a patient with dopaminergic degeneration, there is a 72.7% (95% CI: 54.5 to 86.7%) probability that this patient will be identified as positive (dopaminergic deficit), and there is a 70.0% (95% CI: 50.6 to 85.3%) probability that those who test negative by the PET scan truly have no dopaminergic degeneration.

Historically, thousands of F-18 FDOPA PET scans have been performed without any reported incidence. Online Resource 5 lists adverse events observed in this study, none of which were causally related to F-18 FDOPA injection. No serious adverse events were reported.

The usefulness of F-18 FDOPA PET imaging in confirming PD is based upon a large number of studies dating back to the 1980s. Europe and other countries already have approval for F-18 FDOPA PET imaging in PD [3]. The methodology has been advancing extensively since then and now, and except for research scans that may require mathematical treatment of image/blood data, simple ratio methods have demonstrated excellent support for visual scan reading (Online Resource 6 - Tables 1 and 2).

Our data suggest the high resolution of PET provides increased confidence in visual scan reading even by non-experts after a short training period. Criteria have been established to successfully diagnose early, moderate, and advanced PD based upon the reduction in F-18 FDOPA uptake in posterior putamen with and without asymmetry, involvement of anterior putamen, and finally caudate nuclei.

Clinical Importance for the Approval of F-18 FDOPA PET for Patient Use in the USA

F-18 FDOPA injection has a very short half-life (109.7 min) compared to the current SPECT imaging agent using I-123 (13 h) that can potentially result in higher radiation exposure to the basal ganglia [18]. With the approval of F-18 FDOPA imaging, patients will have access to an alternative procedure compared to the current I-123 FPCIT SPECT at similar cost. Compared to SPECT scanners, current PET scanners have higher resolution and reduced imaging time (20 min scan duration, which can be significantly reduced with the newer generation of PET scanners that offer larger axial field-of-view and increased sensitivity resulting in increased patient comfort/compliance. Additionally, some patients may be allergic to the iodine (Lugol’s solution) used for SPECT imaging. For a direct comparison of I-123 FPCIT and F-18 FDOPA, see Eshuis et al. 2009 [4].

We used very liberal exclusion criteria to reflect real life patient population in whom the referring physician wanted dopaminergic scans for diagnosis/confirmation. Additionally, we used an expert reader as well as two non-expert nuclear medicine readers and performed inter-rater reliability analysis. The results in Table 1 demonstrate high sensitivity/specificity/PPV/NPV (70–90% for the expert reader and 61–87% for majority reading) similar to previously reported results [19–21]. As expected, the results for majority decision are slightly lower because they may or may not have included the expert reader. However, the inter-rater reliability is relatively high, kappa values from 0.634–0.747, suggesting that routine imaging with F-18 FDOPA in both academic/non-academic settings will be useful for patients and referring physicians. Comparison with previously published data provides us with confidence that our current imaging technique is as robust as any other dopaminergic uptake or transporter imaging radiotracer currently in use.

One limitation of this study was the use of clinical gold standard instead of post-mortem data. In our prospective study, waiting for autopsy data would have taken an unpredictably long time with limited number of subjects. Moreover, the use of clinical gold standard, except in very early-stage PD, is considered an acceptable alternative [22].

Conclusion

The results from our prospective study demonstrate high specificity and moderate sensitivity of F-18 FDOPA PET in the diagnosis of dopaminergic deficit parkinsonian syndromes. We are currently pursuing insurance reimbursement to make F-18 FDOPA a widely available radiotracer for the clinical work-up of parkinsonian syndromes.

Appendix

A Fluorodopa F-18 FDOPA PET scan has the unique capability to quantify striatal neuronal loss. This SOR quantification may be helpful as an additional tool to evaluate the scans. Even though we have used SOR quantification procedure in the past for routine clinical interpretation of the PET scan results, for this clinical study the SOR values were not used for the visual determination of the PET scan results. The SOR values were blinded to the image readers and the physician collecting SOT. This SOR quantification is straightforward and does not require any special software program (see below).

We calculated the SOR values for all patients to see how the PET scan results correlate with the SOR values to assess loss of dopaminergic neurons. If the SOR values can be used to further confirm the visual interpretation of the PET scan results, it may be used as a supportive tool for the differential diagnosis of Parkinsonian syndromes. The average values of the left posterior putamen for the positive scans are 1.790 compared to the average values of the left posterior putamen for the negative scans is 2.627. The negative scans are normal, and the values agree with the SOR values for normal subjects (2.545) from our previously unpublished retroactive study (study #17). L = left; R = right.

We did not use ipsilateral/contralateral classification because the images were only read visually without any additional quantification or clinically relevant data. Moreover, instead of focusing on the contralateral posterior putamen (the site of early-stage dopaminergic deficit) this also represents the worst-case scenario for image interpretation.

SOR estimation:

Even though we have an automated MRI-supported ROI selection methodology for F-18 FDOPA PET data analysis available, for this study we used our manual method that is in use for more than 20 years that we find very straightforward and easy to use and does not require any special software except what is supplied by the PET scanner manufacturers for image processing, thereby making it easy to implement anywhere.

Region-of-interest (ROI) analysis is carried out on the F-18 FDOPA PET images by first adding the central five slices on which the striatum is visible (usually 8–9 slices on most scanners; composite slice thickness is 16–18 mm). Large circular ROIs (20 cm²) are placed on both left and right occipital lobes. The average count value is calculated and then the whole image is divided by this value to generate the striatal/occipital ratio (SOR) image. Four small circular ROIs (1.2 cm²) are placed on the transaxial slice along the axis of the striatum covering the caudate and anterior, middle, and posterior sections of the putamen. The ratio values are then compared with the values obtained from a database of normal control subjects. Anything less than two standard deviations below the normal value is considered abnormal

Distribution of striatal/occipital ratio (SOR) values for the 56 subjects with F-18 FDOPA PET imaging and SOT data in the right (left panel) and left (right panel) posterior putamen.

Author Contribution

The study was designed by Vijay Dhawan, Martin H Niethammer, Matthew Hellman, David Eidelberg, and Thomas Chaly. Material preparation and data collection were performed by Vijay Dhawan, Martin H Niethammer, Matthew Hellman, Toni M Fitzpatrick, David Bjelke, Jaskirat Singh, Loreta M Quatarolo, and Yoon Young Choi. The data analysis was performed by Vijay Dhawan, Martin L Lesser, Karalyn N Pappas, David Bjelke, Alice Oh, and Thomas Chaly. Project administrative support was provided by Toni M Fitzpatrick, Loreta M Quatarolo, Yoon Young Choi, and Alice Oh. Supervision was provided by David Eidelberg. The first draft of the manuscript was written by Vijay Dhawan, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This research was supported by funds from the Department of Radiochemistry/Cyclotron directed by Thomas Chaly. He is the senior author and had a role in the study design, data analysis, interpretation and preparation of the manuscript.

Data Availability

Contact the corresponding author for data requests.

Declarations

Ethics Approval and Consent to participate

This study was performed with Institutional Review Board approval of the Feinstein Institutes for Medical Research, and in compliance with the HIPAA Privacy Rule. Informed consent was obtained from all individual participants included in the study. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 2013 Helsinki declaration and its later amendments or comparable ethical standards.

Consent for Publication

Informed consent was obtained from all individual participants included in the study.

Competing Interests

Vijay Dhawan, Martin H Niethammer, Martin L Lesser, Karalyn N Pappas, Matthew Hellman, Toni M Fitzpatrick, David Bjelke, Jaskirat Singh, Loreta M Quatarolo, Yoon Young Choi, Alice Oh, David Eidelberg, and Thomas Chaly declare that they have no competing financial interests.