‘It knocked me for six . . . I became very low . . . I thought it can’t be me . . . it’s just elderly people who got it.’
(patient)2
‘I found it hard to cope with life . . . I didn’t tell anyone . . . I couldn’t face the reality of it.’
(patient)2
5.1. Definition and differential diagnosis
There are many manifestations of PD but the classical diagnostic symptoms are:
The physical signs of PD include:
poverty of movement (
hypokinesia), eg loss of facial expression and arm swing, difficulty with fine movements
rest tremor.
At diagnosis, these signs are usually unilateral, but they become bilateral as the disease progresses. Later in the disease additional signs may be present including postural instability (eg tendency to fall backwards after a sharp pull from the examiner: the ‘pull test’), cognitive impairment and orthostatic hypotension (OH).
There is no single way to define Parkinson’s disease or what is often called idiopathic Parkinson’s disease in order to differentiate it from other causes of parkinsonism, such as multiple system atrophy (MSA) and progressive supranuclear palsy (PSP).
PD is traditionally defined, pathologically, by the finding of Lewy bodies and degeneration of catecholaminergic neurones at post-mortem. Using a pathological definition of PD is problematic for a number of reasons:
In recent years, attempts to define PD genetically have become possible with the discovery of monogenic forms of the disease. However, such families account for a very small proportion of cases.
Another potential way to diagnose PD is using the response to dopaminergic medication. However, this dopaminergic responsiveness can be seen in conditions other than PD such as MSA.
The decline in dopaminergic neurones identified by radionuclide positron emission tomography (PET) or single photon emission computed tomography (SPECT) has also been proposed as a method of defining PD. Unfortunately, this decline is seen in conditions other than PD such as MSA and PSP.
Given these difficulties, it is generally accepted that the diagnosis of PD should be based on clinical findings. The most widely accepted clinical criteria for the diagnosis of PD are those introduced by the UK PDS Brain Bank Criteria ().35
UK PDS Brain Bank Criteria for the diagnosis of PD.
It is important to make an accurate diagnosis in a person with suspected PD as this has an important bearing on prognosis. People with PD will have a longer life expectancy than those with MSA or PSP and will respond better to dopaminergic medication.
PD must also be differentiated from other conditions presenting with tremor (). This can be particularly difficult as PD can present with a postural and action tremor similar to that seen in essential tremor.
In addition, PD must be differentiated from other causes of a parkinsonian syndrome or parkinsonism (). The most common problems arise with multiple cerebral infarction and degenerative parkinsonian syndromes such as MSA and PSP. Differential diagnosis can also be difficult in elderly people since extrapyramidal symptoms and signs are common.34
Causes of a parkinsonian syndrome.
RECOMMENDATION
- R8.
PD should be suspected in people presenting with tremor, stiffness, slowness, balance problems and/or gait disorders.
[D (GPP)]
5.1.1. Methodological limitations of the diagnostic studies
When interpreting the literature about PD diagnosis, the following methodological issues should be considered:
lack of long-term prospective clinical and pathological
follow-up as a reference standard
lack of operational definitions such as defining specialists or clinical diagnostic criteria
unclear whether investigators were blinded to initial diagnosis
sample sizes necessarily limited by the number of cases available with neuropathological outcomes
PD trial age groups are often young as studies were performed by neurologists who see a younger population of people with PD
most studies included people with established disease lasting some years
varying geographical locations
some studies are in specialised units and may not reflect the diagnostic accuracy of other units in the UK
exclusion of some studies using magnetic resonance volumetry and magnetic resonance spectroscopy (MRS) as they lacked appropriate population, intervention and outcome criteria
lack of statistical details of diagnostic accuracy such as sensitivity, specificity and positive predictive values
lack of economic evaluations of SPECT.
5.2. Clinical versus post-mortem diagnosis
Most experienced specialists have adopted the UK PDS Brain Bank Clinical Criteria () for the diagnosis of PD.
How do these compare with the accuracy of pathological diagnosis?
5.2.1. Methodology
Three diagnostic studies were found that assessed the accuracy of clinical diagnosis in parkinsonism compared with autopsy.36–38 These studies compared clinical diagnosis, at various stages of disease progression, to a final diagnosis including details of autopsy findings. The clinical diagnosis was determined using the UK PDS Brain Bank Criteria () in two of three studies.37,38 A third study determined a diagnosis of PD when at least two of the three cardinal signs (bradykinesia, rigidity and resting tremor) were present.36
5.2.2. Evidence statements
Two studies (N=5936 and N=10037) examined people with a terminal diagnosis of PD and found the frequency of people misdiagnosed with PD (ie they did not meet the pathological criteria at post-mortem) was 35% and 24% respectively.36,37 When recommended diagnostic criteria (UK PDS Brain Bank) were retrospectively applied, diagnostic accuracy increased from 70% to 82%.37 (DS II)
A more recent UK PDS Brain Bank study38 examined the brains of 143 people with parkinsonism. These people had previously been seen by a neurologist, with five dedicated movement disorder specialists seeing 92% of the cases, and been given a clinical diagnosis of PD or alternative parkinsonian condition. The clinical diagnosis was later revised in 44 of 122 cases where full follow-up information was available after a mean of 3.4 (range 0.5–12) years. The sensitivity of the final PD clinical diagnosis was 91%, a specificity of 98% and a positive predictive value of 99% (72 out of 73 correctly diagnosed). (DS II)
5.2.3. From evidence to recommendation
The pathological studies emphasise the need for particular care in making a clinical diagnosis of PD. There is limited evidence to suggest that the UK PDS Brain Bank Criteria have adequate sensitivity and specificity in comparison with post-mortem findings. The accuracy of diagnosis using the Brain Bank criteria increases as the condition progresses.
The availability of PD brain tissue has fostered much valuable research in recent years and should be encouraged in the future. Diagnostic information derived from post-mortem examination can also be of value to the families of individual patients.
RECOMMENDATIONS
- R9.
PD should be diagnosed clinically and based on the UK Parkinson’s Disease Society Brain Bank Criteria.
[B (DS)]
- R10.
Clinicians should be encouraged to discuss with patients the possibility of tissue donation to a brain bank for purposes of diagnostic confirmation and research.
[D (GPP)]
5.3. Expert versus non-expert diagnosis
The diagnosis of PD could be made in primary care by the person’s GP or in secondary care by a neurologist, geriatrician or general physician. More recently, PDNSs and other health professionals are developing diagnostic skills. Each may have different levels of expertise in evaluating people with possible PD.
What is the evidence that someone with special expertise is more accurate in diagnosing PD than someone with little experience?
5.3.1. Methodology
Four diagnostic studies39–42 were found looking at the accuracy of PD diagnosis in a community-based population. The specialist diagnosis was based on the UK PDS Brain Bank criteria in four of the studies.39,40,42 In one study41 the expert diagnosis was based on the investigator’s confidence in the diagnosis of PD, presence of atypical features, findings of imaging studies, response to levodopa and results of autopsy examinations. The criteria for the initial diagnoses were not specified in any of the trials. These studies were also performed on prevalent rather than incident PD populations.
5.3.2. Evidence statements
One study39 (N=126) assessed the diagnostic accuracy of neurologist and geriatrician clinical expert diagnosis versus existing clinical diagnosis of parkinsonism from medical records by a non-expert clinician. The standard for comparison was diagnosis according to strict clinical diagnostic criteria (the UK PDS Brain Bank Criteria) after a detailed neurological interview and examination. The study found that neurologists and geriatricians had a sensitivity of 93.5% (95% CI 86.3 to 97.6) and specificity of 64.5% (95% CI 45.4 to 80.8) compared with ‘non-specialist’ sensitivity of 73.5% (95% CI 55.6 to 87.1) and specificity of 79.1% (95% CI 64.0 to 90.0) for diagnostic accuracy. While the positive predictive value of specialists was greater than for other doctors, negative predictive values were equivalent. (DS II)
Another study40 applied the UK PDS Brain Bank criteria to 402 cases derived from a computerised list of people with PD receiving anti-parkinsonian medication from 74 general practices in North Wales. In 59% of cases, the GP made the initial diagnosis of PD. The people with PD were seen either at home or in a specialist movement disorder clinic where a neurological examination was performed. A definite PD diagnosis was made in 53% of all cases, thus the error rate in the community-ascertained cases was 47%. (DS II)
DATATOP (Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism) was a large, multi-site clinical trial41 in the USA and Canada involving 800 people with early-stage PD who were cared for by 34 investigators with a major interest in movement disorders. A secondary analysis examined the number of people with PD with a change in diagnosis after a mean follow-up of 6 years. The study showed that only 8% had a revised diagnosis. The revised diagnosis was clinical and not based on strict criteria or pathology. (DS II)
The UK-PDRG study,42 which investigated the long-term effectiveness of bromocriptine, selegiline and levodopa therapy, found a total of 49/782 people with PD (6%) had their diagnosis changed during the course of the trial. Individuals were eligible for inclusion in the study if they fulfilled criteria for a clinical diagnosis of PD. The authors do not state whether the revised diagnosis was made by one of the specialists performing the study, although this is likely. The authors also do not state whether a specialist or non-specialist conducted the initial diagnostic examination. (DS II)
5.3.3. From evidence to recommendation
These studies provide only circumstantial evidence on the diagnostic ability of experts versus non-experts. However, they show that the diagnosis of PD is wrong in around 47% of community-ascertained cases, 25% of non-expert secondary care diagnosed cases, and 6–8% of cases diagnosed by an expert in movement disorders.
Since medication can mask the symptoms and signs of PD, the GDG felt that people with suspected PD should be referred before treatment is commenced. This can be achieved only if people are seen quickly by experts, for an accurate diagnosis and commencement of treatment, if necessary.
The GDG also had experience that delay in making an accurate diagnosis can lead to psychological stress for the patient and their carer. Similarly, the need to revise an incorrect diagnosis that has, initially, been made by a non-expert can be stressful for patients.
The GDG acknowledges the timeline that the Department of Health and NHS are currently working towards for completion of diagnosis and treatment (18-week target). However, the GDG felt that in the case of PD it should not necessarily mean that patients would have to ‘start’ treatment within 18 weeks from GP referral but rather that this was when a ‘treatment decision’ was made for initial consultation and diagnosis.
5.4. Review of diagnosis
Given the error rate in making a diagnosis of PD, even in expert hands, it is apparent that the diagnosis should be kept under regular review.
What is the most appropriate frequency of follow-up after an initial diagnosis of PD?
5.4.1. Methodology
No trials were found which addressed the most appropriate frequency of follow-up of people with PD.
5.4.2. Evidence statements
No evidence was found on the most appropriate frequency of follow-up after the initial diagnosis of the disease.
5.4.3. From evidence to recommendation
In the absence of any evidence on the issue of frequency of follow-up, the GDG concluded that this should be based on clinical priority. In people with early mild symptoms of PD who may not even be on treatment yet, follow-up to check on the diagnosis and the need for treatment may be infrequent (every 6–12 months). Once treatment is commenced, follow-up may need to be more frequent (every 2–3 months) to assess the response to medication, titrate dosage and re-visit the diagnosis. In later disease, people with PD have more complex problems which require changes in medication. This may require review at frequent intervals (every 2–3 months).
RECOMMENDATION
- R12.
The diagnosis of PD should be reviewed regularly** and reconsidered if atypical clinical features develop.
[D (DS)]
5.5. Single photon emission computed tomography
In single photon emission computed tomography (SPECT), a gamma ray-emitting radioactive isotope is tagged to a molecule of interest (a tracer), which is given to the person with PD by intravenous injection. The labelled cocaine derivatives 123I-β-CIT and 123I-FP-CIT (N-ω-fluoropropyl-2β-carboxymethoxy-3β-(4-iodophenyl)tropane) have most commonly been used, although only the latter is licensed in the UK. These label the presynaptic dopamine re-uptake site and thus the presynaptic neurone, which can be visualised in two-dimensional images. These demonstrate normal uptake in the caudate and putamen in controls and in people with essential tremor, neuroleptic-induced parkinsonism or psychogenic parkinsonism, but reduced uptake in those with PD, PD with dementia, MSA or PSP.
How useful is SPECT in discriminating PD from alternative conditions?
5.5.1. Methodology
Fifteen studies addressed the diagnostic accuracy of SPECT scanning.43–58 The reference standard was clinical diagnosis: eight out of the 16 studies43,45–51 used the UK PDS Brain Bank Criteria, five studies44,52–55 used ‘established’ clinical criteria and three studies56–58 did not state the clinical criteria used to determine the diagnosis. Although many tracers are listed in the evidence statements, only 123I-FP-CIT is licensed for use in the UK. The 123I-β-CIT studies were included as it has a similar structure and labels the same receptors as the 123I-FP-CIT tracer. The GDG agreed that this evidence is supportive of 123I-FP-CIT studies and provides a consistency of effect.
5.5.2. Health economic methodology
Only one study met quality criteria that addressed the economic evaluation of SPECT.59 This study was based on 123I-FP-CIT SPECT effectiveness data, specificity and sensitivity of clinical examination and prevalence of PD were based predominantly on UK data. However, costs were based on German 2002 data.59
5.5.3. Evidence statements
For the differentiation of people with parkinsonism (ie PD, MSA or PSP) from people with essential tremor or controls using SPECT, all studies produced a high sensitivity (range 87% to 98.3%) and specificity (range 80% to 100%).43,45,49,52,53 A summary of the evidence produced in these five studies is provided in and . (DS Ib)
Diagnostic accuracy of SPECT imaging: differentiation of tremulous disorders.
Diagnostic accuracy of SPECT imaging: differentiation of PD and controls.
Three studies (N=80,47,48,54 N=17,47,48,54 N=18347,48,54) attempting to differentiate PD from other parkinsonian conditions (eg MSA, PSP) had insufficiently high levels of sensitivity (range 77% to 97%) and specificity (range 75% to 83%).47,48,54 (DS Ib)
One study58 found, by comparing the 123I-β-CIT SPECT imaging diagnosis for people with parkinsonian syndrome with a clinical diagnosis (based on 6 months’ follow-up), that there was disagreement between only three out of 35 cases (8.6%) with visual diagnosis and two out of 35 cases (5.7%) with quantitative imaging diagnosis. (DS Ib)
5.5.4. Health economic evidence statements
The economic findings indicated:59
For approximately an additional €733 in Euro 2002 (approximately £511), for the equivalent of a patient-month with adequate treatment, SPECT could be used to confirm a PD diagnosis in people with a positive clinical examination before the initiation of treatment.59 Adequate treatment month equivalents (ATME) were used to reflect both duration of adequate treatment and severity of incorrect treatments. The authors indicated that a 0.55 ATME gain per patient is equivalent to approximately 17 additional days of treatment to a PD patient or withholding approximately 2 days of treatment and side effects to a patient who does not have PD.
The specificity of clinical examination and frequency of PD in the clinic population of PD had the greatest relative impact on the incremental cost-effectiveness ratio (ICER) of SPECT following positive clinical examination compared with clinical examination alone. In the sensitivity analysis, when the specificity of clinical examination is reduced to 0.80 (from 0.984) the ICER drops to €63 (approximately £44).59 This suggests that as more non-PD cases are incorrectly classified as PD cases in clinical examination, the greater the cost-effectiveness of SPECT. When the frequency of PD in the clinic population is increased to 74% (from 53%) the ICER increases to €2,411 (approximately £1,697).59 This suggests that the cost-effectiveness of SPECT decreases when the frequency of PD in the clinic population increases. In these populations, there may be fewer false-negative results and therefore fewer people incorrectly being treated for PD. This would mean there are fewer cost-savings from withholding incorrect treatment and therefore an increase in the relative cost-effectiveness of SPECT.
5.5.5. From evidence to recommendation
Considerable evidence supports the use of 123I-FP-CIT SPECT in people with postural and/or action tremor of the upper limbs in the differentiation of essential tremor from a dopaminergic deficiency state. 123I-FP-CIT SPECT cannot, with high accuracy, differentiate PD from other dopaminergic deficiency states such as MSA and PSP. Future work may demonstrate the value of this technique in differentiating parkinsonism due to neuroleptic medication and psychogenic parkinsonism from a dopaminergic deficiency state.
Several clinical trials using SPECT or PET to follow the progression of PD found that 4%,60 11%61 and 14%62 with a clinical diagnosis of PD had normal imaging at the start of the trial. Further long-term clinical follow-up of these people is required.
Due to the subjectivity of the effectiveness measurement, the GDG decided the economic study59 does not support or refute the clinical recommendations. Further development of comparable effectiveness outcomes in diagnostic economic evaluations is required.
RECOMMENDATIONS
- R13.
123I-FP-CIT SPECT should be considered for people with tremor where essential tremor cannot be clinically differentiated from parkinsonism.
[A (DS)]
- R14.
123I-FP-CIT SPECT should be available to specialists with expertise in its use and interpretation.
[D (DS)]
5.6. Positron emission tomography
In positron emission tomography (PET), a positron-emitting radioactive isotope is tagged to a tracer molecule, which is administered by intravenous injection. The most frequently used positron-emitting isotope in this field is 18fluorine, which is attached to dopa or deoxyglucose. 18F-fluorodopa is taken up by the presynaptic dopaminergic neurones of the caudate and putamen (corpus striatum). 18F-fluorodeoxyglucose (FDG) is taken up by all metabolically active cells and phosphorylated to a metabolite, which is trapped in the tissue for the time course of the study.
How valuable is PET in the differential diagnosis of parkinsonism?
5.6.1. Methodology
Six diagnostic studies63–68 were found which addressed the effectiveness of PET scanning compared with clinical diagnosis in the differential diagnosis of a parkinsonian syndrome. No studies were found which compared the effectiveness of PET in the differentiation of PD from essential tremor.
5.6.2. Evidence statements
In one study68 the diagnostic accuracy of 18F-desmethoxy-fallypride PET imaging for the differential diagnosis of atypical (N=16) versus idiopathic (N=16) parkinsonian syndromes showed a threshold value of 2.495 (caudate uptake ratio). The sensitivity, specificity and accuracy were 74%, 100% and 86% respectively. Using this threshold, the positive and negative predictive values for the diagnosis of atypical parkinsonian syndromes were 100% and 76%. (DS Ib)
In one study67 the multi-diagnosis group discriminate analysis from 18F-FDG PET scan images found sensitivity of 75% and specificity of 100% in the PD group (N=8), sensitivity of 100% and specificity of 87% in the MSA group (N=9), and sensitivity of 86% and specificity of 94% in the PSP group (N=7). (DS II)
One study,69 using 18F-FDG uptake, reported 74% of all participants (early PD (N=15), atypical PD (N=9) and controls (N=15)) were correctly classified when regional cerebral glucose metabolism (rCMRGIc) was analysed. This diagnostic accuracy increased to 95% using topographical profile rating, which is a method for calculating participant scores for abnormal regional metabolic co-variance patterns in individual people with PD. (DS II)
One study (N=90),63 using 18F-fluorodopa uptake, found people with clinically diagnosed PD were correctly classified by PET in 64% of the cases and those with atypical parkinsonism (MSA or PSP) in 69% of the cases. (DS II)
In another study70 the probability of the correct diagnosis by 18F-fluorodopa PET was ≥99% for the majority of people with PD (40/41) and controls (26/28). (DS II)
5.6.3. From evidence to recommendation
PET has better spatial resolution than SPECT, so it might be anticipated that PET should be of value in differential diagnosis. However, the evidence for PET’s role in differentiating PD from other parkinsonian conditions using FDG requires further confirmation. No work was found on PET’s ability to differentiate PD from essential tremor. This lack of evidence stems from the high cost and poor availability of PET. Further research is required in this area.
RECOMMENDATIONS
- R15.
PET should not be used in the differential diagnosis of parkinsonian syndromes, except in the context of clinical trials.
[B (DS)]
5.7. Magnetic resonance imaging
Structural magnetic resonance imaging (MRI) provides two- and three-dimensional images of intracranial structures using high magnetic field strengths to excite the hydrogen atoms in water molecules. In PD this technique has been used to examine various structures known to be involved in the pathology of the condition in the hope that it may prove of value in differential diagnosis.
How useful is structural MRI in the differential diagnosis of parkinsonian conditions and essential tremor?
5.7.1. Methodology
Eight diagnostic studies64,66,71–76 were found which addressed the effectiveness of MRI compared with long-term clinical follow-up in diagnosing people with a parkinsonian syndrome. Various MRI scanning sequences were used.
5.7.2. Evidence statements
Seven of these studies64,71–76 provided diagnostic accuracy data for MRI using various techniques. The results are summarised in .
Diagnostic accuracy of MRI.
Another study66 found non-concordance between neuroradiological diagnosis and clinical diagnosis in 2/21 people with PD, 5/14 people with MSA-P and 1/4 people with MSA-C. (DS II)
One study75 reported only 15% of people with PD and 24% of those with PSP had abnormal T2 hypointensity in the posterolateral putamen and none had abnormal putaminal proton density hyperintensity. (DS Ib)
One study74 found two false negatives in the PSP group (one had a diagnosis of clinically probable PSP and one clinically definite PSP) and five false positives (two were non-diseased controls and three had a diagnosis of PD). (DS II)
5.7.3. From evidence to recommendation
In expert hands structural MRI has proved of some value in differentiating PD from other types of parkinsonism, but further research is required before it can be recommended in routine clinical practice.
5.8. Magnetic resonance volumetry
Magnetic resonance volumetry uses the same principles as structural MRI to measure the size of three-dimensional volumes of tissue. This technique has been used to examine the size of various structures involved in the pathology of PD.
Can magnetic resonance volumetry be used in the differential diagnosis of parkinsonism?
5.8.1. Methodology
Two studies76,77 addressed the diagnostic effectiveness of magnetic resonance volumetry against retrospective clinical diagnosis in determining an accurate diagnosis in people with parkinsonian syndrome.
5.8.2. Evidence statements
One study77 (N=61) found no differences between people with PD and controls on any of the magnetic resonance volume measures. However, individuals with PSP were distinguished from people with PD and controls with a sensitivity of 95.2% and a specificity of 90.9% (mainly due to frontal grey matter volume measure). (DS Ib)
Another study76 (N=53) found that mean superior cerebellar peduncle volume atrophy on visual image analysis differentiated PSP from PD, MSA and controls with a sensitivity of 74% and a specificity of 94%, whereas in quantitative analysis the best sensitivity and specificity of the volumetric analysis were 74% and 77%. (DS II)
5.8.3. From evidence to recommendation
While two studies suggest that volumetric MRI can help in the differentiation of PD from other types of parkinsonism, further work is required before it can be recommended.
RECOMMENDATION
- R18.
Magnetic resonance volumetry should not be used in the differential diagnosis of parkinsonian syndromes, except in the context of clinical trials.
[D (DS)]
5.9. Magnetic resonance spectroscopy
Proton MRS measures the concentrations of intermediary metabolites in small volumes of brain tissue. N-acetylaspartate is found in the highest concentration in neurones and their processes, whereas creatine is a marker of energy status and choline is an indicator of membrane synthesis and degradation.
Can MRS be helpful in the correct diagnosis of parkinsonism?
5.9.1. Methodology
A systematic review78 of mixed study designs assessed the diagnostic accuracy of MRS against a clinical diagnosis of a range of parkinsonian syndromes.
5.9.2. Evidence statements
The review78 concluded that due to the heterogeneous nature of the available evidence no comments on the variability in metabolite concentrations and ratios between people with parkinsonian disorders could safely be made. (DS II)
5.9.3. From evidence to recommendation
Contradictory results have been found on the value of MRS in differentiating PD from controls and other types of parkinsonism.
RECOMMENDATION
- R19.
Magnetic resonance spectroscopy should not be used in the differential diagnosis of parkinsonian syndromes.
[B (DS)]
5.10. Acute levodopa and apomorphine challenge tests
Many people with PD respond to single doses of oral levodopa and/or subcutaneous apomorphine.
Can such responses be assessed using clinical rating scales to provide a diagnostic test for PD?
5.10.1. Methodology
A systematic review79 and an additional diagnostic study80 addressed the effectiveness of acute levodopa and apomorphine testing in determining an accurate diagnosis of people with a parkinsonian syndrome. Another review81 published prior to the included systematic review79 was excluded because it summarised the same papers.
5.10.2. Evidence statements
The systematic review79 included 13 studies, four of which examined people with de novo PD and nine others which examined people with well-established PD and with other parkinsonian syndromes. These two groups are presented separately in and . The diagnostic study80 followed people with PD for 3 years to investigate whether an acute challenge of carbidopa/levodopa had better diagnostic accuracy compared with the acute apomorphine challenge test. These results are also included in .
Diagnostic accuracy of acute apomorphine and levodopa challenge testing in de novo PD cases.
Diagnostic accuracy of acute apomorphine and levodopa challenge testing in established PD cases.
The systematic review used logistic regression analysis to determine whether there was a significant difference between the three tests for the misclassification of participants. Two studies82,83 demonstrated no significant difference between the acute apomorphine challenge test and chronic levodopa therapy. However, two other studies82,84 provided evidence that there was a difference between the acute levodopa challenge test and chronic levodopa therapy, in favour of chronic levodopa (p<0.001). (DS II)
The diagnostic study80 commented on the adverse reactions to acute apomorphine challenges. Drowsiness, nausea, vomiting, hypotension and sweating were reported to such an extent that these effects prevented an increased dosage in some people with PD. Levodopa was better tolerated than apomorphine, with vomiting and nausea still occurring, but infrequently. No statistics were provided on whether the better tolerance of the levodopa challenge over the apomorphine challenge was significant. (DS III)
5.10.3. From evidence to recommendation
The evidence demonstrates that acute challenge tests with levodopa and apomorphine add nothing to standard chronic levodopa therapy in the differentiation of established cases of PD from other causes of parkinsonism. Furthermore, when used in the early stages of the disease, as they would be in clinical practice, acute challenges with levodopa and apomorphine are less discriminatory than the standard practice of treating people with levodopa as outpatients. This does not preclude the use of acute apomorphine challenges to assess whether a person with later PD will still respond to dopaminergic medication.
RECOMMENDATION
- R20.
Acute levodopa and apomorphine challenge tests should not be used in the differential diagnosis of parkinsonian syndromes.
[B (DS)]
5.11. Objective smell testing
Around 80% of people with PD may have an impaired sense of smell (hyposomia).85
Since smell can be objectively tested with a battery of different odours, is it possible that objective smell identification may be useful in PD differential diagnosis?
5.11.1. Methodology
We found six diagnostic studies looking at the effectiveness of smell testing in PD differential diagnosis. Two techniques were employed: the ‘Sniffin Sticks’ test86 and the University of Pennsylvania Smell Identification Test (UPSIT). The tests were used to differentiate parkinsonian syndromes86–88 and people with PD from healthy controls.85,89,90
5.11.2. Evidence statements
A separate summary of the five diagnostic accuracy studies is listed in and . One study90 found the discriminatory test scores decreased as a function of age for each of the participant groups and that, on average, lower UPSIT scores are needed to clinically define PD in males than in females. (DS II)
Diagnostic accuracy of smell-testing techniques in differentiating parkinsonian syndromes.
Diagnostic accuracy of smell-testing techniques in differentiating parkinsonian syndromes from non-parkinsonian syndromes.
Another study89 reported that of the 40 odorants in the UPSIT test, the combined smell of pizza and wintergreen was the best discriminator. In addition, pizza (oregano smell) alone specifically indicates anosmia for people with PD with a very high sensitivity and specificity (). (DS II)
A third study85 found abnormal olfactory function in 82% of the PD participants tested compared with 23% of controls. (DS II)
5.11.3. From evidence to recommendation
Objective smell testing has a moderate sensitivity and specificity in differentiating people with PD from controls. However, there are few data on its ability to differentiate PD from other parkinsonian syndromes. Smell is also diminished in Alzheimer’s disease.92 At present, smell identification adds little in the differential diagnosis of parkinsonism but this situation may change with further research.
RECOMMENDATION
- R21.
Objective smell testing should not be used in the differential diagnosis of parkinsonian syndromes, except in the context of clinical trials.
[B (DS)]
- *
The GDG considered that people with suspected mild PD should be seen within 6 weeks, but new referrals in later disease with more complex problems require an appointment within 2 weeks.
- **
The GDG considered that people diagnosed with PD should be seen at regular intervals of 6–12 months to review their diagnosis.
