Abstract
Objective: To study the use of brain parenchyma sonography (BPS) in discriminating between patients with idiopathic PD (IPD) and atypical parkinsonian syndromes (APS).
Methods: Twenty-five patients with APS, 9 with progressive supranuclear palsy (PSP) and 16 with multiple-system atrophy (MSA), and 25 age-matched patients with IPD were prospectively studied with BPS according to a standardized protocol.
Results: Twenty-four of the 25 (96%) IPD patients exhibited hyperechogenicity of the substantia nigra (SN) but only 2 of 23 (9%) APS patients (Mann–Whitney U test, p < 0.001). In those two APS patients, SN hyperechogenicity was moderate only, whereas the remaining 21 APS patients had normal SN echogenicity. The specificity of SN hyperechogenicity in detection of clinically diagnosed IPD patients was 96%, and the sensitivity was 91%. If SN hyperechogenicity was marked, APS could be excluded because of a positive predictive value of 100% for IPD. Nucleus lentiformis hyperechogenicity was found in 17 of 22 (77%) APS patients but in only 5 of 22 (23%) IPD patients (Mann–Whitney U test, p < 0.001). Nucleus caudatus and thalamus echogenicity and widths of the third ventricle and of the frontal horns of the lateral ventricles did not discriminate between IPD and APS. Two patients with PSP could not be assessed because of an insufficient bone window.
Conclusions: BPS is a novel and noninvasive method to differentiate highly specifically between IPD and APS. Therefore, BPS might become a standard investigation in parkinsonian disorders.
The clinical differentiation between idiopathic PD (IPD) and atypical parkinsonian syndromes (APS) such as multiple-system atrophy (MSA) and progressive supranuclear palsy (PSP) is often difficult, especially in the early course of the disease.1,2⇓ Also, dopa sensitivity as tested by the apomorphine test is not always conclusive. Sophisticated neuroimaging methods such as MRI, SPECT, and PET may help to discriminate between IPD and APS, but characteristic findings are often lacking in early stages.3,4⇓ Furthermore, these methods are expensive, not readily available, and may involve exposure to radiation.
Brain parenchyma sonography (BPS) is a new ultrasound technique displaying the tissue echogenicity of the brain through the intact skull. The advantages of this method are its potentially wide availability, short investigation times, low costs, and noninvasiveness. Movement artifacts may be easily compensated by the investigator. BPS proved to be reliable and sensitive in detecting basal ganglia alterations such as substantia nigra (SN) degeneration in IPD,5-7⇓⇓ lesions of lenticular nucleus in idiopathic dystonia,8 and degeneration of SN and caudate nucleus in Huntington’s disease.9 BPS measurements of the widths of the third and lateral ventricles correspond well with CT data.9,10⇓
In the current study, we prospectively evaluated the use of BPS in discriminating patients with APS from patients with IPD.
Methods.
Patients.
Altogether 25 APS patients and 25 age-matched IPD patients were studied. Table 1 shows their demographic data. APS patients were divided in two groups. Group MSA consisted of 11 women and 5 men with clinically probable MSA,11 all of whom had parkinsonism-predominant MSA. Their age was 66.0 ± 9.8 years, their disease duration 3.7 ± 1.9 years, and their disease severity 36.3 ± 16.6 on the motor part of the Unified Parkinson’s Disease Rating Scale (UPDRS-III).12 Group PSP consisted of five women and four men. Eight of these patients met the National Institute of Neurological and Communication Disorders and Stroke criteria for probable PSP.1 One case was classified as possible PSP because of prominent postural instability with early falls and other supportive criteria together with typical MR findings despite absence of vertical gaze palsy. In Group PSP, age was 71.0 ± 5.7 years, disease duration 4.8 ± 1.7 years, and disease severity 56.5 ± 19.0 on the UPDRS-III. Group IPD consisted of 14 women and 11 men. Their age was 67.7 ± 6.8 years, disease duration 5.7 ± 3.1 years, and IPD severity 28.7 ± 14.5 on the UPDRS-III.
Table 1 Demographic data of patients studied
BPS.
BPS was performed as previously described by using a color-coded phased-array ultrasound system equipped with a 2.5-MHz transducer (Sonoline Elegra; Siemens, Erlangen, Germany).13 Ultrasound scanning was performed through the right and left temporal acoustic bone windows using a penetration depth of 16 cm and a dynamic range of 45 dB. Image brightness and time gain compensation were adapted as needed. SN echogenic size measurements were performed on axial BPS scans automatically after manually encircling the outer circumference of the SN echogenic area. SN echogenic sizes of <0.2 cm2 were classified as normal, sizes of ≥0.25 cm2 as markedly hyperechogenic, and sizes in between as moderately hyperechogenic.7,14⇓ For intergroup comparisons, the measurements of both SN echogenic sizes were used. Additionally, echogenicity of the thalami, the lenticular nuclei, and the heads of the caudate nuclei was investigated and classified as hyperechogenic when it was more intense than the surrounding white matter.5 Classification of patients with respect to echogenicity of SN, lenticular nucleus, caudate nucleus, and thalamus was based on the most affected side of the investigated brain structure. The widths of the third ventricle and of the frontal horns of the lateral ventricles were measured on a standardized diencephalic axial scanning plane. All BPS examinations were performed by one investigator who was blinded to the diagnoses. BPS imaging of the mesencephalic brainstem and the basal ganglia was continuously video recorded. BPS video data were subsequently re-evaluated by a second independent investigator who was blinded to the diagnoses and the clinical features, including quantitative measurement of SN echogenic size and semiquantitative classification of echogenicity of the basal ganglia as listed above. For further evaluation, a given structure was accepted as hyperechogenic only if it was classified as hyperechogenic by both investigators.
Statistics.
Descriptive statistics are given as medians with lower (25th percentile) and upper (75th percentile) quartiles. For group comparison, the Mann–Whitney U test was used. Comparison of means was performed with the t-test for independent samples.
Results.
Table 2 summarizes the results of BPS of all brain structures studied in Groups IPD and APS. Figure 1 shows typical sonographic images of a patient with IPD and a patient with APS. Figure 2 shows the SN echogenic sizes calculated from combined bilateral measurements in Groups IPD and APS. BPS data of two PSP patients (4% of all patients studied) had to be excluded because of a bilaterally insufficient acoustic temporal bone window.
Table 2 Qualitative assessment of brain parenchyma echogenicity of substantia nigra, thalamus, lenticular nucleus, and caudate nucleus in patients with idiopathic PD and atypical parkinsonian syndromes
Figure 1. Sonographic images of identical midbrain axial sections in two patients. The butterfly-shaped midbrain section of low echogenicity is surrounded by the hyperechogenic basal cisterns. (A) Patient with atypical parkinsonian syndrome (multiple-system atrophy) exhibiting normal, nearly invisible substantia nigra (thick arrowheads) (thin arrowheads = red nuclei). (B) Patient with idiopathic PD. Note the marked bilateral hyperechogenicity of the substantia nigra (arrowheads). Echogenic area of the left substantia nigra was encircled for computerized measurement.
Figure 2. Echogenic substantia nigra (SN) size in patients from idiopathic PD (IPD), multiple-system atrophy (MSA), progressive supranuclear palsy (PSP), and atypical parkinsonian syndromes (APS; combined data of MSA and PSP patients) groups. Calculations were based on combined measurements of both SN sizes in each patient. Data are given as medians and 25 and 75% percentiles. *p < 0.001, Mann–Whitney U test.
IPD.
The median SN echogenic size of all IPD patients was 0.25 (0.21, 0.29) cm2. Twenty-four (96%) of the 25 IPD patients exhibited SN hyperechogenicity. In 19 (76%), the hyperechogenicity was classified as marked. In 12 (50%), it was unilateral, in 5 (21%) bilateral, and in 2 (8%) only unilateral measurement could be performed owing to insufficient temporal bone window. In 5 (21%) of the 24 patients with hyperechogenic SN, the hyperechogenicity was moderate. In two (8%) it was unilateral, in one (4%) bilateral, and in two (8%) only unilateral measurement could be performed owing to insufficient temporal bone window. IPD patients with marked SN hyperechogenicity had a mean UPDRS-III score of 27.9 ± 15.6, and patients with moderate hyperechogenicity had a mean score of 29.7 ± 12.9, which was not different (Mann–Whitney U test, p = 0.7). IPD duration of patients with marked hyperechogenicity was 5.6 ± 3.3 years and of patients with moderate hyperechogenicity 6.6 ± 2.7 years, which again was not different (t-test, p = 0.51).
Echogenicity of the thalamus was normal in all IPD patients. BPS of lenticular and caudate nuclei was not possible in three IPD patients because of insufficient temporal acoustic bone windows. Lenticular nucleus BPS demonstrated hyperechogenicity in five (23%) patients of Group IPD. Caudate nucleus BPS detected hyperechogenicity in 15 (68%) patients of Group IPD. Width of the third ventricle could be measured in 23 of 25 IPD patients and widths of the frontal horns of the lateral ventricles in 20. Width of the third ventricle was 7.4 ± 2.6 mm in IPD patients. Frontal horn width was 15.1 ± 3.4 mm in IPD patients.
APS.
The median SN echogenic size of all MSA patients was 0.15 (0.12, 0.17) cm2, 0.16 (0.12, 0.18) cm2 for all PSP patients, and 0.15 (0.12, 0.17) cm2 for all APS patients. Of the 23 patients from Group APS with adequate SN assessability, 2 (9%) exhibited SN hyperechogenicity that was bilaterally moderate in both patients (1 MSA, 1 PSP patient), and 21 (91%) had normal SN echogenicity. In one of the latter patients, SN echogenicity could be assessed only unilaterally.
BPS demonstrated normal echogenicity of the thalamus in all APS patients, whereas it revealed hyperechogenicity of the lenticular nucleus in 17 (77%) and of the caudate nucleus in 14 (64%) of the APS patients. BPS of lenticular and caudate nuclei was not possible in one MSA patient owing to insufficient temporal acoustic bone windows.
Widths of the third ventricle were 8.6 ± 1.5 mm in MSA patients, 12.9 ± 3.7 mm in PSP patients, and 10.0 ± 3.1 mm in all APS patients. Frontal horn widths were 15.1 ± 2.3 mm in MSA patients, 19.3 ± 4.2 mm in PSP patients, and 16.5 ± 3.7 mm in all APS patients. Widths of the third ventricle could not be measured in 2 of the 16 MSA patients, and widths of the frontal horns of the lateral ventricles could not be measured in 4 MSA patients.
BPS findings discriminating between IPD and APS.
Table 3 summarizes BPS findings differentiating IPD from APS. Of all 48 patients with adequate SN assessability, SN echogenicity correctly classified 45 (94%) cases as IPD or APS. The median SN echogenic size was different between Groups IPD and MSA (Mann–Whitney U test, p < 0.001), IPD and PSP (p < 0.001), and IPD and APS (p < 0.001) but was not different between Groups MSA and PSP (p = 0.8). Moderate or marked SN hyperechogenicity had a positive predictive value of 92% for the clinical diagnosis of IPD and a negative predictive value of 95% for not having the clinical diagnosis of IPD, that is, being diagnosed as APS. Sensitivity and specificity of moderate or marked SN hyperechogenicity for detecting patients with the clinical diagnosis of IPD were 96 and 91%. For marked SN hyperechogenicity, sensitivity, specificity, and positive predictive value for the clinical diagnosis of IPD were 76, 100, and 100%. Conversely, normal SN echogenicity indicated the clinical diagnosis of APS with a negative predictive value of 95%.
Table 3 Brain parenchyma sonography findings indicating idiopathic PD rather than atypical parkinsonian syndromes
Lenticular nucleus hyperechogenicity was more frequent in Group APS than in Group IPD (Mann–Whitney U test, p = 0.001). Normal echogenicity of lentiform nucleus had a positive predictive value, sensitivity, and specificity of 77% each for the clinical diagnosis of IPD. Conversely, hyperechogenicity of the lentiform nucleus had a positive predictive value, sensitivity, and specificity of 77% each for the clinical diagnosis of APS.
Widths of the third ventricle were larger in Group PSP than in Group IPD (t-test, p < 0.01) and in Group APS than in Group IPD (p < 0.01) but not different between Groups MSA and IPD (p > 0.05). A third ventricle width of <9 mm had a positive predictive value of 71% for the clinical diagnosis of IPD, with a sensitivity of 74% and a specificity of 67%.
Echogenicity of thalami and caudate nuclei and widths of frontal horns were not significantly different between Groups IPD and APS and did not contribute to discrimination between these groups. Furthermore, the combination of SN echogenicity and any other BPS findings did not improve the sensitivity or the specificity of differentiation between IPD and APS.
Discussion.
BPS data obtained in this study show that SN hyperechogenicity discriminates IPD from APS correctly in 94% of the assessable patients. SN hyperechogenicity has a positive predictive value of 92% for the diagnosis of IPD rather than APS. If SN hyperechogenicity is marked, APS can be excluded because of a positive predictive value of 100% for IPD. Conversely, when parkinsonian syndromes are investigated, normal SN echogenicity indicates APS with a high negative predictive value of 95%. In <5% of the patients, SN could not be assessed owing to bilaterally insufficient acoustic temporal bone windows.
BPS in APS patients revealed a median SN echogenic size of 0.15 cm2. This corresponds to median SN echogenic sizes of 0.14 cm2 for healthy subjects aged 50 to 59 years, 0.10 cm2 for subjects 60 to 69 years, and 0.15 cm2 for subjects 70 to 79 years and to median SN echogenic sizes for patients with nonparkinsonian cerebral disorders of 0.13 cm2.7,13⇓ BPS in our IPD patients showed a median SN echogenic size of 0.25 cm2. This confirms the previously reported median SN echogenic size of IPD patients of 0.25 cm2.6,7⇓ Assuming a normal SN echogenic size limit of 0.20 cm2,7,13,14⇓⇓ the sensitivity for IPD detection is 96% in our study as compared with 91 and 100% previously reported.6,7⇓ Comparing patients with IPD with age-matched control subjects, the specificity of SN hyperechogenicity was 93%5; comparing patients with IPD with nondegenerative cerebral disorders, it was 77%7; and comparing patients with IPD with APS in this study, it was 91%, suggesting a high overall specificity of SN hyperechogenicity for IPD.
SN hyperechogenicity in IPD was suggested to reflect abnormal iron accumulation in the SN,13,15⇓ a process associated with degeneration of the SN neurons.16,17⇓ An increase in total iron content in the SN is not specific to IPD but is also present in MSA and PSP.18 However, in MSA and PSP, the increased SN iron is suggested to be bound by increased amounts of ferritin, but in IPD, the SN iron increase is accompanied by a ferritin decrease, thus forcing iron into alternative binding.18 Therefore, increased iron content alone cannot be the reason for the increased SN echogenicity in IPD. Other factors such as iron-binding proteins may be important in generating SN hyperechogenicity.19
In the current study, hyperechogenicity of the lentiform nucleus was found in 77% of the APS patients but in only 23% of the IPD patients. This corresponds to recently reported lentiform nucleus hyperechogenicity in 75% of PSP patients and in 79% of MSA patients but in none of the IPD patients.20 Although lentiform nucleus hyperechogenicity and normal SN echogenicity are both highly specific for APS, their combination does not further improve discrimination of APS from IPD. For this purpose, BPS of SN alone, therefore, seems to be sufficient in routine BPS studies. Other structures investigated in this study were not contributory.
BPS is a novel and noninvasive method to differentiate highly specifically between IPD and APS. Therefore, BPS might become a standard investigation in parkinsonian disorders.
- Received May 23, 2002.
- Accepted September 10, 2002.
References
Letters: Rapid online correspondence
- Brain parenchyma sonography discriminates Parkinson’s disease and atypical parkinsonian
- Reply to Letter to the Editor
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Topics Discussed
Alert Me
Recommended articles
Sonographic discrimination of corticobasal degeneration vs progressive supranuclear palsy
Sonographic detection of basal ganglia lesions in asymptomatic and symptomatic Wilson disease
Transcranial brain sonography findings predict disease progression in multiple sclerosis
Relationship of substantia nigra echogenicity and motor function in elderly subjects