Transcranial Imaging of Substantia Nigra Hyperechogenicity ina Taiwanese Cohort of Parkinson’s Disease

Yu-Wen Huang, MD,1 Jiann-Shing Jeng, MD, PhD,1 Chung-Fen Tsai, MD,2 Li-Ling Chen, BS,1

and Ruey-Meei Wu, MD, PhD1*

1Department of Neurology, National Taiwan University Hospital, College of Medicine, National Taiwan University,Taipei, Taiwan

2Department of Neurology, Cardinal-Tien Hospital, Taipei, Taiwan

Abstract: Transcranial Doppler imaging (TCDI) has been usedas a noninvasive diagnostic tool to differentiate Parkinson’sdisease (PD) from atypical parkinsonism by detecting hypere-chogenicity in the substantia nigra (SN). To our knowledge, noTCDI data are available for Asian populations, and TCDIsensitivity is uncertain across populations. Early-onset PD(EOPD) represents a specific PD subtype based on clinicalfeatures and pathogenic mechanisms. It is not known if EOPDpatients have abnormal echogenicity in SN comparable tolate-onset PD (LOPD) patients. We assessed the area of SNhyperechogenicity (hyper-SN) and a ratio of hyper-SN overipsilateral midbrain (S/M ratio) with TCDI in 164 healthy

Taiwanese, 40 EOPD patients, and 40 LOPD patients. Theupper 95th percentile values for hyper-SN and S/M ratio were0.20 cm2 and 0.07. Our results indicate that S/M ratio is a moresensitive measure than hyper-SN in diagnosing PD. Approxi-mately 92.5% of the LOPD patients and 57.5% of the EOPDpatients had S/M ratios � 0.07. Enlarged hyperechogenicity ofSN is a common finding in LOPD, but not in EOPD. Iron-independent mechanisms of SN cell degeneration in EOPDdistinct from that in LOPD might contribute to the sonographicfindings. © 2007 Movement Disorder Society

Key words: transcranial doppler image; echogenicity; sub-stantia nigra; midbrain; Parkinson’s disease; Taiwanese.

Transcranial Doppler imaging (TCDI) has long beenused as a diagnostic tool for cerebrovascular disorders.Its recent application in patients with extrapyramidaldisorders has identified varying degrees of altered echo-genicity in basal ganglia.1,2 An enlarged area of hypere-chogenic signal in substantia nigra (SN) region is char-acteristic in Parkinson’s disease (PD) and rarely seen inother atypical parkinsonian syndromes.3 However, sono-graphic measurement of hyperechogenic areas in SNmay be challenged due to the small target size. To date,the use of TCDI for diagnosing PD in Asian populationshas not been evaluated. Racial differences in brain size4

and temporal skull thickness5 in Asians may yield dif-ferent values when using TCDI to measure midbrain.

An early-onset form of PD (EOPD), arbitrarily clas-sified as occurring before age 45 or 50, is recognized asa unique subtype of PD, frequently manifested withdystonia, slow progression, and early appearance oflevodopa-related dyskinesia.6 Our recent hospital-basedstudy found that approximately 16% of PD patients di-agnosed in our Movement disorders clinic had onsetprior to age 50.7 Recent work has shown that a subgroupof these patients has an autosomal recessive juvenileonset PD (AR-JPD) with parkin or PINK1 gene muta-tions.8,9 Neuropathological reports of several EOPDcases with parkin mutations showed absence of Lewybodies, and led to speculation that the etiology of AR-JPD differs significantly from that of PD.10 Therefore, itwas worthwhile to know whether SN hyperechogenicityis commonly observed in EOPD.

The present study represents the first comprehensiveevaluation of transcranial sonographic data for measur-ing SN hyperechogenic regions (hyper-SN) and midbrainin a large group of healthy Taiwanese. Furthermore, weinvestigated the sensitivity of SN hyperechogenicity for

*Correspondence to: Ruey-Meei Wu, Department of Neurology,National Taiwan University Hospital, No. 7 Chung-Shan South Road,Taipei 100, Taiwan. E-mail: rmwu@ha.mc.ntu.edu.tw

Received 14 November 2006; Accepted 19 November 2006Published online 26 January 2007 in Wiley InterScience (www.

interscience.wiley.com). DOI: 10.1002/mds.21372

Movement DisordersVol. 22, No. 4, 2007, pp. 550–555© 2007 Movement Disorder Society

550

diagnosing PD in LOPD and EOPD patients. Many dis-tinguishable features of EOPD were identified by clinicalphenotype or genetic etiology.

PATIENTS AND METHODS

Subjects

A total of 282 individuals, including 192 control sub-jects and 90 Parkinson patients followed up at MovementDisorders Clinic of National Taiwan University, Taipei,Taiwan, were included for this study. All participantsgave informed consent according to the Declaration ofHelsinki. Volunteers with no evidence of CNS disorderswere selected from the hospital’s neurological clinic ascontrol subjects for this TCDI study. The diagnosis ofPD was established in accordance with the PD criteria,11

which includes the presence of at least two (possible PD)or three (probable PD) of the four cardinal features(resting tremor, bradykinesia, rigidity, and asymmetriconset), with substantial and sustained response to L-dopaor a dopamine agonist. In addition, PD patients werecategorized as LOPD (onset � 50 years) or EOPD (on-set � 50 years) for further clinical and sonographicstudies. All subjects underwent a detail history takingand thorough clinical examinations. Patient evaluationsincluded the Unified PD Rating Scale, Hoehn & Yahrstaging, the Folstein Mini Mental State Examination, astandard neurological examination, head CT or magneticresonance imaging (MRI), and a TCDI examination.

Transcranial Doppler Imaging

TCDI was performed by one of the authors (Y.W.H.and C.F.T.) who were blind to the clinical information. Acolor-coded, phased-array ultrasound system equippedwith a 2.25 MHz transducer (HP-4500 system) was usedto examine echogenicity in the midbrain and SN throughthe temporal bone with a penetration depth of 16 cm anda dynamic range of 45 dB. As described previously,12 theexamined scanning plane was standardized to display themargin of butterfly-shaped hypoechogenic midbrain sur-rounded by hyperechogenic basal cisterns as clear aspossible by mildly adjusting the probe angle. We man-ually encircled the midbrain outline and the interiorhyperechogenic areas of the SN through the echo-tracingmethod. The midline of midbrain was determined bydrawing an intersectional line passing through the inter-cerebral peduncle and the point of hyperechogenic aque-duct. The respective areas of SN and ipsilateral midbrainwere measured bilaterally (shown in Fig. 1). Moreover,considering that the imaged cross-section of midbrainmight vary in sequential studies due to shifts in theposition and angle of the probe as well as atrophic

changes in midbrain due to age,13 we simultaneouslyused the ipsilateral midbrain as an internal control pa-rameter, and calculated a hyper-SN/ipsilateral midbrainratio (S/M ratio) for further analysis.

Statistical Analysis

An intra-class correlation coefficient (ICC) was usedto determine intra- and inter-observer reliability of mid-brain measurements. Reliability coefficients with 95%confidence intervals (CIs) were calculated for 20 con-trols. The hyper-SN and ipsilateral midbrain areas wererecorded, and an S/M ratio was calculated. Data arereported as means and standard deviation. Estimates ofsensitivity, specificity, positive predictive value (PPV),and negative predictive value (NPV) were calculated forhyper-SN and S/M ratio in LOPD and EOPD,respectively.

The Pearson correlation coefficient (�) was used toevaluate correlations between age and S/M ratio. To givethe high specificity of TCDI in diagnosing PD, the toptwo standard deviations (upper 95th percentile) for hy-per-SN and S/M ratio were used as cut-off points fordefining abnormal extension of hyperechogenic signalsin the SN region (denoting as SN�). To evaluate theassociation between clinical manifestations and hyper-SN, LOPD and EOPD patients were classified into twogroups depending on presence or absence of SN�. Inter-

FIG. 1. Transcranial Doppler imaging (TCDI) of the midbrain througha temporal bone window in a control subject. We identified the but-terfly-shaped, hyperechogenic midbrain (dotted line), and manuallyencircled the interior hyperechogenic area of substantia nigra (thicksolid line) and also ipsilateral midbrain (thin solid line) by the midbrainoutline and an intersectional line passing through the aqueduct (arrow)and inter-cerebral peduncle (arrow head). The respective areas weremeasured bilaterally.

SUBSTANTIA NIGRA ECHOGENICITY IN PARKINSONISM 551

Movement Disorders, Vol. 22, No. 4, 2007

group comparisons were conducted by two-sample, un-paired t tests for continuous variables with normal dis-tributions, and �2 tests for discrete variables. The datawere analyzed with the SPSS 12.0 statistical package(Chicago, IL). Statistical significance was set at P �0.05.

RESULTS

Data from 20 controls were used to evaluate intra-and inter-observer variability. Intra-observer ICCs forhyper-SN, ipsilateral midbrain measurements, andS/M ratios were 0.98 (0.96 – 0.99), 0.79 (0.63– 0.88),and 0.98(0.96 – 0.99) and the corresponding interob-server ICCs were 0.98 (0.96 – 0.99), 0.68 (0.35– 0.86),and 0.96 (0.89 – 0.98).

A total of 192 controls were examined, and 28 wereexcluded due to insufficient temporal bone windows. Thecontrol group, a total of 164 subjects, consisted of 96men and 68 women (52.4 � 14.7 years; range 19–83).On the right side, hyper-SN was 0.066 � 0.066 (0–0.256) cm2, and the S/M ratio was 0.023 � 0.023 (0–0.091). Corresponding values for the left side were0.066 � 0.068 (0–0.266) cm2 and 0.023 � 0.024 (0–0.087). There were no significant differences betweenright and left sides. As shown in Figure 2 there was apositive correlation between S/M ratio and age (� �0.546, P � 0.001). This finding is compatible reports ofage-related changes in the midbrain.13 The upper 95%values for hyper-SN and S/M ratio were 0.20 cm2 and0.07. As aforementioned, we defined abnormal extensionof SN hyperechogenicity (SN�) as any subject with ahyper-SN exceeding 0.20 cm2 or an S/M ratio higherthan 0.07 on one or both sides.

Fifty LOPD and 40 EOPD patients were examined.Ten LOPD patients were excluded due to poor temporalbone windows. Demographic and sonographic data for40 LOPD and 40 EOPD patients and their age- andsex-matched controls are shown in Table 1. Both LOPDand EOPD patients had a significantly larger hyper-SNand S/M ratio than controls, but showed no obviousdifference in midbrain size. Considering SN�, a cutoffof hyper-SN �0.20 cm2 showed high sensitivity (85%),specificity (85%), PPV (85%), and NPV (85%) inLOPD; and high specificity (100%) and PPV (100%), butrelatively low sensitivity (50%) and high NPV (80%) inEOPD. A cutoff of S/M ratio � 0.07 showed highsensitivity (92.5%), specificity (82.5%), PPV (92%), andNPV (82.1%) in LOPD; and high specificity (97.5%) andPPV (92%), but again, relatively low sensitivity (57.5%)and high NPV (82.1%) in EOPD. The S/N ratio has

FIG. 2. Relationship between S/M ratio and age in 164 control sub-jects. The Pearson correlation coefficient is 0.546 (P � 0.001). TheS/M ratio represents the area ratio of the substantia nigra (SN) hypere-chogenic area and the size of the ipsilateral midbrain detected bytranscranial Doppler imaging (TCDI).

TABLE 1. Demographic and sonographic data in late-onset (LOPD) and early-onset (EOPD)Parkinson’s disease patients and controls

Variables LOPD (n � 40) Controls (n � 40) EOPD (n � 40) Controls (n � 80)

Age (yr) 69.0 � 6.5 68.7 � 6.7 49.2 � 7.8 48.9 � 7.8Sex, M/F 25/15 25/15 24/16 48/32Onset age (yr) 62.5 � 6.6 40.8 � 8.6Disease duration (yr) 6.5 � 4.9 8.4 � 6.4Hoehn & Yahr stage 2.350 � 1.027 2.025 � 0.947Hyper-SN, right (cm2) 0.274 � 0.135* 0.108 � 0.069 0.126 � 0.137* 0.049 � 0.049Hyper-SN, left (cm2) 0.276 � 0.119* 0.118 � 0.064 0.180 � 0.143* 0.047 � 0.051Midbrain, right (cm2) 2.847 � 0.307 2.713 � 0.459 2.574 � 0.472 2.708 � 0.345Midbrain, left (cm2) 2.646 � 0.295 2.834 � 0.445 2.612 � 0.496 2.727 � 0.409S/M ratio, right 0.095 � 0.044* 0.041 � 0.026 0.050 � 0.053* 0.018 � 0.019S/M ratio, left 0.104 � 0.046* 0.042 � 0.023 0.072 � 0.055* 0.017 � 0.019

Hyper-SN: substantia nigra (SN) area of hyperechogenicity; S/M ratio: Hyper-SN/ipsilateral midbrain area ratio (seeFig. 1).

*P � 0.001, unpaired t test, as compared with age- and gender-matched controls.

552 Y.-W. HUANG ET AL.

Movement Disorders, Vol. 22, No. 4, 2007

higher sensitivity and positive predictive value than thatof absolute value of hyper-SN alone. We thus use S/Mratio � 0.07 as SN� cutoff in further analyses.

To evaluate the relationship between clinical manifes-tation and SN�, we compared demographic and clinicalfeatures in LOPD and EOPD patients with or withoutabnormal extension of SN (Table 2). In both LOPD andEOPD groups, patients with SN� were older than thosewithout SN� (LOPD: 69.9 � 5.9 years vs. 58.0 � 4.4years, P � 0.001; EOPD: 51.6 � 6.8 years vs. 45.9 � 8.2years, P � 0.023). Patients with SN� also had longerdisease duration than those without SN� (LOPD: 6.9 �4.9 years vs. 1.7 � 0.6 years, P � 0.001; EOPD: 10.8 �6.8 years vs. 5.2 � 4.1 years, P � 0.004). Other clinicalfeatures were not significantly different for LOPD pa-tients (Table 2). However, in EOPD patients, dyskinesiawas significantly more frequent in patients with SN�(60.9% vs. 29.4%, P � 0.049).

DISCUSSION

The present study is the first to investigate the size ofthe midbrain and SN hyperechogenic area by TCDI in alarge group of Asians. In agreement with previous stud-ies,14 we found that TCDI was a reliable and reproduc-ible method. Our concern that the scanned axial plane ofthe midbrain would vary on each examination was sup-ported by our finding of relatively lower intra- and inter-observer ICCs for midbrain measurements (0.79 and0.68) than for hyper-SN (0.98 and 0.98) and S/M ratio(0.98 and 0.96). The consistency in S/M ratios indicatesthat hyper-SN also varies with ipsilateral midbrain size.

S/M ratio was as reliable for assessing SN� as hyper-SNalone. In addition, our LOPD data regarding sensitivity,specificity, PPV and NPV are about the same as thatrecently described by Prestel et al., who assessed 6 of 35controls and 36 of 42 LOPD patients and defined SN >0.2 as an abnormal extension of hyperechogenicity.15 Intheir study, the sensitivity is 85.7%, specificity 82.9%,positive predictive value (PPV) 85.7%, and negativepredictive value (NPV): 82.9%. These findings furthersupport the reliability and reproducibility of TCD studyas a useful tool in the diagnosis of sporadic and old-onsetParkinson’s disease.

We calculated a smaller hyper-SN in Asian controls(0.20 cm2, 95th percentile) than was reported in Cauca-sian populations. Berg and coworkers found that the 90thpercentile for hyper-SN in their 330 healthy volunteerswas 0.25 cm2,14 and the 75th percentile in 93 subjectsover 60 years was 0.20 cm2.16 Our data might suggestthat there are racial–ethnic differences in hyper-SN inAsians and Caucasians, although the values of controlsshould be established in individual laboratories. Thepresent study provides a valuable reference for TCDI asan adjuvant tool for diagnosing PD in Asian populations.However, control values should be obtained in eachsituation, and it should be understood that differences inmeasuring technique might be associated with differ-ences in sonographic results.

In this study, we showed that SN� in PD patients waspositively correlated with age and disease duration butnot with motor symptoms, disease severity, L-dopa dos-

TABLE 2. Comparison of clinical features of PD patients with or without abnormal extension ofSN hyperechogenicity (S/M ratio � 0.07)

Clinical characteristics

LOPD EOPD

S/M � 0.07(n � 37)

S/M � 0.07(n � 3) P value

S/M � 0.07(n � 23)

S/M � 0.07(n � 17) P value

Age (yr) 69.9 � 5.9 58.0 � 4.4 0.001* 51.6 � 6.8 45.9 � 8.2 0.023*Age at onset (yr) 63.0 � 6.5 56.3 � 4.0 0.092 40.7 � 8.3 40.8 � 9.3 0.993Duration (yr) 6.9 � 4.9 1.7 � 0.6 �0.001* 10.8 � 6.8 5.2 � 4.1 0.004*Hoehn & Yahr stage 2.4 � 1.1 2.0 � 0 0.551 2.2 � 0.9 1.8 � 0.9 0.252L-dopa dose (mg/day) 452.7 � 247.2 333.3 � 152.8 0.423 556.5 � 313.1 379.4 � 301.6 0.080Akinesia-rigidity form 12 (32.4%) 1 (33.3%) 0.974 23 (100%) 16 (94.1%) 0.239Tremor-predominant form 25 (67.6%) 2 (66.7%) 0.974 0 (0%) 1 (5.9%) 0.239Family history 6 (16.2%) 1 (33.3%) 0.453 7 (30.4%) 4 (23.5%) 0.629Asymmetric symptoms 20 (54.1%) 3 (100%) 0.122 8 (34.8%) 8 (47.1%) 0.433Tremor 21 (56.8%) 2 (66.7%) 0.738 16 (69.6%) 9 (52.9%) 0.283Dystonia 20 (54.1%) 2 (66.7%) 0.673 9 (39.1%) 4 (23.5%) 0.298Postural instability 11 (29.7%) 0 (0%) 0.267 5 (21.7%) 3 (17.7%) 0.749Motor fluctuation 8 (21.6%) 0 (0%) 0.368 16 (67.0%) 8 (47.0%) 0.151Dyskinesia 8 (21.6%) 0 (0%) 0.368 14 (60.9%) 5 (29.4%) 0.049*

S/M ratio: Hyper-SN/ipsilateral midbrain area (see Table 1).*Statistically significant difference between groups, P � 0.001, unpaired t test.PD, Parkinson’s disease; LOPD, late-onset PD; EOPA, early-onset PD.

SUBSTANTIA NIGRA ECHOGENICITY IN PARKINSONISM 553

Movement Disorders, Vol. 22, No. 4, 2007

age, or L-dopa-related motor complications. This is inaccordance with previous reports.17 However, the findingthat patients with SN� had longer disease duration thanthose without SN� is not supported in the literature. Arecent paper by Berg’s group revealed that the area ofhyper-SN did not change at a five-year follow-up in their27 Parkinson patients.18 This sonographic finding wasthus proposed a trait marker for nigrostriatal vulnerabil-ity. The finding that PD patients with SN� had longerdisease duration than patients without SN may simply bedue to the fact that in patients with shorter diseaseduration the diagnosis was not entirely clear. It is wellknown that, even in movement disorder specialist clinics,up to 15–18% of patients with a Parkinsonian syndromeare given the wrong diagnosis (idiopathic PD vs. atypicalPD) during lifetime. Therefore, it may well be that therewere subjects with atypical PD among the group withoutSN hyperechogenicity, who were misdiagnosed as idio-pathic PD. As diagnostic accuracy increases with dura-tion of the disease, in the group with longer lastingdisease progression, SN hyperechogenicity and diagnosisof idiopathic PD may have been more concordant. How-ever, the relationship between the change of hyper-SNarea and the course of disease duration need furtherclarification in larger cohorts of PD with longitudinalfollow up.

The origin of SN hyperechogenecity is not fully un-derstood. Animal studies19 and postmortem TCDI withhistological examination2 suggest that hyperechogenicityis most likely a site of iron deposition undergoing Fentonreactions to form free radicals. The increased iron con-tent in SN in parkinsonian brains is well established, butthere are inconsistencies in reports of iron concentrationand the time of its emergence in the disease.20 Our datasupport the notion that SN iron levels reflecting as thehyperechogenicity area increase with age.

The frequency of SN� in EOPD patients (57.5%) wassignificantly lower than in LOPD patients (92.5%; Table2). This finding might suggest a pathogenic mechanismof nigral cell death that is not attributable to increases ofiron in EOPD brains. This assumption is bolstered bycases of parkinsonism induced by MPTP21 and pesticiderotenone22 in young persons and AR-JPD.

Among our EOPD patients, 28 were screened forparkin and PINK1 genes. Four patients were positive forparkin mutations on both alleles (all compound-het-erozygous). Two patients had alternate allelic deletionsof parkin exons 3 and 4,23 one patient had a parkin exon2 deletion and an exon 7 G284R substitution,7 and onepatient harbors a deleted exon 4 and a missense pointmutation in C441R. All four patients exhibited expansionof SN� bilaterally. In terms of PINK1 mutations, one

patient had one mutated allele, and did not display SN�.A recent report detected increased hyper-SN in symp-tomatic parkin mutation carriers.24 In a few autopsiedcases with parkin mutation, there was an absence ofLewy bodies10 and aggregates of �-synuclein in thecytoplasm that were probably promoted by iron.25 There-fore, we speculate that Lewy bodies are not a majorcausative factor in hyper-SN in parkin-proven cases.Further studies are necessary to elucidate the role ofparkin in the generation of SN hyperechogenicity. PDpatients with PINK1 mutations have a loss of presynapticdopaminergic function,26 although no autopsy studies areyet available. PINK1, a protein kinase with an unknownsubstrate, appears to protect cells against mitochondrialdamage.9 There are still some gaps in our knowledgeabout the mechanisms of PINK1-related parkinsonism.Further investigation will help clarify the change of SNechogenicity in EOPD patients with PINK1 mutations.

In conclusion, our results show that TCDI is reliablefor assessing SN echogenicity. Furthermore, we demon-strate the usefulness of S/M ratios for defining SN hy-perechogenicity. Assessment of SN hyperechogenicity isuseful for the diagnosis of PD in LOPD patients, butshould be carefully evaluated in EOPD patients.

REFERENCES

1. Becker G, Seufert J, Bogdahn U, Reichmann H, Reiners K. De-generation of substantia nigra in chronic Parkinson’s disease visu-alized by transcranial color-coded real-time sonography. Neurol-ogy 1995;45:182-184.

2. Berg D, Roggendorf W, Schroder U, et al. Echogenicity of thesubstantia nigra: association with increased iron content andmarker for susceptibility to nigrostriatal injury. Arch Neurol 2002;59:999-1005.

3. Walter U, Niehaus L, Probst T, Benecke R, Meyer BU, Dressler D.Brain parenchyma sonography discriminates Parkinson’s diseaseand atypical parkinsonian syndromes. Neurology 2003;60:74-77.

4. Rushton JP, Rushton EW. Brain size, IQ, and racial-group differ-ences: evidence from musculoskeletal traits. Intelligence 2003;31:139-155.

5. Halsey JH. Effect of emitted power on wave-form intensity intranscranial doppler. Stroke 1990;21:1573-1578.

6. Schrag A, Ben Shlomo Y, Brown R, Marsden CD, Quinn N.Young-onset Parkinson’s disease revisited—clinical features, nat-ural history, and mortality. Mov Disord 1998;13:885-894.

7. Wu RM, Bounds R, Lincoln S, et al. Parkin mutations and early-onset parkinsonism in a Taiwanese cohort. Arch Neurol 2005;62:82-87.

8. Kitada T, Asakawa S, Hattori N, et al. Mutations in the parkin genecause autosomal recessive juvenile parkinsonism. Nature 1998;392:605-608.

9. Valente EM, Abou-Sleiman PM, Caputo V, et al. Hereditary early-onset Parkinson’s disease caused by mutations in PINK1. Science2004;304:1158-1160.

10. Mori H, Kondo T, Yokochi M, et al. Pathologic and biochemicalstudies of juvenile parkinsonism linked to chromosome 6q. Neu-rology 1998;51:890-892.

11. Gelb DJ, Oliver E, Gilman S. Diagnostic criteria for Parkinsondisease. Arch Neurol 1999;56:33-39.

554 Y.-W. HUANG ET AL.

Movement Disorders, Vol. 22, No. 4, 2007

12. Becker G, Berg D. Neuroimaging in basal ganglia disorders: per-spectives for transcranial ultrasound. Mov Disord 2001;16:23-32.

13. Sohmiya M, Tanaka M, Aihara Y, Hirai S, Okamoto K. Age-related structural changes in the human midbrain: an MR imagestudy. Neurobiol Aging 2001;22:595-601.

14. Berg D, Becker G, Zeiler B, et al. Vulnerability of the nigrostriatalsystem as detected by transcranial ultrasound. Neurology 1999;53:1026-1031.

15. Prestel J, Schweitzer KJ, Hofer A, Gasser T, Berg D. Predictivevalue of transcranial sonography in the diagnosis of Parkinson’sdisease. Mov Disord 2006;21:1763-1765.

16. Berg D, Siefker C, Ruprecht-Dorfler P, Becker G. Relationship ofsubstantia nigra echogenicity and motor function in elderly sub-jects. Neurology 2001;56:13-17.

17. Berg D, Siefker C, Becker G. Echogenicity of the substantia nigrain Parkinson’s disease and its relation to clinical findings. J Neurol2001;248:684-689.

18. Berg D, Merz B, Reiners K, Nauman M, Becker G. Five-yearfollow-up study of hyperechogenicity of the substantia nigra inParkinson’s disease. Mov Disord 2005;20:383-385.

19. Berg D, Grote C, Rausch WD, et al. Iron accumulation in thesubstantia nigra in rats visualized by ultrasound. Ultrasound MedBiol 1999;25:901-904.

20. Berg D, Hochstrasser H. Iron metabolism in Parkinsonian syn-dromes. Mov Disord 2006;21:1299-1310.

21. Di Monte DA, Lavasani M, Manning-Bog AB. Environmentalfactors in Parkinson’s disease. Neurotoxicology 2002;23:487-502.

22. Betarbet R, Sherer TB, MacKenzie G, Garcia-Osuna M, PanovAV, Greenamyre JT. Chronic systemic pesticide exposure repro-duces features of Parkinson’s disease. Nat Neurosci 2000;3:1301-1306.

23. Wu RM, Shan DE, Sun CM, et al. Clinical, F-18-dopa PET, andgenetic analysis of an ethnic Chinese kindred with early-onsetparkinsonism and parkin gene mutations. Mov Disord 2002;17:670-675.

24. Walter U, Klein C, Hilker R, Benecke R, Pramstaller PP, DresslerD. Brain parenchyma sonography detects preclinical parkinsonism.Mov Disord 2004;19:1445-1449.

25. Ostrerova-Golts N, Petrucelli L, Hardy J, Lee JM, Farer M, Wolo-zin B. The A53T �-synuclein mutation increases iron-dependentaggregation and toxicity. J Neurosci 2000;20:6048-6054.

26. Khan NL, Valente EM, Bentivoglio AR, et al. Clinical and sub-clinical dopaminergic dysfunction in PARK6-linked parkinsonism:an F-18-dopa PET study. Ann Neurol 2002;52:849-853.

SUBSTANTIA NIGRA ECHOGENICITY IN PARKINSONISM 555

Movement Disorders, Vol. 22, No. 4, 2007