Substantia nigra fractional anisotropy is not a diagnostic biomarker of Parkinson’s disease: A diagnostic performance study and meta-analysis
- 473 Downloads
Our goal was to estimate the diagnostic accuracy of substantia nigra fractional anisotropy (SN-FA) for Parkinson’s disease (PD) diagnosis in a sample similar to the clinical setting, including patients with essential tremor (ET) and healthy controls (HC). We also performed a systematic review and meta-analysis to estimate mean change in SN-FA induced by PD and its diagnostic accuracy.
Our sample consisted of 135 subjects: 72 PD, 21 ET and 42 HC. To address inter-scanner variability, two 3.0-T MRI scans were performed. MRI results of this sample were pooled into a meta-analysis that included 1,432 subjects (806 PD and 626 HC). A bivariate model was used to evaluate diagnostic accuracy measures.
In our sample, we did not observe a significant effect of disease on SN-FA and it was uninformative for diagnosis. The results of the meta-analysis estimated a 0.03 decrease in mean SN-FA in PD relative to HC (CI: 0.01–0.05). However, the discriminatory capability of SN-FA to diagnose PD was low: pooled sensitivity and specificity were 72 % (CI: 68–75) and 63 % (CI: 58–70), respectively. There was high heterogeneity between studies (I2 = 91.9 %).
SN-FA cannot be used as an isolated measure to diagnose PD.
• SN-FA appears insufficiently sensitive and specific to diagnose PD.
• Radiologists must be careful when translating mean group results to clinical practice.
• Imaging protocol and analysis standardization is necessary for developing reproducible quantitative biomarkers.
KeywordsParkinson’s disease/Parkinsonism Essential tremor Magnetic resonance imaging (MRI) Diffusion-weighted imaging (DWI) Sensitivity and specificity
The authors thank the Brazil Parkinson-Association for helping to contact volunteers during the recruitment phase; Adelinda da Silva Arruda Gonçalves, Karina Fernandes Dias Correa e Alda Fernandes Castro for administrative support; and Dr. Marcelo Buarque Gusmão Funari, Head of the Radiology Department of the Hospital Israelita Albert Einstein for supporting this study. We also thank the R-Community for providing and maintaining a free software environment for statistical computing and graphics and full financial support from FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo).
The scientific guarantor of this publication is Ellison Fernando Cardoso. The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article. This study has received support from FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) - Grant number: 2011/18747. Three of the authors have significant statistical expertise (João Ricardo Sato, Gilson Vieira and Ellison Fernando Cardoso). Institutional Review Board approval was obtained. Written informed consent was obtained from all subjects in this study. Methodology: prospective, diagnostic study, multicentre study.
- 26.Zhang G, Zhang Y, Zhang C et al. Research article diffusion kurtosis imaging of substantia nigra is a sensitive method for early diagnosis and disease evaluation in Parkinson’s disease. Parkinson’s disease. Hindawi Publishing Corporation. 2015; Article ID 207624: 1–5 doi: 10.1155/2015/207624
- 40.Pepe MS (2003) The statistical evaluation of medical tests for classification and prediction. Oxford University Press, OxfordGoogle Scholar