Dorsolateral nigral hyperintensity for differentiating Parkinson’s disease from atypical Parkinsonian syndromes: a meta-analysis
Abstract
Background: Parkinson’s disease (PD) is a progressive neurodegenerative disorder caused by the loss of dopaminergic neurons in the substantia nigra, leading to cardinal motor symptoms such as bradykinesia, rigidity, resting tremor, and postural instability. Differentiating idiopathic PD from atypical parkinsonian syndromes (APS) remains challenging, particularly in early stages, despite established clinical diagnostic criteria. The swallow-tail sign, corresponding to dorsolateral nigral hyperintensity (DNH) on susceptibility-sensitive MRI, has emerged as a potential imaging biomarker reflecting nigrosome-1 integrity. However, the absence of DNH has also been reported in APS, raising concerns regarding its specificity for PD. Objective: We aimed to determine the diagnostic accuracy of the DNH in distinguishing idiopathic PD from APS, providing pooled estimates of sensitivity and specificity. Methods: This systematic review and meta-analysis systematically searched in PubMed, Embase, CENTRAL, and clinical trial registries from their inceptions through February 2026, without language or date restrictions. All studies reporting diagnostic accuracy measures of the absence of DNH in patients with PD and APS syndromes were eligible regardless of study design. For inclusion, diagnoses of PD and APS had to be established according to current diagnostic criteria or confirmed by functional imaging using single-photon emission computed tomography or positron emission tomography with 18F-fluorodopa. Data were synthesized using a random-effects meta-analysis with both univariate and bivariate models based on logit-transformed data. The primary outcomes included sensitivity, specificity, negative (NPV) and positive (PPV) predictive values, negative (LR–) and positive (LR+) likelihood ratios, diagnostic odds ratio (DOR), and overall diagnostic accuracy. Results: In 17 studies (N = 1,803 patients), we estimated a pooled sensitivity, specificity, NPV, and PPV of 0.92 (95% CI, 0.87–0.94), 0.12 (95% CI, 0.05–0.28), 0.42 (95% CI, 0.20–0.67), 0.71 (95% CI, 0.64–0.78), respectively. Additionally, we found a LR– of 0.63 (95% CI, 0.35–1.10), a LR+ of 1.10 (95% CI, 0.99–1.30), a DOR of 1.94 (95% CI, 0.89–3.70), and an overall accuracy of 0.69 (95% CI, 0.62–0.76). APS etiologies was a significant covariate in the meta-regression model (P = .004), showing that the absence of DNH had a higher specificity for differentiating PD from cerebellar type multiple system atrophy (0.58 [95% CI, 0.34–0.79]) and corticobasal degeneration (0.34 [95% CI, 0.11–0.67]) compared with other APS. The overall area under the curve was 0.79 (95% CI, 0.65–0.86). Conclusion: Although we found high sensitivity for the absence of DNH, its consistently poor specificity limits its applicability in differentiating PD from APS. Additional diagnostic methods are required to rule out differential diagnoses.