Usefulness of nigrosomes imaging at 3T in daily practice

The loss of the “swallow tail sign” on susceptibility weighted images (SWI) has been described to be a reliable marker of the neurodegeneration of the nigrosomes within the substantia nigra. This sign was observed in Parkinson's and other related diseases such as dementia with Lewy bodies, progressive supranuclear palsy, multisystem atrophy (PD&RD with an approximative prevalence of 0.5% into the general population). High diagnostic performances were reported in paired comparative studies. However, we hypothetized that positive predictive values (PPV) were overestimated due to an excessive PD&RD prevalence within studies samples. The main objective of the study was to estimate the diagnostic performance of this sign in daily practice. We conducted a single center retrospective study on SWI acquired on two 3T MR scans for various medical reasons in our department of neuroradiology. All examinations were reviewed independently by 4 readers, blinded of the diagnosis. The sign was defined by the loss of at least one nigrosome, assessed by at least 3 readers. Examinations were excluded in case of excessive artefacts or obvious midbrain lesions. After selection, 1010 individual examinations were reviewed for patients referred for neuro-cognitive disorders unlikely of PD&RD (n=957; 94.8%); neuro-cognitive disorders compatible with PD&RD (n=53; 5.2%). The PD&RD prevalence was 5%, still 10x higher than in the general population, but much lower than in previous paired comparative studies. The sign had a sensitivity of 59.6%, a specificity of 90.0%, a PPV of 25.0%, and a negative predictive value (NPV) of 97.5% Even if the sign is often observed in PD&RD, inconstantly though, the low risk to observe it in non-PD&RD patients, drastically weaken the PPV because of their much higher prevalence in daily practice and of individual or artifactual reasons. Also, normal nigrosomes appearance is highly predictive of non-PD&RD, it could be observed in PD&RD. Nigrosome imaging has to be taken into account with additional data.