Sensitivity and specificity of transcranial magnetic stimulation for differential diagnosis of Alzheimer’s disease and frontotemporal dementia

Background: Early and differential diagnosis of Alzheimer’s disease (AD) and Frontotemporal Dementia (FTD) still remains problematic. Cerebrospinal fluid (CSF) biomarkers, PET amyloid imaging and hippocampal atrophy on MRI have shown the highest accuracy in post-mortem studies of AD cases. Notwithstanding, these biomarkers present considerable drawbacks. In this view, transcranial magnetic stimulation (TMS) allows the assessment, non-invasively and in vivo, of specific neurotransmission signatures. Methods: We performed a prospective study on AD and FTD patients, in accordance with the Standards for Reporting of Diagnostic Accuracy. Paired pulse TMS was used to investigate short-interval intracortical inhibition (SICI) and facilitation (ICF), long-interval intracortical inhibition (LICI) and short-afferent latency inhibition (SAI) in order to measure the activity of different intracortical circuits, with the application of current reference standards, to differentiate patients with AD from those with FTD and healthy controls (HC). The primary outcome measures were sensitivity and specificity of TMS measures, derived from receiver operator curve analysis. Results: A total of 175 subjects met inclusion criteria. We diagnosed 79 AD patients, 64 FTD patients and 32 HC. We found that while AD patients are characterized by a specific impairment of SAI, FTD show a remarkable dysfunction of SICI-ICF intracortical circuits. Using the best index incorporating these measurements (SICI-ICF / SAI ratio), TMS differentiated FTD from AD with a sensitivity of 91·8% and specificity of 88·6%, AD from HC with a sensitivity of 84·8%, and specificity of 90·6%, and FTD from HC with a sensitivity of 90·2% and specificity of 78·1%. These results were confirmed also in patients with a mild disease. Conclusions: TMS is a non-invasive procedure which reliably distinguishes AD from FTD and HC and, if these findings are replicated in larger studies, could represent a useful adjunctive diagnostic tool to be used in clinical practice