Understanding and improving the diagnosis of dementia with Lewy bodies

Abstract

Background Accurate diagnosis of dementia with Lewy bodies (DLB) has important implications for treatment and prognosis, but it is not currently clear how frequently DLB is diagnosed in routine clinical practice, nor how frequently they are assigned an alternative dementia diagnosis. 123I-metaiodobenzylguanidine (MIBG) may be capable of improving DLB diagnostic accuracy but it has not been investigated in clinically representative populations that include patients with comorbidities or interfering medications. Methods We conducted a cross-sectional survey of 5 569 patients attending three Psychiatry of Old Age services. From this cohort, 51 DLB and 51 matched non-DLB cases consented to extraction of data relating to the diagnostic process from their clinical case notes. We enrolled a clinically representative cohort of 17 patients with DLB and 16 with Alzheimer’s disease (AD) to a MIBG utility study. Each patient underwent detailed clinical examination, cardiac MIBG and FP-CIT SPECT imaging. Results DLB represented 5.6% of dementia cases but prevalence varied across services (3.5-5.9%). DLB cases were often given a different dementia subtype diagnosis (39%) and experienced a longer time from referral to diagnosis (265 days) than non-DLB patients (154 days). MIBG had a sensitivity and specificity of 71% and 75% for differentiating DLB from AD, but a lower HMR threshold enhanced specificity (100%) without compromising sensitivity. No significant relationships between HMR and either myocardial infarction, or medication prescription, were identified. ii Conclusions Variation in DLB prevalence across services may suggest differences in detection rather than in the true prevalence of the disease. The higher frequency of clinical contacts seen in DLB may provide opportunities to improve both diagnostic accuracy and time to diagnosis. Our findings support the use of representative cohorts in further MIBG research, particularly in determining appropriate HMR cut-offs. Our finding that three DLB patients had normal MIBG, but abnormal FP-CIT results challenges the Braak hypothesis of DLB pathogenesis