Three Stories of Ischemic Stroke: Aging Dependence, Carotid Procedures, and Asymptomatic Carotid Disease

Aging-dependent diseases are largely a consequence of the underlying chronic degenerative accompaniments of aging with incidences expected to increase exponentially with age. Ischemic stroke should be aging because much of the underlying pathology is a consequence of aging. National Hospital Discharge Survey (NHDS) from 1970 to 2009 were analyzed to define secular trends in sex- and age-specific ischemic stroke incidence. Over the 40 years, sex-specific distributions demonstrated exponentially increasing ischemic stroke incidence with age supporting the notion of ischemic stroke as an aging-dependent disease. Using NHDS data we next examined carotid endarterectomies (CEA) performed in the United States from 1970 through 2010, and the recent introduction of carotid angioplasty and stenting (CAS). Two substantial rises in CEA rates followed by subsequent falls were evident over the four decades. Adoption of CAS was limited. Juxtaposing rates against highly cited studies suggests enthusiasm driven increases tempered by observational study and trial results. Lastly, we examined the potential benefit of CEA for asymptomatic patients in the setting of improved medical care since completion of the Asymptomatic Carotid Surgery Trial in 2004. Nine studies were identified reporting ipsilateral stroke rates in patients with asymptomatic carotid disease published after 2004. Seven included patients who were not likely candidates for CEA. Other biases were prominent in all but one study. The pooled annual ipsilateral stroke rate was 0.98% (95% Credible Interval [CrI]: 0.54 to 1.57); excluding the three most biased studies 1.36% (95% CrI: 0.88 to 1.83). The best estimate of ipsilateral stroke incidence in patients with significant asymptomatic carotid stenosis is somewhat greater than 1%. Using a Monte Carlo approach, we then synthesized evidence and combined it with decision models to compare CEA outcomes with contemporary medical care. Estimates from Bayesian meta-analyses of ipsilateral stroke rates and relative risk reduction of stroke following CEA were incorporated in Markov cohort models over a five-year time horizon. Compared with medical care, to gain expected quality adjusted life years, periprocedural CEA mortality must approach 0% with stroke rates lower than 1.5%.