Cerebrovascular Reserve, Dementia Risk and Renal Function
Harold T. Pretorius, M.D., Ph.D^+, Christopher Kircher, M.D.^, Deborah R. Gustafson, M.S., Ph.D.**, Luis F. Pagani, M.D.^
Abstract Background: Relationships among dementia risk factors are well established: diabetes and related complications, nephropathy and retinopathy, coexist with hypertension, obesity and coronary disease. Increasing evidence implicates brain hypoperfusion as an early and potentially treatable aspect of dementia risk. In this context, the relation of renal dysfunction to cerebrovascular reserve is pertinent in vascular risk subgroups, as well as brain trauma and metal toxicity.
Methods: Outpatients (60% women, 40% men; 65% Caucasian; ages 25-91 years, average 55+-10 years,) with cerebral ischemia or related symptoms (eg. TIA, cognitive impairment, post concussion, complex migraine) were injected with radiolabelled tracers, followed brain SPECT. Cortical metabolic and perfusion indices (CMi, CPi) for basal and perfusion-stimulated SPECT were scaled similarly to rest and stress ventricular ejection fractions: normal CMi 50-72%, normal CPi 53-75% and normal cerebrovascular reserve (CR) if CPi > (CMi+3)%. Quantitative urine porphyrins screened for metal toxicity. Renal function was calculated by MRDD. Patients with type 2 diabetes were treated to average HbA1c 8%; blood pressures were treated to < 140/90 mm Hg.
Results: CR was abnormal in 64% (51/80) of hypertensives, 61% (54/89) of diabetics, 64% (32/50) of insulin resistant nondiabetics, 57% (16/28) of mild brain trauma patients and 58% (42/78) of metal toxic patients. CR was abnormal in 50% (1/2) of patients with GFR < 15 ml/min; but normal in 90% (18/20) of patients with GFR 15 to 60 ml/min. In each risk group, mean age did not differ between patients with normal versus abnormal CR.
Conclusions: Abnormal CR is a fundamental aspect of disorders predisposing to dementia, including diabetes, hypertension, insulin resistance, metal toxicity and brain trauma. Renal dysfunction unexpectedly correlates inversely with abnormal CR, possibly because of an artifact (eg. a greater perfusion-stimulated cerebral arterial input function, if uncorrected for renal failure). Hence, high prevalence of abnormal CR typical of dementia may be underestimated, particularly in conditions associated with renal failure, such as metal toxicity, diabetes and hypertension.
Fig. 1 (Above, Left): Axial SPECT slices are defined parallel to the brain long axis from occipital to prefrontal. For the Cortical Metabolic index (CMi), one or more axial slices are centered one third of the way from the top of the brain, just superior to the roof of the normal-sized lateral ventricles. Activity display uses a Sokoloff color scale, with white for peak brain, black for zero and spectral colors for intermediate activities. Computer-selected isocontours define areas that contain activity above a certain fraction of the peak activity. The 30% isocontour represents total brain activity in that axial slice, chosen slightly outside the actual external edge of the brain (usually near a 60% isocontour) to correct for attenuation. The 60% isocontour represents the edge of normally functioning neurons, and the Cortical Metabolic index (CMi), calculated as the ratio of activity within the 60% to that within the 30% isocontours, is a measure of the fraction of brain function due to normal neuronal function. The Cortical Perfusion index (CPi) is similarly calculated from 60% and 30% isocontours after the patient has receives a cerebral perfusion stimulant such as 0.5 to 1 g acetazolamide IV or 0.4 to 0.8 mg nitroglycerin sublingual.
Fig. 2 (Above, Right): The Cortical Meetabolic index (CMi) 38.58%, for a 91 year-old moderately demented woman is demonstrated using a color scale, available on nearly all commercial SPECT instruments. In patients with low likelihood of disease the values for Cortical Metabolic index (performed with patients injected with metabolic or basal blood flow tracers such as Tc-99m-HMPAO, Tc-99m-ECD or F-18-FDG) are 50 to 72% and for CPi increase to 53 to 75%. The Cerebral Flow Reserve index CRis simply CPi minus CMi, which is normally (defined in low likelihood disease patients) a positive number > 3%. We found previously that abnormal CR values < 3% or even negative numbers, are typical of cerebrovascular or associated disease such as diabetics, prediabetics, oxidative metal exposed and traumatic brain injured.
1. Summary Dementia is a clinical syndrome whose etiology is multifactorial. Multiple interrelated dementia risk factors coexist but may not have equal or simply additive pathophysiologic effects nor equal influence on quantitative biomarkers of dementia. Among the dementia risk factors, including cerebrovascular disease, hypertension, diabetic or nondiabetic insulin resistance (which tend to coexist with obesity), and even traumatic brain injury and neurotoxic metal exposure, there is a common denominator of cerebral hypoperfusion. Brain SPECT, particularly using protocols that compare basal metabolism and stimulated perfusion, is well suited to detect cerebral hypoperfusion. We focused not only on absolute measures but also the cerebral vascular reserve (CR), defined as the difference of Cortical Perfusion and Metabolic indices (CPi minus CMi). Renal dysfunction is of increasing interest as a dementia risk factor, particularly in light of its well recognized pathophysiologic interactions with hypertension, insulin resistance, and metal toxicity. We found that patients with even mild renal dysfunction had a remarkable tendency to preserved (CR) measured by brain SPECT. This effect depended on total renal function and was not observed in patients with only unilateral renal dysfunction. Patients with bilateral renal dysrunction, despite usually normal CR, had a high risk of stroke and compromised cognitive function. Moreover, following significant stroke, brain SPECT measure of CR typically reverts to abnormal, even in the presence of continued renal dysfunction.
Although treatment results remain preliminary, the initial impression is that agents which relieve cerebral hypoperfusion have positive effects. These include antihypertensives, vasodilators such as nitrates and possibly cilostazol (recently recognized as effective stroke prevention in diabetics) and omega-3 unsaturated marine oil. An additional observation in this work was that either acute or chronic disruption of glycemic control contributes to cerebral hypoperfusion. Moreover, in both diabetics and nondiabetics, agents with positive modulation of glucose metabolism and possibly other neuroendocrine effects on body weight, including incretins, also have both acute and chronic effects on regional and global cerebral perfusion.
** From the Section for Psychiatry and Neurochemistry, Neuropsychiatric Epidemiology Unit at the Sahlgrenska Academy, University of Gothenburg, Sweden and; the Departments of Neurology and Medicine, SUNY-Downstate Medical Center, Brooklyn, New York, USA
^ From the Cincinnati Cognitive Collaboration, 4743 Cornell Rd, Cincinnati, OH, 45241 USA and + The Christ Hospital, Cincinnati, OH, 45219, USA.
2. Table 1. A demographic summary of the patients