Hepatic aging as an etiological factor in the development of Alzheimer's disease

Clinical trials with velnacrine: (PROPP) the physician reference of predicted probabilities--a statistical model for the estimation of hepatotoxicity risk with velnacrine maleate

Safety observations during the clinical development of Mentane (velnacrine maleate) have included the occurrence of generally asymptomatic liver enzyme elevations confined to patients with Alzheimer's disease (AD). The clinical presentation of this reversible hepatocellular injury is analogous to that reported for tetrahydroaminoacridine (THA). Direct liver injury, possibly associated with the production of a toxic metabolite, would be consistent with reports of aberrant xenobiotic metabolism in Alzheimer's disease patients. Since a patient related aberration in drug metabolism was suspected, a biostatistical strategy was developed with the objective of predicting hepatotoxicity in individual patients prior to exposure to velnacrine maleate. The method used logistic regression techniques with variable selection restricted to those items which could be routinely and inexpensively accessed at screen evaluation for potential candidates for treatment. The model was to be predictive (a marker for eventual hepatotoxicity) rather than a causative model, and techniques employed "goodness of fit", percentage correct, and positive and negative predictive values. On the basis of demographic and baseline laboratory data from 942 patients, the PROPP statistic was developed (the Physician Reference Of Predicted Probabilities). Main effect variables included age, gender, and nine hematological and serum chemistry variables. The sensitivity of the current model is approximately 49%, specificity approximately 88%. Using prior probability estimates, however, in which the patient's likelihood of liver toxicity is presumed to be at least 30%, the positive predictive value ranged between 64-77%. Although the clinical utility of this statistic will require refinements and additional prospective confirmation, its potential existence speaks to the possibility of markers for idiosyncratic drug metabolism in patients with Alzheimer's disease.

Hippocampal RAGE immunoreactivity in early and advanced Alzheimer's disease

Microvascular accumulation and neuronal overproduction of amyloid-beta peptide (Abeta) are pathologic features of Alzheimer's disease (AD). In this study, we examined the receptor for advanced glycation endproducts (RAGE), a multi-ligand receptor found in both neurons and cerebral microvascular endothelia that binds Abeta. RAGE expression was assessed in aged controls (n = 6), patients with early AD-like pathology (n = 6), and severe, Braak V-VI AD (n = 6). Human hippocampi were stained with a specific polyclonal antibody directed against RAGE (Research Diagnostics, Flanders, NJ). Immunoreactivity was localized in both neurons and cerebral endothelial cells. Quantitative image-analyses were performed on grayscale images to assess the total surface area of endothelial RAGE immunoreaction product in cross sections of cerebral microvessels (5-20 microm). Confocal images were acquired for confirmation of RAGE immunoreactivity in both microvessels and neurons by coupling RAGE with CD-31 and neurofilament, respectively. A significant increase in endothelial RAGE immunoreactivity was found in severe Braak V-VI AD patients when compared to aged controls (p < 0.001), and when compared to patients with early AD pathology (p = 0.0125). In addition, a significant increase in endothelial RAGE immunoreactivity was witnessed when comparing aged controls having no reported AD pathology with patients having early AD-like pathology (p = 0.038). Our data suggest that microvascular RAGE levels increase in conjunction with the onset of AD, and continue to increase linearly as a function of AD pathologic severity (p < 0.0001).