Stephen T. Turner

Evaluation and clinical implications of aortic valve calcification measured by electron-beam computed tomography.

Background—Electron-beam computed tomography (EBCT) is used to measure coronary calcification but not for aortic valve calcification (AVC). Its accuracy, association with aortic stenosis (AS) severity, and diagnostic and prognostic value with respect to AVC are unknown. Methods and Results—In 30 explanted aortic valves, the AVC score by EBCT (1125±1294 Agatston units [AU]) showed a strong linear correlation (r =0.96, P <0.0001) with valvular calcium weight (653±748 mg) by pathology that allowed estimation of calcium weight as AVC score/1.7, with a small standard error of the estimate (53 mg). In 100 consecutive clinical patients, we measured AVC by EBCT and AS severity by echocardiographic aortic valve area (AVA). The AVC score was 1316±1749 AU (range 0 to 7226 AU). Intraobserver and interobserver variabilities were excellent (4±4% and 4±10%, respectively). AVC and AVA were strongly associated (r =0.79, P <0.0001) but had a curvilinear relationship that suggested that AVC and AVA provide complementary information. AVC score ≥1100 AU provided 93% sensitivity and 82% specificity for diagnosis of severe AS (AVA <1 cm2), with a receiver operator characteristic curve area of 0.89. AVC assessment by echocardiography was often more severe than by EBCT (P <0.0001). During follow-up, 22 patients either died, developed heart failure, or required surgery. With adjustment for age, sex, symptoms, ejection fraction, and AVA, the AVC score was independently predictive of event-free survival (risk ratio 1.06 per 100-AU increment [1.02 to 1.10], P <0.001), even after adjustment for echocardiographic calcifications. Conclusions—AVC is accurately and reproducibly measured by EBCT and shows a strong association and diagnostic value for severe AS. The curvilinear relationship between AVC and AVA suggests these measures are complementary, and indeed, AVC provides independent outcome information. Thus, AVC is an important measurement in the evaluation of patients with AS.

Genome-Wide Linkage Analysis Reveals Evidence of Multiple Regions That Influence Variation in Plasma Lipid and Apolipoprotein Levels Associated With Risk of Coronary Heart Disease

Results of genome-wide linkage analyses to identify chromosomal regions that influence interindividual variation in plasma lipid and apolipoprotein levels in the Rochester, Minn, population are reported. Analyses were conducted for total cholesterol (total-C), triglycerides (TGs), high density lipoprotein cholesterol (HDL-C), apolipoprotein A-I, apolipoprotein A-II, apolipoprotein B, apolipoprotein C-II, apolipoprotein C-III, apolipoprotein E, the total-C/HDL-C ratio, and the TG/HDL-C ratio. Genotypes were measured for 373 genome-wide marker loci on 1484 individuals distributed among 232 multigeneration pedigrees sampled without regard to health status. LOD scores and estimates of additive genetic variance associated with map locations were obtained by using the variance-component method of linkage analysis. No evidence of linkage with genes influencing variation in age served as a negative control. Plasma apolipoprotein E levels and the apolipoprotein E gene served as a positive control (LOD score 4.20). Evidence (LOD score >2.00) was provided that was suggestive of a gene or genes on chromosomes 4 and 5 influencing variation in the apolipoprotein A-II level, on chromosome 12 influencing variation in the apolipoprotein A-I level, and on chromosome 17 influencing variation of total-C/HDL-C. These analyses provide new information about genomic regions in humans that influence interindividual variation in plasma lipid and apolipoprotein levels and serve as a basis for further fine-mapping studies to identify new genes involved in lipid metabolism.

Association of Fibrinogen With Quantity of Coronary Artery Calcification Measured by Electron Beam Computed Tomography

Increased plasma fibrinogen concentration is an independent risk factor for cardiovascular disease. Fibrinogen is the main coagulation protein in plasma, a determinant of blood viscosity, and can act as a cofactor for platelet aggregation. In this study of middle-aged men and women, we examined the association between plasma fibrinogen concentration and coronary artery calcification (CAC), a marker of preclinical coronary atherosclerosis. Two hundred twenty-eight participants were selected from the community-based Epidemiology of Coronary Artery Calcification Study, in which CAC was measured noninvasively by electron beam computed tomography. One hundred fourteen participants (57 men) were selected because they had high quantities of CAC; the remaining 114 participants (57 men) were selected because they had no detectable CAC. Logistic regression models were used to investigate the association between plasma fibrinogen concentration and high quantity of CAC. In men, an increase of 1 standard deviation in fibrinogen concentration was associated with a statistically significant odds ratio of 1.6 (95% CI 1.1 to 2.5) for a high quantity of CAC. In women, the corresponding odds ratio was 2.5 (95% CI 1.6 to 4.1). Inferences from sex-specific bivariate logistic models for odds ratios adjusted individually for each coronary risk factor and C-reactive protein were similar to those from the univariate models. In women, there was also a significant interaction between fibrinogen concentration and age. According to the models, younger women with high plasma fibrinogen were more likely to have high quantities of CAC than were younger women with low plasma fibrinogen. The strength of this association was diminished in older women.

Sodium-lithium countertransport genotype and the probability of hypertension in adults.

The objective of the present study was to determine whether information about a biometrically inferred single gene with large effects on erythrocyte sodium-lithium countertransport is useful in predicting the probability of having hypertension. We used multivariate logistic regression to model the relationship between the probability of having hypertension and predictor traits in a sample of 382 unrelated adult women and 347 unrelated adult men from Rochester, Minn. First, we identified a set of demographic, biochemical, and physiological predictors. Second, we analyzed whether the relationship between the probability of having hypertension and the identified predictor traits was heterogeneous between the biometrically inferred single locus genotypes with large effects on sodium-lithium countertransport level. Third, if there was no heterogeneity, we assessed whether sodium-lithium countertransport genotypes made an additional contribution to predicting the probability of having hypertension after other predictors were considered. In women, the predictors of the probability of having hypertension were age, plasma apolipoprotein CIII, body mass index, and an interaction term involving age and body mass index. The relationship between the probability of having hypertension and the identified predictors was not heterogeneous between sodium-lithium countertransport genotypes, and genotype did not contribute to the prediction of the probability of having hypertension after the identified predictors were considered. In men, predictors of the probability of having hypertension were age, plasma levels of high-density lipoprotein cholesterol, apolipoproteins AI and CII, sodium-lithium countertransport level, and sodium-lithium countertransport genotype. The relationship between the probability of having hypertension and sodium-lithium countertransport level and age were heterogeneous between biometrically inferred sodium-lithium countertransport genotypes.(ABSTRACT TRUNCATED AT 250 WORDS)

Towards interpreting models to orchestrate IaaS multi-cloud infrastructures

One challenge to the cloud computing paradigm is the task complexity associated with designing and managing multi-cloud solutions based on operational objectives. Heterogeneous vendor interfaces and a lack of standardization compounds this complexity and may eventually lead to vendor lock-in. In this article we present a model driven approach to allowing network administrators to intuitively describe and rapidly realize non-trivial IaaS behavior in realtime. We have developed iCloudML, an interpreted domain-specific modeling language and its interpreter as tooling support for the domain.