Abstract P156: A Novel Gene-based Tool to Predict the Risk of Essential Hypertension and Initial Validation
Background: According to the CDC, 70 million American adults suffer from high blood pressure. With hypertension costing the United States $46 billion annually, efforts to screen, prevent, and treat hypertension are warranted.
Objective: To create and evaluate the ability of a novel scoring algorithm derived from genotypic and phenotypic factors to predict the risk of essential hypertension.
Subjects: For this multi-center, observational, retrospective study, 462 subjects were chosen from 14 clinical research sites across the U.S from the period of September 30, 2014 to December 17, 2014. The first study group of 126 subjects had a diagnosis of essential hypertension (ICD-9 codes 401, 401.1, and 401.9); and the first study set of 131 controls had no diagnosis of essential hypertension. The second study consisted of 95 subjects with a diagnosis of essential hypertension and 110 controls were used.
Methods: Subjects were genotyped using TaqMan® SNP genotyping assays (Life Technologies, Carlsbad, CA). A scoring algorithm was developed using a logistic regression model with age as a phenotypic factor and 6 single nucleotide polymorphisms (COMT rs4680; PTGS1 rs1330344; SHMT1 rs1979277; LEPR rs1137101; and VKORC1 rs8050894). A second study was conducted to independently apply the algorithm and validate results.
Results: For the first study group, the receiver operating characteristic (ROC) for the algorithm was statistically significant (AUC 0.917, p=0.000). From a scoring range of 0 - 17, 8.5 was determined to be the best cut-off score indicating a high risk of essential hypertension with a sensitivity of 83.17% (84 of 101); specificity 91.45% (107 of 117); PPV 89.36% (84 of 94); and NPV 86.29% (107 of 124). For the second study group, the ROC was also statistically significant (AUC 0.969, p=0.000). As in the first study group, a score of 8.5 was the best cut-off score for essential hypertension risk with a sensitivity of 91.58% (87 of 95); specificity 95.45% (105 of 110); PPV 94.57% (87 of 92); and NPV 92.92% (105 of 113).
Conclusion: This scoring algorithm can reliably predict the risk of essential hypertension in two separate study groups. Implementing such a tool in clinical practice may guide treatment decisions for patients at risk of essential hypertension.
Author Disclosures: T. Onojighofia: None. N. Anand: None. B. Meshkin: A. Employment; Modest; I am the CEO of Proove Biosciences. F. Ownership Interest (includes any stock, stock option, partnership, membership or other equity position in an entity regardless of the form of the entity, or any option or right to acquire such position, and any rights in any patent or other intellectu; Modest; I own stocks in Proove Biosciences. S. Silverman: None. D. Holman: None. J. Hubbard: None. M. Hafez: None. S. Kantorovich: None.
- © 2015 by American Heart Association, Inc.