Mean blood pressure decrease during HCTZ treatment. Upper panel: Data are analyzed according to ACE I/D polymorphism. Lower panel: Data are analyzed according to ␣ -adducin Gly460Trp polymorphism. Statistical analyses performed with the Student t test (I allele for ACE and 460Trp allele for ␣ -adducin dominant). 

Contexts in source publication

Context 1

... he between therapy the of extensive hypertension analysis suffers and reports from a of discrepancy the beneficial effects of blood pressure (BP) lowering in the whole population and the relative lack of such an extensive effort and scientific rigor in investigating the heterogeneity of individual response to pharmacological treatment. Variation in individual response to therapy is a well-known phenome- non in clinical practice. This discrepancy contributes to the high rate of therapy discontinuation and/or underdosing because of poor efficacy or unwanted side effects. 1–3 The study of the renin profile and the vasoconstriction-volume analysis by Laragh et al 4,5 represents pioneering work in the identification of the sources of interindividual variation in response to therapy. Blood pressure response to an antihypertensive agent can be considered a combination of the pharmacological effect of the drug and the physiological reactions of the patient. Salt loss with consequent volume depletion induced by diet or diuretic treatment in some patients may be counterbalanced by hyperactivation of the renin-angiotensin system (RAS) determining vasoconstriction and increased adrenergic tone. 6 Therefore, considering the relationship between volume and the RAS in the patho- genesis of essential hypertension 7 and in the response to drugs, the main question is: What determines the variability from one patient to another in the response to diuretics? What is the respective contribute of pharmacokinetics or pharmacodynamics to this variability? Genetics may help in defining the respective contributes of these 2 mechanisms. When the contribution of the former is defined, the latter may also help in dissecting the genetic complexity of hypertension, since the association between a genetic polymorphism and the response to treatment may provide insight in the pathogenetic mechanisms. Because of the importance of volume and vasoconstriction, sodium balance, and RAS activation as mechanisms underlying hypertension, diuretics can be a good example of application of this approach. 8 –11 We chose ACE I/D and ␣ -adducin Gly460Trp polymorphisms, even though we are aware that their role in the pathogenesis of hypertension is not yet fully elucidated, because of our previous demonstration of their interaction on renal sodium handling and blood pressure regulation. 12 This interaction may be due to the fact that ACE I/D polymorphism affects RAS activation after sodium depletion, 6 thus limiting the BP fall. Conversely, ␣ -adducin Gly460Trp polymorphism influences the constitutive capacity of the kidney to reabsorb sodium, 8 thus implying a modulation of the BP responsiveness to a drug such as HCTZ, which inhibits such renal mechanism. In our work, such polymorphisms were considered both separately and together. Because many factors affect the genotype-phenotype rela- tionshp (environment, stage of disease, previous therapy, population admixture, and so forth), we studied only newly discovered and never-treated patients with mild hypertension to avoid the interference of previous therapy and of the stage of the disease. All participants were recruited in our outpa- tient clinic. We found that hypertensive individuals carrying at least one I allele of ACE and at least one 460Trp allele of ␣ -adducin had the largest BP decrease after 2 months of HCTZ treatment. We enrolled 87 patients (75 men and 12 women). One patient withdrew his consent before the end of the first month of treatment (HCTZ 12.5 mg). The main clinical and biochemical characteristics of the patients who completed the study as a whole and according to ACE and ␣ -adducin genotype are summarized in Tables 1 through 4. A marginal significance was observed for height when data were analyzed according to ACE genotype and for PRA, when they were analyzed according to ␣ -adducin genotype, the latter confirming previous observations. 8 Control for deviation from the Hardy-Weinberg equilib- rium gave nonsignificant results ( P ϭ 0.995 for ACE and P ϭ 0.997 for ␣ -adducin), and the frequency of the ACE and ␣ -adducin alleles (I ϭ 0.40; D ϭ 0.60, for ACE and 460Gly ϭ 0.81; 460Trp ϭ 0.19 for ␣ -adducin) was similar to one obtained in a random sample of 200 individuals of similar age that we collected in the Milano area (I ϭ 0.38; D ϭ 0.62; 460Gly ϭ 0.79; 460Trp ϭ 0.21). A significantly different trend in MBP decrease after 2 months of HCTZ was observed for ACE and for ␣ -adducin (RM-ANOVA, P ϭ 0.049, codominant model for ACE and P ϭ 0.002 for ␣ -adducin, with 460Trp dominant) when ana- lyzed alone. The identical MBP decrease of II and ID patients (Figure 1) led us to test a recessive model for the D allele. II ϩ ID were then analyzed together and compared with DD patients. Also in this case, a significantly different trend was observed (RM-ANOVA P ϭ 0.013 recessive model). The analysis was then repeated, entering ACE (D recessive) and ␣ -adducin (460Trp dominant) together as between-subject factors. The combination of genotypes was chosen according to the fact that carriers of 460Trp respond more to diuretic treatment (460Trp dominant) 8 and that carriers of the D allele have increased PRA with volume depletion 17 and blood pressure after angiotensin I infusion. 18 In this way, we speculated that the hypotensive effect of incremental diuretic doses would be damped in carriers of the D allele. The independent contribution of the 2 genes remained significant, with no interaction (Table 5), indicating a pure additive effect without epistasis. The absolute MBP decrease was significant after 1 month for ␣ -adducin ( P ϭ 0.005) and after 2 months for ␣ -adducin and ACE ( P ϭ 0.003 and P ϭ 0.02, respectively, see Figure 1). In our previous study, we arbitrarily defined responders to this experimental protocol as patients whose MBP decreased Ͼ 15 mm Hg at the end of the second month of treatment. 8 By applying the same criteria to the present data set, 20 of 86 patients were classified as responders (23%). The responder rate doubled when considering only individuals carrying at least one I and one 460Trp allele (9 of 21, 43%, with a positive predictive value of 52%) but was reduced to 1 of 22 (4%) in those with the Gly460Gly ϩ DD genotype. Compared with the latter, whose MBP decrease was only 3.43 Ϯ 1.7 mm Hg, those carrying at least one I and one 460Trp allele had an odds ratio of being respondent to HCTZ of 15.75 (2.06 to 57.63, 95% CI, ␹ 2 ϭ 8.63), with a MBP decrease of 12.7 Ϯ 1.9 mm Hg (Figure 2). There were no significant differences between body weight changes according to the different ACE and ␣ -adducin genotypes after treatment. Our data show that the magnitude of blood pressure fall after HCTZ is strongly influenced by ACE I/D and ␣ -adducin Gly460Trp polymorphisms. These findings may have both theoretical and practical implications, provided that confounding factors may be reasonably excluded or properly understood. These results were obtained in a homogeneous group of newly discovered white hypertensive patients, all selected and followed up in the same unit by the same 2 physicians throughout all the study. Therefore the stage of the disease, the carry over effects of previous therapy, and some environmental factors could reasonably be excluded as a source of variation. Moreover, body weight, age, gender, and basal BP were not different among the various subgroups. To use the magnitude of the blood pressure fall as an intermediate phenotype to dissect genetic complexity, we have to exclude an influence of these 2 genotypes on the pharmacokinetics and metabolism of hydrochlorothiazide. This exclusion is probable, since the dose-effect relationship of HCTZ is very flat and poorly related to plasma levels, implying a relatively modest influence of the metabolism on pharmacodynamics. 19 The major finding of our work is the significantly greater blood pressure fall observed in carriers of the genotype DI or II ϩ Gly460Trp or Trp460Trp compared with carriers of the genotype DD ϩ Gly460Gly ( Ϫ 12.7 Ϯ 1.9 mm Hg versus Ϫ 3.43 Ϯ 1.7 mm Hg) after chronic diuretic treatment. Consid- ering arbitrary definition of responders and a MBP decrease Ͼ 15 mm Hg (identical to the one already proposed 8 ), the whole group had 23% responders, but this value became 4% in the least responsive group (DD ϩ Gly460Gly) and 43% in the most responsive group (DI or II ϩ Gly460Trp or Trp460Trp), with an odds ratio of being as responder of 15.75 when the latter group was compared with the first. In fact, our data confirm the already known remarkable heterogeneity in the individual BP response to sodium depletion or antihypertensive drugs that still represents an obstacle in the therapeu- tic approach to hypertension. Therefore, the BP fall according to the different combinations of genotypes may be due to an interaction of HCTZ, with the pressor mechanism triggered by the genotypes. ACE and ␣ -adducin polymorphisms should work at 2 different biological levels. The activation of the RAS is probably one of the most powerful mechanisms limiting the BP fall after sodium depletion induced by a low salt diet 6 or diuretic therapy. There is a general agreement that the D allele of ACE is associated with increased plasma 20 and tissues levels of ACE. 18,21 – 23 The RAS is mostly a paracrine and autocrine system, and the plasma level of its constituents poorly reflect its activity. Also, if the biological meaning of larger ACE activity is still controversial, the D allele has been often associated with the worsening of cardiovascular and renal prognosis, 24,25 whereas ACE inhibitors ...

Context 2

... variation in response to therapy. Blood pressure response to an antihypertensive agent can be considered a combination of the pharmacological effect of the drug and the physiological reactions of the patient. Salt loss with consequent volume depletion induced by diet or diuretic treatment in some patients may be counterbalanced by hyperactivation of the renin-angiotensin system (RAS) determining vasoconstriction and increased adrenergic tone. 6 Therefore, considering the relationship between volume and the RAS in the patho- genesis of essential hypertension 7 and in the response to drugs, the main question is: What determines the variability from one patient to another in the response to diuretics? What is the respective contribute of pharmacokinetics or pharmacodynamics to this variability? Genetics may help in defining the respective contributes of these 2 mechanisms. When the contribution of the former is defined, the latter may also help in dissecting the genetic complexity of hypertension, since the association between a genetic polymorphism and the response to treatment may provide insight in the pathogenetic mechanisms. Because of the importance of volume and vasoconstriction, sodium balance, and RAS activation as mechanisms underlying hypertension, diuretics can be a good example of application of this approach. 8 –11 We chose ACE I/D and ␣ -adducin Gly460Trp polymorphisms, even though we are aware that their role in the pathogenesis of hypertension is not yet fully elucidated, because of our previous demonstration of their interaction on renal sodium handling and blood pressure regulation. 12 This interaction may be due to the fact that ACE I/D polymorphism affects RAS activation after sodium depletion, 6 thus limiting the BP fall. Conversely, ␣ -adducin Gly460Trp polymorphism influences the constitutive capacity of the kidney to reabsorb sodium, 8 thus implying a modulation of the BP responsiveness to a drug such as HCTZ, which inhibits such renal mechanism. In our work, such polymorphisms were considered both separately and together. Because many factors affect the genotype-phenotype rela- tionshp (environment, stage of disease, previous therapy, population admixture, and so forth), we studied only newly discovered and never-treated patients with mild hypertension to avoid the interference of previous therapy and of the stage of the disease. All participants were recruited in our outpa- tient clinic. We found that hypertensive individuals carrying at least one I allele of ACE and at least one 460Trp allele of ␣ -adducin had the largest BP decrease after 2 months of HCTZ treatment. We enrolled 87 patients (75 men and 12 women). One patient withdrew his consent before the end of the first month of treatment (HCTZ 12.5 mg). The main clinical and biochemical characteristics of the patients who completed the study as a whole and according to ACE and ␣ -adducin genotype are summarized in Tables 1 through 4. A marginal significance was observed for height when data were analyzed according to ACE genotype and for PRA, when they were analyzed according to ␣ -adducin genotype, the latter confirming previous observations. 8 Control for deviation from the Hardy-Weinberg equilib- rium gave nonsignificant results ( P ϭ 0.995 for ACE and P ϭ 0.997 for ␣ -adducin), and the frequency of the ACE and ␣ -adducin alleles (I ϭ 0.40; D ϭ 0.60, for ACE and 460Gly ϭ 0.81; 460Trp ϭ 0.19 for ␣ -adducin) was similar to one obtained in a random sample of 200 individuals of similar age that we collected in the Milano area (I ϭ 0.38; D ϭ 0.62; 460Gly ϭ 0.79; 460Trp ϭ 0.21). A significantly different trend in MBP decrease after 2 months of HCTZ was observed for ACE and for ␣ -adducin (RM-ANOVA, P ϭ 0.049, codominant model for ACE and P ϭ 0.002 for ␣ -adducin, with 460Trp dominant) when ana- lyzed alone. The identical MBP decrease of II and ID patients (Figure 1) led us to test a recessive model for the D allele. II ϩ ID were then analyzed together and compared with DD patients. Also in this case, a significantly different trend was observed (RM-ANOVA P ϭ 0.013 recessive model). The analysis was then repeated, entering ACE (D recessive) and ␣ -adducin (460Trp dominant) together as between-subject factors. The combination of genotypes was chosen according to the fact that carriers of 460Trp respond more to diuretic treatment (460Trp dominant) 8 and that carriers of the D allele have increased PRA with volume depletion 17 and blood pressure after angiotensin I infusion. 18 In this way, we speculated that the hypotensive effect of incremental diuretic doses would be damped in carriers of the D allele. The independent contribution of the 2 genes remained significant, with no interaction (Table 5), indicating a pure additive effect without epistasis. The absolute MBP decrease was significant after 1 month for ␣ -adducin ( P ϭ 0.005) and after 2 months for ␣ -adducin and ACE ( P ϭ 0.003 and P ϭ 0.02, respectively, see Figure 1). In our previous study, we arbitrarily defined responders to this experimental protocol as patients whose MBP decreased Ͼ 15 mm Hg at the end of the second month of treatment. 8 By applying the same criteria to the present data set, 20 of 86 patients were classified as responders (23%). The responder rate doubled when considering only individuals carrying at least one I and one 460Trp allele (9 of 21, 43%, with a positive predictive value of 52%) but was reduced to 1 of 22 (4%) in those with the Gly460Gly ϩ DD genotype. Compared with the latter, whose MBP decrease was only 3.43 Ϯ 1.7 mm Hg, those carrying at least one I and one 460Trp allele had an odds ratio of being respondent to HCTZ of 15.75 (2.06 to 57.63, 95% CI, ␹ 2 ϭ 8.63), with a MBP decrease of 12.7 Ϯ 1.9 mm Hg (Figure 2). There were no significant differences between body weight changes according to the different ACE and ␣ -adducin genotypes after treatment. Our data show that the magnitude of blood pressure fall after HCTZ is strongly influenced by ACE I/D and ␣ -adducin Gly460Trp polymorphisms. These findings may have both theoretical and practical implications, provided that confounding factors may be reasonably excluded or properly understood. These results were obtained in a homogeneous group of newly discovered white hypertensive patients, all selected and followed up in the same unit by the same 2 physicians throughout all the study. Therefore the stage of the disease, the carry over effects of previous therapy, and some environmental factors could reasonably be excluded as a source of variation. Moreover, body weight, age, gender, and basal BP were not different among the various subgroups. To use the magnitude of the blood pressure fall as an intermediate phenotype to dissect genetic complexity, we have to exclude an influence of these 2 genotypes on the pharmacokinetics and metabolism of hydrochlorothiazide. This exclusion is probable, since the dose-effect relationship of HCTZ is very flat and poorly related to plasma levels, implying a relatively modest influence of the metabolism on pharmacodynamics. 19 The major finding of our work is the significantly greater blood pressure fall observed in carriers of the genotype DI or II ϩ Gly460Trp or Trp460Trp compared with carriers of the genotype DD ϩ Gly460Gly ( Ϫ 12.7 Ϯ 1.9 mm Hg versus Ϫ 3.43 Ϯ 1.7 mm Hg) after chronic diuretic treatment. Consid- ering arbitrary definition of responders and a MBP decrease Ͼ 15 mm Hg (identical to the one already proposed 8 ), the whole group had 23% responders, but this value became 4% in the least responsive group (DD ϩ Gly460Gly) and 43% in the most responsive group (DI or II ϩ Gly460Trp or Trp460Trp), with an odds ratio of being as responder of 15.75 when the latter group was compared with the first. In fact, our data confirm the already known remarkable heterogeneity in the individual BP response to sodium depletion or antihypertensive drugs that still represents an obstacle in the therapeu- tic approach to hypertension. Therefore, the BP fall according to the different combinations of genotypes may be due to an interaction of HCTZ, with the pressor mechanism triggered by the genotypes. ACE and ␣ -adducin polymorphisms should work at 2 different biological levels. The activation of the RAS is probably one of the most powerful mechanisms limiting the BP fall after sodium depletion induced by a low salt diet 6 or diuretic therapy. There is a general agreement that the D allele of ACE is associated with increased plasma 20 and tissues levels of ACE. 18,21 – 23 The RAS is mostly a paracrine and autocrine system, and the plasma level of its constituents poorly reflect its activity. Also, if the biological meaning of larger ACE activity is still controversial, the D allele has been often associated with the worsening of cardiovascular and renal prognosis, 24,25 whereas ACE inhibitors improve it. 26 The relation between salt-sensitive hypertension and I/D ACE gene polymorphism has been previously tested in several studies with controversial results. Kojima et al 17 reported a lack of association between the ACE I/D genotype and salt sensitivity in patients with essential hypertension, but they observed that PRA increase after salt restriction is greater in patients with the DD genotype. Hiraga et al 9 instead reported a significant association between salt sensitivity and ACE genotype that has been confirmed by Giner et al, 10 who found that II patients are the most responsive to acute changes in sodium intake. These different results may be related to the protocols used as in all these trials; the antihypertensive therapy was stopped before the beginning of the study, and the patients underwent a washout period from 10 days to 4 weeks. In fact, the carryover effect of a treatment influences not only blood pressure level but also involves other pressure mechanisms. Swart et al 27 reported that the plasma renin response to bendrofluazide was ...