A previous genetic analysis of the Cholesterol and Recurrent Events (CARE) trial found that carriers of the 719Arg allele of the kinesin family member 6 gene ( KIF6 ) (rs20455), but not noncarriers, received significant event reduction from pravastatin therapy. However, that previous analysis of CARE included only Caucasian patients and was limited to the myocardial infarction components of the primary end point. Therefore, the aim of this study was to investigate whether pravastatin therapy reduced primary end point events in KIF6 719Arg carriers and noncarriers, separately, in the combined ethnic groups of CARE. The effect of pravastatin therapy on primary end point events (fatal coronary event or nonfatal myocardial infarction) was investigated in Cox regression models that adjusted for population structure using either self-reported ethnicity or the principal components of genetic heterogeneity. After adjustment for age, gender, and self-reported ethnicity, pravastatin therapy reduced events in carriers of KIF6 719Arg (hazard ratio [HR] 0.63, 95% confidence interval [CI] 0.49 to 0.83) but not in noncarriers (HR 1.01, 95% CI 0.69 to 1.45) (p for interaction = 0.049). After adjustment for age, gender, traditional risk factors, and principal components, pravastatin therapy reduced events in carriers of 719Arg (HR 0.64, 95% CI 0.49 to 0.85) but not in noncarriers (HR 0.90, 95% CI 0.62 to 1.32) (p for interaction = 0.14). In conclusion, in an analysis that included CARE patients of all ethnic groups, pravastatin therapy significantly and substantially reduced primary end point events in carriers of the KIF6 719Arg allele but not in noncarriers.
Variability in response to statin therapy suggests that genetic variants could contribute to response to therapy, and several studies have suggested genetic variants that could affect response to statin therapy by influencing low-density lipoprotein (LDL) cholesterol reduction, coronary artery disease (CAD) event reduction, disease progression, pharmacokinetics, or adverse events (reviewed by Schmitz and Langmann and Mangravite et al ). One such variant is the Arg allele of the Trp719Arg polymorphism of the kinesin family member 6 gene ( KIF6 ), which was found to be associated with response to statin therapy in several randomized trials of statins. Carriers of the 719Arg allele, but not noncarriers, received substantial and significant reduction of coronary events from pravastatin therapy in the Cholesterol and Recurrent Events (CARE) study and West of Scotland Coronary Prevention Study (WOSCOPS). Moreover, carriers of the 719Arg allele also received greater reduction of coronary events than did noncarriers from atorvastatin 80 mg/day, compared to pravastatin 40 mg/day, in the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE-IT)–Thrombolysis In Myocardial Infarction (TIMI) 22 study. Because there was no placebo group in PROVE-IT–TIMI 22, it was not possible to estimate the extent to which coronary events were reduced by pravastatin treatment in that study.
The 719Arg allele of KIF6 was also found to be associated with risk for CAD in 3 large prospective studies. In the Atherosclerosis Risk in Communities (ARIC) study, a prospective study of middle-aged subjects in North American communities, the 719Arg allele was associated with increased risk for CAD. In the Cardiovascular Health Study (CHS), a prospective study in older North Americans, carriers of the 719Arg allele, compared to noncarriers, had increased risk for myocardial infarction (MI). Similarly, in the Women’s Health Study (WHS), a prospective study of initially healthy female health professionals, carriers of the 719Arg allele, compared to noncarriers, had increased risk for major cardiovascular events. In contrast to these studies of CAD events, no association between angiographically defined coronary disease and KIF6 carrier status was observed in a retrospective case-control study. The previous analysis of CARE was focused on the genetics of MI and therefore analyzed only the MI components (confirmed fatal or nonfatal MI) of the composite primary end point, rather than the primary end point of CARE (fatal coronary event or nonfatal MI). Additionally, to reduce potential confounding by population stratification, the previous genetic analysis of CARE included only Caucasian patients. However, because there is no evidence that the association between KIF6 Trp719Arg and response to statin therapy differs according to ethnic group, a more inclusive analysis that adjusted for population structure would be appropriate. Therefore, to provide a genetic analysis of KIF6 Trp719Arg that used the original CARE inclusion criteria and end point, we asked whether pravastatin therapy in the CARE study significantly reduced primary end point events (fatal coronary event or nonfatal MI) in KIF6 719Arg carriers, but not noncarriers, when all ethnic groups were included in the analysis.
Methods
The study population for this genetic study was derived from the CARE study, which has been described previously. Briefly, CARE was a double-blind trial of 4,159 patients with MIs (3,583 men and 576 women) who were randomly assigned to pravastatin 40 mg/day or placebo. Deidentified deoxyribonucleic acid (DNA) for genetic studies was available from 3,109 CARE patients, and the study population for this genetic study was the 3,103 patients who had their KIF6 Trp719Arg genotypes successfully determined.
The end point for this genetic study was the same as the primary end point of the original CARE study: time to first fatal coronary event or nonfatal MI. This study included 336 primary end point events, compared with 245 fatal or nonfatal MI events in the previous genetic analysis of CARE.
All laboratory measurements were made in a core laboratory. KIF6 Trp719Arg genotypes as well as the genotypes of 746 single-nucleotide polymorphisms used to estimate population stratification were determined using an allele-specific real-time polymerase chain reaction or allele-specific oligonucleotide ligation at a core facility.
All reported p values are 2 sided. Differences between baseline characteristics were assessed using Wilcoxon’s rank-sum test (for continuous variables) or by Fisher’s exact test (for discrete variables). Cox proportional-hazards regression models were used to assess the effect of pravastatin 40 mg/day compared to placebo on incident events in KIF6 719Arg carriers and in noncarriers. To estimate hazard ratios while accounting for the effects of potential confounders, baseline covariates were included in the Cox models and were coded either categorically (gender, smoking [current smoker vs never or past smoker], history of hypertension, history of diabetes, and self-reported ethnicity) or continuously (age, LDL cholesterol, high-density lipoprotein cholesterol, body mass index, and principal components of genetic variability). To assess the genetic structure of the CARE patient population, we used EIGENSTRAT software to calculate the principal components of the genetic variability using the genotypes of 746 that were not in linkage disequilibrium with one another (r 2 <0.1), autosomal, single-nucleotide polymorphisms among 1,000 single-nucleotide polymorphisms that were previously genotyped in CARE. To adjust for population structure, we included the first 5 principal components (those with eigenvalues that indicated evidence of population structure with p values <10 −7 ) as covariates in Cox regression models. Likelihood ratio tests were used to evaluate the addition of an interaction term between KIF6 719Arg carrier status and pravastatin therapy to regression models. The Kaplan-Meier method was used to estimate the cumulative fraction of fatal coronary events or nonfatal MI events, and differences were assessed using the log-rank test. Absolute risk reduction by pravastatin therapy was calculated from Kaplan-Meier estimates of the cumulative incidence of events at 5 years. R statistical software (R Foundation for Statistical Computing, Vienna, Austria) was used for all regression models and for generating Kaplan-Meier estimates.
Results
The baseline characteristics according to KIF6 719Arg carrier status for this genetic study of CARE are listed in Table 1 . Risk factors for CAD had similar prevalence in carriers and noncarriers of the 719Arg allele. Baseline risk factors for CAD were also similar in the 2 treatment groups (p >0.12; data not presented). Overall, the frequency of 719Arg carriers was 60% in this study. Carrier frequency was highest in African Americans ( Table 2 ). Caucasians constituted most patients in this study; Hispanics and African Americans were the second and third largest ethnicities in this study. Triglyceride levels, histories of diabetes, and histories of hypertension differed dramatically (p <0.0001) among ethnic groups. Several other baseline characteristics were also different (p <0.05; Table 2 ).
| Characteristic | Arg/Arg + Arg/Trp (n = 1,873) | Trp/Trp (n = 1,230) | p Value ⁎ |
|---|---|---|---|
| Men | 1,607 (86%) | 1,071 (87%) | 0.34 |
| Age (years) | 58.4 ± 9 | 58.6 ± 9 | 0.39 |
| Body mass index (kg/m 2 ) | 27.7 ± 4 | 27.4 ± 4 | 0.10 |
| Smokers | 316 (17%) | 194 (16%) | 0.43 |
| History of diabetes | 261 (14%) | 164 (13%) | 0.67 |
| History of hypertension | 802 (43%) | 523 (42%) | 0.88 |
| LDL cholesterol (mg/dl) | 138.7 ± 15 | 139.1 ± 15 | 0.58 |
| High-density lipoprotein cholesterol (mg/dl) | 38.5 ± 9 | 38.9 ± 9 | 0.27 |
| Total cholesterol (mg/dl) | 208.5 ± 17 | 209.3 ± 17 | 0.18 |
| Triglycerides (mg/dl) | 156.7 ± 60 | 156.9 ± 61 | 0.74 |
| Self-reported ethnicity | <0.0001 | ||
| African American | 70 (3.7%) | 8 (0.65%) | |
| Asian Pacific islander | 14 (0.75%) | 7 (0.57%) | |
| Caucasian | 1,722 (91.9%) | 1,185 (96.3%) | |
| Hispanic | 62 (3.3%) | 23 (1.9%) | |
| Other | 5 (0.27%) | 7 (0.23%) |
⁎ Wilcoxon’s rank-sum test for continuous variables and Fisher’s exact test for discrete variables.
| Characteristic | African American (n = 78) | Asian or Pacific Islander (n = 21) | Caucasian (n = 2,907) | Hispanic (n = 85) | Other (n = 12) | p Value |
|---|---|---|---|---|---|---|
| KIF6 719Arg carriers | 70 (89.7%) | 14 (66.7%) | 1,722 (59.2%) | 62 (72.9%) | 5 (41.7%) | <0.0001 |
| Men | 58 (74.4%) | 18 (85.7%) | 2,518 (86.6%) | 73 (85.9%) | 11 (91.7%) | 0.056 |
| Age (years) | 58.1 ± 9.3 | 57.3 ± 10.0 | 58.5 ± 9.3 | 57.9 ± 9.6 | 53.8 ± 7.8 | 0.35 |
| Body mass index (kg/m 2 ) | 29.9 ± 5.3 | 26.1 ± 3.8 | 27.5 ± 4.3 | 28.1 ± 4.6 | 28.0 ± 4.0 | 0.0003 |
| Smokers | 12 (15.4%) | 0 (0%) | 483 (16.6%) | 12 (14.1%) | 3 (25.0%) | 0.19 |
| History of diabetes | 22 (28.2%) | 4 (19.0%) | 367 (12.6%) | 27 (31.8%) | 5 (41.7%) | <0.0001 |
| History of hypertension | 58 (74.4%) | 9 (42.9%) | 1,213 (41.7%) | 40 (47.1%) | 5 (41.7%) | <0.0001 |
| Low-density lipoprotein cholesterol (mg/dl) | 140.6 ± 14.5 | 135.3 ± 14.2 | 138.9 ± 14.6 | 134.5 ± 15.5 | 134.3 ± 15.8 | 0.0094 |
| High-density lipoprotein cholesterol (mg/dl) | 41.6 ± 10.6 | 38.6 ± 9.3 | 38.7 ± 8.8 | 36.6 ± 8.2 | 36.2 ± 6.4 | 0.021 |
| Total cholesterol (mg/dl) | 206.8 ± 17.2 | 202.8 ± 16.5 | 209.0 ± 17.3 | 205.1 ± 17.1 | 203.5 ± 16.6 | 0.042 |
| Triglycerides (mg/dl) | 123.1 ± 51.6 | 144.3 ± 58.7 | 157.2 ± 60.3 | 174.6 ± 63.2 | 165.3 ± 45.5 | <0.0001 |
The genetic structure of the CARE patient population was estimated by a principal-component analysis. The eigenvalues of the first 10 principal components of genetic variability are listed in Table 3 ; the null hypothesis of the p values for each eigenvalue is that the data do not show evidence of additional population structure over and above what has already been detected. The correspondence between self-reported ethnicity of CARE patients and the first 3 principal components of genetic variability is shown in Figure 1 .
| Principal Component | Eigenvalue | p Value |
|---|---|---|
| 1 | 41.8 | <1 × 10 −200 |
| 2 | 12.9 | 7.3 × 10 −190 |
| 3 | 10.6 | 2.0 × 10 −44 |
| 4 | 9.87 | 9.1 × 10 −15 |
| 5 | 9.64 | 1.2 × 10 −8 |
| 6 | 9.33 | 0.008 |
| 7 | 9.31 | 0.005 |
| 8 | 9.22 | 0.04 |
| 9 | 9.10 | 0.4 |
| 10 | 9.08 | 0.4 |
Carriers of KIF6 719Arg received substantial and significant benefit from pravastatin therapy compared to placebo, but noncarriers did not. After adjustment for age, gender, and self-reported ethnicity, pravastatin therapy reduced the risk for fatal coronary events or nonfatal MI by 37% in carriers of the 719Arg allele (p = 0.0009; Table 4 ). However, pravastatin therapy did not significantly reduce the risk for events in noncarriers of the 719Arg allele (p = 0.98). The difference in benefit from pravastatin therapy between carriers and noncarriers was significant (p = 0.049 for interaction between KIF6 719Arg carrier status and treatment). Further adjustment for traditional risk factors (smoking status, LDL cholesterol, high-density lipoprotein cholesterol, history of hypertension, history of diabetes, and body mass index) and for population structure (using the principal components of genetic variability instead of self-reported ethnicity) did not appreciably change the benefit from pravastatin therapy in carriers (36% reduction, p = 0.0016) or noncarriers (10% reduction, p = 0.59), although the difference in benefit between carriers and noncarriers was no longer statistically significant (p = 0.14 for interaction between KIF6 719Arg carrier status and treatment). In those whose ethnicity was not Caucasian, pravastatin therapy reduced the risk for events in carriers of the 719Arg allele by 70% (p = 0.012), after adjustment for age, gender, and self-reported ethnicity, but did not significantly reduce the risk for events in noncarriers (36% reduction, p = 0.75).
| KIF6 Trp719Arg Subgroup | Treatment | Events | Total | Model 1 ⁎ | Model 2 † | Model 3 ‡ | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | p Value | HR | 95% CI | p Value | HR | 95% CI | p Value | ||||
| Arg/Arg + Arg/Trp | Pravastatin | 89 | 951 | 0.63 | 0.49–0.83 | 0.0009 | 0.62 | 0.47–0.82 | 0.00067 | 0.64 | 0.49–0.85 | 0.0016 |
| Placebo | 133 | 922 | Reference | Reference | Reference | |||||||
| Arg/Arg | Pravastatin | 25 | 240 | 0.69 | 0.40–1.17 | 0.17 | 0.71 | 0.41–1.23 | 0.22 | 0.62 | 0.35–1.08 | 0.094 |
| Placebo | 29 | 204 | Reference | Reference | Reference | |||||||
| Arg/Trp | Pravastatin | 64 | 711 | 0.61 | 0.45–0.84 | 0.0022 | 0.59 | 0.43–0.82 | 0.0014 | 0.63 | 0.46–0.87 | 0.0047 |
| Placebo | 104 | 718 | Reference | Reference | Reference | |||||||
| Trp/Trp | Pravastatin | 57 | 616 | 1.01 | 0.69–1.45 | 0.98 | 0.96 | 0.66–1.40 | 0.85 | 0.90 | 0.62–1.32 | 0.59 |
| Placebo | 57 | 614 | Reference | Reference | Reference | |||||||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
