In the present study, conducted in a North American HIV-positive cohort including 2 major races, white and black, 3 main observations were made regarding TVLS as the response to treatment with HAART: First, the treatment response did not differ significantly by race. Disparate virologic responses to HAART between races or ethnicities may [41
] or may not occur [42
]. Interethnic differences in certain allele frequencies (eg, higher frequencies of the CYP2B6
516T and 983C alleles in black populations and of the CCR5
Δ32 allele in white populations) may be correlated with differences in response to HAART. In our study, despite significant differences in the allele frequencies of the drug-metabolizing enzyme, transporter, and chemokine receptor genes between white and black patients (), we did not observe a significant difference in TVLS between the 2 groups.
Second, in the drug-metabolizing enzyme and transporter set of genetic polymorphisms, the treatment response differed significantly only by CYP2B6
516G>T genotype in black patients, but not in white patients. Among black patients, those who carried the 516G allele achieved virologic success significantly earlier. Interestingly, this association was nonsignificant in the Cox model that included age, sex, baseline CD4+
T-cell count, and baseline viral load. Many studies involving ethnically diverse patients have analyzed the association between CYP2B6
genotypes and the pharmacokinetics of EFV and NVP [8
] and also between CYP2B6
genotypes and EFV-related treatment responses [8
]. Although the results pertaining to the association between CYP2B6
genotypes and the pharmacokinetics of these drugs are mostly consistent across these studies, those pertaining to the association between CYP2B6
genotypes and EFV-related treatment responses remain unclear. Furthermore, the results of these studies [8
] suggest that the interrelationship among ethnicity, CYP2B6
genotypes, and the phenotypes considered therein is far from clear; ethnicity may [8
] or may not [9
] influence genotype-phenotype associations. Our study was not designed to analyze the pharmacokinetics of any of these drugs or any treatment response other than TVLS. However, when we consider our results together with those of other studies [8
], it seems either that the drug-metabolizing enzyme/transporter genotypes considered so far do not substantially influence response to HAART or that the extent of their influence may be cohort dependent.
To our knowledge, whether UGT2B7
SNPs affect antiretroviral treatment response has not been reported elsewhere. Several promoter variants were reported to alter expression of UGT2B7
], and it is still unknown whether coding SNPs other than 802C>T result in impaired catalytic activity toward EFV. Recently, a synonymous coding SNP, 735A>G, was found to be associated with faster AZT clearance in patients coinfected with HIV and tuberculosis, and with higher AZT glucuronidation in vitro [45
]. Therefore, additional polymorphisms in UGT2B7
as well as relationships between UGT2B7
polymorphisms and treatment responses to AZT and EFV in other cohorts should be investigated.
Third, in the chemokine receptor set of genetic polymorphisms, the treatment response differed significantly only by CCR5
−2459G>A genotype in black patients, but not in white patients. Among black patients, those who carried the −2459G allele achieved virologic success significantly earlier. No significant differences in baseline viral load (P
= .61) or baseline CD4+
T-cell count (P
= .48) were observed among the 3 CCR5
−2459 genotypes in black patients. The association between the −2459G allele and TVLS remained significant even when CCR2
190G>A as well as all the drug-metabolizing enzyme and transporter genotypes, including CYP2B6
516G>T, were included in the Cox model, suggesting that CCR5
−2459G>A genotype had a stronger influence on TVLS in black patients. As described above, studies that have analyzed the effects of chemokine receptor gene polymorphisms on response to HAART have yielded inconsistent results [20
]. Because these studies were conducted in various populations under a variety of designs, it is difficult to compare our findings directly with the findings of these studies. Nevertheless, the race-specific influence of the −2459G allele, observed in black patients in our study, was not reported in any of the other studies.
Some of the studies reported elsewhere found a significant association of CCR5
Δ32 with improved responses to HAART [24
]. In our white patients, we did not observe significant association between this deletion and TVLS. In addition to probable differences in cohort and design between those studies and ours, it may be that the frequency of the Δ32 allele in our white patients (.06) was too low for us to detect a difference in the outcome between wild-type (wtwt) homozygotes (n
= 155 [88.57%]) and wtΔ32 heterozygotes (n
= 20 [11.43%]).
The observation that CCR5
−2459G exerted a race-specific influence on response to HAART raises the question of whether the genetic characteristics of this cohort are in any way unique. In an attempt to answer this question, we further analyzed patterns of LD at the CCR2
locus in both groups of patients, by including other known SNPs in the CCR5
promoter region (−2733A>G, −2554G>T, −2135T>C, −2132C>T, −2086A>G, and −1835C>T). When we performed pairwise LD analysis of all the 9 CCR2
polymorphisms, the overall patterns of LD were similar between whites and blacks (data not shown) and were similar to the patterns seen in comparable populations [46
]. We also quantified admixture in both groups of patients by using the Duffy blood group antigen (FY)
as a population-specific marker. Among the 3 most common FY
alleles, FY*A, FY*B,
and FY*BES, FY*BES
is a key marker for African ancestry [47
]. Among white patients, frequency of the FY*BES
allele was .01, reflecting the extent (1%) of the African ancestry contribution to European American populations in the continental United States [47
]. Among black patients, frequency of the FY*BES
allele was .81 (FY*A,
.11), reflecting the extent (19%) of the European ancestry contribution to African American populations in the continental United States [47
]. Thus, our analysis of the CCR2
locus and admixture proportions, based on FY
alleles, did not reveal any unique genetic characteristics of this cohort. However, analyses should take into consideration genetic characteristics based on other genes potentially important in HIV/AIDS pathogenesis, located on chromosome 3 along with CCR2
, or on other chromosomes [40
]. Finally, to determine whether the race-specific influence of CCR5
−2459G in black patients is unique to this HIV-positive cohort requires future study of various North American and African cohorts.
We acknowledge that our study has some limitations. Most of the patients in our study had prior exposure to antiretrovirals, and this exposure did not differ significantly between races. Thus, it is important to note that for these patients, the first HAART regimen was not necessarily their initial therapy. However, we did not consider data regarding prior antiretroviral exposure into our genotype-phenotype association analyses. Adherence is a significant factor in response to HAART [50
], and the tool we used to assess adherence, self-reported 72-hour recall of missed doses, is admittedly less precise than other measures, such as pill counts or electronic monitoring. Nevertheless, this tool has been clinically validated in our patient population [35
], and we therefore feel confident that it provides a meaningful measure of adherence in this cohort. We intentionally selected a stringently high level of adherence as an inclusion criterion (≥90%), in an attempt to minimize the confounding effect of adherence on the results.
In conclusion, a strong, race-specific influence of CCR5 −2459G>A genotype on TVLS was observed in a North American, treated, adherent HIV-positive cohort. Understanding the possible mechanisms underlying this influence is important, because it would probably enhance our understanding of genetic factors influencing response to antiretroviral drugs in diverse worldwide populations, complementing our current efforts to further lessen the morbidity and mortality due to this global killer.