Selected genetics studies from CROI

Svilen Konov, HIV i-Base

Hemocromatosis (HFE) gene mutations have been associated with a decreased risk of peripheral neuropathy (PN) in HIV-1 seropositive patients during ART. Silvia Costarelli and colleagues from the University of Brescia, Italy , looked into this possible association. [1]

Until recently, the pathogenesis of PN was related to mitochondrial injury and oxidative stress but the actual mechanism of development of this condition is not yet fully understood. A study from 2006 showed that the patients with hemocromatosis (HFE) mutations (C282Y and H63D) are at a lower risk of peripheral neuropathy. [2]

That study, however, had limitations, particularly that the diagnosis of PN was not confirmed by electromyography. In order to confirm or reject the findings, the team from Italy conducted a cross-sectional study in 57 patients affected by peripheral neuropathy. Baseline demographcs included 82% male, mean age 46 years, 67% HCV co-infected, 9% alcohol users, 5% diabetic, median CD4 nadir 75 cells/mm3, median exposure to d-drugs 554 days..

Among the group, 36/57 (63%) had no HFE mutations, 14 (25%) had H63D heterozygosis, 3 (5%) had H63D homozygosis, and 4 (7%) had C282Y heterozygosis. No statistically significant differences were found with respect to demographic and risk factors for peripheral neuropathy comparing patients with or without HFE gene mutations.

Most of patients (35/57; 61%) had 3 or more risk factors for PN (alcohol use, diabetes, HCV/HBV co-infection, CD4 nadir <100 cells/mm3, HIV RNA zenith >5 log/mL, cumulative exposure to d-drugs >554 days). The group of patients with no HFE mutations was not significantly different for number of risk factors for PN (67% and 58% with >/=3 risk factors in patients with and without mutations, respectively; p = 0.53). Since both patient groups with and without HFE mutations had similar and comparable numbers of peripheral neuropathy risk factors, these mutations did not appear to protect them from peripheral neuropathy.

Lipptrott and colleagues from the Liverpool Pharmacology Group and the Royal Free Medical School looked into a possible influence of PD1 genotype on virological and immunological response in patients initiating HAART [3].

The regulatory single nucleotide polymorphism (SNP) located in an enhancer within the fourth intron of the PD1 gene (PD1.3) has been associated with disease progression in various autoimmune disorders such as rheumatoid arthritis and systemic lupus erythematosis. PD1.3 (A allele) disrupts binding of the Runx1 transcription factor to the enhancer and thereby alters the regulation of PD1 gene expression. The above study investigated the effect of PD1.3 genotype in patients initiating ART.

77 treatment-naive subjects, starting on efavirenz were included. The median viral load and CD4 count at baseline were 4.9 log and 202 cells/mm3 respectively. Genomic DNA was extracted from whole blood and normalised to 20 ng.mL. PD1.3 and delta-32 genotype were then assessed for their influence on the time to virological suppression and differences in CD4 and CD8 counts (and percentage) from baseline to 8 months after initiation of therapy. For

PD1.3, 15 heterozygotes and 1 mutant homozygote were observed; for CCR5 delta-32, 7 heterozygotes were observed. Both SNP were in Hardy-Weinburg equilibrium. There were no associations (virological or immunological) with CCR5 delta-32 heterozygosity. Similarly, there were no associations with PD1.3 genotype at baseline. However, there was a trend in time to undetectable viral load between GG and GA individuals for PD1.3 (16 [8 to 16] vs. 12 [8 to 16] weeks, respectively; p = 0.07). CD4 counts were significantly lower in GA than GG individuals at 2, 4, 6, and 8 months (eg, 240 [85 to 628] vs. 379 [10 to 949], at 8 months; p = 0.009). There was no association with absolute CD8 counts, but a higher percentage of CD8 cells was observed in GA vs. GG from 2 to 8 months (e.g. 61 [38 to 77] vs. 56 [17 to 83]; p = 0.04). The researchers concluded that the PD-1 genotype is related to immunological progression in response to therapy.

A paediatric study using the data from the Pediatric AIDS Clinical Trials Group (PACTG) attempted to shed light on the effect of the CYP2B6-516 genetic variants on the pharmacokinetics of nevirapine and clinical responses. [4]

The study deserves commending as it is among the few pharmacogenomics studies in this population and has an impact on the expertise of clinical management of HIV infection in children. Genetic polymorphisms in drug metabolising enzymes and transporters affect ART drug disposition. These factors, along with maturational changes, likely affect ART pharmacokinetics in children; however, their potential influence on pharmacokinetics and clinical outcomes are not well characterised.

It has been shown that the gene polymorphism cytochrome P450 2B6 (CYP2B6)-G516T is associated with decreased activity of CYP2B6 in liver and increased plasma efavirenz (EFV) levels, but no data are available regarding the impact of the CYP2B6-G516T on the pharmacokinetics of nevirapine (NVP) in children.

Real-time polymerase chain reaction (RT-PCR) was used to genotype polymorphisms of CYP2B6 in 126 children, who received NVP and protease inhibitors (PI, nelfinavir [NFV], or ritonavir [RTV]) as part of their HAART regimens. Pharmacokinetic data at week 4 was used to calculate area under the curve (AUC) and determine clearance. Immunologic and virologic data were also collected during HAART.

Of the 126 children, 50 (40%) had the CYP2B6-516-G/G (wild type), 63 (50%) had the G/T (heterozygous), and 13 (10%) had the T/T genotype (homozygous). Age, gender, race/ethnicity, concomitant PI regimens, baseline HIV-1 RNA, and CD4 T cell percentage were not different among the 3 groups.

NVP AUC in children with the T/T genotype (95.0 ug*hr/mL) was significantly higher than those with the G/G (57.8 ug*h/mL, p = 0.004) and G/T genotype (58.3 ug*h/mL, p = 0.003). NVP clearance in children with the T/T genotype (1.6 L/h/m2) was significantly decreased compared to those with the G/G (2.3 L/h/m2, p = 0.001) and G/T genotype (2.1 L/h/m2, p = 0.003). Children with the T/T genotype had an increase in CD4 T cell percentages (+7.0%) compared to those with the G/G (+3.4%, p = 0.008) and G/T genotype (+5.5%, p = 0.04) at week 12. This trend was observed at week 24 (p = 0.01).

Furthermore, in a subset of 58 subjects whose HIV-1 RNA data were available, children with the T/T genotype had a more rapid decline of log HIV-1 RNA from baseline to week 12 (–39%) compared to those with the G/G (–18%, p = 0.002) and G/T genotype (–14%, p = 0.006). These results clearly show that the G516T genotype alters NVP pharmacokinetics, as well as immunologic and virologic responses to NVP-containing HAART regimens in children.

A study from Celeste Pearce and colleagues from the Women’s Interagency HIV Study (WIHS) explored a broader set of 10 common genetic variants on immunological and virological responses to ARVs in 537 women who initiated HAART between 1995 and 2004. [5]

The researchers defined response to therapy as achieving an undetectable viral load within 12 months of starting HAART. In this group, 322 were responders and 215 were non-responders. Genotyping was carried out using the TaqMan Allelic Discrimination Assay and the Sequenom MassARRAY system. Data were analysed using logistic regression. Genotype was examined using a log additive model as well as assuming no mode of inheritance.

The 10 variants analysed were in ADH1B, CYP2B6, CYP2D6, CYP2E1, MDR1, and UGT1A. The minor allele frequencies of the variants ranged from 9.3% to 49.1%, with substantial ethnic differences. The 2 variants that have been studied fairly extensively with response to HIV therapy are: rs3745274 (G516T) in CYP2B6 and rs1045642 (C3435T) in MDR1.

After adjusting for ethnicity, no association was found between G516T and response to therapy (p = 0.88). After adjusting for ethnicity, peak CD4 cell count, and peak viral load prior to HAART initiation, there was a marginally statistically significant association between response to therapy and MDR1 C3435T (OR = 0.74, 95%CI 0.55 to 1.01, p = 0.057). Some evidence indicates that this association may be restricted to protease-inhibitor-based regimens (OR = 0.67, 95%CI 0.47 to 0.96, p <0.05), but there was no statistically significant interaction between type of HAART and MDR1 C3435T. The analysis of the remaining eight variants has not yet been completed.


  1. Costarelli S., Benerini L., et al. No Evidence of Relationship between HIV Peripheral Neuropathy and the Presence of Hemocromatosis Gene Mutations. Abstract 378.
  2. Kallianpur AR., Canter JR., et al. Hemochromatosis (HFE) gene mutations and peripheral neuropathy during antiretroviral therapy. AIDS. 2006 Jul 13;20 (11):1503-13.
  3. Lipptrott N., Mahungu T., et al. Influence of PD1 Genotype on Virological and Immunological Response in Patients Initiating ART. Abstract 452.
  4. Akihiko S., Sarles E., et al. CYP2B6-516 Genetic Variants Affect the Pharmacokinetics of Nevirapine and Clinical Responses in HIV-1-infected Children Receiving HAART. Abstract 735.
  5. Pearce C., Gandhi M., et al. Pharmacogenomics of HAART Metabolism and Transport: The Women’s Interagency HIV Study. Abstract 785.

Links to other websites are current at date of posting but not maintained.