Absence of HLA-B*57:01 among patients on antiretroviral therapy in Nigeria: Implications for use of Abacavir

Abstract

Background

Presence of Human Leukocyte Antigen (HLA) B*57:01 allele predicts hypersensitivity reaction (HSR) to Abacavir (ABC), a nucleoside reverse transcriptase inhibitor used for HIV treatment. However, prevalence of this allele among Nigerians with HIV is yet to be established. We aimed to determine the prevalence of HLA B*57:01 allele among Nigerians with HIV infection.

Methods

We conducted a multicenter cross-sectional epidemiologic survey. Between April 2016 and April 2017, patients were enrolled across 5 HIV treatment facilities in Nigeria. Participants’ demographic information and history of ABC exposure were obtained, and venous blood obtained for HLA typing.

Results

One thousand five hundred and four (1,504) adults were enrolled, with a mean age of 44.6 ± 10.7 years, 1,078 (71.7%) were female. 1,463 (97.3%) were on antiretroviral therapy. ABC use was reported by 12 (0.8%) participants and none reported HSR. Of 1,500 blood samples that were processed, 1,458 (97.2%) were successfully typed. Out of these 132 (9.1%) were HLA-B*57 positive using non-specific low-resolution HLA-B*5701 primer mix. On further analysis, none of the 132 samples (0%) had the HLA-B*5701allele.

Conclusion

HLA-B*5701allele is rare among Nigerians.

Introduction

Abacavir (ABC) is a nucleoside reverse transcriptase inhibitor (NRTI) used for the treatment of human immunodeficiency virus (HIV) infection. On account of its efficacy and favourable safety profile, ART combinations which include ABC have become a preferred first-line treatment for HIV treatment^1–3. While few long term adverse effects have been reported with the use of ABC, hypersensitivity to ABC still remains a major challenge ⁴. Hypersensitivity to ABC occurs in about 5–8% of persons who initiate ABC^{5, 6}, and this incidence could vary based on the population as well as mode of diagnosis. Occurrence of hypersensitivity following exposure to ABC precludes the subsequent use of ABC as a treatment option. Re-exposure to ABC after an initial hypersensitivity reaction often results in more severe reaction which may be fatal.

The association between ABC hypersensitivity reaction and the Human Leukocyte Antigen (HLA) B*57:01 was first demonstrated among 200 Caucasians in an Australian HIV cohort in 2002⁷. This association has been supported by other studies^{8, 9}. Genetic screening for HLA B*57:01 prior to initiating ABC showed no cases of hypersensitivity to the drug in persons who tested negative to HLA B*57:01^{10, 11}. On account of this, HIV treatment guidelines now recommend screening for HLA B*57:01 prior to initiating ABC and persons who test positive should not receive ABC¹².

The prevalence of HLA B*57:01alelle has been shown to vary widely among populations tested, with low prevalence rates reported among people of African descent ^13–15. In North America, prevalence has been reported to be higher among Caucasians (7.2%) than African Americans (2.8%)¹⁶. Few prevalence studies have been carried out in sub-Saharan African where about 80% of persons living with HIV (PLHIV) reside and none has been done in Nigeria, the country with the second highest number of PLHIV. The objective of this study was to determine the prevalence of the genetic marker HLA B*57:01 amongst Nigerians living with HIV/AIDS.

Methods

Ethics Statement

Ethical approval was obtained from all participating sites (Federal Medical Center Yenagoa: From the Head of Clinical Services, dated: 4^th November 2014; Lagos University Teaching Hospital: From Health Research and Ethics Committee, dated 3^rd November, 2014, Reference Number: ADM/DCST/HRC/2198; Ahmadu Bello University Teaching Hospital: the Health Research Ethics Committee, dated 18^th November, 2014, Reference Number: ABUTH/HREC/L33/2014; Federal Medical Center Makurdi: From the Health Research Ethics Committee, dated 29^th October, 2014, Reference Number: FMH/FMC/MED. 108/ VOL. I/X; Jos University Teaching Hospital: From the Institutional Health Research Ethic Committee, dated 6^th November, 2014, Reference Number: JUTH/DCS/ADM/127/XIX/6026. In addition written informed consent was obtained from all study participants and all study data was treated with confidentiality.

Study Design

We conducted a multicenter cross-sectional survey among adults with HIV-infection across five HIV treatment facilities in Nigeria. Study sites were chosen based on geographic distribution to capture the ethnic diversity of the country. Study locations were Jos (North-Central), Lagos (South-West), Makurdi (Middle belt), Yenagoa (South-South) and Zaria (North).

Sample Size

The minimum sample size for the study was obtained using an online sample size calculator (select-statistics.co.uk/calculators/sample-size-calculator-population-proportion), accessed on the 3^rd of September 2013. A minimum sample size of 753 was required based on an estimated prevalence of HLA B27 01 allele of 2%, an error margin of 1%, and a 95% confidence interval¹⁶. Our target enrolment was to 1 500 participants, with 300 participants recruited at each of the study sites.

Patient Recruitment and Data & Sample Collection

Eligible participants were adults >18 years, with confirmed HIV-infection. Between April 2016 and April 2017, consecutive adults presenting to the study sites were approached to participate in the study during their routine clinic visits. Each study participant gave written informed consent prior to study enrolment. Demographic information and history of ABC exposure were obtained from each participant and three millilters (mls) of venous blood was obtained from each participant in bottles with ethylenediaminetetraacetic acid (EDTA) for HLA typing. Blood samples at each study site were batched and transported once a week to a central laboratory for analysis.

DNA extraction

The DNA from the blood samples were extracted using the Gentra Puregene Plus (1000ml) Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol¹⁷. This kit was best suited for the study because of its high-volume sample processing and high DNA elution yield. The ratio of sample volume, lysis solutions, protein precipitation solution and wash buffers used were in accordance to the manufacturer’s protocol. In addition, to completely eliminate all traces of RNA, RNase A solution was added during extraction, therefore yielding only pure DNA.The final elution volume of 300μl was quantified with a spectrophotometer at A260/280nm ratio (SmartSpec Plus, Bio-Rad, Spain). All samples were stored at −40 degrees centigrade (° C) prior to further analyses.

HLA Typing

Samples were initially assessed using non-specific low-resolution HLA-B*5701 primer mix containing various subtypes of the B57 gene (OLERUP GmbH,Vienna, Austria). Samples found to be HLA-B*57 positive were further subjected to a high-resolution screening using sequence specific primer methodology. Low-resolution reaction included 30–60ng/μl of DNA, bulk primer solution and PCR Master Mix (OLERUP GmbH, Vienna, Austria) in accordance to manufacturer’s protocol. DNA samples with concentrations below 15ng/ μl were exempted from PCR amplification due to failed amplification revealed during our pilot analysis for this study.

For high-resolution analysis, PCR-Sequence-Specific-Primer methodology was used. 50ng/μl of DNA, PCR Master Mix complete with Taq (OLERUP GmbH, Vienna, Austria) and water was added to each well pre- aliquoted primer solutions in 0.2 ml wells cut PCR plate. Thermocycling parameters for both low-resolution and high-resolution amplification included initial denaturation at 94° C for 2 minutes (min) followed by 10 cycles of 94° C for 10 seconds (sec) and 10 cycles of annealing and extension at 65° C for 60 sec. Afterwards, a second phase of denaturation at 94° C for 10 sec, annealing at 61° C for 50 sec and extension at 72° C for 30 sec in 20 cycles. Ramp rate of thermal cycler was reduced to 1°c per sec at every step. PCR products were electrophoresed on a 2% agarose gel (Bio-Rad, Spain).

Data Analysis

Demographic information was summarized using descriptive statistics. Prevalence of HLA-B*57:01:01 allele was reported as a percentage. Statistical analysis was carried out using STATA version 14, College Station, Texas, USA.

Results

A total of 1504 patients were enrolled in the study. The mean age was 44.6 ± 10.7 years, and 1078 (71.7%) were female. A total of 1463 (97.3%) were on antiretroviral therapy. ABC use was documented in 12 (0.8%) patients. None of the patients reported a diagnosis of HSR to ABC. Table 1 summarizes demographic characteristics of study population.

Table 1:

Demographic Characteristics of 1504 Nigerians with HIV evaluated for HLA-B*57 01 Allele

Characteristic	Number N=1504	(Percentage)
Sex
Male	426	(28.3)
Female	1078	(71.1)
Age group (years)
≤ 29	137	(9.1)
30 – 39	542	(36.0)
40 – 49	498	(33.1)
50 – 59	223	(14.8)
≥ 60	104	(6.9)
Ethnicity
Hausa/Fulani	202	(13.4)
Igbo	200	(13.3)
Yoruba	128	(8.5)
Others	974	(64.8)
On Antiretroviral Therapy
Yes	1463	(97.3)
No	41	(2.7)
Ever taken Abacavir
Yes	12	(0.8)
No	1492	(99.2)

One thousand five hundred (1,500) blood samples were suitable for analysis (4 sample spilled) and 1458 (97.2%) were successfully typed. A total of 132 (9.1%) were HLA B57* positive using non-specific low-resolution HLA-B*57:01 primer mix. On further analysis none of the 132 samples (0%) had the HLA-B*57:01allele. Table 2 shows the frequency of HLA B57* identified in our patient population.

Table 2.

Frequency of HLA-B*57 allele among 132 Nigerians with HIV

HLA-B*57 allele	Number (Percentage)
*5701	0 (0)
*57:02:01– 57:02:02, 57:12, 57:42	21 (15.9)
*57:03:01– 57:03:02, 57:17, 57:66, 57:70	93 (70.5)
*57:04	7 (5.3)
Unidentifiable allele	11 (8.3)
Total	132

Discussion

Our aim was to determine the prevalence of HLA-B*57:01 allele in the Nigerian population. To the best of our knowledge, this is the first published report of HLA HLA-B*57:01 allele in Nigeria. None of the patients in our study population had the HLA-B*57:01, although other closely related alleles were identified. These closely related alleles are not associated with HSR to ABC. However, HLA B*57:03 which was one of the most frequent alleles found is associated with slow progression of HIV¹⁸. The clinical significance of the other identified alleles is yet to be determined.

Our results support emerging data indicating a low prevalence of HLA-B*57:01allele among Africans. The PREDICT-1 study reported which enrolled 1956 participants across 19 countries reported a prevalence of 2.8% in blacks and 7.2% whites in North America¹⁶. In another large epidemiologic study in the UK which enrolled 1,494 subjects, 770 (52%) being black, prevalence was 0.26% in Blacks compared to 7.93% in Whites. Few prevalence studies have been carried out in Africa; in Kenya, a prevalence of 0.8% was recently reported among 1004 samples analysed¹⁹, while none (0%) of the 161 samples from HIV-infected patients in Uganda had the HLA-B*57:01 allele²⁰.

Very few participants in our study cohort (0.8%) had prior exposure to ABC and none reported symptoms suggestive of HSR. The low frequency of exposure to ABC in this population may be related to the uncertain risk of ABC hypersensitivity. Given the strong association between HLA-B*57:01and ABC HSR, our study suggests that the risk for ABC HSR in our population is low. While current guidelines recommend routine screening for HLA-B*57:01prior to ABC initiation, the cost effectiveness of this recommendation needs to be evaluated in Nigeria and other countries with very low prevalence of HLA-B*5701.

Our study had some limitations. While our study sites were chosen to capture the ethnic diversity of Nigeria, we cannot confirm that our participants included members of the over 250 ethnic groups in Nigeria. However, with the large sample size and the diverse locations, we believe our findings are generalizable to the whole population.

In conclusion, our findings show that HLA-B*57:01 allele is very rare among Nigerians. While current guidelines recommend routine screening for HLA-B*5701prior to ABC initiation, the cost effectiveness of this recommendation needs to be re-evaluated in Nigeria and other countries with very low prevalence of HLA-B*5701.