Randomized Trials Results Back Gene Screening for Abacavir Reaction: PREDICT 1 and SHAPE Studies
4th IAS Conference on HIV Pathogenesis, Treatment, and Prevention
July 22-25, 2007
PREDICT-1, a randomized, double-blind trial, determined that screening for the genetic marker HLA-B*5701 reliably prevents hypersensitivity reactions to abacavir . Principal investigator Simon Mallal argued that the findings should sweep aside any lingering doubts about making this gene test a routine part of HIV care. A separate study found that HLA-B*5701 screening works as well in blacks as in whites .
For PREDICT-1 an international team recruited 1956 abacavir-naive people and randomized them to start an abacavir regimen under the standard of care--clinically diagnosed hypersensitivity--or to start abacavir with the standard of care plus HLA-B*5701 screening. In the second group anyone with a positive B*5701 test got excluded, while those with a negative test continued.
The trial had two primary endpoints: (1) incidence of clinically suspected hypersensitivity during 6 weeks of observation and (2) incidence of hypersensitivity immunologically confirmed with skin patch testing during 6 weeks of observation. Of the first 100 abacavir-tolerant people through week 6, none had a positive patch test. That meant the specificity of patch testing in spotting the reaction came to 100%.
For the primary endpoints Mallal analyzed 1650 people--everyone randomized who started abacavir and either completed the 6-week observation or stopped abacavir early because of diagnosed hypersensitivity. About three quarters of this group were men, 84% were white, and 12% were African in heritage. About 20% started the trial with no antiretroviral experience.
None of 802 gene-screened people had patch-test-confirmed hypersensitivity, compared with 27 of 842 people (2.7%) who did not have B*5701 screening. That meant screening lowered the risk of an immunologically confirmed reaction 97% (P < 0.0001). While only 27 of 803 people (3.4%) in the B*5701-screened group had a clinically suspected reaction, 66 of 847 (7.8%) in the control group had clinically suspected hypersensitivity. That meant screening cut the risk of a clinically suspected reaction 60% (P < 0.0001). For immunologically confirmed hypersensitivity, B*5701 had a negative predictive value of 100%. And for clinically suspected hypersensitivity, the gene test had a negative predictive value of 96%.
Mallal concluded with two recommendations:
- Where validated B*5701 screening is available, clinicians should consider screening any HIV-infected person who has not taken abacavir and for whom an abacavir regimen is being considered.
- Where validated B*5701 screening is not available, it is reasonable to begin abacavir with appropriate clinical vigilance.
A separate study by the same group determined that gene screening is 100% sensitive for HLA-B*5701 in both whites and blacks . Sensitivity in identifying HLA-B*5701 fell when physicians relied on clinical signals alone to diagnose abacavir hypersensitivity. The Australian team cautioned that "HLA-B*5701 screening may augment clinical management of hypersensitivity but must never substitute for clinical vigilance."
1. Mallal S, Phillips E, Carosi G, et al. PREDICT-1: a novel randomised prospective study to determine the clinical utility of HLA-B*5701 screening to reduce abacavir hypersensitivity in HIV-1 infected subjects (study CNA106030). 4th IAS Conference on HIV Pathogenesis, Treatment, and Prevention. July 22-25, 2007. Sydney. Abstract WESS101.
2. Saag M, Balu R, Brachman P, et al. High sensitivity of HLA-B*5701 in whites and blacks in immunologically-confirmed cases of abacavir hypersensitivity. 4th IAS Conference on HIV Pathogenesis, Treatment, and Prevention. July 22-25, 2007. Sydney. Abstract WEAB305.
PREDICT-1: a novel randomised prospective study to determine the clinical utility of HLA-B*5701 screening to reduce abacavir hypersensitivity in HIV-1 infected subjects (study CNA106030)
Reported by Jules Levin
4th IAS Conference, 22-25 July, Sydney, Australia
Mallal S, Phillips E, Carosi G, Molina J-M, Workman C, Toma_ic J, JŠgel-Guedes E, Rugina S, Kozyrev O, Flores Cid J, Hay P, Nolan D, Hughes S, Hughes A, Thorborn D & Benbow A on behalf of the PREDICT-1 study team.
Wednesday 25th July 2007
--In clinical settings where validated screening methods are available, clinicians should consider screening for HLA-B*5701 in any HIV-infected patient who has not previously been exposed to ABC and for whom ABC therapy may be considered
--Where screening is not readily available, it is reasonable to initiate ABC with appropriate clinical vigilance
Clinical vigilance remains the cornerstone of successful ABC risk management
Key inclusion criteria:
--Naive to ABC
--Pre-study need for ABC treatment
Key exclusion criteria:
--Contraindications to ABC
--Known HLA-B*5701 status
Double-Blind Study Design
Abacavir-naive subjects randomized to:
--ABC-containing regimen, HSR monitoring according to Standard of Care (control arm), or
--ABC-containing regimen, HSR monitoring according to Standard of Care plus HLA-B*5701 screening (screening arm): HLA-B*5701 positive subjects were excluded; HLA-B*5701 negative subjects continued.
1. retrospective high resolution typing
2. prospective high resolution typing
Incidence of clinically-suspected ABC HSR during the 6-week observation period
--90% power to detect >50% reduction in HSR rate from 7.3 to 3.6%
Incidence of immunologically-confirmed (clinical diagnosis plus positive skin patch testing1) ABC HSR during the 6-week observation period
--99% power to detect 80% reduction in HSR rate from 4.6 to 0.91%
1 Assessment by a central independent clinical evaluation committee
--Immune cell-mediated reaction
--Research tool used to identify patients with immune-mediated abacavir HSR
Specificity: 100/100 = 100% 95% CI (96.4%, 100%)
Note: all ABC tolerant subjects patch-tested were HLA-B*5701 negative
- The first 100 ABC-tolerant patients through week 6
-- ITT (n=1956) - number of subjects randomised
-- ITT exposed (n=1772) - number of subjects with at least one dose of ABC-containing medication
PRIMARY POPULATION FOR ANALYSIS
ITT evaluable (n=1650) - all randomised patients who started abacavir and either completed the six-week observation period or stopped ABC early due to HSR
Test 100 subjects:
--Treat 94 subjects at low risk for ABC HSR
--Prevent 4 ABC HSR events
--Inappropriately exclude ABC in 2 subjects
--Open screening may also reduce clinical overdiagnosis of HSR
Example shown is based upon PPV calculated from PREDICT-1