Earlier HIV detection with prototype Abbott assay

Create PDF

The prototype assay reduced the second diagnostic window, or trough effect. Several trough bleeds in three of 14 seroconversion panels had between two- and sevenfold higher signal compared with the Architect and Elecsys tests. “The higher signal is driven by enhanced antigen and antibody sensitivity of the prototype,” Dr. Qiu said.

The enhanced antigen sensitivity of the prototype assay, she noted, “did not compromise antibody detection.”

The assay’s antibody sensitivity was tested against the Architect in 543 samples from patients who were at different stages of HIV: acute, symptomatic, and AIDS. The prototype and Architect assays detected 100 percent of the samples, but three of the Architect results showed low reactivity. The sensitivity of low-titer samples was improved two- to 13-fold by the prototype assay over the Architect assay. “The signal-to-cutoff for those three samples is 3.8 to 16,” Dr. Qiu said. “But in the prototype, the 3.8 becomes 48.9, and you can see the high signal. This indicates the antibody sensitivity is also improved by the prototype.”

The prototype assay maintained broad detection of antibodies against different HIV subtypes. In a study of 473 HIV subtype samples, which included 327 HIV-1 group M samples, 50 HIV-1 group O samples, and 96 HIV-2 samples, the prototype assay detected all samples with a very high signal, Dr. Qiu said.

“In addition, we compared the prototype performance between Architect and Abbott’s new Alinity platform, and we found excellent correlation.” This comparison involved HIV-1 groups M and O, HIV-1 URF, and HIV-2 samples for antibody detection and, for antigen detection, 52 dilutions of HIV-1 groups M and O and HIV-2 virus isolates.

“From the 473 antibody subtype panels, the slope is 1.1, and the correlation coefficient R2 is .99,” she said. “It was similar for the antigen detection. The slope is one, and the R2 is .99 for the results from the 52 diluted HIV virus isolate samples.”

The 473 specimens in the HIV genotype panel used in this study were collected and characterized by Abbott’s Global Viral Surveillance Program, launched 25 years ago. It has a library of more than 70,000 samples containing HIV and hepatitis viruses from six continents and 45 countries. “These samples are used to track virus evolution, challenge the performance of Abbott’s tests, and drive continual improvement of testing technology,” Dr. Qiu says.

The prototype assay also demonstrated high specificity. In an evaluation performed on 4,316 fresh blood donors, it demonstrated a specificity of 99.93 percent. When tested on 1,690 specimens from a low-risk population, specificity was 99.76 percent. Of those 1,690 samples, 14 were found to be repeatedly reactive, 10 were confirmed HIV positive, and four were falsely reactive.

In addition, “The signal distributions of both populations were tight,” Dr. Qiu said. “For the blood donor population, the standard deviation to cutoff is 31, and for the low-risk population it is 22. In general, a standard deviation to cutoff greater than 10 is suitable for blood screening, so a standard deviation to cutoff of 31 is considered a very specific assay.”

Another specificity improvement demonstrated by the prototype is reduced heterophilic antibody interference, she said. “There are nine Architect false reactive samples due to heterophilic antibody interference. Some of them had a quite high signal, greater than 400.” All false reactive samples became completely inactive in the prototype. “This shows the improvement.”

The assay will undergo external evaluation by key laboratories. Abbott plans to make it available on the Architect and Alinity platforms.

“This fourth-generation prototype assay is designed to detect highly infectious people who may be missed by existing tests and can play an important role in helping the global community end the HIV epidemic by 2030,” Dr. Qiu says.

Amy Carpenter Aquino is CAP TODAY senior editor.