The Evolution of HIV Testing Continues

HIV testing is a key component of HIV prevention. It is this critical clinical encounter that serves as the starting point for diagnosing and treating persons who are infected and delivering preventive services to those who are uninfected. Because HIV testing is so important to prevention strategies for controlling the HIV epidemic in the United States, we read with great interest the article by Hurt and colleagues1 in this issue, which provides an excellent overview of the current options available for HIV testing in clinical, nonclinical, and research settings. Their update highlights recent changes to nomenclature, updated data—particularly on the window period of HIV tests—and updates to the laboratory algorithm for diagnosis of HIV infection, at a time when this information is changing rapidly. Hurt et al. refer to changes in the “official nomenclature” of HIV tests. Although the Centers for Disease Control and Prevention (CDC) does not determine official nomenclature for HIV test types, the CDC Division of HIV/AIDS Prevention has recently made changes to Web sites and other documents that refer to the different types of HIV tests. As discussed at the 2016 HIV Diagnostics Conference,2,3 the term “generations” began to appear in the literature shortly after HIV tests that used recombinant peptides instead of viral lysate antigens (the “2nd generation”) were developed.4–6 However, the “official” nomenclature likely gained traction when Owen et al.7 published an article including a discussion of generations, and CDC and others largely adopted the term for use in presentations, Web pages, and other documents. Indeed, a complete description of test generations appears in both the updated Clinical & Laboratory Standards Institute standards8 and the CDC/Association of Public Health Laboratories (APHL) guidelines for the laboratory diagnosis of HIV infection.9 However, as new HIV tests continued to become available, the lines between generations began to blur. In the 2008 article,7 the term generation was reserved for laboratory-based, instrumented immunoassays. As Hurt et al. reviewed, single-use, point-of-care rapid tests use different technology and probably should be considered separately. Nevertheless, both test manufacturers and authors evaluating these tests began to use the term generations to describe rapid tests. Originally, the generations described incremental improvements in test sensitivity and specificity. However, some of the newer tests within the same generation have different sensitivity for early infection.10 These differences can largely be explained by other aspects of test design, for example, whether they are lateral flow or immunconcentrating rapid tests, reagents used for detection of analytes, or the volume of sample required to perform the test.1 In addition, there are also IgG-sensitive rapid tests that differentiate HIV-1 from HIV-2, and new tests that differentiate p24-antigen detection from antibody detection, but have the same sensitivity during early infection as tests that report only one signal as “reactive for p24-antigen and/or HIV antibody.”1,3,10 As a result, in the article documenting seroconversion sensitivity on plasma specimens that Hurt et al. referenced,10 tests were described in terms of the analytes they can detect and the types of technology (instrumented, laboratory-based, vs. single-use, rapid) that they use to do so. These changes have been implemented in CDC Web pages and documents contained therein.11 In particular, the advantages/disadvantages of Food and Drug Administration–approved HIV tests guide12 may be particularly useful for clinicians and others who need to understand differences in characteristics of the tests available in the United States. In addition to changing the way we refer to types of HIV tests, CDC has also updated messaging about test window periods.13 On the basis of new data10 highlighted by Hurt and colleagues, information related to retesting after an exposure, when the initial test result is negative, but within the “window period” for the test being performed has been revised. These changes include defining the window period as the time between when a person may have been exposed to HIV and when a test can tell for sure if they have HIV or not, and defining that time as 45 days for laboratory-based antigen/antibody tests performed on plasma or serum specimens.13 This is much shorter than the previously recommended 90 days that was used for all HIV tests, regardless of specimen type, Clinical Laboratory Improvement Amendments–complexity level, or the analytes that the test could detect. However, it is important to note that the 45-day window period is only for laboratory-based HIV tests performed on serum or plasma. For point-of-care tests (single-use devices used on unprocessed blood or oral fluid specimens), there are not yet sufficient data to recommend reducing the window period to less than 90 days. We are currently assessing the sensitivity of point-of-care tests through the CDC-funded Project DETECT, which evaluates the performance of rapid tests when performed on unprocessed specimens in real time through serial follow-up of participants identified during the process of seroconversion.14 One goal of this study is to provide the data needed to update the window periods for these tests in the near future. Evaluations of new tests in Project DETECT might also inform future updates to the recommended algorithm for laboratory diagnosis of HIV infection. Although the HIV diagnostic algorithm for the United States remained largely unchanged from 1989 through June 2014, CDC has already updated information related to the 2014 algorithm twice.15,16 The first update concerned the change from the Multispot HIV-1/HIV-2 differentiation test that was available in 20149 to the Geenius test that replaced it permanently in late 2016. The Geenius test has additional test interpretations that Multispot did not,15 including HIV-2 and HIV indeterminate results. Because Hurt et al. refer to the 2014 recommendation in Figure 4 of their work,9 they fail to capture these new Geenius results.1,15 HIV-2 indeterminate results occur when only 1 of the 2 HIV-2 antigens (gp36 and gp140) is detected. In this instance, the manufacturer recommends repeating the Geenius test. HIV-2 indeterminate results arising from detection of gp140 only have often proved to not be reproducible.17 In this case, where Geenius is negative on repeat testing, the diagnostic algorithm would lead laboratories to move on to the HIV-1 nucleic acid test (NAT).9,15 However, even if the gp140-only HIV-2 indeterminate result occurs repeatedly, CDC and APHL now recommend that the next step in the algorithm should be an HIV-1 NAT.15 This is recommended because gp140-only HIV-2 indeterminate results have been observed in persons with early HIV-1 infection and because true HIV-2 infection is less common than early HIV-1 infection in the United States. If the HIV-1 NAT result is negative, CDC recommends that laboratories either reflex to testing with a validated HIV-2 test or repeat testing starting at the beginning of the algorithm on a new sample collected in 2 to 4 weeks. An additional complication with the Geenius test is how test results are reported. The final assay result (or assay interpretation18), an overall result, and individual results for HIV-1 and HIV-2 are included in the Geenius result report. For example, the assay interpretation could be “HIV-1 positive” but with individual results of “HIV-1 positive/HIV-2 indeterminate.” The CDC has received reports from both clinical laboratories and individual clinicians that seeing both “HIV-1 positive” and” HIV-2 indeterminate” results on the same report can be confusing. To address these issues, CDC and APHL produced a document describing suggested reporting language that reflects all possibilities for outcomes with the Geenius test, which is available from both the APHL18 and CDC HIV laboratory testing Web sites.19 The CDC recently produced a second technical update for the laboratory algorithm16 that provides information on the use of the Alere Determine HIV 1/2 Ag/Ab Combo single-use rapid test (Determine) in laboratories where it is not feasible to conduct an instrumented antigen/antibody test as the initial test in the algorithm. This update was prompted by data on the performance of Determine that became available after the 2014 Laboratory Guidance was published. In 2014, sufficient data were only available to recommend instrumented antigen/antibody tests. However, as alluded to by Hurt et al., the data from 3 CDC studies of Determine performance10,20,21 suggest that performing the Determine test with serum or plasma may be a useful option, particularly for smaller laboratories that perform a low volume of HIV tests where it may not be feasible to add a large instrumented platform to perform antigen/antibody HIV screening tests. Hurt et al. have provided a valuable update on HIV testing options through the summer of 2017. However, it should be noted that the field of HIV testing is still changing rapidly. Recently published data document that patients who become infected while taking suboptimal doses of antiretrovirals as preexposure prophylaxis22 and those treated with antiretroviral therapy before reaching peak viremia during acute infection23 may experience delayed seroconversion. In addition, both an updated version of the Alere antigen/antibody rapid test24 and several new platforms designed for detecting25–27 or quantifying28–30 HIV nucleic acid at the point of care, including using unprocessed whole blood specimens,25–30 are now widely used internationally. Manufacturers have recently brought similar technology to the United States for detection of viral pathogens other than HIV.31,32 However, the barriers, whether real or perceived, that delayed approval of antigen/antibody tests for use in the United States33 persist for these newer technologies. The CDC continues to work with our colleagues at the Food and Drug Administration to improve the efficiency of the approval process for new tests, and we are optimistic that these new options for HIV testing will arrive in the United States in an expeditious manner. The article by Hurt et al. is a useful summary of the current toolbox for HIV testing and highlights the need for ongoing evaluations and updates to advance the field of HIV testing and synopsize best practices. As the options for HIV testing continue to evolve, CDC will continue to gather data and provide additional updates to our testing recommendations and disseminate them through our Web site and other resources.9,11–13,15,16,19,34,35