Before prescribing PrEP, the absence of acute or chronic HIV infection should be confirmed [3]. During the earliest weeks of HIV infection (ie, acute phase of infection), antibody to HIV may not yet be measurable but viral antigens can be. While the patient is taking PrEP, HIV testing should be repeated at least every 3 months before PrEP prescriptions are refilled or reissued. Testing is also recommended when there has been low medication adherence and when signs or symptoms consistent with acute HIV infection occur [14], recognizing that during the earliest stages of infection there is increased probability for false-negative testing. This frequent testing schedule is intended to ensure that HIV infection is diagnosed and treated early to limit immune system injury, minimize the risk of inducing HIV drug resistance from exposure to incompletely suppressive PrEP regimens, and prevent onward HIV transmissions.

For HIV testing, the Centers for Disease Control and Prevention (CDC) and the Association of Public Health Laboratories (APHL) recommend that infection be diagnosed using an instrumented, laboratory-based Ag/Ab screening immunoassay, followed, when reactive, by an HIV-1/HIV-2 Ab differentiation immunoassay [15]. When the differentiation assay interpretation is negative or indeterminate for HIV-1, an HIV-1 nucleic acid test (NAT) for HIV RNA should be performed. Instrumented Ag/Ab tests are preferred over rapid POC tests (which can be either Ab only or Ag/Ab tests) because the former are more sensitive to HIV during acute infection. All positive POC test results should be confirmed by drawing blood for testing in the laboratory [15].

Characteristics of currently available Food and Drug Administration–approved HIV tests are available on the CDC website [16]. These tests can be grouped into 3 major categories: NATs, Ag/Ab tests, and Ab-only tests [17]. HIV tests are very accurate, but no test can detect the virus during the first few days after infection, when replication is limited to tissues near the site of infection. Once HIV enters the circulatory system, infection is detectable in blood. The earliest that HIV infection can be detected is with a NAT, typically between 10–33 days after an exposure. However, NATs are expensive compared with other tests and are usually used when either recent exposure or acute HIV infection is suspected. Ag/Ab blood tests performed by a laboratory can usually detect HIV infection 18–45 days after an exposure. Ab-only tests can also typically detect HIV within the first weeks after infection but in some cases may take up to 90 days before detectable levels of antibody are present. Most POC HIV tests are rapid Ab-only tests that use either whole blood or oral fluid; using blood can detect infection earlier than oral fluid [18], so oral fluid testing is not recommended for persons taking PrEP. HIV infection is considered established when both HIV RNA and antibody are present.

The performance characteristics of the HIV tests described above were derived using specimens from individuals not exposed to ARVs during infection [19]. The presence of ARVs in PrEP patients during infection may alter these trajectories [20], suppressing or slowing viral replication and the development of antibodies, which may also occur out of synchrony, such as detectable antibody in the absence of detectable antigen. In persons who become HIV-infected while taking PrEP, HIV testing may produce results that could be interpreted as ambiguous or falsely negative (ie, erroneously interpreted as uninfected) [21], thus delaying accurate diagnosis. Technical issues related to false-positive testing (ie, erroneously interpreted as infected), specifically repeated false-positive Ag/Ab test results in an HIV-uninfected person taking PrEP, can also pose a challenge, taking time to reconcile and creating psychological stress.

At least 4 breakthrough PrEP infections with ambiguous HIV test results that could not be resolved with the HIV testing algorithms used have been reported among persons using oral daily FTC-TDF or TDF (Table 1) [4, 5, 9, 22]. In all 4 cases, the screening test was an Ag/Ab assay, which was reactive. Supplemental testing, when done, was not always consistently reactive and included Western blot (WB), HIV-1/2 differentiation assays, and/or HIV RNA tests. The presence of infection was confirmed in 1 of 2 ways: either by (1) continuing PrEP in 3 cases and adding a protease and/or integrase inhibitor for treatment of presumptive HIV infection while performing additional testing, including HIV RNA and/or DNA tests, or (2) stopping PrEP in 1 case and subsequently documenting the presence of both a stable antibody response and presence of HIV RNA. Resolving ambiguous test results created delays in the diagnosis of HIV infection and may have contributed to an opportunity for ARV resistance to develop, such as the K65R mutation, which can confer resistance to tenofovir and cross-resistance to FTC; this mutation was observed in 2 persons. One patient also had the M184V mutation, which confers resistance to FTC. Two persons also did not have high concentrations of HIV RNA, possibly as a result of suppression of viral replication by PrEP or combined with the addition of an integrase inhibitor to the treatment regimen of 1 patient while testing continued [4]. In the second patient, PrEP was stopped, and the infection was allowed to progress, producing diagnostic levels of HIV RNA but obviating the potential benefit of blunting early viral dissemination.

Two PrEP studies, the HPTN 067/ADAPT and iPreX studies, that retrospectively studied stored blood specimens also demonstrated the increased sensitivity of plasma NAT for detecting breakthrough HIV-1 infection in persons for whom parallel rapid tests are falsely nonreactive or discordant [12, 21]. Interestingly, Ag/Ab combo tests were also nonreactive for 7 persons in the HPTN 067/ADAPT study at the first HIV testing visit, and development of drug resistance mutations (M184I, K65R) was seen in 3 seroconvertors.

Repeated false-positive Ag/Ab tests are rare but are problematic when they occur in the context of PrEP. This phenomenon was observed in 1 participant in Project DETECT, an ongoing study funded by the CDC and being conducted at the Public Health Seattle King County STD clinic (J Steckler MD, et al. accepted, OFID, August 2018). Similar cases have also been reported to the CDC by providers who administer PrEP outside of this study. Documenting patterns of HIV test results in these patients will be informative. Promptly establishing the absence of infection in these cases is critical, not only to relieve psychological stress but to be able to continue the protection that PrEP use confers.

Two issues arise from these situations with unusual HIV testing results. First, how to confirm the presence or absence of infection. Second how to manage antiretroviral medications while infection status is being confirmed, specifically whether to continue PrEP, transition to an ART regimen, or discontinue PrEP. There are pros and cons to each of these antiretroviral management strategies.

Persons prescribed PrEP merit a high index of suspicion for possible HIV infection. Ambiguous HIV test results appear to be rare using the recommended CDC/APHL laboratory HIV testing algorithm. Rapid and accurate resolution of ambiguous test results can minimize harm. Persons with ambiguous test results require counseling about options to manage antiretroviral medications while ambiguous results are resolved. By ensuring that such persons understand the potential benefits and harms of each option to their own health and to their risk of transmitting HIV to others, they can, together with their health care provider, make informed choices most appropriate to their individual circumstances.

More experience is needed in the management of PrEP, as is a broader understanding of how HIV tests perform in the context of the rare breakthrough HIV infection while taking PrEP. In addition, more data are needed to provide a clearer understanding of how to quickly resolve false positive test results for PrEP patients eventually determined not to have acquired HIV infection. In the meanwhile, providers in the United States with questions about management of specific testing results for PrEP patients, or who wish to report ambiguous HIV test results in PrEP patients, can call the PrEPline toll-free at 855-448-7737 (11 am–6 pm EST). Providers may also want to consult with local infectious disease specialists or colleagues with more expertise in HIV diagnosis options.