A simple algorithm for selecting cases to investigate acute and early HIV infections in low‐ and middle‐income countries

We documented the outcome of an over 10‐year (2011–2021) effort to diagnose acute and early HIV infections (AEHI) in an Infectious Diseases Outpatient Clinic with limited resources. Of a total of 132, 119 HIV‐RNA tests were performed from 2017 to 2020, 12 cases were identified, using a simple algorithm: risk exposure of 6 weeks or less before the visit and/or symptoms compatible with acute retroviral syndrome 7–30 days after exposure and/or undetermined 3rd generation rapid diagnostic test or serology. AEHI diagnoses varied from 2.4% among asymptomatic to 25% for undetermined serology cases using this simple screening applicable to different settings.

antibodies are undetectable (acute infection) or inconclusive (early infection), 11 representing cases that may be missed by current serology tools. The detection of viremia after the first week, and of the p24 antigen after the second week, may improve diagnosis in this phase of greater transmissibility. 12 Ideally, all people at risk of HIV acquisition should be tested for AEHI with a molecular test, as viral load, 13 but this is not feasible in resource-limited settings due to logistical issues and the high costs of molecular testing. 11 Tests to diagnose AEHI are not routine in Brazil and elsewhere, and the use of a 3rd generation rapid antibody test (RDT3) in the routine of Brazilian public services makes diagnosis even more difficult.
Our study was conducted in an Infectious Diseases Outpatient Clinic  With increasing access to knowledge about infection and recommendations for immediate treatment, the closer health systems are to these goals, the more acute/early undiagnosed individuals will proportionally contribute to a larger share of newer infections. Several studies assessed the ability of algorithms to screen which subpopulations should be subjected to molecular testing for AEHI, especially among men who have sex with men. 14,15 In a systematic review and meta-analysis, the overall pooled AEHI yield was 6.3% (95% confidence interval [CI]: 2.1-12.4; five studies); but varying among the different targeting strategies used, from 11.1% (95% CI: 5.9-17.6; three studies) to 1.6% (95% CI: 0.8-2.4; two studies) in universal testing strategies. 11 In the first years of our study, the use of HIV-RNA tests for diagnostic purposes was rare and not systematically recorded. From 2017 to 2020, in which the search for suspected cases of AEHI was incorporated more consistently, the yield was 6.2% among those screened by the algorithm. To account for this, we calculated the rate considering different denominators. Using different scenarios, it appears that the algorithm proposed was useful for case selection, as all but one, were diagnosed by HIV-RNA tests in the AEHI had suspected symptoms of the acute retroviral syndrome in the last 30 days of collection. This one exception (case 12) was screened due to a discordant RDT3. This young woman was the only AEHI identified among 41 asymptomatic individuals, an elite controller (undetected viral load, that could have been dispensed as not infected with HIV RDT4 had not been performed). After 1 month her plasma was reactive to 4th generation serology (Chemiluminescence, architect ® ; Abbott) and Western blot indeterminate. Important to note that a way to improve the sensitivity of these RDT is the potential benefit of performing the test in plasma. Although blood collection and processing may not be feasible in some places, testing plasma may improve detection. One example of this was a negative RDT3 and RDT4 at fingerstick, that was, however, RDT4 positive in plasma sample collected for HIV-RNA, that showed a viral load of 3 430 777 copies/ml (case 10). This case illustrates what has been suggested in antibody detection to various agents, 16 that plasma or serum from processed blood may be more sensitive than whole-blood from fingerstick testing. If we take this into account, RDT4 diagnosed all four cases of AEHI that RDT4 was applied, including the asymptomatic elite controller case.
Another interesting finding was that all the reactive cases in RDT4 showed the line of reactive antibodies and only one with antigen and antibodies line (case 8). Therefore, it was expected that they would be diagnosed with RDT3 if the differential of RDT4 was only the addition of antigen line, implying that the sensitivity for detecting RDT4 antibodies is higher than that of RDT3. Despite the limited number of cases, our study suggests that the plasma RDT4 technology could be a reasonable alternative to HIV-RNA test, at a much lower cost, point of care (POC), with easy execution and immediate diagnosis, allowing linking to a specific treatment on the same day, as an important strategy to reduce the high viremias of highly infectious variants of acute/early infection. The better sensitivity of RDT4 could be an even more important alternative to RDT3, in candidates for HIV pre-exposure prophylaxis, as many, by definition, could be at an RDT3 immunological window from a recent risk event. Improving AEHI identification with simple screening based on symptom/timing of risk assessment and the use of at least an RDT4 test seems a reasonable task for HIV testing sites, but still many cases would not be reached. The increasing use of POC viral load tests may further favor the ability of HIV testing centers to improve AEIH diagnosis, but the cost is still a limiting factor.
Other environments that may receive individuals unaware of being at early HIV infection, such as emergency units, or even settings outside regular health care system, should be stimulated to use strategies to reach out for AEHI patients, either referring to evaluation or conducting some level of laboratory testing. Return for reevaluation in some days of suspect cases might be warranted.
In conclusion, our 10-year experience suggests the feasibility of strategies to target potential acute/early HIV infections. With the growing identification and incorporation of chronic infections with improvement in policies to treat all persons living with HIV, the acute/early cases will increasingly represent a larger pool that, if properly identified, treated, and suppressed, may favor the control of AIDS pandemic, not restricted to health care units.

ACKNOWLEDGEMENTS
The study was supported in part by grants Conselho Nacional de De-  Abbreviations: AEHI, acute and early HIV infections; ARS, acute retroviral syndrome.