Inconsistent detection of an evolving HIV-1 infection by a popular high-throughput screening assay.

A 27 year old female patient presented with a 2-week history of sore throat, tender cervical adenopathy, fever and vomiting, following unprotected sex with a new partner. She had also recently developed painful mouth ulcers and a rash. Investigations revealed generalised lymphadenopathy, splenomegaly, an atypical lymphocytosis, transient hepatitis (alanine transferase 86 IU/ml) and thrombocytopaenia (119×10/L). Her clinical presentation was consistent with a differential diagnosis of glandular fever, as well as possible HIV seroconversion illness, and an acute viral hepatitis. Blood samples were taken for HIV, EBV, CMV, and viral hepatitis screening. This patient had been a regular attender over the past few months for high risk bloodborne virus screening, so previous samples were still available for comparative testing. Serological screening on current and previous samples for viral hepatitis (HAV, HBV, HCV, HEV) were all negative for recent, acute infection. However, serological screening for CMV and EBV revealed a positive CMV IgM and CMV IgG, and testing on archived samples indicated that a primary CMV infection had occurred. Current CMV infection was confirmed with a CMV DNA level of> 36,000 IU/ml (on whole blood). Similar testing on current and archived sera showed current EBV VCA IgM reactivity, with detectable EBV VCA IgG, but an undetectable EBV DNA. This suggested that the EBV VCA IgM positivity was more likely due to cross-reactivity with the CMV IgM antibodies. For HIV, there was some initial low level reactivity result on our frontline screening enzyme immunoassay (EIA, Siemens ADVIA Centaur® HIV Ag/Ab Combo, Siemens Healthcare Ltd., Camberley, UK), prompting further testing on our supplementary Ag/Ab combo EIA (Siemens Enzygnost ® HIV Integral 4) and immunoblot (GeeniusTM HIV 1/2 Supplemental, Bio-Rad Laboratories Ltd., Watford, UK) assays (Table 1). Followup HIV-1 RNA testing showed a viral load of> 100,000 copies/ml. However, the results of the immunoblot and HIV-1 RNA testing were not available for several days, which caused something of a diagnostic dilemma at the time. Baseline genotyping using Sanger sequencing was performed at the national reference laboratory (Public Health England, Birmingham) showed a fully susceptible recombinant A1, G HIV-1 subtype. This patient was started on dolutegravir (50mg, od) and Kivexa (1 tablet, od) about 4 weeks after (on Day 137) of her first confirmed positive viral load result (on Day 109) (Table 1). Surprisingly, routine repeat samples received for confirmatory testing one week later, showed that the reactivity of this virus on the Centaur HIV combo screening assay remained borderline, fluctuating between low positive and negative. This remained unchanged on a further 8 serum samples taken over 6 months on this Centaur assay, despite the other (Integral and Geenius) assays showing the expected increase in reactivity during this time. Primary HIV infection is often asymptomatic and the final diagnosis relies on laboratory-based testing. This now includes fourth generation combined Ag/Ab screening assays to narrow the seroconversion window [1], with an additional screening test if this is positive, before serological confirmation and typing with immunoblot, Western blot and/or nucleic acid testing for HIV RNA [2–4]. Ideally, any positive result on a single sample should also be confirmed on a second sample. Since the availability of fourth generation combined Ag/Ab screening assays, a ‘second diagnostic window’ has been described, which can lead to diagnostic dilemmas [5,6]. However, additional samples taken from the patient usually leads to a definitive result, despite the ever-present risks of sample mix-ups or cross-contamination [7,8]. This initial unclear diagnosis of HIV infection on the Centaur screening assay caused many problems for the laboratory and clinical team in the first few days, including checking for any sample mix-ups, and explaining the need for several follow-up samples to the patient for further testing. The underlying cause of the persistent lack of reactivity on the Centaur HIV combo assay is still unknown. The response from Siemens was to quote from their kit insert’s ‘Limitations’ section: “Currently available assays for the detection of p24 antigen and/or antibodies to HIV-1 and/or HIV-2 may not detect all infected individuals. A negative test result does not exclude the possibility of exposure to or infection with HIV. HIV antibodies and/or p24 antigen may be undetectable in some stages of the infection and in some clinical conditions”. This response emphasises the necessity of using an alternative EIAs to confirm HIV infection, which forms part of the current guidance.