Objectives: We assessed the performance of the DC RT against third generation RT in antenatal and post-partum women.

Methods: Third generation RT Advance Quality and Acon were used in a serial algorithm to diagnose HIV infection in antenatal and post-partum women over six months at a tertiary hospital in Johannesburg, South Africa. This data provided the reference against which the DC RT was compared on plasma and whole bloodsamples.

Results: The 1019 participants comprised 345 (34%) antenatal and 674 (66%) post-partum women. Ninety women (8.8%) tested HIV-positive of whom 59 (66%)were tested antenatally, and 31 (34%) post-partum yielding prevalence rates of 17.1% and 4.6% respectively. The sensitivity and specificity of the Ab component of DC on plasma antenatally was 100% (93.8% – 100%) and 100% (98.6% – 100%) respectively and post-partum was 100% (88.9% – 100%) and 99.6% (98.8% – 99.9%) respectively. One false positive and not a single true positive p24Ag was detected. Of 505 post-partum women who tested HIV-negative 6–12 months prior toenrolment, 12 (2.4%) seroconverted.

Conclusion: The fourth generation DC offered no advantage over current third generation RT in the diagnosis of HIV infection.

In South Africa, pregnant women are offered counselling and testing for HIV at their first antenatal clinic visit and at 34 weeks of pregnancy if their initial HIV test was negative.² Women at the Rahima Moosa Mother and Child Hospital (RMMCH) in Johannesburg, South Africa are also offered an HIV test immediately post-partum if their HIV status is unknown or more than six weeks have elapsed since their last negative HIV test. The importance of HIV retesting has been demonstrated by a South African study in which 3.4% of women who tested HIV-negative at their first antenatal visit, subsequently seroconverted during pregnancy or within a year after delivery.³ The Advance Quality HIV Rapid Test (In Tec Products, Inc. Xianen, China) and Acon HIV 1/2/0 Tri-line Rapid Test (Acon Laboratories, Inc., San Diego, USA) are currently used to diagnose HIV infection in women in Prevention of Mother-to-Child transmission (PMTCT) programmes in Gauteng province, South Africa. These third generation RT detect HIV Ab that are produced in response to the virus by a serial testing algorithm as recommended by the South African PMTCT guidelines² (Figure 1). Advance Quality is used to screen for HIV Ab and if positive, the Acon test is performed to confirm HIV status. Laboratory-based HIV ELISA and, less commonly, viral detection assays can be used as a tiebreaker to confirm an HIV status if serial RT results are discordant.²

The Determine Combo HIV-1/2 Ag/Ab Combo Test (DC) (Inverness Medical, Japan Co., Ltd) is the first fourth generation RT and can be performed on either plasma or whole blood samples.⁴ It is an enhancement of the third generation tests as it can detect both HIV Ab and p24Ag in a single test.⁵ The p24Ag is a marker of early HIV infection and is detectable in blood during the window period before HIV Ab become detectable.⁶ The DC RT is reported to have the potential to identify HIV infection five days (range 2–20 days) earlier than third generation RT. The reported sensitivity and specificity of the DC Ab component is 100% (95% confidence interval 98% – 100%) and 100% (95% confidence interval 98.2% – 100%) respectively and the sensitivity of the p24Ag component is 86.6% (95% confidence interval 76% – 93.7%).⁷

Detection of early HIV infection during the window period of third generation RT would allow more maternal HIV infections to be detected. Furthermore, during early maternal HIV infection the levels of the virus in the blood are at their peak and the risk of transmission to the infant during birth and breast-feeding is very high.⁸ Detection of early infection would allow more women and their infants at a high risk of vertical transmission to access PMTCT.

The performance of the fourth generation HIV DC RT in diagnosing HIV infection status antenatally and in the early post-partum period in comparison to the third generation HIV RT in routine use at public healthcare facilities in Gauteng, South Africa was assessed. The advantage offered by the DC RT over third generation RT of detecting HIV infection earlier to increase identification of women at a high risk of vertical transmission for PMTCT was investigated.

FIGURE 1: Study methodology: The Determine Combo rapid test (DC RT) was performed at the same time as the routine algorithm.

The women received their RT results and post-test counselling within four hours of blood sampling. Patients that were HIV-positive on both third generation RT were referred to antiretroviral treatment clinics. The results of the DC RT were not disclosed to the patient. However, where the third generation and DC HIV Ab RT results were discordant, samples were referred for confirmatory fourth generation ELISA (ARCHITECHT® HIV Ag/Ab Combo assay; Abbott Diagnostics; Wiesbaden, Germany). Patients that had detectable p24 Ag on DC were followed up with three confirmatory tests that is viral detection assay (Vironostika HIV-1 Antigen; bioMerieux; Bosiend, The Netherlands) , viral load testing (NucliSENS EasyQ-EasyMag HIV-1, version 1.2 assay; bioMerieux; Boxtel, The Netherlands) as well as fourth generation ELISA. Disclosure of the patient’s HIV status was delayed for 48 hours.

Knowledge of the women’s HIV status prior to undergoing HIV testing on the study was documented for those tested in the early post-partum period only (Table 1). Of the 505 women that had reported or tested HIV-negative between 6 and 12 weeks prior to study enrolment, 12 (2.4%) tested positive demonstrating that new HIV infections were occurring in this population. Not all of these women had their negative HIV status documented on their maternal record and the possibility that some reported their status incorrectly cannot be excluded.

The DC RT was performed on plasma samples of the 1019 women and on whole blood samples on a subset of 380 women. All 1399 tests performed demonstrated positive control strips therefore no DC RT was invalid. Sensitivity, specificity and likelihood ratios for the DC RT were calculated separately for women who were tested antenatally and post-partum (Table 2). The sensitivity of the DC RT Ab component was 100% in all groups of women tested irrespective of the sample type. The specificity of the DC RT Ab component in women tested post-partum was less than those tested antenatally.

The DC RT results concurred with the third generation RT results in all patients except for three post-partum women on whom both plasma and whole blood samples were tested. In two patients the Ab component of the DC RT was positive on both their plasma and whole blood samples but negative on both third generation RT. Laboratory based fourth generation ELISA tests on both cases were negative confirming two false positive DC HIV Ab results. The DC p24Ag was reactive on plasma and the whole blood of only one patient of all 1019 women tested. This patient was suspected of having an early infection since both third generation RT and the DC Ab component tested negative. However the laboratory based fourth generation HIV ELISA was negative and the viral load was undetectable indicating a false positive DC p24Ag result. No p24Ag bands on DC were obtained on any of the 90 HIV-infected women nor was a single case of early HIV infection detected.

Sensitivity and specificity of the DC RT was comparable to plasma and whole blood in contrast to a previous report that demonstrated lower sensitivity of the DC RT in whole blood compared to serum samples.¹² The sensitivity of the fourth generation DC RT in detecting HIV Ab was comparable to that of the reference third generation RT in antenatal and post-partum women in plasma and whole blood samples. The specificity of the DC RT in detecting HIV Ab was slightly reduced in whole blood and plasma in post-partum women owing to the false positive results in two women, but was still within the World Health Organization recommended range of more than 98%.¹³

The fourth generation DC RT did not detect a single true positive p24Ag, even in the 90 HIV-infected women. The sensitivity in detecting p24Ag in HIV-infected women was 0% as compared with the claim of 86.6% obtained on HIV-infected samples.⁷ The reason for this may be that p24Ag forms immune complexes with HIV Ab and thus no free p24Ag is present for detection by the DC.⁶ The DC RT therefore did not identify any new cases of maternal HIV infection over the 6 month study period. Possible reasons for this include that the sensitivity of the DC p24 Ag component is poor or that no women with acute HIV infection were enrolled. The former concurs with previous reports that the DC RT p24 Ag component lacks sensitivity particularly where levels of p24 Ag are below 400 pg/mL.^12,14 Additionally, the p24Ag component of the DC RT has reduced sensitivity in comparison to other fourth generation viral detection assays,⁷ some of which are able to detect p24Ag at levels of 4 pg/mL – 5 pg/mL.¹⁵ The initial laboratory based p24Ag assays also demonstrated poor sensitivity that was subsequently improved by denaturation of the immune complex and signal amplification to enhance p24Ag detection.¹⁶

A limitation of this study is that the incidence of HIV infection in women undergoing HIV testing at RMMCH is unknown therefore, although new maternal infections were demonstrated, it is possible that no women with acute HIV infection were enrolled during the short window period before HIV Ab and subsequent immune complex formation. Nevertheless, the DC RT did not identify any new infections over those identified by the 3rd generation RT assays over six months in the RMMCH PMTCT programme.