Potential Influence of More-Sensitive HIV-1 Load Detection by the new Roche Cobas AmpliPrep/Cobas TaqMan Version 2.0 Assay on Clinical Management of HIV-Positive Pregnant Women

Recently developed real-time reverse transcription-PCR (RT-PCR)-based quantitative HIV-1 assays are more sensitive than previous assays and are being introduced into clinical practice worldwide (6-8). As recently reported in this journal by Pas et al., viral loads of more than 50 copies per milliliter (cop/ml) were detected with the new Cobas AmpliPrep/Cobas TaqMan HIV-1 version 2.0 assay (CAP/CTM v2.0) in samples from a cross-sectional retrospective data set in which the viral load was reported to be undetectable when tested by the previously used Cobas Ampliprep/Cobas Amplicor HIV-1 Monitor test version 1.5 (CAP/CA v1.5) (both tests from Roche Molecular Systems, Branchburg, NJ) (5). Moreover, in that study, the magnitude of these viral loads ranged to an upper limit of 2,600 cop/ml, and detection was associated with the time the patient had been on combination antiretroviral therapy (cART) but also with treatment failure. However, apart from treatment monitoring, the implications of a detectable viral load are particularly noteworthy in the clinical management of pregnant women with HIV who are on cART. International guidelines state that an undetectable viral load needs to be achieved in HIV-1-infected pregnant women to minimize the odds of mother-to-child-transmission (MTCT) of HIV during vaginal delivery (4). These recommendations are based mostly on studies performed in times when less-sensitive tests for HIV-1 viral load detection with a higher limit of detection were used (1). As a consequence of using new highly sensitive HIV viral load assays, physicians may more frequently choose to intensify cART or perform caesarean sections. To assess the possible clinical repercussions, the frequency and extent of discordance between viral load detection by the more-sensitive CAP/CTM v2.0 assay and that by the previously used CAP/CA v1.5 method were retrospectively studied with a cohort of pregnant women.

All HIV-1-positive women who gave birth between 1997 and 2010 in the Leiden University Medical Center, a tertiary referral and teaching hospital in the Netherlands, were identified. Inclusion of multiple pregnancy episodes for the same woman was allowed. Plasma samples that had been drawn prior to each woman having given birth and had been stored at −80°C were collected. Directly after the samples were thawed, the HIV-1 viral load was measured in a blinded manner by using both the CAP/CA v1.5 and the CAP/CTM v2.0 assays, with detection thresholds at 50 and 20 cop/ml, respectively. Results were measured as detectable or undetectable as well as in absolute numbers (cop/ml) and compared between tests. In addition, results were linked with the outcome for the child (HIV-1 positive/negative) and potential treatment decisions. Forty-two pregnancy episodes from 33 women were identified for possible inclusion in this study. All women were receiving cART at the time of delivery, and their median age was 32 years (interquartile range [IQR], 24 to 37 years). Samples for assessment of the HIV-1 RNA load were available for 30 of the 42 (71%) pregnancy episodes (from 2003 onward). Based on maternal indication, cART was prescribed during the complete pregnancy in 13 episodes (43%). In 17 episodes (57%), the prevention of MTCT was the indication for initiation of cART. In the latter group, the median time between initiation of cART and delivery was 116 days (IQR, 85 to 137 days). Twenty-five births occurred through vaginal delivery. Caesarean section was performed at the end of 5 pregnancies (17%); this was indicated due to obstetrical complications. According to U.S. Department of Health and Human Services (DHHS) guidelines, all infants received 4 weeks of postexposure prophylaxis (PEP). The median time between sample date and delivery was 16 days (IQR, 7 to 26 days). In 2 of the 30 predelivery samples (7%), HIV-1 RNA was detected above the threshold by the CAP/CA v1.5 detection method, compared to 12 (40%) when the samples were tested by the new CAP/CTM v2.0 assay. Of the 10 samples in which the latter method discordantly yielded positive results, six were attributable to the lower detection threshold, but four had viral loads in the range of 50 to 150 cop/ml (Fig. 1). In the period of study, no MTCT of HIV occurred.

We found that an HIV-1 viral load was more frequently detected by the new CAP/CTM v2.0 assay, with the lower limit of detection accounting for approximately half of the discordant samples. No difference with regard to risk of transmission was observed in this setting. Major limitations of this study include the relatively small sample size and incomplete information about the frequency of non-B subtypes. Despite these drawbacks, the data demonstrate that attending physicians will be challenged more frequently with detectable viral loads in their patients’ last month of pregnancy. These test results have to be considered when making treatment decisions, e.g., with respect to the pursued mode of delivery. Since cART is initiated only 3 to 4 months prior to delivery in the majority of women, these detectable loads probably represent viral replication and cannot simply be ascribed to the release of virus RNA from the decay of latently infected cells (3). The results suggest that HIV-1 viral load in this specific setting will mainly be found to be under the level of 500 cop/ml. Additional actions undertaken for patients with results in this range, e.g., intensifying the therapeutic regimen with another antiretroviral agent or performing a caesarean section, will probably not result in a further decrease in MTCT of HIV-1 in the presence of infant PEP (2). Nonetheless, an HIV-1 viral load in the range of 20 to 500 cop/ml detected by newer methods still warrants confirmation of compliance to cART and intensified follow-up for the timely recognition and prevention of situations in which a higher risk of MTCT of HIV during delivery exists. Rarely, but more frequently than when the CAP/CA v1.5 method (5) is used, a detectable load of more than 500 cop/ml may be detected and would legitimately require intervention to prevent MTCT. Although in general the new methods provide increasing insight into the dynamics of HIV-1 in the lower ranges of detectable HIV-1 viral loads, uncertainty persists regarding their weight and mode of interpretation in clinical decision making during the peripartum period in HIV-1-infected pregnant women. Thus, further prospective evaluation of the performance of the new generation of HIV-1-RNA viral load assays and definitions of their role in updated clinical guidelines, with emphasis on patient groups in which the results may have major consequences, are warranted.

• Open in new tab
• Download powerpoint

FIG. 1.

Concordance between one of the newest methods that measures HIV-1 viral load (CAP/CTM v2.0) and the previously used method (CAP/CA v1.5) for plasma samples drawn from pregnant women prior to delivery (n = 30 pregnancy episodes).

• Copyright © 2010 American Society for Microbiology