Share this article on:

Changes in plasma HIV-1-RNA viral load and CD4 cell counts, and lack of zidovudine resistance among pregnant women receiving short-course zidovudine

Ekpini, René-Anatolea; Nkengasong, John N.a,b; Sibailly, Toussainta; Maurice, Chantala; Adjé, Christianea; Monga, Ben B.a; Roels, Thierry H.a,b; Greenberg, Alan E.b; Wiktor, Stefan Z.a,c


Objective To describe changes in HIV-1 plasma viral load (VL) and CD4 cell counts and to assess zidovudine resistance associated with a short course of oral zidovudine during late pregnancy.

Methods From April 1996 to February 1998 in Abidjan, Côte d'Ivoire, 280 HIV-1-seropositive women were randomly assigned at 36 weeks’ gestation to receive zidovudine (300 mg) or placebo twice a day, and then one tablet every 3 h from the onset of labor until delivery. Blood samples obtained every 2 weeks until delivery, then at 2 and 4 weeks, and 3 or 6 months after delivery were tested from selected women based on duration of therapy for plasma VL and CD4 cell counts, and samples from 20 women in the zidovudine group were tested by DNA sequencing for the presence of zidovudine resistance mutations.

Results In the zidovudine group, the median reduction in plasma VL (log10 copies/ml) was −0.48 after 2 weeks (P = 0.02 versus placebo), −0.48 after 4 weeks (P = 0.06), −0.80 after 6 weeks (P = 0.29) of treatment, −0.12 at delivery (P = 0.11), +0.21 at 2 weeks (P = 0.83), +0.17 at 4 weeks (P = 0.69), and +0.21 at 3 months (P = 0.56) postpartum. Median CD4 cell counts were higher in the zidovudine than in the placebo group after 2, 4, and 6 weeks of treatment (P < 0.05). No mutations associated with zidovudine resistance were identified in any of the samples tested.

Conclusion These findings suggest that a short course of zidovudine has no adverse HIV-1 virological consequences for the mother.

From the aProjet RETRO-CI, Abidjan, Côte d'Ivoire; bDivision of HIV/AIDS Prevention, Surveillance and Epidemiology, and cGlobal AIDS Program, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia, USA.

The use of trade names and commercial sources is for identification only, and does not imply endorsement by the Public Health Service or the US Department of Health and Human Services.

Correspondence and requests for reprints to: Dr John Nkengasong, Projet RETRO-CI, 01 BP1712, 01 Abidjan, Côte d'Ivoire.

Received: 13 July 2001;

revised: 5 October 2001; accepted: 16 October 2001.

Back to Top | Article Outline


Clinical trials in Thailand and West Africa have documented that a short course of oral zidovudine administered in late pregnancy reduces the risk of mother-to-child transmission of HIV-1 by 37–50% at 6 months of age or less [1–3]. Other studies have demonstrated the efficacy of intrapartum nevirapine and of zidovudine combined with lamivudine [4,5]. The exact mechanism by which zidovudine prevents transmission is unknown; however, a reduction in the maternal HIV-1 plasma RNA viral load is thought to play an important role [6–9]. The dynamics of viral load reduction for women receiving these therapies are not known; in particular, it is not known how rapidly the viral load reaches a nadir after the initiation of therapy, whether this reduction is maintained throughout therapy, and whether there is an overshoot of viral load above baseline after medication is stopped. Such an overshoot has been documented among patients after stopping highly active antiretroviral therapy (HAART) [10–12]. If it occurs after short-course zidovudine and results in a higher viral load in breast milk, it could increase the risk of postnatal HIV-1 transmission in countries such as in Africa, where most HIV-1-infected women breast-feed.

Ideally, as in more developed countries, these interventions to prevent the mother-to-child transmission of HIV-1 would be administered in addition to antiretroviral therapy to treat the mother's HIV disease. Unfortunately, the high cost of this type of therapy precludes its long-term use, and in fact, short courses were designed to be as short as possible so that they could be used for the greatest number of women. Therefore, after delivery, women do not continue to take antiretroviral therapy. There is concern as to whether the prenatal administration of a short course of an antiretroviral drug may negatively affect the clinical course of the mother's HIV disease, as measured by a persistently elevated viral load or the development of resistance to zidovudine. Resistance to nevirapine has been documented among women treated with a single dose of intrapartum nevirapine [13]. This type of resistance could limit the therapeutic options for later pregnancies and for long-term antiretroviral therapy.

Back to Top | Article Outline


Study population

The study protocol was approved by the Institutional Review Board of the Centers for Disease Control and Prevention and by the Ethical Committee of the Côte d'Ivoire Ministry of Public Health. Details concerning enrolment and follow-up have been published [2]. From 14 April 1996 to 17 February 1998, all consenting, eligible HIV-1-seropositive pregnant women attending a public antenatal clinic in Abidjan were enrolled at 36 weeks’ gestation and were randomly assigned to receive zidovudine (300 mg tablets; n = 140) or placebo (n = 140). Women who were HIV-2 seropositive or HIV-1 and HIV-2 seropositive were excluded. Women were instructed to take one tablet of the study drug twice a day until the onset of labour, one tablet at the onset of labour, and one tablet every 3 h until delivery, at which time they stopped taking the study drug. Gestational age was estimated by the date of last menstrual period and physical examination. However, there was considerable variability in the duration of therapy as a result of the lack of precision of this estimate. Consequently, the duration of treatment varied greatly (median 29 days; range 1–60 days). After enrolment, women were seen bi-weekly until delivery, at 2, 4, and 12 weeks after delivery, and then every 3 months. A blood specimen was obtained at each visit.

As we were interested in the issue of post-therapeutic overshoot in viral load and the possible effect of the duration of therapy on this overshoot, we selected women for this analysis on the basis of the duration of therapy and the availability of samples in the early postpartum period. Therefore, of the 280 enrollees, we selected the 84 women for whom blood specimens were available from all of the completed prenatal visits, from the delivery, and from the postpartum visits at 2, 4, and 12 weeks. These women were classified into three groups (2, 4, and 6 weeks) on the basis of the duration of therapy, and we then randomly selected women from within each group: 12 (nine zidovudine, three placebo) of the 37 women with 2 weeks of therapy, 21 (18 zidovudine, three placebo) of the 30 women with 4 weeks of therapy, and 16 (seven zidovudine, nine placebo) of the 17 women with 6 weeks of therapy. All available blood specimens for the selected 49 women (34 in the zidovudine group and 15 in the placebo group) were analysed.

Back to Top | Article Outline

Laboratory testing

Maternal serum samples were tested for the presence of HIV-1 and HIV-2 antibodies by using standard testing algorithms that included enzyme-linked immunosorbent assays, a synthetic peptide line immunossay, and Western blot testing, as previously described [14–16]. Lymphocyte subtypes were enumerated by standard flow cytometry (Facscan, Becton Dickinson, San Jose, CA, USA) on fresh peripheral blood collected into ethylenediamine tetraacetic acid tubes. Commercially available monoclonal antibodies (Becton Dickinson) were used for cell staining.

The HIV-1-RNA viral load was quantified in masked plasma specimens by using the Amplicor HIV-1 Monitor assay (version 1.5; Roche Diagnostic Systems, Branchburg, NJ, USA). HIV-1-RNA extraction, amplification, and detection was performed according to the manufacturer's instructions.

For zidovudine resistance testing, we sequenced a 1.2 kb fragment of complementary DNA that included the HIV-1 protease gene and the first 300 codons of reverse transcriptase from RNA in plasma. The amplification products were assessed for the presence of resistance mutations at codons M41L, D67N, K70R, and T215Y/F. Briefly, purified viral RNA (Qiagen Viral RNA Extraction Kits, Qiagen, Valencia, CA, USA) was reverse-transcribed and amplified using polymerase chain reaction (PCR) (R. Chiengsong, Division of AIDS, STD, and TB Laboratory Research, National Center for Infectious Diseases, CDC, Atlanta, Georgia, USA, unpublished data, 1999). Approximately 200 ng of the 1.2 kb amplified product [covering the whole protease gene, reverse transcriptase codons 39–142 (beginning), and codons 135–244 (middle)] were sequenced by using dye-labelled dideoxyterminator cycle sequencing (Clip, Visible Genetics, Toronto, Ontario, Canada) [17]. The sequence for each sample was compared with sequences from a database of known drug-resistance mutations according to the consensus statement on antiretroviral drug resistance [17].

Back to Top | Article Outline

Statistical analysis

To obtain an approximately normal distribution, plasma HIV-1-RNA viral load results were log-transformed before analysis. All the samples for which the viral load results were below the detection limit of the assay [200 (2.3 log10) copies/ml] were assigned a value of 100 (2.0 log10) copies/ml. The RNA viral load result at each visit was subtracted from the baseline viral load value at enrolment to obtain the change in the number of RNA copies (log10/ml plasma). Median differences in viral load were calculated to minimize the influence of data points with undetectable viral loads. Similarly, median changes in CD4 T cell counts were calculated by subtracting the values obtained at each visit from the baseline value. Median differences within and between groups in viral load and CD4 cell count levels were compared by using the Wilcoxon signed rank test and the Mann–Whitney U test, respectively. A 95% confidence interval for the prevalence of zidovudine resistance was calculated assuming a Poisson distribution. For all the statistical tests, the level of significance was set at a two-tailed α value of 0.05.

Back to Top | Article Outline


Baseline characteristics

The 34 women in the zidovudine group were similar to the 15 women in the placebo group with respect to age [26.0 years (range 19–37) versus 24.0 years (range 17–31);P = 0.20], median CD4 T cell count at enrolment [515 cells/μl (range 207–885) versus 488 cells/μl (range 198–985);P = 0.66], and median HIV-1 viral load at enrolment [4.11 log10 RNA copies/ml (range 2.63–4.96) versus 4.30 log10 RNA copies/ml (range 2.0 {non-detectable, n = 2}–6.08);P = 0.96]. Among women in the zidovudine group, median CD4 cell counts and HIV-1 viral loads were similar for women in the 2 week group (551 cells/μl; 4.6 log10 copies/ml), 4 week group (445 cells/μl; 4.1 log10 copies/ml) and the 6 week group (621 cells/μl; 3.9 log10 copies/ml) (P = 0.77 for differences in CD4 cell count and P = 0.46 for differences in viral load). Women in the zidovudine group took 87% (range 66–100%) of the expected number of prenatal doses, and 88% of these women took at least one intrapartum drug dose (mean 1.7 doses).

Back to Top | Article Outline

Changes in HIV-1-RNA viral load

In the placebo group during the prenatal period, the median change from baseline viral load ranged from −0.08 to +0.23 log10 copies/ml (Fig. 1, Table 1). Starting at delivery, the viral load was 0.26 log10 copies/ml above baseline and remained elevated at the 2 and 4 week visits before returning to +0.10 log10 copies/ml above baseline at the 12 week visit. However, none of these variations were significantly different from the baseline level.

Fig. 1.

Fig. 1.

Table 1

Table 1

In the zidovudine group after 2 weeks of therapy, the median maternal viral load was 0.48 log10 copies/ml lower than at baseline (P = 0.02) (Table 1). The magnitude of the reduction after 2 weeks of therapy varied considerably in the three subgroups of women receiving zidovudine: the median viral load reduction was 0.11 log10 copies/ml in the 2 week therapy group, 0.51 log10 copies/ml in the 4 week therapy group, and 0.73 log10 copies/ml in the 6 week group (Fig. 1).

At later visits, the reduction in viral load differed in the three zidovudine subgroups: in the 6 week group the reduction continued at more than 0.7 log10 copies/ml until delivery; in the 4 week group there was a partial return towards the baseline value; and in the 2 week group, viral load was above baseline at the time of delivery (Fig. 1).

At the postpartum visits, viral loads in the women having received zidovudine remained between 0.17 and 0.21 log10 copies/ml above baseline. This increase was significantly above the baseline value at the 2 and 4 week postpartum visits (P < 0.05), but was not higher than the corresponding values in the placebo group (P > 0.69).

Back to Top | Article Outline

Changes in CD4 T cell counts

In the placebo group, CD4 cell counts remained stable during the prenatal period: changes 23 cells/μl or less (Table 1, Fig. 2). At delivery, however, the CD4 cell count increased to 157 cells/μl, which was significantly higher than at baseline (P = 0.03), and then declined gradually at the 2 and 4 week visits. At 6 months postpartum, the CD4 cell count was 28 cells/μl above the baseline value (P = 0.39).

Fig. 2.

Fig. 2.

In the zidovudine group, CD4 cell counts increased gradually after the initiation of therapy. CD4 cell counts after 2 and 4 weeks of therapy were significantly higher (P < 0.01) than at baseline, and higher than the corresponding values in the placebo group (P < 0.01). After delivery, CD4 cell counts declined gradually (remaining slightly above baseline), but were no higher than those in the placebo group at 4 weeks (P = 0.60) and 6 months (P = 0.78).

Back to Top | Article Outline

Resistance to zidovudine

To assess the presence of zidovudine resistance, we tested all women in the zidovudine group with an available DNA sample obtained at delivery (n = 20). We found no known mutations associated with zidovudine resistance (M41L, D67N, K70R, and T215Y/F) in any of the specimens tested (upper bound of 95% confidence interval = 15%).

Back to Top | Article Outline


In these HIV-1-infected women who started zidovudine therapy at between 34 and 38 weeks’ gestation, the viral load declined by approximately a half-log within the first 2 weeks. However, the magnitude and persistence of this reduction varied considerably by the gestational age at which women started therapy. Those with a longer duration of therapy (6 week therapy group), who presumably began therapy earlier in gestation, had a more marked reduction than those who delivered after only 2 or 4 weeks of therapy. It is unknown whether this variability is related to physiological changes occurring in late pregnancy, for example, impaired drug absorption. If it is real, and as a lower viral load is associated with a lower risk of transmission [6–9], this finding would suggest that it is important to initiate therapy earlier than 38 weeks’ gestation to obtain a maximal reduction in viral load.

A particular concern for women who take antiretroviral therapy in late pregnancy and then stop taking it at delivery is the possibility of an overshoot of viral load above baseline. Because most HIV-1-infected women in Africa breast-feed, this overshoot, if it occurs and if it results in an increase in breast-milk viral load, could result in an increased risk of transmission of HIV-1 through breast milk. Although the viral load increased above that at baseline, this increase also occurred in the placebo group, and could have been caused partly by immunological changes near the time of delivery [18]. As breast-milk samples were not collected, we are unable to assess the impact of prenatal zidovudine on breast-milk viral load; however, a separate analysis from this cohort indicated that breast milk and plasma HIV-1 viral loads are closely correlated [19].

Descriptions of the timing and magnitude of viral load overshoots from other studies have not been consistent. In some studies, this overshoot occurred within the first week [10]; in others the overshoot peaked 2–3 weeks after therapy was stopped [11,12]. These differences could be partly caused by the different antiretroviral treatments used and in the timing of blood sampling. It is possible that a significant overshoot did occur in our study, but that we did not observe it because our first data point occurred 2 weeks after delivery. However, the follow-up of the children in this cohort suggests that if such an overshoot occurs, it does not result in a substantial increase in postnatal transmission. The rate of new HIV infections between birth and 4 weeks of age was lower in infants born to the mothers in the zidovudine group than in those born to mothers in the placebo group [2]. More detailed studies with more frequent postpartum viral load measurements are needed to resolve this question.

Another concern is whether the short course negatively affects the mother's health by accelerating the clinical course of her HIV infection. It was therefore reassuring that the viral load at 12 weeks’ postpartum was not significantly higher among women receiving zidovudine. Similarly, after 6 months of postpartum follow-up, the CD4 cell counts in the zidovudine and placebo groups were almost identical, suggesting that short-course zidovudine does not adversely affect women's health.

The CD4 cell counts of women in the placebo group increased significantly near the time of delivery. This increase has been documented in other studies, and is thought to be caused by immunological changes related to pregnancy and delivery [18].

In view of the high fertility rate in many African nations, even among HIV-infected women, a woman might require zidovudine to prevent mother-to-child transmission of HIV-1 during more than one pregnancy. It was reassuring that after this short course we did not find any zidovudine resistance. This result is consistent with results from the ACTG 076 study, in which after approximately 12 weeks of zidovudine therapy, only 2.6% of women developed zidovudine resistance [20], but differs from the recent report of resistance after a single intrapartum dose of nevirapine [13].

Back to Top | Article Outline


Zidovudine initiated late in pregnancy resulted in a rapid decline in HIV-1 viral load, but not in a significant overshoot in viral load after the women stopped taking medication. In addition, during the postpartum period, viral loads and CD4 cell counts were not significantly different in women receiving zidovudine and in those receiving placebo, and no zidovudine-resistant strains were identified. These findings suggest that a short course of zidovudine administered late in the pregnancy of HIV-1-infected women to reduce mother-to-child transmission has no adverse health consequences for the women.

Back to Top | Article Outline


1. Shaffer N, Chuachoowong R, Mock PA. et al. Short-course zidovudine for prevention of mother-to-child transmission of HIV-1 in Bangkok, Thailand: a randomised trial. Lancet 1999, 353: 773–780.
2. Wiktor SZ, Ekpini E, Karon JM. et al. Short-course oral zidovudine for prevention of mother-to-child transmission of HIV-1 in Abidjan, Côte d'Ivoire: a randomised trial. Lancet 1999, 353: 781–785.
3. Dabis F, Msellati P, Meda N. et al. 6-Month efficacy, tolerance, and acceptability of a short regimen of oral zidovudine to reduce vertical transmission of HIV in breastfed children in Côte d'Ivoire and Burkina Faso: a double-blind placebo-controlled multicentre trial. Lancet 1999, 353: 786–792.
4. Guay LA, Musoke P, Fleming T. et al. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: HIVNET 012 randomised trial. Lancet 1999, 354: 795–802.
5. Gray G. The PETRA study: early and late efficacy of three short ZDV/3TC combination regimens to prevent mother-to-child transmission of HIV-1. In:XIIIth International Conference on AIDS, 2000, 9–14 July; Durban, South Africa. [Abstract LbOr5].
6. Sperling RS, Shapiro DE, Coombs RW. et al. Maternal viral load, zidovudine treatment, and the risk of transmission of human immunodeficiency virus type 1 from mother to infant. N Engl J Med 1996, 335: 1621–1629.
7. Mofenson LM, Lambert JS, Stiehm ER. et al. Risk factors for perinatal transmission of human immunodeficiency virus type 1 in women treated with zidovudine. N Engl J Med 1999, 341: 385–393.
8. Shaffer N, Roongpisuthipong A, Siriwasin W. et al. Maternal virus load and perinatal human immunodeficiency virus type 1 subtype E transmission, Thailand. J Infect Dis 1999, 179: 590–599.
9. Leroy V, Montcho C, Manigart O. et al. Maternal plasma viral load, zidovudine and mother-to-child transmission of HIV-1 in Africa: DITRAME ANRS 049a trial. AIDS 2001, 15: 517–522.
10. de Jon MD, de Boer RJ, de Wolf F. et al. Overshoot of HIV-1 viraemia after early discontinuation of antiretroviral treatment. AIDS 1997, 11: F79–F84.
11. Jubault V, Burgard M, Le Corfec E, Costagliola D, Rouzioux C, Viard J-P. High rebound of plasma and cellular HIV load after discontinuation of triple combination therapy. AIDS 1998, 12: 2358–2359.
12. Staszewski S, Miller V, Sabin C, Berger A, Hill AM, Phillips AN. Rebound of HIV-1 viral load after suppression to very low levels. AIDS 1998, 12: 2360.2360.
13. Jackson JB, Becker-Pergola G, Guay LA. et al. Identification of the K103N resistance mutation in Ugandan women receiving nevirapine to prevent HIV-1 vertical transmission. AIDS 2000, 14: F111–F115.
14. De Cock KM, Porter A, Kouadio JC. et al. Rapid and specific diagnosis of HIV-1 and HIV-1 infections: an evaluation of testing strategies. AIDS 1990, 4: 874–878.
15. Nkengasong J, Maurice C, Koblavi S. et al. Evaluation of HIV serologic parallel and serial testing algorithms in Abidjan, Côte d'Ivoire. AIDS 1999, 13: 109–117.
16. Nkengasong J, Maurice C, Koblavi S. et al. Field evaluation of a combination of monospecific ELISAs for the type-specific serodiagnosis of HIV-1 and HIV-2 infections in seropositive persons in Abidjan, Côte d'Ivoire. J Clin Microbiol 1998, 36: 123–127.
17. Durant J, Clevenbergh P, Halfon P. et al. Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial. Lancet 1999, 353: 2195–2199.
18. Biggar RJ, Pahwa S, Minkoff H. et al. Immunosuppression in pregnant women infected with human immunodeficiency virus. Am J Obstet Gynecol 1989, 161: 1239–1244.
19. Bertolli J, Byers B, Sibailly TS, et al. Two distinct patterns of HIV-1 viral load in early breast milk of HIV+ women treated with short-course zidovudine (ZDV) in Abidjan, Cote d'Ivoire. In:XIIIth International AIDS Conference. Durban, July 2000 [Abstract MoPeB2222].
20. Eastman PS, Shapiro, DE, Coombs RW. et al. Maternal viral zidovudine resistance and infrequent failure of zidovudine therapy to prevent perinatal transmission of human immunodeficiency virus type 1 in Pediatric AIDS Clinical Trials Group Protocol 076. J Infect Dis 1998, 177: 557–564.

Côte d'Ivoire; HIV-1; mother-to-child transmission; viral load; zidovudine

© 2002 Lippincott Williams & Wilkins, Inc.