Martin, Fabiola MD*†; Navaratne, Lesley MBBS‡; Khan, Wahid MBChB§; Sarner, Liat MBBS‖; Mercey, Danielle MBChB¶; Anderson, Jane MBBS†; Noble, Heather MBBS‖; Fakoya, Ade MBBS‖; Hawkins, David A MBBS§; Ruiter, Annemiek De MBBS‡; Taylor, Graham P MBChB*#
The British HIV Association Guidelines (1998-2005) on the management of HIV infection in pregnancy have 3 treatment strategies for mothers infected with HIV:1-3
* Zidovudine monotherapy (ZDVm) remains an option for asymptomatic mothers with CD4 counts >200 cells/μL and an HIV viral load (VL) <10,000 HIV RNA copies/mL of plasma who are intending to deliver by prelabor cesarean section (PLCS).
* Highly active antiretroviral therapy (HAART) is recommended for mothers who require therapy for themselves (ie, CD4 counts ≤200 cells/μL or symptomatic infection). HAART is a combination of antiretroviral therapy (ART) sufficient to suppress viral replication fully and usually contains at least 3 effective agents.
* Short-term HAART (START) is recommended for mothers not requiring therapy for their own health but who have a VL >10,000 HIV RNA copies/mL and for mothers with lower VLs who propose to deliver vaginally. START is HAART commenced during pregnancy and discontinued at delivery.
There are no data comparing these selective approaches in cohorts, and prescribing ZDVm in pregnancy has become a debated strategy. Some argue that PLCS could be avoided if HAART is given to all mothers. Great controversy still remains around the safety of HAART for mother and child. In some cohorts, HAART has been associated with increased rates of adverse events, including preterm delivery,4,5 low or very low birth weight,6 pre-eclampsia7,8 and hyperglycemia,9 but in other studies, this has not been found.10,11
Although many studies have focused on pregnancy outcomes, there is a paucity of data on long-term clinical outcome of mothers after the use of ZDVm, START or continuous HAART (cHAART) during pregnancy. There are concerns that ZDVm could jeopardize future treatment options by the selection of ZDV resistance, and mutations associated with resistance to nonnucleoside analogue reverse transcriptase inhibitors (NNRTIs) have been reported after START, even when the NNRTI was discontinued before the nucleoside backbone.12 We are therefore conducting an ongoing prospective multicenter audit of HIV-infected pregnant women, assessing their clinical outcome and response to future therapy postpartum.
PATIENTS AND METHODS
The audit is conducted at Chelsea and Westminster NHS Trust, Guy's and St. Thomas' NHS Foundation Trust, Homerton University NHS Trust, Mortimer Market Center University College Healthcare Trust, Newham Healthcare NHS Trust, and St. Mary's NHS Trust, London. Clinical outcomes and laboratory results up to September 2004 were collated on all HIV-infected women who delivered between January 1, 1998 and December 31, 2002.
Assessment of the clinical effectiveness of a medical protocol is part of the audit cycle that needs to be performed at regular intervals at NHS hospitals in the United Kingdom. Because audits examine data that have been collected during medical follow-up of patients, the Central Office for Research Ethics Committees does not require these to be reported.13 Anonymity of patient data is ensured by restricting data identification to dedicated clinic identifiers that are distinct from hospital and NHS numbers and known only at the respective clinics.
Depending on their health, CD4 count, and HIV VL, mothers were offered 1 or more of the aforementioned options recommended in the British HIV Association Guidelines. This resulted in 3 audit cohorts:
1. Asymptomatic mothers choosing ZDVm and PLCS because of good health, CD4 counts >200 cells/μL, and VLs ≤10,000 HIV RNA copies/mL (ZDVm group).
2. Patients requiring HAART for their own health, which is to be continued after delivery (cHAART group).
3. Patients in the START group, which is defined as HAART commenced during pregnancy and discontinued at delivery, requiring HAART not for their health but receiving HAART because of a high VL >10,000 HIV RNA copies/mL or because they wished to deliver vaginally.
All HIV-infected pregnant women presenting antenatally before labor who accepted antiretroviral therapy (ART) before delivery and delivered after 24 weeks' gestation were included. For each patient, a minimum follow-up period of 3 months postpartum was required, and the last follow-up was the time the mother was last seen after delivery by her HIV physician. The following data were analyzed:
1. CD4 counts and HIV VLs at three time points: first antenatal clinic (ANC) attendance, delivery, and last follow-up.
2. Centers for Disease Control and Prevention (CDC) status and death from all causes at last follow-up visit at the clinic up to September 2004.
3. Whether the mother was taking HAART at last follow-up (the type of HAART was not specified).
Although not part of the audit question, the HIV infection status of the babies is included as part of the description of the cohorts. The babies were considered to be uninfected with HIV if they had no antibodies against HIV or HIV DNA could not be detected after amplification by polymerase chain reaction (PCR) assay on two occasions and to be infected with HIV if they had 2 positive HIV DNA PCR assay results.
CD4 lymphocyte counts, HIV RNA quantification, and HIV DNA PCR assay results for all patients are those used in routine clinical practice in each hospital and are therefore not separately described. Mode of delivery was spontaneous vaginal delivery (SVD), planned PLCS, or emergency cesarean section.
Data were entered using Microsoft Excel software at each center and analyzed at St. Mary's NHS Trust. Presentation and outcomes were compared by treatment: ZDVm versus cHAART (all patients taking any HAART during pregnancy and continuing after delivery) and ZDVm versus START (all patients who commenced and stopped HAART at delivery). Qualitative variables were compared by use of the χ2 test for trend. Normally distributed variables were compared by the Student t test.
Between 1998 and 2002, 360 HIV-infected women delivered at the participating centers. Forty-six women stopped attending the HIV clinic within 3 months of delivery, 1 patient declined care, and 2 patients delivered before 24 weeks of gestation. These 49 mothers, who were not eligible for inclusion in the analysis, had higher nadir CD4 counts (326 vs. 270 cells/μL; P = 0.03) and rates of SVD (22% vs. 11%; P = 0.04) than those eligible for the study. Data on 311 women are audited in the final analysis.
Baseline demographics, as recorded at the first ANC visit, are presented in Table 1. Most of the mothers were of black-African origin and multiparous, and the most common risk factor, as documented in clinical records, for acquiring HIV was heterosexual intercourse. The median age at delivery was 31 years, but the women who took short-course therapy (ZDVm or START) were significantly younger (median = 29 years) than the mothers on cHAART.
Seventy-two percent of all the mothers delivered by PLCS, and SVD was more common in the cHAART and START groups. Twenty-nine women had 1 or more subsequent pregnancies. Of the 343 babies born, 298 have a confirmed HIV status. Only 1 baby, whose mother did not adhere to HAART, is infected with HIV-1.
Baseline (First Antenatal Clinic and Delivery) Data
Mothers Who Took Zidovudine Monotherapy During Pregnancy
Data are summarised in Table 2. As expected, the 85 mothers choosing to receive ZDVm had higher CD4 counts and lower VLs than those who were treated with HAART. At the first ANC visit, 49 (58%) women had a CD4 count of 201 to 500 cells/μL and 34 (40%) had a CD4 count >500 cells/μL; 10 (12%) women had a VL <50 HIV RNA copies/mL of plasma, 60 (71%) had a VL from 50 to 10,000 HIV RNA copies/mL of plasma, and 14 (16%) had a VL >10,000 HIV RNA copies/mL of plasma. One patient's VL was not available. Sixteen women received ZDVm outside of treatment guidelines: 1 woman with a CD4 count <200 cells/μL and in CDC category C requested monotherapy, but the reasons for ZDVm in a second mother classified as CDC category C and in the 14 women with a VL >10,000 HIV RNA copies/mL of plasma are not documented.
By delivery, 16 (19%) of 85 women had a VL <50 HIV RNA copies/mL and 36 (47%) had a VL from 50 to 1000 HIV RNA copies/mL.
Mothers Continuing HAART Postpartum
One hundred fifty-five mothers took HAART during pregnancy and continued this postpartum. At the first ANC visit, CD4 counts were ≤200 cells/μL in 70 (45%) women, 201 to 500 cells/μL in 71 (46%), and >500 cells/μL in 13 (8.4%); 42 (28%) mothers (including those taking HAART before conception) had no detectable viremia.
At delivery, 110 (74%) women had a VL <50 HIV RNA copies/mL of plasma and the median VL of those with detectable viremia was 399 HIV RNA copies/mL of plasma.
Mothers Who Received Short-Term HAART Until Delivery
At the first ANC visit, CD4 counts of mothers in the START group were ≤200 cells/μL in 9 (13%) women, from 201 to 500 cells/μL in 46 (66%), and >500 cells/μL in 15 (21%). One (1%) mother had no detectable viremia.
At delivery, 44 (68%) women had a VL <50 HIV RNA copies/mL of plasma and the median VL of those with detectable viremia was 143 HIV RNA copies/mL of plasma.
Comparing the two groups that received short-course therapy, the CD4 counts were significantly lower in the START group when compared with the ZDVm group (363 vs. 462 cells/μL; P < 0.001) at the first ANC visit. As expected, the median VLs were significantly higher in the START (log10 4.1 vs. log10 3.3; P < 0.001) group at the first ANC visit, because this was the reason why START was recommended to these women according to the British HIV Association Guidelines. These differences were no longer apparent by the time of delivery.
Data are summarised in Table 3. The analyzed follow-up period (Fig. 1) for all women starts at delivery of the first pregnancy after January 1, 1998. The mean follow-up period for 310 mothers after delivery was 33 (range: 3-72) months. Two hundred sixty-nine (87%) mothers were followed up for more than 12 months. There was a trend to a shorter period of follow-up in the ZDVm group. The total follow-up is 847 person-years.
At the last follow-up of these women:
* The entire cohort showed a significant rise (100 cells/μL) in median CD4 counts and a reduction (almost 2 log10) in median VL when compared with the first ANC visit.
* Ninety-eight percent of all mothers had an AIDS progression-free survival (Fig. 2).
* One hundred ninety-three (63%) mothers were receiving HAART, and 161 (83%) of these women had a VL <50 HIV RNA copies/mL of plasma.
Mothers Who Took Zidovudine Monotherapy During Pregnancy
At the last follow-up of these women:
* The median CD4 count of those treated with ZDVm remained greater than 500 cells/μL (CD4 counts were ≤200 cells/μL in 6 [7%] women, from 201 to 500 cells/μL in 35 [42%], and >500 cells/μL in 42 [51%]).
* Nineteen (23%) mothers had a VL <50 HIV RNA copies/mL of plasma, 49 (58%) had a VL from 50 to 10,000 HIV RNA copies/mL of plasma, and 15 (18%) had a VL >10,000 HIV RNA copies/mL of plasma.
* Twenty-two (26%) mothers were receiving HAART, and 20 (91%) of these women had a VL <50 HIV RNA copies/mL of plasma.
During follow-up, 3 women had progressed from CDC category A to category C and were diagnosed, respectively, with pulmonary tuberculosis, military tuberculosis, and Pneumocystis pneumonia.
Mothers Continuing HAART Postpartum
At the last follow-up of these women:
* The median CD4 counts were ≤200 cells/μL in 14 (9.2%) women, from 201 to 500 cells/μL in 75 (49%), and >500 cells/μL in 64 (42%). Compared with measurements at delivery, the median CD4 count of the cHAART group fell to less than 500 cells/μL but remained greater than 200 cells/μL (median = 360 cells/μL).
* One hundred twenty (78%) patients had a VL <50 HIV RNA copies/mL of plasma, 20 (13%) had a VL from 50 to 10,000 HIV RNA copies/mL of plasma, and 13 (8.5%) had a VL >10,000 HIV RNA copies/mL of plasma.
* One hundred forty-three (92%) patients were taking HAART; of these, 119 (78%) had a VL <50 HIV RNA copies/mL of plasma. One patient in this group who was taking didanosine, stavudine, and nevirapine during pregnancy developed severe acute lactic acidosis and died of adult respiratory distress syndrome 14 days after her emergency cesarean section.14
During follow-up, 2 patients progressed to stage B and 2 progressed to stage C. The AIDS-defining illness (ADI) of 1 mother was pulmonary tuberculosis, and the other is not known.
Mothers Who Received Short-Term HAART Until Delivery
At the last follow-up of these women:
* All but 5 mothers in the START group had CD4 counts >200 cells/μL, 75 (49%) had a CD4 count between 201 and 500 cells/μL, and 64 (42%) had CD4 counts >500 cells/μL.
* Twenty-two (31%) mothers had a VL <50 HIV RNA copies/mL of plasma, 22 (26%) had a VL from 50 to 10,000 HIV RNA copies/mL of plasma, and 27 (32%) had a VL >10,000 HIV RNA copies/mL of plasma.
* Twenty-eight (40%) mothers were receiving HAART, and of these, 22 (79%) had a VL <50 HIV RNA copies/mL of plasma.
During follow-up, 1 mother progressed to CDC category C with pulmonary tuberculosis.
Comparison Between the Groups
Zidovudine Monotherapy Group Compared With the Continuous HAART Group
Whereas at the first ANC visit and at delivery, there is a significant difference in CD4 counts between the ZDVm and cHAART groups, this is lost by the time of last follow-up, mainly because of the improvement in CD4 counts in those continuing or restarting ART in each group.
At last follow-up, 20 (24%) women had no detectable viremia in the ZDVm group compared with 119 (77%) in the cHAART group. The difference in median VL (2067 vs. 6860 HIV RNA copies/mL of plasma; P < 0.001) between the ZDVm and cHAART groups at the first ANC visit was lost at delivery (391 vs. 49 HIV RNA copies/mL of plasma; P = 0.8) and restored (1886 vs. 49 HIV RNA copies/mL of plasma; P = 0.04) at last follow-up.
At last follow-up, 22 (26%) mothers in the ZDVm group versus 143 (92%) women in the cHAART group were taking HAART for their own health. All women in the ZDVm group and 135 (88%) in the cHAART group who were taking HAART at last follow-up had a CD4 count >200 cells/μL. Of those taking HAART at last follow-up, 20 (91%) of 22 in the ZDVm group and 119 (78%) in the cHAART group had no detectable viremia (P < 0.001).
Zidovudine Monotherapy Group Compared With the Short-Term HAART Group
All patients in the ZDVm group and the START group were treatment naive (or previously ZDVm-only exposed) and accepted ART for prevention of mother-to-child transmission of HIV. Mothers in the START group were followed up slightly longer (33 vs. 30 months; P = 0.06). At last follow-up, the median CD4 count was significantly lower in the START group, reflecting the lower nadir CD4 counts of the mothers in this group. The median VL in the START and ZDVm groups was undetectable. At last follow-up, 28 (40%) women in the START group compared with 22 (26%) in the ZDVm group had commenced regular HAART and 22 (71%) of 28 had no detectable viremia compared with 20 (91%) of 22, but neither difference was statistically significant.
Over the years, there has been concern that pregnancy might lead to increased mortality and morbidity in HIV-infected women, but many of the studies have been small, had short follow-up periods, or have focused on immunologic parameters rather than on clinical outcomes. In the era before protease inhibitors and NNRTIs, a decline or no change in CD4 counts during pregnancy was reported.15-17 Similarly, in the SEROGEST and SEROCO cohorts (241 pregnant and 124 nonpregnant women with known dates of seroconversion between 1988 and 1996 and matched for CD4 counts were studied), there was no evidence that pregnancy after seroconversion resulted in a more rapid progression to AIDS after 39 months of follow-up.18 Although these data are encouraging, some questions remain. Do the results apply to other cohorts or to women presenting later in the natural history of HIV infection? What effect does HIV therapy have on maternal outcome, especially short-course therapy, be it ZDVm or HAART, with the possible development of drug resistance, and in light of the SMART study,19 is there any danger in starting and stopping HAART for the prevention of mother-to-child transmission of HIV?
We have presented data on morbidity, mortality, and surrogate markers for a cohort of 311 women with 847 person-years follow-up. One mother died, whose death was attributed to drug toxicity, and only 6 mothers developed an ADI, at a rate of 1 case per 141 person-years, with no significant difference between the 3 treatment groups.
Since 1997, the use of HAART in pregnancy has, where affordable, become widespread. This may be for maternal indications, to reduce transmission risk further, or to give the mother the option of a vaginal delivery. Nevertheless, this approach exposes mothers and babies to polypharmacy and increases the risk of drug toxicity and poor adherence.
In the European Mode of Delivery Study,20 combining ZDVm with PLCS reduced the risk of HIV transmission to the babies to <1% overall. ZDVm is cost-effective, easy to adhere to, and avoids polypharmacy. Although many believe that lower rates of transmission may be achieved with fully suppressive HAART without the need for PLCS, there are no robust data to confirm this to date. Nevertheless, there is concern that ZDVm might have an adverse impact on future maternal therapeutic options through the emergence of ZDV resistance mutations. In our cohort, only 22 (26%) women in the selected ZDVm group had commenced HAART, for their own health, after a mean of 30 months. All had good immunologic and viral responses. Only 2 had low-level detectable viremia. Although 3 had progressed to CDC category C, all were well on HAART at follow-up, and in 2 women, the ADI was pulmonary tuberculosis, which occurred at CD4 counts greater than 200 cells/μL. Pulmonary tuberculosis is prevalent in our black-African population, can develop at any CD4 count, and is not routinely considered an indication to start HAART in the United Kingdom. Most ZDVm patients had remained well 30 months after delivery with no need for HAART.
Although it might be assumed that short courses of HAART, through effective suppression of viral replication, might give better clinical protection and prevent future treatment failures more effectively than ZDVm, this has not been the case in our cohort.12,21 Because of the selection criteria for the British HIV Association therapeutic options, the START group obviously differed in their surrogate markers (but not in CDC staging) from the ZDVm group at the first ANC visit. At last follow-up, more women in the START group had restarted HAART. This is to be expected, because one of the criteria for START during pregnancy is that the maternal VL is usually >10,000 HIV copies/mL. Although there was no statistical difference, women in the ZDVm group tended to be more successful in suppressing their VL when eventually started on HAART postpartum than those in the START group (91% vs. 79%). This is most probably attributable to the significantly lower VL of the ZDVm group at the initiation of therapy. From this study, there is no evidence that previous treatment with ZDVm has an adverse impact on the clinical response to subsequent HAART. This is consistent with an earlier observation that ZDV-associated mutations may be uncommon in ZDVm-exposed mothers when this approach has been selected on the basis of CD4 counts >200 cells/μL, VL <10,000 log copies/mL, and good physical health.22 The possibility that ZDV was selectively avoided when starting HAART during follow-up could not be excluded in this cohort, however.
Although not biased by the process of recruiting to a clinical trial, which might select more motivated, socially secure, and well-supported mothers, our audit is limited by the relatively short follow-up duration, the rate of loss to follow-up, and the lack of comparison with a matched cohort of women who are infected with HIV but not pregnant. Our cohort is composed mostly of heterosexual non-drug injecting black-African women. Our experience may not be reproduced in cohorts with a higher proportion of other ethnic backgrounds or injecting drug users. The components of HAART prescribed were not audited, but most women received 2 NNRTIs combined with nevirapine, whereas the most common protease inhibitor used was nelfinavir (personal communication with centers).
In summary, these data can reinforce the growing confidence of women with HIV infection, who wish to start or extend their family, that transmission rates less than 1% are achievable and that, with access to HAART, progression to AIDS is infrequent, regardless of the type of therapy taken during pregnancy. Although this is likely to be the common experience, we believe that this is the first documentation of maternal medium-term clinical outcome after pregnancy in a general clinic population during the HAART era. Although our audit did not look for the protective effects of PLCS, our data suggest that ZDVm initiated at 24 to 30 weeks of gestation remains a valid option for selected healthy HIV-infected pregnant women prepared to deliver by PLCS, with CD4 counts >200 cells/μL and VLs <10,000 HIV RNA copies/mL, and it does not seem to jeopardize future therapy options for the mothers. Similarly, no adverse effect of starting HAART in pregnancy to prevent mother-to-child transmission and stopping HAART once the baby has been delivered was seen. The outcome of our patient cohort, which reflects evolving practice in the general clinical setting, provides clinicians with evidence of healthy maternal survival after pregnancy for up to 6 years. A second audit cycle is planned for September 2006.
1. Taylor GP, Lyall E, Mercey D, et al. British HIV Association guidelines for prescribing antiretroviral therapy in pregnancy (1998). Sex Trans Infect. 1999;75:90-97.
2. Lyall E, Blott M, de Ruiter A, et al. Guidelines for the management of HIV infection in pregnant women and the prevention of mother-to-child transmission, British HIV Association. HIV Med. 2001;12:314-334.
3. The British HIV Association Pregnancy Guidelines Writing Committee 2005. Guidelines for the management of HIV infection in pregnancy and the prevention of mother-to-child transmission. 2001. Available at: http://www/bhiva.org/
4. European Collaborative Study, Swiss Mother and Child HIV Cohort Study. Combination antiretroviral therapy and duration of pregnancy. AIDS. 2000;14:2913-2930.
5. European Collaborative Study. Increased risk of adverse pregnancy outcomes in HIV-infected women treated with highly active antiretroviral therapy in Europe. AIDS. 2004;18:2337-2339.
6. Tuomala RE, Shapiro DE, Mofenson LM, et al. Antiretroviral therapy during pregnancy and the risk of an adverse outcome. N Engl J Med. 2002;346:1863-1870.
7. Wimalasundera RC, Smith JH, Thom S, et al. Is pre-eclampsia in HIV positive women treated with antiretroviral therapy a manifestation of immune reconstitution. Lancet. 2002;360:1152-1154.
8. Suy A, Martinez E, Coll O, et al. Increased risk of pre-eclampsia and fetal death in HIV-infected pregnant women receiving highly active antiretroviral therapy. AIDS. 2006;20:59-66.
9. Gonzalez-Tome MI, Ramos J, Solis I, et al, for the Spanish Cohort of HIV Infected Mother-Infant Pairs. Gestational diabetes and ART in pregnant HIV-1-infected women [abstract 68]. Presented at: 12th Conference on Retroviruses and Opportunistic Infections; 2005; Boston.
10. Tuomala RE, Watts DH, Li D, et al. Improved obstetric outcomes and few maternal toxicities are associated with antiretroviral therapy, including highly active antiretroviral therapy during pregnancy. J Acquir Immune Defic Syndr. 2005;38:449-473.
11. Beckerman K, Covington D, Garcia P, et al. Association between antiretroviral therapy during pregnancy and prematurity/low birth weight. 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, CA, February 8-11, 2004.
12. Lyons F, Coughlan S, Byrne C, et al. Emergence of antiretroviral resistance in HIV-positive women receiving combination antiretroviral therapy in pregnancy. AIDS. 2005;19:63-67.
14. Sarner L, Fakoya A. Acute onset lactic acidosis and pancreatitis in the third trimester of pregnancy in HIV-1 positive women taking antiretroviral medication. Sex Trans Infect. 2002;78:58-59.
15. Burns DN, Nourjah P, Minkoff H, et al. Changes in CD4+ and CD8+ cell levels during pregnancy and post partum in women seropositive and seronegative for human immunodeficiency virus-1. Am J Obstet Gynecol. 1996;174:1461-1468.
16. 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.
17. Miotti PG, Liomba G, Dallabetta GA, et al. T lymphocyte subsets during and after pregnancy: analysis in human immunodeficiency virus type 1-infected and -uninfected Malawian mothers. J Infect Dis. 1992;165:1116-1119.
18. Saada M, Le Chenadec J, Berrebi A, et al. Pregnancy and progression to AIDS: results of the French prospective cohorts. SEROGEST and SEROCO Study Groups. AIDS. 2000;14:2355-2360.
19. El-Sadr W, Neaton J. Episodic CD4 guided use of ART is inferior to continuous therapy: results of the SMART study [abstract 106LB]. Presented at: 13th Conference on Retroviruses and Opportunistic Infections, Denver, CO, February 5-8, 2006.
20. The European Mode of Delivery Collaboration. Elective caesarian-section versus vaginal delivery in prevention of vertical HIV-1 transmission: a randomised clinical trial. Lancet. 1999;353:1035-1039.
21. 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.
22. Larbalestier N, Mullen J, O'Shea S, et al. Drug resistance is uncommon in pregnant women with low viral loads taking zidovudine monotherapy to prevent perinatal HIV transmission. AIDS. 2003;17:2665-2667.
© 2006 Lippincott Williams & Wilkins, Inc.