Cesarean delivery for HIV-infected women: recommendations and controversies

Benefits of Cesarean Delivery and Evolution of Recommendations Regarding the Mode of Delivery for HIV-Infected Women 

Early observational studies suggested there might be a role for cesarean delivery in the prevention of MTCT of HIV. Several studies of twins who were born to HIV-infected women noted that, among twins who were born vaginally, first-born twins were more likely to be infected with HIV compared with second-born twins.6, 7 Because first-born twins are the first to pass through the birth canal, it was assumed that they would have the greatest exposure to infectious blood and genital tract secretions. In addition, first-born twins remain in the birth canal for a longer period of time.6 This supported the theory that exposure to HIV in the birth canal may play an important role in MTCT of HIV and that cesarean delivery could be protective by limiting exposure to blood and genital secretions in the birth canal. Additional retrospective and prospective studies yielded inconsistent results; some studies reported that cesarean delivery was associated with a decreased risk of MTCT of HIV,8, 9, 10, 11, 12 and other studies failed to show any association of the mode of delivery with transmission.13, 14, 15, 16

These early observational studies may have had conflicting results for several reasons. One critical issue was that it is often not easy to differentiate between cesarean deliveries that are performed before labor from those that are performed after the onset of uterine contractions, when microtransfusions of maternal blood into fetal circulation may occur.17 In addition, to have maximal effect, cesarean deliveries should be performed before the rupture of membranes, because once the integrity of the membranes is compromised the risk of infection increases.18 To standardize terminology, the term elective cesarean section was adopted widely to refer to a cesarean delivery that was performed before the onset of labor and before the rupture of membranes.19 Alternatively, the American College of Obstetricians and Gynecologists (ACOG) uses the term scheduled cesarean delivery so that it is not confused with other uses of the term elective in obstetrics.20, 21 Hereafter in this article, the term elective cesarean delivery is used to denote cesarean delivery before labor and before ruptured membranes.

A turning point came in June 1998, when the preliminary results from both a multicenter randomized clinical trial22 and a large individual patient data meta-analysis23 were presented at the 12th World AIDS Conference in Geneva, Switzerland, and published subsequently in 1999.18, 24 Results from the randomized trial, which was conducted in 6 European countries, demonstrated an 80% reduction in the rate of MTCT of HIV among women who were allocated to the elective cesarean delivery group. When the actual mode of delivery was analyzed, cesarean delivery after labor and/or after ruptured membranes resulted in an intermediate rate of MTCT of HIV (8.8%), compared with vaginal delivery (10.2%) and to elective cesarean delivery (2.4%). In this study, although elective cesarean delivery was associated with a decreased odds of transmission compared with vaginal delivery (odds ratio, 0.3; 95% CI, 0.1-0.8), cesarean delivery after labor and/or after ruptured membranes was not associated with a significant decreased odds of transmission compared with vaginal delivery (odds ratio, 1.0; 95% CI, 0.3-3.7).24 Results from the large metaanalysis of individual patient data from 15 prospective cohort studies indicated that elective cesarean delivery was associated with an approximately 50% reduction in the risk of MTCT of HIV.18 The results from the randomized trial and the individual patient data metaanalysis were sufficient for ACOG to issue new guidance regarding the role of cesarean delivery in perinatal HIV prevention. In August 1999, ACOG issued a committee opinion that recommended that HIV-infected women be offered a scheduled cesarean delivery at 38 completed weeks of gestation.21 The original 1999 ACOG committee opinion was updated in 2000.20 The current recommendations by ACOG20 and the US Public Health Service25 recommend that HIV-infected pregnant women with plasma loads of >1000 copies/mL be counseled regarding the benefits of an elective cesarean delivery. Elective cesarean deliveries should be performed at 38 completed weeks of gestation, based on the best clinical estimate of gestational age. A woman’s prenatal antiretroviral regimen should not be interrupted around the time of delivery. In addition, for most women, an infusion of zidovudine should be started at least 3 hours before the operation.20

Trends in Cesarean Delivery Rates Among HIV-Infected Women 

Since the release of the results of the randomized clinical trial in Europe24 and the individual patient data metaanalysis from North America and Europe18 and of subsequent guidelines,20, 25 the cesarean delivery rate among HIV-infected women has increased dramatically.5 In the United States, the cesarean delivery rates in a pediatric surveillance system and a pediatric HIV longitudinal cohort study demonstrated a doubling of cesarean delivery rates after June 1998, from 20% to nearly 50%.5 Unfortunately, it was not possible in this study to distinguish between elective cesarean deliveries and cesarean deliveries performed after the onset of labor or rupture of membranes. In addition, such studies of trends in the mode of delivery in the United States have not been updated recently, so it is unknown whether cesarean delivery rates among HIV-infected women have continued to increase or have stabilized over the last few years. In Europe, where cesarean delivery rates among HIV-infected women have been higher traditionally than in the United States,18 similar increases in cesarean delivery rates have been reported after 1998. For example, in a recent report from Sweden, the cesarean delivery rate for HIV-infected women increased from 8% in 1985-1993 to 44% in 1994-1998 and to 80% in 1999-2003.26 These higher rates may reflect a more aggressive policy of offering cesarean delivery to all women, regardless of viral load.27 The European Collaborative Study reported that the elective cesarean delivery rate among HIV-infected women increased from 1997-2000, then decreased slightly, and finally began increasing again in 2003. The authors speculate that these fluctuations were, in part, due to an increase in nonelective cesarean deliveries among women who had planned an elective procedure but were seen in labor or after rupture of membranes. These results led to a policy shift towards the scheduling of elective cesarean deliveries earlier in gestation.28

Associated Risks and Cost-Effectiveness of Cesarean Delivery for the Prevention of the MTCT of HIV 

In weighing the risks and benefits of elective cesarean delivery, the benefit of the prevention of MTCT must be carefully weighed against any increase of morbidity or mortality for either the woman or her infant, as well as increased costs and recovery time. A number of studies have addressed the question of whether HIV-infected women have higher postcesarean complication rates compared with HIV-uninfected control subjects. Most studies demonstrate an increased risk of postoperative morbidity, mostly infectious, in HIV-infected women compared with uninfected control subjects, and the risk of complications is correlated with the degree of immunosuppression.27, 29, 30, 31, 32, 33, 34

However, from a clinical perspective, the pertinent question is whether elective cesarean delivery increases an HIV-infected woman’s risk of complications compared with a vaginal delivery or with a nonelective cesarean delivery. Six studies24, 35, 36, 37, 38, 39 that address this issue were recently summarized in a Cochrane review.19 This review concluded that, among HIV-infected women, nonelective cesarean delivery was associated with the highest rate of postpartum morbidity, that elective cesarean delivery was intermediate in risk, and that vaginal delivery had the lowest risk of morbidity. Much of the postpartum morbidity was relatively minor, including postoperative fever, anemia, endometritis, and wound infection. Findings from this review reinforce the importance of the ACOG recommendations that all women who undergo cesarean delivery, regardless of HIV infection status, should receive prophylactic antibiotics.40 Maternal deaths are rare, and these studies did not have adequate sample sizes to assess potential differences in maternal mortality rates. Although short-term postoperative morbidity may be increased among HIV-infected women, it does not appear that mode of delivery is associated with more long-term effects, such as subsequent HIV disease progression.41

In resource-limited settings, there are limited data to suggest that the risks of postpartum morbidity42 and mortality43 rates among HIV-infected women who undergo cesarean delivery may be magnified. In addition, there may be inadequate resources available to provide cesarean deliveries for all HIV-infected women in settings of high HIV seroprevalence among pregnant women.

In terms of risks to the infant that are associated with cesarean deliveries, there are no studies to address this among infants born to HIV-infected women. However, we know from studies of HIV-uninfected women that the primary risk to infants that is associated with elective cesarean delivery is iatrogenic prematurity and its sequelae.44 To reduce the likelihood of onset of labor or rupture of membranes before delivery, ACOG recommends that elective cesarean delivery for HIV-infected women be scheduled at 38 completed weeks of gestation, which is 1 week earlier than for HIV-uninfected women.20 Because amniocentesis should be avoided in HIV-infected women, clinicians should rely on best clinical estimates of gestational age, rather than documentation of fetal lung maturity.20 In HIV-infected pregnant women, earlier delivery (38 vs 39 weeks) without documentation of fetal lung maturity may lead theoretically to more iatrogenic prematurity, although there are no data to support or refute this. It will be important to monitor the rate of infant morbidity because of iatrogenic prematurity in this setting.

Elective cesarean delivery has been shown to be relatively cost-effective45, 46, 47 and, in some cases, cost-saving.45, 48 However, these analyses focused primarily on women who were receiving prenatal and intrapartum zidovudine only and assumed relatively high transmission rates.45, 46, 47, 48 Because these cost-effectiveness models are very sensitive to changes in the baseline rate of MTCT,45, 46 if assumptions about the increased effectiveness of combination antiretroviral regimens are included, the cost/benefit of cesarean delivery is reduced markedly.45, 46 By contrast, because the cost of treating postpartum morbidity is relatively low compared with treatment of pediatric HIV disease, these models are relatively stable over a wide range of postpartum morbidity rates.45, 46 However, over a broad range of assumptions, elective cesarean delivery remains relatively cost-effective.

Unanswered Clinical Questions 

The initial ACOG guidelines21 were circumspect about the role of elective cesarean delivery in pregnant women with low HIV viral loads, and neither the randomized trial24 nor the metaanalysis18 could address this issue. Because these studies were conducted before the advent of viral load testing, these studies could not incorporate adjustment for viral load, which is now known to be critically important in the determination of the risk of MTCT.49 When the ACOG guidelines were updated in May 2000, they further specified that women with viral loads of >1000 copies/mL should be counseled regarding the benefits of cesarean delivery.20 These updated guidelines cited results from the Women and Infants Transmission Study, a prospective cohort study. In this analysis, no transmissions were reported among the 57 women with viral loads of <1000 copies/mL.50 For women with viral loads of <1000 copies/mL, the ACOG guidelines state that there are insufficient data with which to make recommendations regarding mode of delivery. The US Public Health Service guidelines51 are consistent with the ACOG guidelines and support elective cesarean delivery for women with viral loads of >1000 copies/mL25 and encourage additional clinical research regarding the potential role of elective cesarean delivery in decreasing MTCT among women with undetectable viral loads.

Another related unanswered question is whether there is any benefit of elective cesarean delivery in the prevention of MTCT of HIV among women who receive combination antiretroviral regimens, including highly active antiretroviral therapy (HAART). Both the randomized clinical trial24 and the individual patient data metaanalysis18 included mostly women who were receiving either no antiretrovirals or zidovudine only. In the randomized trial, although more than one half of the women received zidovudine during pregnancy, few women received combination antiretroviral regimens.24 In the individual patient data metaanalysis, >70% of mother-child pairs did not receive any antiretroviral drugs during the prenatal, intrapartum, or neonatal periods.18 Because the risk of transmission already is reduced substantially for women who take HAART prenatally (ie, 1%-2% transmission rate),52, 53 a study with a very large sample size would be needed to detect a further reduction in transmission because of elective cesarean delivery.

However, there are some studies that shed some light on the issue of whether elective cesarean delivery confers additional benefit among women with low viral loads while receiving HAART. In an individual patient data metaanalysis of 1202 women with plasma viral loads of <1000 copies/mL, cesarean delivery was an independent predictor of transmission risk in analyses that controlled for maternal receipt of antiretrovirals. Of note, among women who underwent cesarean deliveries, there were no transmissions among 270 women who received antiretroviral therapy, whereas there were 5 transmissions among 66 women who had not received antiretrovirals. However, this study was unable to distinguish elective from nonelective cesarean deliveries.54 In a recent report from the European Collaborative Study,28 elective cesarean delivery was associated independently with transmission risk in analyses that adjusted for maternal viral load and maternal antiretroviral therapy. When restricted to 560 women with undetectable viral loads, elective cesarean delivery was protective in univariate analyses. However, when adjusted for maternal antiretroviral use (none vs any), the association between elective cesarean delivery and transmission risk was no longer statistically significant (adjusted odds ratio, 0.52; 95% CI, 0.14-2.03). It is unclear whether these results mean there is no true protective effect or whether the study lacked adequate statistical power to reveal an association because of the small sample size. Because of their limitations, neither of these studies28, 54 definitively answers the question of whether elective cesarean delivery is associated with a decreased risk of MTCT of HIV among women with undetectable viral loads in the era of HAART.

Another unanswered clinical question is how soon after the onset of labor or the rupture of membranes that the benefit of cesarean delivery is lost. Although early studies dichotomized the length of membrane rupture and found that rupture of membranes for >4 hours was associated with a nearly 2-fold increase in transmission risk,15 a subsequent individual patient data metaanalysis demonstrated a continuously increasing risk of MTCT, with the transmission risk increasing approximately 2% for every additional hour of ruptured membranes.55 Therefore, how does one counsel a woman who planned for an elective cesarean delivery but who arrives in early labor or shortly after rupture of membranes? In such a situation, if a long period of labor is anticipated, some clinicians may choose to proceed with cesarean delivery; others may choose to proceed with an expedited vaginal delivery. What about the case of an HIV-infected woman with a high viral load who arrives with preterm labor or with premature preterm rupture of membranes? In these cases, the preferred mode and timing of delivery should be individualized on the basis of the specific clinical situation.

Other questions arise from clinical situations in which there is incomplete information about plasma viral load. For example, suppose a woman arrives late in pregnancy; she has not been receiving antiretroviral therapy, and viral load results are unlikely to be available before delivery. In this case, it is unlikely that her viral load will be suppressed adequately before delivery, and the woman should be counseled that elective cesarean delivery is likely to reduce her risk of transmission.

Comment 

Among HIV-infected pregnant women, cesarean delivery before labor and before the rupture of membranes has been shown to be safe and effective in reducing the risk of MTCT of HIV. However, the benefits of cesarean delivery in preventing MTCT of HIV must be weighed against potential increases in maternal and infant morbidity and the costs of cesarean delivery. In the United States, the benefits of cesarean delivery for women with viral loads of >1000 copies/mL generally outweigh the increased risk of minor postoperative morbidity. However, a number of important unanswered questions remain, such as how soon after labor onset or rupture of membranes the benefit of cesarean delivery is lost and whether cesarean delivery has a role in women with low HIV viral loads while receiving HAART. Furthermore, the appropriate role, if any, of elective cesarean delivery among HIV-infected women in various resource-limited settings with variable degrees of medical infrastructure and HIV prevalence rates will need to be better defined, particularly given the expanding availability of HAART in these settings.