Single-Dose Perinatal Nevirapine plus Standard Zidovudine to Prevent Mother-to-Child Transmission of HIV-1 in Thailand
http://www.100md.com
《新英格兰医药杂志》
ABSTRACT
Background Although zidovudine prophylaxis decreases the rate of transmission of the human immunodeficiency virus (HIV) type 1 substantially, a large number of infants still become infected. We hypothesized that the administration, in addition to zidovudine, of a single dose of oral nevirapine to mothers during labor and to neonates would further reduce transmission of HIV.
Methods We conducted a randomized, double-blind trial of three treatment regimens in Thai women who were receiving zidovudine therapy during the third trimester of pregnancy. In one group, mothers and infants received a single dose of nevirapine (nevirapine–nevirapine regimen); in another, mothers and infants received nevirapine and placebo, respectively (nevirapine–placebo regimen); and in the last, mothers and infants received placebo (placebo–placebo regimen). The infants also received one week of zidovudine therapy and were formula-fed. The end point of the study was infection with HIV in the infants, established by virologic testing.
Results Between January 15, 2001, and February 28, 2003, a total of 1844 Thai women were enrolled. At the first interim analysis, the independent data monitoring committee stopped enrollment in the placebo–placebo group. Among women who delivered before the interim analysis, the as-randomized Kaplan–Meier estimates of the transmission rates were 1.1 percent (95 percent confidence interval, 0.3 to 2.2) in the nevirapine–nevirapine group and 6.3 percent (95 percent confidence interval, 3.8 to 8.9) in the placebo–placebo group (P<0.001). The final per-protocol transmission rate in the nevirapine–nevirapine group, 1.9 percent (95 percent confidence interval, 0.9 to 3.0), was not significantly inferior to the rate in the nevirapine–placebo group (2.8 percent; 95 percent confidence interval, 1.5 to 4.1). Nevirapine had an effect within subgroups defined by known risk factors such as viral load and CD4 count. No serious adverse effects were associated with nevirapine therapy.
Conclusions A single dose of nevirapine to the mother, with or without a dose of nevirapine to the infant, added to oral zidovudine prophylaxis starting at 28 weeks' gestation, is highly effective in reducing mother-to-child transmission of HIV.
Mother-to-child transmission of the human immunodeficiency virus (HIV) remains a major health problem worldwide, even though the preventive efficacy of zidovudine was established in 1994.1,2,3 In Thailand, where 1.3 percent of pregnant women are infected with HIV, since 1993 the Ministry of Public Health has operated programs for voluntary counseling and testing of pregnant women and has provided free infant formula to mothers infected with HIV.4 Since 1999, this national program has provided zidovudine therapy during the last trimester of pregnancy, at delivery, and to the newborn. Even with these interventions, however, transmission rates remain at or above 6 percent.5,6 When a clinical trial demonstrated the efficacy of a single dose of nevirapine (Viramune, Boehringer Ingelheim) during labor plus a single dose to the newborn,7 we hypothesized that, without additional toxicity, logistic complications, or significant cost, perinatal nevirapine therapy added to zidovudine therapy could further reduce mother-to-child transmission of HIV in Thailand.
Methods
Trial Design
We conducted a multicenter, phase 3, double-blind, randomized, placebo-controlled trial. The primary objective was to assess the safety and efficacy of a single dose of nevirapine administered to women at the onset of labor and given to infants between 48 and 72 hours after birth, added to zidovudine prophylaxis, for the prevention of mother-to-child transmission of HIV type 1. A secondary objective was to evaluate the incremental effect of the administration of perinatal nevirapine to newborns.
The ethics committees of the Thai Ministry of Public Health, Chiang Mai University, and the Harvard School of Public Health approved the protocol used in the study, including the amendments. All study sites complied with regulations of the Department of Health and Human Services for the protection of human research subjects.
Patients
Eligible subjects were pregnant women participating in the national program of voluntary counseling and testing who were receiving zidovudine prophylaxis (starting at 28 weeks' gestation or as soon as possible thereafter) at any of 37 study sites in Thailand. The subjects were enrolled between January 15, 2001, and February 28, 2003. Pre-enrollment evaluations included HIV testing; history taking and physical examination; ultrasonographic uterine imaging; measurement of hemoglobin, white-cell and differential counts; CD4 count; measurements of plasma HIV RNA, alanine aminotransferase, and creatinine; and hepatitis B surface antigen and hepatitis C virus serologic testing.
The subjects were randomly assigned to one of three study groups at enrollment if they had received zidovudine for at least 2 weeks, had provided written informed consent, agreed not to breast-feed, and had the following laboratory values within 21 days before randomization: hemoglobin level greater than 8.0 g per deciliter, absolute neutrophil count greater than 750 cells per cubic millimeter, alanine aminotransferase level less than five times the upper limit of normal (i.e., less than 35 U per liter), and creatinine level less than 1.5 mg per deciliter (132.6 μmol per liter). Exclusion criteria included a maternal or fetal condition or a concomitant treatment that contraindicated treatment with zidovudine or nevirapine, oligohydramnios, unexplained polyhydramnios or in utero anemia, or a medical condition that required immediate use of highly active antiretroviral therapy. Open-label nevirapine was available to women who did not enroll in the study.
Until delivery, the women in the study had an obstetrical examination every two weeks; every other week, blood was drawn for hematologic, virologic, and biochemical testing. Adherence to zidovudine therapy was assessed at each visit by counting pills. After delivery, the women were seen at 10 days, 6 weeks, and 4 months for a physical examination and for hematologic and virologic monitoring. They were referred to an internist for further follow-up.
Infants were examined at birth, and follow-up visits were scheduled at 10 days (for measurement of hematocrit and alanine aminotransferase levels), 6 weeks (for hematologic tests), and 4, 6, 9, and 12 months. At each visit, a supply of powdered formula was provided to the mother; the child's health history since the previous visit was recorded, including the measurement of weight, recumbent length, and head circumference; and a clinical examination of the infant was performed. To determine the presence or absence of HIV infection, peripheral blood obtained at birth, six weeks, and four and six months was spotted onto filter papers, dried, and stored at –20°C to await shipping to the New England Newborn Screening Program for testing with a polymerase-chain-reaction (PCR) DNA assay for HIV (Amplicor HIV-1 DNA, version 1.5, Roche Molecular Systems).5 Serious adverse events were reported to the ministry of public health and Boehringer Ingelheim.
All the women received 300 mg of zidovudine twice daily starting at 28 weeks' gestation or as soon as possible thereafter; the dosing regimen was changed to 300 mg every three hours from the onset of labor until delivery. Infants received 2 mg of zidovudine per kilogram of body weight in an oral suspension every six hours for one week after birth. In accordance with national guidelines, if the mother had received zidovudine for a period of less than four weeks, the infant was treated for four to six weeks.8
Prophylaxis against opportunistic infections was provided as necessary, including the administration of oral trimethoprim–sulfamethoxazole to immunocompromised women9 and to infants from six weeks of age until it was established that they were not infected.10 After delivery, immunocompromised women and HIV-infected, immunocompromised infants were offered antiretroviral treatment that was provided by the Ministry of Public Health.8
Treatment
Women were randomly assigned to one of three treatment groups. In the nevirapine–nevirapine group, women received one dose of oral nevirapine (200 mg) at the onset of labor and neonates received nevirapine oral suspension (6-mg fixed dose, 0.6 ml) between 48 and 72 hours after birth. In the placebo–placebo group, women received placebo at the onset of labor and neonates received placebo between 48 and 72 hours after birth. In the nevirapine–placebo group, women received one dose of oral nevirapine (200 mg) at the onset of labor and neonates received placebo between 48 and 72 hours after birth.
In instances of prolonged and false labor, additional doses were provided to women every 48 hours until delivery. Infants born less than one hour after the mother had ingested nevirapine or placebo while in labor, according to the random assignment of treatment, were to be given the assigned treatment within six hours after birth.
Randomization and Blinding
On enrollment, women were provided with the study drug, either nevirapine or matched placebo, to be taken at home at the onset of labor. The study drugs were identified by random numbers with the use of permuted blocks of six in a ratio of 1:1:1. Twins were assigned to receive the same study drug. Additional doses were accessible in a blinded fashion at any hospital participating in the study with the use of a shared code. The study drugs were provided by Boehringer Ingelheim and were repackaged at the pharmacy of Children's Hospital, Boston, in individual blisters or oral syringes (Exacta-Med dispenser, Baxa).
Primary End Point
Infants were considered to be infected with HIV if the results of the PCR test were positive in blood obtained on two separate occasions, and infants were considered to be uninfected if the test results were negative on two occasions after one month of age.11 Infants who did not meet this definition were evaluated and classified as infected (six infants) or uninfected (three infants) by a clinician and a virologist before unblinding. In twins, infection in one or both was counted as one transmission.
Statistical Analysis
The study was designed primarily to test the superiority of the nevirapine–nevirapine regimen over the placebo–placebo regimen. We calculated that 484 mother–infant pairs were needed in each treatment group for a 0.05 one-sided type I error with 0.8 power, assuming a transmission rate of 8 percent in the placebo–placebo group and 4 percent in the nevirapine–nevirapine group. The nevirapine–placebo group was added in order to estimate, as a secondary objective, the proportion of overall transmission that was prevented by treating infants with nevirapine. Assuming that 5 percent of the cases could not be evaluated, the initial target sample size was 1530.
Interim safety analyses were planned when 40 percent and 70 percent of the total number of women had been enrolled. An increased rate of transmission associated with any of the three treatment regimens would be considered significant if any nominal P value was less than 0.0004 at the first interim analysis (with the use of O'Brien–Fleming boundaries as modified by Lan and DeMets and the use of one-sided tests with Bonferroni adjustments for multiple comparisons).12,13,14
After the first interim analysis, on May 2, 2002, the independent data and safety monitoring committee recommended stopping enrollment in the placebo–placebo group. The trial design was modified, and the target sample size was increased to 695 per treatment group to ensure 0.8 power to test for noninferiority of the nevirapine–placebo regimen as compared with the nevirapine–nevirapine regimen. Noninferiority would be declared if the upper limit of the 95 percent one-sided confidence interval of the observed difference was lower than 2.5 percent.
Two separate efficacy analyses were performed. The first efficacy analysis included all three treatment groups and all mothers who delivered before the first interim analysis. The second tested the noninferiority of the nevirapine–placebo regimen as compared with the nevirapine–nevirapine regimen. In addition to an as-randomized analysis, a per-protocol analysis was performed, primarily to test for noninferiority, in which the populations compared were defined according to actual intake of the assigned study drug, major deviations from the protocol, and adherence to zidovudine prophylaxis.15 Infants were analyzed according to the actual intake of the maternal or infant study drug. For the per-protocol analysis, criteria for exclusion were the intake of placebo by the mother and nevirapine by the infant, an infant's receipt of nevirapine more than 72 hours after birth or more than 6 hours after birth if the mother delivered less than 1 hour after receiving nevirapine, undocumented timing of intake by the mother or infant of the study drug, intake by the mother of the study drug more than 96 hours before delivery, less than 50 percent adherence to zidovudine prophylaxis or adherence that was not documented, breast-feeding, and maternal initiation of highly active antiretroviral therapy during pregnancy.
Characteristics of the mother, the delivery, and the infant were compared among the treatment groups with the use of the chi-square and Kruskal–Wallis tests. The Kaplan–Meier method and Greenwood's formula were used to estimate transmission rates and their standard errors. The time to the first positive HIV test was considered the time to the end point. Data on infants with negative HIV test results were censored at the time of their last HIV test. The method developed by Balasubramanian and Lagakos, which does not require a predefined operational definition of HIV infection, was used for confirmation; the results are not presented, because they were not different from the transmission rates estimated with the Kaplan–Meier method and Greenwood's formula.16
The differential effect of nevirapine was tested by a comparison within subgroups, defined according to known risk factors for transmission, of the difference in transmission rates between the placebo–placebo group and the nevirapine–nevirapine group (before the first interim analysis) and between the nevirapine–placebo group and the nevirapine–nevirapine group (at the final analysis). In case of rejection of homogeneity of the effect of nevirapine across subgroups with the use of a chi-square test,12 the differential effect was also examined within each subgroup. Except for noninferiority testing, all reported P values are two-sided, without adjustment for multiple testing. Adverse events in mothers and infants were analyzed from enrollment until six months after the birth of the infant.
Results
Enrollment
Of 3061 pregnant women who underwent testing for HIV at the study sites and whose results were positive, 1844 were randomly assigned to study treatment and 1807 were followed through to delivery. Of those not enrolled, 316 women declined enrollment, 395 moved to another town or were lost to follow-up, 50 enrolled in a perinatal nevirapine pharmacokinetics substudy, 104 delivered before pre-enrollment evaluations, 129 terminated their pregnancy, 5 had a spontaneous abortion, 3 started highly active antiretroviral therapy, 2 died before inclusion, and 213 did not meet the inclusion criteria, including 184 who delivered after less than two weeks of zidovudine prophylaxis and were offered open-label nevirapine. Thirty-nine women randomly assigned to the placebo–placebo group were excluded from the analysis because they were provided with open-label nevirapine after the first interim analysis. There were 12 twin pregnancies and 14 fetal deaths (including 1 twin). Enrollment, loss to follow-up, pregnancy outcomes, and available end points for each treatment group in the study are summarized in Figure 1. End points were available for 97.6 percent of the 1766 live-born infants included in the analysis.
Figure 1. Treatment Assignment, Loss to Follow-up, and Outcome of Pregnancy.
Thirty-nine women assigned to the placebo–placebo group were excluded from the analyses, because they were given open-label nevirapine (NVP) after the first interim analysis. PL denotes placebo. Delivery of twins counted as only one delivery.
Characteristics of the Women, Deliveries, and Infants
Baseline characteristics of the women, the deliveries, and the newborns were similar across treatment groups (Table 1). Zidovudine prophylaxis was initiated at a median of 28.7 weeks' gestation. The median duration of zidovudine prophylaxis was 9.7 weeks. The median CD4 count among the women was 372 cells per cubic millimeter, and the median viral load was 4.0 log10 copies per milliliter. Fifty-six women had received antiretroviral therapy during a previous pregnancy.
Table 1. Characteristics of Mothers, Deliveries, and Infants in the Study, According to the Treatment Assignment.
Thirteen percent of the women delivered at or before 37 weeks' gestation. The median interval between the onset of labor and administration of a dose of the assigned treatment was 2.1 hours, and the median interval between receipt of the dose of the assigned treatment and delivery was 6.5 hours. Seventy women received two doses of the assigned treatment because of false or prolonged labor. Twenty percent of the deliveries were by cesarean section, of which 28 percent were performed before the onset of labor.
Nine percent of the infants weighed less than 2500 g (5.5 lb) at birth. The median interval between birth and administration of the first dose of the assigned treatment was 48.4 hours. Eight percent of infants received zidovudine for at least one month, because their mothers had received less than four weeks of zidovudine prophylaxis. There was one report of breast-feeding by the mother.
Analysis of Efficacy
First Interim Analysis
In women who delivered by May 2, 2002, the as-randomized transmission rate among those in the nevirapine–nevirapine group was 1.1 percent (95 percent confidence interval, 0.3 to 2.2), which was significantly lower than the rate of 6.3 percent (95 percent confidence interval, 3.8 to 8.9) in the placebo–placebo group (P<0.001). This finding confirmed the results of the first interim analysis, at which time the data and safety monitoring board recommended stopping the placebo–placebo regimen. In the nevirapine–placebo group, the rate was 2.1 percent (95 percent confidence interval, 0.6 to 3.7). Per-protocol transmission rates were very similar (Table 2).
Table 2. Kaplan–Meier Estimates of the Rate of Transmission of HIV at Six Months, According to Assigned and Actual Treatment.
Final Analysis
The final per-protocol transmission rate among those in the nevirapine–nevirapine group (1.9 percent; 95 percent confidence interval, 0.9 to 3.0) was not statistically inferior to the rate in the nevirapine–placebo group (2.8 percent; 95 percent confidence interval, 1.5 to 4.1). The upper limit of the one-sided confidence interval of the difference between these observed rates was 2.2, which was below the predefined 2.5 percent inferiority threshold. The rates in the as-randomized analysis were similar — 2.0 percent (95 percent confidence interval, 1.0 to 3.1) in the nevirapine–nevirapine group and 2.8 percent (95 percent confidence interval, 1.6 to 4.1) in the nevirapine–placebo group.
Other Analyses of Transmission
Among infants, the rate of testing positive for HIV within three days after birth in the placebo–placebo group (3.1 percent) was significantly higher than the rate among mothers in the study groups assigned to receive nevirapine before the interim analysis (0.4 percent) (P<0.001) (Table 2). In the final analysis, the overall rate of testing positive within three days after birth was 1.3 percent (95 percent confidence interval, 0.7 to 1.9). An analysis of the rate of death or HIV infection in infants before six months of age did not modify the conclusions of the efficacy analysis (Table 3).
Table 3. Estimates of the Rate of Transmission of HIV or Death at Six Months, According to Treatment Assignment.
Effect of Nevirapine According to Known Risk Factors for Transmission
Nevirapine had an effect according to the known risk factors for transmission, including viral load and CD4 count at baseline, gestational age at the initiation of zidovudine treatment, gestational duration, and the duration of zidovudine prophylaxis, although the effect often varied across subgroups (Table 4).2,5,17,18,19,20,21 In addition, the nevirapine–nevirapine regimen was somewhat superior to the nevirapine–placebo regimen in most subgroups.
Table 4. Effect of Nevirapine among Infants with Known Risk Factors for Transmission of HIV.
Maternal Safety
No severe rashes were observed. At 10 days, mild rashes were reported in 3.8 percent of mothers. Alanine aminotransferase levels were 36 to 42 IU per liter in 8.5 percent of mothers and 43 to 69 IU per liter in 0.9 percent (none of the mothers had levels above 70 IU per liter); there was no significant difference among the treatment groups. Rates of serious adverse events were similar across the groups (216 reports) and were related to pregnancy (59 percent), infection or HIV (26 percent), possibly zidovudine prophylaxis (anemia in 7 percent), possibly nevirapine (e.g., allergic reaction during therapy in one woman), and other conditions (7 percent).
Infant Safety
In infants, the rates of rash were 15.9 percent and 7.5 percent at 10 and 45 days of life, respectively, with no significant difference among treatment groups. Likewise, increases in alanine aminotransferase levels (36 to 69 IU per liter in 4.8 percent of infants and 260 IU per liter in one infant; none had levels above 350 IU per liter) and the presence of hyperbilirubinemia were not associated with exposure to nevirapine. The rates of serious adverse events were similar across treatment groups (598 reports) and were related to neonatal or obstetrical conditions (11 percent), to congenital abnormalities (6 percent), to any infection, including HIV (72 percent), possibly to zidovudine (anemia in 2 percent), and to other causes (9 percent). Prolonged neonatal icterus (1 percent) was classified as possibly related to nevirapine even in the placebo–placebo group, since classification was performed by persons blinded to treatment assignment. Nineteen infants died within six months after birth (Table 3).
Discussion
This study demonstrates the high efficacy of adding a single dose of nevirapine in the mother, with or without a dose in the infant, to oral zidovudine prophylaxis for the reduction of perinatal transmission of HIV. The observed reduction by 80 percent that led to the early interruption of enrollment in the placebo–placebo group was much higher than we had hypothesized when we designed the study. The transmission rate observed in the placebo–placebo group was consistent with rates observed in other studies in which zidovudine prophylaxis was initiated at the beginning of the last trimester of pregnancy,1,19,20,21 in particular in our previous trial, which was performed in a similar population.5
Although the efficacy of a single maternal dose of nevirapine was declared statistically noninferior to that of a maternal dose plus an infant dose with the use of our predetermined criteria (2.5 percent as the upper limit of the one-sided 95 percent confidence interval for the difference between the observed rates of transmission), in fact, two doses appeared consistently better than one throughout our analysis (Table 4). Moreover, because the nevirapine–placebo group had a somewhat lower median viral load at baseline than the nevirapine–nevirapine group, the difference in transmission rates might be an underestimate. On the basis of these findings and for additional logistic reasons, the ministry of public health in Thailand has recommended administration of nevirapine to both mother and infant, with a slight modification, that of giving nevirapine to the mother on arrival at the hospital and to the infant as soon as possible after birth.22
The percentage of infants with a positive HIV test within three days after birth was significantly lower in the nevirapine–nevirapine and the nevirapine–placebo groups than in the placebo–placebo group. This surprising finding suggests either that infants become HIV-positive earlier after infection than previously suspected or that perinatal nevirapine may delay the detection of HIV positivity or even abort infection completely when transmission occurs in the last few weeks of pregnancy.
Newborns in this study were formula-fed. This recommendation, made by the Ministry of Public Health in 1993, is strongly adhered to and has been shown to have no negative effect either on the health of the infant or on the promotion of breast-feeding among HIV-negative women.5,23,24
In the South African Intrapartum Nevirapine Trial, the rate of intrapartum transmission of HIV was three times as high when the mother had received nevirapine less than two hours before delivery as when the mother had received it earlier.25 Our study did not confirm this finding. However, since the administration of nevirapine to infants appears to be safe whatever the delay between birth and receipt of the drug7,25 and given that in clinical practice women may take their nevirapine dose later during labor and stay in the hospital for a shorter period, it may be preferable to administer the dose of nevirapine to the infant shortly after birth, with the first dose of zidovudine.
In the Pediatric AIDS Clinical Trials Group study known as PACTG 316, which evaluated the efficacy of perinatal nevirapine therapy added to standard antiretroviral therapy, no benefit of nevirapine was shown.26 The effect of nevirapine was not observed even in the subgroup of women who received only zidovudine in addition to nevirapine. This subgroup, however, may have been selected to receive zidovudine monotherapy because of the subjects' low viral load. In addition, 43 percent of the women in the group delivered by planned cesarean section. We also found that in the subgroup of women whose viral load at entry was less than 25,000, the transmission rate was very low in the placebo–placebo group (2 percent) and only marginally lower in the nevirapine groups. Although a subgroup analysis has the potential for bias, the effect of perinatal nevirapine was observed in most of the subgroups, a finding that suggests that there is no population for whom nevirapine prophylaxis would not be beneficial.
The discovery of drug-resistance mutations in a substantial proportion of the recipients of perinatal nevirapine27,28,29,30,31,32,33,34 has raised questions of a possible risk to mothers who subsequently receive nonnucleoside reverse-transcriptase inhibitors as therapy for their own disease. The resistance consequences among the participants of our study are described by Jourdain et al. elsewhere in this issue of the Journal.35 The maternal risk is juxtaposed with the clear benefit to the babies. The relative value of this or other interventions in any particular setting will depend on considerations of cost, medical logistics, and drug access, as well as on data from future studies of outcomes in women receiving both perinatal prophylaxis and antiretroviral therapy.36 Nevirapine at delivery added to third-trimester zidovudine results in very low transmission rates with minimal medical and financial burdens — rates that are very similar to those achieved with more complex, expensive, and potentially toxic multidrug maternal regimens.37,38,39,40,41,42,43
Supported by grants from the National Institutes of Health (5 R01 HD 39615); the Agence Nationale de Recherches sur le Sida (ANRS 12-08); the Ministry of Public Health, Thailand; the Institut de Recherche pour le Développement, France; the Institut National d'Etudes Démographiques, France; the Department of Technical and Economic Cooperation, Thailand; Boehringer Ingelheim; GlaxoSmithKline (which provided the zidovudine); and Roche Molecular Systems.
Dr. Jourdain reports having equity in Aventis.
We are indebted to S. Hammer and M. Tod for their contribution to the design of the study; to M.P. Kieffer for assistance in preparing the initial trial proposal; to the data and safety monitoring committee members V. Phoolcharoen, D. Costagliola, V. Vithayasai, A. Chaovavanich, for their advice and support throughout the trial; to the staff of the Perinatal HIV Prevention Trial, as follows: for administrative support: N. Chaiboonruang, Y. Thita, P. Pirom, D. Tiptepin, N. Pawadee, R. Klaijak, W. Ruangphornwisut; tracking and supplies: K. Taninard, P. Asongchaiwatanakul, S. Phonin; safety monitoring: S. Chalermpantmetagul, R. Peongjakta; sites monitoring: P. Sukrakanchana, A. Saiphueng, B. Yapuang, W. Pongchaisit, Y. Yensuang, B. Subhamitra, C. Kanabkaew, C. Puttimit, J. Wallapachai, N. Kruenual, N. Kongstan, N. Krapunpongsakul, R. Wongchai, S. Thammajitsagul, S. Jina, N. Seetapun; data management: R. Seubmongkolchai, A. Maleesatharn, J. Suewan, N. Kawee, S. Chailert, S. Tanasri, A. Wongja, K. Yoddee, M. Nuchniyom, K. Chaokasem, P. Supinya, J. Maneechan, K. Suebmongkolchai, S. Tansenee, A. Seubmongkolchai, C. Pholhinkong, C. Chimplee, K. Saopang, N. Karbkam, P. Chusut, P. Onnoy, S. Aumtong; and laboratory: T. Cressey, W. Khamduang, A. Preedam, A. Kummee, K. Janjaroen, M. Sumonwiriya, P. Chanthubthong, W. Boonprasit, J. Kamkorn, P. Tungyai, S. Kanthawong, S. Kunkeaw; to the Thai Ministry of Public Health for advice and assistance: Office of the Permanent Secretary, Department of Health, Department of Communicable Diseases Control, and Provincial Hospitals Division, especially, M. Teeratantikanont, P. Amornwichet, V. Chokevivat, T. Siraprapasiri, T Liewsaree, W. Liewsaree, P. Kantipong, C. Bowonwatanuwong, P. Leenasirimakul, A. Nilmanat, A. Chutanunta, O. Kamsao, W. Panitsuk, P. Pathipvanich; and Chiang Mai University: W. Sirirungsi, P. Lechanachai, U. Haesungcharern, A. Matanasaravoot; to the microbiology reference laboratories: R. Sutthent, P. Auewarakul (Mahidol University, Thailand); A.M. Comeau, J. Gerstel-Thompson (University of Massachusetts), F. Barin (Tours University, France); to the National Institute of Child Health and Human Development: A. Willoughby, B. Nugent, L. Mofenson; to Boehringer Ingelheim: E. Van Steenberge; to GlaxoSmithKline: S. Chinnapha, W. Rungsiyapornrat; to the following colleagues who contributed to this project in many critical ways: M. Essex, P. Matangkhasombut, A. Patterson, P. Casky, P. Jacob, J. Sullivan, C. Bellec, E. Kiley, J. Morcos, M.L. Bezencon, U. Thisyakorn, M. Mirochnich, A. Kuo, D. Bray, K. Young, A. Kuo, N. Briand, R. Balasubramanian; to all members of the hospital teams; and to the women and children who participated in this study.
* The coinvestigators and study coordinators of the Perinatal HIV Prevention Trial (Thailand) are listed in the Appendix.
Source Information
From Epidémiologie Clinique, Santé Maternelle et Infantile et Sida, Institut de Recherche pour le Développement, Paris (M.L., G.J., S.L., N.N.-G.-H.); the Department of Immunology and Infectious Diseases, Harvard School of Public Health, Boston (M.L., G.J., N.N.-G.-H.); Institut National d'Etudes Démographiques, Paris (S.L.); INSERM, Erm 0321, Paris (J.Y.M.); the Family Health Research Center, Mahidol University, Bangkok, Thailand (S. Koetsawang); the Ministry of Public Health, Bangkok, Thailand (S. Kanshana, V.T.); and Children's Hospital and the Department of Pediatrics, Harvard Medical School — both in Boston (K.M.).
Address reprint requests to Dr. Lallemant at the Perinatal HIV Prevention Trial — Institut de Recherche pour le Développement, Unité 54, 29/7-8 Samlan Rd., Soi 1 Prasing Muang, Chiang Mai 50200, Thailand, or at marc@phpt.org.
References
Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994;331:1173-1180.
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.
Mofenson LM, McIntyre JA. Advances and research directions in the prevention of mother-to-child HIV-1 transmission. Lancet 2000;355:2237-2244.
Kanshana S, Thewanda D, Teeraratkul A, et al. Implementing short-course zidovudine to reduce mother-infant HIV transmission in a large pilot program in Thailand. AIDS 2000;14:1617-1623.
Lallemant M, Jourdain G, Le Coeur S, et al. A trial of shortened zidovudine regimens to prevent mother-to-child transmission of human immunodeficiency virus type 1. N Engl J Med 2000;343:982-991.
Amornwichet P, Teeraratkul A, Simonds RJ, et al. Preventing mother-to-child HIV transmission: the first year of Thailand's national program. JAMA 2002;288:245-248.
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.
Ministry of Public Health. National guidelines for the clinical management of HIV infection in children and adults. Nonthaburi, Thailand: AIDS Division, Department of Communicable Disease Control, 2000.
1999 USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virus. MMWR Recomm Rep 1999;48:1-59, 61.
1995 Revised guidelines for prophylaxis against Pneumocystis carinii pneumonia for children infected with or perinatally exposed to human immunodeficiency virus. MMWR Recomm Rep 1995;44:1-11.
Revised recommendations for HIV screening of pregnant women. MMWR Recomm Rep 2001;50:63-85.
Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: John Wiley, 1981:161-5.
O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics 1979;35:549-556.
Lan KKG, DeMets DL. Changing frequency of interim analysis in sequential monitoring. Biometrics 1989;45:1017-1020.
Ellenberg JH. Intention-to-treat analysis. In: Redmond CK, Colton T, eds. Biostatistics in clinical trials. Chichester, England: John Wiley, 2001:245-9.
Balasubramanian R, Lagakos SW. Estimation of the timing of perinatal transmission of HIV. Biometrics 2001;57:1048-1058.
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.
Shapiro DE, Sperling RS, Coombs RW. Effect of zidovudine on perinatal HIV-1 transmission and maternal viral load. Lancet 1999;354:156-156.
Wade NA, Birkhead GS, Warren BL, et al. Abbreviated regimens of zidovudine prophylaxis and perinatal transmission of the human immunodeficiency virus. N Engl J Med 1998;339:1409-1414.
Matheson PB, Abrams EJ, Thomas PA, et al. Efficacy of antenatal zidovudine in reducing perinatal transmission of human immunodeficiency virus type 1. J Infect Dis 1995;172:353-358.
Simpson BJ, Shapiro ED, Andiman WA. Reduction in the risk of vertical transmission of HIV-1 associated with treatment of pregnant women with orally administered zidovudine alone. J Acquir Immune Defic Syndr Hum Retrovirol 1997;14:145-152.
Guidelines for the management of prevention of mother to child transmission of HIV in Thailand: policy of prevention of mother to child transmission of HIV in Thailand. Nonthaburi, Thailand: Ministry of Public Health, Department of Health, December 2003.
Kanshana S, Simonds RJ. National program for preventing mother-child HIV transmission in Thailand: successful implementation and lessons learned. AIDS 2002;16:953-959.
Talawat S, Dore GJ, Le Coeur S, Lallemant M. Infant feeding practices and attitudes among women with HIV infection in northern Thailand. AIDS Care 2002;14:625-631.
Moodley D, Moodley J, Coovadia H, et al. A multicenter randomized controlled trial of nevirapine versus a combination of zidovudine and lamivudine to reduce intrapartum and early postpartum mother-to-child transmission of human immunodeficiency virus type 1. J Infect Dis 2003;187:725-735.
Dorenbaum A, Cunningham CK, Gelber RD, et al. Two-dose intrapartum/newborn nevirapine and standard antiretroviral therapy to reduce perinatal HIV transmission: a randomized trial. JAMA 2002;288:189-198.
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.
Eshleman SH, Mracna M, Guay LA, et al. Selection and fading of resistance mutations in women and infants receiving nevirapine to prevent HIV-1 vertical transmission (HIVNET 012). AIDS 2001;15:1951-1957.
Cunningham CK, Chaix ML, Rekacewicz C, et al. Development of resistance mutations in women receiving standard antiretroviral therapy who received intrapartum nevirapine to prevent perinatal human immunodeficiency virus type 1 transmission: a substudy of Pediatric AIDS Clinical Trials Group protocol 316. J Infect Dis 2002;186:181-188.
Morris L, Pillay C, Chezzi C, et al. Low frequency of the V106M mutation among HIV-1 subtype C-infected pregnant women exposed to nevirapine. AIDS 2003;17:1698-1700.
Kantor R, Lee E, Johnston E, et al. Rapid flux in non-nucleoside reverse transcriptase inhibitor resistance mutations among subtype C HIV-1-infected women after single dose nevirapine. Antiviral Ther 2003;8:S85. abstract.
Chaix ML, Montcho C, Ekouevi DK, et al. Genotypic resistance analysis in women who received intrapartum nevirapine associated to a short course of zidovudine to prevent perinatal HIV-1 transmission: the Ditrame Plus ANRS 1201/02 Study, Abidjan, C?te d'Ivoire. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:657. abstract.
Chalermchokcharoenkit A, Asavapiriyanont S, Teeraratkul A, et al. Combination short-course zidovudine plus 2-dose nevirapine for prevention of mother-to-child transmission: safety, tolerance, transmission, and resistance results. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:96. abstract.
Martinson N, Morris L, Gray G, et al. HIV resistance and transmission following single-dose nevirapine in a PMTCT cohort. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:38. abstract.
Jourdain G, Ngo-Giang-Huong N, Le Coeur S, et al. Intrapartum exposure to nevirapine and subsequent maternal responses to nevirapine-based antiretroviral therapy. N Engl J Med 2004;351:229-240.
World Health Organization. Antiretroviral drugs and the prevention of mother-to-child transmission of HIV infection in resource-limited settings: expert consultation, Geneva, 5-6 February 2004: a summary of main points from the meeting. (Accessed June 18, 2004, at http://www.who.int/3by5/arv_pmtct/en/.)
Bucceri AM, Somigliana E, Matrone R, et al. Combination antiretroviral therapy in 100 HIV-1-infected pregnant women. Hum Reprod 2002;17:436-441.
Cooper ER, Charurat M, Mofenson L, et al. Combination antiretroviral strategies for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr 2002;29:484-494.
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.
Wimalasundera RC, Larbalestier N, Smith JH, et al. Pre-eclampsia, antiretroviral therapy, and immune reconstitution. Lancet 2002;360:1152-1154.
Shapiro D, Tuomala R, Pollack H, et al. Mother-to-child HIV transmission risk according to antiretroviral therapy, mode of delivery, and viral load in 2895 U.S. women (PACTG 367). In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:99. abstract.
Shepard KV. Important new safety information: clarification of risk factors for severe, life-threatening and fatal hepatotoxicity with VIRAMUNE (nevirapine). Ridgefield, Conn.: Boehringer Ingelheim Pharmaceuticals, February 2004.
Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Washington, D.C.: Department of Health and Human Services, March 23, 2004.(Marc Lallemant, M.D., Gon)
Background Although zidovudine prophylaxis decreases the rate of transmission of the human immunodeficiency virus (HIV) type 1 substantially, a large number of infants still become infected. We hypothesized that the administration, in addition to zidovudine, of a single dose of oral nevirapine to mothers during labor and to neonates would further reduce transmission of HIV.
Methods We conducted a randomized, double-blind trial of three treatment regimens in Thai women who were receiving zidovudine therapy during the third trimester of pregnancy. In one group, mothers and infants received a single dose of nevirapine (nevirapine–nevirapine regimen); in another, mothers and infants received nevirapine and placebo, respectively (nevirapine–placebo regimen); and in the last, mothers and infants received placebo (placebo–placebo regimen). The infants also received one week of zidovudine therapy and were formula-fed. The end point of the study was infection with HIV in the infants, established by virologic testing.
Results Between January 15, 2001, and February 28, 2003, a total of 1844 Thai women were enrolled. At the first interim analysis, the independent data monitoring committee stopped enrollment in the placebo–placebo group. Among women who delivered before the interim analysis, the as-randomized Kaplan–Meier estimates of the transmission rates were 1.1 percent (95 percent confidence interval, 0.3 to 2.2) in the nevirapine–nevirapine group and 6.3 percent (95 percent confidence interval, 3.8 to 8.9) in the placebo–placebo group (P<0.001). The final per-protocol transmission rate in the nevirapine–nevirapine group, 1.9 percent (95 percent confidence interval, 0.9 to 3.0), was not significantly inferior to the rate in the nevirapine–placebo group (2.8 percent; 95 percent confidence interval, 1.5 to 4.1). Nevirapine had an effect within subgroups defined by known risk factors such as viral load and CD4 count. No serious adverse effects were associated with nevirapine therapy.
Conclusions A single dose of nevirapine to the mother, with or without a dose of nevirapine to the infant, added to oral zidovudine prophylaxis starting at 28 weeks' gestation, is highly effective in reducing mother-to-child transmission of HIV.
Mother-to-child transmission of the human immunodeficiency virus (HIV) remains a major health problem worldwide, even though the preventive efficacy of zidovudine was established in 1994.1,2,3 In Thailand, where 1.3 percent of pregnant women are infected with HIV, since 1993 the Ministry of Public Health has operated programs for voluntary counseling and testing of pregnant women and has provided free infant formula to mothers infected with HIV.4 Since 1999, this national program has provided zidovudine therapy during the last trimester of pregnancy, at delivery, and to the newborn. Even with these interventions, however, transmission rates remain at or above 6 percent.5,6 When a clinical trial demonstrated the efficacy of a single dose of nevirapine (Viramune, Boehringer Ingelheim) during labor plus a single dose to the newborn,7 we hypothesized that, without additional toxicity, logistic complications, or significant cost, perinatal nevirapine therapy added to zidovudine therapy could further reduce mother-to-child transmission of HIV in Thailand.
Methods
Trial Design
We conducted a multicenter, phase 3, double-blind, randomized, placebo-controlled trial. The primary objective was to assess the safety and efficacy of a single dose of nevirapine administered to women at the onset of labor and given to infants between 48 and 72 hours after birth, added to zidovudine prophylaxis, for the prevention of mother-to-child transmission of HIV type 1. A secondary objective was to evaluate the incremental effect of the administration of perinatal nevirapine to newborns.
The ethics committees of the Thai Ministry of Public Health, Chiang Mai University, and the Harvard School of Public Health approved the protocol used in the study, including the amendments. All study sites complied with regulations of the Department of Health and Human Services for the protection of human research subjects.
Patients
Eligible subjects were pregnant women participating in the national program of voluntary counseling and testing who were receiving zidovudine prophylaxis (starting at 28 weeks' gestation or as soon as possible thereafter) at any of 37 study sites in Thailand. The subjects were enrolled between January 15, 2001, and February 28, 2003. Pre-enrollment evaluations included HIV testing; history taking and physical examination; ultrasonographic uterine imaging; measurement of hemoglobin, white-cell and differential counts; CD4 count; measurements of plasma HIV RNA, alanine aminotransferase, and creatinine; and hepatitis B surface antigen and hepatitis C virus serologic testing.
The subjects were randomly assigned to one of three study groups at enrollment if they had received zidovudine for at least 2 weeks, had provided written informed consent, agreed not to breast-feed, and had the following laboratory values within 21 days before randomization: hemoglobin level greater than 8.0 g per deciliter, absolute neutrophil count greater than 750 cells per cubic millimeter, alanine aminotransferase level less than five times the upper limit of normal (i.e., less than 35 U per liter), and creatinine level less than 1.5 mg per deciliter (132.6 μmol per liter). Exclusion criteria included a maternal or fetal condition or a concomitant treatment that contraindicated treatment with zidovudine or nevirapine, oligohydramnios, unexplained polyhydramnios or in utero anemia, or a medical condition that required immediate use of highly active antiretroviral therapy. Open-label nevirapine was available to women who did not enroll in the study.
Until delivery, the women in the study had an obstetrical examination every two weeks; every other week, blood was drawn for hematologic, virologic, and biochemical testing. Adherence to zidovudine therapy was assessed at each visit by counting pills. After delivery, the women were seen at 10 days, 6 weeks, and 4 months for a physical examination and for hematologic and virologic monitoring. They were referred to an internist for further follow-up.
Infants were examined at birth, and follow-up visits were scheduled at 10 days (for measurement of hematocrit and alanine aminotransferase levels), 6 weeks (for hematologic tests), and 4, 6, 9, and 12 months. At each visit, a supply of powdered formula was provided to the mother; the child's health history since the previous visit was recorded, including the measurement of weight, recumbent length, and head circumference; and a clinical examination of the infant was performed. To determine the presence or absence of HIV infection, peripheral blood obtained at birth, six weeks, and four and six months was spotted onto filter papers, dried, and stored at –20°C to await shipping to the New England Newborn Screening Program for testing with a polymerase-chain-reaction (PCR) DNA assay for HIV (Amplicor HIV-1 DNA, version 1.5, Roche Molecular Systems).5 Serious adverse events were reported to the ministry of public health and Boehringer Ingelheim.
All the women received 300 mg of zidovudine twice daily starting at 28 weeks' gestation or as soon as possible thereafter; the dosing regimen was changed to 300 mg every three hours from the onset of labor until delivery. Infants received 2 mg of zidovudine per kilogram of body weight in an oral suspension every six hours for one week after birth. In accordance with national guidelines, if the mother had received zidovudine for a period of less than four weeks, the infant was treated for four to six weeks.8
Prophylaxis against opportunistic infections was provided as necessary, including the administration of oral trimethoprim–sulfamethoxazole to immunocompromised women9 and to infants from six weeks of age until it was established that they were not infected.10 After delivery, immunocompromised women and HIV-infected, immunocompromised infants were offered antiretroviral treatment that was provided by the Ministry of Public Health.8
Treatment
Women were randomly assigned to one of three treatment groups. In the nevirapine–nevirapine group, women received one dose of oral nevirapine (200 mg) at the onset of labor and neonates received nevirapine oral suspension (6-mg fixed dose, 0.6 ml) between 48 and 72 hours after birth. In the placebo–placebo group, women received placebo at the onset of labor and neonates received placebo between 48 and 72 hours after birth. In the nevirapine–placebo group, women received one dose of oral nevirapine (200 mg) at the onset of labor and neonates received placebo between 48 and 72 hours after birth.
In instances of prolonged and false labor, additional doses were provided to women every 48 hours until delivery. Infants born less than one hour after the mother had ingested nevirapine or placebo while in labor, according to the random assignment of treatment, were to be given the assigned treatment within six hours after birth.
Randomization and Blinding
On enrollment, women were provided with the study drug, either nevirapine or matched placebo, to be taken at home at the onset of labor. The study drugs were identified by random numbers with the use of permuted blocks of six in a ratio of 1:1:1. Twins were assigned to receive the same study drug. Additional doses were accessible in a blinded fashion at any hospital participating in the study with the use of a shared code. The study drugs were provided by Boehringer Ingelheim and were repackaged at the pharmacy of Children's Hospital, Boston, in individual blisters or oral syringes (Exacta-Med dispenser, Baxa).
Primary End Point
Infants were considered to be infected with HIV if the results of the PCR test were positive in blood obtained on two separate occasions, and infants were considered to be uninfected if the test results were negative on two occasions after one month of age.11 Infants who did not meet this definition were evaluated and classified as infected (six infants) or uninfected (three infants) by a clinician and a virologist before unblinding. In twins, infection in one or both was counted as one transmission.
Statistical Analysis
The study was designed primarily to test the superiority of the nevirapine–nevirapine regimen over the placebo–placebo regimen. We calculated that 484 mother–infant pairs were needed in each treatment group for a 0.05 one-sided type I error with 0.8 power, assuming a transmission rate of 8 percent in the placebo–placebo group and 4 percent in the nevirapine–nevirapine group. The nevirapine–placebo group was added in order to estimate, as a secondary objective, the proportion of overall transmission that was prevented by treating infants with nevirapine. Assuming that 5 percent of the cases could not be evaluated, the initial target sample size was 1530.
Interim safety analyses were planned when 40 percent and 70 percent of the total number of women had been enrolled. An increased rate of transmission associated with any of the three treatment regimens would be considered significant if any nominal P value was less than 0.0004 at the first interim analysis (with the use of O'Brien–Fleming boundaries as modified by Lan and DeMets and the use of one-sided tests with Bonferroni adjustments for multiple comparisons).12,13,14
After the first interim analysis, on May 2, 2002, the independent data and safety monitoring committee recommended stopping enrollment in the placebo–placebo group. The trial design was modified, and the target sample size was increased to 695 per treatment group to ensure 0.8 power to test for noninferiority of the nevirapine–placebo regimen as compared with the nevirapine–nevirapine regimen. Noninferiority would be declared if the upper limit of the 95 percent one-sided confidence interval of the observed difference was lower than 2.5 percent.
Two separate efficacy analyses were performed. The first efficacy analysis included all three treatment groups and all mothers who delivered before the first interim analysis. The second tested the noninferiority of the nevirapine–placebo regimen as compared with the nevirapine–nevirapine regimen. In addition to an as-randomized analysis, a per-protocol analysis was performed, primarily to test for noninferiority, in which the populations compared were defined according to actual intake of the assigned study drug, major deviations from the protocol, and adherence to zidovudine prophylaxis.15 Infants were analyzed according to the actual intake of the maternal or infant study drug. For the per-protocol analysis, criteria for exclusion were the intake of placebo by the mother and nevirapine by the infant, an infant's receipt of nevirapine more than 72 hours after birth or more than 6 hours after birth if the mother delivered less than 1 hour after receiving nevirapine, undocumented timing of intake by the mother or infant of the study drug, intake by the mother of the study drug more than 96 hours before delivery, less than 50 percent adherence to zidovudine prophylaxis or adherence that was not documented, breast-feeding, and maternal initiation of highly active antiretroviral therapy during pregnancy.
Characteristics of the mother, the delivery, and the infant were compared among the treatment groups with the use of the chi-square and Kruskal–Wallis tests. The Kaplan–Meier method and Greenwood's formula were used to estimate transmission rates and their standard errors. The time to the first positive HIV test was considered the time to the end point. Data on infants with negative HIV test results were censored at the time of their last HIV test. The method developed by Balasubramanian and Lagakos, which does not require a predefined operational definition of HIV infection, was used for confirmation; the results are not presented, because they were not different from the transmission rates estimated with the Kaplan–Meier method and Greenwood's formula.16
The differential effect of nevirapine was tested by a comparison within subgroups, defined according to known risk factors for transmission, of the difference in transmission rates between the placebo–placebo group and the nevirapine–nevirapine group (before the first interim analysis) and between the nevirapine–placebo group and the nevirapine–nevirapine group (at the final analysis). In case of rejection of homogeneity of the effect of nevirapine across subgroups with the use of a chi-square test,12 the differential effect was also examined within each subgroup. Except for noninferiority testing, all reported P values are two-sided, without adjustment for multiple testing. Adverse events in mothers and infants were analyzed from enrollment until six months after the birth of the infant.
Results
Enrollment
Of 3061 pregnant women who underwent testing for HIV at the study sites and whose results were positive, 1844 were randomly assigned to study treatment and 1807 were followed through to delivery. Of those not enrolled, 316 women declined enrollment, 395 moved to another town or were lost to follow-up, 50 enrolled in a perinatal nevirapine pharmacokinetics substudy, 104 delivered before pre-enrollment evaluations, 129 terminated their pregnancy, 5 had a spontaneous abortion, 3 started highly active antiretroviral therapy, 2 died before inclusion, and 213 did not meet the inclusion criteria, including 184 who delivered after less than two weeks of zidovudine prophylaxis and were offered open-label nevirapine. Thirty-nine women randomly assigned to the placebo–placebo group were excluded from the analysis because they were provided with open-label nevirapine after the first interim analysis. There were 12 twin pregnancies and 14 fetal deaths (including 1 twin). Enrollment, loss to follow-up, pregnancy outcomes, and available end points for each treatment group in the study are summarized in Figure 1. End points were available for 97.6 percent of the 1766 live-born infants included in the analysis.
Figure 1. Treatment Assignment, Loss to Follow-up, and Outcome of Pregnancy.
Thirty-nine women assigned to the placebo–placebo group were excluded from the analyses, because they were given open-label nevirapine (NVP) after the first interim analysis. PL denotes placebo. Delivery of twins counted as only one delivery.
Characteristics of the Women, Deliveries, and Infants
Baseline characteristics of the women, the deliveries, and the newborns were similar across treatment groups (Table 1). Zidovudine prophylaxis was initiated at a median of 28.7 weeks' gestation. The median duration of zidovudine prophylaxis was 9.7 weeks. The median CD4 count among the women was 372 cells per cubic millimeter, and the median viral load was 4.0 log10 copies per milliliter. Fifty-six women had received antiretroviral therapy during a previous pregnancy.
Table 1. Characteristics of Mothers, Deliveries, and Infants in the Study, According to the Treatment Assignment.
Thirteen percent of the women delivered at or before 37 weeks' gestation. The median interval between the onset of labor and administration of a dose of the assigned treatment was 2.1 hours, and the median interval between receipt of the dose of the assigned treatment and delivery was 6.5 hours. Seventy women received two doses of the assigned treatment because of false or prolonged labor. Twenty percent of the deliveries were by cesarean section, of which 28 percent were performed before the onset of labor.
Nine percent of the infants weighed less than 2500 g (5.5 lb) at birth. The median interval between birth and administration of the first dose of the assigned treatment was 48.4 hours. Eight percent of infants received zidovudine for at least one month, because their mothers had received less than four weeks of zidovudine prophylaxis. There was one report of breast-feeding by the mother.
Analysis of Efficacy
First Interim Analysis
In women who delivered by May 2, 2002, the as-randomized transmission rate among those in the nevirapine–nevirapine group was 1.1 percent (95 percent confidence interval, 0.3 to 2.2), which was significantly lower than the rate of 6.3 percent (95 percent confidence interval, 3.8 to 8.9) in the placebo–placebo group (P<0.001). This finding confirmed the results of the first interim analysis, at which time the data and safety monitoring board recommended stopping the placebo–placebo regimen. In the nevirapine–placebo group, the rate was 2.1 percent (95 percent confidence interval, 0.6 to 3.7). Per-protocol transmission rates were very similar (Table 2).
Table 2. Kaplan–Meier Estimates of the Rate of Transmission of HIV at Six Months, According to Assigned and Actual Treatment.
Final Analysis
The final per-protocol transmission rate among those in the nevirapine–nevirapine group (1.9 percent; 95 percent confidence interval, 0.9 to 3.0) was not statistically inferior to the rate in the nevirapine–placebo group (2.8 percent; 95 percent confidence interval, 1.5 to 4.1). The upper limit of the one-sided confidence interval of the difference between these observed rates was 2.2, which was below the predefined 2.5 percent inferiority threshold. The rates in the as-randomized analysis were similar — 2.0 percent (95 percent confidence interval, 1.0 to 3.1) in the nevirapine–nevirapine group and 2.8 percent (95 percent confidence interval, 1.6 to 4.1) in the nevirapine–placebo group.
Other Analyses of Transmission
Among infants, the rate of testing positive for HIV within three days after birth in the placebo–placebo group (3.1 percent) was significantly higher than the rate among mothers in the study groups assigned to receive nevirapine before the interim analysis (0.4 percent) (P<0.001) (Table 2). In the final analysis, the overall rate of testing positive within three days after birth was 1.3 percent (95 percent confidence interval, 0.7 to 1.9). An analysis of the rate of death or HIV infection in infants before six months of age did not modify the conclusions of the efficacy analysis (Table 3).
Table 3. Estimates of the Rate of Transmission of HIV or Death at Six Months, According to Treatment Assignment.
Effect of Nevirapine According to Known Risk Factors for Transmission
Nevirapine had an effect according to the known risk factors for transmission, including viral load and CD4 count at baseline, gestational age at the initiation of zidovudine treatment, gestational duration, and the duration of zidovudine prophylaxis, although the effect often varied across subgroups (Table 4).2,5,17,18,19,20,21 In addition, the nevirapine–nevirapine regimen was somewhat superior to the nevirapine–placebo regimen in most subgroups.
Table 4. Effect of Nevirapine among Infants with Known Risk Factors for Transmission of HIV.
Maternal Safety
No severe rashes were observed. At 10 days, mild rashes were reported in 3.8 percent of mothers. Alanine aminotransferase levels were 36 to 42 IU per liter in 8.5 percent of mothers and 43 to 69 IU per liter in 0.9 percent (none of the mothers had levels above 70 IU per liter); there was no significant difference among the treatment groups. Rates of serious adverse events were similar across the groups (216 reports) and were related to pregnancy (59 percent), infection or HIV (26 percent), possibly zidovudine prophylaxis (anemia in 7 percent), possibly nevirapine (e.g., allergic reaction during therapy in one woman), and other conditions (7 percent).
Infant Safety
In infants, the rates of rash were 15.9 percent and 7.5 percent at 10 and 45 days of life, respectively, with no significant difference among treatment groups. Likewise, increases in alanine aminotransferase levels (36 to 69 IU per liter in 4.8 percent of infants and 260 IU per liter in one infant; none had levels above 350 IU per liter) and the presence of hyperbilirubinemia were not associated with exposure to nevirapine. The rates of serious adverse events were similar across treatment groups (598 reports) and were related to neonatal or obstetrical conditions (11 percent), to congenital abnormalities (6 percent), to any infection, including HIV (72 percent), possibly to zidovudine (anemia in 2 percent), and to other causes (9 percent). Prolonged neonatal icterus (1 percent) was classified as possibly related to nevirapine even in the placebo–placebo group, since classification was performed by persons blinded to treatment assignment. Nineteen infants died within six months after birth (Table 3).
Discussion
This study demonstrates the high efficacy of adding a single dose of nevirapine in the mother, with or without a dose in the infant, to oral zidovudine prophylaxis for the reduction of perinatal transmission of HIV. The observed reduction by 80 percent that led to the early interruption of enrollment in the placebo–placebo group was much higher than we had hypothesized when we designed the study. The transmission rate observed in the placebo–placebo group was consistent with rates observed in other studies in which zidovudine prophylaxis was initiated at the beginning of the last trimester of pregnancy,1,19,20,21 in particular in our previous trial, which was performed in a similar population.5
Although the efficacy of a single maternal dose of nevirapine was declared statistically noninferior to that of a maternal dose plus an infant dose with the use of our predetermined criteria (2.5 percent as the upper limit of the one-sided 95 percent confidence interval for the difference between the observed rates of transmission), in fact, two doses appeared consistently better than one throughout our analysis (Table 4). Moreover, because the nevirapine–placebo group had a somewhat lower median viral load at baseline than the nevirapine–nevirapine group, the difference in transmission rates might be an underestimate. On the basis of these findings and for additional logistic reasons, the ministry of public health in Thailand has recommended administration of nevirapine to both mother and infant, with a slight modification, that of giving nevirapine to the mother on arrival at the hospital and to the infant as soon as possible after birth.22
The percentage of infants with a positive HIV test within three days after birth was significantly lower in the nevirapine–nevirapine and the nevirapine–placebo groups than in the placebo–placebo group. This surprising finding suggests either that infants become HIV-positive earlier after infection than previously suspected or that perinatal nevirapine may delay the detection of HIV positivity or even abort infection completely when transmission occurs in the last few weeks of pregnancy.
Newborns in this study were formula-fed. This recommendation, made by the Ministry of Public Health in 1993, is strongly adhered to and has been shown to have no negative effect either on the health of the infant or on the promotion of breast-feeding among HIV-negative women.5,23,24
In the South African Intrapartum Nevirapine Trial, the rate of intrapartum transmission of HIV was three times as high when the mother had received nevirapine less than two hours before delivery as when the mother had received it earlier.25 Our study did not confirm this finding. However, since the administration of nevirapine to infants appears to be safe whatever the delay between birth and receipt of the drug7,25 and given that in clinical practice women may take their nevirapine dose later during labor and stay in the hospital for a shorter period, it may be preferable to administer the dose of nevirapine to the infant shortly after birth, with the first dose of zidovudine.
In the Pediatric AIDS Clinical Trials Group study known as PACTG 316, which evaluated the efficacy of perinatal nevirapine therapy added to standard antiretroviral therapy, no benefit of nevirapine was shown.26 The effect of nevirapine was not observed even in the subgroup of women who received only zidovudine in addition to nevirapine. This subgroup, however, may have been selected to receive zidovudine monotherapy because of the subjects' low viral load. In addition, 43 percent of the women in the group delivered by planned cesarean section. We also found that in the subgroup of women whose viral load at entry was less than 25,000, the transmission rate was very low in the placebo–placebo group (2 percent) and only marginally lower in the nevirapine groups. Although a subgroup analysis has the potential for bias, the effect of perinatal nevirapine was observed in most of the subgroups, a finding that suggests that there is no population for whom nevirapine prophylaxis would not be beneficial.
The discovery of drug-resistance mutations in a substantial proportion of the recipients of perinatal nevirapine27,28,29,30,31,32,33,34 has raised questions of a possible risk to mothers who subsequently receive nonnucleoside reverse-transcriptase inhibitors as therapy for their own disease. The resistance consequences among the participants of our study are described by Jourdain et al. elsewhere in this issue of the Journal.35 The maternal risk is juxtaposed with the clear benefit to the babies. The relative value of this or other interventions in any particular setting will depend on considerations of cost, medical logistics, and drug access, as well as on data from future studies of outcomes in women receiving both perinatal prophylaxis and antiretroviral therapy.36 Nevirapine at delivery added to third-trimester zidovudine results in very low transmission rates with minimal medical and financial burdens — rates that are very similar to those achieved with more complex, expensive, and potentially toxic multidrug maternal regimens.37,38,39,40,41,42,43
Supported by grants from the National Institutes of Health (5 R01 HD 39615); the Agence Nationale de Recherches sur le Sida (ANRS 12-08); the Ministry of Public Health, Thailand; the Institut de Recherche pour le Développement, France; the Institut National d'Etudes Démographiques, France; the Department of Technical and Economic Cooperation, Thailand; Boehringer Ingelheim; GlaxoSmithKline (which provided the zidovudine); and Roche Molecular Systems.
Dr. Jourdain reports having equity in Aventis.
We are indebted to S. Hammer and M. Tod for their contribution to the design of the study; to M.P. Kieffer for assistance in preparing the initial trial proposal; to the data and safety monitoring committee members V. Phoolcharoen, D. Costagliola, V. Vithayasai, A. Chaovavanich, for their advice and support throughout the trial; to the staff of the Perinatal HIV Prevention Trial, as follows: for administrative support: N. Chaiboonruang, Y. Thita, P. Pirom, D. Tiptepin, N. Pawadee, R. Klaijak, W. Ruangphornwisut; tracking and supplies: K. Taninard, P. Asongchaiwatanakul, S. Phonin; safety monitoring: S. Chalermpantmetagul, R. Peongjakta; sites monitoring: P. Sukrakanchana, A. Saiphueng, B. Yapuang, W. Pongchaisit, Y. Yensuang, B. Subhamitra, C. Kanabkaew, C. Puttimit, J. Wallapachai, N. Kruenual, N. Kongstan, N. Krapunpongsakul, R. Wongchai, S. Thammajitsagul, S. Jina, N. Seetapun; data management: R. Seubmongkolchai, A. Maleesatharn, J. Suewan, N. Kawee, S. Chailert, S. Tanasri, A. Wongja, K. Yoddee, M. Nuchniyom, K. Chaokasem, P. Supinya, J. Maneechan, K. Suebmongkolchai, S. Tansenee, A. Seubmongkolchai, C. Pholhinkong, C. Chimplee, K. Saopang, N. Karbkam, P. Chusut, P. Onnoy, S. Aumtong; and laboratory: T. Cressey, W. Khamduang, A. Preedam, A. Kummee, K. Janjaroen, M. Sumonwiriya, P. Chanthubthong, W. Boonprasit, J. Kamkorn, P. Tungyai, S. Kanthawong, S. Kunkeaw; to the Thai Ministry of Public Health for advice and assistance: Office of the Permanent Secretary, Department of Health, Department of Communicable Diseases Control, and Provincial Hospitals Division, especially, M. Teeratantikanont, P. Amornwichet, V. Chokevivat, T. Siraprapasiri, T Liewsaree, W. Liewsaree, P. Kantipong, C. Bowonwatanuwong, P. Leenasirimakul, A. Nilmanat, A. Chutanunta, O. Kamsao, W. Panitsuk, P. Pathipvanich; and Chiang Mai University: W. Sirirungsi, P. Lechanachai, U. Haesungcharern, A. Matanasaravoot; to the microbiology reference laboratories: R. Sutthent, P. Auewarakul (Mahidol University, Thailand); A.M. Comeau, J. Gerstel-Thompson (University of Massachusetts), F. Barin (Tours University, France); to the National Institute of Child Health and Human Development: A. Willoughby, B. Nugent, L. Mofenson; to Boehringer Ingelheim: E. Van Steenberge; to GlaxoSmithKline: S. Chinnapha, W. Rungsiyapornrat; to the following colleagues who contributed to this project in many critical ways: M. Essex, P. Matangkhasombut, A. Patterson, P. Casky, P. Jacob, J. Sullivan, C. Bellec, E. Kiley, J. Morcos, M.L. Bezencon, U. Thisyakorn, M. Mirochnich, A. Kuo, D. Bray, K. Young, A. Kuo, N. Briand, R. Balasubramanian; to all members of the hospital teams; and to the women and children who participated in this study.
* The coinvestigators and study coordinators of the Perinatal HIV Prevention Trial (Thailand) are listed in the Appendix.
Source Information
From Epidémiologie Clinique, Santé Maternelle et Infantile et Sida, Institut de Recherche pour le Développement, Paris (M.L., G.J., S.L., N.N.-G.-H.); the Department of Immunology and Infectious Diseases, Harvard School of Public Health, Boston (M.L., G.J., N.N.-G.-H.); Institut National d'Etudes Démographiques, Paris (S.L.); INSERM, Erm 0321, Paris (J.Y.M.); the Family Health Research Center, Mahidol University, Bangkok, Thailand (S. Koetsawang); the Ministry of Public Health, Bangkok, Thailand (S. Kanshana, V.T.); and Children's Hospital and the Department of Pediatrics, Harvard Medical School — both in Boston (K.M.).
Address reprint requests to Dr. Lallemant at the Perinatal HIV Prevention Trial — Institut de Recherche pour le Développement, Unité 54, 29/7-8 Samlan Rd., Soi 1 Prasing Muang, Chiang Mai 50200, Thailand, or at marc@phpt.org.
References
Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994;331:1173-1180.
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.
Mofenson LM, McIntyre JA. Advances and research directions in the prevention of mother-to-child HIV-1 transmission. Lancet 2000;355:2237-2244.
Kanshana S, Thewanda D, Teeraratkul A, et al. Implementing short-course zidovudine to reduce mother-infant HIV transmission in a large pilot program in Thailand. AIDS 2000;14:1617-1623.
Lallemant M, Jourdain G, Le Coeur S, et al. A trial of shortened zidovudine regimens to prevent mother-to-child transmission of human immunodeficiency virus type 1. N Engl J Med 2000;343:982-991.
Amornwichet P, Teeraratkul A, Simonds RJ, et al. Preventing mother-to-child HIV transmission: the first year of Thailand's national program. JAMA 2002;288:245-248.
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.
Ministry of Public Health. National guidelines for the clinical management of HIV infection in children and adults. Nonthaburi, Thailand: AIDS Division, Department of Communicable Disease Control, 2000.
1999 USPHS/IDSA guidelines for the prevention of opportunistic infections in persons infected with human immunodeficiency virus. MMWR Recomm Rep 1999;48:1-59, 61.
1995 Revised guidelines for prophylaxis against Pneumocystis carinii pneumonia for children infected with or perinatally exposed to human immunodeficiency virus. MMWR Recomm Rep 1995;44:1-11.
Revised recommendations for HIV screening of pregnant women. MMWR Recomm Rep 2001;50:63-85.
Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: John Wiley, 1981:161-5.
O'Brien PC, Fleming TR. A multiple testing procedure for clinical trials. Biometrics 1979;35:549-556.
Lan KKG, DeMets DL. Changing frequency of interim analysis in sequential monitoring. Biometrics 1989;45:1017-1020.
Ellenberg JH. Intention-to-treat analysis. In: Redmond CK, Colton T, eds. Biostatistics in clinical trials. Chichester, England: John Wiley, 2001:245-9.
Balasubramanian R, Lagakos SW. Estimation of the timing of perinatal transmission of HIV. Biometrics 2001;57:1048-1058.
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.
Shapiro DE, Sperling RS, Coombs RW. Effect of zidovudine on perinatal HIV-1 transmission and maternal viral load. Lancet 1999;354:156-156.
Wade NA, Birkhead GS, Warren BL, et al. Abbreviated regimens of zidovudine prophylaxis and perinatal transmission of the human immunodeficiency virus. N Engl J Med 1998;339:1409-1414.
Matheson PB, Abrams EJ, Thomas PA, et al. Efficacy of antenatal zidovudine in reducing perinatal transmission of human immunodeficiency virus type 1. J Infect Dis 1995;172:353-358.
Simpson BJ, Shapiro ED, Andiman WA. Reduction in the risk of vertical transmission of HIV-1 associated with treatment of pregnant women with orally administered zidovudine alone. J Acquir Immune Defic Syndr Hum Retrovirol 1997;14:145-152.
Guidelines for the management of prevention of mother to child transmission of HIV in Thailand: policy of prevention of mother to child transmission of HIV in Thailand. Nonthaburi, Thailand: Ministry of Public Health, Department of Health, December 2003.
Kanshana S, Simonds RJ. National program for preventing mother-child HIV transmission in Thailand: successful implementation and lessons learned. AIDS 2002;16:953-959.
Talawat S, Dore GJ, Le Coeur S, Lallemant M. Infant feeding practices and attitudes among women with HIV infection in northern Thailand. AIDS Care 2002;14:625-631.
Moodley D, Moodley J, Coovadia H, et al. A multicenter randomized controlled trial of nevirapine versus a combination of zidovudine and lamivudine to reduce intrapartum and early postpartum mother-to-child transmission of human immunodeficiency virus type 1. J Infect Dis 2003;187:725-735.
Dorenbaum A, Cunningham CK, Gelber RD, et al. Two-dose intrapartum/newborn nevirapine and standard antiretroviral therapy to reduce perinatal HIV transmission: a randomized trial. JAMA 2002;288:189-198.
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.
Eshleman SH, Mracna M, Guay LA, et al. Selection and fading of resistance mutations in women and infants receiving nevirapine to prevent HIV-1 vertical transmission (HIVNET 012). AIDS 2001;15:1951-1957.
Cunningham CK, Chaix ML, Rekacewicz C, et al. Development of resistance mutations in women receiving standard antiretroviral therapy who received intrapartum nevirapine to prevent perinatal human immunodeficiency virus type 1 transmission: a substudy of Pediatric AIDS Clinical Trials Group protocol 316. J Infect Dis 2002;186:181-188.
Morris L, Pillay C, Chezzi C, et al. Low frequency of the V106M mutation among HIV-1 subtype C-infected pregnant women exposed to nevirapine. AIDS 2003;17:1698-1700.
Kantor R, Lee E, Johnston E, et al. Rapid flux in non-nucleoside reverse transcriptase inhibitor resistance mutations among subtype C HIV-1-infected women after single dose nevirapine. Antiviral Ther 2003;8:S85. abstract.
Chaix ML, Montcho C, Ekouevi DK, et al. Genotypic resistance analysis in women who received intrapartum nevirapine associated to a short course of zidovudine to prevent perinatal HIV-1 transmission: the Ditrame Plus ANRS 1201/02 Study, Abidjan, C?te d'Ivoire. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:657. abstract.
Chalermchokcharoenkit A, Asavapiriyanont S, Teeraratkul A, et al. Combination short-course zidovudine plus 2-dose nevirapine for prevention of mother-to-child transmission: safety, tolerance, transmission, and resistance results. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:96. abstract.
Martinson N, Morris L, Gray G, et al. HIV resistance and transmission following single-dose nevirapine in a PMTCT cohort. In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:38. abstract.
Jourdain G, Ngo-Giang-Huong N, Le Coeur S, et al. Intrapartum exposure to nevirapine and subsequent maternal responses to nevirapine-based antiretroviral therapy. N Engl J Med 2004;351:229-240.
World Health Organization. Antiretroviral drugs and the prevention of mother-to-child transmission of HIV infection in resource-limited settings: expert consultation, Geneva, 5-6 February 2004: a summary of main points from the meeting. (Accessed June 18, 2004, at http://www.who.int/3by5/arv_pmtct/en/.)
Bucceri AM, Somigliana E, Matrone R, et al. Combination antiretroviral therapy in 100 HIV-1-infected pregnant women. Hum Reprod 2002;17:436-441.
Cooper ER, Charurat M, Mofenson L, et al. Combination antiretroviral strategies for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr 2002;29:484-494.
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.
Wimalasundera RC, Larbalestier N, Smith JH, et al. Pre-eclampsia, antiretroviral therapy, and immune reconstitution. Lancet 2002;360:1152-1154.
Shapiro D, Tuomala R, Pollack H, et al. Mother-to-child HIV transmission risk according to antiretroviral therapy, mode of delivery, and viral load in 2895 U.S. women (PACTG 367). In: Proceedings of the 11th Conference on Retroviruses and Opportunistic Infections, San Francisco, February 8–11, 2004:99. abstract.
Shepard KV. Important new safety information: clarification of risk factors for severe, life-threatening and fatal hepatotoxicity with VIRAMUNE (nevirapine). Ridgefield, Conn.: Boehringer Ingelheim Pharmaceuticals, February 2004.
Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Washington, D.C.: Department of Health and Human Services, March 23, 2004.(Marc Lallemant, M.D., Gon)