Original Articles Indian Pediatrics 2001; 38: 132-138 |
|||||||||||||||||||||||||||||||||||||||||||||||||||
Strategy for preventing vertical transmisssion of HIV : Bombay Experience |
|||||||||||||||||||||||||||||||||||||||||||||||||||
Key words: HIV, Vertical transmission, Zidovudine. SEVERAL studies have demonstrated that the rate of mother to child transmission of HIV ranges from 15 to 40%(1-3). The need for early diagnosis and treatment of perinatally infected infants is pressing as pediatric HIV infection remains a serious disease. The AIDS Clinic Trial Group (ACTG-076) showed that in the absence of breastfeeding, 500 mg oral AZT per day to HIV infected pregnant women started between 14-34 weeks gestation followed by intravenous AZT during labor and oral chemoprophylaxis with AZT to their infants, lowered the risk of perinatal transmission by approximately two-thirds(4). This regimen is the standard practice followed in Western countries but its complexity and cost makes it difficult to implement in countries like India. Since January 1993, we have tested more than 82,590 pregnant housewives and have noted an average seroprevalence of 1.44% in this low risk population (Table I). At our institute, we have noted a mother to child transmission rate of approximately 24% in those who had not received any form of peri-natal intervention. In view of the increasing seroprevalence of HIV that we noted, a significant rate of vertical transmission and the financial constraints of our population, we devised a protocol (Wadia Model) in an effort to reduce vertical transmission. In the present manuscript we discuss the benefits accruing from this strategy.
All pregnant women attending the ante-natal clinic received group orientation and education about HIV in their vernacular language. They were then offered the option of HIV testing and a blood sample taken after a verbal informed consent. Blood samples of 5 women were pooled together and tested as one pool as a cost saving strategy(5). A pool that tested positive was then subjected to individual samples testing using a third generation ELISA kit (MUREX, Welcome Diagnostic, UK). Results of tests were explained in total confidentiality to the women and her spouse when they returned for the next clinic visit. Rationale and importance of retesting was explained by a social counselor in simple language and a second blood sample obtained after an informed consent from any woman who had reacted positive to the first ELISA test. This was done using a different ELISA kit (DETECT-HIVTM)(v.2), Biochem Immuno systems Inc., Montreal, Quebec, Canada H3M 3A2). Women who tested posi-tive on the second retest were then explained, along with their spouses, the intervention strategy that we planned (post-test counseling). It was explained to each couple that the intervention could have started soon after the first trimester but in view of the high cost and late antenatal registry of case, it was being initiated at 32 weeks gestation. Although, all HIV positive women were offered interven-tion, only those who could afford the inter-vention and who agreed to deliver at our hospital were included in the study.
After an informed consent, 400 mg AZT was initiated daily in four divided doses on an outpatient basis and the women followed up every two weeks. At every visit, each woman was interviewed for drug compliance and tolerance. All women had a baseline hemo-gram before starting therapy and a repeat hemogram and liver function test (LFT) at the end of 4 weeks of AZT therapy. All women were scheduled to deliver by elective Caesa-rian section (CS) at 38 weeks [as calculated by last menstrual period (LMP) and ultrasound]. Delivery characteristics were recorded. No postpartum AZT was given to these women but their babies received 8 mg per kilogram per day of AZT powder for 6 weeks. Breast- feeding was withheld. Infants were scheduled for follow up at 1,3,6,9,12,15 and 18 months. Complete blood count (CBC) and LFT was done at the end of 6 weeks of AZT therapy and HIV ELISA tests done at 9 months and between 15-18 months using two different ELISA kits (as mentioned earlier). A definitive diagnosis of HIV infect-ivity was made if the infant tested positive on atleast 2 occasions. Infants not followed up beyond 18 months were excluded from the analysis. All ELISA tests were done at the laboratory of the hospital free of cost and in accordance with the manufacturer’s instruc-tions. In view of the cost, none of the women or infants had routine viral load or CD4 cell counts evaluated.
Over a four and half year period from July ’96, 372 pregnant women were identified as being HIV infected. Of these, 229 did not receive perinatal intervention either because of non-affordability or because they had registered beyond 32 weeks of gestation. All the 143 women enrolled in this study were below 30 years of age, all were married, had no associated high-risk behavior and no clinical evidence of immune suppression at the time of enrollment. However, 4 women were diagnosed to have pulmonary tuberculosis. Baseline hemogram in all was within acceptable limits. The minimum duration of antepartum treatment was 6 weeks and there was no loss to follow up. There was no interruption of treatment with the exception of one woman who developed an adverse reaction to AZT in the form of paraesthesias of the extremities, necessitating discontinua-tion of the drug. This woman was excluded from the data analysis. Of the 143 women, 36 delivered vaginally before the fixed date for Caesarian section, while the remaining 107 delivered by elec- tive Caesarian section prior to rupture of membranes at a mean gestation of 38.5 weeks. Of the 107 women who underwent Caesarian section, one developed abdominal wound dehisence which was appropriately managed. In view of this complication, a CD4 count was done to evaluate her immune system, which was within normal limits. The mean birth weight of infants was 2,650 +150 g. Three infants were noted to have congenital defects: a small ventricular septal defect, unilateral polydactyly and cleft lip with cleft palate. The number of congenital malformations comparable to the general population of infants at our center. All infants were started on AZT powder soon after birth, tolerated the drug well and had normal CBC and serum aminotransferase at the end of 6 weeks of therapy. Of the 107 infants delivered by Caesarian section, 22 were lost to follow-up and 2 expired in the first year of life without confirmation of their infectivity while 15 were less than 18 months of age at the time of this analysis. In the group of 68 infants who fulfilled all 4 study criteria and a follow-up of 18 months, only 4 tested positive for HIV by ELISA (transmission rate of 5.8%). Both these infants were confirmed to be HIV-1 infected on Western blot analysis (Immunoblot Assay, Ranbaxy Specialty Laboratory, Mumbai). Of the 229 women-infant pairs who did not receive perinatal intervention, 55 infants followed up to 15-18 months, were found to be infected (transmission rate of 24%; p = 0.0005 by Fischer exact test).
Perinatal transmission can occur in-utero, during labor and delivery, or through breast- milk(1-3) although current evidence shows that most of the transmission occurs peri-partum(6-10). There is a need to develop low cost strategies to identify cases and prevent transmission from the infected mother to her infant. AZT a nucleoside reverse transcriptase inhibitor decreases the maternal viral load and hence reduces the risk of fetal infection(4), although some infants become infected either because transmission has occurred before initiation of therapy or due to insufficient suppression of maternal viral replication. The optimal time for initiating antepartum AZT has still not been defined. A cost-effective method to study mass seroprevalence is screening for HIV antibody in pooled serum samples(11-14). Pooling methods have been assessed in various studies and been found to be as sensitive and specific as individual sample testing and yield a cost-saving of approximately 55% in a low prevalence population(15). WHO recom-mends a maximum pool of five samples(16) where the seroprevalence is less than 2%, as noted in our population. The crux of our intervention consisted of the use of a shorter course with a lower dose of AZT (as compared to the original ACTG 076 protocol), coupled with elective Caesarian section, oral AZT powder to the infant (we used oral AZT powder due to unavailability of oral AZT suspension at the time of this study) and complete omission of breast milk. The dose of AZT administered to the woman was 400 mg per day. We believe that this dose not only reduced the cost by one-fifth (the current cost of therapy is approximately Rs. 2100) but would be adequate for most of our women who have a pre-pregnancy weight of 40 to 45 kg. Studies conducted in other developing countries(17) which used an even shorter regimen of AZT starting from 36 weeks of gestation showed that the risk of vertical transmission was lowered by about 50%. A substantial proportion of transmission occurs during labor and delivery due to contamination from cervicovaginal secretions and blood and may be circumvented by elective Caesarian section(18). A reduction in the HIV transmission rate was noted after elective Caesarian section in the European Collaborative study(19). A Swiss study(20) found that combination of AZT therapy to pregnant women, and delivery by Caesarian section totally eliminated vertical transmission. Even if a reduction in the risk of transmission is confirmed, it still remains controversial whether Caesarian section should be under-taken solely for the purpose of preventing HIV transmission(2). Few developing countries have the capacity to perform safe operatives deliveries on such a large scale. At our institute with good laboratory and operative facilities, delivery by this mode as a potential method of reduction of HIV trans-mission was justifiable. We were however unable to decipher the extent to which elective Caesarian section alone reduced vertical transmission. Postpartum transmission of HIV through breast milk was first reported in 1985(21). Van de Perre et al. postulated that transmission of HIV through breast milk could be favored by the presence of infected cells or deficiency of anti-infective substances in breast milk(22). Two large European studies have shown breastfeeding to be associated with a doubling of the risk of transmission(23), and a metaanalysis of several prospective studies indicates that risk of transmission attributable to breastfeeding varies between 7% and 22%(23). Breastfeeding is thus a significant and preventable mode of perinatal trans-mission. Therefore, although breast milk is the best infant food, in the case of a HIV infected mother in may be beneficial to offer breast milk substitutes to her infant to reduce the small, albeit definite, risk of postnatal transmission. Within the Indian context also, it may be worthwhile to offer a safe alternative to breastfeeding in order to contain postnatal transmission. All mothers in our study were counseled regarding this and the final decision was left to them. Serological diagnosis in infants is problematic because maternal anti-HIV IgG antibodies cross the placenta into fetal circulation. As a result, ELISA and Western blot tests may be positive upto 15 to 18 months of life(24). Routine ELISA or Western blot testing therefore does not distinguish infants under 15 months of age who are infected or not infected(25). As the median time to sero-conversion in non-infected is 9 months and some antibodies persist upto 15-18 months, infants at our center were tested twice once at 9 months and then at 15-18 months by the ELISA technique [having a sensitivity of about 99.5%(26)] to confirm infective status. This study did not aim to compare the transmission rate between those who delivered vaginally versus those who delivered by Caesarian section. Limitations of the study include: (a) A high drop out rate which restricts the precision of our estimates. The high drop out rate of infants could have been due to the stigma associated with the disease and the fact that a proportion of our patients lived away from the city and had come to Bombay only for safe delivery. Hence, our findings need to be substantiated by larger studies using similar models; (b) The viral load and CD4 counts in these women could not be estimated due to financial constraints. We therefore could not analyze the implication of immune status of these women to the rate of transmission; and (c) Multivariate analysis was not done to assess the beneficial effect of any one selected strategy for its protective effect. Despite these shortcomings, the striking success in reduction of perinatal transmission to 5.8% has significant potential public health implications. Our data reinforces recommendations from other public health services(17,27,28) that AZT should be offered to HIV infected pregnant women. Nevirapine in HIVNET 012 has been found to reduce the risk of HIV transmission by 47% in a breast feeding population. In a recently published editorial, Laurence(29) reviewed the strategy for prevention of vertical transmission and mentioned that potential problems with the use of Nevirapine include lack of time for volun-tary counseling (when the woman comes to the hospital for the first time in active labor), lack of facilities for rapid testing at all centers and unknown potential long term toxicities to the infant. In our setting where this drug was not available, the use of AZT was still beneficial, more practical and have a high success rate. Contributors: RHM co-ordinated the study, its design and interpretation; he will act as the guarantor for the study. KD was involved in designing the study and was responsible for the direct care of antenatal cases, their follow up and safe delivery. ISG helped in initiation of study. JSK was responsible for counsel-ing at all stages of each of the study participants. RUB helped in data collection, analysis and drafting the paper. SRM and JSO helped in data collection. SC helped in data collection and analysis. Competing interests:
None stated.
|