|Year : 2012 | Volume
| Issue : 4 | Page : 344-348
|Approaches to effectively documenting prior single dose nevirapine exposure among women in Africa
Francesca M Conradie1, Yu Zheng2, Mina C Hosseinipour3, John W Mellors4, Fredrick K Sawe5, Susan H Eshleman6, Lerato Mohapi7, Tsungai Chipato8, Aida Asmelash9, Elizabeth M Stringer10, Abraham M Siika11, Sandra D Rwambuya12, P Ramnarain13, Evelyn P Hogg14, Christina L Blanchard-Horan15, Judith Currier16, Michael D Hughes17, James McIntyre18, Kityo Cissy19, Ann Walawander20, Apsara Nair20, Shahin Lockman21
1 Deputy Director, Clinical HIV Research Unit (CHRU), University of the Witwatersrand, Johannesburg, South Africa
2 Center for Biostatistics in AIDS Research, Harvard School of Public Health, Boston, M.A., USA
3 Clinical Director, University of North Carolina Project, Lilongwe, Malawi
4 Division of Infectious Diseases, University of Pittsburgh Medical Centre, P.A, USA
5 Deputy Director, HIV Care and Research, Kenya Medical Research Institute/Walter Reed Project, Kericho, Kenya
6 Virologist, Johns Hopkins University School of Medicine, Baltimore, MD, USA
7 CRS Leader, Soweto ACTG CRS, Perinatal HIV Research Unit (PHRU), University of the Witwatersrand, Johannesburg, South Africa
8 University of Zimbabwe, College of Health Sciences, Harare, Zimbabwe
9 Botswana-Harvard AIDS Institute, Gaborone, Botswana
10 Associate Professor and Site P.I., University of Alabama at Birmingham/CIDRZ, Lusaka, Zambia
11 Department of Medicine, Moi University School of Medicine, Eldoret, Kenya
12 Study Coordinator, Joint Clinical Research Centre, Kampala, Uganda
13 University of KwaZulu Natal, South Africa
14 Clinical Trials Specialist, Social and Scientific Systems Inc, Silver Spring, MD, USA
15 ACTG International Site Liaison, Social and Scientific Systems Inc, Silver Spring, MD, USA
16 Professor of Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA
17 Center for Biostatistics in AIDS Research, Harvard School of Public Health, Boston, MA, USA
18 Executive Director, Perinatal HIV Research Unit (PHRU) and Associate Professor, Department of Obstetrics and Gynaecology, University of the Witwatersrand, Johannesburg, South Africa
19 JCRC Kampala, Uganda
20 Data Manager, Frontier Science and Technology Research Foundation, Amherst, NY 14226, USA
21 Assistant Professor in Medicine, Brigham and Women's Hospital and Harvard School of Public Health, Boston USA; and Botswana Harvard School of Public Health AIDS Initiative Partnership, Gaborone, Botswana
Click here for correspondence address and email
|Date of Web Publication||8-Oct-2012|
| Abstract|| |
Background: Single dose nevirapine (sdNVP) is widely used in resource-limited settings for the prevention of mother to child transmission of HIV, but can result in NVP resistance that negatively impacts the subsequent efficacy of maternal antiretroviral therapy (ART). It is important to determine prior sdNVP exposure status to help guide treatment decisions, but systematic data on approaches to documenting previous sdNVP ingestion are lacking. Aim: With the growing body of evidence of the effects of sdNVP exposure on subsequent choices of ART, we aim to highlight some of the practical challenges that exist in documenting prior sdNVP exposure or lack thereof. Materials and Methods: ACTG A5208 Optimized Combination Therapy after Nevirapine Exposure (OCTANE) is a randomized treatment trial of protease inhibitor vs. NVP-based ART that enrolled 745 HIV-infected women in 7 African countries. Documentation of previous exposure to sdNVP (or lack thereof) was collected prospectively and intensively, as were locally-available sources of such data. Results: All 243 women who were exposed to sdNVP recalled having taken sdNVP; written documentation of sdNVP exposure was found for 73% and an additional 20% identified having ingested a NVP tablet when the tablet was shown to them. Among 502 women not exposed to sdNVP, only 10 (2%) had written documentation of lack of sdNVP exposure. NVP resistance was detected in 33 (13.8%) of sdNVP-exposed and 1 of non-exposed women. Conclusion: Maternal self-report of prior sdNVP exposure was corroborated by supporting evidence in the majority of women participating in the trial.
Keywords: Pregnancy, prevention of mother to child transmission, single-dose nevirapine
|How to cite this article:|
Conradie FM, Zheng Y, Hosseinipour MC, Mellors JW, Sawe FK, Eshleman SH, Mohapi L, Chipato T, Asmelash A, Stringer EM, Siika AM, Rwambuya SD, Ramnarain P, Hogg EP, Blanchard-Horan CL, Currier J, Hughes MD, McIntyre J, Cissy K, Walawander A, Nair A, Lockman S. Approaches to effectively documenting prior single dose nevirapine exposure among women in Africa. Ann Trop Med Public Health 2012;5:344-8
|How to cite this URL:|
Conradie FM, Zheng Y, Hosseinipour MC, Mellors JW, Sawe FK, Eshleman SH, Mohapi L, Chipato T, Asmelash A, Stringer EM, Siika AM, Rwambuya SD, Ramnarain P, Hogg EP, Blanchard-Horan CL, Currier J, Hughes MD, McIntyre J, Cissy K, Walawander A, Nair A, Lockman S. Approaches to effectively documenting prior single dose nevirapine exposure among women in Africa. Ann Trop Med Public Health [serial online] 2012 [cited 2018 May 21];5:344-8. Available from: http://www.atmph.org/text.asp?2012/5/4/344/102051
| Introduction|| |
The HIVNET 012 trial demonstrated a 41% reduction in mother to child transmission (MTCT) of HIV-1 with the use of a single dose of nevirapine (sdNVP) to mother and neonate.  As a result of this trial; sdNVP was adopted in many resource-poor countries to prevent transmission of HIV-1. However, NVP-resistant virus is detected in up to 75% of women and 87% of infants following sdNVP exposure. ,, Although this resistance "fades" from detection over time, it has a negative impact upon the subsequent efficacy of treatment regimens that include NVP or other non-nucleoside reverse transcriptase inhibitors (NNRTIs), particularly when treatment is initiated within 1 to 2 years after exposure to sdNVP. , Women who took sdNVP within the previous 2 years should initiate a protease-inhibitor-containing (rather than NNRTI-containing) regimen when possible. Therefore, it is important to be able to identify women and infants who have previously taken sdNVP when subsequently initiating antiretroviral therapy.
In MTCT prevention programs, sdNVP is generally given to HIV-positive mothers to self-administer at the onset of labor. The availability of documentation of provision (and ultimately of ingestion) of sdNVP months or years after it was provided may be limited in many settings. When initiating ART, it is, therefore, important to be able to determine with confidence whether any prior sdNVP exposure has occurred.
In this report, we describe how we approached the ascertainment of prior sdNVP exposure among women enrolling in a randomized clinical trial (ACTG A5208/OCTANE) undertaken in 7 countries in southern and eastern Africa. This trial comprises 2 concurrent randomized treatment open-label ART trials. Trial 1 was conducted among 243 women who previously ingested sdNVP and Trial 2 among 502 women without prior sdNVP exposure (target enrolment 240 and 500, respectively).
Our aim is to highlight some of the practical challenges that exist in ascertaining such documentation and mechanisms that might be used to overcome them.
| Materials and Methods|| |
Both trials in OCTANE compared the virologic response to non-nucleoside reverse transcriptase inhibitor (NNRTI)-based versus protease inhibitor (PI)-based antiretroviral treatment (ART) in HIV-infected treatment-naοve women with CD4 < 200 cells/mm 3 .  The primary study end point was time to virologic failure or death. Virologic failure was defined as a confirmed plasma HIV-1 RNA level less than 1 log 10 copy per milliliter below baseline at 12 weeks after treatment initiation or a confirmed HIV-1 RNA level that was 400 log 10 copies or more per milliliter 24 weeks or more after treatment was begun.
Trial 1 evaluated the superiority of PI-based ART over NNRTI-based ART in 243 women who reported having taken sdNVP (6 or more months previously) for the prevention of MTCT of HIV. Trial 2 evaluated the equivalence of PI- and NNRTI-based ART in 502 women with no prior sdNVP exposure. Women were enrolled at 10 sites in 7 African countries: 3 sites in South Africa (2 in Johannesburg, 1 in Durban); 2 sites in Kenya (Kericho and Eldoret); and Gaborone, Botswana (including a local subunit); Harare, Zimbabwe; Lusaka, Zambia; Kampala, Uganda and Lilongwe, Malawi. Women were enrolled to OCTANE between November 2005 and July 2008, and completed study treatment in August 2009. Sites in Johannesburg, Soweto, Durban, Harare, Lusaka, Lilongwe, and Kampala are in urban areas. Sites in Kericho and Eldoret serve more of a rural population, and the Botswana site enrolled at 1 urban and 1 rural location.
We attempted to prospectively and intensively collect the best available information on previous exposure to sdNVP (or lack thereof) for study participants, as well as locally-available sources of such data. At study entry, detailed information related to all prior exposures to sdNVP and the level/type of documentation related to each exposure was collected for each participant.
Each participating site completed a structured written survey regarding local practices related to sdNVP documentation in March 2007. The survey included questions on the location of deliveries, availability, and type of PMTCT services and dates of MTCT prevention roll-out. Information on timing of initiation of MTCT prevention regimens in each region was gathered. Open-ended questions also queried local record keeping and documentation practices related to MTCT prevention provision and receipt by patients.
Genotyping was not available at many sites, baseline genotyping is not standard of care at any site, and it was not deemed to be safe for participants to wait for return of the genotype result prior to treatment assignment, given advanced HIV disease. Viroseq TM genotype (Celera Diagnostics, Alameda, California) was performed on samples that were obtained from women just prior to initiation of antiretroviral therapy in the study; all Trial 1 participants and a random sample of 25% of Trial 2 participants and an ad hoc specimen taken from early enrollees. Baseline samples were stored and genotyped retrospectively.
| Results|| |
Nature of documentation of sdNVP exposure (or lack thereof)
243 women were enrolled and randomized in Trial 1 (sdNVP-exposed group), 241 of whom started study treatment. All Trial 1 participants recalled having previously taken sdNVP, and 93% of Trial 1 participants had some additional supporting evidence for prior sdNVP exposure.
[Table 1] presents the types of evidence supporting prior sdNVP exposure among the 243 women in Trial 1; if a participant had more than 1 type of evidence supporting prior sdNVP (in addition to current self-report), then only the most rigorous type of evidence is counted for that participant. 64 (26%) women had 2 forms of evidence supporting prior sdNVP ingestion in addition to participant recall, and another 21 (9%) had 3 additional forms of evidence. Written documentation that sdNVP had been provided or taken was available for 73% of women [Table 1].
|Table 1: Overall levels of evidence for sdNVP ingestion for Trial 1 (n = 243)|
Click here to view
The single most frequent type of supporting evidence was written documentation of the participant having been prescribed or given a tablet of NVP during at least 1 previous pregnancy (n = 176, 72%). Current recall was the sole supporting evidence of prior sdNVP exposure in only 16 (7%) women, with an additional 50 (21%) of women confirming prior sdNVP ingestion by correctly identifying a NVP tablet in the study clinic.
There was a trend toward a shorter interval since most the recent sdNVP exposure and the availability of written documentation supporting sdNVP receipt (median time since last sdNVP exposure and study enrolment was 16 months among women with written documentation vs. 20 months for women without documentation, P = 0.07).
502 women were randomized to Trial 2 (no prior sdNVP). 1 woman who reported prior sdNVP exposure (with written evidence thereof) was incorrectly assigned to Trial 2. [Table 2] presents the evidence supporting lack of sdNVP exposure. 208 (41.8%) women either had never been pregnant or have had their last pregnancy before 1 st Jan 1998. The remaining 293 (58.3%) have had a pregnancy since 1 st Jan 1998. One third of this group (169 or 33.7%) had been diagnosed with HIV infection subsequent to their last pregnancy. 64 (12.8%) have had their last pregnancy between 1 st Jan 1998 and the introduction of sdNVP in their region. 30 (6.0%), although they have had a pregnancy subsequent to the introduction of sdNVP, stated they had not been given it. 20, (4.0%) when shown the NVP tablet, denied having been given it at the time of their pregnancies and deliveries.
|Table 2: Overall levels of evidence for lack of prior sdNVP ingestion for Trial 2 (n = 501)|
Click here to view
Drug resistance and sdNVP exposure status
Among the 239 (98%) women in Trial 1 with baseline genotype results available, 33 (13.8%) had NVP resistance mutations detected [Table 3].
|Table 3: Presence of NNRTI resistance mutations among participants in Trials 1 and 2|
Click here to view
Baseline resistance testing was undertaken on specimens from 41 women enrolling early in Trial 2, and an additional random sample of 126 (25%) of Trial 2 participants stratified by sites (7 of the 126 samples failed to genotype test).
The K103N mutation was detected in 1 (0.8%) Trial 2 participant of the 160 specimens tested. This participant had been pregnant 3 times, in 2000, 2002, and 2006. Except for the pregnancy in 2000, which was prior to the use of sdNVP in the local health care facility, there was only patient's recall for having not taken sdNVP during the other 2 pregnancies. The K103N mutation was the most common in the sdNVP exposed group found in 23 (9.6%) women. G190A, V181 C/I, and V108I were found alone or in combination in the remainder
Reported site practices related to delivery and documentation of MTCT prevention provision
Most sites reported that the majority of women in their area delivered in a hospital or clinic although there was some variability across sites. 68 (28%) of women delivered where less than 50% of the deliveries occurred in hospitals or clinic in Lilongwe, Kericho, and Eldoret. Deliveries at a health facility were likely to be registered at that site in a post-natal log. In 9 out of 10 sites, information on dispensing and/or maternal self-report of ingestion of intrapartum antiretroviral drugs was supposed to be documented in maternity logbooks. However, in busy labor wards, this information was often not documented thoroughly.
Patient-retained records for obstetrics and child health were common in the countries that enrolled this trial, in part because of challenges with record filing at health facilities. There were usually no other notes available, and antenatal care and delivery often occurred at different sites. Many women were said to have retained these cards after delivery, and these proved a helpful source of information regarding previous pregnancies. However, the proportion was not collected.
Once the rollout of sdNVP was started, at all sites, women were given or prescribed sdNVP in the antenatal period, to keep for self-administration at the onset of labor; sdNVP was generally not first prescribed and/or dispensed at presentation in the labor ward unless women were first diagnosed with HIV during labor. The only site reporting regular testing during labor was Lusaka; this practice has been recommended in most guidelines subsequently. At one site, women were also given a pre-filled syringe for the baby NVP dose. This "pack" of tablet plus syringe may be easier to remember. Dispensing or prescription of the drug was usually well-documented in the antenatal clinic, but the actual taking of the dose (either by maternal self-report at the time, or witnessed by health care workers during delivery) was often not documented in writing.
In most of the countries, there were no population registers that could be inspected to determine whether a participant had delivered a baby between 1998 and screening for the trial. Only in Lusaka, there were district clinic notes regarding prior pregnancies, which could be cross referenced with participant self-report.
| Discussion|| |
We found that corroborating evidence for sdNVP exposure (or lack thereof) could be found for the majority of women, and that documenting previous sdNVP exposure among women in 10 sites in Africa was generally challenging but feasible. Study staff needed to exert considerable effort to try to document (or exclude) prior sdNVP exposure -information that was key to the validity of the A5208/OCTANE trial.
Sites that participated in the A5208/OCTANE had dedicated staff for the trial. It is very unlikely that in a resource-limited setting this could be mirrored. However, we believed that the lessons learnt in the exercise can be of value in establishing sdNVP exposure.
Ultimately, written documentation supporting prior sdNVP ingestion could be found for 73% of women participating in Trial 1 (sdNVP-exposed). Correct classification of prior sdNVP exposure is supported by the presence of NNRTI resistance in 14% of women in Trial 1 but only in 0.8% in Trial 2. It is also supported by trial findings related to the increasing efficacy of NVP-based ART with longer time since last sdNVP exposure. There are definite limitations to the standard genotype. Mutations may have faded over time. However, that there was still detectable resistance in so many cases, adds to the weight of the evidence the sdNVP exposure or lack thereof was accurate.
Further evidence supporting accurate documentation of sdNVP is in the striking difference in the findings of Trial 1 and Trial 2. In Trial 1, where women were exposed to sdNVP, the primary modified intention-to treat analysis 42 women reached an end point: 32 (26%) in the nevirapine group and 10 (8%) in the ritonavir-boosted lopinavir group (HR 3.6; adjusted 95% CI, 1.7 to 7.5; adjusted P = 0.001).  These numbers included 28 women (23%) and 9 women (8%), respectively who had virological failure and 4 women (3%) and 1 woman (1%) who died without virological failure. However, in Trial 2, 42 (17%) women in the NVP arm and 50 (20%) in the LPV/r arm reached the primary endpoint (HR 0.85; 95% CI 0.56 - 1.29). When the composite endpoint was further analyzed, 37 (15%) vs. 43(17) % experienced virological failure, and 2% vs. 3% died in the NVP and LPV/r arms, respectively. 
This study has limitations. Although the presence of written documentation among 3 quarters of exposed women, we cannot know with certainty that all of the Trial 1 participants had indeed ingested sdNVP. Stringer et al. demonstrated that only 68% of HIV-infected women who were given sdNVP antenatally in a city-wide PMTCT program in Lusaka had measurable concentrations of NVP in the cord blood at delivery. This means that 32% of women in this study may have been non-adherent.  Conversely, much of the available documentation related to absence of prior sdNVP receipt among Trial 2 participants rested upon the presence and dates of prior pregnancies, as well as upon the date of HIV diagnosis. We relied largely upon participant recall of these events and dates. Pregnancy and HIV diagnosis are important events that individuals are likely to recall with some accuracy, and participants did not have an incentive to report prior sdNVP receipt (or lack thereof) to study staff. Irrespective of whether a woman had received sdNVP, provided she fulfilled the other criteria for the trial, she would be able to participate. Nevertheless, written documentation was infrequently available to support lack of prior sdNVP exposure.
A further limitation is that the study sites were "handpicked." They were placed in relatively well-resourced areas. 173 (72%) of women delivered their infants in regions where over 85% of deliveries were conducted in a health care facility. This is higher than in many other regions and may thus limit the generalizability.
Patient retained records have been used with some success. Since the concept of Directly Observed Therapy for TB was introduced, most patients with TB retain their cards as proof of taking their TB medication. Road to health card documenting immunization status, growth and intercurrent illnesses are often kept up to date. A cross-sectional study done in Kampala in 2005 and in Dar es Salaam indicate between 67 and 74% rate of retention.  The introduction of these interventions was accompanied by political commitment, extensive health care worker training, and the creation of community awareness. Creating a mother to child dyad card alongside this already successful model may be viable. Accurate assessment of prior antiretroviral exposure is clearly important for clinical management, drug resistance surveillance, and research studies. We suggest that PMTCT programs maintain complete and accessible written records regarding prescription or dispensing of sdNVP, and ideally, ingestion of sdNVP (either witnessed or by patient self-report at the time). We would further suggest that if an HIV-infected woman is not given sdNVP for any reason, this should also be documented. This documentation could be maintained in population registers, standardized "health passports," or in maternal labor and delivery registers. Importantly, documentation of prior sdNVP receipt should be accessible to practitioners who are prescribing ART to patients.
| References|| |
|1.||Jackson JB, Musoke P, Fleming T, Guay LA, Bagenda D, Allen M, et al. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: 18-month follow-up of the HVNET 012 randomized trial. Lancet 2003;362:859-68. |
|2.||Eshleman SH, Hoover DR, Chen S, Hudelson SE, Guay LA, Mwatha A, et al. Resistance after single-dose nevirapine prophylaxis emerges in a high proportion of Malawian newborns. AIDS 2005;19:2167-9. |
|3.||Lockman S, Shapiro RL, Smeaton LM, Wester C, Thior I, Stevens L, et al. Response to antiretroviral therapy after a single, peripartum dose of Nevirapine. N Engl J Med 2007; 356:135-47. |
|4.||Jourdain G, Ngo-Giang-Huong N, Le Coeur S, Bowonwatanauwong C, Kantipong P, Leechanachai P, et al. Intrapartum exposure to Nevirapine and subsequent maternal responses to Nevirapine-based antiretroviral therapy. N Eng J Med 2004;351:229-40. |
|5.||Kuhn L, Semrau K, Ramachandran S, Sinkala M, Scott N, Kasonde P, et al. Mortality and virologic outcomes after access to antiretroviral therapy among a cohort of HIV-infected women who received single-dose Nevirapine in Lusaka, Zambia. J Acquir Immune Defic Syndr 2009;52:132-6. |
|6.||Chi BH, Chintu N, Lee A, Stinger EM, Sinkala M, Stringer JS. Expanded services for the prevention of mother-to-child HIV transmission. J Acquir Immune Defic Syndr 2007;45:125-7. |
|7.||Lockman S, Hughes MD, McIntyre J, Zheng Y, Chipato Y, Conradie FM, et al. Antiretroviral therapies in women after single-dose nevirapine exposure. N Engl J of Med 2010;363:1499-509. |
|8.||McIntyre J, Hughes M, Mellors J, Zheng Y, Hakim J, Asmelash A, et al. 'Efficacy of ART with NVP+TDF/FTC vs LPV/r+TDF/FTC among Antiretroviral-naïve Women in Africa: OCTANE Trial 2/ACTG A5208'. |
|9.||Stringer J, Sinkala M, Maclean C, Levy J, Kankasa C, DeGroot A, et al. Effectiveness of a city-wide program to prevent mother-to-child HIV transmission in Lusaka, Zambia. AIDS 2005;19:1309-15. |
|10.||Simba DO. Towards a sustainable community database: Taking advantage of the Road-to-Health cards to monitor and evaluate health interventions targeting under fives. Tanzan J Health Res 2009;11:46- 50. |
Francesca M Conradie
Deputy Director, Clinical HIV Research Unit (CHRU), Helen Joseph Hospital, Perth Road, Westdene, Johannesburg
Source of Support: None, Conflict of Interest: None
[Table 1], [Table 2], [Table 3]