British HIV Association guidelines for the management ofHIVinfection in pregnant women 2012

Abstract

The overall purpose of these guidelines is to provide guidance on best clinical practice in the treatment and management of human immunodeficiency virus (HIV)-positive pregnant women in the UK. The scope includes guidance on the use of antiretroviral therapy (ART) both to prevent HIV mother-to-child transmission (MTCT) and for the welfare of the mother herself, guidance on mode of delivery and recommendations in specific patient populations where other factors need to be taken into consideration, such as coinfection with other agents. The guidelines are aimed at clinical professionals directly involved with, and responsible for, the care of pregnant women with HIV infection. The British HIV Association (BHIVA) revised and updated the Association's guideline development manual in 2011 (www.bhiva.org/GuidelineDevelopmentManual.aspx; see also Appendix 1). BHIVA has adopted the modified GRADE system for the assessment, evaluation and grading of evidence and the development of recommendations. Full details of the guideline development process including selection of the Writing Group and the conflict of interest policy are outlined in the manual. The guidelines were commissioned by the BHIVA Guidelines Subcommittee who nominated the Chair of the Writing Group and deputy. They then nominated a Writing Group of experts in the field based on their knowledge, expertise and freedom from conflicts of interest. The scope, purpose and guideline topics were agreed by the Writing Group. Questions concerning each guideline topic were drafted and a systematic literature review undertaken by an information scientist. Details of the search questions and strategy (including the definition of populations, interventions and outcomes) are outlined in Appendices 2 and 3. The literature searches for the 2012 guidelines covered the period up until September 2011 and included abstracts from selected conferences. For each topic and healthcare question, evidence was identified and evaluated by Writing Group members with expertise in the field. Using the modified GRADE system (see Appendix 1), members were responsible for assessing and grading the quality of evidence for predefined outcomes across studies and developing and grading the strength of recommendations. All Writing Group members received training in use of the modified GRADE criteria before assessing the evidence. Owing to the lack of data from randomized controlled trials (RCTs) in several important areas the Writing Group were unable to assign high grades (in areas such as mode of delivery); however, they have made recommendations on best practice where decisions need to be made on the balance of available evidence. Recommendations are summarized and numbered sequentially within the text. The guidelines were published online for public consultation and external peer review was commissioned, comments from which resulted in minor revision before final approval by the Writing Group. BHIVA views the involvement of patient and community representatives in the guideline development process as both important and essential. The Writing Group included a patient representative who was involved in all aspects of guideline development. The guidelines will be next fully updated and revised in 2014. However, the Writing Group will continue to meet regularly to consider new information from high-quality studies and publish amendments and addendums to the current recommendations before the full revision date where this is thought to be clinically important to ensure continued best clinical practice. In the event that a woman who has initiated HAART during pregnancy has not achieved a plasma VL of <50 HIV RNA copies/mL at 36 weeks the following interventions are recommended: Review adherence and concomitant medication. Perform resistance test if appropriate. Consider therapeutic drug monitoring (TDM). Optimize to best regimen. Consider intensification. One of the major successes in the management of HIV-positive patients has been the PMTCT of HIV-1. With the widespread implementation of routine antenatal screening for HIV-1, transmission of HIV-1 from mother to child is now a rare occurrence in the UK. Despite few recent RCTs regarding the use of ART in pregnancy or obstetric intervention, practice continues to evolve. This is largely informed by observational data, theoretical considerations and expert opinion. At the outset, the aim of the Writing Group was to make these guidelines as clinically relevant and as practical as possible. The Writing Group drew up a list of questions reflecting day-to-day practice and queries. It was acknowledged that the level of evidence for many of these topics was poor but recognized that there was a need to provide guidance. These guidelines have expanded on all areas relevant to the clinical care of HIV-positive pregnant women. The guidelines are intended to inform and aid healthcare workers in the management of pregnant women with HIV. They are not intended to be prescriptive or restrictive and it is recognized that situations will arise where the optimum management may deviate from these recommendations and new data will emerge to better inform practice. A particular focus has been obstetric management. An increasing number of women are aiming for and achieving a vaginal delivery but the rate of emergency CSs has increased. It is hoped that the recommendations contained within these guidelines will enable a further increase in the proportion of vaginal deliveries and a reduction in the number of emergency CSs. Linked to this is the proposed starting gestation for women temporarily taking HAART in pregnancy, which has been brought forward depending on baseline VL. It is anticipated that this will result in a larger proportion of women achieving a VL <50 HIV RNA copies/mL by 36 weeks' gestation, thereby allowing them to plan for a vaginal delivery. Additional guidance has been provided with regard to conception on HAART, the choice of specific drugs or drug classes and the management of women with HBV or HCV coinfection. For the first time these guidelines have addressed the issue of continuation of HAART post delivery in women with a baseline CD4 cell count >350 cells/μL. The paediatric section provides further guidance on infant PEP, drug dosing and safety. It is clear that there exists an urgent need for paediatric syrup preparations for a wider variety of ARV drugs because the current options, particularly in the case of maternal viral resistance, are limited. In key areas, the National Study of HIV in Pregnancy and Childhood (NSHPC) informs the management of HIV in pregnancy through the comprehensive data collection, collation and analysis, and the need to interrogate the data continues as practice changes. Prevalence of HIV infection among women giving birth in the UK is monitored through an unlinked anonymous survey based on residual neonatal dried blood spots. This has been in place in London since 1988, other selected English regions since 1990 and Scotland between 1990 and 2008. The survey provides an estimate of overall HIV prevalence in women giving birth regardless of whether or not they have been diagnosed. Nationally, estimated prevalence increased gradually during the 1990s, more rapidly between 2000 and 2005, and has since stabilized. In 2009 the survey covered over 400 000 births, and estimated HIV prevalence was 2.2 per 1000 women giving birth (1 in every 449). Prevalence in London was about 1 in 350 in 2000, rising to 1 in 250 by 2003 and has been relatively stable since then. In the rest of England, about 1 in 3500 women giving birth was HIV positive in 2000, rising to 1 in 700 by 2006, and remaining stable since then. In Scotland prevalence increased from about 1 in 2150 in 2000 to 1 in 1150 in 2008 1,2. The majority of HIV-positive pregnant women are from sub-Saharan Africa with prevalence stable between 2004 and 2007 at about 2–2.5% among sub-Saharan African mothers giving birth in London, and slightly higher at 3–3.5% among sub-Saharan women giving birth elsewhere in England. Although prevalence among UK-born women giving birth remained low at about 0.46 per 1000 women (1 in 2200) in 2009, a gradual increase has been seen since 2000 when it was 0.16 per 1000. In the UK, the rate of HIV MTCT from diagnosed women was 25.6% in 1993, at which time interventions were virtually non-existent 3. Between 2000 and 2006, with high uptake of interventions, the overall transmission rate from diagnosed women was 1.2%, and <1% among women who had received at least 14 days of ART. Among more than 2000 women who had received HAART and delivered with an undetectable VL, there were only three transmissions, an MTCT rate of 0.1% 4. These very low transmission rates persist. A small proportion of HIV-positive women remain undiagnosed at delivery in the UK, which probably means that currently about 2% of all HIV-exposed infants (born to diagnosed and undiagnosed women) are vertically infected 1. By 2010, over 98% of all diagnosed women received some form of ART before delivery: the proportion of those who were taking zidovudine monotherapy dropped from about 20% in 2002–2003 to <5% since 2006, and only about 2% in 2009–2010. Over the same period the proportion of women delivering by elective CS declined from about two-thirds to just over one-third, while vaginal deliveries increased from <15% of all deliveries to almost 40%. Although planned vaginal delivery is now common for women who are on HAART with undetectable VL close to delivery, the increase in planned vaginal deliveries may have contributed to a rise in reported emergency CS, from about 20% to 25% 5. Between 2005 and 2010 between 1100 and 1300 children were born each year in the UK to diagnosed HIV-positive women. Since virtually all diagnosed women in the last decade have taken ART to reduce the risk of MTCT, almost all of these children are uninfected. However, this means there are, in 2011, over 11 000 HIV-exposed uninfected children in the UK whose mothers conceived on combination ART (cART), or started ART during pregnancy 5. The number of children diagnosed with vertically acquired HIV infection in the UK increased from about 70 a year in the early 1990s to a peak of 152 in 2004, and declined to 82 in 2009 6. During the last decade, about two-thirds of newly diagnosed children were born abroad. Owing to the increasing prevalence of maternal infection, combined with increasing maternal diagnosis rates and decreasing MTCT rates, the estimated number of infected children born in the UK has remained stable over the last decade, at about 30–40 a year. More than 300 children have also been reported, mostly in the early years of the epidemic, with non-vertically acquired infection, the majority from blood or blood products. Among HIV-positive children with follow-up care in the UK and Ireland, the rate of AIDS and mortality combined declined from 13.3 cases per 100 person years before 1997 to 2.5 per 100 person years in 2003–2006 7. With improving survival, the median age of children in follow-up increased from 5 years in 1996 to 12 years in 2010, by which time over 300 young people had transferred to adult care 8. Pregnancies in vertically infected young women are now occurring 9. Before the widespread implementation of the routine offer and recommendation of antenatal HIV screening in the UK, detection rates before delivery were poor. In the mid-1990s only about one-third of infected pregnant women were diagnosed, and most of those were aware of their infection status before they became pregnant 10. In England, the routine offer and recommendation policy was implemented in 2000, and similar policies were subsequently adopted elsewhere in the UK. By the end of 2003, virtually all maternity units had implemented the antenatal screening policy, and over two-thirds had achieved >80% uptake, with about one-third reaching the 90% target 11. Standards for monitoring antenatal screening were revised and updated in 2010 12. National uptake of antenatal HIV screening was reported to be 95% in 2008, up from 89% in 2005, and all regions reported at least 90% 13. Between 2000 and 2004 the majority of HIV-positive women diagnosed before delivery were identified through antenatal screening. However, since 2005 the situation has reversed and in 2010 about three-quarters of women diagnosed before delivery were already aware of their infection before they conceived, many of them diagnosed in a previous pregnancy 5. Nevertheless, some HIV-positive women remain undiagnosed at delivery, leading to potentially avoidable cases of MTCT. Since 2000, about 10 transmissions from diagnosed women have been recorded each year in the UK, against a background of increasing prevalence. However, another 20–30 UK-born children are also diagnosed each year, at various ages, whose mothers were not known to have been infected at the time of their birth 5. An audit of the circumstances surrounding nearly 90 perinatal transmissions in England in 2002–2005 demonstrated that over two-thirds of these infants were born to women who had not been diagnosed before delivery 14. About half of those undiagnosed women had declined antenatal testing. A smaller proportion had tested negative: these women presumably seroconverted in pregnancy, or while they were still breastfeeding. In 2009, the National Screening Committee considered the introduction of a routine repeat screening test in the third trimester to identify seroconversions in pregnancy, but concluded that a universal re-offer should not be introduced at that time. However, it was reiterated that women who declined the initial offer should be re-offered screening at about 28 weeks' gestation, and that repeat tests could be offered to any woman who was thought to be at continuing risk of infection, and to any woman who requested a second or subsequent test 12. It is the responsibility of clinicians caring for women with HIV and their children to report them prospectively to the NSHPC. Aggregated data tables from the UK and Ireland of ARV exposure and congenital malformations are regularly sent to the Antiretroviral Pregnancy Registry (APR). Individual prospective reports should also be made to the APR antenatally with postnatal follow-up. Antiretroviral Pregnancy Registry Research Park, 1011 Ashes Drive, Wilmington, NC 28405, USA In UK call Tel: 0800 5913 1359; Fax: 0800 5812 1658; For forms visit: www.apregistry.com This is the UK and Ireland's surveillance system for obstetric and paediatric HIV, based at the Institute of Child Health, University College London. HIV-positive children and children born to HIV-positive women are reported through the British Paediatric Surveillance Unit of the Royal College of Paediatrics and Child Health, or in the case of some units with large caseloads, direct to the NSHPC. Diagnosed pregnant women are reported prospectively through a parallel reporting scheme run under the auspices of the Royal College of Obstetricians and Gynaecologists. Longer-term data on infected children are subsequently collected through the Collaborative HIV Paediatric Study (CHIPS). For further information see the NSHPC website (http://www.nshpc.ucl.ac.uk), the CHIPS website (http://www.chipscohort.ac.uk) or email NSHPC ([email protected]). 4.1.1 Sexual health screening is recommended for pregnant women newly diagnosed with HIV. Grading: 1B 4.1.2 For HIV-positive women already engaged in HIV care that become pregnant sexual health screening is suggested. Grading: 2C 4.1.3 Genital tract infections should be treated according to BASHH guidelines. Grading: 1B There are few data regarding the prevalence of genital infections in HIV-positive women in the UK 15. At present, the majority of pregnant HIV-positive women in the UK come from, and mostly acquired HIV in, sub-Saharan Africa where the prevalence of genital infections, particularly in the HIV-positive population, can be high 16. Data from the unlinked anonymous survey of newborn infant dried blood spots show that, while the prevalence of HIV infection among pregnant women born in sub-Saharan Africa has remained relatively stable in recent years, there has been a fourfold increase in prevalence among women born in Central America and the Caribbean rising from 0.21% in 2000 to 0.78% in 2009 2. A high prevalence of genital infections in women of Afro-Caribbean origin has been reported 17. The diagnosis and treatment of genital infections in any individual have clear benefits in terms of both individual morbidity and possible infectivity to any sexual partner. In pregnancy, the welfare of the baby is an additional issue. However, apart from the recommendation that all pregnant women should be screened for HIV, HBV and syphilis, asymptomatic HIV-uninfected pregnant women in the UK are not routinely screened for genital infections. In HIV-positive pregnant women, additional considerations are the potential effects of the presence of a genital infection on MTCT of HIV-1. This could occur through an increase in the HIV-1 VL in the genital tract and/or the presence of chorioamnionitis. In addition, certain infections may be linked to premature birth, an event that occurs more frequently in HIV-positive women when compared with HIV-uninfected women. VL in cervicovaginal specimens has been shown to correlate with HIV-1 MTCT 18. Genital tract VL will usually mirror the plasma VL 19, but there is increasing evidence of compartmentalization of HIV-1 between the plasma and genital tract. Genital tract HIV-1 has been detected in women with an undetectable plasma VL 20,21 and genetic diversity of virus from the two compartments has been reported 22. A number of factors may be responsible for this, including differential drug penetration into body compartments and the presence of genital tract infections. With increasing numbers of women in the UK aiming for and achieving a vaginal delivery an increasing number of fetuses are exposed to the cervicovaginal secretions of HIV-positive women. The clinical significance of this is not clear. Data from the UK and Ireland 4 and France 23 showing no difference in MTCT associated with mode of delivery in women with an undetectable VL provide some reassurance that potential discordance may not be clinically relevant but further research is warranted. It has long been recognized that genital infections, in particular ulcerative diseases, are associated with an increased risk of sexual transmission of HIV 24,25. This may be a consequence of an increase in local HIV replication resulting in a higher VL in genital secretions, secondary to the presence of specific microorganisms, and/or ulceration and inflammation 26,27. Organisms associated with bacterial vaginosis (BV) have been shown to stimulate HIV expression in vitro 28,29. A study from Kenya demonstrated a reduction in cervical mucosal shedding of HIV-1 RNA following treatment of both gonococcal and chlamydial cervicitis 30. A study from Zimbabwe has shown a correlation between herpes simplex virus type 2 (HSV-2) antibody status and HIV-1 MTCT 31. A study from Thailand of perinatal cervicovaginal lavages showed that HSV-2 shedding was associated with increased risk of intrapartum HIV transmission and that the effect was independent of perinatal cervicovaginal lavage and plasma HIV VL. However, this study was carried out in the context of either zidovudine monotherapy from 36 weeks or placebo 32. That there may still be an increased risk associated with HSV shedding with patients on HAART is suggested by a randomized, double-blind, placebo-controlled trial of herpes-suppressive therapy in HIV-1/HSV-2-infected women taking HAART in Burkina Faso, which demonstrated that valaciclovir 500 mg twice a day further reduced genital HIV replication in those women with residual HIV shedding despite HAART 33. A study from the USA reported greater rates of HSV-2 shedding at delivery in HSV-2 seropositive women with HIV compared with HIV-negative women, 30.8% vs. 9.5% (RR 3.2, 95% CI 1.6–6.5) 34. Future studies are needed to evaluate whether valaciclovir can reduce the risk of HIV MTCT during late pregnancy, the intrapartum period and breastfeeding. Chorioamnionitis may lead to premature rupture of the membranes with the possibility of premature birth 35,36. Chorioamnionitis, prolonged ROMs and premature birth have all been associated with MTCT of HIV and may be interlinked 37-39. However, a Phase III clinical trial of antibiotics to reduce chorioamnionitis-related perinatal HIV-1 transmission showed no benefit in reducing MTCT in the context of single-dose nevirapine prophylaxis 40. Although both Chlamydia trachomatis and Neisseria gonorrhoeae have been associated with chorioamnionitis, the organisms usually implicated are those associated with BV, including Ureaplasma urealyticum 41,42. A strong association between BV and premature delivery has been reported 43,44. There are data from Malawi that suggest that BV may be associated with an increased risk of maternal HIV infection in pregnancy as well as premature delivery and MTCT of HIV 42. A study in which mothers received zidovudine from 34 weeks of pregnancy reported that maternal fever >38 °C and BV were associated with in utero transmission of HIV with 2.6-fold and 3-fold risks, respectively 45. It is not known how applicable this is in settings where mothers receive HAART from earlier in pregnancy. A large meta-analysis assessing the effects of antibiotic treatment of BV in pregnancy does not support the routine screening for, and treatment of, BV in pregnant HIV-negative women 43,44. However, the available evidence cannot rule out a small benefit in pregnancy outcome associated with the screening and treatment of BV. The latest Cochrane analysis concludes that there is little evidence that screening and treating all HIV-1-uninfected pregnant women with asymptomatic BV will prevent preterm delivery (PTD) and its consequences. However, there is some suggestion that treatment before 20 weeks' gestation may reduce the risk of PTD 46. In HIV-1-uninfected women, data regarding the effect of screening for and treating BV on premature delivery are conflicting. As outlined above, in HIV-positive pregnant women, there are additional considerations regarding the potential effect of genital infections on MTCT of HIV-1, but these data are largely from the pre-HAART era. In the setting of full virological suppression on HAART, it is unclear to what extent, if any, the presence of any genital infection will contribute to HIV MTCT. Newly diagnosed HIV-positive pregnant women should be screened for sexually transmitted infections as per the routine management of newly diagnosed patients 47. For pregnant HIV-1-positive women already engaged in HIV care, in the absence of RCTs but for the reasons outlined above, the Writing Group suggests screening for genital tract infections, including evidence of BV. This should be done as early as possible in pregnancy and consideration should be given to repeating this at about 28 weeks. Syphilis serology should be performed on both occasions. In addition, any infection detected should be treated according to the BASHH guidelines (www.bashh.org/guidelines), followed by a test of cure. Partner notification should take place where indicated, to avoid reinfection. With regard to cervical cytology, HIV-positive pregnant women should be managed as per Guidelines for the NHS Cervical Screening Programme 2010 48. Routine cytology should be deferred until after delivery, but if follow-up cytology or colposcopy is advised because of a previously abnormal result, then this should be undertaken. 4.2.1 Newly diagnosed HIV-positive pregnant women do not require any additional baseline investigations compared with non-pregnant HIV-positive women other than those routinely performed in the general antenatal clinic. Grading: 1D 4.2.2 HIV resistance testing should be performed before initiation of treatment (as per BHIVA guidelines for the treatment of HIV-1 positive adults with antiretroviral therapy 2012; www.bhiva.org/PublishedandApproved.aspx ), except for late-presenting women. Post short-course treatment a further resistance test is recommended to ensure that mutations are not missed with reversion during the off-treatment period. Grading: 1D In the case of late-presenting women, HAART, based on epidemiological assessment of resistance, should be initiated without delay and modified once the resistance test is available. 4.2.3 In women who either conceive on HAART or who do not require HAART for their own health there should be a minimum of one CD4 cell count at baseline and one at delivery. Grading: 2D 4.2.4 In women who commence HAART in pregnancy a VL should be performed 2–4 weeks after commencing HAART, at least once every trimester, at 36 weeks and at delivery. Grading: 1C Performing a VL test at 2 weeks allows for a more rapid assessment of adherence and may be of particular benefit in a late-presenting woman. 4.2.5 In women commencing HAART in pregnancy, LFTs should be performed as per routine initiation of HAART and then at each antenatal visit. Grading: 1C Hepatotoxicity may occur because of the initiation of HAART and/or the development of obstetric complications such as obstetric cholestasis, pre-eclampsia, HELLP syndrome and acute fatty liver. Close liaison with the obstetric team is recommended. For a woman who conceives on HAART that is not fully suppressive or loses virological control during pregnancy, these interventions should be undertaken as soon as possible. If treatment failure occurs when the infant is likely to be delivered prematurely and may be unable to take medication enterally, intensification should consist of therapies that readily cross the placenta such as double-dose tenofovir, raltegravir and single-dose nevirapine. 5.1.1 It is recommended that women conceiving on an effective HAART regimen should continue this even if it contains efavirenz or does not contain zidovudine. Grading: 1C Exceptions are: (i) PI monotherapy should be intensified to include (depending on tolerability, resistance and previous ARV history) one or more agents that cross the placenta. Grading: 2D (ii) The combination of stavudine and didanosine should not be prescribed in pregnancy. Grading: 1D Despite the lack of licence for the use of ART in pregnancy, with the exception of zidovudine in the third trimester, there is global consensus that women who conceive on effective HAART should continue this throughout the pregnancy. Where the risk of treatment failure due to reduced or intermittent drug exposure with hyperemesis gravidum exceeds the risk of treatment interruption the Writing Group recommends the latter option although there are no data that specifically address this issue. The APR provides the best data on teratogenicity and first trimester ART exposure. This prospective database records rates of congenital birth defects in babies born to women with first-trimester exposure to ART in comparison with background rates of congenital birth defects and second and third trimester-only exposures to the same compounds. The congenital malformation rate observed in babies exposed to a specified drug is reported once a minimum of 200 prospective first-trimester exposures to an individual ARV have been reported. In prospectively reported cases, zidovudine, lamivudine and ritonavir have been shown to have congenital malformation rates within the expected range and a congenital malformation rate >1.5-fold higher than the general population has been excluded. Among other currently used agents (abacavir, tenofovir, emtricitabine, lopinavir, atazanavir nevirapine and efavirenz) there are now more than 200 prospective reports of first-trimester exposure with no signal of increased risk (and a greater than twofold higher rate than in the general population has been excluded) 49. There are insufficient data to recommend routinely switching from efavirenz to another agent. The earlier recommendation that efavirenz be avoided in women who may conceive 50 was based on preclinical animal studies that had not been conducted on any other ART, the FDA reclassification of efavirenz to category D and the paucity of human data. Three of 20 offspring of cynomolgus macaques exposed to efavirenz in the first trimester had significant abnormalities at birth: one had anencephaly and unilateral anophthalmia; the second microphthalmia; and the third a cleft palate 51. Subsequently four anecdotal cases of myelomeningocoele and two of Dandy Walker syndrome were reported following human first-trimester efavirenz exposure. No prospective data were available, causation was not proven and a lack of data on the number of cases reported compared with the number of exposures meant that the relative risk of the putative association could not be calculated. Based on the emerging prospective data in which no evidence of human teratogenicity has been seen, the Writing Group consider that there are insufficient data to support the former position and furthermore recommend that efavirenz can be both continued and commenced (see below) in pregnancy. Thus, it is the recommendation of the Writing Group, based on current evidence, that efavirenz can be used in pregnancy without additional precautions and considerations over and above those of other ARTs. Non-pregnant adults in the UK are now rarely prescribed zidovudine as part of HAART. Despite the proven efficacy of zidovudine in PMTCT, particularly in the pre-HAART era 61, there are no data to support routinely switching to zidovudine, or adding zidovudine to a combination of ARVs that is suppressing HIV replication to <50 HIV RNA copies/mL plasma. Analysis of data combined from two observational studies, the European Collaborative Study (ECS) and the