Abstract

AIDS-related complications are a common cause of maternal death worldwide and are responsible for a high proportion of maternal deaths in low-income countries; they are a significant contributing cause of maternal death in high-income countries, although the absolute numbers are small [1, 2]. Their medical management is complicated by the requirement to balance the needs of the mother and the fetus. As opportunistic infections in pregnant women living with human immunodeficiency virus (HIV) in the UK are rare, they should be managed with close collaboration between HIV specialists, obstetricians, neonatologists, paediatricians and pharmacists. Any case of confirmed or suspected opportunistic infection in pregnancy should ideally be discussed within a fetal medicine and infection multidisciplinary team (MDT). If an MDT is not available locally, a national MDT is available and can be contacted ([email protected]; [email protected]; [email protected]; [email protected]; [email protected]). It is important to understand the physiological changes that occur in pregnancy as they can affect the interpretation of test results, clinical findings and the pharmacokinetics of drugs used in pregnant women [1, 3, 4]. Absolute CD4 cell counts characteristically decrease during pregnancy. Furthermore, there is a shift from cell-mediated immunity (Th1 response) towards humoral immunity (Th2 response) which leads to an increased susceptibility to, and severity of, certain infectious diseases in pregnant women, irrespective of HIV infection, including toxoplasmosis, varicella and listeriosis [5]. There are increases in cardiac output, plasma volume, red cell mass and glomerular filtration rate in pregnancy. Absorption of aerosolised medication may be affected by increased tidal volume and pulmonary volume. Placental transfer of drugs, increased renal clearance, altered gastrointestinal absorption and metabolism by the fetus may affect drug levels. Therapeutic drug monitoring should be considered due to altered drug pharmacokinetics in pregnancy, and the potential for complicated multiple interactions between antiretroviral agents and many of the drugs used to treat opportunistic infections [3, 6]. For information about drug safety during pregnancy and breastfeeding, see Appendix 1. While this guidance refers to women and breastfeeding throughout, the writing group fully acknowledges that it is also applicable to pregnant transgender men and gender diverse individuals and also that some may prefer the term chestfeeding to breastfeeding. We endorse the use of person-centered, gender-inclusive language in healthcare settings according to the individual's preferences. Guidance on supporting people living with HIV with opportunistic infections, including in pregnancy, can be found on the British HIV Association (BHIVA) website (https://www.bhiva.org/file/6225e44b53c49/OI-guidelines-supporting-patients.pdf). A full review of these guidelines is due by 2029, with interim updates only if recommendations need updating in line with new data. In general, pregnant women with symptoms suggestive of an AIDS-defining illness should be managed and investigated in the same way as non-pregnant individuals. When choosing an imaging modality for the diagnosis of opportunistic infections in pregnant women, consideration should be given to the need for a rapid diagnosis and the potential harm of the investigation. The potential fetal doses of ionising radiation from different imaging modalities and the risk of childhood cancer were recently outlined by Wiles et al. [7]. Discussion between HIV specialists, obstetricians, senior radiologists and the pregnant woman is recommended. If opportunistic infection in the lung is suspected, a chest X-ray may be carried out with little or no risk to the fetus [7]. An ultrasound scan is a safe option for imaging of the abdomen. A direct computed tomography (CT) scan of the fetus in the pregnant abdomen should be avoided where possible. Magnetic resonance imaging of the fetus and abdomen is considered safe at all stages of pregnancy, although use of gadolinium in pregnancy has been reported to cause an increased risk of inflammatory conditions and should be avoided where possible [8]. CT scans of the brain, thorax or limbs of the mother may be carried out with minimal exposure to the fetus. Modern CT scanners have little radiation scatter to areas outside the scanner itself, so the main radiation scatter that would affect the fetus during a thoracic CT scan would be internally within the body of the mother. The use of contrast with CT scanning is permitted. Pulmonary embolus (PE) is a leading cause of maternal morbidity and death, and suspected PEs need to be investigated and treated promptly. Ventilation and perfusion scans, or in some situations limited 'perfusion' scans, are regarded as acceptable for cases of suspected PE in pregnancy. CT pulmonary angiogram (CTPA) scans are also being used more often and are becoming regarded by many as the investigation of choice for the diagnosis of PEs in pregnancy. Although CTPA is associated with low fetal radiation exposure, it does expose the mother's breast tissue to a relatively high radiation dose [7, 9]. There are no specific contraindications for lymph node biopsy, liver biopsy and lumbar puncture in pregnancy. Endoscopic procedures, including bronchoscopy and upper and lower gastrointestinal endoscopy, may also be undertaken if necessary [10]. Where an opportunistic infection has been diagnosed, the fetus should be closely monitored, for example by serial high-resolution ultrasound scans and fetal cardiac monitoring, so that signs of disease, growth retardation, fetal distress or drug toxicity can be detected early [1, 11]. A paediatrician or neonatologist with expertise in HIV and congenital infection should be involved, prior to delivery, in discussions about the management of opportunistic infections in pregnant women. An MDT approach is advised to include risk–benefit discussions about opportunistic infection treatment choices in pregnancy, fetal monitoring and early review of the neonate for signs of congenital infection, drug toxicity and teratogenicity. Congenital infections in the neonate have been described for a variety of opportunistic pathogens affecting the mother. These include Mycobacterium tuberculosis [12, 13], Cryptococcus neoformans [14-16], cytomegalovirus (CMV) [17], Pneumocystis jirovecii [18, 19] and Toxoplasma [20-22]. In some instances, it may be difficult to distinguish between congenital and early postnatal infection. Neonates born to women living with HIV should be assessed by a paediatrician with expertise in HIV and other congenital infections, and where necessary actively screened for congenital infections using appropriate national and local guidelines and assessed for signs of teratogenicity or drug toxicity. The scope, purpose and guideline topics were agreed by the writing group. The search (population, intervention, comparator and outcome [PICO]) questions were set and an independent systematic literature review carried out. The Medline, Embase and Cochrane Library databases were searched and the literature reviewed to address each question. The PICO questions and search strategies are outlined in Appendix 2. Further details of the methodology can be found on the BHIVA website (https://www.bhiva.org/file/5d514ec9b503d/OI-guidelines-methods-general.pdf), including the use of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system to assess and grade the evidence. Good practice points (GPPs) are recommendations, based on the clinical judgment and experience of the writing group, with which few clinicians are expected to disagree and for which evidence is unlikely to emerge as they are generally considered to be good practice. From Section 6.2 Antifungal treatment for candidiasis and cryptococcal infection. There is some evidence from case studies that PCP in pregnancy may be more aggressive, with increased morbidity and mortality, than in non-pregnant women [23]. Investigation and diagnosis of PCP in pregnant women is the same as for non-pregnant adults. No large, randomised trials investigating the treatment of PCP have included pregnant and breastfeeding women [24]. Trimethoprim-sulfamethoxazole inhibits bacterial and, to a lesser extent, human folate metabolism. Folate is essential for fetal development and folate deficiency has been associated with an increased risk of neural tube defects and other congenital anomalies [25]. Trimethoprim-sulfamethoxazole has been associated with cardiovascular anomalies and defects of the urinary tract in some studies [26]. Other studies, including a meta-analysis, have not found an increased risk of congenital malformations associated with trimethoprim-sulfamethoxazole use in pregnancy, including in the first trimester [27, 28]. A nested control study from the Quebec Pregnancy Cohort (predominantly in women without HIV) found an increase in the rates of spontaneous abortion in pregnant women who took trimethoprim-sulfamethoxazole in pregnancy (adjusted odds ratio 2.94, 95% confidence interval 1.89–4.57) [29]. However, the benefit of trimethoprim-sulfamethoxazole to treat PCP in pregnancy outweighs any potential harm and it should be used, including in the first trimester. Folic acid at 0.4–5 mg/day is recommended for all pregnant women [30]. Some authors have advocated using higher-dose folic acid replacement to prevent fetal anomalies associated with anti-folate drugs [31]. However, others have reported PCP treatment failure in this context [32]. Therefore, doses above 0.4 mg daily should only be considered in the first trimester of pregnancy and on a case-by-case basis. Steroids should be administered as per standard guidelines for the treatment of PCP in non-pregnant women. Steroids are generally considered safe in pregnancy, although the findings of a systematic review of case–control studies suggested a small increased risk of oral clefts [33]. However, other large population-based studies have not found an association between maternal steroids and congenital anomalies [34, 35]. Alternative options are limited to dapsone with trimethoprim or atovaquone. Clindamycin is generally considered safe in pregnancy, but primaquine can cause both maternal and fetal haemolysis in G6PD deficiency and is therefore not recommended. Dapsone has been used in pregnancy to treat leprosy and malaria and appears to be safe [36, 37]. The risk of haemolysis in neonates with G6PD deficiency seems to be very low [38]. Limited data suggest that atovaqone is safe in pregnancy [39]. Chemoprophylaxis for PCP should be prescribed to pregnant women living with HIV in line with standard guidelines for non-pregnant individuals. It is important to remember that there is a false reduction in absolute CD4 count during pregnancy, especially during the third trimester, and in such circumstances a CD4 percentage less than 14% can be used as an indicator for the need to commence PCP prophylaxis. Trimethoprim-sulfamethoxazole is the preferred prophylactic agent against PCP in pregnancy. There are theoretical concerns over the safety of this drug in the first trimester, especially prior to the closure of the neural tube, and an alternative agent could be considered during this time. Possible alternatives include once daily dapsone, atovaquone or nebulised pentamidine. It is unclear whether congenital Pneumocystis infection can occur, although occasional case reports suggest it may be possible [40, 41]. However, it is recommended that infants born to women treated for PCP in pregnancy should be reviewed by a paediatrician or neonatologist, mainly to rule out any teratogenic effects and/or toxicity of maternal drug treatment. If there are signs of possible Pneumocystis infection, the infant should be urgently reviewed and the case discussed with the local paediatric infectious diseases team. Vaginal candidiasis occurs more frequently in pregnancy; however, although it is supposed that other fungal infections are more common in pregnancy, there is little direct evidence to support this. Investigation of suspected fungal infections should be identical in pregnancy, except for considering the risk associated with radiological procedures (see Introduction). Candidiasis Cryptococcal infection Two case series illustrate that cryptococcal infection in pregnancy is associated with high maternal mortality and frequent stillbirths and miscarriages despite antifungal therapy [42, 43]. There have been no reports of teratogenesis or other adverse pregnancy outcomes with liposomal amphotericin B [44]. Flucytosine has been associated with teratogenesis when used in rats at high doses [45]; however, there are case reports of its use to treat cryptococcal meningitis during the second and third trimesters of pregnancy with healthy fetal outcomes [46, 47]. Although single-dose fluconazole has not been associated with any birth defects in pregnancy [48], specific birth defects associated with continuous daily doses of fluconazole of 400 mg/day or more in the first trimester, including cardiac septal defects, have been suggested by one case report and the finding of a large case–control study [49, 50]. Two studies have also demonstrated a higher risk of spontaneous abortion with fluconazole exposure in the first and second trimesters [51]. Hence national agencies have been re-evaluating guidelines for its use in pregnancy. In one case series, 12 pregnant Ugandan women with cryptococcal meningitis were described, five of whom received fluconazole in the second and third trimesters after amphotericin B therapy [43]. Although few of these pregnancies resulted in live births, no congenital abnormalities were detected. Voriconazole has been strongly associated with teratogenicity in rats and there have been no reports of its use in humans during pregnancy [52]. Itraconazole is also not recommended in pregnancy due to teratogenicity in animals, however the findings of one case series suggested it may be safe in humans [53]. The echinocandins are not considered safe, given evidence of teratogenicity in animal studies, although no human data are available [54]. Candida infection in pregnancy is not known to have direct implications for the neonate. Infants should be reviewed if a woman has been treated with antifungal drugs with a potential risk of teratogenicity, especially during the first trimester. Congenital cryptococcal infection has been reported, but appears to be rare [14-16]. Infants born to mothers with cryptococcal disease in pregnancy should be assessed for neonatal cryptococcal disease by a paediatrician or neonatologist soon after birth. If cryptococcal disease is suspected, advice on management should be sought from a local paediatric infectious diseases team. The risk of reactivation of Toxoplasma gondii increases at lower CD4 counts. Pregnant women with a low CD4 count and a history of T. gondii infection should be monitored for signs of T. gondii reactivation. Pregnant women with negative Toxoplasma serology, suggesting no previous exposure to T. gondii infection, should be advised about the behavioural risk factors for primary T. gondii acquisition and associated risk reduction strategies [55, 56]. There is no evidence that toxoplasmosis is more severe when it occurs in pregnancy [57]. Investigation and diagnosis of toxoplasmosis in pregnant women is the same as for non-pregnant adults. Toxoplasmosis in pregnancy should be treated in the same way as in non-pregnant individuals. Pyrimethamine and sulfadiazine are the drugs of choice in pregnancy, especially after the first trimester [58]. Pyrimethamine is a folic acid antagonist (inhibitor of dihydrofolate reductase). Folinic acid 15 mg daily is recommended as co-administration, as high-dose folic acid (5 mg) compromises the efficacy of sulfadiazine-pyrimethamine in pregnancy [59]. Although some folic acid antagonists are human teratogens, this does not seem to include pyrimethamine [60-66]. A few single case reports of defects have been documented following exposure in pregnancy to pyrimethamine, but no systematic studies are available. One case report described a severe defect of the abdominal and thoracic wall and a missing left arm in an infant exposed to pyrimethamine, chloroquine and dapsone in the first trimester. However, an association between the drugs and the defect has been questioned [58]. The safety of sulfonamides during pregnancy has not been determined, although sulfonamides do not appear to pose a significant teratogenic risk when used as single agents [58]. One study in humans demonstrated an association with birth defects, but a causative association could not be established as other factors, such as its use in combination with trimethoprim, may have had a role [58]. The sulfonamides readily cross the placenta to the fetus during all stages of gestation [67-74]. Equilibrium with maternal blood is usually established after 2–3 hours, with fetal levels averaging 70–90% of maternal levels. Significant levels may persist in the neonate for several days after birth when given to the mother near term. Sulfonamides are most likely to cause harm to the neonate if administered near to term. Toxicities that may be observed in the neonate include jaundice, haemolytic anaemia and kernicterus. Severe jaundice in the neonate has been related to maternal sulfonamide ingestion at term by several authors [75-80]. Premature infants seem to be especially prone to the development of hyperbilirubinaemia [79]. Haemolytic anaemia has been reported in two neonates and in a fetus following in utero exposure to sulfonamides [75, 76, 80]; both neonates survived. Because of the potential toxicity to the neonate, these agents should be avoided near term. Due to these concerns, we recommend that sulfonamides should not be used after week 32 of pregnancy and that clindamycin should be used instead. Clindamycin crosses the placenta, achieving serum cord levels of approximately 50% of the maternal serum level [81, 82], and is generally considered safe in pregnancy. Clindamycin is an alternative option when there is intolerance to sulfonamides. Secondary prophylaxis should be the same as for non-pregnant individuals. The effect of maternal HIV on the risk of congenital T. gondii infection is unclear. Similarly, the impact of maternal T. gondii infection during pregnancy on the risk of vertical HIV transmission is not known. In the non-immunocompromised host, congenital infection is usually associated with primary infection in pregnancy. There have been case reports of T. gondii transmission to neonates following reactivation during pregnancy in women living with HIV [83, 84], and not always in the context of severe immunosuppression [85]. The neonate known to be at risk of congenital T. gondii infection should be reviewed by a paediatrician or neonatologist soon after birth. When clinical findings or laboratory/radiology investigations are consistent with possible congenital T. gondii infection, the case should be discussed with a local paediatric infectious diseases specialist to guide further investigation and medical management [86, 87]. The main focus of this guidance is the management of symptomatic CMV disease presenting as an opportunistic infection in pregnant women living with HIV. A detailed review of the management of all CMV infection and its impact on the pregnancy is outside the scope of this guidance. There is no evidence to suggest that CMV disease presents differently in pregnant and non-pregnant women [88]. It is estimated that 5–10% of maternal CMV infections are symptomatic. Both primary and non-primary infections (through reactivation or reinfection with a different strain) can lead to CMV disease in the mother and congenital CMV disease in the fetus [89-91]. For pregnant women living with HIV, indications for investigation and treatment of suspected maternal CMV disease follow the same principles as for non-pregnant women. Investigation and interventions for possible in utero CMV infection of the fetus also mirror the guidance for pregnancies in women without HIV, taking into account that CMV infection diagnosed in the context of advanced HIV has a higher risk of transmission to the fetus [92]. All pregnant women should be advised about reducing the risk of acquiring infections such as CMV that are associated with congenital infection [93]. All licensed anti-CMV agents including ganciclovir, valganciclovir, foscarnet and cidofovir show adverse fetal effects in animal studies but there is a lack of adequate and well-controlled studies in humans. Any systemic therapy for the mother, especially in the first trimester, should be subject to a risk–benefit analysis with consideration of localised therapy if appropriate, for example intra-vitreal maintenance treatment of CMV retinitis to minimise the risk to both the mother and fetus [94]. The most clinical experience and case report data are available for ganciclovir and its oral pro-drug valganciclovir. We suggest ganciclovir and valganciclovir as first-line treatment for CMV disease in pregnancy. Minimal safety data exist for the use of ganciclovir and valganciclovir in pregnancy, and their use is recommended after considering the risks and benefits. Ganciclovir has been shown to be associated with teratogenicity in animal studies and is also mutagenic [95, 96]. A 2023 pharmacovigilance study (not specific to women living with HIV) did not find any increased reporting of adverse pregnancy outcomes or birth defects with ganciclovir/valganciclovir when compared with the use of aciclovir/valaciclovir during pregnancy [97]. There is evidence from case reports for the safe use of ganciclovir in the first and second trimesters for pregnant women in the post-transplant setting [98-100]. Of relevance to pregnant women living with HIV, there are case report data showing safe use of ganciclovir in the second and third trimesters to treat maternal CMV disease, with variable success in preventing vertical transmission of congenital CMV infection [101-105]. In one case of ganciclovir use, maternal and fetal death were reported, probably related to underlying CMV disease rather than treatment [106]. Foscarnet has been associated with fetal skeletal abnormalities in animal studies [107, 108]. A single case report demonstrated the safe use of foscarnet in the second trimester of pregnancy for a non-CMV indication with no adverse fetal effects [109]. A further case of foscarnet use in the third trimester for a non-CMV indication reported neonatal death; it is unclear whether this was related to the treatment [110]. Due to the potential for renal toxicity, careful monitoring of amniotic fluid for oligohydramnios should be undertaken, especially in the second and third trimesters. Cidofovir has been shown to be embryotoxic in animal studies, including rat and rabbit models. It is associated with fetal soft tissue and skeletal abnormalities. There are no data on the clinical use of cidofovir in pregnant women and its use is not advised [111]. Outside the context of CMV as an HIV-associated opportunistic infection, there is increasing evidence to suggest that valaciclovir at high doses can reduce the incidence of congenital CMV infection following maternal CMV infection, especially in the first trimester. This is an area of ongoing review and treatment of asymptomatic CMV infection during pregnancy (for those with or without HIV) with the sole aim of preventing congenital infection may be appropriate in some cases and should be discussed with the mother and relevant MDT on a case-by-case basis [112, 113]. There is evidence to suggest that neonates born to women living with HIV not on antiretroviral therapy (ART) are at higher risk of developing congenital CMV infection than those born to mothers on ART, however data on the subsequent impact and rates of congenital CMV disease are scarce [92, 114-116]. We recommend vigilance for signs of congenital CMV in neonates born to women living with HIV, especially those born to mothers who have CD4 counts <200 cells/mm3 (<14%) or who are not on ART. These neonates should be assessed for signs and symptoms of congenital CMV infection. Audiology assessment according to national guidance should be ensured. There is no strong evidence to support CMV PCR screening of all infants exposed to HIV. The same pathways as for the general population should be followed. In view of the higher risk of vertical CMV transmission, all neonates born to mothers with evidence of active CMV disease should be reviewed and investigated for congenital CMV infection with CMV PCR using an appropriate sample (i.e. urine or saliva) as soon as possible after birth and within the first 21 days of life and discussed with the paediatric infectious diseases team. The same pathways for investigation and management of possible congenital CMV should be followed as for the general population. There is no evidence to suggest that disseminated Mycobacterium avium complex (DMAC) disease in pregnant women presents differently to that in non-pregnant individuals. Diagnostic tests and treatment indications are the same for pregnant women as for other adults. Clarithromycin has been associated with birth defects in animal studies. However, data from the Quebec Pregnancy Cohort did not identify an increased risk of congenital malformations in 686 pregnancies with exposure to clarithromycin in the first trimester [117]. Furthermore, 401 women in Denmark who were exposed to clarithromycin in the first trimester showed no increased risk of congenital malformations [118]. However, several studies have shown an increased risk of spontaneous abortion following first trimester exposure to clarithromycin [118, 119]. Of note, Muanda et al. in a nested case–control study within the Quebec Pregnancy Cohort did not control for the severity of infection, and their findings may have been due to maternal infection rather than drug exposure [119]. Azithromycin has not been associated with birth defects in animal studies and human studies have not identified any increased risk of congenital malformations in infants exposed in utero [117-120]. Muanda et al. found an increased risk of spontaneous abortion following maternal exposure to azithromycin in pregnancy. However, as for clarithromycin, maternal infection may have been the cause of their findings. Therefore, azithromycin is preferred as first-line treatment for DMAC disease in pregnant women and clarithromycin is not recommended. Ethambutol and rifampicin have been widely used to treat Mycobacterium tuberculosis in pregnancy without any teratogenic effects being identified and are therefore considered safe to use when treating pregnant women with DMAC infection. There are no data on the safety of rifabutin in pregnancy and therefore rifampicin is preferred when treating pregnant women [121]. Congenital/neonatal Mycobacterium avium complex disease has not been reported. Infants born to women treated for DMAC in pregnancy should be reviewed by a paediatrician or neonatologist to exclude drug toxicity or teratogenicity. If there are any concerns regarding possible infection, review with the local paediatric infectious diseases team is recommended. The writing group thanks Dr Lisa Hamzah, St George's University Hospitals NHS Foundation Trust, London, and Prof Graham Taylor, Imperial College London, for their reviews and helpful comments. The writing group also thanks Mr Ben Cromarty and Prof Nicoletta Policek, both of UK Community Advisory Board (UK-CAB), for reviewing this chapter. 1. Rosenblatt JE. Antiparasitic agents. Mayo Clin Proc 1999; 74: 1161–1175. 2. Pasternak B, Hviid A. Atovaquone-proguanil use in early pregnancy and the risk of birth defects. Arch Intern Med 2011; 171: 259–260. 3. Muanda FT, Sheehy O, Bérard A. Use of antibiotics during pregnancy and the risk of major congenital malformations: a population based cohort study. Br J Clin Pharmacol 2017; 83: 2557–2571. 4. Briggs GG, Freeman RK, Tower CV et al. Briggs Drugs in Pregnancy and Lactation: A Reference Guide to Fetal and Neonatal Risk. 2021, 12th edn. Lippincott Williams & Wilkins. 5. Seidel V, Feiterna-Sperling C, Siedentopf JP et al. Intrauterine therapy of cytomegalovirus infection with valganciclovir: review of the literature. Med Microbiol Immunol 2017; 206: 347–354. The Medline, Embase and Cochrane Library databases were searched for English language papers published between 2010 and the date of the searches (May 2023). Abstracts from selected conferences (Conference on Retroviruses and Opportunistic Infections, International AIDS Society/AIDS Conference, European AIDS Clinical Society, BHIVA, HIV Glasgow, Infectious Diseases Society of America and International Workshop on HIV and Women) between January 2017 and June 2019 were also searched. The databases were searched using the following terms: Search 1. Introduction/investigations HIV/AIDS AND (pregn* OR fetus) AND (diagnostic imaging OR irradiation/radiation dose OR CT scan OR MRI scan OR X-ray OR gadolinium OR contrast medium OR endoscopy OR surgery OR (intensive care AND (ventilation OR pregnancy))) Search 2. Pneumocystis jirovecii pneumonia (PCP) HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (PCP OR co-trimoxazole OR trimethoprim OR sulfamethoxazole OR clindamycin OR primaquine OR atovaquone OR pentamidine) Search 3. Candidiasis HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (candid* OR amphotericin B OR anidulafungin OR caspofungin OR fluconazole OR itraconazole OR micafungin OR voriconazole OR posaconazole OR isavuconazole) Search 4. Cryptococcal infection HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (cryptococci* OR amphotericin B OR flucytosine OR fluconazole) Search 5. Toxoplasmosis HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (toxoplasma* OR pyrimethamine OR sulfonamide OR spiramycin OR atovaquone) Search 6. Cytomegalovirus (CMV) HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (CMV OR valganciclovir OR ganciclovir OR foscarnet OR cidovovir) Search 7. Mycobacterium avium complex (MAC) HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (non-tuberculous/atypical mycobacteria/mycobacterium avium complex/M kansasii OR rifabutin OR clarithromycin OR azithromycin OR ethambutol) Search 8. Antifungal agents HIV/AIDS AND pregn* AND (antifungal OR fluconazole OR itraconazole OR posaconazole OR clotrimazole OR ketoconazole OR miconazole OR isoconazole OR voriconazole OR amphotericin OR echinocandin OR caspofungin OR micofungin OR anidulafungin OR terbinafine OR flucytosine OR griseofulvin OR imidazole OR triazole OR nystatin) Search 9. Neonatal considerations HIV/AIDS AND (perinatal OR congenital OR pregn*) AND (PCP OR non-tuberculous/atypical mycobacteria OR toxoplasmosis OR candidiasis OR cryptococcosis OR CMV) This search was also limited to randomised controlled trials, systematic reviews and observational studies. The literature searches were based on the following PICO questions: Which radiological investigations are safe to use for pregnant women living with HIV and at risk of opportunistic infections? What is the best treatment for PCP for pregnant women living with HIV? For pregnant women living with HIV, with a CD4 count <200 cells/mm3, what is the best agent for PCP prophylaxis? What dose of folic acid should be used for pregnant women receiving trimethoprim-sulfamethoxazole for the treatment of PCP? What is the best treatment for candidiasis for pregnant women living with HIV? What is the best treatment for cryptococcal infection for pregnant women living with HIV? What is the best treatment for toxoplasmosis for pregnant women living with HIV? What is the best treatment for CMV disease for pregnant women living with HIV? What is the best treatment for disseminated MAC disease for pregnant women living with HIV?