Chapter VII
HIV And Reproduction

Jean Anderson, MD

I. Introduction  TOP

The ability to become pregnant and to bear children is uniquely female. With increasing numbers of HIV-infected women, 80% of whom are of childbearing age, and concerns about perinatal transmission of HIV, pregnancy in the setting of HIV infection has been a focus of much interest, research, and often discrimination. From 1989 to 1994 it was estimated that 1.5 to 1.7/1000 U.S. childbearing women were HIV-positive (Davis, 1998); however, this number may grow as more women become infected through sexual exposure, often unaware of their risk, and as more women who know they are infected choose to become pregnant because of therapeutic advances in care and prevention of vertical transmission. Almost one-third of HIV-infected men and women receiving medical care in the US desire children in the future (Chen, 2001). Furthermore, 20% of serodiscordant couples would practice unsafe sex in order to conceive (Klein, 2003).

This chapter will review issues related to contraception and pregnancy and will discuss guidelines for care during pregnancy to optimize the health of both the mother and the fetus and infant.

II. Counseling  TOP

The American College of Obstetricians and Gynecologists (ACOG) advocates reproductive counseling for all women of child bearing age as a part of primary care. For women known to be HIV-infected, education and counseling about pregnancy and HIV should be done early in the course of HIV care, not delayed until the woman is pregnant, so that decisions about contraception and if or when to get pregnant can be most informed and carefully considered. Discussions about pregnancy should be repeated at intervals throughout care, especially when personal circumstances change (e.g. new sexual partner, postpartum); when there is nonuse of effective contraception; where therapies are considered which may have adverse effects in pregnancy; or when the woman expresses a desire to become pregnant. Over one half of pregnancies in U.S. women are unplanned, and many of the risk factors for unintended pregnancy also place women at increased risk for HIV. These include:

• substance abuse (patient or partner)
• mental illness
• domestic violence

Adolescents are at an increased risk of unintended pregnancy and may also be at increased risk for HIV because of frequent unstable sexual relationships and unsafe sexual practices. Women with advanced HIV disease and HIV dementia may be at increased risk for unintended pregnancy if they are dependent on a contraceptive method (such as condom use or oral contraceptives) that requires negotiation with a sexual partner or other ongoing patient action (i.e., remembering to take pills). Issues to discuss when counseling about reproductive issues are listed in Table 7-1.

III. Contraception  TOP

The majority of HIV-infected U.S. women use some form of contraception, most commonly condoms (Wilson, 1999; Watts, 1999). Women using no form of contraception do not necessarily intend to become pregnant but may lack significant power in their sexual relationship, be under pressure from partner or family to have children, may not have disclosed their HIV status to their partner, be unaware of their options concerning contraception or believe they cannot become pregnant, have a disorganized lifestyle that precludes consistent use of contraception, or simply have decided to take their chances. Unplanned also does not necessarily mean unwanted; several studies show low rates of elective pregnancy termination in HIV-positive women (Smits, 1999; Greco, 1999) and no significant difference in repeat pregnancy rates in HIV-positive compared with HIV-negative women from an inner-city population (Lindsay, 1995). Table 7-2 outlines currently available methods of contraception, their effectiveness, side effects and contraindications, and noncontraceptive benefits.

Table 7-2: Contraceptive Methods

Hormonal methods of contraception, particularly combined estrogen-progestin oral contraceptives (OCs), can have significant drug interactions, resulting in either decreased contraceptive effectiveness or increased or decreased concentrations of the coadministered drug. Use of nelfinavir, ritonavir, lopinavir, and nevirapine are associated with decreases in ethinyl estradiol (estrogen component of OCs) with possible decrease in effectiveness (and possible increase in breakthrough bleeding); an alternative or additional method should be used. Indinavir, atazanavir and efavirenz are associated with increases in ethinyl estradiol and indinavir and atazanavir are associated with increases in norethindrone (progestin component in many OCs). Norethindrone levels are increased over 100% with atazanavir. The clinical significance of these increases in hormonal blood levels is unclear, but they raise concerns about potential increase in estrogen- or progestin-related side effects. Alternative methods of contraception should be considered; if OCs are used, lowest effective doses of affected hormonal components should be prescribed and an additional method is recommended. Amprenavir (and probably fos-amprenavir) not only increases blood levels of ethinyl estradiol and norethindrone, but OCs decrease amprenavir levels as well; these drugs should not be co-administered and an alternative contraceptive method should be used. Other medications known to interact with oral contraceptives (and in some cases with progestin-only contraceptives) include tetracyclines, penicillin, oral hypoglycemic agents, rifampin, tricyclic antidepressants, oral anticoagulants, ß-blockers, methyldopa, vitamin C, benzodiazepines, and seizure medications. Clinicians treating women who are at risk for drug interactions should review the need for possible use of alternative methods of contraception or dose adjustment for the interacting agent. There is minimal information about drug interactions with use of newer hormonal contraceptive methods (patch, vaginal ring, estrogen-progestin injection).

Concerns have been raised about possible increased risk of HIV transmission or acquisition in hormonal contraceptive users. There is evidence that both combined oral contraceptives and progestin-only contraceptives may increase genital tract HIV shedding; furthermore, oral contraceptives have been associated with increased cervical ectopy, which has also been linked with genital tract HIV shedding. Similarly, ectopy or other epithelial changes secondary to hormonal contraception or associated effects on immune response may increase susceptibility to HIV, and animal studies have suggested a link between progesterone implants and vulnerability to simian immunodeficiency virus (Mostad, 1998; Plummer, 1998). Data from epidemiologic studies are conflicting and inconclusive regarding the relationship of these methods of contraception and HIV transmission (Martin, 1998; Stephenson, 1998; Kiddugavu, 2003; Wang, 1999; Kapiga, 1998). At the current time, given their effectiveness, overall safety, and ease of use, hormonal methods of contraception remain an appropriate option for HIV-infected or at-risk women. These women should be advised that these contraceptives do not protect against HIV transmission and consistent condom use should be emphasized.

Use of the intrauterine device (IUD) has been accompanied by concerns about a potential increased risk for HIV susceptibility. However, a recent large prospective cohort study of almost 2500 HIV-uninfected women found no association between IUD use and HIV transmission; there was also no increased risk associated with increasing duration of use (Kapiga, 1998). Furthermore, another study from Kenya found no increase in overall complications or infection-related complications in HIV-positive IUD users as compared to HIV-negative IUD users and complications did not differ by CD4 count (Morrison, 2001). Use of the IUD was not associated with increased rate of cervical HIV shedding 4 months after insertion over baseline pre-insertion shedding rates (Richardson, 1999). However, risk of pelvic inflammatory disease is increased in IUD users who are at increased risk for acquiring other sexually transmitted infections (STIs) and a recent study found an association between IUD use and bacterial vaginosis, also a risk factor for PID (Joesoef, 2001). Furthermore, copper IUDs are associated with increased menstrual flow and duration, possibly contributing to transmission risk and anemia in HIV-positive women. The IUD should be used cautiously in the setting of HIV infection.

Spermicides have in vitro activity against HIV; however, standard spermicidal doses of nonoxynol-9 (N-9) have been associated with an increase in irritation, colposcopic and histologic evidence of inflammation, and decreased numbers of vaginal lactobacilli in N-9 users, compared with placebo recipients (Stafford, 1998). In a randomized placebo-controlled clinical trial of N-9 conducted among commercial sex workers with high rates of sexual activity, N-9 did not protect against HIV infection, resulted in increased vaginal lesions, and possibly caused increased transmission (Richardson, 2002). Although these adverse effects might not occur with less frequent use, given current evidence, spermicides containing N-9 should not be recommended as an effective means of HIV prevention. A meta-analysis of randomized controlled trials using N-9 also found no evidence of protection against HIV acquisition (Wilkinson, 2002) and N-9 appears to offer no protection against sexually transmitted infections such as gonorrhea or chlamydia (WHO, 2002).

Condoms — used consistently — reduce HIV transmission risk by 80% and provide the best known protection against sexual transmission of HIV. They should be emphasized for all HIV-infected and at-risk women to decrease risk of HIV transmission/acquisition and transmission/acquisition of other STIs. Other barrier contraceptive methods provide limited STI protection and have not been shown to offer significant protection against HIV transmission.

Because male and female condoms are used for both prevention of infection and prevention of pregnancy, these two separate issues should be distinguished when counseling patients. There is some evidence that condom use is less likely in HIV-infected women using other methods of contraception. Condom use should be reinforced for HIV-positive or at-risk women when prevention of pregnancy is not a concern: postmenopausal women, during pregnancy, despite infertility, and with the use of other methods of contraception. As with use of contraception in general, use of condoms for HIV prevention is related to education, relationship to sexual partner, and chaos in life. There is some indication that use of HAART and decrease in viral load may lead to drop-off in condom use.

IV. Pregnancy Testing  TOP

Indications for pregnancy testing in currently or recently sexually active women:

• missed menses (unless on Norplant or Depo-Provera)
• irregular bleeding (unless on Norplant or Depo-Provera)
• new onset of irregular bleeding after prolonged amenorrhea on Norplant/
  Depo-Provera
• new onset pelvic pain
• enlarged uterus or adnexal mass on exam
• consider before instituting new therapies

Pregnancy tests are performed on blood or urine and may be qualitative (positive/negative) or quantitative. Quantitative tests are useful in early pregnancy when ectopic pregnancy or abnormal intrauterine pregnancy (e.g., missed abortion) is suspected. Several qualitative urine pregnancy tests are available over the counter. Most pregnancy tests in current use are positive before the first missed menses with normal intrauterine pregnancy. Table 7-3 lists types of available pregnancy tests and their sensitivity.

V. HIV and Fertility  TOP

Recent studies in Africa, as well as in developed countries, have suggested that HIV may have an adverse effect on fertility in both symptomatic and asymptomatic women (Desgrees, 1999; L.M. Lee, 2000; Zaba, 1998). A cross-sectional study from Uganda found likelihood of pregnancy lower in HIV-positive women compared with HIV-negative women and lowest in women who were symptomatic from HIV or were coinfected with syphilis. A prospective study in the same population found that pregnancy rates were lower and pregnancy loss was more common in HIV-infected women (Gray, 1998). There are longer intervals between births for HIV-positive women compared to HIV-negative women (Glynn, 2000) and higher HIV viral loads have been associated with longer times to achieve pregnancy in women trying to conceive (Nguyen, 2003). In addition, both advanced disease stage and HIV-related therapies may be associated with abnormal sperm counts in HIV-infected men and menstrual dysfunction in HIV-infected women.

VI. Effects of Pregnancy and HIV Infection  TOP

A. CD4 COUNT AND HIV RNA LEVELS IN PREGNANCY

In both HIV-positive and HIV-negative women the response of CD4 cell counts to pregnancy is variable (Tuomala, 1997). Many studies have suggested that there is a decline in absolute CD4 cell counts in pregnancy, which return to baseline at the end of pregnancy or during the postpartum period. The decline in CD4 count is thought secondary to hemodilution; on the other hand, percentage of CD4 cells remains relatively stable. Therefore, percentage, rather than absolute number, may be a more accurate measure of immune function for HIV-infected pregnant women (Brettle, 1995; European Collaborative Study and the Swiss HIV Pregnancy Cohort, 1997; Miotti, 1992). When comparing changes in CD4 count/percentage over time, there is no difference between HIV-positive pregnant and nonpregnant women (O’Sullivan, 1995), suggesting that pregnancy does not accelerate decline in CD4 cells. HIV RNA levels (viral load) remain relatively stable throughout pregnancy in the absence of treatment (Burns, 1998). However, recent data suggest that HIV-RNA levels increase during the postpartum period regardless of ARV treatment (although use of HAART appears to blunt the effect), possibly due to immune activation associated with hormonal changes or to unmasking of a pregnancy-related viral load suppression. The implications for risk of transmission and treatment recommendations in the early postpartum period are unclear (Cao, 1997; Truong, 2003; Watts, 2003). The increase in viral load postpartum does not appear to reflect a long-lasting effect of pregnancy on viral load (Minkoff, 2003).

B. CLINICAL COURSE OF HIV IN PREGNANCY

Most studies to date examining the impact of pregnancy on HIV disease have been small but have not shown significant differences in HIV progression or survival between pregnant women and nonpregnant women with HIV infection. A recent meta-analysis of seven prospective cohort studies found no overall significant differences in death, HIV disease progression, progression to an AIDS-defining illness, or fall in CD4 count to below 200/mm³ between cases and controls (French, 1998). A subsequently reported prospective study of 331 women with known dates of seroconversion were followed for a median of 5.5 years; during this time 69 women were pregnant. There were no differences in progression between those who were and were not pregnant during follow-up (Alliegro, 1997). In addition, a long-term observational study showed no difference in viral load, CD4, or clinical disease progression in women with repeat pregnancy, compared to those with only one pregnancy (Minkoff, 2003).

VII. Effect of HIV on Pregnancy Course and Outcome  TOP

Adverse pregnancy outcomes may occur secondary to underlying disease processes (or their treatment), as well as for unknown reasons. Approximately 10% of U.S. pregnancies end prematurely, and preterm birth is the leading cause of perinatal morbidity and mortality. Data have accumulated that HIV, especially when more advanced, may result in increases in certain pregnancy complications. However, results of studies are conflicting. Some studies suggest that HIV-infected women have an increase in other risk factors for adverse pregnancy outcome (such as smoking, drug use, poor prenatal care) and if these risk factors are controlled for, there is no independent effect of HIV on adverse outcomes (Lambert, 2000). Furthermore, concerns have been raised that antiretroviral treatment itself may increase some adverse outcomes in pregnancy (see discussion under Antiretroviral Treatment). A recent study of 497 HIV-infected pregnant women enrolled in a perinatal clinical trial found that risk factors for adverse pregnancy outcomes (preterm birth, low birth weight, and intrauterine growth retardation) in antiretroviral-treated women are similar to those reported for uninfected women (Lambert, 2000). Table 7-4 summarizes the relationship between common pregnancy-related complications and HIV (Brocklehurst, 1998a; D’Ubaldo, 1998; van Bentham, 2000; Ngweshemi, 2003; Dreyfuss, 2003; Coley, 2001; Ladner, 1998).

Both HIV and pregnancy may affect the natural history, presentation, treatment, or significance of certain infections, and these, in turn, may be associated with pregnancy complications or perinatal infection.

A. VULVOVAGINAL CANDIDIASIS

Pregnancy is associated with both increased rates of colonization and an increase in symptomatic infections with species of Candida. HIV infection is also associated with an increase in colonization and possible increased infection rates, especially with declining immune function (Burns, 1997; Cu-Uvin, 1999; Duerr, 1997; Schuman, 1998; Spinillo, 1994). Therefore, pregnant women with HIV infection may be particularly susceptible to yeast infections. Only topical azole agents should be used during pregnancy and should be given for at least 7 days. Prophylactic topical therapy should be considered during courses of systemic, especially broad-spectrum, antibiotics.

B. BACTERIAL VAGINOSIS

Bacterial vaginosis (BV) has been associated with several adverse pregnancy outcomes, including preterm labor and birth, premature rupture of membranes, low-birth-weight infants, chorioamnionitis and amniotic fluid infection, postpartum and postabortal endometritis, and perinatal HIV transmission. HIV infection has been associated with increased prevalence and persistence of BV, and prevalence, persistence, and severity increase with lower CD4 cell counts (Jamieson, 2001). If BV is diagnosed during pregnancy, preferred therapies are metronidazole 250 mg po tid x 7 days or clindamycin 300 mg po bid x 7 days, since only oral agents have been shown to reduce preterm births in women with BV (Hauth, 1995; McGregor, 1995; Morales, 1994). Because BV is more common in the setting of HIV, and because both BV and HIV have been linked to increased risk of preterm birth, pregnant women with HIV should be regularly asked about signs or symptoms of vaginal infection and, if present, evaluated for possible BV. Infection should be treated if identified. Currently, there are insufficient data to suggest that screening for and treating BV during pregnancy in the general population reduces the overall rate of preterm birth (Berg, 2001). A recent meta-analysis (Caro-Paton, 1997) found no relationship between metronidazole exposure during the first trimester of pregnancy and birth defects.

C. GENITAL HERPES SIMPLEX

Primary herpes simplex virus (HSV) infection during pregnancy has been associated with spontaneous abortion and prematurity. Congenital or intrauterine infection is uncommon but maternal HSV shedding at delivery is associated with neonatal HSV infection, which is almost always symptomatic (including skin, eye, and central nervous system involvement, or disseminated infection involving multiple organ systems) and frequently lethal. The risk of neonatal herpes is greatest with primary HSV, especially when acquired close to delivery (30–50%), whereas only 0–3% of neonates become infected with recurrent maternal disease at delivery; however, because recurrent HSV is more common than primary disease, most neonatal infections are associated with recurrent HSV. Two thirds or more of mothers with infected infants are asymptomatic during pregnancy; only one third have a history of HSV in themselves or their sexual partner. Because most neonatal infection occurs during vaginal delivery, if genital lesions or prodromal symptoms are present at the time of labor or membrane rupture, cesarean section should be performed. Cesarean section is not indicated for recurrent HSV distant from the genital tract (e.g., thigh, buttocks) (ACOG, 1999b).

HIV infection, particularly with evolving immune compromise and higher plasma HIV viral load (Wright, 2003), is associated with increased HSV shedding and more frequent, severe, and prolonged episodes of genital or perianal herpes (Augenbraun, 1995). Higher doses and/or longer courses of antiviral agents may be required and suppressive therapy is often beneficial in nonpregnant individuals. Infection with HSV-2 is common among pregnant HIV-infected women and reactivation of herpes in labor occurs more frequently in the setting of HIV infection (Hitti, 1997).

Treatment of symptomatic HSV infections and suppressive therapy for frequent recurrences should be offered during pregnancy to HIV-infected women (USPHS/IDSA, 2003). The risk for herpes is high in infants of women who acquire genital HSV in late pregnancy and such women should be managed in consultation with an expert. The use of oral acyclovir prophylactically in late pregnancy has been shown to suppress genital HSV outbreaks and HSV shedding in HIV- women and may reduce the need for cesarean section for recurrent HSV (Brocklehurst, 1998a; Watts, 2001; Scott, 2001); however, this strategy has not been evaluated in HIV+ women, who are more likely to have HSV-2 antibodies and to have both symptomatic and asymptomatic reactivation of genital HSV. Therefore, use of acyclovir for the purpose of reducing the need for cesarean section in all HIV+ women is not recommended (USPHS/IDSA, 2003).

Acyclovir is the drug of choice for HSV therapy during pregnancy and there is no current evidence for increased risk for major birth defects or other adverse pregnancy outcomes (Reiff-Eldridge, 2000). While experience with valacyclovir is more limited, its safety profile is expected to be similar to acyclovir, since it is the prodrug of acyclovir. Experience with use of famciclovir in pregnancy is limited and exposures to this drug should be reported to the Famciclovir Registry at 1-888-669-6682. Documented HSV infections during pregnancy which do not respond to these agents should be managed with expert consultation.

Prevention of neonatal herpes should also emphasize prevention of acquisition of herpes in susceptible women in pregnancy. If her sexual partner has a history of oral or genital HSV infection, serologic evidence of HSV infection, or infection status is unknown, the pregnant woman should be counseled to avoid unprotected genital and oral sexual contact during pregnancy. Type-specific HSV serology may be useful to identify the pregnant woman at risk for HSV and to guide counseling, especially if her sexual partner has HSV infection. At the onset of labor, all women should be questioned carefully about HSV symptoms, including prodromal symptoms, and all women should be examined carefully for herpetic lesions, in order to make judicious decisions about the use of cesarean section.

D. HUMAN PAPILLOMAVIRUS

Genital warts may be more frequently seen and often enlarge and become friable during pregnancy and in some cases may mechanically obstruct the vaginal canal in labor; perinatal exposure can result in laryngeal papillomatosis in infants and children, although a recent prospective study suggests that the risk of perinatal transmission of human papillomavirus (HPV) is low (Watts, 1998). Both HPV infection in general and genital warts are more common in HIV-infected individuals, correlated with level of immunosuppression. Imiquimod, podophyllin, and podofilox should not be used in pregnancy. In women with large volume or bulk of genital warts treatment in late pregnancy with laser, excision, or cavitronic ultrasonic aspiration may be considered. Cesarean section is not currently recommended to prevent neonatal exposure to HPV, although in rare instances cesarean section may be indicated when extensive lesions obstruct the vagina. Pregnant women with abnormal Pap smears should undergo colposcopy and cervical biopsy, if indicated; increased bleeding may occur with biopsy during pregnancy. Endocervical currettage should not be performed during pregnancy. Pap smear should be repeated with or without colposcopy at 34–36 weeks gestation in women with initial abnormal Pap smear to rule out progression of dysplasia. Women with preinvasive cervical lesions can deliver vaginally, if otherwise appropriate; women with suspected invasive cervical cancer should be referred to a gynecologic oncologist.

E. SYPHILIS

Syphilis is more prevalent in HIV-infected populations and HIV may affect clinical manifestations, serologic response, or response to treatment for syphilis. Pregnancy does not alter the clinical manifestations of syphilis but untreated primary or secondary syphilis during pregnancy affects essentially all fetuses, with 50% rate of prematurity, stillbirth, or neonatal death (Radolf, 1999). Even with later stages of syphilis, there is a significant increase in adverse pregnancy outcomes, although the frequency and severity of fetal disease decrease with longer duration of untreated maternal infection. Manifestations of congenital syphilis in the newborn include mucocutaneous lesions, hepatosplenomegaly, osteochondritis/periostitis, jaundice, petechiae/purpura, and meningitis.

Congenital syphilis can generally be prevented by identification and appropriate treatment of syphilis during pregnancy. All pregnant women should have serologic testing for syphilis at the beginning of prenatal care and testing should be repeated at 28 wk gestation and at delivery, particularly in women who remain at risk for infection. Any woman with stillbirth after 20 wk gestation should be tested for syphilis. Development of neurologic symptoms mandates evaluation for possible neurosyphilis. Treatment of syphilis during pregnancy should be the penicillin regimen appropriate for the stage of syphilis, although a second injection one week after the first in cases of primary, secondary, or early latent syphilis should be considered, because of concerns about effectiveness of standard therapy in pregnant women and in the setting of HIV infection (USPHS/IDSA, 2003). HIV-positive women with late latent syphilis or syphilis of unknown duration should have cerebrospinal fluid examination before treatment. If there is concern for neurosyphilis, treatment should be with 7–10 days of IV penicillin (CDC, 2002).

Ultrasound evidence of hydrops or hepatosplenomegaly suggesting fetal syphilis increases risk for treatment failure and should be managed with expert consultation. Treatment in the second half of pregnancy is associated with the Jarisch-Herxheimer reaction in up to 40% of cases, with resulting premature labor and/or fetal distress (Myles, 1998); fetal and contraction monitoring for 24 hrs should be considered, especially in the setting of abnormal ultrasound findings, or, alternatively, patients should be advised to seek immediate attention after treatment if contractions or decrease in fetal movements occur (USPHS/IDSA, 2003). Pregnant women with a history of penicillin allergy should be skin tested and, if necessary, desensitized and treated with penicillin, because there are no proven effective alternatives to penicillin for treatment and prevention of congenital syphilis. Even with appropriate treatment of the pregnant woman with syphilis, fetal infection may still occur and neonates should be carefully evaluated for evidence of congenital infection.

Clinical and serologic follow-up should be performed in the third trimester and at delivery and at 3, 6, 9, 12, and 24 mo after treatment. Treatment failure should be managed with cerebrospinal fluid examination and retreatment. Some experts recommend monthly serologic testing after treatment after pregnancy with retreatment if there is a rise in titer.

F. CYTOMEGALOVIRUS

Cytomegalovirus (CMV) is the most common cause of congenital viral infection in the United States: .2–2.2% of liveborn infants acquire this infection perinatally (ACOG, 1993a). Most maternal CMV infections are asymptomatic but may cause a mononucleosis-like illness. Transmission can occur sexually or with injection drug use, because CMV has been recovered from virtually all body fluids. Transmission can also occur with oral contact with infected secretions (i.e. from children). Transmission of CMV from mother to infant may occur in utero (1–2% of infants born to women with CMV prior to pregnancy and up to 50% of infants born to women with primary infection during pregnancy), intrapartum (25–50% of exposed infants), and through breastfeeding (40–60% of exposed infants) (USPHS/IDSA, 2003). In general, it is in utero infection that results in significant neonatal/infant effects. Ninety percent of infected infants are asymptomatic at birth, but symptomatic infection is more likely with maternal infection acquired early in pregnancy. Even if asymptomatic, many infected infants subsequently develop deafness, mental retardation, or delayed psychomotor development. More severe clinical manifestations include symmetric growth restriction, hepatosplenomegaly, chorioretinitis, microphthalmia, hydrocephaly, microcephaly, and cerebral calcifications.

In the setting of HIV infection, in utero infection was detected in 4.5% of infants, compared to 1–2% transmission noted in previous studies of CMV-seropositive HIV-negative women. By six months of age, 40% of HIV-infected infants were CMV-seropositive compared to 15% of HIV-uninfected infants born to HIV-infected mothers (Kovacs, 1999). Because symptomatic infection in the newborn is usually associated with primary CMV infection of the mother during pregnancy, and because >90% of HIV-infected pregnant women are CMV seropositive in most studies, the risk of symptomatic infection in the newborn is low (Kovacs, 1999; Mussi-Pinhata,1998; Quinn, 1987), and treatment of asymptomatic maternal CMV infection in order to prevent infant infection is not indicated (USPHS/IDSA, 2003). There have been some reports that cotransmission of HIV and CMV may be related to more rapid HIV progression (Kovacs, 1999; Mussi-Pinhata, 1998).

Testing for antibody to CMV should be considered in pregnancy, especially if the CD4 count is <100/mm³ for reasons of maternal health evaluation; however, seropositivity is common and does not preclude viral shedding during pregnancy and perinatal transmission. Methods to reduce risk of exposure to CMV include safer sexual practices, careful handwashing, and transmission of only CMV antibody-negative blood products. Primary prophylaxis is not routinely recommended; however, after CMV disease (retinal or invasive CMV disease), chronic suppression is indicated in pregnancy and should be continued with expert consultation concerning choice of agents. (See Opportunistic Infection Prophylaxis below.)

G. TOXOPLASMOSIS

Approximately one third of U.S. women have toxoplasma antibodies, reflecting prior infection. Primary infection occurs in approximately .1–.5% of pregnancies and places the fetus at risk for congenital toxoplasmosis. Congenital infection is more common when infection in the mother occurs during the third trimester (59% in third trimester vs. 9% in first trimester) but is generally more severe when occurring in the first trimester. Although the majority of infected infants are asymptomatic at birth, most will develop some sequelae of congenital toxoplasmosis; two thirds of infants infected after maternal first trimester infection have severe manifestations and 5% are stillborn or die in the perinatal period (ACOG, 1993a).

Congenital toxoplasmosis may affect all systems, but the most common findings are chorioretinitis, microcephaly, hydrocephaly, and cerebral calcifications.

Transmission of toxoplasmosis from a mother with antibody evidence of prior infection can occur in the setting of HIV infection (as opposed to in HIV-uninfected women), but does not seem to be common (0–3.7% in two studies), although there are limited data in more immunosuppressed mothers (European Collaborative Study and Research Network in Congenital Toxoplasmosis, 1996; Minkoff, 1997b).

Testing for IgG antibodies to toxoplasma is recommended for all HIV-infected individuals soon after the diagnosis of HIV is made and should be considered as part of prenatal testing in HIV-positive pregnant women. Primary prophylaxis and prophylaxis against recurrent disease in pregnancy are discussed below (See Opportunistic Infection Prophylaxis). Pregnant women with symptoms including fever, chills, malaise, lymphadenopathy, myalgias, and headache should be evaluated serologically for possible primary toxoplasmic infection. Evidence of primary infection or active toxoplasmosis should be evaluated and managed with expert consultation. Detailed ultrasound examination of the fetus should be performed in this situation to look for evidence of congenital toxoplasmosis. Infants born to women infected with HIV and seropositive for toxoplasma should also be evaluated for evidence of congenital toxoplasmosis if suspected by clinical presentation of the infant.

To prevent exposure to toxoplasmosis, pregnant women should be counseled to avoid raw or undercooked meat, wash hands after contact with raw meat or with soil, and wash fruits and vegetables well before eating them raw. Cats should preferably be kept inside and fed only canned or dried commercial food; litter boxes should be changed daily, preferably by someone who is not HIV-positive or pregnant.

H. HEPATITIS B

Approximately 300,000 new cases of hepatitis B virus (HBV) infection occur each year and more than 1 million Americans are chronic carriers. Most patients who become infected have complete resolution of infection and develop protective levels of antibody (anti-HBs). Chronic HBV infection develops in 1–6% of persons who are infected as adults; they are chronically HBsAg+ and are at risk of chronic liver disease, including cirrhosis and hepatocellular carcinoma (CDC, 1991). The presence of HBeAg indicates active viral replication and increased infectivity. HBV is transmitted parenterally, sexually, perinatally, and through household or institutional contact. Approximately one quarter of regular sexual contacts of infected individuals will become seropositive and sexual transmission accounts for 30–60% of new infections. Perinatal transmission, usually with intrapartum contact with maternal blood and genital secretions, occurs in 10–20% of women who are HBsAg+, but increases to approximately 90% if the mother is also HBeAg+. Chronic HBV infection develops in about 90% of infected newborns, who are at high risk of chronic liver disease (ACOG, 1998).

All pregnant women should be screened for HBsAg. Symptomatic acute HBV infection should be treated supportively, with special attention to maintaining blood glucose levels and clotting function. Risk of preterm labor and birth may be increased with acute HBV infection in pregnancy. Treatment of chronic HBV infection is generally not indicated in pregnancy. Antiretroviral therapy containing lamivudine (3TC) may potentially decrease risk of perinatal HBV transmission in women with high HBV DNA levels, although this has not been examined in the setting of HIV infection.

Infants born to women who are HBsAg+ should receive hepatitis B immune globulin and initiate HBV vaccination within 12 hr after birth. HBV vaccine can be safely administered during pregnancy and should be considered in women who are high risk (injection drug use, STIs, multiple sexual partners, household or sexual contact of HBV carrier) and are anti-HBs- or anti-HBc-negative, indicating susceptibility. Some experts argue for more liberal use of vaccination in HIV-infected individuals, because HBV infection in the setting of HIV infection increases risk for chronic HBV infection. HIV can impair response to HBV vaccine; therefore, testing for hepatitis B surface antibody is recommended 1–2 mo after the third vaccine dose. Full revaccination should be considered for those who are nonresponders (ACOG, 1998; Bartlett, 1999).

I. HEPATITIS C

Hepatitis C virus (HCV) infection is primarily transmitted by injection drug use, but may also be transmitted sexually. Approximately 50% of those with acute HCV infection develop biochemical evidence of chronic liver disease, and 20% or more ultimately have chronic active hepatitis or cirrhosis and are at risk for hepatocellular carcinoma (CDC, 1998). Coinfection with HIV increases the risk and speeds the rate of development of progressive liver disease (Graham, 2001; Soto, 1997). Cofactors influencing disease progression include age, low CD4 cell count, and history of alcoholism. There is evidence that HCV infection may also hasten progression of HIV infection (Piroth, 2000).

Women newly diagnosed with HIV in pregnancy should have testing for antibody to HCV by enzyme immunoassay; positive results should be confirmed with HCV polymerase chain reaction (PCR) and liver function abnormalities should be documented. Negative serologic screening associated with history of risk factors for HCV transmission or unexplained liver function abnormalities is an indication for performance of HCV viral RNA testing, especially with low CD4 cell counts. Studies have shown that serum transaminases tend to decrease during pregnancy in HCV-infected women, but may rise transiently postpartum, while HCV-RNA levels tend to increase during pregnancy (Gervais, 2000; Conte, 2000). Treatment of HCV infection aims to eradicate infection and prevent the long-term complications of progressive liver disease and generally includes combination therapy with interferon plus ribavirin. However, treatment is generally not recommended during pregnancy (USPHS/IDSA, 2003) and evaluation for treatment, including liver biopsy, can be delayed until three months or more after delivery to allow pregnancy-related changes in disease activity to resolve. Ribavirin is teratogenic at low doses in multiple animal species and both women and men of childbearing potential receiving ribavirin should be counseled regarding the need for effective contraception during and for six months after completion of therapy.

Women coinfected with HIV and HCV should avoid alcohol, both during and after pregnancy, because alcohol use increases risk of cirrhosis. Vaccination against hepatitis A, if the woman is anti-HAV-negative, is recommended because the risk for fulminant hepatitis associated with hepatitis A is increased in HCV-infected individuals; this vaccination may be given safely during pregnancy (ACOG, 1998; Bartlett, 1999).

The risk of perinatal transmission of HCV is significantly higher among HIV-infected compared to HIV-uninfected women and has been reported to be approximately 22% or an increase in relative risk of 1.7–7.5-fold (European Paediatric Hepatitis C Virus Network, 2001). This may be related to higher HCV RNA levels seen in the setting of HIV infection, since perinatal transmission in both HIV-infected and –uninfected women is related to higher plasma HCV RNA levels (Yeung, 2001), although HCV transmission is highest with higher HCV RNA levels in the setting of HIV (Thomas, 1998). Furthermore, maternal coinfection with HIV and HCV may also increase risk for perinatal HIV transmission (Hershow, 1997). Scheduled cesarean section may reduce the risk of HCV transmission among HIV-coinfected women; in one large study scheduled C-section was associated with a reduction in transmission of almost two-thirds compared to other modes of delivery, although concomitant mother-to-child transmission of HIV was not controlled for (European Paediatric Hepatitis C Virus Network, 2001). Perinatal HCV transmission may be more likely in HIV-infected infants born to dually infected mothers (Papaevangelou, 1998).

VIII. Perinatal Transmission  TOP

The baseline rate of perinatal HIV transmission without prophylactic therapy is approximately 25%. The timing of transmission is a critical factor impacting on development of preventive interventions. There is evidence that transmission can occur during the course of pregnancy, around the time of labor and delivery, or postpartum through breast-feeding; however, two thirds to three quarters of transmission appears to occur during or close to the intrapartum period, particularly in non-breast-feeding populations (Mofenson, 1997).

POTENTIAL VARIABLES IN TRANSMISSION

A. HIV-RELATED FACTORS

• Plasma HIV RNA level: HIV RNA levels correlate with risk of transmission in both antiretroviral-treated and untreated women. The risk of perinatal transmission appears to be extremely low in women with undetectable plasma viral loads, but transmission has been reported at all levels of maternal HIV RNA. There is no upper limit of HIV RNA above which perinatal transmission always occurs (Garcia, 1999; Mofenson, 1999; Shaffer, 1999; Cooper, 2002).
• Strain variation (genotype): Each HIV-infected individual’s viral pool is composed of a variety of HIV quasispecies. One recent study found that in utero transmission was associated with transmission of major maternal viral variants, whereas intrapartum transmission was associated with transmission of minor maternal viral variants, suggesting that different selective pressures may be involved in determining the pattern of viral strain transmission depending on timing of transmission (Dickover, 2000). HIV in vaginal secretions can be derived from local expression and may have significant genotypic differences from plasma virus, with possible implications for perinatal transmission (Subbarao, 1998).
• Biologic growth characteristics (phenotype): Fetal blood mononuclear cells may be more susceptible to macrophage-tropic, non-syncytium-inducing HIV phenotypes and this may influence mother-to-infant HIV transmission (Palasanthiran, 1994; Reinhardt, 1995).
• Genital tract viral load: There is general correlation between plasma and genital tract viral load but discordance has been reported and may help explain some cases of transmission with undetectable plasma HIV RNA. In the Thai short-course zidovudine (ZDV) clinical trial, both plasma and cervicovaginal HIV RNA levels were suppressed by ZDV treatment and both were independently correlated with transmission (Chuachoowong, 2000). The female genital tract can also be a reservoir for virus with a different drug-resistance pattern than that observed in plasma (Fang, 1998). The use of HAART has been associated with undetectable HIV RNA levels in the genital tract and viral suppression in the genital tract may occur rapidly after initiating therapy (Cu-Uvin, 2000). It is possible that intra-cellular HIV in the genital tract can lead to transmission, even in the presence of antiretroviral treatment (Tuomala, 2003).
• Antiretroviral resistance: Special concerns have been raised about a possible increased risk for mother-to-child transmission associated with the potential development of antiretroviral resistance related to the use of single agents (i.e., ZDV or nevirapine) or dual nucleosides during pregnancy for perinatal prophylaxis. These regimens do not totally suppress viral replication, a common denominator in the development of antiretroviral drug resistance, since the process of reverse transcription necessary for viral replication is mutation prone. In addition, the increasing prevalence of resistance in both ARV treatment-experienced and newly-infected and treatment-naïve individuals (implying transmission of resistant strains) makes the relationship between drug resistance and perinatal transmission even more critical to understand.
  
  The presence of resistance mutations has been described in pregnant women and mother-to-child transmission of resistant virus has been reported, although it appears to be rare (Frenkel 1995; Johnson 2001). Studies to date of resistance mutations in the setting of ZDV monotherapy during pregnancy have shown increasing prevalence of ZDV mutations over time and an association with length of drug exposure and more advanced disease (Eastman, 1998; Welles, 2000; Sitnitskaya, 2001; Palumbo, 1999). Most studies, including PACTG 076, PACTG 185, Swiss cohort, or the PACTS, have not shown an increased risk of perinatal transmission associated with the detection of ZDV or other resistance mutations (Eastman, 1998; Kully, 1999; Palumbo 1999; Mofenson, 2002). However, in a recent Women and Infants Transmission Study (WITS) substudy, 25% of 142 maternal isolates from women receiving ZDV in pregnancy had at least one ZDV-associated resistance mutation and, on multivariate analysis, the presence of resistance mutations was independently associated with perinatal transmission (Welles, 2000). It has been pointed out that the characteristics of women in this cohort (mean CD4 count at delivery 315 cells/mL, usually did not receive ZDV in labor or for the neonate) and the factors associated with development of resistance (use of ZDV prior to pregnancy, higher HIV-RNA level, and lower CD4 count) illustrate the importance of current USPHS guidelines (discussed under Antiretroviral Therapy below) which advise the use of HAART in these circumstances, both for the woman’s health and for prophylaxis against perinatal transmission. There is also some evidence that ARV-resistant virus may have decreased fitness for transmission; in the WITS substudy, when a transmitting mother had a mixed viral population of wild-type and low-level resistant virus, only the wild-type virus was found in the infant, suggesting that virus with low-level ZDV resistance may be less transmissable (Colgrove, 1998).
  
  Selection of nevirapine-resistant virus has also been detected in women who received a single dose of nevirapine for prevention of perinatal transmission. In the HIVNET 012 clinical trial, ARV-naïve Ugandan women received one dose of nevirapine during labor and in 111 women who had detectable viral replication, 21 (19%) had genotypic mutations associated with nevirapine resistance at 6 weeks (Eshleman, 2001). However, maternal drug resistance was not associated with increased risk of perinatal transmission: rates of resistance were similar among mothers whose children were or were not infected.
• CD4 cell count: Lower CD4 count or decreased CD4:CD8 ratio have been consistently associated with increased risk of transmission.
• Maternal immune response: Studies have been inconsistent when evaluating the role of maternal antibodies, including anti-gp120, anti-gp41, anti-p24, and autologous neutralizing antibody titers. ß-chemokine and cytokine responses may affect risk of transmission (Pitt, 2000; Rich, 1998).

B. MATERNAL/OBSTETRIC FACTORS

• Clinical stage: Maternal symptomatic disease or AIDS-defining illness are consistently associated with higher risk for transmission. Women with primary HIV infection in pregnancy, at which time plasma viremia is high, are also at increased risk for transmission (Nesheim, 1996).
• STIs/other coinfections: STIs have been shown to increase genital tract HIV shedding and also increase plasma viremia (Plummer, 1998), both of which may increase risk for perinatal transmission. STIs (Mandelbrot, 1996), syphilis (M.J. Lee, 1998), bacterial vaginosis (Taha, 1998), and placental malaria have been associated with increased risk for vertical transmission, as have increased levels of genital tract inflammatory cells (Panther, 2000; Chandramohan, 1998).
• Vitamin A deficiency: Vitamin A deficiency has been associated with increased risk of perinatal HIV transmission and increased genital tract HIV shedding (Nimmagadda, 1998). However, a recent randomized trial of vitamin A supplementation in South Africa found no overall reduction in mother-to-child transmission of HIV, although vitamin A recipients were less likely to have a preterm delivery, and in preterm deliveries, those infants assigned to the vitamin A group were less likely to be infected (Coutsoudis, 1999). A controlled clinical trial in Malawi of vitamin A supplementation combined with iron and folate vs iron and folate alone also found no effect on HIV perinatal transmission (Kumwenda 2002), although vitamin A did improve pregnancy outcome (increased birth weight and decreased infant anemia). In another randomized trial of vitamin A or multivitamins (B, C, E) from 20 weeks gestation through completion of lactation, vitamin A was actually associated with increased risk of HIV MTCT. Multivitamin supplementation was associated with reduced child mortality and MTCT through breastfeeding among women with immunologic or nutritional compromise (Fawzi, 2002)
• Substance abuse: Illicit drug use during pregnancy has been associated with increased risk for perinatal transmission (Landesman, 1996; Lyman, 1993; Rodriguez, 1996).
• Cigarette smoking: Cigarette smoking has been associated with an increased risk of perinatal transmission (Burns, 1994; Turner, 1997).
• Antiretroviral therapy: Monotherapy with ZDV or with nevirapine, as well as dual nucleoside agents, have demonstrated effectiveness in reducing perinatal transmission in randomized clinical trials (see page 284 Table 7-8). In the PACTG 076 study (ZDV given antepartum/intrapartum/neonatal), reduction in plasma viral load accounted for only 17% of ZDV’s effectiveness, suggesting pre- and/or post-exposure prophylaxis as other possible mechanisms of action (Sperling, 1996). Although there are no completed clinical trials examining effectiveness of HAART regimens in reducing perinatal transmission, prospective cohort data from the WITS found that the protective effect of antiretroviral therapy increased with the complexity and duration of the regimen, and the use of HAART was associated with the lowest rates of transmission, 1.2% of 250 women (Cooper, 2002). Preliminary results from the PACTG 367 study, a combined retrospective and prospective chart analysis of over 2000 HIV-infected pregnant women at 67 U.S. clinical sites, found that in all subgroups of viral load, lowest transmission rates were seen with multiagent ARV therapy (Shapiro, 2004), both dual therapy and HAART.
• Sexual behavior: Unprotected sex with multiple partners has been associated with increased risk for perinatal transmission (Bulterys, 1997).
• Preterm delivery: Delivery at preterm gestational age has been associated with increased risk for perinatal transmission (Kuhn, 1997, 1999).
• Duration of membrane rupture: A recent metaanalysis from 15 prospective cohort studies, including almost 5000 deliveries, examined the role of duration of ruptured membranes in perinatal transmission (International Perinatal HIV Group, 2001). The likelihood of transmission increased linearly with increasing duration of ruptured membranes, with a 2% increase in risk for each hour increment. Women with clinical AIDS had the most pronounced increase in risk, with a 31% probability of vertical transmission after 24 hr of ruptured membranes. This study did not include women receiving HAART and did not control for viral load. The effect of duration of membrane rupture with very low viral loads is not clear.
• Placental disruption-abruption, chorioamnionitis: Clinical and histologic chorioamnionitis (Goldenberg, 1998; Mwanyumba, 2002) has been associated with increased risk of transmission. Placental abruption causing disruption of fetal-placental barrier and possible increased exposure of the fetus to maternal blood has also been suggested as a risk factor for transmission.
• Invasive fetal monitoring: Use of fetal scalp electrodes or fetal scalp sampling increases exposure of the fetus to maternal blood and genital secretions and may increase risk of vertical transmission (Maiques, 1999). Amnioscopy and amniocentesis increased risk in the French Perinatal Cohort (Mandelbrot, 1996).
• Episiotomy, forceps: Use of episiotomy or vacuum extraction or forceps may potentially increase risk of transmission by increasing exposure to maternal blood/genital secretions with trauma to maternal or neonatal tissue. On the other hand, judicious use of these techniques to shorten duration of labor or ruptured membranes with vaginal delivery may decrease likelihood of transmission.
• Vaginal vs. cesarean delivery: Several studies (done before routine use of viral load testing and use of combination antiretroviral therapy in pregnancy) indicate that cesarean delivery performed before the onset of labor and rupture of membranes significantly reduces the risk of perinatal HIV transmission by 55–80% (European Mode of Delivery Collaboration, 1999; International Perinatal HIV Group, 1999; Kind, 1998; Mandelbrot, 1998). Whether cesarean delivery offers any benefit when the mother is receiving HAART and/or if she has low or undetectable viral load is unknown, and cesarean section is not recommended in those circumstances (ACOG, 2000).

C. FETAL/NEONATAL FACTORS

Fetal/neonatal factors, including an immature immune system (particularly in the premature infant) and genetic susceptibility, as expressed by human lymphocyte antigen (HLA) genotype (Just, 1995) or CCR-5 receptor (a coreceptor for macrophage-tropic strains of HIV; a homozygous deletion in this gene confers a high degree of natural resistance to HIV sexual transmission) mutations may play a role in perinatal transmission (Kostrikis, 1999; Mangano, 2000; Philpott, 1999). A recent study from South Africa (Kuhn, 2000) found that early acquired cellular immune responses to HIV, presumably from in utero exposure, were present in over one third of 86 uninfected infants born to HIV-infected mothers. These detectable immune responses appeared to provide complete protection against subsequent HIV transmission at delivery and through breast-feeding.

D. BREAST-FEEDING

Overall breastfeeding appears to increase the risk of perinatal transmission by 5–20% (DeCock, 2000). In 1998 breastfeeding is estimated to have accounted for up to 50% of newly infected children globally (Fowler, 1999). Factors that have been associated with an increased risk of breast milk transmission include the following:

• Maternal factors:
  • acute HIV infection or recent seroconversion (Dunn, 1992), most likely related to high HIV viral loads
  • advanced HIV infection clinically or with low CD4 counts
  • high plasma or breast milk viral load
  • inflammatory breast conditions, such as mastitis or breast abscess
  • cracked nipples
  • vitamin A deficiency
  • colostrum
• Newborn factors:
  • Oral thrush
  • Other mucosal lesions due to trauma or infection
  • Preterm birth or low birthweight
  • Nutritional deficiencies
• Breastfeeding characteristics:
  • Timing – highest in first months, increases with longer duration of breastfeeding (Miotti, 1999)
  • Pattern of breastfeeding-mixed feeding (addition of other solids or liquids to breastmilk) associated with increased risk over exclusive breastfeeding (Coutsoudis, 2000)

A recent randomized clinical trial of breastfeeding vs. formula feeding in Kenya (Nduati, 2000) found that formula feeding prevented 44% of infant infections and was associated with a significantly improved survival.

STRATEGIES FOR PREVENTION OF PERINATAL TRANSMISSION

Based on the potential factors impacting perinatal HIV transmission discussed above, several basic approaches to prevention have been suggested. These include:

• Identification and treatment of modifiable risk factors
• Decreasing viral load
• Decreasing viral exposure
• Stimulation of the immune system (passive or active immunization)

Currently, most major efforts at prevention are aimed at decreasing viral load.

IX. Guidelines for Care  TOP

A. ANTEPARTUM

HISTORY/PHYSICAL EXAMINATION
(See also Chapter IV on Primary Medical Care.)

• HIV history: date of diagnosis; history of HIV-related symptoms or opportunistic infections or malignancies; lowest CD4 cell count; highest and current viral load; complete antiretroviral history, including specific drugs, side effects or toxicity, length of treatment, adherence, results of resistance testing (if performed), and response to treatment
• Pregnancy history: previous pregnancies and outcomes, complications, mode of delivery, use of antiretroviral prophylaxis, and HIV status of other children
• Signs or symptoms of HIV/AIDS: (initial and follow-up evaluations) assess signs or symptoms that suggest symptomatic HIV infection or AIDS (e.g., generalized lymphadenopathy, thrush, constitutional symptoms such as fever [38.5°C] or diarrhea >1 mo, herpes zoster involving two episodes or >1 dermatome, peripheral neuropathy, wasting, dysphagia, shortness of breath, persistent mucocutaneous herpetic ulcerations, cognitive dysfunction, etc.).
• Signs or symptoms of pregnancy-related complications: (the initial and follow-up evaluations) elevated blood pressure, significant edema, severe headache, vaginal bleeding or leakage of fluid, intractable nausea and vomiting, dysuria, abnormal vaginal discharge, persistent abdominal or back pain or cramping, decrease in fetal movement, etc. Gingival disease has recently been identified as a risk factor for preterm labor (Hill, 1998).
• Signs or symptoms of ARV toxicity: (initial and follow-up evaluations) nausea/vomiting, abdominal pain, jaundice, extreme fatigue, skin rash.
  Certain symptoms of HIV disease, antiretroviral toxicity, and normal or abnormal pregnancy may overlap, resulting in possible delay in appropriate diagnosis and management.
• Relevant family history of possible heritable diseases.

LABORATORY EXAMINATION BY TRIMESTER SEE TABLE 7-5.

Table 7-5: Laboratory Evaluation in the HIV-Infected Pregnant Woman

Test	Comment
Entry into Prenatal Care and Ongoing
HIV serology	Unconfirmed HIV infection; + test with other techniques; always repeat in the case of a new diagnosis to rule out false+
CD4 cell count/% HIV RNA	Repeat every 3–4 mo or as indicated to monitor changes with ARV therapy; at milestones for therapeutic decisions, re: ARV therapy/OI prophylaxis
Viral resistance testing (genotyping)	Acute HIV infection, virologic failure, sub-optimal viral suppression after initiation of ARV therapy, or high likelihood of exposure to resistant virus based on community prevalence or source characteristics
CBC	Repeat at minimum of every trimester in women on stable ARV therapy; for changes in regimen, repeat q 2–4 wk until stable; in general, consider more frequent testing if low or receiving marrow-toxic drugs (e.g., ZDV)
Serum chemistry panel (liver enzymes, electrolytes, +/- amylase)	Repeat at minimum of every trimester in women on stable ARV therapy; for changes in regimen, repeat q 2–4 wk until stable. Repeat as indicated with abnormal results or use of hepatotoxic/nephrotoxic drugs
Syphilis serology
Hepatitis serology: HBsAg, anti-HCV, anti-HAV	Order anti-HBs or anti-HBc and anti-HAV to screen for hepatitis B and A vaccine candidates. If anti-HCV+, order HCV-RNA
Rubella, Blood type and Rh, Antibody screen, Urine culture, GC/chlamydia testing, Pap smear	Cytobrush can be used
PPD	+ skin test = 5 mm induration; anergy testing not indicated; obtain CXR if at high risk for TB exposure
Hemoglobin electrophoresis, red blood cell indices	Perform in women at increased risk for hemoglobinopathies
G6PD	Optional — may consider screening black women or those receiving oxidant drugs (e.g., dapsone, sulfonamides)
CMV IgG	Consider especially with CD4 <100 mm3 or in patients at low risk for CMV (non-IDU)*
Toxoplasmosis IgG	Screen all patients with initial HIV diagnosis; repeat with CD4 <100/mm3 and not on TMP-SMZ, or with symptoms suggestive of toxoplasmic encephalitis
Urine toxicology screen	As indicated
Serum screening for Tay-Sachs disease	Consider screening both partners if at increased risk (Ashkenazi Jews, French-Canadian, or Cajun descent)
Bacterial vaginosis screening	Consider in women at high risk for preterm labor (previous preterm birth); women with signs/symptoms of vaginitis
16–20 wk
Ultrasound	Gestational dating, anomaly screen; repeat as indicated to monitor fetal growth
Maternal serum ∂-fetoprotein**	Voluntary; requires counseling; screening test for neural tube and abdominal wall defects; abnormal result (usually >2.5 multiple of the median) requires further evaluation
Triple screen (HCG, unconjugated estriol, ∂-fetoprotein)**	Voluntary; requires counseling; noninvasive test to determine risk of neural tube & abdominal wall defects, Down syndrome, and trisomy 18
24–28 wk
CBC, Syphilis serology, Antibody screen
Diabetes screen	Glucose 1 hr after 50 g glucola — 3 hr oral GTT if abnormal; may need additional glucose monitoring in women on protease inhibitors (consider 20 wk screening and repeat at 24–28 wk if on PIs)
32–36 wk
GC/chlamydia testing
Group B streptococcus culture (35–37 wk) (vaginal and rectal)	Recommend intrapartum chemoprophylaxis with IV PCN G (2.5 million units q 4 hr) if positive (or if GBS bacteriuria during current pregnancy or with previous infant with invasive GBS disease; if unknown GBS status, IP prophylaxis with delivery <37 wk gestation, membrane rupture 18 hr or IP temperature 100.4ºF/38.0ºC (Schrag, 2002).
CD4, HIV-RNA	Results may influence decisions about mode of delivery
Syphilis serology	Consider in high-risk patients or populations
Other Considerations
Liver enzymes, electrolytes	Assess more frequently in third trimester in setting of NRTI therapy, nevirapine, recent changes in ARV therapy
Serum lactate, electrolytes, liver enzymes; consider anion gap, CPK, amylase, lipase	Signs or symptoms suggest possible lactic acidosis in setting of NRTI therapy, especially if long-term
Fasting lipid profile	Consider at baseline and at 3–6 mo after starting PI- or NNRTI-based therapy; subsequent measurements based on initial results and risks
Liver enzymes (ALT, AST) every 2 wks in the first 6 wks of treatment, then every mo for first 12 wks; then every 1-3 mo.; more frequently in patients with pre-existing liver disease	Initiation of nevirapine therapy (does not apply to single dose prophylactic therapy in labor)

* Seroprevalence CMV IgG in US adults is 50–60%; IDU patients 90%
** Accurate gestational age is essential for interpretation of both tests.
** Though not yet considered standard of care, 1st trimester screening for genetic abnormalities can be done with nuchal lucency assessment on ultrasound and with biochemical markers.

ARV, antiretroviral; OI, opportunistic infections; CBC, complete blood count; PPD, purified protein derivative; G6PD, glucose 6-phosphate dehydrogenase; CMV, cytomegalvirus; TMP-SMZ, trimethoprim-sulfamethoxazole; HCG, human chorionic gonadotropin; GTT glucose tolerance test; PCN G, penicillin G; IDU injection drug use. PI, protease inhibitor; CXR, chest x-ray; R/o, rule out; GBS, Group B streptococci; IP, intrapartum CPK, creatine phosphokinase; NRTI, nucleoside reverse transcriptase inhibitor; AST, aspartate aminotransferase; ALT, alanine aminotransferase; PI, protease inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor

ANTEPARTUM FETAL SURVEILLANCE/TESTING

The general purpose of antepartum fetal testing and surveillance is to identify fetal abnormalities or compromise so that appropriate interventions can be undertaken to optimize fetal health and prevent fetal damage or death; or, in some instances, to aid in decisions regarding continuation of pregnancy (ACOG, 1999).

• Fetal surveillance: Indications include:
  • maternal conditions in which risk of fetal death is increased. This includes (but is not limited to) hemoglobinopathies, chronic renal disease, systemic lupus erythematosus, hypertension, and diabetes.
  • pregnancy-related conditions in which risk of fetal death is increased. This includes pregnancy-induced hypertension, decreased fetal movement, oligohydramnios, polyhydramnios, intrauterine growth retardation, postterm pregnancy, mild to moderate isoimmunization, previous fetal death, and multiple gestation.
  • HIV considerations: There are no data specifically on the need for and use of fetal surveillance techniques in the HIV-infected woman during pregnancy, and HIV per se is not an indication for fetal testing. However, HIV-infected women who have coexisting medical conditions placing the fetus at increased risk should have fetal surveillance; furthermore, HIV infection, especially when more advanced or associated with substance abuse, may be associated with increased risk for poor fetal growth, which places the fetus at increased risk. May consider in pregnant women on HAART particularly when containing newer agents with little experience of use in pregnancy. Need for fetal surveillance in the HIV-positive pregnancy should be determined on an individual basis.

Fetal surveillance techniques include:

• Fetal movement assessment: “kick-counts” – perception of 10 distinct movements in a period of up to 2 hr is reassuring.
• Nonstress test (NST): reactive or reassuring test is defined as two or more fetal heart rate accelerations (at least 15 beats/min above baseline and lasting at least 15 sec on fetal monitor) within a 20-min period.
• Contraction stress test (CST): negative or reassuring test is absence of late or significant variable fetal heart rate decelerations with at least three contractions (lasting at least 40 sec) within 10 min.
• Biophysical profile: consists of an NST combined with observations of fetal breathing, fetal movements, fetal tone, and amniotic fluid volume by real-time ultrasonography. Each component is given a score of 2 (normal or present) or 0 (abnormal or absent); a composite score of 8 or 10 is normal.
• Modified biophysical profile: combines NST and amniotic fluid index (AFI), which is the sum of measurements of the deepest amniotic fluid pocket in each abdominal quadrant; normal AFI is >5 cm. This test combines a short-term indicator of fetal acid-base status (NST) and an indicator of long-term placental function (AFI); placental dysfunction often leads to poor fetal growth and oligohydramnios.
• Umbilical artery Doppler velocimetry: evaluation of flow velocity wave forms in the umbilical artery; in the normally growing fetus, characterized by high-velocity diastolic flow; of benefit only in pregnancies complicated by intrauterine growth restriction.

Although there is no data from randomized clinical trials, antepartum fetal surveillance has been consistently associated with lower rates of fetal death than in untested pregnancies from the same institution or than historic controls with similar complicating factors. Testing should be initiated at 32–34 wk gestation, but may be started as early as 26–28 wk in pregnancies at very high risk. When the condition prompting testing persists, testing should be repeated periodically (weekly or, in some cases, biweekly) until delivery. Fetal reevaluation should also be repeated with significant deterioration in maternal medical condition or acute decrease in fetal movement, regardless of the time elapsed since the previous test.

NST, CST, biophysical profile, and modified biophysical profile are the most commonly used forms of testing and have a negative predictive value >99%. However, they are not predictive of acute events, such as placental abruption or umbilical cord accidents. On the other hand, the positive predictive value of an abnormal test can be quite low and the response to an abnormal result should be dictated by the individual clinical situation. Any abnormal test result requires further evaluation or action. Maternal perception of decreased fetal movements should be evaluated by NST, CST, biophysical profile, or modified biophysical profile. If normal, the mother can be reassured that the fetus is in no immediate danger. A nonreactive NST or abnormal modified biophysical profile is usually followed by additional testing with a CST or full biophysical profile. Management will be based on results of these tests, gestational age, degree of oligohydramnios (if assessed), and maternal condition. Oligohydramnios should prompt evaluation for membrane rupture. Depending on the degree of oligohydramnios, the gestational age, and the maternal medical condition, oligohydramnios warrants either delivery or close maternal/fetal surveillance.

• Ultrasound. Indications for obstetric ultrasound are many. Some of the more common include (ACOG, 1993b):
  • pregnancy dating
  • evaluation of fetal growth
  • evaluation of vaginal bleeding during pregnancy
  • determination of fetal presentation
  • suspected multiple gestation
  • significant uterine size/clinical dates discrepancy
  • pelvic mass