Acquired immunodeficiency syndrome (AIDS) was first described in 1981 when a cluster of patients was found to have defective cellular immunity and Pneumocystis carinii pneumonia (Gallo and Montagnier, 2003). Worldwide, it was estimated in 2003 that there were 40 million infected persons with HIV/AIDS, 5 million new cases of HIV, and 3 million HIV-related deaths (Clinton, 2003; Gayle, 2003; Steinbrook, 2004).

In the United States through 2001, the Centers for Disease Control and Prevention (2002b) estimated that there were 1.3 to 1.4 million infected individuals and almost a half million deaths. In 2002, HIV infection was the fifth leading cause of death in persons 25 to 44 years of age (Kochanek and Smith, 2004). Currently, a third of new cases result from heterosexual transmission, and two thirds of these are in women (Centers for Disease Control and Prevention, 2004b). Fortunately, the number of children with AIDS attributed to perinatal transmission has decreased by more than 90 percent in the past 10 years (Centers for Disease Control and Prevention, 2002b).

In addition, during this time, highly active antiretroviral therapy has led to an increasing number of people living with chronic HIV infection (

Fig. 59-5).

Figure 59-5. AIDS cases, deaths, and persons living with acquired immunodeficiency syndrome (AIDS) in the United States from 1985 through 2002. (From Centers for Disease Control and Prevention, 2002b.)

ETIOLOGY.
Causative agents of AIDS are DNA retroviruses termed human immunodeficiency viruses, HIV-1 and HIV-2. Most cases worldwide are caused by HIV-1 infection. Transmission of these viruses is similar to hepatitis B virus, and sexual intercourse is the major mode of transmission. The virus also is transmitted by blood or blood-contaminated products, and mothers may infect their fetuses.

PATHOGENESIS.
The common denominator of clinical illness with AIDS is profound immunosuppression that gives rise to a variety of opportunistic infections and neoplasms. Thymus-derived lymphocytes, or T lymphocytes, defined phenotypically by the CD4 surface antigen, are the principal targets. The CD4 site serves as a receptor for the virus. Co-receptors are necessary for infection, and two chemokine receptors -  CCR5 and CXCR4 -  have been identified to fill this role (Kahn and Walker, 1998; Sheffield and colleagues, 2005). After initial infection, the level of viremia usually decreases to a set-point, and patients with the highest viral burden at this time progress more rapidly to AIDS and death (Kahn and Walker, 1998).

Over time the number of T cells drops insidiously and progressively, resulting eventually in profound immunosuppression. Although it is thought that pregnancy has minimal effects on CD4+ T-cell counts and HIV RNA levels, the latter are often higher 6 months postpartum than during pregnancy (U.S. Public Health Service, 2003). Monocyte-macrophages may also be infected, and microglial brain cell infection may cause neuropsychiatric abnormalities. HIV-infected persons also have an increased incidence of neoplasms, notably Kaposi sarcoma, B-cell and non-Hodgkin lymphomas, and some carcinomas.

CLINICAL MANIFESTATIONS.
The incubation period from exposure to clinical disease is days to weeks. Acute illness is similar to many other viral syndromes and usually lasts less than 10 days. Common symptoms include fever and night sweats, fatigue, rash, headache, lymphadenopathy, pharyngitis, myalgias, arthralgias, nausea, vomiting, and diarrhea. After symptoms abate, the set-point of chronic viremia is established. Stimuli that cause further progression from asymptomatic viremia to AIDS are presently unclear, but the median time is about 10 years (Fauci, 2003).

When HIV-positive assay results are associated with any number of clinical findings, then AIDS is diagnosed. Generalized lymphadenopathy, oral hairy leukoplakia, aphthous ulcers, and thrombocytopenia are common. A number of opportunistic infections that may herald AIDS include esophageal or pulmonary candidiasis; persistent herpes simplex or zoster lesions; condyloma acuminata; tuberculosis; cytomegaloviral pneumonia, retinitis, or gastrointestinal disease; molluscum contagiosum; pneumocystis pneumonia; toxoplasmosis; and others. Neurological disease is common, and about half of patients have central nervous system symptoms. A CD4⁺ count of less than 200/mm³ is also considered definitive for the diagnosis of AIDS.

There are unique gynecological issues for women with HIV, such as menstrual abnormalities, genital neoplasia, other STDs, and contraceptive needs that overlap into pregnancy (Cejtin, 2003; Stuart and Castano, 2003). Repeated pregnancy does not have significant effect on the clinical or immunological course of viral infection (Minkoff and colleagues, 2003).

SEROLOGICAL TESTING.
The enzyme immunoassay (EIA) is used as a screening test for HIV antibodies as a part of a standard testing protocol. A repeatedly positive screening test has a sensitivity of more than 99.5 percent. A positive test is confirmed with either the Western blot or immunofluorescence assay (IFA). Results typically are available in several days.

Although highly specific, the Western blot is less sensitive than immunoassay because more antibody is required for a positive result. Thus, IFA can be used to resolve an EIA-positive, Western blot-indeterminate sample. According to the Centers for Disease Control and Prevention (2001a,b), antibody can be detected in most patients within 1 month of infection, and thus, antibody serotesting may not exclude early infection. For acute primary HIV infection, identification of viral p24 core antigen or viral RNA or DNA is possible. False-positive confirmatory results are rare (Centers for Disease Control and Prevention, 2001a).

In some cases, the first opportunity for HIV testing may occur when women present late for prenatal care or when they are in labor. Rapid tests can detect HIV antibody in 10 to 60 minutes at the point-of-care, and they have sensitivity and specificity comparable with EIAs (Branson, 2003; Centers for Disease Control and Prevention, 2004c,d). A reactive rapid screening test result should be supplemented with a confirmatory Western blot or IFA test. Peripartum and neonatal interventions to reduce perinatal transmission, however, may be based on the initial rapid testing results.

The Mother-Infant Rapid Intervention at Delivery (MIRIAD) multicenter study indicated that rapid HIV testing can be used to identify infected women and their neonates so that peripartum antiretroviral prophylaxis can be administered (Bulterys and colleagues, 2004). A set of practical guidelines for rapid HIV testing in labor and delivery is available from the Centers for Disease Control and Prevention (2004d).

SCREENING.
In 1991, universal, but voluntary, prenatal screening was recommended by the Institute of Medicine, the Centers for Disease Control and Prevention (2002c), the American Academy of Pediatrics, and the American College of Obstetricians and Gynecologists (1999b, 2002, 2004).

A study conducted from 1998 to 2001 in the United States and Canada reported that the rates of acceptance of HIV prenatal testing depend on the approach used (Centers for Disease Control and Prevention, 2002c). Under an opt-in approach, women receive pretest counseling and must specifically consent to HIV testing. In the opt-out approach, women are notified that an HIV test is part of a comprehensive set of antenatal tests and procedures, and they may decline testing (Stringer and colleagues, 2001). Areas that used or switched to the opt-out approach had much higher rates of prenatal serotesting. The Centers for Disease Control and Prevention (2004e) recommended adoption of the opt-out protocol into routine prenatal care. Sansom and associates (2003) suggest that it may be cost effective to repeat HIV testing in the third trimester in areas with a prevalence of 1 per 1000 person years or higher. Several states also recommend or require a second HIV test at delivery.

MATERNAL AND FETAL-NEONATAL INFECTION.
Mother-to-child transmission accounts for most pediatric HIV infections. Transplacental transmission can occur early, and the virus has been identified in specimens from elective abortion (Lewis and co-workers, 1990). According to Blair and associates (2004), pregnancy rates among women with HIV infection increased significantly in the current era of antiretroviral therapy compared with rates before 1996. In most cases, however, the virus is transmitted peripartum, and 15 to 40 percent of neonates born to nonbreast feeding, untreated, HIV-infected mothers are infected. Kourtis and colleagues (2001) have proposed a model for estimation of the temporal distribution of vertical transmission. They estimate that 20 percent of transmission occurs before 36 weeks, 50 percent in the days before delivery, and 30 percent intrapartum (

Fig. 59-6). Transmission rates for breast feeding may be as high as 30 to 40 percent.

Vertical transmission is more common in preterm births, especially those associated with prolonged membrane rupture. Analyzing data from the Perinatal AIDS Collaborative Transmission Study, Kuhn and associates (1999) reported a 3.7 relative risk for intrapartum viral transmission with preterm delivery. Landesman and co-workers (1996) reported that HIV-1 transmission at birth was increased from 15 to 25 percent in women whose membranes were ruptured for more than 4 hours.

Concurrent syphilis infection is common and is also associated with vertical perinatal HIV transmission (Koumans and colleagues, 2000; Schulte and associates, 2001). Finally, there is evidence that placental inflammation and chorioamnionitis may increase HIV-1 transmission by 3 percent, but it is unclear if antimicrobials decrease this risk (Mwanyumba and co-workers, 2002).

Perinatal HIV transmission can be most accurately correlated with measurement of maternal plasma HIV RNA burden (U.S. Public Health Service 2004b, 2003; Watts, 2002). As shown in

Figure 59-7, cohort neonatal infection was 1 percent with less than 400 copies/mL, and it was more than 30 percent when viral RNA exceeded 100,000 copies/mL (Cooper and colleagues, 2002a). Importantly, zidovudine therapy that reduced these levels to less than 500 copies/mL also minimized the risk of transmission. The investigators also reported that maternal infusions of HIV-1 hyperimmune globulin did not alter the risk of transmission.

According to the U.S. Public Health Service Guidelines (2003), maternal morbidity and mortality are not increased by pregnancy in seropositive asymptomatic women. Conversely, adverse fetal outcomes may be increased in these cases. In a review of 634 women delivered after 24 weeks, Stratton and associates (1999) reported that adverse pregnancy outcomes were common in HIV-infected women. Moreover, they reported that adverse outcomes were associated with a CD4+ cell proportion of less than 15 percent. In these otherwise asymptomatic women, the rate of preterm birth was 20 percent and fetal growth restriction was identified in 24 percent. Watts (2002) emphasized that these adverse outcomes were even more prevalent in developing countries.

Figure 59-6. Estimated rates of perinatal human immunodeficiency virus (HIV) transmission for different times of gestation and delivery in nonbreast feeding populations. Estimates are based on a hypothetical cohort of 100 children born to HIV-infected women without any interventions. Numbers in blue indicated number of children at risk for infection. (From Kourtis and colleagues, 2001.)

Figure 59-7. Incidence of perinatal human immunodeficiency virus (HIV) infection plotted against plasma HIV-1 RNA levels in 1542 neonates born to mothers in the Women and Infants Transmission Study. (Data from Cooper and colleagues, 2002a.)

MANAGEMENT DURING PREGNANCY.
Counseling is mandatory for HIV-positive women. This is preferable early in pregnancy, and if the woman chooses to continue pregnancy, ongoing counseling for psychological support is important.

The availability of an increasing number of antiretroviral agents and the rapid evolution of new information has introduced substantial complexity into HIV therapy. In addition, separate strategies are used for pregnant and nonpregnant adults. Therapeutic goals are a maximal and enduring suppression of viral load and restoration and preservation of immunological function.

For nonpregnant patients, treatment is generally offered to adults with less than 350 CD4⁺ T cells/mm³ or plasma HIV RNA levels exceeding 55,000 copies/mL (

Table 59-4). A more aggressive strategy has evolved for the management of HIV infection in pregnancy, because RNA level reduction is paramount to reduce perinatal transmission. Antiretroviral therapy should be offered to all HIV-infected pregnant women to begin maternal treatment as well as to reduce the risk of perinatal transmission regardless of CD4⁺ T-cell count or HIV RNA level.

The many approved antiretroviral agents can be grouped into several classes and used to design antiretroviral regimens (

Table 59-5). Antiretroviral agents may be combined into non-nucleoside reverse transcriptase inhibitor (NNRTI)-based, protease inhibitor (PI)-based, triple nucleoside reverse transcriptase inhibitor (NRTI)-based, or fusion inhibitor (FI)-based regimens (

Table 59-6). The new FI agents -  cell membrane fusion inhibitors -  block HIV-1 cell membrane binding and cell entry. One member of this class, enfuvirtide, has been used in salvage regimens with resistant HIV-1 in settings in which therapeutic options are limited (Kilby and Eron, 2003). Its efficacy and safety in pregnancy are unclear. Information on the safety and toxicity of antiretroviral agents has been reviewed by Money (2003) and also is available as a supplement to the U.S. Public Health Services guidelines (2004a,b, 2003) and through the web site http://AIDSinfo.nih.gov, which is frequently updated.

The perinatal guidelines (U.S. Public Health Service, 2003) recommend measurement of CD4⁺ T-lymphocyte counts approximately each trimester, or about every 3 to 4 months. HIV RNA levels should be monitored 4 weeks after initiation of a change in treatment, then monthly until undetectable, then every 3 months, and finally near term for planning delivery. These results are used to make decisions to alter therapy, to decide route of delivery, or to begin prophylaxis for opportunistic infections such as P carinii pneumonia.

Even with treatment, the incidence of perinatal complications is increased. Lorenzi and colleagues (1998) reported that 78 percent of women treated with two reverse transcriptase inhibitors had one or more adverse events, especially preterm delivery. Half of the neonates in this study had adverse events. Newer drug regimens may diminish these complications. For example, Tuomala and associates (2002) found no association with preterm birth, low birthweight, or stillbirths when combination antiretroviral therapy was used. They did, however, show that women given combination PI regimens had an increased risk of very-low-birthweight infants. Despite this, combined therapy should not be withheld if indicated (Watts, 2002).

At least two follow-up studies of children from the Pediatric AIDS Clinical Trial Group (PACTG) 076 Study found no adverse effects in children at 18 months and up to a mean of 5.6 years after zidovudine exposure (Culnane and associates, 1999; Sperling and colleagues, 1998).

Treatment failures may be due to nonadherence, inadequate drug potency, suboptimal levels of antiretrovirals, or viral resistance. Poor adherence to therapy appears to be a significant problem in pregnancy. In one study of 549 women from New York, poor compliance was reported in a third of women (Laine and associates, 2000).

Most HIV-infected women require additional medical care during pregnancy (Minkoff, 2003; Watts, 2002). They are given hepatitis B, influenza, and pneumococcal vaccines, ideally after viral suppression is achieved. Frequent monitoring is necessary for antiretroviral clinical and laboratory toxicities during the initial 1 to 2 months of treatment; this is particularly important with nevirapine regimens. Thereafter, surveillance for hepatic toxicity and lactic acidosis may be decreased. Reviewing a variety of regimens in the PACTG Protocol 316, Watts and co-workers (2004) found moderate-grade toxicity in 5 percent of women. Gestational diabetes varied by treatment regimen and was highest with early combination PI treatment. These latter women should have early glucose intolerance screening. Careful surveillance is important for antiretroviral drug interactions with other antiretrovirals, as well as therapies for opportunistic infection, methadone, and tuberculosis (Piscitelli and Gallicano, 2001). Information about this issue in pregnancy has been reviewed by Minkoff (2003) and Watts (2002). More detailed updates are available at the previously cited web site (http://AIDSinfo.nih.gov).

The management of some HIV complications can be altered by pregnancy. If the CD4+ T-cell count is below 200/mm3, primary prophylaxis for P carinii pneumonia is recommended with sulfamethoxazole-trimethoprim or dapsone. Pneumonia is treated with oral or intravenous sulfamethoxazole-trimethoprim or dapsone-trimethoprim. Other symptomatic opportunistic infections that may develop include toxoplasmosis, herpesvirus, mycobacteria, and candida. The U.S. Public Health Service, the Infectious Diseases Society of America, and the American Thoracic Society have published guidelines for prevention and treatment of opportunistic infections in persons infected with HIV (Centers for Disease Control and Prevention, 2002a,d, 2004f).

Prevention of Transmission. Precautions for antepartum, peripartum, and pediatric care of infected mothers and newborns are similar to those for hepatitis B, with avoidance of exposure to blood and body fluids. Antepartum antiretroviral therapy can be continued intrapartum. Optimal management of labor is uncertain, but if labor is progressing with intact membranes, artificial rupture and invasive fetal monitoring are avoided. Labor augmentation is used whenever needed to shorten the interval to delivery.

Pooled data from several prospective studies of health care workers suggest that the average risk of HIV transmission is 0.3 percent after percutaneous exposure to infected blood and 0.09 percent after mucous membrane exposure (Gerberding, 2003). Important for obstetricians, suture needles have not been implicated as a source of infection in prospective studies. However, occupational infection has occurred among surgical personnel. For health care workers with significant percutaneous or mucosal HIV exposure, postexposure prophylaxis is recommended (Centers for Disease Control and Prevention, 2001c; Gerberding, 2003).

Prevention of Vertical Transmission. The two principal approaches suggested for prevention of maternal-neonatal transmission of HIV infection are antiretroviral therapy and cesarean delivery. When standard treatment regimens are combined with zidovudine perinatal prophylaxis shown in

Table 59-7, vertical transmission is substantively lowered (Minkoff, 2003; Watts, 2002). The lowest transmission rates are achieved using prenatal combination therapy; they were 1 and 2 percent in two prospective studies reported by Cooper (2002a), Dorenbaum (2002), and all their colleagues. Short-term toxicity appears to be minimal for most women and newborns.

The main long-term problem is development of resistance. This is associated most commonly with prenatal combination therapy with lamivudine and with peripartum nevirapine (Clavel and Hance, 2004; Mandelbrot and associates, 2001; Moodley and co-workers, 2003). At present the known risks are clearly outweighed by the potential benefits of prophylaxis.

Indications for resistance testing for pregnant women are the same as for nonpregnant adults and include acute HIV-1 infection, virological failure, suboptimal viral suppression after initiation of antiretroviral therapy, and high likelihood of exposure to resistant virus based on community prevalence. Women who have presumed or documented zidovudine resistance and are on regimens that do not include zidovudine should still be given intravenous zidovudine intrapartum, and their neonates are givenoral zidovudine (

see Table 59-7).

If women on combination therapy temporarily discontinue treatment during pregnancy, all drugs should be stopped and restarted simultaneously to reduce the potential for emergence of resistance.

A number of clinical scenarios have been anticipated by the U.S. Public Health Service perinatal guidelines (2003):

1. If maternal HIV RNA level is more than 1000 copies/mL, then combination antiretroviral therapy is indicated.

2. When the maternal HIV RNA level is less than 1000 copies/mL, either combination antiretroviral therapy or zidovudine monotherapy can be given.

3. For women with no treatment prior to labor, intrapartum prophylaxis is appropriate with zidovudine, zidovudine with lamivudine, zidovudine with nevirapine, or nevirapine alone.

4. If delivery occurs before treatment is given, the newborn can receive prophylaxis for 6 weeks with zidovudine, or in some cases combination antiretroviral treatment.

A multistate surveillance program determined that by 1996, 80 percent of HIV-infected women were identified before delivery and that most received antepartum and intrapartum prophylaxis (Wortley and colleagues, 2001). By 2001, nearly 60 percent of American women were aware of effective intervention strategies to prevent perinatal HIV transmission (Anderson and associates, 2004). This rapid adoption of clinical practices clearly has resulted in the dramatic reductions in perinatal transmission to the current levels of 1 to 2 percent.

Cesarean Delivery. A meta-analysis of 15 prospective cohort studies by the International Perinatal HIV Group (1999) included 8533 mother-neonate pairs. Vertical HIV transmission was reduced by about one half when cesarean was compared with vaginal delivery. When antiretroviral therapy was given in the prenatal, intrapartum, and neonatal periods along with cesarean delivery, the likelihood of neonatal transmission was reduced by 87 percent compared with that of other modes of delivery and without antiretroviral therapy. The European Mode of Delivery Collaboration (1999) as well as others had similar findings. Based on these, the American College of Obstetricians and Gynecologists (2000) concluded that scheduled cesarean delivery should be discussed and recommended for HIV-infected women whose HIV-1 RNA load exceeds 1000 copies/mL. Scheduled delivery was recommended as early as 38 weeks to lessen the chances of premature membrane rupture. Data are insufficient to estimate any benefits of cesarean delivery for women whose HIV RNA levels were below 1000 copies/mL. Given their lower rate of transmission, it is unlikely that scheduled cesarean delivery would confer additional risk reduction.

Others have expressed concern that morbidity may be significantly increased in HIV-infected women undergoing cesarean delivery (Read and co-workers, 2001). Stringer and colleagues (1999) as well as Star and associates (1999) made a plea for restraint regarding routine cesarean delivery for prevention of vertical transmission. They concluded that combination antiretroviral therapy alone may reduce the risk to less than 2 percent. Our experiences at Parkland Hospital are similar, and perinatal transmission in women taking combination antiretroviral therapy has been 1 to 2 percent.

Breast Feeding. Vertical transmission is increased by breast feeding which generally is not recommended in HIV-positive women in this country (Read and co-workers, 2003). The probability of HIV transmission per liter of breast milk ingested is estimated to be similar in magnitude to heterosexual transmission with unsafe sex in adults (Richardson and colleagues, 2003). The risk is related to the maternal HIV RNA level, HIV disease status, breast health, and duration of breast feeding (De Cock and associates, 2000; John-Stewart and co-workers, 2004). Most transmission occurs in the first 6 months, and as many as two thirds of infections in breast-fed infants are from breast milk. In the Petra study (2002) from Africa, the prophylactic benefits of short-course perinatal antiviral regimens were diminished considerably by 18 months of age due to breast feeding. The World Health Organization (2001) has recommended continuing breast feeding promotion with early weaning by 6 months for women living in developing countries in which infectious diseases and malnutrition are the primary causes of infant deaths.

Postpartum. Many otherwise healthy women with normal CD4⁺ T-cell counts and low HIV RNA levels may discontinue treatment after delivery and be closely monitored according to adult guidelines. Psychosocial support is essential during this time, especially while awaiting diagnostic testing for pediatric infection. Contraceptive needs are complex and also may entail condoms in discordantly infected couples. There are effects on hormone levels with oral contraceptives and possibly with injectable agents (Stuart and Castano, 2003). Intrauterine devices may be an acceptable choice in some women with normal immunocompetence and a low risk for STDs. Gynecological care is complex in HIV-infected women and has been comprehensively reviewed recently by Cejtin (2003).