Herpes, Varicella, and Rubella

Herpesvirus is a large, linear, double-stranded DNA virus. All herpes viruses have similar structural elements arranged in concentric layers.¹ The virion is composed of the DNA-containing core, an icosahedral capsid, and an outer lipid-containing membrane or envelope.² The envelope consists of glycoproteins, lipids, and polyamines. Herpes simplex virus type 1 (HSV-1) and herpes simplex virus type 2 (HSV-2) have considerable overlap in their glycoproteins, but unique glycoproteins exist for each virus that allow differentiation with the use of restriction enzyme analysis.³ The viral envelope recognizes specific sites on the host cell membrane, and envelope glycoproteins mediate attachment of the virus to host cells.

After attachment, the virus enters the cell by pinocytosis. Once the nucleocapsid is within the cytoplasm of the cell, the virus is uncoated, and the viral genome is released and delivered to the nucleus. Viral genes are transcribed by cell polymerases in the nucleus and then exported to the cytoplasm, where HSV protein synthesis occurs. Whole viruses are released by exostosis or cell lysis. During an active infection, herpesviruses ultimately kill the cells in which they replicate.³

During the initial active infection, herpesvirus travels inside the axon of the nerve cell bodies to the sensory and autonomic ganglia of that nerve region. Replication occurs in the ganglia and nerve cells. The virus then migrates back to the skin by the peripheral sensory nerve branches of that sensory ganglion. The virus is also probably capable of replication in the skin and mucosa, where it produces focal necrosis and ballooning degeneration of cells, multinucleated giant cells, and eosinophilic intranuclear inclusions.³

After this primary infection, latent infection occurs.⁴ All herpesviruses possess the ability to establish latency, persist for extended times, and reactivate either spontaneously or in response to a stimulus. The viral DNA of HSV persists in the host for the lifetime of the patient. During latency, viral replication ceases, normal cell activity takes place, and the virus exists within the host cell without killing it. Periodically, the virus will reactivate to produce reinfection. During reactivation, the virus travels down the sensory nerve to the dermatome served by the infected sensory ganglion and is shed from the skin or mucosa. Viral shedding often is asymptomatic, but it also can result in a recognized recurrent infection in which symptoms and visible skin lesions develop.

Several factors influence recurrent infection: the severity and extent of the primary infection, the length of time since a primary infection, and the HSV type. Rates of recurrent infection are highest after a severe and extensive primary infection and within the first several months after a primary infection. HSV-2 more readily establishes latent infection in sacral ganglion than does HSV-1. Thus, both symptomatic and asymptomatic reactivation of HSV-2 infection is more frequent in the genital area than is HSV-1 infection. HSV-2 causes 70% to 75% of primary genital infection and up to 88% of recurrent genital HSV infection.⁵

Herpesvirus is not eliminated by antiviral agents or by the host's defense mechanisms. The virus remains protected from antiviral therapy when in the latent phase within the ganglion cells. Because latent virus probably does not replicate in the latent phase, it is not susceptible to antiviral drugs that affect viral DNA synthesis. The duration, severity, and frequency of recurrent episodes are reduced by antiviral treatment, but the virus is not killed. Thus, reactivation of the virus still occurs when antiviral therapy is stopped.⁶

The diseases caused by herpesviruses are endemic, and transmission usually requires intimate contact with mucosal surfaces or abraded skin.³ Approximately 90% of primary and recurrent genital herpes infection is caused by HSV-2 and the remaining 10% by HSV-1. Conversely, infections of the lips and mouth usually are caused by HSV-1. The prevalence of genital herpes depends on the population studied, the technique used for viral identification, and whether the diagnosis is made on the basis of clinical or serologic evidence.⁴ There is a discrepancy in the estimated prevalence of genital herpes that depends upon whether infection is defined by the prevalence of antibody or by a history of a clinical genital infection. Considerably more persons have antibody to HSV-2 than have a history of clinical infection, indicating that the majority of persons with antibody to the genital HSV-2 serotype have always had asymptomatic primary and asymptomatic recurrent infection.^{7, 8, 9}

HSV usually is transmitted sexually. Children less than age 15 have a low prevalence of HSV-2 antibody. Rates of HSV-2 antibody begin to rise after the onset of intercourse. Estimates by the Centers for Disease Control suggest that the total annual incidence of a new onset of HSV-2 is 50/100,000 persons; however, among 20- to 24-year-old women the incidence of HSV-2 infection is up to 210 new cases each year. As expected, HSV-2 seroprevalence rates vary substantially with the adult population studied: from 4% of college students to 46% of women attending an urban sexually transmitted disease (STD) clinic.¹⁰ Usually HSV-2 seroprevalence rates are between 16% and 33%. As shown in Figure 1, seroprevalence rates begin to rise at ages 15 to 19. Seroprevalence rates are higher among blacks than whites in all age groups. Female blacks have about six times the seroprevalence rate as female whites in all age groups.¹¹ The relative odds of HSV-2 antibody also was found to be twice as great for black women as for black men.¹¹ Among whites, there is no difference in HSV-2 antibody prevalence between women and men.¹¹ Socioeconomic level has not consistently correlated with HSV-2 antibody; however, seroprevalence rates are highest among black divorcees and those with a history of gonorrhea and syphilis.^{11, 12, 13} HSV-2 infection is also more common among homosexual or bisexual men than among heterosexual men, and it is more common among HIV-positive men than among HIV-negative men: Both homosexual men and HIV-positive men have a 20% higher prevalence of antibodies to HSV-2 than do heterosexual men and HIV-negative men.^{14, 15}

In contrast to HSV-2, HSV-1 infection starts to occur very early in life among young children who acquire HSV-1 gingivostomatitis. Also in contrast to HSV-2, which is not consistently correlated with socioeconomic status, HSV-1 antibody prevalence varies substantially by socioeconomic groups. Seroprevalence rates of HSV-1 increase dramatically from age 1 to 4 and peak by 15 to 20 years of age among female blacks. Seroprevalence rates for whites continue in a straight line until the peak at age 60. Thus, by age 15, approximately 70% of black girls but only 25% of white girls have antibody to HSV-1.^{16, 17} Overall, antibody to HSV-1 is present in 75% to 90% of persons in lower socioeconomic groups by age 15 to 20, compared with only 30% to 40% of persons in middle to upper socioeconomic groups by age 25. The high incidence of relatively asymptomatic primary HSV-2 in pregnancy may be explained in part by the high prevalence of HSV-1 antibody, particularly among women in lower socioeconomic groups.^{17, 18} Symptoms and signs of a primary HSV-2 infection in patients with preexisting HSV-1 antibody are considerably less than when a primary HSV-2 infection occurs in patients with no preexisting HSV-1 antibody (primary first-episode infection).

Genital HSV results after genital contact with an infected person. Most sexual partners, however, do not report having had a typical episode of clinically recognizable HSV around the time of viral transmission. In one study, one third of sexual partners reported having had a recent typical genital HSV infection with pain and genital ulcers, one third reported having had symptoms of atypical herpes infection, and the remaining third reported never having had symptoms.¹⁹ Further, spread from one partner to another is not inevitable.²⁰ It is estimated that among discordant sexual partners in whom one has HSV-2 antibody and the other does not, the annual seroconversion (infection) rate is 7%.¹⁹

HSV is the most common cause of genital ulceration, accounting for 20% to 50% of ulcerative lesions among patients attending STD clinics.⁴ Vesicles are particularly typical of herpes. However, the differential diagnosis of a genital ulcer also includes syphilis and chancroid. HSV infection produces an extremely wide range of clinical symptoms dependent upon the viral type, antibodies to heterologous herpes and prior genital HSV infection.²⁰ It is important to reiterate that 50% to 70% of first-episode HSV infections are asymptomatic, or have symptoms unrecognizable to the patient.^{19, 21}

Several definitions are necessary to explain differences among clinical syndromes. Primary first-episode genital herpes is characterized by a negative clinical history for prior genital HSV infection in a patient with no preexisting antibodies to either HSV-1 or HSV-2. Nonprimary first-episode genital herpes is characterized by the first-time acquisition of genital HSV-2 infection in a patient with preexisting antibodies to HSV-1. Recurrent genital herpes is characterized by the reactivation of HSV infection by the same viral type that previously infected the patient (e.g., HSV-2 reinfection in the presence of HSV-2 antibodies). In a cross-sectional study of 648 patients with genital herpes, 286 (45%) had first-episode infection, whereas the remaining 55% had recurrent disease.²² Of the primary genital infections, 209 (73%) were primary first episodes and 76 (27%) nonprimary first episodes.²² A higher proportion of patients with recurrent genital herpes have either mild atypical symptoms and signs or none.²³ It is important to recognize that in research settings, patients can be taught to recognize atypical symptoms and signs with reasonable accuracy, which means that with proper instruction, asymptomatic and atypical episodes of genital HSV can be recognized.²⁴

Primary Genital Herpes

As mentioned, about three quarters of primary genital infection is caused by HSV-2. Patients with primary genital HSV-1 or HSV-2 have a similar frequency of systemic symptoms, genital symptoms, and duration of lesions.²² During a primary infection, HSV-2 cannot be distinguished from HSV-1 on clinical findings. The incubation period of HSV extends from 2 to 20 days, but symptoms usually develop 3 to 7 days after exposure (Fig. 2). Symptomatic primary infection is characterized by prodromal symptoms of a feeling of tingling or itching in the genital area followed by painful multiple bilateral labial lesions, vaginal discharge, external dysuria, sacral paresthesia, and tender inguinal lymph node enlargement. Pain can be severe, but it ranges widely and it can even be nonexistent. Local symptoms increase during the first 6 to 7 days, when the skin lesions are in the vesicular stage (see Fig. 2). Local symptoms continue for 1 to 2 days after healing has occurred.²²

Skin lesions develop first as erythematous papules, become vesicles, and then form ulcers with a very tender base.^{22, 23} The mean number of lesions is approximately 15. Skin lesions tend to coalesce into larger ulcers. Ulcers of the skin continue to develop in approximately 75% of patients for up to 10 days.²² Cervical lesions occur in approximately 90% of patients with primary genital herpes. The appearance of the cervix ranges from diffuse friability to extensive ulceration and necrosis.^{22, 24} Virus has been isolated from the cervix of 88% of women with primary HSV-2, and 80% with primary HSV-1 infection.¹⁹ Vaginal ulcers occur in only approximately 5% of women. In women with primary genital herpes, the mean duration of viral shedding from the cervix is similar to that from lesions of the vulvar skin.¹⁹ Lymphadenopathy appears in the second to third week of infection. Extragenital lesions of the hips, groin, and buttocks occur in approximately 20% of women from autoinoculation of the virus. Symptomatic pharyngitis can occur in association with primary genital herpes, and may be the presenting complaint in up to 20% of patients with either primary HSV-1 or HSV-2.²⁵

Lesions of a primary infection usually persist for 3 weeks (range, 2 to 6 weeks) and then heal completely. Viral shedding occurs for an average of 12 days (see Fig. 2), but shedding can occur during a primary infection for 3 weeks.²²

More than two thirds of patients with primary HSV infection have systemic symptoms including malaise, headache, low-grade fever, myalgia, and abdominal pain.²² Systemic symptoms usually reach a peak within the first 4 days of the onset of lesions (see Fig. 2). Women with primary HSV-2 infection commonly have neurologic symptoms of autonomic dysfunction (e.g., urinary retention, constipation, parenthesis) or transverse myelitis (e.g., lower extremity weakness). Symptoms of aseptic meningitis (e.g., stiff neck, headache, photophobia, fever) also are reported frequently.²² Both HSV-1 and HSV-2 have been isolated from cerebrospinal fluid, but HSV-2 isolation is much more common. Neurologic symptoms resolve with bed rest, analgesics, and hydration, and neurologic sequelae are unusual.²² Dissemination of primary mucocutaneous herpes is rare; this finding is suggestive of an immunocompromised patient. Dissemination of HSV can cause pneumonia, hepatitis, thrombocytopenia, or disseminated intravascular coagulation. Physicians should be aware that pregnancy can predispose a patient to disseminated herpes infection.^{4, 22, 26}

Subclinical, first-episode genital HSV is common. Up to 80% of women with HSV-2 antibody have no history of genital lesions.⁹ One third of 56 asymptomatic women with HSV isolated in early labor had serologic evidence of a subclinical primary first-episode genital infection.²⁷ For the most part, asymptomatic patients are not identified; this substantial rate of subclinical primary infection means that a large number of patients are unaware they have ever had HSV infection.²⁸

Nonprimary First-Episode Genital Herpes

Preexisting heterologous HSV antibodies are present in approximately 25% of patients who develop a first clinical episode of genital herpes. Nonprimary first-episode genital herpes differs from primary illness in several ways. Patients with nonprimary first-episode genital infection have fewer constitutional symptoms, fewer lesions, a shorter duration of pain, fewer newly developing lesions, a shorter duration of viral shedding, and a shorter duration of genital lesions than do persons with a true primary genital herpes infection. The mean number of days of pain, healing, and viral shedding is reduced by about 4 days in a nonprimary first-episode compared with a primary first-episode genital infection.²² As previously mentioned, in the lower socioeconomic groups 75% to 90% of patients have antibody to HSV-1.^{16, 17} Thus, in some populations, the majority of patients acquiring HSV-2 would have a nonprimary first-episode genital infection.

Recurrent Genital Herpes

Recurrence of genital herpes is most likely to occur soon after a primary infection and becomes less frequent over time. In one study, more than 90% of patients with a primary genital HSV-2 infection had a reactivation within the first year (Fig. 3).²² Some patients have multiple recurrent infections in the first year. Approximately 40% of patients have six or more recurrent episodes in the first year after a primary infection.

Recognizable prodromal symptoms of genital itching, irritation, and pain start about 1 day before lesions appear in approximately 50% of patients.²² The virus can be isolated at the time of prodromal symptoms and before the development of lesions. The frequency and severity of recurrence vary considerably among patients and even within a single patient. Compared with primary infections, however, most recurrent infections are accompanied by fewer lesions, less pain, and a reduced duration of pain and viral shedding. Constitutional symptoms are infrequent, occurring in only 12% of women with a recurrence.²² As with the initial genital infection, symptoms of recurrent genital herpes tend to be more severe in women than in men. Skin lesions typically occur during a 3-day period and are healed within 7 to 10 days. Viral shedding occurs for a mean of 4 days. It is important to recognize that many “atypical” genital ulcers represent active HSV infection.⁹ Recurrent genital herpes is characterized by symptoms, signs, and anatomic sites of infection localized to the genital region, and as mentioned, is more likely to be caused by HSV-2 than HSV-1.

In contrast to primary illness, only approximately 10% of women who present with recurrent genital lesions experience concomitant cervical infection.^{9, 22} Viral shedding from the cervix has been documented in 2% of patients when lesions are present remote from the genital area.²⁹ In 4% of cases, patients have asymptomatic shedding, or shedding when no genital lesions are visible. HSV transmission to sexual partners or neonates can occur as a result of asymptomatic shedding.

The salient differences among primary HSV-2, nonprimary first-episode HSV-2, and recurrent HSV infection among nonpregnant women are compared in Table 1. Similar differences would be expected among pregnant women. These differences help explain marked differences in rate of HSV transmission to the neonate during pregnancy. Transmission to the neonate occurs mainly from direct contact with genital HSV during labor and at birth. The likelihood of transmission to the neonate increases when there is a large concentration of virus and when there is no serum antibody from the mother. Because the fetus is in contact with the cervix much longer than with the vulva, virus from cervical lesions is especially likely to be transmitted to the neonate as opposed to virus from vulvar lesions. Systemic symptoms, often indicative of a viremia, are for the most part limited to primary infection. As can be seen in Table 1, the duration of pain, number of lesions, and surface area of the lesions progressively decrease in primary versus nonprimary first-episode versus recurrent HSV infection. Patients with primary first-episode HSV infection compared with those with recurrent disease have higher fetal transmission rates in part because of a 3-fold higher number of lesions (more virus), a 10-fold increase in mean surface area of lesions (more virus), a 4-fold increased duration of viral shedding from both genital and cervical lesions (longer time the virus is present), and a 7-fold increased rate of cervical shedding (a critical area for fetal exposure).²²

TABLE 1. Relationship Between Clinical Characteristics and Primary and Nonprimary First-Episode and Recurrent Herpes Infection

(Corey L, Adams HG, Brown ZA, Holmes KK: Genital herpes simplex virus infections: Clinical manifestations, course and complications. Ann Intern Med 98:958, 1983)

Neonatal HSV infection is the most feared consequence of genital HSV infection because it carries a high rate of mortality and morbidity. The increased prevalence of adult HSV between the 1960s and the 1980s was associated with a parallel increase in the prevalence of neonatal HSV in the United States.^{4, 30, 31} Annually, 1500 to 2000 cases of neonatal HSV occur in the United States. Approximately 86% of neonatal infections are acquired at the time of birth, 10% are acquired after birth, and 4% are congenital infections.³² HSV-2 causes about two thirds of neonatal herpes and HSV-1 the remaining third.

Genital HSV infection is much more common in pregnancy than is generally recognized. Only 2% to 5% of pregnant women are aware of a previous episode of genital herpes; however, characteristic of the asymptomatic nature of both first-episode and particularly recurrent infection, most persons carrying the virus do not recognize having had an episode of HSV infection. Overall, 25% to 35% of pregnant women have serologic evidence of prior HSV-2 infection.^{33, 34} The prevalence of antibody to HSV-2 in pregnant women is similar to the prevalence of HSV-2 antibody in nonpregnant women of reproductive age.¹¹ The prevalence of HSV antibody among women 30 years or older is 25% for whites and 60% for blacks.¹¹ Therefore, if only 2% to 5% of pregnant women are aware of a prior episode of HSV whereas 25% to 35% have antibody to HSV, approximately 80% of pregnant women with latent HSV infection do not report a history of a prior HSV infection and are unaware that they could possibly expose their neonate to HSV.

Neonatal HSV is most likely to occur as a result of a primary maternal infection, rather than from the 25% to 35% of women with a recurrent (latent) HSV infection. Only about 1 in 1000 women have HSV in the genital tract at delivery from a primary HSV infection. A recent study, however, reported that neonatal HSV occurred in 15 (41%) of 37 infants born of mothers with first-episode genital HSV infection at delivery, and most experts believe that approximately 50% of infants will become infected when exposed to a primary maternal infection (Table 2).^{27, 34, 35, 36, 37, 38, 39} The high risk of neonatal HSV from a primary versus a recurrent infection means that the majority of infected infants are born to mothers with first-episode genital herpes.²⁷ In this study by Brown and colleagues,²⁷ of 52 neonates exposed to HSV in whom the maternal antibody level was known, 18 (35%) were exposed to a first-episode genital HSV infection and the remaining 34 (65%) to recurrent HSV. NeonatalHSV infection, however, occurred in 6 (33%) of the 18 infants exposed to first-episode genital herpes versus only 1 (3%) of the 34 infants exposed to recurrent genital herpes (Fig. 4).

TABLE 2. Risk Estimates for Neonatal Herpes with Various Exposure Conditions

*Assumed attack rates.
†Probability for positive culture varies among reports, ranging from 0% to 22%.
(Hensleigh PA: Genital herpes in pregnancy. In Parer JT (ed): Antepartum and Intrapartum Management, pp 108–120. Philadelphia, Lea & Febiger, 1989)

The high rate of neonatal infection during a primary maternal infection is related to an increase in the concentration of virus, a larger surface area of lesions, and a high degree of cervical shedding during primary versus recurrent genital infection, as previously mentioned (see Table 1). A high rate of neonatal infection with exposure to primary HSV also is related to the lack of a protective effect from HSV serum antibody. HSV antibodies neutralize the virus, mediate cytotoxicity, and produce type-specific inhibition. Infected neonates had much lower HSV antibody titers compared with their mothers or exposed but uninfected neonates.¹⁸ Homologous serum HSV antibody also decreases the rate of disseminated neonatal infection.⁴⁰

The presence of even heterologous antibody found in cases of nonprimary first-episode infection influences both pregnancy and neonatal outcome. Adverse pregnancy outcomes occurred in 6 of 15 neonates born of mothers with primary firstepisode infections. One of five first-trimester infections resulted in spontaneous abortion because of herpetic chorioamnionitis, one of five secondtrimester infections resulted in preterm delivery, and four of five third-trimester infections were associated with preterm delivery, growth retardation, and neonatal herpes. No abnormalities were noted in the 14 neonates born of mothers with nonprimary first-episode infection.³⁶

The majority of neonatal infections will not only result from primary infection at or shortly before delivery,³⁴ but the majority of women with a primary infection will be asymptomatic at delivery. These asymptomatic women with a primary infection at delivery may have had a recent asymptomatic primary or, occasionally, a very recent primary infection where symptoms develop only after delivery. In the study by Brown and associates,²⁷ of 29 women who presented with first-episode genital HSV-2, 15 had a primary infection and 14 had a nonprimary first-episode infection. All six infected infants were born of mothers with a primary first-episode infection in which asymptomatic shedding of HSV-2 at delivery accounted for nearly all the cases of neonatal infections. In a second study by Prober and co-workers,³⁸ none of 34 neonates born to mothers with a history of recurrent genital HSV-2 acquired neonatal herpes. In 31 cases of neonatal herpes, 26% of mothers had a history of recurrent genital herpes, 29% had primary infections, 35% had antibody to HSV-2 but no clinical history of genital herpes, and in 10% of cases the mother was not the source of the infection.^{18, 34, 36}

As many as 80% of women with a history of recurrent infection have a recurrence of HSV during pregnancy. Pregnancy does not appear to have much influence on HSV recurrence, although the mean number of recurrences has been reported to increase from 0.6 in the first trimester to 1.6 in the third trimester.^{34, 41} The majority of documented recurrences are associated with external genital lesions.⁴ As in nonpregnant women, HSV-2 causes the majority of primary and nonprimary first-episode genital infections and nearly all recurrent infections in pregnant women.

Most of the recurrences in pregnancy are asymptomatic^{35, 41}; thus, asymptomatic shedding of HSV also is common in pregnancy. The prevalence of positive cultures, mostly from asymptomatic shedding, at any time during pregnancy for women with known recurrent HSV can be as high as 1% to 2% on any one day.^{35, 36, 41} At delivery, a mean of approximately 0.2% of all pregnant women will asymptomatically shed virus,⁴² and approximately 1% of serologically HSV-2-positive pregnant women have HSV isolated from the genital tract.^{34, 43} The site of asymptomatic shedding was the usual site of lesions in 50% to 80%; in 20% to 50% of women, the site was the cervix.^{34, 43} Women at risk of asymptomatic shedding are difficult to identify. Asymptomatic shedding at delivery cannot be predicted on the basis of prior antenatal culture data. Only 1 of 17 women with asymptomatic shedding during pregnancy had viral shedding at delivery, and that woman had a visible lesion.⁴¹ Five other women had asymptomatic shedding of HSV at delivery but were culture-negative during the pregnancy. The duration of asymptomatic shedding is only 1 to 2 days.²⁰ The attack rate of neonatal HSV infection from a mother with recurrent HSV infection (with type-specific antibodies)³⁸ is estimated to be 2% (see Table 2; range, 1% to 5%). A word of caution is in order for patients with a recent primary infection because they are more likely to shed HSV asymptomatically than are those who had HSV before becoming pregnant. Patients with a recent primary infection are also more likely to shed HSV asymptomatically from both the cervix and the vulva than are those with longstanding infection, who tend to shed from the vulva only.^{34, 35, 41}

A summary of the estimated risk of developing neonatal herpes related to exposure to a primary or recurrent lesion or asymptomatic shedding is depicted in Table 2.³⁹ The probability of a positive genital culture is very high with primary HSV infection, approaching 100%. The attack rate of neonatal HSV is estimated to be 50% when exposure to virus in a primary HSV lesion occurs. In contrast to primary HSV, the odds of HSV transmission to the neonate after exposure to a recurrent lesions is about 100-fold less. Only approximately 20% of cultures from recurrent lesions contain HSV, and with the attack rate estimated to be 2%, approximately 1 in 250 of neonates exposed to a recurrent HSV lesion would develop neonatal HSV. Approximately 2.5% of patients with HSV lesions remote from the genital area shed HSV from the genital area. If 5% is used as a high end of the range of attack rate for recurrent HSV, about 1 in 800 neonates exposed to a remote HSV lesion would develop HSV. Asymptomatic shedding of the virus at delivery occurs in approximately 1% of women with recurrent HSV. The attack rate of neonatal herpes is not known, but the attack rate of neonatal herpes is probably even lower in the setting of asymptomatic shedding than with a recurrent HSV lesion. The attack rate is no more than 5%, but it may be as low as 1%. Thus, the rate of neonatal herpes from asymptomatic shedding in patients without lesions would be approximately 1 in 5000 to 1 in 2000 exposures (0.01 × 0.01 to 0.01 × 0.05) This low rate of neonatal infection makes routine cesarean section unnecessary when HSV lesions are not present in the patient with a history of recurrent HSV.

In some reports, primary HSV infection also has been associated with spontaneous abortion.^{36, 37} Congenital infection from the hematogenous spread of virus to the placenta and fetus is not likely, but if it occurs, it does so during the viremia of a primary first-episode infection.³⁶ Criteria to establish the diagnosis of congenital infection include identification of infection within 48 hours of birth, lesions at birth, and exclusion of other pathogens (e.g., cytomegalovirus, syphilis, toxoplasmosis). Neonates with congenital infection also may have elevated levels of immunoglobulin M (IgM) in the cord blood as well as viremia. Congenital infection is unusual, but it was reported to have occurred in 9 (5%) of 192 neonatal infections.⁴⁴

Symptomatic first-episode HSV infection also has been associated with preterm birth, although the number of cases reported has been too small to permit an accurate assignment of the risk. Preterm delivery occurred in as many as 33% to 56% of women with symptomatic primary infection during pregnancy,^{36, 37} including four of five cases involving a primary third-trimester infection.³⁶ Women with an asymptomatic first-episode HSV infection during pregnancy seem to have no increased rate of adverse outcome.⁴⁵

Pregnant women possibly have an increased risk of disseminated HSV infection. Dissemination of HSV can cause serious hepatic and central nervous system disease. Signs of disseminated HSV in the mother should be treated with intravenous acyclovir to reduce the severity of maternal disease.

Management of Recurrent Herpes Simplex Virus in Pregnancy

Pregnant women with recurrent HSV often have periodic outbreaks of genital HSV, and their outbreaks may even become slightly closer together in late pregnancy, but there is a limited chance of an HSV outbreak at the time of labor. Cesarean section for delivery is indicated only for an active HSV lesion, but only about 1 in 500 women in labor have HSV in the genital tract from a recurrent HSV lesion.²⁸ Cesarean section is unnecessary for the patient with prior recurrent HSV who has no prodromal symptoms and no lesions at the onset of labor or with rupture of membranes because the rate of neonatal infection is so small (see Table 2). Women with a history of recurrent HSV should be reassured that both cesarean and neonatal infection are uncommon in this setting.

The routine practice of weekly viral cultures in the last month of pregnancy does not predict asymptomatic viral shedding among patients known to have had genital HSV.^{35, 41} The previous practice of weekly viral cultures for women with known recurrent HSV is no longer recommended.⁴⁶ None of the rapid tests to identify HSV in the asymptomatic woman without lesions is sensitive enough to exclude reliably the presence of HSV, and rapid tests are not indicated for women in labor. It is possible that oral acyclovir suppresses recurrent HSV or asymptomatic shedding, but the use of acyclovir is not recommended for this purpose because of the lack of data on efficacy and safety.⁴² Symptomatic infection is reduced with the use of acyclovir, but asymptomatic shedding still can occur.⁴⁷

Patients with prior recurrent HSV should be advised to report to the hospital early in labor and shortly after membranes rupture for a detailed physical examination to exclude the presence of herpetic lesions. All women with a known history of prior HSV infection require a detailed physical evaluation during labor or upon rupture of membranes to establish whether or not HSV lesions are present. It has been recommended that HSV cultures be obtained from women with a history of prior HSV, even in the absence of lesions.⁴² The quality of these screening cultures for HSV in the absence of lesions has not been established and do not represent the standard of care. A single Dacron applicator can be used to swab the cervix, the vulva, and the genital area usually involved in recurrent lesions. The applicator is placed in transport media and processed as a single culture. A negative culture is reassuring. A positive culture identifies neonates potentially exposed to HSV, but there is a 48-hour waiting period for culture results. It is not known whether prophylactic treatment of asymptomatic infants exposed to HSV is of benefit. These infants do benefit, however, from very close evaluation for early symptoms of neonatal HSV. Some physicians also perform surface and oropharynx cultures of the neonate.⁴² Serial cultures of high-risk infants after birth, including infants born of mothers with an HSV-positive culture taken during labor, also are recommended by some pediatricians. An HSV-positive culture from the neonate 24 hours after delivery is considered evidence of neonatal infection, rather than contamination from secretions from the mother. These infants should be considered candidates for acyclovir treatment.

Approximately 1 in 250 neonates exposed to an active recurrent HSV lesion will acquire HSV (see Table 2), and cesarean section is recommended for the few patients with recurrent HSV who have active genital lesions at the time of labor or rupture of membranes. Cesarean might also be considered if typical prodromal symptoms are present in patients with no genital lesions. Cesarean section should be done as soon as possible after membrane rupture. Cesarean delivery within 4 to 6 hours of membrane rupture reduces the rate of HSV transmission to neonates of mothers with an active recurrent HSV lesion.³⁷ Cesarean section does not prevent all cases of neonatal HSV, even when it is performed within 4 to 6 hours of membrane rupture or even before membrane rupture. In a recent survey, 15 (8%) of 184 infants with neonatal HSV were delivered by cesarean before membrane rupture.¹³ However, patients with active lesions generally are delivered by cesarean section, even after a prolonged period of membrane rupture, because cesarean delivery may slightly reduce HSV transmission to the neonate. If the cesarean section is performed 6 hours or more after membrane rupture, the mother and neonate should be cultured for HSV and counseled that neonatal infection could still occur.

Management becomes problematic for the patient in labor with an active extragenital HSV lesion present some distance from the genital area. Viral shedding from the cervix occurs in approximately 2.5% of women with an extragenital lesion, which is similar to the 3% rate of cervical HSV shedding in patients with recurrent vulvar lesions.^{34, 39, 41} Extragenital HSV lesions can be covered, and because the risk of cervical shedding is estimated to be only 2.5%, vaginal delivery can be considered; however, because the estimated risk of neonatal herpes is still 1:800 (see Table 2), there is no unanimous opinion on the management of delivery in these cases. Many experts recommend that a cesarean section be performed when extragenital lesions are present, and this recommendation appears prudent.⁴²

The management of women with active recurrent HSV and membrane rupture before term is particularly difficult. The physician must weigh the chance of a preterm infant developing an ascending HSV infection versus the chance of serious morbidity or death from respiratory distress syndrome. If the fetus has a low chance of respiratory distress syndrome (i.e., the fetus is more than 34-weeks' gestation) an immediate cesarean section would be appropriate. If the fetus is very immature and has a high chance of either respiratory distress syndrome or death, or both, all of the following can be considered:

1. Delaying delivery until betamethasone has been given (even when there has been a membrane rupture)
   
2. Performing a cesarean section, relying on the administration of topical surfactants to the neonate after birth
   
3. Administering acyclovir therapy to reduce the chance of neonatal HSV exposure.
   

None of these three management regimens have been used extensively, and each case requires individual consideration. A summary of the management of patients with recurrent genital HSV is provided in Table 3.⁴²

TABLE 3. Considerations for the Management of Pregnant Women with Recurrent Genital Herpes

1. Sequential cultures during late gestation are not indicated unless they are done after clinically evident recurrent infection to document cessation of virus excretion.
   
2. Cesarean delivery should not be suggested as a “prophylactic” measure to women who do not have active genital lesions at delivery.
   
3. Symptomatic recurrences of genital herpes during the third trimester will be brief; vaginal delivery is appropriate if no active lesions are present at delivery.
   
4. Obtaining specimens for culture at delivery from women with a history of HSV infection may help in identifying exposed infants, but the clinical benefit of this procedure has not been established.
   
5. Options for the management of women with active genital herpes at onset of labor:
   1. If the membranes have ruptured, the mother is afebrile, and the fetal lungs are immature, the following steps may be taken:
      1. Delay delivery until betamethasone can be given.
         
      2. Proceed with cesarean delivery and give topical surfactant to the infant or
         
      3. Manage the patient expectantly with or without acyclovir therapy.
         
      
      
   2. If term gestation or fetal lung maturity has been established and the membranes have ruptured, cesarean delivery is indicated.
      
   
   

Management of Primary Herpes Simplex Virus Infections in Pregnancy

Neonatal exposure to a primary maternal HSV infection results in a 50% rate of neonatal HSV. First-episode infection is associated with a large number of viral lesions, often comprising a large surface area, and a 90% chance of cervical HSV infection. Thus, the neonate must be protected from exposure to primary HSV. Cesarean section should be offered to all women with active first-episode HSV, even if it was a nonprimary first-episode infection. Transplacentally acquired heterologous antibodies to HSV-1 does little to influence the acquisition of HSV-2 by the neonate.²⁸ Cesarean section should be done immediately in the patient has a membrane rupture. Although most infections of the neonate have occurred after the membranes have been ruptured for more than 4 to 6 hours, neonatal infection after exposure to primary infection has occurred when cesarean section was performed within 4 to 6 hours of membrane rupture.^{36, 41, 42} When the fetus is mature, most physicians will still perform cesarean section beyond 6 hours of membrane rupture for primary HSV because the neonatal infection rate is so high that any incremental reduction of this attack rate would be of benefit.³⁹ The guidelines for women with rupture of membranes in the preterm infant are similar between those with primary and recurrent infections. The much-increased chance of neonatal infection in pregnancies complicated by primary versus recurrent infection, however, needs to be considered; in cases of primary infection, there is an increased pressure toward early delivery.⁴²

Women with a first-episode HSV infection earlier in pregnancy should have a viral culture to document that HSV is the cause of the lesions and to type the strain as either HSV-2 or HSV-1. Accurate serology testing is available in only a few areas of the United States. Commercial serologic testing is discouraged because of inaccuracies in distinguishing between HSV-1 antibody and HSV-2 antibody.^{48, 49} Accurate differentiation of these antibodies has a potential advantage because many first-episode infections will be nonprimary ones, which are associated with less maternal morbidity and fewer complications of pregnancy than are primary infections. The patient with a primary infection should be advised of the limited data on spontaneous abortion, congenital infection, growth retardation, and preterm labor. It is not known whether acyclovir reduces this morbidity.⁴² Amniocentesis for viral culture is not indicated because of the poor predictive value of the results.⁴² To date, the data are too limited to use culture evidence as an indication for the presence of congenital infection: Congenital infection has occurred despite a negative amniotic fluid culture, and it has not occurred in some persons with a positive amniotic fluid culture.

A tenfold increased risk of asymptomatic shedding occurs for several months after a primary infection, compared with infections that have occurred 6 to 12 months or more earlier. ^{36, 41, 50} This high risk of viral shedding immediately after a primary infection means that HSV may still be present in the cervix or genital area if the patient goes into labor within 2 months of a primary HSV infection. Patients with primary genital HSV late in pregnancy should be followed with weekly cervical or vaginal cultures, or both. If viral shedding stops by the time of delivery and no lesions are present, vaginal delivery can be undertaken with the precautions in place for recurrent HSV, including careful examination for lesions and cultures of patients without lesions. A cesarean section should be performed, however, for the patient in labor with a recent primary HSV infection if viral shedding continues until delivery or if lesions are present. It may be wise to favor a cesarean section for women with a primary HSV infection within 4 to 6 weeks of delivery, given the difficulties in predicting HSV shedding immediately after a primary episode.⁴² These guidelines are not intended to represent a standard of care but are offered based on our experience.

Prevention During Pregnancy

Prevention of a primary episode of HSV during pregnancy would reduce the stage of HSV infection to which the infant is most susceptible. Sexual contact obviously should not occur when the sexual partner has an active HSV lesion or prodromal symptoms where the risk of infection increases further. Sexual contact with condom protection could resume once the lesions have healed. Condom use also should be considered for any sexual contact with a male partner with a known history of genital herpes to protect the pregnant woman from acquiring HSV during asymptomatic shedding by the male partner. There is a small but definite rate of primary HSV during pregnancy transmitted even by stable asymptomatic partners. In a study of private obstetric patients, 32% of women and 25% of their husbands had antibody to HSV-2. Two thirds of the women with HSV antibody had no history of genital herpes. Of the 190 couples, 73% were serologically concordant; of these couples, 57% were both antibody negative and 16% were both antibody positive. The discordant couples had been sexually intimate for a mean of 6 years.¹² Annually, 7% to 10% of sexual partners are expected to acquire genital herpes from a partner with a history of HSV, and two in three infections will be acquired from an asymptomatic sexual partner.^{12, 51} Unfortunately, the majority of sexual partners with latent HSV have always been asymptomatic and are unaware that they have ever had HSV in the first place. Rate of condom use by the asymptomatic male with no known history of prior HSV would be very low. Thus, prevention of much of HSV infection becomes problematic short of abstinence.

Postnatal infection appears to account for approximately 10% of neonatal HSV. The higher rate of HSV-1 isolation from neonatal infection (30%) than from genital infection (10% to 25%), is suspected to be related to postnatal transmission from adults with active oral HSV-1. Parents, other relatives, friends, and hospital personnel with active oral labial HSV should avoid direct contact with newborns when they have active lesions, and they should wash their hands with soap to reduce the chance of HSV transmission to the newborn from free virus contaminating the hands. Because postnatal HSV infection is preventable, all attempts should be made to reduce exposure by advising unsuspecting adults about the contagious nature of active oral lesions.

Neonatal Morbidity and Mortality

Neonatal HSV is one of the most life-threatening of all infections in newborns. Disseminated illness is associated with the greatest risk of death, the death rate ranging from 57% to 82%.^{32, 33} Antiviral therapy does little to modify this risk.³² The morbidity and mortality of neonatal HSV are influenced by several factors, classified into four groups:

1. Infection localized to the skin, eyes, or mouth (15%)
   
2. Infection primarily localized to the central nervous system (20% to 30%)
   
3. Disseminated infection with or without central nervous system involvement (70%)
   
4. Asymptomatic infection (rare).
   

Neonatal death depends most on the location of infection (Fig. 5). High death rates (60% to 80%) occur with disseminated infection, moderate rates (20% to 30%) with local central nervous system infection, and low rates (15%) with HSV localized to the skin or eye. Death rates are increased with concomitant prematurity, reduced level of consciousness, and disseminated intravascular coagulopathy.³² Permanent neurologic sequelae are common among survivors of neonatal HSV, but sequelae vary with the invasiveness of infection. The neurologic impairment in survivors is 63% after disseminated disease, 41% after localized central nervous system infection, and 6% after localized skin infections.³²

Inadvertent exposure of the neonate to HSV at delivery should lead to prompt notification of the primary caregiver and the parents. If exposure to a recurrent infection occurred, the rate of neonatal infection is low. The parents should be advised of symptoms and signs of lethargy, poor feeding, or lesions (Table 4). If exposure occurred to a primary maternal infection, where the risk of neonatal infection is high, it is advised that viral cultures be taken of the eyes, throat, urine, stool, and cerebrospinal fluid.⁴² Some physicians culture the infant's eyes, nose, mouth, and skin weekly thereafter for 4 to 6 weeks (see Table 4). Therapy using acyclovir is recommended if any HSV culture is positive 48 hours after delivery or the cerebrospinal fluid is abnormal.⁴² Acyclovir can be given prophylactically to an inadvertently exposed infant, but acyclovir appears to have limited ability to prevent all of the manifestations of neonatal infection. Acyclovir needs to be studied in this situation before it can be routinely recommended. The infrequency of transplacental passage of HSV and the problems of diagnosis of congenital HSV make identification and management of possible congenital HSV experimental.

TABLE 4. The Management of Infants Exposed to Herpes at Delivery

1. If the infant has been exposed to a first episode of maternal genital herpes, the following steps are recommended:
   1. Obtain specimens of urine, stool, and CSF as well as specimens from the eyes and throat for culture.
      
   2. Instruct parents to report signs of lethargy, poor feeding, fever, or lesions.
      
   3. If the result of a culture performed for an infant >48 hours of age is positive or if findings on examination of CSF are abnormal, initiate therapy with IV acyclovir.
      
   
   
2. If the infant has been exposed to recurrent maternal genital herpes, the following steps are recommended:
   1. Instruct parents to report signs of lethargy, poor feeding, fever, or lesions.
      
   2. Perform weekly surveillance cultures of specimens from the eyes, nose, mouth, and skin for 4–6 weeks after delivery on an optional basis.
      
   
   

Viral cultures provide the most sensitive and specific method to diagnose HSV. Cultures are expensive, and positive results usually take 1 to 2 days. The culture is not interpreted as negative until it has incubated for 4 days. HSV-1 and HSV-2 can be distinguished by culture techniques.⁵² To establish the diagnosis of HSV with certainty in pregnancy, cultures should be performed on persons with typical lesions and no prior history of herpes and on patients with atypical lesions. Cultures are not necessary during pregnancy for patients with atypical lesions and a prior history of herpes.

Vesicles are the lesions most likely to contain virus. Virus is isolated from approximately 90% of vesicles, approximately 70% of ulcers, and only approximately 25% of lesions with a crusted base.²² The vesicle is unroofed with a needle, the base swabbed, and the swab placed in transport media, which is maintained at 4°C until cultured. Freezing at -20°C kills the virus. For the asymptomatic patient in labor with a history of recurrent HSV, one swab should be used to culture the cervix, and the same swab should be used to swab vigorously the usual site of recurrence as well as the labia minora. The single swab is then placed in transport media.

The most sensitive rapid test uses monoclonal antibody to HSV. For vesicular and ulcerated lesions, the direct monoclonal antibody test for HSV is 65% to 90% more sensitive compared with the culture technique. Tests are available within 2 hours. Positive tests are helpful, but negative tests do not exclude HSV infection. Tsank and Papanicolaou smears are only approximately 50% sensitive compared with the culture, and these older rapid tests should no longer be used. A negative Tsank or Papanicolaou smear offers little reassurance that HSV is not present. The polymerase chain reaction test is the most promising for rapid identification, but this method is not yet widely available.⁵³ Rapid tests should not be relied on to detect asymptomatic viral shedding because in this setting the viral load is low and the sensitivity of rapid tests have not been determined.

Serologic tests can detect antibody to HSV. Most commercially available serologic tests consist of enzyme-linked immunosorbent assays or latex agglutination, which cannot distinguish between HSV-1 and HSV-2 antibodies. Reports of titers for HSV-1 and HSV-2 antibody using commercial serologic tests usually are inaccurate and misleading and therefore of little use in the management of HSV in pregnancy.⁴⁸ Type-specific serologic assays for HSV are available that utilize differences in the glycoprotein between HSV-1 and HSV-2 in an enzyme immunoassay or differences in antigens in a Western blot assay.⁴⁹ Both of these assays accurately distinguish between HSV-1 and HSV-2 antibodies. These assays have been used by researchers to distinguish primary from nonprimary first-episode infections and to study the effect of maternal antibody on neonatal outcome.

Acyclovir is the drug recommend for genital HSV. Acyclovir is activated by specific viral thymidine kinase, and activated acyclovir inhibits HSV-specific DNA polymerase. Inhibition occurs at lower concentrations than is required to inhibit the patient's cellular DNA. Acyclovir has 150 times the affinity for viral versus cellular DNA.⁵⁴ Peak levels appear to be approximately 50% lower in pregnant as opposed to nonpregnant patients because of the increased extracellular volume and renal excretion in pregnancy.^{54, 55} Fetal levels are lower, but breast milk levels are higher than serum levels.^{54, 55, 56} Acyclovir is well tolerated, even in pregnancy, but it should be used with caution in pregnancy until its safety is better defined.^{54, 57, 58} In 1988, 239 pregnant women with first-trimester exposure were reported from a postmarketing pregnancy registry.⁵⁹ The rate of congenital anomalies were at an expected level, but the number of women followed to date is small and more data are needed to establish safety for routine use in pregnancy. Another theoretic but unknown concern of acyclovir use is a blunting of the maternal antibody response, which could reduce the protective effect of antibody for the neonate.

Intravenous acyclovir, however, should be given to women with disseminated HSV to reduce morbidity. Acyclovir also could be considered in an attempt to reduce viremia when pregnant women have severe systemic symptoms during a primary infection. The oral dose is 200 mg, five times daily; the intravenous dose is 5 mg/kg every 8 hours. Systemic acyclovir promotes lesion healing, reduces viral shedding, and reduces new lesion formation. Acyclovir does not appear to reduce the frequency of recurrent HSV.^{47, 60}

For patients with recurrent HSV, oral acyclovir reduces the healing time by 2 days and viral shedding by 1 day.^{47, 60} This effect is too small to justify routine treatment of recurrent HSV at term in an attempt to prevent neonatal contact during labor. Nonpregnant patients with six or more annual episodes of HSV are candidates for suppressive acyclovir therapy. Suppressive regimens of 400 mg, two to three times daily, and 200 mg three to five times daily cause a 70% to 90% reduction in the frequency of recurrent HSV. Suppressive acyclovir regimens have been used in research settings in an attempt to reduce recurrent HSV among pregnant patients with frequent recurrences,⁶¹ but this treatment is currently experimental.⁴²

Varicella-zoster virus (VZV) is a herpesvirus that causes varicella (chickenpox) and herpes zoster (shingles). It most commonly occurs in children: 90% of VZV cases occur in persons less than 15 years of age. Infection develops after respiratory inhalation of or direct contact with varicella lesions. VZV has an incubation period of 11 to 21 days (mean, 15 days), which is followed by a transient asymptomatic primary viremia.⁶² During the primary viremia, VZV spreads from the lymph nodes to internal organs. After replication in these organs, a secondary viremia occurs, and by day 14, virus in the skin produces an exanthem.⁶³ Fever, malaise, intensely pruritic vesicles, and a rash then develop. The vesicles are produced in various stages: while one breaks open, others are forming. The rash lasts 7 to 10 days.

VZV is highly contagious, with an attack rate of 60% to 90% among susceptible persons and up to 90% among household contacts.⁶⁴ Adults are less susceptible to the virus than are children. Communicability ranges from 1 to 2 days before rash onset and up to 6 days after rash onset, or until all lesions are crusted.

VZV is usually a benign infection in children with normal immunity; however, it can be a serious infection in adults. The rate of varicella complications for adults is 9- to 25-fold higher than that for children. The most frequent complications are pneumonia and encephalitis. Although only 2% of varicella cases occur among adults older than age 20, adults constitute one quarter of all VZV-related deaths. Smokers are particularly susceptible to pneumonia secondary to varicella.⁶⁵

Virus-specific IgM antibodies develop in the first 5 days of infection and peak between 2 and 3 weeks. IgG antibody to varicella persists indefinitely. Approximately 95% of American women⁶⁶ and 84% of women from the subtropics have antibody to varicella; subclinical infection is rare. Because such a high number of persons are immune, the history of clinical varicella in an American woman virtually guarantees immunity.

Varicella during pregnancy is unusual, with rates estimated from 2/1000 to 5/10,000.⁶⁷ The rate and severity of complications from varicella in pregnant women, however, appear even higher than in other adults. The death rate among pregnant women with varicella ranges in the literature from 0.7%⁶⁸ to 2.4%.⁶⁹ Pneumonia occurs in up to 9% and premature labor in 9% of pregnant women with varicella.⁶⁹ Pneumonia is a serious VZV complication in all adults, but particularly in pregnant women, with death rates as high as 12% in nonpregnant⁷⁰ versus 35% in pregnant VZV patients.⁷¹ Pneumonia develops 1 to 6 days after rash onset, and symptoms range in severity from none to cough, dyspnea, pleuritic chest pain, hemoptysis, and sudden respiratory failure.

The probability of pneumonia development appears to correlate with the extent of the rash. Acyclovir should be given to pregnant patients with extensive rashes. Varicella pneumonia in pregnancy requires blood gases, supportive therapy, respiratory support in some cases, and intravenous acyclovir. Early administration of acyclovir reduces the mortality from VZV pneumonia in adults from 7% to 0%,⁷² and acyclovir given to pregnant women reduces the mortality from 40% to 14%.⁷¹

Varicella also can cause serious infection and death in term neonates who acquire varicella in utero and are born before they have had a chance to receive antibody passively from the mother. Term infants are susceptible to varicella when the mother develops a rash 4 days before delivery to 2 days after delivery. The attack rate with exposure during this time is 20%, and the mortality is 30% for infected neonates.^{73, 74} Administration of varicella-zoster immune globulin (VZIG) to infants exposed to VZV shortly after delivery ameliorates the disease and reduces mortality.^{73, 75} Infants born to mothers who develop a rash 5 or more days before delivery have a benign infection because of antibody transferred from the mother. Infants born to mothers who develop a rash 3 or more days after delivery have not been exposed to the in utero maternal viremia.

Spontaneous abortion^{68, 76} and preterm delivery^{68, 69} secondary to VZV appear to be uncommon. It is unusual for varicella in the first 20 weeks of pregnancy to cause stigmata of intrauterine infection. In fact, the frequency of stigmata from congenital varicella is too small for reliable estimates. Combining data from several studies, where the rate of congenital abnormalities ranged from 0% to 9%, a rough estimate is that approximately 2% of first-trimester congenital infections result in congenital abnormalities (95% confidence interval = 0.5% to 6.5%).⁷⁷

Well-documented cases of congenital varicella have been described with a variety of abnormalities that constitute the congenital varicella syndrome. The abnormalities associated with congenital VZV infection include skin lesions (vesicles and cicatricial lesions), hypoplastic limbs from viral infection of the ganglion nerve supply to these muscles, hydrocephalus, chorioretinitis, and optic atrophy.

After primary varicella infection, VZV enters the dorsal root ganglia to produce latent infection. Reactivation of the virus produces herpes zoster, which is characterized by a unilateral painful vesicular rash in the skin innervated by that dorsal root. The cervical, thoracic, and lumbar areas are affected most commonly, but facial nerve involvement with palsy can occur. Reactivation of VZV is uncommon in pregnancy, most common in the sixth decade, and associated with aging and immunosuppression.

Herpes zoster in the pregnant woman appears infrequently to cause neonatal abnormalities.⁶⁹ This is to be expected because mothers with herpes zoster have antibody that is passively transferred to the baby, and viremia is not present. In utero exposure to VZV also can produce neonatal herpes zoster, which is generally a mild disease.⁷⁸ Fetal herpes zoster early in pregnancy, however, can result in limb atrophy as a result of nerve damage to those muscles.

The rash of varicella is characteristic enough to provide an accurate diagnosis on the basis of appearance alone. Varicella can be recovered from vesicles by culture.

Serology can be used to determine susceptibility, but only a limited number of serologic tests are reliable.⁷⁹ Immunofluorescent assays (Pharmacia ENI, Merck and Co, West Point, PA) and enzyme-linked immunosorbent assays (Pharmacia ENI) are reliable, but complement-fixation tests are insensitive (i.e., suggest susceptibility in an immune individual).⁷⁹ To document acute infection, IgM can be measured within 3 days of acute infection. A fourfold rise in IgG titer can be documented in two sera specimens drawn 7 to 10 days apart. IgG can be formed as early as 7 days from the onset of symptoms.

The prenatal diagnosis of fetal varicella can be established with a combination of ultrasound, cordocentesis (IgM is present only after 20-weeks' gestation), or chorionic villi sampling using polymerase chain reaction identification of VZV before 20-weeks' gestation.^{80, 81}

VZV vaccine (Varivax) became available in 1995 for administration to susceptible persons. VZIG can be used within 96 hours of exposure to ameliorate symptoms. VZIG is expensive and in limited supply, so before deciding to administer VZIG, it should be determined whether the patient is likely to be susceptible, whether exposure is likely to result in infection, and whether an increased rate of complication is to be expected.

A history of chickenpox is reliable, and patients with such a history should be considered immune. Even without having a known history of chickenpox, most adults older than age 15 (85% to 95%) are immune. Thus, a history or age is usually sufficient to establish immunity, but serologic testing can be used in unusual circumstances.

Neonatal varicella infection is likely to occur after exposure to household contacts, playmates, or hospital contact⁷⁵ as well as to the mother during the contagious period. Candidates for VZIG include newborns of mothers who acquired varicella 5 days before to 2 days after delivery, premature infants less than 28-weeks' gestation with postnatal exposure regardless of maternal history (who may not yet have had antibody passed from the mother), pregnant women, and given the high complication rate, perhaps all susceptible adults.

VZIG reduces the rate of clinical infection and the severity of complications when infection ensues. It should be given to pregnant women to prevent maternal complications, rather than to prevent congenital infection, because it is not known whether VZIG effectively prevents neonatal infection.⁷⁵ Newborns born to mothers in whom varicella developed within 5 days of delivery should receive VZIG, even if VZIG was administered to the mother. The intramuscular dose is 125 U (per vial)/10 kg body weight, up to a total of five vials. VZIG should be given within 48 to 72 hours of exposure; it may not be effective when given after 96 hours from exposure. The neonatal dose is 125 U (one vial) given intramuscularly. Mothers or neonates with varicella should be isolated and discharged as soon as possible to prevent nosocomial spread.

The recommended intravenous dose of acyclovir to treat varicella pneumonia is 10 mg/kg body weight every 8 hours. There are too few data to support the use of acyclovir to prevent congenital varicella.

The rubella virus is a single-stranded RNA virus belonging to the Togavirus family. Rubella is only moderately contagious. It is spread by the inhalation of droplets from respiratory secretions or from direct contact. Rubella enters the respiratory tract and disseminates to regional lymph nodes, where it replicates. Viremia occurs 7 to 9 days after exposure; it is during the viremia stage that the placenta can become infected. Rubella causes a distinctive maculopapular rash. The infection is most contagious when the rash is erupting, but nasopharyngeal viral shedding continues to occur for 2 weeks. The rash is preceded by malaise, fever, headache, conjunctivitis, and postauricular and posterior cervical lymphadenopathy. The rash typically appears on the face and trunk and moves peripherally to the extremities within a 1- to 2-day period. One quarter to one half of all rubella infections are subclinical, and the rash can mimic other viral syndromes.

The rash appears as immunity develops and the virus disappears from the bloodstream, suggesting an immune-mediated mechanism. The most important determinant of the severity of rubella infection is the age at which it is acquired. Infections in children and adults are usually mild or subclinical, and only rarely do complications such as arthritis, encephalitis, hepatitis, or hemorrhagic complications occur in these persons. Severe fetal manifestations of congenital rubella, however, have been recognized since 1941.⁸²

Universal vaccination against rubella was implemented in 1969.⁸³ There has been a decline in reported cases of rubella among all age groups from 28/100,000 in 1969 to 0.23/100,000 in 1985. In persons 15 years of age and older, reported cases have declined from 4.8/100,000 in 1979 to 0.2/100,000 in 1985.⁸⁴

The current rubella vaccine is the live, attenuated strain RA2713.⁸⁵ The primary goal of vaccination was to control rubella infection among preschool-age and younger children, who are the primary source of transmission. The secondary emphasis was to vaccinate susceptible adolescents and adults, especially women, to decrease the risk of congenital rubella.^{85, 86, 87} Despite these efforts, up to 20% of pregnant women remain susceptible to rubella.⁸⁸ Persistence of immunity varies depending on the method used to confirm immunity. With the use of hemagglutination-inhibition antibody testing, there is an 8% loss of immunity 16 years after vaccination.⁸⁹ After vaccination with rubella virus, antibody can eventually decrease to undetectable levels; however, revaccination shows an amnestic response, and viremia has not been detected.⁹⁰ Thus, after vaccination, immunity persists for at least 10 years and probably is permanent.⁸⁷

In some persons, a brief rubella-like illness (fever, rash, lymphadenopathy) develops after vaccination. Arthralgias occur in 25% of vaccine recipients, but frank arthritis in only 1%.⁸⁷ Immunosuppressed and pregnant patients should not be vaccinated against rubella. The live virus in the vaccine can cross the placenta and potentially infect the fetus if administered to a pregnant woman; however, congenital rubella has not occurred after vaccination for rubella to date. Defects consistent with congenital rubella were not found in more than 700 women vaccinated in the 3 months before and after conception, although 9 of 143 term infants born to vaccinated mothers had serologic evidence of rubella infection.⁸⁷ Of 307 maternal exposures to rubella vaccine either before conception or during the first trimester, there were no cases of illness or defects consistent with congenital rubella. The theoretic risk of fetal disease after rubella vaccine is 1.7%.⁸⁴ Vaccine virus can be detected in breast milk and secretions for 1 to 4 weeks, but the virus is not transmitted, and close contact and breastfeeding is allowed after vaccination.⁸⁴

Rubella infection during pregnancy can be devastating. In the United States during the last rubella epidemic (1964 to 1965), more than 11,000 cases of fetal death resulted from either miscarriage or therapeutic abortion, and approximately 20,000 infants were born with congenital rubella syndrome (CRS).⁹¹ Of those with CRS, 2100 died as neonates, and nearly 12,000 infants were born deaf. The costs of caring for a child with CRS can be substantial. A wide range of abnormalities occur with CRS, including (in increasing order of frequency) neurosensory deafness, mental retardation, cardiac anomalies (e.g., patent ductus arteriosus, pulmonary artery stenosis), ocular abnormalities (e.g., cataracts, retinopathy, microphthalmia, chorioretinitis) and intrauterine growth retardation.^{92, 93} Hepatosplenomegaly and thrombocytopenic purpura also occur. Of adult survivors of CRS, 40% have been noted to have insulin-dependent diabetes mellitus.⁹³

The effects of rubella infection on the fetus vary with maternal age and parity as well as the gestational age at the time of infection. The incidence of CRS increases as maternal age and parity increase.⁹⁴ Congenital infection documented by a positive serology in the fetus also varies according to the trimester in which the mother is infected. Congenital infection occurred in 81% of infants of mothers infected before 12 weeks gestation, in 54% between 13 and 16 weeks, in 36% between 17 and 22 weeks, in 30% between 23 and 30 weeks, in 60% between 31 and 36 weeks, and in 100% after 36 weeks.⁹⁴ Infection rates by trimester of pregnancy are provided in Table 5.

TABLE 5. Congenital Rubella Infection Confirmed by Seropositivity in Each Trimester of Maternal Illness

(Adapted from Miller E, Cradock-Watson JE, Polock TM: Consequences of confirmed maternal rubella at successive stages of pregnancy. Lancet 2:781, 1982)

The risk of fetal seropositivity translating to a fetal defect in the first trimester is high; however, in the second and third trimesters, although fetal seropositivity increases, the risk of fetal defects decreases.⁹⁴ Deafness, the most common defect associated with CRS,^{94, 95} varies with the trimester of maternal infection. In one study,⁹⁴ of 13 infants with positive serology after first-trimester maternal infection, 85% had defects consistent with CRS, including 5 infants with cardiac defects and 6 infants with deafness. Of 70 infants who were seropositive after exposure to infection in the second trimester, only 9 (13%) had defects, but all 9 had deafness. Thirty-four infants were seropositive after exposure to infection in the third trimester, and no defects were found (see Table 5). Of all infants exposed to maternal rubella infection, infants seropositive for rubella also had a significantly lower birth weight and smaller head circumference than seronegative infants. In another study,⁹⁵ no defects were noted if maternal infection occurred after 18-weeks' gestation. Of offspring with microcephaly, 50% were noted to have normal intelligence, leading the authors to conclude that mental handicap is not an inevitable sequel to microcephaly.

The following recommendations can be used to assess CRS risk:

1. If a maternal rubella-like illness/rash or exposure to rubella infection occurs during pregnancy, blood saved after syphilis serology testing (early sera) taken at the first prenatal visit should be tested for prenatal rubella.
   
2. If the early sera are positive for rubella, the patient is immune and no further testing is necessary.
   
3. If early sera for rubella testing is not available and there is an unknown vaccination history, serologic testing for both IgM and IgG should be performed to document acute infection; a positive IgM result indicates acute infection.
   
4. If the specific rubella IgM is negative or not available, testing of paired acute and convalescent sera for IgG should be performed.
   
5. Acute sera should be drawn as soon as possible after maternal rash develops, followed by convalescent sera 2 to 3 weeks later. If only a suspected exposure has occurred, the acute sera should be drawn immediately and the convalescent sera 4 to 5 weeks later.
   
6. If the prenatal rubella status is unknown, the rash occurred more than 4 weeks ago, or exposure occurred more than 5 weeks ago, and if serologic testing shows no IgG antibody, the patient is at risk for acute infection.
   
7. If rubella-specific IgG is present (indicating immunity to rubella), the date of infection is unknown. A low IgG level indicates distant infection, but repeat testing to determine an increase or decrease in levels may be warranted.
   
8. If the prenatal rubella antibody status is negative, the patient is susceptible to maternal infection and the offspring to CRS. Serologic testing should be performed as for exposed patients with an unknown antibody status.
   
9. Counseling of the mother regarding CRS risk is based on the trimester of exposure. If serologic testing is consistent with a recent infection, the risk of infection of the fetus is high in any trimester (see Table 5).
   
10. To prevent rubella exposure, all children older than 12 months should be vaccinated. All adolescents and adults not known to be immune to rubella, especially women of childbearing age, should also be vaccinated. All prenatal patients should be screened for antibody to rubella, and seronegative patients should receive rubella vaccination in the postpartum period.
    

The diagnosis of acute rubella is difficult, and a clinical diagnosis based on the rash is not reliable. Recent infection must be documented serologically. Acute infection can be documented by IgM specific to rubella or by a fourfold rise in IgG to rubella in paired acute and convalescent sera collected 2 to 3 weeks apart. The rubella IgM level rises early in the illness, peaks 7 to 10 days after the onset of symptoms, and persists for up to 4 weeks after the rash. Rubella-specific IgM can cross-react with rheumatoid factor to create a false-positive result. Levels of rubella-specific IgM also can decline early, causing a false-negative result. A single positive rubella-specific IgG test is usually not helpful in confirming the diagnosis of an acute infection. Low IgG levels do not exclude the possibility of recent infection, although IgG levels are usually high after acute rubella. High levels of IgG, however, have been found in up to 15% of the healthy population.⁹⁶

Congenital rubella infection has been documented by placental biopsy and by isolation of the virus from amniotic fluid.^{96, 97} Rubella-specific IgM can be detected in fetal blood obtained by cordocentesis after 20-weeks' gestation,^{98, 99} but because IgM is not produced by the fetus until this gestational age, diagnosis of rubella by cordocentesis is not accurate before this time.

Despite the persistence of immunity after a natural rubella infection, reinfection with rubella virus can occur. The majority of reinfections are asymptomatic and can be documented by serologic testing. Viremia during a reinfection is rare. Reinfection is more common after vaccination-induced immunity than after naturally acquired immunity. Reinfection during pregnancy, whether after vaccination or natural infection, can cause symptoms suggestive of congenital rubella, but this is an exceedingly rare event.