Howard L. Minkoff
Ronald S. Gibbs
Human Immunodeficiency Virus
Introduction
In this chapter, we will discuss four important perinatal infections—human immunodeficiency virus (HIV), toxoplasmosis, cytomegalovirus (CMV), and herpes—and two common bacterial infections—intraamniotic infection (IAI) and postpartum endometritis.
It has been more than 20 years since an article appearing in the Morbidity and Mortality Weekly Report drew the nation's attention to a previously unrecognized illness. That report described five gay males who had been diagnosed with Pneumocystis carinii pneumonia, an illness theretofore seen only among debilitated hosts. An accompanying editorial noted, “the fact that these patients were all homosexuals suggests an association between some aspect of a homosexual lifestyle or disease acquired through sexual contact and Pneumocystis pneumonia in this population” and “all the above observations suggest the possibility of a cellular–immune dysfunction related to a common exposure that predisposes individuals to opportunistic infections such as pneumocystosis and candidiasis.” Although that editorial showed remarkable insight, almost no one at that time perceived that report to be the harbinger of an epidemic that would devastate large swaths of the globe, become a leading cause of infant mortality, and change the practice of medicine and obstetrics around the world.
The ensuing years have been marked by dramatic progress that resulted from the harnessed energies of the nation's scientific community. In rapid sequence an etiologic agent was discovered, diagnostic tests were developed, public health interventions were instituted, and highly active pharmaceutical agents were developed. For those fortunate enough to be able to afford these interventions, the transformation of the epidemic could not have been more remarkable. No longer did a diagnosis ensure a short painful life and a certain death and for pregnant women mean a 25% chance of birthing an HIV-infected child. With access to state-of-the-art treatment there began to be concrete reasons for hope. Although the diagnosis still signals the beginning of a difficult, life-long struggle against the virus, there are effective, although toxic, therapies and the prevention of perinatal transmission can be ensured in almost 99% of cases. In this section of the chapter, developments in the field will be summarized in a fashion that should allow their integration into the practice of obstetrics and, thereby, ensure HIV-infected women the best possible prognosis for themselves and for their children. The focus of practice must be on two components of care, ensuring the health of women and minimizing the risks of transmission.
Microbiology
Retroviruses constitute a large and diverse family of enveloped RNA viruses that use as a replication strategy the transcription of virion RNA into linear double-stranded DNA, with subsequent integration into the host genome. The characteristic enzyme used for this process, an RNA-dependent DNA polymerase that reverses the flow of genetic information, is known as reverse transcriptase. The unique lifestyle of the retrovirus involves two forms, a DNA provirus and an RNA-containing infectious virion. Infection is initiated by the binding of a protein on the surface of the virus (gp120 Env protein) to the CD4 molecule found on some T cells, macrophages, and microglial cells. CD4 was first shown to be a viral receptor in a number of studies showing the susceptibility of CD4-bearing cells to infection and the ability to block infection with anti-CD4 monoclonal antibodies in culture.
HIV is composed of core (p18, p24, and p27) and surface (gp120 and gp 41) proteins, genomic RNA, and the reverse transcriptase enzyme surrounded by a lipid bilayer envelope. The virion contains three structural genes (gag, pol, and env) and a complex set of regulatory genes, including tat, vif, nef, vpu, and ref that control the rate of virion production. As noted, it preferentially infects cells with the CD4 antigen, particularly helper lymphocytes, but also macrophages, cells of the central nervous system and, according to some evidence, cells of the placenta. At least two other cell surface molecules help HIV enter cells. These co-receptors for HIV, called CXCR4 and CCR5, are receptors for chemokines. Individuals who are homozygous for a deletion at the CCR5gene appear less likely to acquire HIV, while deletion heterozygotes progress less rapidly if infected.
Epidemiology
Approximately 25% of the 800,000 to 900,000 Americans living with HIV disease are women. Not coincidentally, the vast majority of cases of pediatric acquired immunodeficiency syndrome (AIDS) are secondary to vertical transmission of HIV from mother to fetus. Although AIDS rates are lower in women than men (9.3 per 100,000 women versus 32.4 per 100,000 men in 1999), the cumulative percentage of AIDS cases in women almost tripled from 6.7% in 1986 to 18% in 1999. Women of color constitute an increasing majority of AIDS cases in women. In 1999, African American women accounted for 63% of newly reported AIDS cases, and the case rate for Latinas (14.9 per 100,000) was more than 6 times the rate for white women. By 1998, HIV was the third leading cause of death among African American women ages 25 to 44.
The mode of acquisition of HIV in American women has evolved over time. Early in the epidemic, the majority of female AIDS cases were due to injection drug use (IDU). However, by 1995, the proportion of women infected through the heterosexual route surpassed that of IDU. As of 1999, the proportion of women who had sex with a known HIV-infected partner or an individual at high risk had increased to approximately 40%, while only 30% acknowledged using injection drugs. The proportion of women with no known risk increased to slightly greater than 30% in that same survey. Studies suggest that at least 50% of the latter individuals may have acquired HIV through heterosexual means. A report by the Centers for Disease Control and Prevention (CDC) estimates that 81% of those reporting no ascertained risk were infected via heterosexual means. Although noninjected drugs are not a direct vehicle for infection, cohort studies have revealed that much of the heterosexual risk of HIV is associated with use of drugs, such as alcohol and cocaine, which may mediate risk-taking behaviors.
Other important trends characterize the epidemic in women. For example, an increasing proportion of AIDS cases are occurring in women in the South, perhaps reflecting the dramatic increase in other sexually transmitted diseases (STDs) first seen in that region a decade ago. Poverty status might also vary with gender, with women substantially more likely to be covered by Medicaid and less likely to be privately insured. Poverty, in turn, is associated with risk behaviors (e.g., drug use) that are linked to HIV acquisition. These data demonstrate, yet again, that poverty, drug use, and STDs continue to fuel the HIV epidemic among women in the United States.
Pathophysiology
HIV infection leads to a progressive debilitation of the immune system, rendering infected individuals susceptible to opportunistic infections (e.g., P. cariniipneumonia and central nervous system toxoplasmosis) and neoplasias (e.g., Kaposi sarcoma) that rarely afflict patients with intact immune systems. An HIV-infected patient with one of several specific opportunistic infections, neoplasia, dementia encephalopathy, or wasting syndrome, is diagnosed as having AIDS. The diagnosis of AIDS also can be made in the absence of laboratory evidence of infection if the patient has no other known cause of immune deficiency and has the definitive diagnosis of one of a number of indicator diseases. In 1993 the CDC changed the case definition to include all individuals with HIV infection whose CD4 cell counts drop below 200 per cubic millimeter, as well as HIV-infected individuals with advanced cervical cancer, pulmonary tuberculosis, and recurrent pneumonia.
At the time of initial infection, an individual may be asymptomatic or may develop an acute mononucleosis-like syndrome that can be accompanied by aseptic meningitis. There is then an immediate viremia of substantive proportions (up to ten billion viral particles turned over per day) and an equally impressive immune responsive with similar levels of T-cell turnover. Antibodies can be detected in almost all individuals 6 to 12 weeks after exposure but, in rare circumstances, this latent period (the so-called window phase) can be longer. After seroconversion has occurred, an asymptomatic period of variable length usually follows. The median clinical latency in the absence of effective therapy was estimated at approximately 11 years. Very few infected persons (<5%) develop AIDS within 3 years. Evidence of immune dysfunction may be followed by clinical conditions ranging from fever, weight loss, malaise, lymphadenopathy, and central nervous system dysfunction to infections such as herpes simplex virus or oral candidiasis. These nonspecific conditions are usually progressive and are a prelude to an opportunistic infection that is diagnostic of AIDS. Studies of infected individuals have noted that 5 years after infection was confirmed, up to 35% had progressed to AIDS. A study of subjects with hemophilia demonstrated that the incidence rate of AIDS after seroconversion was 2.67 per 100 person-years and was related directly to age (younger individuals developed AIDS at a slower rate). It should be noted that all these statistics antedate the use of new, more powerful antiretroviral agents, which have a significant effect on both clinical outcomes and surrogate markers of disease progression. The level of virus in the plasma can provide an estimate of the probability that an individual will develop AIDS within 5 years.
Once AIDS is diagnosed, the short-term prognosis is improving (e.g., surviving an initial episode of P. carinii pneumonia). The effect that the newest antiretroviral agents, including protease inhibitors (PIs), will have on survival cannot yet be clearly gauged, but they do provide reason for cautious optimism. Conversely, the long-term morbidity related to medications that have been linked to lipodystrophy, among other problems, also remains to be determined. Viral load and CD4 cell counts can be used to predict the likelihood that an individual will develop AIDS during a given follow-up period. Appropriately controlled studies reveal no convincing evidence that the natural history of HIV infection is influenced by gender or pregnancy (Fig. 19.1).
FIG. 19.1. Summary odds ratio of studies which attempted to control for confounding by matching or restriction techniques compared with those which had not used matching or restriction. (From French R, Brocklehurst P. The effect of pregnancy on survival in women infected with HIV: a systemic review of the literature and meta-analysis. Br J Obstet Gynecol 1998;105:827–835, with permission.) |
Management
Monitoring
The guiding principle in the care of HIV-infected pregnant women continues to be strict adherence to the standards of care that apply to all other HIV-infected individuals. The first step in that care is the monitoring of the woman's immune status with CD4 cell counts and viral load measurements. Monitoring of resistance is becoming a frequent addendum to monitoring regimes (vide infra). Viral loads can be checked every 3 months. Once a decision to initiate therapy is made, viral load monitoring should occur monthly until virus is no longer detectable and then can be cut back to every 2 or 3 months. With appropriate therapy, one can anticipate a drop of 1.5 to 2.0 logs within 1 month of initiation of treatment. The recommended timing of initiation of therapy (vís a vís CD4 cell counts and viral loads) has undergone several revisions over the last few years. Despite the clear benefits of therapy, well-recognized toxicities may make it difficult to maintain strict adherence. Some data suggest that no harm befalls patients who delay therapy until viral loads rise and CD4 cell counts drop further than had previously been recommended. Guidelines are to start therapy in nonpregnant women when the viral load is over 30,000 by HIV RNA (55,000 by polymerase chain reaction [PCR]) or when the CD4 cell count drops below 350 per cubic millimeter. In the nonpregnant state, when those thresholds are passed, highly active antiretroviral therapy (HAART) should be initiated. Initiation during pregnancy is discussed below.
Resistance testing has become a staple of care for the HIV-infected woman. Most randomized trials of resistance testing have demonstrated that those assigned to study arms with access to resistance test results had a greater reduction in viral load after the initiation of salvage therapy, although follow-up generally has been short. These tests are recommended for individuals who have just seroconverted or who have failed therapy. Treatment failure is defined as the failure to attain an undetectable level of virus or the persistent presence of virus after it had become undetectable. Transient low-level viremia may not be the same as a drug failure, and sustained response to treatment can occur even with occasional low-level viremia. Blood for testing should be obtained before a failing regimen is discontinued lest wild-type virus overgrow before the test is performed. In that circumstance, an individual with no apparent resistance would still fail therapy when reexposed to drugs that favor the growth of the resistant virus over the wild-type strain. In essence, resistance testing is more useful for ruling out, than for ruling in, therapies to be used in a given patient. That is because, as noted, the absence of resistance may merely reflect the reemergence of a wild-type strain after an antiretroviral agent has been withdrawn. In that circumstance, the assays will not detect a low volume of a minority mutant strain. However, if the patient is reexposed to the offending agent, the resistant strain may again attain dominance. It is not possible to perform resistance studies if there are fewer than 1,000 copies of virus detectable.
Two types of testing are available, genotypic and phenotypic, each with distinct advantages and disadvantages. Phenotypic testing compares the ability of the virus to replicate in various concentrations of an antiretroviral drug with its replication in the absence of the drug. In general, if the amount of drug required to inhibit viral production by 50% is four-fold or greater for the patient's virus than the control strain, then the patient's strain is considered resistant. Specific cutoffs for resistance for each drug are being developed based on clinical correlation and the serum levels usually attainable for a given drug. Measurement of drug levels per se has not yet been shown to be a useful addendum to standard monitoring.
Genotypic testing is directed at detecting mutations in the genes that encode reverse transcriptase and protease formation by the virus. Point mutations in the virus result in the substitution of amino acids in the proteins produced, namely reverse transcriptase or protease. The significance of these point mutations has to be determined by correlating specific mutations with phenotypic resistance, as measured by viral susceptibility assays and correlation with clinical response to therapy. Genotypic changes leading to resistance are believed to result from the combination of the rapid turnover of HIV (107–108 rounds of replication per day) and the high error rate of reverse transcriptase when replicating the nearly 10,000 nucleotides present in the HIV genome. Genetic mutants with resistance to antiretroviral agents are then selected by evolutionary pressure when incompletely suppressive drug regimens are used. The rate at which resistance develops will depend on the number of mutations necessary for a significant change in susceptibility to occur. The genetic basis for resistance must be understood before the impact of a specific mutation can be predicted. In addition, mutational interactions may make prediction of phenotype (i.e., susceptibility) difficult when multiple mutations are present. Prediction of cross-resistance to other drugs within a class such as protease inhibitors can be difficult to predict based only on genotype. In clinical practice, most clinicians will rely upon algorithms developed by panels of experts or found in on-line databases. Obstetricians should interpret and act upon these results in consultation with an expert in the field.
It has been suggested that HIV-infected pregnant women will benefit routinely from resistance testing, but a few caveats are necessary in that regard. The goal in all cases of HIV infection in pregnancy is to maintain a viral load under 1,000, because that would both allow a vaginal delivery and minimize the mother-to-child transmission rate. However, if a regimen were successful in dropping the load to or below that level, it would be impossible to perform resistance testing. If the regimen is unsuccessful, then resistance testing is warranted, regardless of the patent's pregnancy status. That said, there are considerations that are unique to pregnancy and they relate to the fitness of resistance virus for mother-to-child transmission of HIV.
In general, resistance to zidovudine (ZDV, also commonly known as AZT) and other resistance mutations have not been associated with an increased risk of perinatal transmission in most perinatal transmission studies. In one study, detection of ZDV resistance was associated with transmission when adjustment was made for duration of ruptured membranes and total lymphocyte count. Factors associated with resistance at delivery included ZDV use prior to pregnancy, higher log HIV RNA, and fewer CD4+ lymphocytes present. Although perinatal transmission of resistant virus has been reported, it appears to be unusual, and it is not clear that the presence of mutations increases the risk of transmission. In a multisite study of HIV-infected pregnant women (WITS), it was reported that when a transmitting mother had a mixed viral population of wild-type and low-level resistant virus, only the wild-type virus was found in the infant, suggesting that virus with low-level ZDV resistance may be less transmissible.
Pharmaceutical Therapy
Although the number of HAART regimens that have been shown to achieve persistently low viral loads continues to increase, they generally fall into one of a few broad categories. The original regimens described in 1996 included dual nucleoside therapy accompanied by a protease inhibitor. There are seven nucleoside reverse transcriptase inhibitors (NRTIs), three nonnucleoside reverse transcriptase inhibitors (NNRTIs), and six PIs approved for therapy, with many others in the pharmaceutical “pipeline,” so theoretically a large number of choices within these categories exists. However, certain medications should not be used in combination. For example, ZDV and D4T have overlapping toxicities, ddI and ddC should not be used in combination, and ddI and d4T have been linked to several fatal cases of mitochondrial toxicity in pregnancy.
Because the number of new drugs continues to multiply, as does information about their benefits and toxicities, it has become increasingly difficult for busy clinicians to stay abreast. To assist in that endeavor, the U.S. Department of Health and Human Services has established a Web site (http://www.aidsinfo.nih.gov) that is regularly updated. The expert panel that authors those update guidelines “strongly recommends” regimens that include either one or more PI(s) (indinavir, nelfinavir, ritonavir + saquinavir, ritonavir + indinavir, ritonavir + lopinavir) or an NNRTI (efavirenz in the nonpregnant patient) in combination with one of several two-NRTI combinations. There is a great deal of clinical outcome data that support the use of a PI in combination with two NRTIs. Some of the PIs, such as ritonavir, are not usually chosen as first-line therapy, because patients often have difficulty tolerating standard dosages and because of the drug's many interactions. Other PIs, such as saquinavir-SGC, create difficulties for many patients because of the large pill burden associated with their use; however, if a woman can maintain the dosage and it is working well, then there is no reason to change drugs.
A critical factor to recognize is that poor results often occur with antiretroviral regimens used in a patient who has experienced virologic failure with a previous regimen. That suggests that the first regimen affords the best opportunity for long-term control of viral replication. Because the genetic barrier to resistance is greatest with PIs, many would consider a PI + two NRTIs to be the preferred initial regimen. However, efavirenz (NNRTI) + two NRTIs appears to be at least as effective as a PI + two NRTIs in suppressing plasma viremia and increasing CD4+ T-cell counts, and many would prefer such a regimen initially because it may spare the toxicities of PIs for a considerable time. However, concerns about teratogenicity that have been demonstrated in animal models, as well as a case report of a myelomeningocele after in utero exposure in a human, make this a poor choice for use in early pregnancy. Thus, although the demonstrated ability of efavirenz in combination with two NRTIs to suppress viral replication and increase CD4+ T-cell counts to a similar degree as a PI with two NRTIs supports a preference for efavirenz over the other available NNRTIs, in pregnancy the choice of a different NNRTI may be preferable. Abacavir, a new NRTI + two other NRTIs (i.e., a triple NRTI regimen), has been used with some success, as well. Such a regimen, however, may have short-lived efficacy when the baseline viral load is more than 100,000 copies per milliliter, and has recently been shown to be less effective than efavirenz. Using two NRTIs alone does not achieve the goal of suppressing viremia to below detectable levels as consistently as do the other regimens discussed above and should be used only if more potent treatment is not possible. Use of antiretroviral agents as monotherapy is contraindicated, except when there are no other options, or in pregnancy to reduce perinatal transmission as noted below. When initiating antiretroviral therapy, all drugs should be started simultaneously at full dosage with the following three exceptions: dosage escalation regimens are recommended for ritonavir, nevirapine and, in some cases, ritonavir plus saquinavir.
Considerations Related to HAART Therapy for Pregnant Women
Although therapeutic recommendations should not be modified a priori because of pregnancy, a few comments deserve particular mention. First, in regard to the dosage of ZDV, a drug that has been a cornerstone of therapy in pregnancy since 1994, it should be remembered that the antenatal dosing regimen used in the PACTG 076 trial (100 mg administered orally 5 times daily) was selected on the basis of standard ZDV dosage for adults at the time of the study. However, administration of ZDV 3 times daily will maintain intracellular ZDV triphosphate at levels comparable with those observed with more frequent dosing. Comparable clinical response also has been observed in some clinical trials among persons receiving ZDV twice daily. Thus, the standard dosing for ZDV, whether used as a part of a HAART regimen or as single-drug therapy for transmission prevention (discussed below), is 200 mg 3 times daily, or 300 mg twice daily. Although it is possible that these dosing regimens may not have efficacy equivalent to that observed in PACTG 076, a regimen of 2- or 3-times daily is likely to enhance maternal adherence.
NRTIs were the first class of drug used for treatment of HIV, and there is a greater experience with their use in pregnancy than with other classes of drug. Most information is available for ZDV, the first agent to be used widely in pregnancy. Most data regarding the safety of ZDV have been reassuring. Almost 1,000 children have been tracked for 4 years with no increase in incidence of neurodevelopmental delay or carcinogenesis. Some concern had been raised, however, as a result of mouse data reported from the National Cancer Institute. That report suggested that mice exposed to high dose rates of ZDV in the third trimester birthed pups that were subsequently found to have high rates of liver, lung, skin, and genitourinary tumors in midlife. Although other investigators have not reported similar findings, the National Cancer Institute data should chasten those who would dismiss all concerns regarding these therapies and reinforce the need for thorough counseling and long-term follow-up. In regard to monitoring of the mother on ZDV, it is necessary to measure the blood count and liver functions only on a monthly basis. The only abnormality that occurs with any frequency is anemia.
Other NRTIs have been well tolerated generally and have not been demonstrated to be teratogenic in humans. However, concerns have been raised about potential adverse effects on both mothers and infants related to the avidity of these drugs for mitochondria. By binding to mitochondrial γ-DNA polymerase and interfering with replication, these drugs can induce mitochondrial dysfunction. The greatest inhibition of mitochondrial γ-DNA polymerase is demonstrated by ddC, followed in order by ddI, d4T, 3TC, ZDV, and abacavir. Clinical disorders associated with mitochondrial toxicity include neuropathy, myopathy, cardiomyopathy, pancreatitis, hepatic steatosis, and lactic acidosis.
In regard to potential maternal toxicity, several cases of lactic acidosis, three of which were fatal and two of which were accompanied by pancreatitis, have been reported among pregnant or recently delivered women who had been on ddI and d4T therapy along with a variety of third agents since before conception. Two cases of fatal liver failure in pregnant women on ZDV, lamivudine, and nelfinavir also have been reported. These cases developed in late pregnancy, and in several cases the clinical picture was similar to that seen with acute fatty liver of pregnancy, a condition that has been linked to mitochondrial fatty oxidation disorders in the fetus (homozygotic) and mother (heterozygotic). This has led to speculation that the metabolic changes of late pregnancy may enhance susceptibility to complications of nucleoside agents, especially those with greater inhibition of mitochondrial γ-DNA polymerase. That susceptibility is suggested both by the syndrome of acute fatty liver of pregnancy and animal data that demonstrate reduced mitochondrial fatty acid oxidation in late pregnancy and in animals treated with exogenous estradiol and progesterone to mimic pregnancy levels. However, many cases of death outside of pregnancy have been reported related to these medications as well. In any event, although these serious morbidities appear to be rare, providers caring for HIV-infected women receiving nucleoside analog agents should be cognizant of the risk and monitor accordingly. One approach would be to monitor hepatic enzyme levels during the last trimester and to investigate aggressively all new symptoms. Women with substantial elevations in transaminase levels above baseline or other new abnormalities, in the absence of other explanations such as preeclampsia, should have their nucleoside agents discontinued, either with substitution of agents from another class of antiretroviral drugs or discontinuation of all antiretroviral medications. In view of the reports of maternal deaths and toxicity associated with prolonged use of d4T and ddI in pregnancy, this combination should be used in pregnancy with caution and only if other nucleoside agents cannot be used because of resistance or toxicity.
In regard to infants, concern about mitochondrial toxicity was raised initially by a French group that reported eight cases of HIV-uninfected infants (among 1,754 exposed fetuses) with abnormalities potentially related to mitochondrial dysfunction, which developed after exposure to prophylactic nucleoside therapy. Two infants had progressive neurologic symptoms and died several months after completing in utero and neonatal courses of ZDV/3TC. Three other infants had mild to moderate symptoms, and three had asymptomatic laboratory abnormalities. Mitochondrial abnormalities were not proven to be the cause of the abnormalities, and the relationship between these findings and in utero and neonatal nucleoside exposure has not been established. In response to these concerns, investigators from several large cohort studies in the United States reviewed all 353 deaths among more than 20,000 children born to HIV-infected women and found no deaths similar to those in the French cohort, although only 6% of the children had been exposed to the combination of ZDV/3TC. Review of data on living children from these cohorts for diagnoses or conditions suggestive of mitochondrial dysfunction is ongoing.
Combination therapy also has been linked to increased rates of prematurity in a few European studies, but these have not been confirmed in larger American series. Reports also have warned of an increased rate of neonatal febrile seizures in infants who had been exposed to antiretroviral therapy. Finally, as noted above, efavirenz should not be recommended because of reported teratogenic effects in monkey models.
In regard to prophylaxis for opportunistic infections, if the count drops below 200 cubic millimeters, P. carinii pneumonia prophylaxis should be instituted, and below 50 cubic millimeters, Mycobacterium avium complex prophylaxis should be given and an ophthalmology consult obtained. The specific details of management of the myriad infections to which these women are prone are beyond the scope of this chapter. If HAART therapy is successful in returning CD4 cell counts to levels above the threshold for prophylaxis of opportunistic infections and those levels are maintained for 6 months, then the prophylaxis can be discontinued.
Preventing Perinatal Transmission
It has been almost a decade since the results of PACTG Protocol 076 were released. Those results demonstrated that a regimen of ZDV given during pregnancy, labor, and to the newborn for 6 weeks was able to reduce remarkably the risk of perinatal HIV transmission; an almost 70% reduction in transmission risk was observed, from 25% in the placebo group to 8% in the ZDV group. Subsequently, ACTG 185 demonstrated that similar results could be seen, even among women who had previous exposure to ZDV or who had CD4 cell counts below 200 cubic millimeters. ZDV has now been integrated successfully into clinical practice in the United States and Europe, and its widespread use has been accompanied by dramatic declines in perinatal transmission rates in these countries. The regimen consists of ZDV 200 mg 3 times a day in the antepartum period (beginning after 14 weeks). In the intrapartum period, a loading dose of 2 milligrams per kilogram over the first hour is followed by a maintenance dose of 1 milligram per kilogram per hour thereafter. During the neonatal period the infant receives ZDV syrup for 6 weeks at a dosage of 2 milligrams per kilogram every 6 hours.
There are insufficient data to justify the substitution of any agent for ZDV in most circumstances, save perhaps for times when women have not received therapy prior to labor (vide infra). Even if there is evidence of ZDV resistance, which might make ZDV therapy for the prevention of mother-to-child transmission seem futile, ZDV might still have a role. The demonstration of transmission of escape mutants from mother to child suggests that an individual ZDV-susceptible virus might be transmitted, even if the predominant strain(s) in the mother is (are) not susceptible. If the mother is intolerant of ZDV, it may be reasonable to consider using an alternative agent, which potentially could reduce viral load during the antepartum period, and then adding ZDV for the intrapartum (depending on the type of toxicity experienced by the mother) and neonatal periods. Thus, if the therapeutic regimen being used to treat a woman's HIV disease does not include ZDV, it would seem judicious to incorporate it if she becomes pregnant.
The recommendation to use ZDV, as opposed to other antiretroviral agents, for the prevention of mother-to-child transmission of HIV is based on two considerations. First, there are empiric data that strongly support its use in that regard. No other agent has been studied similarly. Second, pharmacokinetic studies demonstrate that high levels of drug are available in the fetal compartment. Similar information is unavailable for most other agents. The importance of placental passage of drug is not certain. If viral load thresholds truly are central to transmission risk, than viral dynamics on the maternal side of the placenta might be as important as placental passage of drug. Several publications (Fig. 19.2) suggest that a correlation exists between maternal viral load and rates of mother-to-child HIV transmission and, further, hint at the existence of a clinically relevant threshold. Although other researchers disagree, most data support the thesis that dropping viral loads as low as possible in pregnancy is a useful strategy. There is clear evidence that, even if the viral load is low, those women on antiretroviral therapy will fare better than women with similar viral loads who are not on therapy, and data suggest that combination therapy is even more effective than ZDV alone in reducing rates of transmission. A French team reported that adding 3TC to ZDV reduced transmission rates to 1.6%. A longitudinal study in the United States found that 20% of women with no antiretroviral therapy transmitted HIV to their infant as compared with 10.4% who received ZDV alone, 3.8% who received combination therapy without a PI, and 1.2% who received a combination that included a PI.
FIG. 19.2. Left: Maternal human immunodeficiency virus–type 1 (HIV-1) RNA levels at delivery in infected nontransmitting and transmitting mothers. Mothers who received zidovudine during gestation or labor and delivery are indicated by the open circles; mothers who did not receive zidovudine during gestation, labor, or delivery are indicated by the darkened circles. Horizontal bars indicate the median for each of the measured variables. Right: Perinatal transmission rate according to HIV-1 RNA levels at delivery. (From Dickover RE, Garratty EM, Herman SA, et al. Identification of levels of maternal HIV-1 RNA associated with risk of perinatal transmission. JAMA 1996;275:599–605, with permission.) |
The vast majority of perinatal HIV-1 infection occurs in the developing world. Unfortunately the PACTG 076 ZDV regimen is too costly and logistically complex for many nonindustrialized countries to implement on a widespread scale, and its efficacy in a breast-feeding population (where postpartum transmission remains a real concern) is unknown. An ideal preventive intervention would be cheap, nontoxic to mother and fetus, easy to administer, only need to be given once or for a limited period, and have utility in preventing postpartum transmission. The results of PACTG 076 have spurred the worldwide evaluation of multiple other modalities to reduce transmission.
Many of the regimens designed for countries with fewer resources have attempted to achieve reduced numbers of infected children with shortened courses of antepartum and neonatal therapy. These regimens have included a Thai regimen of 4 weeks of antepartum ZDV combined with oral ZDV therapy during labor that reduced transmission to 9% compared with 19% in a placebo arm. Similar findings have been reported from the use of an abbreviated course of ZDV and 3TC in an African trial. An even more dramatic finding emerged from the HIVNET 012 trial which revealed that a single does of nevirapine during labor and again to the neonate could reduce transmission at 6 weeks from the 21% seen in those receiving a short course of ZDV to only 12%. Unfortunately, the benefits of all these regimens are attenuated in populations that breast-feed after the regimens have been completed.
Perhaps of greater relevance to American populations are data that demonstrate that even patients whose serologic status is unknown until the peripartum period may benefit from pharmaceutical interventions. In a review of the New York state experience, Wade and colleagues found that the transmission rate from mothers who received no therapy was lowered significantly with ZDV even when therapy was delayed until the first 24 hours postpartum. It was lowered to an even greater extent when therapy was begun during the intrapartum period.
Other efforts to prevent transmission have focused on attempts to reduce peripartum and postpartum exposure to the virus. Reduction of intrapartum exposure could include attempts to minimize the duration of ruptured membranes, which has been related to the rate of mother-to-child transmission of HIV (Fig. 19.3). Cesarean delivery is the one strategy that can guarantee that the infant will not be exposed to the risks associated with ruptured membranes during parturition. Evidence has accumulated that cesarean section may be a beneficial addendum to pharmaceutical therapy. Read and her colleagues published a meta-analysis performed on primary data from 15 prospective cohort studies, including more than 7,800 mother-child pairs. In that meta-analysis, the rate of perinatal HIV-1 transmission in women undergoing elective cesarean delivery was 8.2% and 2.0% in those receiving no antiretroviral agents and in those receiving ZDV, respectively. Both rates were significantly lower than those seen among women delivered by either nonelective cesarean or vaginal delivery.
FIG. 19.3. Probability of HIV-1 transmission in relation to the duration of ruptured membranes. The dots at the top represent women who transmitted HIV-1 to their infants and those at the bottom, women who did not transmit. (From Landesman S, Kalish L, Burns D, et al. The relationship of obstetrical factors to the mother-to-child transmission of HIV-1. N Engl J Med 1996;334:1617–1623, with permission.) |
Subsequently, the results of a European randomized trial of cesarean section were reported (Table 19.1). The results were remarkably consistent with those that had been found by Read. In the European study, HIV-infected women between 34 and 36 weeks of gestation were randomly assigned elective cesarean delivery at 38 weeks or vaginal delivery. Three (1.8%) of 170 babies born to women assigned cesarean delivery were infected compared with 21 (10.5%) of 200 born to women assigned to vaginal delivery (P < .001). Seven (3.5%) of 203 infants of women who gave birth by cesarean section were infected compared with 17 (10.2%) of 167 born vaginally (P = .009). Unfortunately, the number of participants was too small to allow a separate analysis of the benefit of cesarean section for a patient receiving antiretroviral therapy, and HAART use was not reported from the cohort. There were few postpartum complications, and no serious morbidity was noted in either group. Despite the shortcomings of the trial, it does suggest that among women who were not optimally treated with HAART, cesarean section could have an important effect on reducing rates of mother-to-child transmission of HIV.
TABLE 19.1. HIV-1 infection status of children according to allocated and actual mode of delivery |
In sum, the above-cited studies indicate that compared with other types of delivery, cesarean delivery performed prior to the onset of labor and prior to rupture of membranes (elective, or scheduled, cesarean) significantly reduces the rate of perinatal HIV-1 transmission by odds ratios of 0.2 to 0.4. On the basis of the data, the American College of Obstetricians and Gynecologists published a committee opinion that concluded that HIV-infected women should be offered scheduled cesarean section in order to reduce the rate of transmission beyond that which could be achieved with ZDV alone. They also pointed out that data are insufficient to demonstrate a benefit for women with viral loads less than 1,000 copies per milliliter of plasma. They suggested that scheduled cesarean sections should be performed at 38 weeks of gestation and that amniocentesis should not be performed in order to avoid contamination of the amniotic cavity with viral antigen from maternal blood. They also suggested that prophylactic antibiotics be employed because of concerns of heightened risks of postoperative infectious morbidity. The data supporting a heightened risk for postoperative morbidity among HIV-infected women are not robust; however, the evidence that all women undergoing an operative delivery face a greater risk of infectious morbidity is overwhelming. The liberal use of antibiotics in the setting of cesarean section and HIV infection would, therefore, seem reasonable.
Although not as compelling as the data just cited, there is some evidence that cesarean section could be beneficial even when viral loads are under 1,000 copies. That data come from a meta-analysis of studies that focused exclusively on women who had viral loads lower than 1,000 copies. In those women, antiretroviral therapy still played a major role in reducing transmission, dropping rates from approximately 10% to approximately 1%. Cesarean section apparently dropped the rate from 6% to 1.5%, but there was no control in that analysis for the use of antiretroviral therapy. In sum, there is no strong evidence that cesarean section will provide additional benefit to the HIV-infected woman on therapy who has an undetectable viral load.
TOXOPLASMA
Introduction
Primary maternal toxoplasmosis occurs in approximately 1 of every 900 pregnancies in the United States. This estimate is based on a prospective study of sera from 23,000 pregnant women done in early and late gestation. That study, which was conducted by the National Institutes of Health, also showed that 38% of the women tested had antibodies to Toxoplasma gondii, indicating previous infection with the organism. More recent data suggest that the seroprevalence may now be somewhat lower (15% among women of childbearing age). In the earlier data, the presence of antibodies correlated with increasing patient age and was twice as frequent among blacks as among whites. None of the mothers tested had evidence of significant clinical disease. It has been estimated that between 400 and 4,000 babies are born with congenital T. gondii infections each year. Some data suggest that congenital infections occur in approximately 1 in 10,000 births.
Microbiology: Transmission
Cats
The cat is the definitive host for T. gondii, a protozoan parasite. About one half of the cats tested in the United States have antibodies to T. gondii. It is thought that cats acquire infection by eating infected wild rodents and birds. A week after infection, the cat begins to shed oocysts in its feces. Shedding of the oocysts persists for about 2 weeks before the cat recovers spontaneously. These animals are susceptible to reinfection and may also shed toxoplasma oocysts when infected with other organisms.
Although the cats excrete unsporulated (i.e., noninfectious) oocysts in their feces, within days to weeks these oocysts sporulate and become extremely infectious. The fecal oocysts are an important source of infection to humans through inadvertent ingestion. Because sporulation of the organism occurs after days to weeks in the litter, it may be prevented by having a nonpregnant person change the litter daily. Care must be taken in disposing of cat litter, because the oocysts can remain infectious for long periods in favorable climates.
Meat
In Europe, where the use of refrigeration is more limited and meat usually is not frozen, ingestion of infected meat is an important cause of toxoplasmosis. In contrast, in the United States, much of the meat is frozen at some point during storage or transport. Freezing is probably one of the factors responsible for the difference in incidence of toxoplasmosis here and in Europe. Worldwide, about 1% of cattle, 20% of hogs, and 30% of sheep have toxoplasmosis, according to estimates based on isolation of the organism from animal muscle tissue. To avoid contagion from meat, it should be cooked thoroughly at adequate temperatures.
Epidemiology
T. gondii has a worldwide distribution and has been reported from wherever cats are found. It is somewhat more common in tropical and coastal regions and is less common in regions that are either cold, warm and arid, or at high elevation. The infection rate in the United States is significantly lower than that in France. Within the United States, seroprevalence rates are lower in the west central and mountain states, and higher in east Atlantic and east central states.
Pathophysiology
T. gondii exists in three forms: trophozoites or proliferative form (Fig. 19.4), tissue cysts, and the oocytes. The organism undergoes a substantial portion of its life cycle in the cat where, after between 5 and 8 days of infection, there is peak oocyte production. As many as ten million oocysts per day can be shed in feces for periods varying between 7 and 20 days. These oocytes sporulate within 1 to 3 days and can remain infectious for several months in moist soil. At that point they may be carried from point of deposition by other animals (e.g., flies) and deposited in food. It has also been suggested that they can also become airborne from a dried out litter box and lead to human infection. Eating undercooked or raw meat that contains tissue cysts causes approximately one half of infections in most humans. In the human, the trophozoite form of T. gondii is seen in the acute phase of the infection, and it is during this phase that host cells are invaded. Thereafter, the organism multiplies every 4 to 6 hours until the cytoplasm becomes so filled with trophozoites that the cells rupture, releasing organisms to invade other cells.
FIG. 19.4. T. gondii in trophozoites or proliferative form. (From Jones JL, Lopez A, Wilson M, et al. Congenital toxoplasmosis: a review. Obstet Gynecol Surv 2001;50:296–305, with permission.) |
In Utero Transmission
Newborns with congenital toxoplasmosis become infected in utero by transplacental passage of the parasite when the mother has acute infection. Chronic infections (onset precedes pregnancy) do not lead to congenital infection except in the rare circumstance of an immunocompromised host (e.g., patient with systemic lupus erythematosus taking steroids, HIV disease). In general, the likelihood of fetal infection increases with each trimester of pregnancy, being approximately 15%, 25%, and 60% in the first, second, and third trimester, respectively. The severity of damage associated with congenital toxoplasmosis also is related to the timing of maternal infection, but in this situation, the risks decrease toward term. Severe fetal disease or fetal death occurs in about 10% of cases when infection occurs during the first trimester and is extremely rare with infection during the third trimester. Mild damage is more frequent in the second and third trimesters (about 5%). Subclinical infections increase from about 2% with first-trimester infections to 50% with third-trimester infections.
The results of a case-control study of women with poor pregnancy outcomes and controls suggested that acute infection could be associated with preterm delivery and stillbirth but not with spontaneous abortion. Chronic infections were not associated with any untoward outcomes.
Diagnosis in Pregnancy
Mother
Maternal infection with T. gondii is usually asymptomatic, although 10% to 20% of infected mothers have lymphadenopathy. Posterior cervical lymphadenopathy is the most frequent finding associated with acute maternal toxoplasmosis. The infection also can result in a mononucleosis-like syndrome with fatigue and lassitude and, rarely, can cause encephalitis. Acute toxoplasmosis should be considered in any pregnant woman who has lymphadenopathy, particularly involving the posterior cervical chain, or mononucleosis-like symptoms. The vast majority, however, of those acutely infected with T. gondii are asymptomatic. The clinical picture can be much more severe in immunocompromised adults.
In the absence of symptoms, clinicians are forced to rely on serologic tools for the diagnosis of toxoplasmosis in pregnancy. The diagnosis of primary infection with T. gondii during pregnancy requires either (a) the demonstration of a seroconversion to this organism, (b) a significant rise in antibody titer obtained from maternal sera taken at two different times, or (c) the detection of toxoplasma-specific immunoglobulin M (IgM) antibody. Adults with primary infection develop immunoglobulin G (IgG) and IgM antibody to toxoplasma rapidly. Toxoplasma-specific IgG antibody develops within 2 weeks after infection, peaks in 6 to 8 weeks, drops down over the subsequent several months, and then persists for life. Toxoplasma-specific IgM develops within 10 days after infection and remains elevated for 6 months to more than 6 years. Some researchers have suggested that the presence of high-avidity IgG antibodies exclude acute infection within the previous 3 months, thereby making it a useful test for first-trimester infection.
Because IgM antibody remains elevated for many months, this test may not provide useful information to document recent primary infection in pregnant women. The enzyme-linked immunosorbent assay (ELISA) test for IgM frequently shows the development of high titers of antibody that can persist for many years. Indirect immunofluorescence antibody (IFA) tests for toxoplasma-specific IgM usually show high titers for only about 6 months after infection, following which the titer rapidly drops. The IFA test, then, frequently is more useful than ELISA in differentiating remote from recent primary infection of a pregnant woman. In any case, the presence of IgG and the absence of IgM suggest an infection that is probably at least a year old. Because of difficulties with the reliability of some commercially available tests for IgM, it has been recommended that all positive test results be confirmed in a reference laboratory.
Approximately 50% of placentas of congenitally infected infants will show T. gondii cysts on histologic slides, and their presence supports the diagnosis of acute infection in the mother during pregnancy. Additional cases can be detected by the presence of parasites in the cord blood. The organism also has been isolated from placental tissue of acutely infected mothers in 2% to 25% of cases. Recovery was more frequent when infection occurred later in pregnancy. Isolation of organisms from tissue specimens, buffy coat heparinized blood, and body fluids can be used, also, for diagnosis. These specimens produce the organism after inoculation into the intraperitoneal cavities of mice or into tissue culture.
Prenatal Diagnosis
Antenatal diagnosis of fetal toxoplasmosis used to rely on culture of amniotic fluid (15–20 cc) or fetal blood (1.5–3.0 cc) obtained at the time of diagnostic amniocentesis or cordocentesis, respectively. The specimen was cultured commonly in mice or fibroblast cells. The main difficulties with culture techniques have been that some assays may take up to several weeks to get complete results, and very few laboratories are able to perform the assay.
Toxoplasma-specific IgM, when present in fetal blood from cordocentesis, has also been used to diagnose fetal infection prenatally. Unfortunately, fetal-specific IgM antibody frequently does not develop until after 21 to 24 weeks gestation and is positive in only about 50% of infected cases. Additionally, cordocentesis is a procedure that entails some risk.
More recently, the PCR has been used to detect T. gondii in amniotic fluid and has been shown to be useful in the detection of fetal infections in utero. In one large series, PCR performed better than conventional tests (sensitivity 97.4% versus 89.5%; negative predictive value 99.7% versus 98.7%). PCR of amniotic fluid has, to a large extent, rendered cordocentesis obsolete for the purpose of diagnosing toxoplasma infections.
Child
Most congenitally infected newborns are asymptomatic at birth. Literature has shown that detection of toxoplasma-specific immunoglobulin A (IgA) may be a reliable method for the diagnosis of toxoplasmosis in the newborn. A number of these asymptomatic, untreated infants will go on to have delayed and potentially serious manifestations. The 20% with clinically obvious symptoms at birth will exhibit multiple findings. The most frequent clinical findings are chorioretinitis, jaundice, fever, and hepatosplenomegaly. Hydrocephaly or microcephaly and cerebral calcifications can be seen in severe cases. Demonstration of toxoplasma-specific IgM infection may be diagnostic, although in newborns, approximately 20% of infections are not detectable by toxoplasma-specific IgM at birth.
Treatment and Prevention
Treatment of acute toxoplasmosis is primarily supportive. The prognosis, in general, following acute infection is good, except in cases of profound immunosuppression.
The treatment in pregnancy is a bit more complex. In Europe, where the seroprevalence and, hence, the clinical experience is greater, spiramycin is the first-line agent used. However, that agent generally does not cross the placenta, and if fetal infection is detected, women are also treated with a combination of pyrimethamine, folinic acid, and a sulfonamide. Although not definitive, treatment with these regimens may prevent maternal-to-fetal transmission of the infection or improve the outcome among infected fetuses. In one study from France, 163 mothers diagnosed with toxoplasmosis prior to 28 weeks were treated with spiramycin (23 also received pyrithiamine and sulfadiazine). Three fetuses died in utero and 27 were diagnosed with congenital toxoplasmosis. All 27 were free from symptoms and had normal neurologic development at 15 to 71 months. Most studies that control for gestational age at the time of infection suggest that transmission rates are not altered markedly by therapy, but the degree of fetal sequelae may be.
The standard dosage is 25 mg of pyrimethamine by mouth given daily and 1 g of sulfadiazine by mouth 4 times daily for 1 year. Pyrimethamine is a folic acid antagonist and, therefore, may have teratogenic effects when given in the first trimester. Whenever possible, treatment with pyrimethamine in the first trimester should be weighed against potential risk of drug teratogenicity to the infant. Folinic acid, 6 milligrams given intramuscularly or by mouth every other day, should be used to correct the depletion of folic acid induced by pyrimethamine.
Spiramycin can be obtained in the United States through the CDC. It is used more commonly in Europe and, hence, there are no good controlled studies of its efficacy in this country. It has not been found to be teratogenic in humans or animals.
Case-control studies in France, where women routinely receive serial assessments to detect seroconversion, have revealed that risk factors for acquisition of infection include poor hand washing, eating undercooked beef or lamb, and owning a pet cat. Based on these sorts of data, the CDC has recommended that sero-susceptible pregnant women should be counseled to avoid eating raw or undercooked meats that may contain the T. gondii cysts. This can be accomplished by using a food thermometer to ensure that the meat is cooked all the way through. Fruits and vegetables should be peeled and washed before eating. Gloves should be used for gardening and during any contact with soil or sand, because cat waste might be found there. Pregnant women should avoid close contact with cat feces, such as changing cat litter. CDC also highlighted the need for obstetricians to educate their pregnant patients about these important preventive steps.
In countries such as France, with extremely high seroprevalence rates, routine serologic screening programs have proven successful in diagnosing recent seroconverters, allowing for prenatal diagnosis of fetal status and either termination of pregnancy or prenatal therapy. There is no consensus for routine screening in the United States. Because of the low prevalence of the disease and the possibility of false-positive results, the American College of Obstetricians and Gynecologists does not recommend routine screening.
Most recently, programs have been undertaken that focus on screening of newborns and the institution of treatment during the neonatal period to minimize the morbidity that would otherwise accrue to congenitally infected children. Many infections in children, that otherwise would be missed on routine clinical examination, can be detected with IgM assays. Treatment of these infected infants has been associated with very low rates of subsequent neurologic or retinal disease.
CYTOMEGALOVIRUS
Introduction
CMV, a member of the herpes virus family, is the largest virus to infect humans. It can code for over 200 proteins, through which it can lead to substantial down-regulation of the immune system and many diseases, such as infectious mononucleosis in young adults. It also causes the most common congenital viral infection, affecting approximately 1% of all live births, about 35,000 infants annually in this country. Congenital CMV infection is acquired by the fetus in utero when the mother develops primary CMV infection while pregnant or with reactivation of a prior maternal infection. Although reactivated disease in the pregnant woman accounts for more than one half of the congenital infections, primary maternal CMV infection is much more likely to result in a severely affected and symptomatic infant. Passive in utero transfer of maternally derived CMV-specific IgG antibody appears to provide some protection to the fetus when CMV is reactivated during pregnancy.
The great majority of infants with congenital CMV are asymptomatic (90%), but some have evidence of disease during the newborn period (10%). About 10% of the symptomatic group have full-blown cytomegalic inclusion disease. An additional 10% of infected infants who are asymptomatic at birth later develop symptoms related to CMV infection. The most common late-onset symptoms in these cases are mental retardation and deafness. These infants usually shed high titers of virus in the urine and saliva for a number of months.
CMV infections can be acquired by the child during the postpartum period through (a) exposure of the infant to infectious maternal body fluids such as cervical secretions, urine, saliva, and breast milk, (b) following blood transfusion or tissue transplantation from an infected donor, or (c) most frequently, through contact with infected individuals, such as in the newborn nursery, day care centers, or from other family members in the household. These cases are called acquired CMV infections, in contrast to congenital infections.
Microbiology
Cytomegaloviruses are members of the herpes virus group. The virus has a diameter of about 180 nanomeres and is encapsulated by an icosahedral capsid containing 162 capsomeres. Within the capsid is double-stranded DNA of about 230 kilobases that is enclosed by a lipid bilayer envelope. The diagnosis of CMV requires laboratory confirmation and cannot be made on clinical grounds alone. Seroconversion, monoclonal antibody, and PCR are among the tools available for diagnosis.
Antibody response (whether maternal or fetal) may play a less important role in transmission and, subsequently, in the development of symptomatic disease than other viral factors, such as the amount of maternal viremia or the timing in pregnancy during which infection and subsequent maternal-to-fetal transmission occurs.
Epidemiology
Seroprevalence of CMV among adult populations shows significant geographic variability. Rates range from 40% to 100% depending on the region surveyed, with most infections being detected solely by serologic screening (i.e., asymptomatic). Countries or regions with low socioeconomic status usually demonstrate very high seroprevalence rates. The risk of seroconversion during pregnancy for a seronegative woman is approximately 2%. However, because congenital infections can follow either primary disease (maternal seroconversion) or reactivation of disease, congenital infections may occur in a much higher percentage of cases. There are two periods in life when infection rates are particularly high-perinatal (related to mother-to-child transmission, breast-feeding, and child-to-child) and reproductive age (putatively related to sexual transmission).
Pregnant women acquire CMV infection either through exposure to infected children (the infected children shed virus in urine, saliva, and nasopharyngeal secretions and are infectious for a prolonged period) or through sexual contact. About 1% to 2% of American women shed CMV from their cervix at any given time, although that number is higher in women with multiple sexual partners.
Pathophysiology
Following primary infection, the virus goes into a latent phase. Intermittent periods of reactivation frequently occur, and virus is again excreted in the nasopharynx, cervix, urine, saliva, and breast milk. Maternal viral shedding increases throughout pregnancy, and neonatal infection becomes increasingly likely as cervical shedding rates increase toward term.
Acquisition of disease by neonates at the time of birth is much more common than congenital infection, but much less devastating. The former types of infection come from CMV carried in the cervix during the late stages of pregnancy and from CMV in breast milk. After primary infection during the prenatal period, the congenital infection rate has been reported to be as high as 55%. Although the great majority of infants with congenital CMV are asymptomatic, infants with symptomatic infection at birth usually have findings associated primarily with the reticuloendothelial and central nervous systems. The severe cases of congenitally acquired symptomatic infection result from primary rather than recurrent infection in the pregnant woman and from infections occurring early in pregnancy. The most common findings in these cases include hepatosplenomegaly, jaundice, a generalized petechial rash, and microcephaly. Less common findings include chorioretinitis with or without optic atrophy, pneumonitis, cerebral calcifications, microphthalmia, microcephaly, seizures, and cerebral and cerebellar atrophy. The mortality of newborns with symptomatic disease is approximately 30%.
Diagnosis
Mother
The great majority of primary maternal infections with CMV are asymptomatic and unrecognized. When symptoms do occur, they may go unrecognized because they appear as a mild, infectious, mononucleosis-like illness, with lymphadenopathy, fatigue, and slight fever. Even though CMV is hepatotropic, elevations in liver enzyme levels rarely are seen in primary or recurrent infections. About one third to two thirds of all pregnant women have IgG antibodies to CMV, indicating previous infection. Detection of CMV-specific IgM during the acute phase of infection is useful for making the diagnosis of CMV infection, but only 80% of women with primary infection demonstrate this antibody. In addition, more than a third of mothers with recurrent CMV from latent infection will be positive for CMV-specific IgM. CMV-specific IgM antibodies may persist from 4 to 9 months, and some test methods may give false-positive results because of cross-reactions with other herpes viruses, antinuclear antibody, or rheumatoid factor. A microneutralization assay may be of some utility. In one study, the antibodies did not appear until 15 weeks after infection and persisted thereafter.
Thus its absence, shortly after a presumed infection, can help to rule out infection.
Viral culture is the gold standard for the diagnosis of CMV infection. Virus usually is detected in the cervix, nasopharynx, and urine of infected individuals. However, culture results are positive with both primary and recurrent infections. PCR has also proven to be quite useful in the detection of CMV. Quantitative PCR testing has proven particularly useful in the management of AIDS and transplant in patients whose high serum levels augur complications, and it offers an early opportunity for the initiation of antiviral therapy.
Prenatal
Sonography may be useful for identifying some abnormalities in the fetus that may be related to CMV infection. Nonspecific sonographic findings of fetal hydrops, intrauterine growth retardation, polyhydramnios, fetal ascites, and specific central nervous system anomalies (e.g., ventriculomegaly, periventricular calcifications) suggest an intrauterine infection, possibly CMV, and should be evaluated further with amniocentesis or cord blood sampling.
Lynch and others reported the successful prenatal diagnosis of fetal infection by a combination of amniotic fluid culture and measurement of total and CMV-specific IgM and γ-glutamyl transpeptidase in fetal blood samples. In a study of 189 pregnancies with known outcome, Enders and colleagues reported 89.5% sensitivity with the use of amniotic fluid and fetal blood assessments for virus and anti-CMV IgM. They noted that the correct diagnosis of in utero infection with CMV by amniotic fluid analysis could be expected after 21 weeks of gestation and at least 6 weeks after the diagnosis of infection in the mother. Other authors have reported similar sensitivities and have pointed out that those instances of false-negative results often are associated with infants with minimal stigmata of disease. Further studies are needed on the sensitivity and specificity of these methods in identifying infected infants prior to birth.
The definitive diagnosis is best made through the detection of viruria during the first week of life. The presence of IgM antibodies in cord serum is suggestive but not sufficiently specific to make a definitive diagnosis.
Management
Mother
Susceptible pregnant women have a 2% risk of seroconverting while pregnant. Because this is a relatively low risk of infection and there is a somewhat low rate of possible damage to the fetus, and because no effective therapy is available to infected infants, routine serologic testing of pregnant women generally is not recommended.
Probably the single most important method of preventing primary infection during pregnancy is minimizing exposure in high-risk areas, such as nurseries, day care centers, and other places that have a high concentration of young children. Careful hand-washing techniques, as well as proper handling of potentially infectious body fluids, should be instituted to minimize spread of the infection.
There are no protocols for the use of antiviral agents (acyclovir and ganciclovir) during pregnancy to decrease the risk of mother-to-child transmission of CMV. It is hoped that a vaccine ultimately will be developed that will prevent CMV. Some authors have suggested that in order to target both mothers most likely to transmit to infants and those most likely to transmit to pregnant women, that young unmarried mothers and all toddlers and preteen children should be considered for vaccination.
Infant
Clinically evident infection in newborns and infants is treated symptomatically. In the most severe cases of neonatal infection, antiviral agents such as acyclovir and ganciclovir have been used to suppress the infection, but discontinuation of the medication results in reappearance of the infection. Foscarnet (phosphonoformic acid) has been used in primary symptomatic neonatal disease and has been effective in reducing viral shedding.
HERPES SIMPLEX VIRUS
Introduction
Herpes simplex virus (HSV) causes an extremely common STD with potentially devastating consequences for the perinatally infected neonate. It also is the cause of more calls to the national STD hotline than any other STD. Management of these infections in pregnancy has undergone substantive evolution over the last decades, reducing reliance on operative delivery.
Microbiology
HSV is an encapsulated double-stranded DNA virus that infects susceptible mucosal surfaces. The DNA-containing core is surrounded by an envelope that contains viral proteins that are important for attachment to host cell receptors, cell penetration, and immune escape mechanisms. There is approximately a 50% homology in the DNA sequences of HSV type 1 (HSV-1) and HSV type 2 (HSV-2), the latter being predominantly genital and the former predominantly oral. Either virus can cause serious illness in the neonate. Immunity to one virus may lead to an attenuated illness if an individual becomes infected with the other.
Epidemiology
Whereas only approximately 5% of the reproductive-age population gives a history of clinical herpes infections, serologic surveys suggest that up to 20% of the sexually active population has had genital herpes. That represents a 30% rise in the age-adjusted risk since the late 1970s. The National Health and Nutrition Evaluation Survey revealed that slightly over 25% of women are seropositive for HSV-2 (Fig. 19.5). A large majority of those individuals who are seropositive will shed virus intermittently for many years after the initial infection. The rate of shedding does not vary significantly between those with and those without a history of genital herpes. The reason that most seropositive women are unaware of their status may be related to the presence of antibody to HSV-1 that offers sufficient protection from HSV-2 to mute any symptoms. However, with proper education a large percentage of women can be trained to recognize recurrences. The infection is transmitted from both symptomatic and asymptomatic individuals, with the latter being the cause of the majority of new infections. Transmission seems to occur with greater facility from men to women. However, condoms seem to be more effective in preventing transmission from men to women than from women to men. There is little evidence to suggest that pregnancy, in itself, increases either the frequency or the severity of genital HSV infections. In one study, in an unselected patient population, HSV shedding occurred in only 0.1% to 0.4% of all deliveries. In other studies of pregnant women with histories of HSV, positive culture results have been found in 0.2% to 7.4% of asymptomatic women. It has been estimated that if PCR is used in lieu of cultures, shedding would be detected about 8 times as frequently. The rate of shedding at term is no higher than at other times in pregnancy. Unfortunately, from the therapeutic perspective, most women who shed viruses are not aware that they are shedding or that they are seropositive for HSV.
FIG. 19.5. HSV-2 seroprevalence according to the lifetime number of sexual partners, adjusted for age, for black and white men and women in National Health and Nutrition Evaluation Survey (1988–1994). Bars indicate 95% confidence intervals. (From Fleming DT, McQuillan GM, Johnson RE, et al. Herpes simplex virus type 2 in the United States, 1976–1994. N Engl J Med 1997;337:1105–1111, with permission.) |
Pathophysiology
HSV has an incubation period of 2 to 10 days, followed by a primary infection that is characterized by focal vesicle formation and a pronounced cellular immune response. The infection enters a latent phase, with the virus ascending peripheral sensory nerves and coming to rest in nerve root ganglia. Recurrent exacerbations occur intermittently, stimulated by poorly understood mechanisms. Infection may be primary, with a 2- to 3-week course; recurrent, with a 5- to 10-day course; or asymptomatic. Primary infection poses the greatest risk to both mother and infant. Clinically it is extremely difficult to distinguish primary from recurrent disease. Fifty percent of infants born vaginally to mothers with a primary infection will themselves have HSV infection, compared with only 4% of those born to mothers with recurrent infection. Primary infection also may be associated with abortion, low birth weight, preterm birth and, rarely, congenital infection, although less commonly than was once thought. As expected, viral shedding occurs for a significantly longer period with primary infection (1–2 weeks) than with recurrent infection (3–6 days).
Antibodies appear approximately 7 days following the onset of primary infection, reaching a peak in 2 to 3 weeks, and generally remain detectable for life. Titers do not rise significantly with recurrent infections. The significant difference in infection rates in neonates born to mothers with primary infections (50%) versus those born to women with recurrent infections (4%) suggests that maternal antibodies provide some protection. With recurrent infection, the infection rate has been estimated to range from zero to 8% when virus is present at the time of vaginal delivery.
Fortunately, disseminated primary HSV infections in pregnant women are rare. When these infections become disseminated during pregnancy, mortality is high for both the mother and fetus, and systemic antiviral therapy is recommended.
Although HSV shedding occurs in approximately 0.1% to 0.4% of deliveries, the frequency of neonatal infection is actually much lower, occurring in approximately 0.01% to 0.4% of deliveries. The reason for this 10-fold difference in infection between mother and newborn is unknown but probably is related to both a protective benefit of maternal antibodies and to the size of the viral inoculum to which the fetus is exposed during birth.
The viral inoculum associated with asymptomatic shedding is several logs less than that associated with primary infection. Because of the inability to mount a timely immune response, maternal seroconversion around the time of delivery may be particularly likely to result in an infected neonate.
Neonatal infection may result from either HSV-1 or HSV-2. The majority of cases of neonatal HSV infections, however, are caused by HSV-2, with approximately 76% of isolates from infected neonates being HSV-2. It is estimated that approximately 90% of cases of neonatal HSV infections can be traced to a maternal source of infection and are secondary to either ascending infection with ruptured membranes or to colonization during the actual birth process.
As noted, approximately 50% of mothers delivering vaginally with a primary genital HSV infection will give birth to an infant with HSV infection. Of these infected neonates, 60% will die during the neonatal period. Of equal concern is that approximately 50% of the survivors will have significant sequelae, such as microcephaly, mental retardation, seizures, microphthalmia, retinal dysplasia, chorioretinitis, meningitis, encephalitis, hypertonicity, apnea, and coma. Although less common, the neonate may also acquire infection from exposure during the neonatal period from either or both parents or from other family members, from other infected infants, or from infected health care workers.
Although there have been isolated reports of “congenitally infected” fetuses with malformations, other investigators have found no such associations. Congenital infections have been defined by the presence of skin vesicles or scarring, chorioretinitis, hydranencephaly, microphthalmia, microcephaly, or an abnormal CT scan of the brain within the first week of life.
Diagnosis
The diagnosis of primary infection is relatively easy to make from a clinical standpoint. There generally are multiple painful ulcers on an inflamed surface. The lesions progress to ulcers, and there may be accompanying painful adenopathy and fevers. However, recurrent infections may be more difficult to diagnose solely on clinical grounds, and other ulcerative lesions (e.g., syphilis, chancroid) can occur, also. Virus isolation by tissue culture remains the most accurate method of confirming the diagnosis of HSV infection and should be utilized if there is any lack of certainty on clinical grounds. Cultural sensitivity decreases with the duration of the lesion. Cytologic tests, such as Pap smear and Tzanck preparations, tend to be less accurate. Rapid accurate tests for HSV, such as PCR, have been demonstrated to have utility in research settings and undoubtedly will eventually establish a role in clinical medicine.
Management
In the nonpregnant individual there are many treatments now available that will shorten the course or ameliorate symptoms. Even before antiviral therapy is instituted, there are some therapies that can be offered. Application of topical cool compresses with Burow solution for 15 minutes 4 to 6 times daily may be helpful for patients with extensive erosions on the genitalia. For an initial case of genital herpes, several medications are available.
· Acyclovir ointment can be applied every 3 to 6 hours (6 times daily) for 7 days. Severe cases may be treated with intravenous acyclovir 5 milligrams per kilogram, infused at a constant rate over 1 hour, every 8 hours for 7 days in patients with normal renal function) or oral acyclovir 200 milligrams 5 times daily for 7 to 10 days.
· Valacyclovir tablets (1 g b.i.d. for 10 days).
For the woman with a recurrent lesion there are also several choices. One is acyclovir (200 mg p.o. 5 times a day for 5 days, 400 mg p.o. b.i.d. for 5 days, or 800 mg p.o. b.i.d. daily for 5 days), generally started during the prodrome or within 2 days of onset of lesions. Famciclovir also is useful for treatment of recurrent genital herpes (dose is 125 mg q12h for 5 days in patients with normal renal function) started at the first sign of symptoms. Finally, valacyclovir (dose is 500 mg q12h for 5 days in patients with normal renal function) can be used.
In pregnancy there are additional considerations: First, it should be remembered that HSV is an STD, so those individuals diagnosed during pregnancy should be screened for other STDs. In regard to the prevention of mother-to-child transmission of HSV, management has changed greatly over the last decades. It was once recommended that pregnant women suspected of having a genital HSV infection or who had a history of prior HSV infection or sexual partners with HSV infections be monitored closely, with third-trimester cultures, for recurrent infection or for asymptomatic HSV shedding. Those women who had a positive culture result near to term were scheduled for elective cesarean section in order to avoid contact between the neonate and the virus during delivery.
That protocol had little impact on the overall incidence of neonatal HSV infection for several reasons. Most women who delivered infants with neonatal HSV infections had no history of infection and no lesions at the time of delivery, and thus would not have met the criteria for monitoring in the first place. It was calculated that a weekly screening strategy would cost approximately $1.8 million for each case of neonatal HSV infection averted. It is not clear whether the neonatal morbidity and mortality actually prevented by such a protocol outweighed the maternal mortality that would result from complications of cesarean delivery. Moreover, cesarean delivery, even with intact membranes, will not prevent all cases of neonatal herpes infections secondary to maternal genital HSV infection. There are many reported cases of neonatal HSV infections in infants that were delivered by cesarean with intact membranes.
Recommendations for pregnant women with HSV infections include the following:
· Cultures should be performed to confirm the diagnosis when a woman has active HSV lesions during pregnancy. If there are no visible lesions at the onset of labor, vaginal delivery is acceptable.
· Weekly surveillance cultures of pregnant women with a history of HSV infection, but no visible lesions, are not necessary and vaginal delivery is acceptable.
· Amniocentesis in an attempt to rule out intrauterine infection is not recommended for mothers with HSV infection at any stage of gestation.
· Cesarean section should be performed if lesions are present at the onset of labor.
It has been suggested that oral acyclovir prophylaxis might have a role in the prevention of recurrence and, thereby, could reduce the need for cesarean section. Several studies have pointed to lowered shedding and cesarean section rates when women with outbreaks during pregnancy are started on acyclovir in later stages of pregnancy. Cost savings with that strategy have been suggested, based upon decision analysis. Although acyclovir is a nucleoside that incorporates into DNA, registries have not revealed an untoward rate of adverse outcomes among exposed infants. Preliminary trials with acyclovir in pregnancy have reported that the need for cesarean section was reduced, with no increase in morbidity for exposed neonates.
Intrapartum
It is recommended that term patients who have visible lesions, are in labor, and have ruptured membranes should undergo cesarean delivery. There has been a suggestion that the policy of cesarean section in such circumstances for women with recurrent lesions should be rethought. The presence of antibody in such women may make the mother-to-child HSV transmission rate so low as make the risks of routine cesarean section unjustified. However, most clinicians still opt for cesarean section with visible lesions regardless of the mother's history.
Although it has been classically taught that cesarean delivery of women with visible lesions should be performed only if membranes have been ruptured for less than 4 to 6 hours, not all neonates born to mothers with HSV infections become infected, even with membranes ruptured for more than 24 hours. In contrast, neonates born by cesarean delivery within 2 hours of rupture of membranes have developed HSV infection.
For patients with active HSV infections and premature rupture of membranes remote from term, there are not enough data to recommend a management protocol that would apply in all clinical situations. The risk of extreme immaturity must be weighed against the risk of neonatal HSV infection. There have been several case reports of infants born without sequelae after prolonged periods of conservative management.
Although the use of scalp electrodes has been implicated as a rare etiology of neonatal infection, monitoring by fetal scalp electrode is not contraindicated if needed to adequately assess fetal condition in women with history of HSV infection but without lesions or symptoms.
It has been suggested that cultures be obtained at delivery in women with a history of HSV infection to aid the pediatrician in assessing the need for therapy in the newborn.
Treatment
There is no known cure for HSV infection. The purine analogue acyclovir has been used to treat both primary HSV infections and to prevent recurrent HSV infections in nonpregnant women.
Prevention
Although it is not necessary to isolate the infant from the infected mother, the infected parturient should be counseled regarding hand washing and good hygiene to prevent infection of the infant. Every effort should be made to avoid direct contract of the newborn with herpetic lesions.
INTRAAMNIOTIC INFECTION
Introduction
Clinicians continue to apply a number of terms to this entity, including chorioamnionitis, amnionitis, intrapartum infection, amniotic fluid infection, and IAI. We use the last designation to distinguish this clinical syndrome from bacterial colonization of amniotic fluid and from histologic inflammation of the cord or placenta. Prospective studies published in the last few years report rates of 4% to 10% among non-privately insured and privately insured patients, respectively.
Pathogenesis
The ascending route is the most common pathway for development of IAI. A hematogenous or transplacental route of infection occurs with Listeria monocytogenes. Other virulent organisms, such as group B streptococci, may lead to a similar blood-borne infection. IAI may occur infrequently as a complication of invasive diagnostic procedures, such as amniocentesis or intrauterine transfusion.
In large investigations using logistic regression analysis, risk factors for IAI were identified as follows: low parity, prolonged duration of membrane rupture, prolonged duration of labor, larger number of vaginal examinations, and duration of internal fetal monitoring. In regard to the use of internal fetal monitoring, we feel that this technique should be used if it enables the practitioner to diagnose and treat labor abnormalities more efficiently.
Once IAI develops, the fetus may swallow and aspirate the infected fluid and is prone to develop pneumonia, enteritis, meningitis, and sepsis. Indeed, a clinical diagnosis of IAI is one of the most important risk factors for neonatal sepsis.
Microbiology
IAI is a polymicrobial infection, involving both aerobic and anaerobic bacteria. Table 19.2 shows the most common isolates in the amniotic fluid of over 400 cases of IAI.
TABLE 19.2. Amniotic fluid isolates in 404 cases of intraamniotic infection |
Diagnosis
The most common clinical criteria are fever, leukocytosis, and ruptured membranes; fetal and maternal tachycardia are noted in variable percentages of cases. Foul-smelling amniotic fluid and uterine tenderness, although more specific signs, occur in a minority of cases. In cases of clinical IAI, maternal fever was present in 85% to 99%, fetal tachycardia in 37% to 82%, maternal tachycardia in 19% to 37%, uterine tenderness in 13% to 16%, and foul-smelling amniotic fluid in 9% to 22%. The determination of amniotic fluid glucose concentration is a practical test for diagnosing clinical IAI. With an amniotic glucose concentration less than 5 milligrams per deciliter, the likelihood of amniotic fluid culture results being positive is approximately 90%. On the other hand, when the amniotic fluid glucose concentration is greater than 20 milligrams per deciliter, then the likelihood of a positive culture result is approximately 2%. At intermediate values (for example 14 and 15 mg/dL), the likelihood of a positive amniotic culture result is 30% to 50%.
Management
When IAI is diagnosed, there is a need for delivery of the fetus and for antibiotics. With regard to timing of delivery, there has been excellent maternal–neonatal outcome without use of arbitrary time limits. Cesarean delivery has been performed for standard obstetric indications, not for IAI alone. In nearly all cases, delivery occurred within 8 hours after diagnosis of IAI, and the mean time was between 3 and 5 hours. No critical interval from diagnosis of amnionitis to delivery could be identified. Cesarean section rates are higher among patients with IAI, running 2 to 3 times greater than in the general population. The reason for the increase most likely results from two things. First, IAI commonly develops in the patients with dystocia as an underlying problem. Second, the uterus with IAI is less sensitive to oxytocin.
The benefits of intrapartum treatment have been well established. Intrapartum initiation of antibiotic treatment improves maternal outcome, decreases neonatal bacteremia, and does not result in delayed sepsis (Table 19.3).
TABLE 19.3. Outcome by maternal treatment group in cases of intraamniotic infection: results of a randomized study |
As noted above, the traditional antibiotic approach to treatment has been with combination therapy, primarily broad-spectrum penicillin with an aminoglycoside, plus clindamycin in some cases (such as cesarean delivery or apparent sepsis). Because of the expense and complexity of such therapy, there has been recent interest in single-drug treatment of IAI. In view of the well-described microbes involved, there would be several reasonable choices of single-agent therapy, but there have not been sufficient comparative trials to recommend alternative single-agent therapy. Studies have addressed the issue of duration of antibiotic therapy after delivery in the treatment of IAI. In one study, all patients received ampicillin plus gentamicin intrapartum, and those patients delivering vaginally where randomized to receive cefotetan 2 gram intravenously as a single dose or as multiple doses every 12 hours for 48 hours. Patients randomized to single-dose therapy also left the hospital more quickly (33 versus 57 hours, P = .001). Patients who received single-dose therapy also had a higher rate of “failed therapy” (11% versus 3.7%, P = .27). Although this difference did not achieve statistical significance, a three-fold increase in failed therapy raises concern about the limitation of single-dose therapy.
In a study of antibiotic therapy after cesarean delivery for IAI, all patients received ampicillin during labor and clindamycin and gentamicin, one dose each, preoperatively. Patients were then randomized to receive no scheduled postoperative antibiotics versus clindamycin and gentamicin until they were afebrile (for a minimum of at least 24 hours). No patients in either group developed an abscess or readmission for endometritis. Although there was no significant difference in endometritis (14.8% in those with no routine antibiotic versus 21.8% in those treated with clindamycin and gentamicin), there was a nonsignificant, but still 2.5-fold, increase in wound infection rate in patients randomized to no routine antibiotics (5% versus 1.8% with those randomized to clindamycin and gentamicin postpartum). Based upon the limited observations of these two trials, the nagging concerns about failed therapy in the first study, and the wound infection rates in the second study, it would appear premature to conclude that therapy, after either vaginal delivery or cesarean delivery, in the presence of IAI, is unnecessary.
Outcome
Short-term Outcomes
From descriptive studies largely from the early 1980s, several strong consistent observations may be made. Maternal outcome was excellent, with bacteremia occurring in only 2% to 6%. Maternal outcome was more complicated in patients having cesarean delivery. No critical diagnosis to delivery interval was demonstrable. Specifically, neither prenatal mortality nor maternal complications correlated with more prolonged intervals from diagnoses of IAI to delivery. Yet all patients delivered within 4 to 12 hours, and nearly all patients received intrapartum antibiotics. Although prenatal mortality was relatively high, little of this was directly attributable to infection, because most of the prenatal mortality was due to accompanying prematurity. Prior to term pregnancy, neonates have a higher frequency of complications if they are delivered of mothers with IAI. The group with IAI had a significantly higher number with respiratory distress syndrome and any diagnosis of infection.
Thus, IAI has a significant adverse effect upon the mother and neonate. Short-term outcome is dependent largely on the organisms in the amniotic fluid (with Escherichia coli and group B streptococci more likely to result in maternal or neonatal bacteremia), low birth weight (with low-birth-weight infants faring more poorly), and timing of antibiotic therapy (with intrapartum administration improving outcome).
Long-term Outcomes
Traditional complications of IAI have been maternal and neonatal sepsis, neonatal pneumonia and meningitis, and neonatal death. Studies indicate that complications of IAI should be expanded to include periventricular leukomalacia, cerebral palsy, respiratory distress syndrome and, perhaps, other neonatal complications. The hypothesized mechanism is that ascending infection leads to placental and congenital infection. This leads to an overly exuberant production of inflammatory cytokines which, in turn, leads to cell damage. Evidence for these expanded neonatal complications after IAI are as follows:
1. Intrauterine exposure to maternal or placental infection is associated epidemiologically with an increased risk of cerebral palsy.
2. Levels of inflammatory cytokines are increased in the amniotic of infants with brain white matter lesions or respiratory distress syndrome.
3. Experimental intrauterine infection has led to brain white matter lesions in rabbits.
Prevention
Within in the last few years, several intervention strategies for preventing clinical IAI have been evaluated. These are summarized in Table 19.4.
TABLE 19.4. Clinical measures to prevent IAI |
POSTPARTUM ENDOMETRITIS
Introduction
Seven decades into the antibiotic era, genital tract infections continue to pose a common and occasionally severe threat to women after childbirth. Although substantial progress has been made in the control of puerperal sepsis, infection still ranks as the fourth most common cause of maternal death.
The most common cause of puerperal fever is uterine infection, occurring in approximately 1% to 3% of women after vaginal delivery and up to 27% after cesarean delivery, even when prophylactic antibiotics are used. The infection is variously called endometritis, endoparametritis, or simply metritis. Criteria for endomyometritis include fever, uterine tenderness, and purulent or foul lochia, peripheral leukocytosis, and exclusion of another infected site. Nonspecific signs and symptoms such as malaise, abdominal pain, chills, and tachycardia may be present. In the vast majority of cases of uterine infection, initial signs and symptoms develop within the first 5 days after delivery.
Pathophysiology
The major risk factors are summarized in Table 19.5. Cesarean section is the most important predisposing clinical factor for pelvic infection. The severity of infection also is increased in abdominal delivery. Those patients with electively scheduled operations (with no labor and no rupture of membranes) have lower infection rates than those with emergency or nonelective procedures. This observation has been made nearly universally in a large number of studies.
TABLE 19.5. Major risk factors for postpartum infection |
Risk factors of labor, rupture of membranes, vaginal examinations, and internal fetal monitoring are intricately interrelated, and even sophisticated statistical techniques may not be able to discern which of these factors is an independent variable. Whether amniotomy increases postpartum infection has been the subject of many reports. Amniotomy, as part of a plan of active labor management, probably does not increase the risk. Regardless of race, indigent patients have higher puerperal infection rates than do middle-class patients. The cause is unclear, but differences in flora, hygiene, and nutrition have all been postulated as reasons. Diabetes is also a risk factor for endometritis, with diabetic patients having approximately a four-fold increase in risk. Among patients without labor or ruptured membranes, 9.1% (5/55) of diabetics developed endometritis or wound infection versus 1.8% (2/110) of nondiabetics (P = .042). Among patients with either labor or rupture of membranes or both, the infection rate for diabetics was 25% (6/24) versus 6.3% (3/48) for nondiabetics (P = .032). Endometritis most often is caused by a mixture of aerobic and anaerobic bacteria from the genital tract. Table 19.6 shows the isolates from one carefully done study.
TABLE 19.6. Endometrial isolates (collected by a triple-lumen catheter) from 51 patients with postpartum endometritis |
Diagnosis
The diagnosis of endomyometritis usually is based on symptoms of fever, malaise, abdominal pain, and purulent or foul lochia. Other sources of fever should be excluded. In clinical investigations, various temperature criteria have been used in the definition of endometritis. These include 100°F (37.8°C) and 100.4°F (38°C) on two or more occasions. When there are signs of infection, multiple risk factors for infection, or persistence of low-grade fever in the puerperium, it is reasonable to presume a genital tract infection is present and proceed with the workup and treatment. We believe that appropriate specimens include a complete blood count, blood cultures, and an aerobic uterine culture. Gram staining of the genital specimen may be helpful when hemolytic streptococci, clostridia, or other anaerobes are suspected.
Treatment
With supportive therapy and appropriate antibiotics, the vast majority of patients improve within 1 to 3 days. For patients with endomyometritis after vaginal delivery, the treatment should include good anaerobic coverage, probably using a single agent (such as broad-spectrum penicillin or cephalosporin or a penicillin–β-lactamase inhibitor combination).
Among patients with endomyometritis after cesarean section, response to antibiotics is poorer. Initial therapy for endometritis after cesarean section should consist of broad-spectrum antibiotics, with activity against anaerobes as well as Gram-positive and Gram-negative aerobes. In many clinical trials, the combination of clindamycin and aminoglycoside has been considered the standard for comparison for treatment in genital tract infections after cesarean section. A large number of new penicillins and cephalosporins have become available to treat postpartum infection. Although no single agent provides activity against the entire bacterial spectrum, most have sufficient aerobic and anaerobic activity to merit use in endometritis. Table 19.7 gives specific regimens.
TABLE 19.7. Selected regimens for initial parenteral therapy of postpartum endometritis |
The new penicillins and cephalosporins usually are tolerated very well and have few side effects. Administration of a single agent requires less time and equipment, but the higher cost of most new agents must also be considered. Metronidazole has excellent anaerobic activity. Because it has little activity against aerobes, its use as single-agent therapy is unwise. Its use in combination with gentamicin still leaves the Gram-positive aerobes (notably group B streptococci and enterococci) uncovered. New antibiotics are being formulated that may replace gentamicin as initial therapy. The monobactams (e.g., aztreonam) have exquisite Gram-negative activity, but they have little activity in the rest of the bacterial spectrum. These agents have the same spectrum as the aminoglycosides and have fewer side effects. Clinical trials of these monobactams in combination with clindamycin have shown excellent results, equivalent to clindamycin plus gentamicin.
For patients who respond promptly to parenteral antibiotics, some questions arise: How long should therapy be continued? Is oral therapy needed as an adjunct? It has been recommended that parenteral therapy should be continued for 24 to 48 hours after the patient becomes afebrile and asymptomatic. Then intravenous antibiotics may be discontinued and the patient discharged without oral antibiotics, unless the patient has had staphylococcal bacteremia. The approach is based upon several descriptive studies plus a randomized trial of oral antibiotic therapy after successful intravenous therapy in women with puerperal endometritis. Patients given placebo had similar subsequent courses to those of patients given oral amoxicillin (after parenteral therapy). Side effects were also similar in both groups (Table 19.8).
TABLE 19.8. Oral amoxicillin after intravenous antibiotics for postpartum endometritis |
At the other end of the patient response spectrum, there are patients who do not respond within 48 to 72 hours of appropriate antibiotic therapy. Diagnostic considerations are: (a) an infected “mass” such as abscess or hematoma of the wound or pelvis, extensive pelvic cellulitis, septic pelvic thrombophlebitis, or retained placenta; (b) a resistant organism such as enterococci in patients treated with cephalosporin-like antibiotics or with clindamycin plus gentamicin; (c) a nongenital source of infection such as pyelonephritis, pneumonia, or intravenous catheter phlebitis; (d) a noninfectious fever such as drug fever or factitial fever; or (e) inadequate dosage or inadequate route of otherwise correct antibiotics (Fig. 19.6). Appropriate bedside examination and review of the chart and cultures often reveal the cause. In general, in patients who are stable and not seriously ill, an appropriate change in antibiotics (such as adding penicillin to clindamycin plus gentamicin to add coverage for enterococci) is effective in about 80% of those who had not responded in the initial 48 hours.
FIG. 19.6. Causes of poor response to antibiotics. (From Sweet RL, Gibbs RS. Postpartum infection. In: Sweet RL, Gibbs RS. Infectious diseases of the female genital tract. Philadelphia: Lippincott Williams & Wilkins, 2002, with permission.) |
In about 20% of cases, the initial failure to respond is due to resistant organisms, and in another 30%, no cause for the failure is identified. We believe that 48 to 72 hours is an appropriate point to change therapy unless the patient is unstable, when changes are needed more promptly. In the remaining 40% to 50%, the reason for poor response to initial therapy is an infected mass. If the mass or collection is present in the wound after cesarean section, physical examination usually identifies the source. In other cases, radiographic studies are helpful in identifying pelvic masses or deep-seated wound infection.
Ultrasonography or a CT study may reveal a mass in the pelvis. In patients with persistent pelvic infection after vaginal delivery, an ultrasonographic examination is helpful, because it may reveal a pelvic mass, retained placental tissue, or septic pelvic thrombophlebitis. Sonography is readily available, is inexpensive, and requires no special preparation. When patients are obese or have an open wound, ultrasonography provides a limited examination, and computed tomography is useful (Fig. 19.7 and Fig. 19.8).
FIG. 19.7. Left parauterine abscess in a patient after cesarean delivery. (From Sweet RL, Gibbs RS. Postpartum infection. In: Sweet RL, Gibbs RS. Infectious diseases of the female genital tract. Philadelphia: Lippincott Williams & Wilkins, 2002, with permission.) |
FIG. 19.8. Computed tomography scans showing percutaneous placement of a pigtail catheter in an abscess after cesarean section. (From Sweet RL, Gibbs RS. Postpartum infection. In: Sweet RL, Gibbs RS. Infectious diseases of the female genital tract. Philadelphia: Lippincott Williams & Wilkins, 2002, with permission.) |
Prevention
Principles for use of antibiotic prophylaxis to prevent endometritis after cesarean delivery are well established. These are summarized in Table 19.9.
TABLE 19.9. Recommendations for use of prophylactic antibiotics in cesarean section |
SUMMARY POINTS FOR HIV, TOXOPLASMOSIS, CMV, AND HERPES
· All pregnant women should have HIV testing recommended to them.
· HIV-infected pregnant women should have regular monitoring of their CD4 cell count and viral load.
· All HIV-infected women should receive ZDV per the PACTG 076 protocol.
· Standards of care for HIV infection should be upheld during the prenatal period.
· During the intrapartum period, attempts should be made to minimize the duration of ruptured membranes among women with HIV infection.
· All pregnant women should be counseled regarding ways in which to minimize the risk of acquiring infection with T. gondii.
· Pregnant women who acquire T. gondii infections should have studies performed to ascertain the status of the fetus.
· If congenital infection with T. gondii is diagnosed and the patient continues the pregnancy, antiparasitic therapy should be instituted.
· CMV is the most common congenital infection in the United States.
· Although patients with antibodies may have reactivation resulting in infected newborns, these children are less frequently and less severely infected then those whose mothers have primary infections.
· Primary, first episode, genital HSV infection may increase the risk of untoward perinatal events, including preterm birth, low birth weight, and congenital infection.
· Viral shedding in labor should be determined by clinical assessment and should be managed with an operative delivery.
· Nucleoside analogues may be used to reduce the frequency of HSV viral shedding and, thereby, the need for cesarean section.
SUMMARY POINTS FOR INTRAAMNIOTIC INFECTION AND POSTPARTUM ENDOMETRITIS
· Both clinically evident IAI and postpartum endometritis are polymicrobial, upper genital tract infections, involving anaerobes, aerobes, and genital mycoplasmas most commonly.
· The benefits of intrapartum treatment of IAI with broad-spectrum antibiotic therapy are well established.
· In addition to causing maternal and neonatal sepsis, neonatal pneumonia, meningitis, and neonatal death, intraamniotic infection is now strongly associated with long-term neonatal adverse outcomes including cerebral palsy, respiratory distress syndrome, and other complications.
· Several effective measures have been identified to prevent clinical IAI.
· Response of postpartum endometritis is very good to a variety of antibiotic regimens, which have coverage against likely aerobes and anaerobes.
· Oral therapy generally is not needed after successful parenteral therapy for endometritis.
· Antibiotic prophylaxis strategies are well established for cesarean section.
SUGGESTED READINGS
HIV
American College of Obstetricians and Gynecologists. Scheduled cesarean delivery and the prevention of vertical transmission of HIV infection. Committee Opinion No. 234, May 2000.
Centers for Disease Control and Prevention. HIVAIDS Surveillance Report. Vol 11. Atlanta: US Dept. of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, 1999.
Connor EM, Sperling RS, Gelber R, et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994;331:1173–1180.
Dickover RE, Garratty EM, Herman SA, et al. Identification of levels of maternal HIV-1 RNA associated with risk of perinatal transmission. JAMA1996;275:599–605.
French R, Brocklehurst P. The effect of pregnancy on survival in women infected with HIV: a systemic review of the literature and meta-analysis. Br J Obstet Gynecol 1998;105:827–835.
Hader SL, Smith DK, Moore JS, et al. HIV infection in women in the United States. Status at the millennium. JAMA 2001;285:1186–1192.
Ioannidis JPA, Abrams EJ, Ammann A, et al. Perinatal transmission of human immunodeficiency virus type 1 by pregnant women with RNA virus loads <1,000 copies/mL. J Infect Dis 2001;183:539–545.
Johnson VA, Woods C, Hamilton CD, et al. Vertical transmission of multi–drug-resistant human immunodeficiency virus type 1 (HIV-1) and continued evolution of drug resistance in an HIV-1–infected infant. J Infect Dis 2001;183:1688–1693.
Landesman S, Kalish L, Burns D, et al. The relationship of obstetrical factors to the mother-to-child transmission of HIV-1. N Engl J Med 1996;334:1617–1623.
Martin R, Boyer P, Hammill H, et al. Incidence of premature birth and neonatal respiratory distress in infants of HIV-positive mothers. The pediatric pulmonary and cardiovascular complications of vertically transmitted human immunodeficiency virus study group. J Pediatr 1997;131:851–856.
Shaffer N, Chauwoowang R, Mock PA, et al. Short course zidovudine for perinatal HIV-1 transmission in Bangkok, Thailand: a randomized control trial. Lancet 1999;353:773–780.
The European Mode of Delivery Collaboration. Elective caesarean section versus vaginal delivery in prevention of vertical HIV-1 transmission: a randomized clinical trial. Lancet 1999;353:1035–1039.
The International Perinatal HIV Group. The mode of delivery and the risk of vertical transmission of human immunodeficiency virus type 1: a meta-analysis of 15 prospective cohort studies. N Engl J Med 1999;340:977–987.
Wade N, Birkhead GS, Warren BL, et al. Abbreviated regimens of zidovudine prophylaxis and perinatal transmission of human immunodeficiency virus. N Engl J Med 1998;339:1409–1414.
Watts H, Minkoff H. Managing pregnant patients. In: Dolin R, Masur H, Saag M, eds. AIDS therapy, second edition. Churchill Livingstone, New York (in press).
Toxoplasmosis
American College of Obstetrician and Gynecologists. Perinatal viral and parasitic infections. ACOG Practice Bulletin No. 20, Washington DC, ACOG, 2000.
Centers for Disease Control and Prevention. Preventing congenital toxoplasmosis. MMWR 2000;49(RR-2):57–75.
Foulon W, Villena N, Stray-Pedersen B, et al. Treatment of toxoplasmosis during pregnancy: a multicenter study of impact of treatment on fetal transmission and child's sequelae at age 1 year. Am J Obstet Gynecol 1999;180:410–415.
Guerina NG, Hsu HW, Meissner HC, et al. Neonatal serologic screening and early treatment for congenital Toxoplasma gondii infection. The New England regional working group. N Engl J Med 1994;330:1858–1863.
Hohfeld P, Daffos F, Costa JM, et al. Prenatal diagnosis of congenital toxoplasmosis with a polymerase-chain-reaction test on amniotic fluid. N Engl J Med1994;331:695–699.
Jones JL, Lopez A, Wilson M, et al. Congenital toxoplasmosis: a review. Obstet Gynecol Surv 2001;50:296–305.
Lisenfeld O, Montoyua JG, Kinney S, et al. Effect of testing for IgG avidity in the diagnosis of Toxoplasma gondii infection in pregnant women. J Infect Dis2001;183:1248–1253.
Minkoff H, Remington J, Holman S, et al. Vertical transmission of toxoplasma by HIV infected women. Am J Obstet Gynecol 1997;176:555–559.
Cytomegalovirus
Azam AZ, Vial Y, Fawer CL, et al. Prenatal diagnosis of congenital cytomegalovirus. Obstet Gynecol 2001;97:443–448.
Enders G, Bader U, Lindemann L, et al. Prenatal diagnosis of congenital cytomegalovirus in 189 pregnancies with known outcome. Prenat Diagn2001;21:362–377.
Lynch L, Daffos F, Emanuel D, et al. Prenatal diagnosis of fetal cytomegalovirus infection. Am J Obstet Gynecol 1991;165:714.
Murph JR, Bale JF. The natural history of acquired cytomegalovirus infection among children in group day-care. Am J Dis Child 1988;142:843.
Pass RF. Immunization strategy for the prevention of congenital cytomegalovirus infection. Infect Agents Dis 1996;5:240–244.
HSV
Braig S, Luton D, Sibony O, et al. Acyclovir prophylaxis in late pregnancy prevents recurrent genital herpes and viral shedding. Eur J Obstet Gynecol Reprod Biol 2001;96:55–58.
Brown ZA, Benedetti J, Selke S, et al. Asymptomatic shedding of herpes simplex virus at the onset of labor: relationship to preterm labor. Obstet Gynecol1996;87:483–488.
Brown ZA, Vontver LA, Benedetti J, et al. Effects on infants of a first episode of genital herpes during pregnancy. N Engl J Med 1987;317:1246–1251.
Brown ZA, Vontver LA, Benedetti J, et al. Genital herpes in pregnancy: risk factors associated with recurrences and asymptomatic viral shedding. Am J Obstet Gynecol 1985;153:24–30.
Fleming DT, McQuillan GM, Johnson RE, et al. Herpes simplex virus type 2 in the United States, 1976–1994. N Engl J Med 1997;337:1105–1111.
Hensleigh PA, Andrews WW, Brown Z, et al. Genital herpes during pregnancy: inability to distinguish primary and recurrent infections clinically. Obstet Gynecol 1997;89:891–895.
Randolph AG, Hartshorn RM, Washington AE. Acyclovir prophylaxis in late pregnancy to prevent neonatal herpes: a cost-effectiveness analysis. Obstet Gynecol 1996;88:603–610.
Scott LL, Sanchez PJ, Jackson GL, et al. Acyclovir suppression to prevent cesarean delivery after first-episode genital herpes. Obstet Gynecol 1996;87:69–73.
Wald A, Langenberg AGM, Link K, et al. Effect of condoms on reducing the transmission of herpes simplex virus types 2 from men to women. JAMA2001;285:3100–3106.
Walt A, Zeh J, Selke S, et al. Reactivation of genital herpes simplex virus type 2 infection in asymptomatic seropositive persons. N Engl J Med2000;342:844–850.
Intraamniotic Infection
Alexander JM, Gilstrap LC, Cox SM, et al. Clinical chorioamnionitis and the prognosis for very low birth weight infants. Obstet Gynecol 1998;91:725–729.
Chapman SJ, Owen J. Randomized trial of single-dose versus multiple-dose cefotetan for the postpartum treatment of intrapartum chorioamnionitis. Am J Obstet Gynecol 1997;177:831–834.
Gibbs RS, Blanco JD, St Clair PJ, et al. Quantitative bacteriology of amniotic fluid from patients with clinical intra-amniotic infection at term. J Infect Dis1982;145:1.
Gibbs RS, Dinsmoor MH, Newton ER, et al. A randomized trial of intrapartum versus immediate postpartum treatment of women with intra-amniotic infection. Obstet Gynecol 1988;74:562.
Gibbs RS. Chorioamnionitis and bacterial vaginosis. Am J Obstet Gynecol 1993;169:460–462.
Mercer BM, Miodovnick M, Thurnau GR, et al. Antibiotic therapy for reduction of infant morbidity after preterm premature rupture of membranes. A randomized controlled trial. National Institute of Child Health and Human Development, Maternal-Fetal Medicine Units Network. JAMA 1997;278:989–995.
Mozurkewich EL, Wolf FM. Premature rupture of membranes at term: a meta-analyses of three management schemes. Obstet Gynecol 1997;89:1035–1043.
Romero R, Yoon BH, Mazor M, et al. The diagnostic and prognostic value of amniotic fluid white blood cell count, glucose, interleukin-6, and gram stain in patients with preterm labor and intact membranes. Am J Obstet Gynecol 1993;169:805–816.
Sperling RS, Newton E, Gibbs RS. Intra-amniotic infection in low-birth-weight infants. J Infect Dis 1988;157:113.
Turnquest MA, How HY, Cook CR, et al. Chorioamnionitis: is continuation of antibiotic therapy necessary after cesarean section? Am J Obstet Gynecol1998;179:1261–1265.
Postpartum Endometritis
Berg CJ. Pregnancy-related mortality in the United States. Obstet Gynecol 1996:88;161–167.
Brown CE, Stettler RW, Twickler D, et al. Puerperal septic pelvic thrombophlebitis: incidence and response to heparin therapy. Am J Obstet Gynecol1999;181:143–148.
Diamond MP, Entman SS, Salyer SL, et al. Increased risk of endometritis and wound infection after cesarean section in insulin-dependent diabetic women. Am J Obstet Gynecol 1986;155:297.
Dinsmoor MJ, Newton ER, Gibbs RS. A randomized, double-blind, placebo-controlled trial of oral antibiotic therapy following intravenous antibiotic therapy for postpartum endometritis. Obstet Gynecol 1991;77:60–62.
Garite TJ. Amniotomy in labor: routine or reserved for specific indications? Contemp Ob Gyn 1994;(May):39–58.
Newton ER, Prihoda TJ, Gibbs RS. A clinical and microbiologic analysis of risk factors for puerperal endometritis. Obstet Gynecol 1990;75:402–406.
Rosene K, Eschenbach DA, Tompkins LS, et al. Polymicrobial early postpartum endometritis with facultative and anaerobic bacteria, genital mycoplasmas and C. trachomatis: treatment with piperacillin or cefoxitin. J Infect Dis 1986;153:1028.