Ultrasound Characteristics of In Utero Infection

In utero infection of the fetus has become recognized as an important cause of fetal and neonatal morbidity and mortality. Since both anatomic and functional abnormalities have been described in the fetus related to various infections, ultrasonography may be a valuable diagnostic tool in this regard. A complete review of the current literature was undertaken to report available information on this topic. Common pathogens or clinical conditions were selected. The identified data were confounded by the way in which each case originally presented for study. Although certain anomalies were frequently associated with individual organisms, their incidence could not be determined, nor were most specific to that infectious agent. Representative ultrasound images are presented for common and unusual cases.

number of pathogenic organisms increases, so too does our appreciation of the potential for in utero diagnosis of these infections. Both anatomic and functional abnormalities in the fetus have been described for many types of infection. The focus of this report is on the use of ultrasonography as a tool to recognize, and monitor, these infectious complications. Although a great deal of recent literature has become available concerning invasive means (e.g., amniocentesis, cordocentesis) to diagnose in utero infection, this is beyond the scope of the current review.
Most of the data to be presented are derived from limited sources, including case reports or small case series. These sources are biased by the way the case presents, i.e., if infection was recognized prospectively, or if a given anomaly led to a thorough workup of the patient which included potential infectious agents. Therefore, associations may be drawn, but absolute cause and effect relationships should be tempered by the quality of the data. Similarly, an accurate incidence of a given ultrasound-based anomaly is hard to establish due to the rarity of most infectious processes.

Cytomegalovirus (CMV)
CMV remains the most common congenital virus infection, affecting an estimated 0.4-2.3% of all liveborn infants. 1,e The virus itself is a doublestranded DNA herpes virus with multiple strains identified. An increased incidence of seropositivity is encountered in lower socioeconomic strata with variation between regions of the country. Diagnosis of a primary maternal infection requires demonstration of the presence of CMV-IgM antibodies in the absence of CMV-IgG, as well as seroconversion of CMV-specific IgG antibody from negative to positive. Presence of IgG antibodies without documented seroconversion is not helpful in diagnosing the primary maternal infection thought responsible Ultrasound manifestations of non-immune fetal hydrops: (a) hydrothorax, (b) ascites, () polyhydramnios, (d) scalp for in utero infection. Approximately 2% of seronegative patients will convert during pregnancy. Of these primary CMV infections, 40% result in fetal infection. Only 10% of fetal infections result in significant disease. " It is in this group that ultrasound findings consistent with fetal infection would be expected. The role of prenatal diagnosis of fetal infection is not clear in the absence of ultrasound evidence of fetal disease. Isolation of CMV from amniotic fluid was first reported in 1971, 4 and amniotic fluid viral culture remains the gold standard for diagnosis of fetal infection. The use of fetal blood sampling for culture and detection of CMV IgM is not as sensitive as amniocentesis, s Non-specific serologic findings including anemia, thrombocytopenia, elevated liver function tests, and elevated total IgM levels support a diagnosis of in utero infection but do not necessarily correlate with severity of fetal disease.
Many of the manifestations of CMV infection in the neonate may be detected prenatally and warrant further evaluation. The ultrasound findings associated with CMV are non-specific and include hepatosplenomegaly (74%), microcephaly (50%), intrauterine growth retardation (IUGR) (41%), and cerebral calcifications. 6 The presence of fetal ascites and/or non-immune hydrops ( Fig. 1) is more frequently noted with severe infections. The intracranial calcifications may be accompanied by ventriculomegaly ( Fig. 2) or intrahepatic calcifications, however, mild ventriculomegaly may be an isolated finding. Subtle signs include oligoor polyhydramnios. Less common ultrasound findings described include pulmonary hypoplasia, isolated pericardial or pleural effusions, and echogenic bowel ( As with all fetal infections, the ultrasound diagnosis of abnormalities indicates a more severe disease and worse prognosis. The findings associated with CMV are not unique to this infection, however, the relative frequency of CMV infection makes it more commonly encountered. Any patient with ultrasound findings of in utero infection warrants inclusion of CMV in the differential. Rubella Rubella (German measles) is a well-characterized, mild exanthematous disease of childhood caused by a single-stranded RNA virus of the togavirus   family. In the pregnant host, this disease can have catastrophic effects on the fetus, as first described by Gregg in 1941.11 The last epidemic in the prevaccine era (1964)(1965) resulted in approximately 11,000 miscarriages, abortions, and stillbirths, and approximately 20,000 cases of congenital rubella syndrome in the newborn, le Since the introduction of an effective vaccine in 1969, the incidence of clinical infection has dramatically declined, although there has been a recent resurgence of rubella identified by the Centers for Disease Control and Prevention (CDC). 1  These findings are tempered by others noting that these lesions develop over time and fetuses at risk should be followed serially. Several authors have cautioned that too early an examination of the fetus may miss evidence of fetal infection. Prctorius et al. zl performed serial ultrasound examinations in VZV-infccted gravidas and noted a latency period of at least 5 weeks before ultrasound findings of infection were detected. Similarly, Harding and Baumcr z4 reported on a case of maternal varicella infection at 14 weeks gestation with a normal ultrasound examination at 17 weeks, and a severely compromised neonate with contracturcs and posrural deformities involving all extremities at delivery. Additionally, a report of maternal varicella at 11 weeks gestation, with serial ultrasonography every 4 weeks, did not reveal abnormalities until 22 weeks, when atrophy of the muscles of the left leg and malposition of the left foot were noted, z- The second type of in utero presentation, nonimmune hydrops, also appears long after the initial maternal VZV infection has subsided. Three gravi-das with varicella infection before 20 weeks of gestation were noted to have ultrasound features in the fetus consistent with viral infection only in the third trimester, z6-zs These findings included hepatomcgaly, ascitcs, pleural effusion, pcricardial effusion, and liver calcifications.

Parvovirus
Parvovirus B19 may infect both children and adults, and generally has a benign course in these patients. Significant disease in the fetus may be caused by in utcro infection. Sonographic signs of fetal infection have been reported from 3 to 12 weeks after maternal infection, z9 Numerous case reports have documented a broad range of antenatal ultrasound findings in infected fetuses. [3][4] When abnormalities are detected during antenatal ultrasound, they are almost exclusively abnormal fluid collections and are believed to be due to fetal anemia and/or myocarditis. Therefore, placental thickening, polyhydramnios, pleural and pericardial effusions (Fig. 7), ascites, skin edema, and cardiomegaly are the most frequently found manifestations.
Others cases may present as a fetal demise, 35 twin gestation with clinical findings of twin-twin transfusion syndrome, 36 or oligohydramnios. 37 Congenital anomalies have rarely been associated with documented parvovirus infection. Although anomalies are suggested by a report of a fetus diagnosed at 13 weeks (cystic hygroma and no midline intracranial echo) with progression at 16 weeks (no midline intracranial echo, thick nuchal fold, fetal hands and feet in abnormally flexed position, growth at less than the 5th percentile), these may be coincidental. Despite a normal karyotype, the pregnancy was terminated and parvovirus was found in multiple tissues, though none was found in maternal blood. 3s A recent report of two cases of hydrocephalus in association with this fetal infection, after normal second trimester ultrasound examinations, 4 does raise the possibility that congenital anomalies may bc in the spectrum of disease.
The in utero natural history of this disease process, in particular the ultrasound-related findings, is poorly defined. This is due to the relative rarity of this condition, the variable gcstational age at diagnosis, and the now common attempt at therapeutic intervention (e.g., in utero transfusion).
However, serial ultrasound studies of even severely affected fetuses have shown resolution of the hydropic changes. 33 Humphrey and colleagues 39 reported on a 26 week fetus with parvovirus infection and hydrops, anhydramnios, thickened placenta, bilateral pleural effusions, cardiomcgaly, pericardial effusion, and marked ascitcs. With observation, these findings had all resolved by 32 weeks gestation.

BACTERIAL INFECTIONS
Syphilis The incidence of syphilis in the general U.S. population has increased significantly over the past few years, with a current rate of approximately 14-15 cases per 100,000 persons. 4 Although the exact incidence in pregnancy is undetermined, a 4-fold increase in congenital syphilis cases has recently been reported. 4e Unfortunately, congenital syphilis is first suspected in many cases when the ultrasound report demonstrates a fetal demise, 4 consistent with the 22% incidence of fetal demise in mothers with syphilis reported in the 1930s. 4e These features of fetal demise have been well described elsewhere. 44 Fetal infection can be diagnosed in utero, and may be the cause of ultraso.nographic findings of non-immune hydrops. Classi- cally, the feature of placental thickening (Fig. 8) is reported in the literature as consistent with this in utero infection. Additional findings may include ascites and skin edema. 4s, 46 Although the data are limited, Nathan et al. 47 reported the incidence of ultrasound-detected hepatomcgaly to be 57%, placental thickening 62%, and ascites 17% in infected fetuses. Ultrasound detection of dilatation of fetal bowel has been reported and is associated with fibrotic and infiltrative processes causing obstruction. 4s, 49 The characteristic bone findings of congenital syphilis have also been detected in utero, s Both abnormal bone morphometry (three standard deviations below normal for gcstational age) and morphology (thickening, curvature, and multiple areas of bowing) were described. Despite the numerous other potential observable ultrasound features of congenital syphilis, none has been reported in the recent literature. Data are lacking with regard to the effect of aggressive antimicrobial therapy and the resolution of syphilis-related ultrasonographic abnormalities.

Intraamniotic Infection (Chorioamnionitis)
As previously mentioned, in addition to anatomic abnormalities, ultrasound-based functional aberrations in the fetus have bccn reported which arc associated with infection. Vintzileos and colleagues s were the first to introduce the concept of using biophysical profile testing to detect occult fetal infection. In their study of 73 gravidas with premature rupture of the membranes, a low biophysical score (-<7) was a good predictor of impending fetal infection. When subdivided into individual components, the first manifestations of impending infection were loss of fetal breathing and non-reactive non-stress testing. Conversely, the presence of fetal breathing had the highest specificity in predicting absence of fetal infection. In this series, testing was done on a daily basis, and it appears that this frequency is necessary to provide accurate fetal assessment related to infection.
Use of biophysical profile testing for subclinical infection remains controversial. Miller et al. se were not able to identify a relationship between any component of the biophysical profile and subsequent clinical chorioamnionitis. This study was limited by the absence of any identified neonatal sepsis in the study group. This has prompted the suggestion that biophysical profile testing may be more accurate in its ability to detect impending subclinical fetal infection as opposed to maternal infection. The rationalization for this latter perspective is that the loss of reactivity, movement, breathing, and tone in the fetus is similar to these same non-specific signs seen with newborn sepsis in the nursery. In a recent study by Carroll et al., s study subjects with premature rupture of the membranes underwent ultrasound testing, cordocentesis, and amniotic fluid sampling in an attempt to identify in utero infection. Although there was a tendency for lower biophysical profile scores and amniotic fluid indices, the majority of subjects with documented in utero infection had normal ultrasound-based testing.

PARASITIC INFECTIONS Toxoplasma
Antenatal infection with Toxop/asma gondii is seen frequently in Europe and may cause significant sequelae. The initial infection leads to focal tissue necrosis and destruction, most often in the brain, heart, lungs, and liver. Classic manifestations of severe disease include choriorctinitis, convulsions, hydrocephalus, and intracranial calcifications. Of these, the latter two may be diagnosed by antenatal ultrasound. Abnormalities of the cerebral ventricles ( Fig. 9), ranging from mild ventriculomegaly to severe hydrocephalus, s4 and in rare instances to hydranencephaly, ss are the most frequent prenatal ultrasound findings in this infection. The dilatation of the ventricles is usually symmetrical s and may develop weeks after the prenatal diagnosis has been madc.S7-s9 Intracranial calcifications without distinct pattern may be seen in utero, s Characteristic intracranial findings and their progression have also been noted in utero, with documented fetal toxoplasmosis infection. Mild ventriculomegaly and periventricular hyperechoic areas of -<2 mm were first detected at 30 weeks of gestation and repeat ultrasound at 33 weeks revealed progressive hydrocephalus, increase in the size (10 mm) and number of hyperechoic areas, and polyhydramnios. Nonspecific findings of infection have also been documented: ascites, hepatomegaly, increased placental thickness, calcifications in the fetal liver, pericardial and pleural effusions, oligohydramnios, and intrauterine growth restriction, s<61-6 Though the presence of ultrasound findings may prompt investigation, it should be noted that the majority (approximately 80%) of prenatally diagnosed cases of congenital toxoplasmosis do not have detectable ultrasound abnormalities, s4,s9,l'4 CONCLUSIONS Although some infections acquired in utero have classic ultrasound findings, there is a significant overlap between infectious agents. Non-immune hydrops, intrauterine growth restriction, ventriculomegaly, and fluid abnormalities form a significant component of the ultrasound findings in these infections. Therefore, in the absence of more specific features, any patient warranting evaluation for these findings would benefit from the inclusion of infectious etiologies in the diagnostic process.