Although the vessel walls of aorta and inferior vena cava are shown to have antithrombotic properties in the young compared to adults (524,525), these vessels can often be damaged by intravascular catheters in sick newborns. Indeed, intravascular catheters have become the single most important risk factor for neonatal thrombotic disease (522). The
placement of such catheters carries a potential risk of thrombosis, regardless of which vessel is used; possibly because of their widespread use, umbilical catheters have received the most attention to date. In a retrospective review of 4,000 infants who underwent umbilical artery catheterization in the 1970s, severe symptomatic vessel obstruction was observed in approximately 1% of patients (526). Clinically silent thromboses have been detected more often at autopsy or during contrast angiography in asymptomatic infants with an umbilical artery catheter in place. Between 3% and 59% of cases have had postmortem evidence of catheter-related thrombosis (527,528,529,530,531,532,533). In six prospective angiographic studies, thrombosis was demonstrated in 10% to 95% of patients (534,535,536,537,538,539). Very recently, thrombi associated with a correctly placed umbilical venous catheter were documented venographically at the time of elective catheter removal in 30% (14/47) of asymptomatic infants (540). In the same study, it was found that real-time and Doppler flow ultrasonography was not sufficiently accurate to diagnose such thrombi. Before this report, neonatal thrombosis, which was associated with central venous catheters, had been estimated to occur in approximately 14% of asymptomatic infants, based on serial echocardiography (541,542).

Abnormal chorionic vessels are present in a range of maternal disorders such as PIH or placental infection. These abnormal blood vessels predispose to the development of chorionic thrombi, which can embolize into the fetal circulation, particularly into the pulmonary arteries and portal veins (543). Paradoxic emboli have also been reported as the cause of neonatal cerebral infarction (544,545).

Neonatal TE can occur in either the venous or arterial system. In the Canadian Registry, thrombosis was venous, arterial, and mixed in 62, 34, and 4 %, respectively (522). The vessels affected are numerous; they include renal, adrenal, portal, hepatic, and cerebral veins; peripheral, cerebral, pulmonary, coronary, renal, and mesenteric arteries; and the aorta, vena cava, and right atrium (546). The clinical presentation of thrombosis is variable. Signs and symptoms may be fairly specific, such as a cool, pale, and pulseless limb in the case of peripheral arterial obstruction. Often, however, the clinical presentation is very nonspecific: examples include respiratory insufficiency and profound hypoxemia caused by pulmonary emboli (556) and seizures caused by arterial stroke or cerebral venous thrombosis (557,558,559,560,561). Occasionally, thrombosis is suspected only retrospectively when sequelae of vessel occlusion become apparent (e.g., portal hypertension as a consequence of portal vein thrombosis). Table 46-21 summarizes the major sites of thrombotic vessel obstruction and their clinical symptoms and signs in the newborn infant.

The most common predisposing factor for thrombosis is the presence of a catheter (522,523). In thrombosis unrelated to a catheter, renal vein thrombosis is the most common (523). Thrombocytopenia often accompanies thrombosis in newborns. Patients should be evaluated for a thromboembolic disorder if thrombocytopenia cannot be explained by other conditions.

Purpura fulminans is characterized by ecchymotic lesions that increase in a radial fashion and become purplish black with bullae, and then necrotic and gangrenous. In the presence of purpura fulminans, protein C or protein S deficiencies should be considered.

The literature on neonatal thrombotic disease consists almost entirely of case reports and case series (562). Such anecdotal observations are poor guides to management, as they are uncontrolled and fraught with bias (562). Extrapolation of findings in adult patients to newborn infants may be inappropriate because the etiology, the localization of thrombi, the coagulation system and its response to antithrombotic and fibrinolytic agents differ markedly between the two age-groups. Current approaches to the management of neonatal thrombosis remain to be validated in future clinical trials. Updated information on the use of antithrombotic agents are published regularly by American College of Chest Physicians (563).

Whenever thrombotic disease is suspected in the newborn, every effort should be made to confirm or refute the
diagnosis. Contrast angiography with the use of nonionic contrast media is the “gold standard” imaging technique for the confirmation of thrombosis before embarking on thrombolytic or surgical therapy. Other, less invasive tests such as real-time ultrasonography or Doppler flow studies may be helpful adjunctive measures, but their precision and accuracy in neonatal thrombotic disease are still uncertain (564). The advantages of this technique are that it is noninvasive, does not require exposure to ionizing radiation, and can be performed at the bedside.