Intrahepatic Administration of Liposomal Amphotericin B (Ambisome) for the Management of a Liver Abscess from Candida albicans in a Preterm Infant

Case report.A preterm neonate (gestational age, 28 weeks; body weight, 1,000 g; Apgar score, 7 to 8) was born from twin pregnancy by cesarean section due to life-threatening labor. The baby was intubated at birth and treated with surfactant, and an umbilical venous catheter (UVC) was placed for parenteral nutrition and drug administration. At 20 h of life, the baby was extubated and ventilated by nasal continuous positive airway pressure.

On the 8th day of life (DOL), clinical conditions worsened, the baby required reintubation and mechanical ventilation, and the UVC was replaced by a central venous catheter (CVC) through the left arm. On DOL 12, a right abdominal tender elastic mass (d = 48 × 50 mm) with perihepatic fluid was detected by abdominal ultrasound (US) and computed tomography (CT) scanning. Therefore, the baby was referred to our neonatal intensive care unit (NICU). At admission, abdominal US and CT scans showed an intrahepatic, ovalar, tender mass (d = 65 × 45 × 35 mm) that appeared multilobulate with a nonuniform colliquated core and occupied most of the right hepatic lobe, with dislocation of the hepatic vessels and gallbladder (Fig. 1A, B, C). Presumably, parenteral nutrition extravasation through the UVC represented the causative mechanism, with subsequent hemorrhagic infarction. Neonatal conditions were severe; the neonate was mechanically ventilated and assisted with inotropes. A two-lumen CVC through the left jugular vein was placed. On DOL 17, respiratory and cardiocirculatory conditions critically worsened, and peripheral and CVC blood cultures turned positive for C. albicans. Susceptibility tests were performed according to Clinical and Laboratory Standards Institute (CLSI) guidelines M27-A3 (7) and M27-S4 (8). The isolated Candida strain was susceptible to micafungin (MIC, 0.015 µg/ml), anidulafungin (MIC, 0.03 µg/ml), fluconazole (MIC, 0.25 µg/ml), itraconazole (MIC, 0.06 µg/ml), and voriconazole (MIC, <0.008 µg/ml) and wild type to amphotericin B (MIC, 0.5 µg/ml). Among the effective antifungals, we chose micafungin because our NICU developed a method for monitoring plasma concentrations of such drugs via heelstick microsampling and the baby did not present with liver failure (9). Therefore, we promptly started intravenous micafungin (Mycamine 8 mg/kg daily), but no improvement in clinical conditions was achieved. Peripheral and CVC blood cultures continued to be positive after 14 days of micafungin therapy, despite CVC replacement and micafungin blood concentrations within therapeutic ranges (10), i.e., 3.1, 18.0, 16.0, and 10.4 µg/ml 1 h before and 1, 2, and 8 h after infusion. Because the infection persisted, we added intravenous liposomal amphotericin B (3 to 5 mg/kg daily) to micafungin and started the lock therapy of CVC with micafungin (5 mg/liter in 70% ethanol, closing each lumen for 12 h) (11), with the aim to preserve the CVC considering the critical conditions. Due to continued positive blood cultures, the CVC was newly replaced, but blood cultures once more remained positive for C. albicans, and clinical conditions did not improve. US scans of the brain, kidneys, and spleen and echocardiography did not detect localizations of the infection or endocardial vegetation that could justify the persistence of the fungemia. On DOL 25, due to an increasing amount of ascitic fluid, a tunneled intraperitoneal catheter (Argyle, 43 cm) was placed, which resulted in the exit of ∼100 ml of citrine liquid. However, the clinical conditions again did not improve, blood cultures were persistently positive, and peritoneal fluid grew the same species of C. albicans. On DOL 35, suspecting a Candida superinfection of the preexisting hepatic lesion, after obtaining informed consent for the procedure from the patient's parents, we placed an intralesional 6F pigtail catheter under US guidance; as expected, 4 ml of purulent material was drained, and this grew C. albicans (Fig.1E). Therefore, we decided to administer 5 ml of amphotericin B diluted in 0.9% saline solution (1 mg/ml) once daily directly inside the liver abscess, closing the pigtail catheter for 30 min. Ten days later, blood and intralesional fluid cultures were finally negative, and neonatal conditions progressively improved. The intrahepatic lesion appeared markedly reduced at the next hepatic US assessment (d = 27 × 35 × 32 mm). The entire course of intravenous antifungal therapy lasted 4 weeks. Brain magnetic resonance imaging detected a bilateral subependymal hemorrhage, but the neurologic development of the baby at the 6-month follow-up was normal for the corrected chronological age.

• Open in new tab
• Download powerpoint

FIG 1

(A) X-ray imaging showing an intrahepatic lesion caused by parenteral nutrition extravasation through the umbilical venous catheter. Ultrasound (B), Doppler (C), and computed tomography (D) imaging showing an ovalar intrahepatic lesion involving the right lobe, with dislocation of the hepatic vessels and gallbladder. (E) X-ray imaging showing the left peritoneal catheter and intrahepatic drainage.

Fungal liver abscesses are rare during the neonatal period and often constitute a severe complication of umbilical vessel catheterization, central parenteral nutrition catheters, necrotizing enterocolitis, and prematurity. Untreated, they lead to serious clinical sequelae and high mortality rates (12).

Preterm infants are more predisposed than others in relation to physiological immunodeficiency and prolonged exposure to NICU risk factors, such as invasive procedures to resistant nosocomial germs requiring prolonged antibiotic therapies and steroids.

Currently, liver abscesses in adult patients are treated with antibiotics along with percutaneous needle aspiration or percutaneous catheter drainage, and surgical drainage is used only in patients who fail to respond to such treatment. In abscesses >5 cm, catheter drainage performs better than repeated needle aspiration (13). In neonates, the optimal treatment of fungal liver abscesses is still uncertain, and clinical experiences are poor and limited to sporadic reports (1–3). The difficulties of repeated needle aspirations in infants with very low weight (our patient weighed 1.4 kg) and the underlying severe clinical conditions should be considered.

The diagnosis of liver abscess is difficult at first and requires a high degree of suspicion, especially in neonates with very low birth weight. In the early stage of abscess formation, radiography is not conclusive. In the more advanced stages, the suspicion may arise due to the presence of intralesional air, which is revealed by radiographic examination of the abdomen. The availability of US has been an important step toward diagnosis of this condition because it allows for repeatable, noninvasive screening in case of suspicion; guarantees an early diagnosis; and, in relation to this precocity, offers a higher survival rate. CT scan is the definitive tool for the diagnosis of hepatic abscesses.

The baby described in the present case report was transferred to our NICU with a suspected hepatic abscess, so we did not have firsthand knowledge of the medical situation. It was difficult to achieve infection resolution because the abscess was extensive and probably constituted a continuous source of contamination also of the peritoneum, where C. albicans was also present; and the drugs administered systemically could not penetrate the lesion in concentrations sufficient to sterilize it. In our case, the placement of an echo-guided catheter into the hepatic abscess and the intralesional administration of liposomal amphotericin B allowed the resolution of a critical life-threatening situation that had not responded to any other intensive treatment. Further investigations are required to confirm such practice in neonates.