Propranolol Therapy for Congenital Chylothorax Free

A retrospective chart review of prenatally diagnosed CC treated with prenatal or postnatal propranolol was performed (Columbia University institutional review board AAAS2707). Medical charts were evaluated for CC course, time to resolution, length of hospital stay, concurrent treatments, and complications. Severe CC was defined as pleural effusions that contained >1000 cells/mL (>80% lymphocytes)²⁶  that caused mediastinal shift, hydrops, Doppler abnormalities, and/or severe/rapidly progressing effusions. Pregnant women were initiated at 20 mg 4 times daily and increased to 40 mg 4 times daily.²⁷  Neonates who received postnatal propranolol were started at 0.3 mg/kg/d in divided doses 3 times a day and increased to 1 to 2 mg/kg/d based on clinical response. Patients were monitored for adverse effects of propranolol, including bronchospasm, bradycardia, and hypoglycemia.

Demographics and prenatal and postnatal courses for 4 patients treated with propranolol are summarized in Fig 1 and Tables 1 and 2.

Case 1 presented with ultrasonographic (USG) evidence of CC at 18 weeks 4 days (18w4d), confirming unilateral pleural effusion with mediastinal shift of the fetal heart and lungs and compression of the right thoracic wall (Fig 2A, Table 1). Thoracentesis fluid was consistent with CC. The mother was initiated on propranolol 20 mg 4 times daily and increased to 40 mg 4 times daily by 19w0d (Fig 1, Table 1). USG confirmed CC resolution at 24w3d, 38 days after dose of 40 mg 4 times daily was achieved. Propranolol was continued for the remainder of the pregnancy until repeat cesarean delivery at 37w0d.

Apgar scores were 8 and 9 and respiratory support was not required. Pleural effusions were not apparent on chest radiograph (Fig 2A, Table 2). On day of life 2 (DOL2) the neonate was transferred to the neonatal ICU (NICU) for hypoglycemia thought to be due to low birth weight and unrelated to propranolol, returned to the well-baby nursery on DOL3, and discharged on DOL4 on room air (RA) without medications. The child is well 5 years later and has not received further treatment.

Case 2 presented at 21w5d with left pleural effusion causing mediastinal shift, lung compression, ascites, and hydrops requiring multiple thoracenteses (Fig 2B, Table 1). Thoracentesis fluid was consistent with CC. Maternal propranolol was initiated at 20 mg 4 times daily at 22w3d and increased to 40 mg 4 times daily at 26w0d (Fig 1, Table 1). USG demonstrated CC resolution at 30w4d, 32 days after reaching the maximum dose. Propranolol was continued for the remainder of the pregnancy, with delivery at 39w3d.

Apgar scores were 7 and 8. The neonate was admitted to the NICU for respiratory distress. Postpartum chest radiograph showed residual effusion (Fig 2B, Table 2). The infant was treated with continuous positive airway pressure (CPAP) and was weaned to RA after 2 hours, transferred to the well-baby nursery on DOL2, and discharged on DOL5 without medications. The child is well 5 years later and has not received further treatment.

Case 3 presented at 29w6d with bilateral pleural effusions with lung compression, midline shift of the heart, and scalp edema on prenatal USG; hydrops developed at 30w2d. Thoracentesis fluid was consistent with CC, which persisted despite multiple interventions (Fig 1, Table 1). The mother developed preterm premature rupture of membranes at 34w3d and delivered vaginally.

Apgar scores were 7 and 8. The neonate was admitted to the NICU for respiratory distress (Table 2) and was treated with CPAP until DOL4. Pleural effusion persisted despite repeated thoracenteses. Bilateral chest tubes were placed on DOL4 and 7 and remained in situ through DOL15. Octreotide was administered from DOL10-38. The patient was discharged from the hospital after 1.5 months on RA and MCT formula without medications.

The patient was readmitted to the hospital (DOL85) for respiratory syncytial virus infection and pleural effusion. He was discharged after his respiratory status stabilized (DOL88), but readmitted (DOL90) for persistent respiratory distress and unilateral pleural effusion requiring intubation and chest tube placement. Pleural fluid was consistent with CC and octreotide drip was initiated on DOL93 (Table 2). The thoracic duct was ligated (DOL100) for persistent high drainage and octreotide was discontinued. The patient was extubated to CPAP on DOL103. Chest tube output temporarily decreased, but reaccumulated on DOL107 (Fig 3A,C). Propranolol was initiated on DOL108 at 0.3 mg/kg/d and increased to 2 mg/kg/d (DOL130; Table 2). Chest tube output decreased and was removed on DOL120. Significant radiographic improvement of pleural effusion was observed 29 days after propranolol initiation, and 7 days after 2 mg/kg/d was achieved (Fig 3A,C; Table 2). The patient was discharged from the hospital on RA (DOL171) with high MCT formula, furosemide, and propranolol.

The patient was readmitted (DOL232-246) for pleural effusion requiring a chest tube. The patient was discharged from the hospital on furosemide and propranolol with a clinically insignificant persistent right pleural effusion on RA. Furosemide was discontinued by 2 years and propranolol by 3 years. At 8 years, he is healthy and performing sports.

Case 4 was diagnosed with bilateral pleural effusion and trace ascites on USG at 27w1d requiring multiple thoracenteses (Fig 1, Table 1). Thoracentesis fluid was consistent with CC. Propranolol was not offered because of a prolonged fetal PR interval. The mother developed preterm premature rupture of membranes at 33w6d and delivered by repeat cesarean section at 34w0d.

Apgar scores were 4, 5, and 7. The neonate was intubated because of respiratory distress, admitted to the NICU, and underwent bilateral chest tube placement for pleural effusions (Fig 3B, Table 2) for 14 days. The neonate was extubated on DOL4 to CPAP. Because of worsening pleural effusion, propranolol was initiated at 0.3 mg/kg/d (DOL22) and increased to 0.7 mg/kg/d (DOL25; Table 2). Radiologic improvement of pleural effusions was observed by DOL32, which stabilized by DOL35, 10 days after 0.7 mg/kg/d of propranolol was achieved (Fig 3B, Table 2). Because of residual effusions, propranolol was increased to 1 mg/kg/d at DOL54. The neonate required ongoing diuretic therapy and CPAP until DOL57, when he was weaned to RA 35 days after propranolol initiation. The baby was discharged from the hospital on RA on DOL68 on diuretics and propranolol. Chest radiograph at 8 months demonstrated resolution of pleural effusions, 206 days after 1 mg/kg/day (Fig 3B, Table 2). Propranolol was discontinued at 28 months of age and absence of pleural effusions was confirmed on echocardiography at 31 months.

Our case series suggests that treatment with prenatal or postnatal propranolol was beneficial for severe refractory CC. Fetal patients demonstrated complete resolution of pleural effusion by USG 38 and 32 days after 40 mg 4 times daily was achieved. Although postnatal CC stabilization was observed 10 and 7 days after 0.7 to 2 mg/kg/d propranolol was achieved, CC resolution by imaging occurred months later. Prenatally treated patients delivered at term, required no postnatal intervention, and were discharged from the hospital from the well-baby nursery. By contrast, postnatally treated neonates had persistent prenatal pleural effusions, were delivered preterm, and experienced persistent and recurrent pleural effusions that required prolonged hospitalizations.

Current postnatal CC management can include use of MCT feeds and treatment with octreotide or sirolimus.^1,9,10  Octreotide has been described as both efficacious and inefficacious in treating postnatal CC.^9–14  Octreotide can affect hormonal balance leading to hypoglycemia and hypothyroidism and has been implicated in the development of necrotizing enterocolitis.^1,11  Octreotide administration also requires injections that may be challenging in the outpatient setting. Sirolimus treatment of vascular anomaly patients has shown risks for developing hematopoietic, gastrointestinal, and metabolic toxicities.²⁸  Sirolimus has been used to treat fetal cardiac rhabdomyomas, but its use to treat prenatal CLAs has not been reported.²⁹  A review of patients with CLA treated postnatally with sirolimus demonstrated that 3 patients with chylothorax did not respond.³⁰ 

When compared with octreotide and sirolimus, propranolol was noninvasive, conservative, and well-tolerated, which is consistent with propranolol being safe in the neonatal and pediatric populations.^27,31  Propranolol has been widely used for infantile hemangiomas, and its adverse effect profile in infants is well described, including bronchospasm, hypoglycemia, sleep disturbances, and acrocyanosis.³²  Our prenatal cases had transient hypoglycemia thought to be related to low birth weight and need for transient CPAP. Thus, prenatally treated newborns need to be closely monitored for hypoglycemia and respiratory status. Ongoing outpatient monitoring is critical for patients with CC and requires a multidisciplinary team to adjust dosage with growth and monitor effusions with regular imaging. Parents should be advised to seek medical evaluation promptly if symptoms of respiratory illness develop.

Previous studies have demonstrated symptomatic improvement of CLAs in response to propranolol,^20–25  and 3/4 of our CC cases had comorbidities consistent with CLA, such as ascites, edema, and hydrops. The mechanism of action of propranolol on the lymphatics is not well-understood.^22,23,25  In a patient with CLA, propranolol was associated with reduced circulating pro-lymphangiogenic factors, vascular endothelial growth factor-A, -C, and -D.²³  Propranolol has also been shown to improve lymphatic pumping in sheep.³³  Finally, propranolol has demonstrated inhibitory effect on inflammatory macrophages, which contribute to pathologic lymphangiogenesis.^34,35 

Although our case series suggests that propranolol is effective for severe CC, it is hard to draw conclusions from a small number of patients without a control group. Furthermore, the optimal dose for prenatal treatment needs to be determined because the bioavailability to the fetus is unknown. Finally, it is difficult to predict which CC require treatment because neonatal CCs have been reported to resolve spontaneously.^36,37 

Based on our experience, prenatally diagnosed, large pleural effusions that need transuterine thoracentesis may benefit from a course of prenatal propranolol given to the mother, with a maximum dose of 40 mg 4 times per day. Our 2 cases experienced resolution in 30 to 40 days. In infants, we found a range of propranolol from 0.7 to 2 mg/kg/d induced a response consistent with previous studies in which propranolol at 0.7 to 1.5 mg/kg/d improved pleural effusions in cases of CLA and CC.^25,38  Unlike the prenatal cases, clinical improvement may be seen as early as 1 to 2 weeks in postnatal CCs, but full resolution by imaging may take months. Contraindications would include cardiac abnormalities (bradycardia, heart block, prolonged PR interval) or parental reluctance. Although more studies are needed to assess the efficacy of propranolol, optimal dosing, duration of administration, and mechanisms of action in CCs, our case series supports a potential role for propranolol as a conservative and low-risk option in the prenatal and postnatal management of severe CC that is worthy of further investigation.

The Lymphatics Work Group at Columbia University Irving Medical Center includes Thomas Starc, MD, Erica M. Fallon, MD, Julie Monteagudo, MD, Angela Kadenhe-Chiweshe, MD, and Sheryl Tulin-Silver, MD.

CC

congenital chylothorax

CLA

congenital lymphatic anomaly

CPAP

continuous positive airway pressure

DOL

day of life

MCT

medium chain triglyceride

NICU

neonatal ICU

RA

room air

USG

ultrasonography