Spontaneous neonatal pulmonary arterial thrombosis – cases, mechanisms, and literature review

Abstract

Spontaneous pulmonary artery thrombosis in neonates is rare and can be life-threatening. Clinical presentation may mimic pulmonary hypertension or CHD. Further, not all children present with identifiable risk factors. We report the case of two infants with pulmonary artery thromboses who underwent rapid diagnosis and therapy, one with percutaneous intervention and the other with anticoagulation. We also conducted a literature review to highlight the importance of early identification and referral to a centre capable of performing appropriate medical and interventional therapies.

Background

Neonatal thrombosis ranges from incidentally diagnosed “clinically unsuspected” thrombi to life-threatening thromboembolism. Neonatal thrombosis, in any form, affects approximately 0.2% of neonatal ICU patients, of which ∼ 35% are in the venous circulation. ¹ Only ∼ 45% of patients have an identified aetiology, which range from extrinsic issues such as sepsis and perinatal asphyxia to intrinsic problems including thrombophilias and CHD. ¹ Neonatal pulmonary artery thrombosis is rare and can be life-threatening. We report two infants with left pulmonary artery thrombi of unclear aetiology. We also conducted a systematic review of all available literature for this rare condition.

Case report

The Institutional Review Board approved this report, and the patients’ families provided informed consent for publication.

Patient I

Patient 1 was a male born at 33 + 4 weeks gestation via induced, vaginal delivery due to maternal pre-eclampsia to a G1P1 mother with type 1 diabetes mellitus treated with insulin. Birth weight was 3.024 kg, and he was initially asymptomatic. On day 10 of life, he developed tachypnoea and hypoxaemia requiring oxygen at 1 L/min via nasal cannula. Echocardiogram demonstrated normal intracardiac anatomy and ventricular function, though no flow could be demonstrated in the left pulmonary artery. Subsequent MRI was concerning for a proximal left pulmonary artery thrombus with differential perfusion of 80% to the right pulmonary artery and 20% left pulmonary artery with no flow to the left upper pulmonary artery. Laboratory analysis demonstrated normal complete blood count, blood chemistry, blood/urine cultures, inflammatory markers (c-reactive protein and procalcitonin), prothrombin time, fibrinogen, and antithrombin 3 with elevated activated partial thromboplastin time of 36.2 and factor VIII of 370.1. His condition deteriorated slightly on day 13, ultimately stabilising on oxygen at 6 L/min via high-flow nasal cannula.

He was then transferred to our institution and underwent cardiac catheterisation on day of life 14. Diagnostics demonstrated mildly elevated right ventricular pressure with no gradient to the right pulmonary artery. A large, nearly occlusive filling defect, consistent with thrombus, was identified in the left pulmonary artery, originating near the ductal insertion and extending into the lower left pulmonary artery. Overall left pulmonary artery flow was severely diminished, with minimal-to-no left upper pulmonary artery visualisation. Injection past the defect demonstrated normal distal left pulmonary artery arborisation and normal pulmonary venous return. Mechanical aspiration with the Indigo Cat 3 system (Penumbra Inc., Alameda, CA, USA) was considered but was not readily available. ² Further, we felt that other reported techniques, such as using vascular plugs to extract the thrombus, had unpredictable risk of vascular damage given the location of this thrombus near the pulmonary artery bifurcation. ³ Left pulmonary artery angioplasty was performed using a 4-mm balloon with complete angiographic resolution of the obstruction and mild residual gradient ∼ 15 mmHg (Fig. 1 and Videos 1–3).

Figure 1. Fluoroscopic scenes of the cardiac catheterisation of patient 1. a ) Initial angiogram in the main pulmonary artery demonstrates a filling defect in the left pulmonary artery (}) consistent with a nearly occlusive thrombus; b ) single frame of the balloon thromboplasty; and c ) final angiogram in the MPA demonstrating resolution of the filling defect with normal flow in both the right and left pulmonary arteries.

He was weaned to air the evening of the procedure. In collaboration with haematology, therapeutic enoxaparin was initiated the night of the intervention with a plan for at least 6 months of therapy. Subsequent workup has revealed normal Factor V Leiden, Protein S, and Protein C levels; no prothrombin gene mutations were identified.

Patient 2

Patient 2 was a male newborn born at 38 weeks gestation after an uncomplicated pregnancy to a G1P1 mother. There were no identifiable maternal risk factors for neonatal thrombosis. His mother was GBS + and did not receive appropriate intrapartum antibiotics, so he underwent screen for sepsis – blood and urine cultures remained negative after 72 hours – and received prophylactic ampicillin and gentamicin for 48 hours. He did well immediately after birth but developed tachypnoea and hypoxaemia 8 hours after birth, being stabilised on oxygen at 2 L/min via nasal cannula. Laboratory analyses were all normal, including complete blood count, urinalysis, cultures, and inflammatory markers. Chest X-ray did not demonstrate acute pulmonary pathology, and a systolic murmur was auscultated at the left upper sternal border radiating to the left axillae, prompting an echocardiogram. The echocardiogram demonstrated normal cardiac anatomy with a large echogenic mass, consistent with thrombus, at the base of the left pulmonary artery; a patent ductus arteriosus was not identified (Fig. 2 and Videos 4–5).

Figure 2. Echocardiographic images of the left pulmonary artery thrombus in patient 2. a) Initial 2D echocardiography (left) demonstrates a large echogenic mass (arrow) at the base of the LPA, in the region of the patent ductus arteriosus insertion. Colour Doppler (right) demonstrates significant flow aliasing (arrow), consistent with a significant obstruction. b) Subsequent images after 5 days of anticoagulation demonstrating a smaller thrombus with improved flow (arrows).

He was transferred to our hospital and started on a therapeutic heparin infusion, which was transitioned to enoxaparin after 4 days. His symptoms slowly improved, and he was weaned to air on day-of-life 3. Repeat echocardiogram on day 5 demonstrated a decrease in thrombus size, so he was discharged with close follow-up. As with patient 1, screening haematologic workup was unrevealing with normal Protein C, Protein S, activated partial thromboplastin time, prothrombin time, and INR levels with no identified prothrombin mutations.

Systematic review

We utilised the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) methodology to conduct a systematic review. ⁴ Data from published reports, which are limited and disparate, were inadequate to perform a meta-analysis. Table 1 provides a concise review of published cases.

Table 1. Review of clinical characteristics, therapies, and outcomes for all identified infants with neonatal pulmonary artery thrombosis.

AGA = appropriate for gestational age; C/S = caesarean section; CTA = CT angiography; CXR = chest X-ray; DM = diabetes mellitus; DOL = day of life; ECMO = extra-corporeal membrane oxygenation; F = female; FHx = family history; FT = full term; gtt = infusion; HFOV = high-frequency oscillatory ventilation; HTN = hypertension; LPA = left pulmonary artery; M = male; NCO2 = nasal cannula oxygen; PDA = patent ductus arteriosus; PE = pulmonary embolism; PICC = peripherally inserted central catheter; PROM = premature rupture of membranes; RPA = right pulmonary artery; RV = right ventricle; SAB = spontaneous abortion; tPA = tissue plasminogen activator; wk = week.

* Perinatal risks include central venous lines if noted.

Discussion

Though thrombosis in children is fairly common, isolated pulmonary artery thrombus is rare with no reliable incidence estimate. Therapies depend on the clinical condition as well as the child’s age. With our cases, we opted for an interventional procedure for patient 1 given that he was nearly 2 weeks old at the time of diagnosis. Conversely, we opted for medical management in patient 2, who was diagnosed just hours after birth and only requiring nasal cannula oxygen.

Virchow’s triad provides insight into the development of pulmonary artery thrombi. A plausible mechanism in common with all pulmonary artery thrombi is endothelial disruption related to closure of the patent ductus arteriosus at its insertion in the PA roof, serving as the nidus for the thrombus. In most reported cases, another process was identified that contributed to stasis of flow (e.g., intracardiac lines) or created a prothrombotic state (e.g., serious peripartum bacterial infection and inherited thrombotic disorder). ⁵

Our systematic review identified 20 reported cases of pulmonary artery thrombi in neonates, with a general review provided in Table 1. ^6–22 Aetiologies varied but can be broadly categorised into those with intrinsic and extrinsic predisposing factors. Intrinsic issues include anatomic anomalies and thrombotic disorders, while extrinsic issues largely comprised either a serious infection, maternal conditions that predispose to neonatal thrombosis (e.g., diabetes mellitus), or perinatal asphyxia with severe stress.

As outlined in the table, anatomic anomalies associated with pulmonary artery thrombus include premature closure of the patent ductus arteriosus, pulmonary artery/ductal vascular anomalies, as well as cystic pulmonary disease. ^7,9,22 Patients with any form of thrombophilia are potentially at risk, with reports of Protein C and Factor V Leiden deficiencies. ^10,13 Per our review of the literature, a notable number of cases did not have a clear identifiable risk factor. ^7,8,19–21 And others were noted to be associated with risks that are unclear, such as meconium-stained amniotic fluid or very brief birth asphyxia requiring “minimal resuscitation.” ^6,12,20 The therapies were also significantly disparate, ranging from no intervention, to systemic anticoagulation and tissue plasminogen activator, transcatheter angioplasty, and even surgical thrombectomy. Some patients were also initially stabilised on extra-corporeal membrane oxygenation. No significant therapy-related complications were described in the reports, including no bleeding complications with tissue plasminogen activator and no procedure-related complications after catheterisation or surgery.

Diagnosing a pulmonary artery thrombus begins with recognising it as part of the differential. Asymmetric chest X-ray vascular markings are often the first clue. An echocardiogram will typically identify the thrombus, though some obstructions may occur beyond the pulmonary artery hilum so may be missed by echocardiography. Though diagnostic angiography was reported in many older cases, we recommend proceeding to CT angiography if echocardiographic findings are unusual (e.g., decreased flow to one branch pulmonary artery or pulmonary veins). ^10,13

Finally, management should be directed by the clinical condition as well as identified aetiologies. Medical management with anticoagulants may be reasonable for patients with mild symptoms requiring minimal treatments (e.g., nasal cannula oxygen), while systemic tissue plasminogen activator, transcatheter angioplasty, surgical thrombectomy, and even extra-corporeal membrane oxygenation may be indicated for more ill patients with larger occlusive/near-occlusive thrombi or evidence of right ventricle strain. ^10,15,17

Conclusion

Pulmonary artery thrombus is a rare and serious condition in neonates with variable aetiologies. Clinicians should be vigilant for this issue, especially since the condition may masquerade as primary pulmonary hypertension. Earlier diagnosis will allow for appropriate therapy which can range from medical management to surgical thrombectomy and even extra-corporeal membrane oxygenation support. Ongoing larger sample size studies need to be conducted to determine the significance of such associations.